You are on page 1of 344

Applied Linguistics Review 2 2011

Applied Linguistics Review 2 2011


Li Wei

De Gruyter Mouton

ISBN 978-3-11-023932-4 e-ISBN 978-3-11-023933-1 ISSN 1868-6303 ISSN online 1868-6311

Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at 2011 Walter de Gruyter GmbH & Co. KG, Berlin/New York Typesetting: PTP-Berlin Protago TEX-Production GmbH, Berlin Printing: Hubert & Co. GmbH & Co. KG, Gttingen Printed on acid-free paper Printed in Germany

Li Wei Department of Applied Linguistics & Communication University of London, Birkbeck College 26 Russel Square, Bloomsbury London WC1B 5DQ UK E-mail:

Associate Editors
David Block University of London, Institute of Education Bencie Woll University College London Itesh Sachdev University of London, SOAS

International Advisory Board

Keiko Abe Kyoritsu Womens University, Japan Kingsley Bolton City University of Hong Kong, China Vivian Cook Newcastle University, UK Donna Christian Center for Applied Linguistics, USA Annick De Houwer University of Erfurt, Germany Patricia Duff University of British Columbia, Canada Diana Eades University of New England, Australia Yihong Gao Beijing University, China Ofelia Garcia City University of New York Graduate Center, USA Susan Gass University of Michigan, USA Fred Genesee McGill University, Canada Nancy Hornberger University of Pennsylvania, USA Alex Housen Vrije Universiteit Brussel, Belgium Andy Kirkpatrick Hong Kong Institute of Education, China Claire Kramsch University of California, Berkeley, USA Mayouf Ali Mayouf University of Sebha, Libya Tim McNamara University of Melbourne, Australia Ben Rampton Kings College London, UK Elana Shohamy Tel Aviv University, Israel David Singleton Trinity College Dublin, Ireland Anna Verschik Tallinn Pedagogical University, Estonia Terry Wiley Center for Applied Linguistics, USA Lawrence Zhang National Institute of Education, Singapore


Li Wei Editors note Suresh Canagarajah Translanguaging in the classroom: Emerging issues for research and pedagogy Massimiliano Spotti Modernist language ideologies, indexicalities and identities: Looking at the multilingual classroom through a post-Fishmanian lens Kingsley Bolton Language policy and planning in Hong Kong: Colonial and post-colonial perspectives Lin Pan English language ideologies (ELI) in Olympic Beijing Andy Kirkpatrick English as a medium of instruction in Asian education (from primary to tertiary): Implications for local languages and local scholarship Sin Preece Universities in the Anglophone centre: Sites of multilingualism Anne Pauwels Risking or boosting masculinity? Mens language choices in multilingual settings



51 75

99 121




Larissa Aronin, Muiris O Laoire and David Singleton The multiple faces of multilingualism: Language nominations Raphael Berthele On abduction in receptive multilingualism. Evidence from cognate guessing tasks Annick De Houwer Language input environments and language development in bilingual acquisition Margo DelliCarpini The role of phonemic awareness in early L2 reading for adult English language learners: Pedagogical implications Piia Varis, Xuan Wang and Caixia Du Identity repertoires on the Internet: Opportunities and constraints Crispin Thurlow and Adam Jaworski Tourism discourse: Languages and banal globalization Jeff Bezemer, Alexandra Cope, Gunther Kress and Roger Kneebone Do you have another Johan? Negotiating meaning in the operating theatre




241 265 285


Editors note

The inaugural issue of Applied Linguistics Review was published in June 2010. All the evidence suggests that it has been very positively received. Some of the articles have not only been cited by other researchers but also made onto the reading lists for students. This second volume includes another ne selection of articles on a range of topics in applied linguistics, from critical analysis of classroom interaction and language teaching ideologies to tourism discourse and interactions in the operating theatre. Together the articles demonstrate the breadth and strength of applied linguistics as a eld of transdisciplinary, critical enquiry. This year we received a lot more articles than we could possible include. It is always a shame not to be able to accept all the submissions. But we do want to ensure high quality of the published articles. I am very grateful to the contributors for the most interesting work they present in their articles. I am also grateful to my co-editors, especially David Block, for their work in putting this volume together. Several of my immediate colleagues at Birkbeck have reviewed the papers for this volume, along with other invited reviewers and members of the advisory board. Brigid OConnor acted as the copy editor. I thank them all. As ever, the work by the staff at the Mouton ofce must be acknowledged. Without their enthusiasm, this publication would not have been possible. London January 2011

Translanguaging in the classroom: Emerging issues for research and pedagogy


Abstract This article attempts to synthesize the scholarship on translanguaging conducted in different academic disciplines and social domains, and raises critical questions on theory, research and pedagogy to take the orientation forward. The literature review highlights the overly cognitive and individualistic focus on translanguaging competence, the need to explore this communicative practice in domains other than conversation, and the failure to develop teachable strategies of translanguaging. Findings from a classroom ethnography of a writing course are marshaled to develop teaching strategies for the co-construction of meaning and orientations for assessing effective translanguaging practices.

1. Introduction Advances in our understanding of multilingual communication have solidied academic interest around the term translanguaging (see Canagarajah forthcoming; Creese and Blackledge 2010; Garcia 2009). A neologism, it has come to stand for assumptions such as the following: that, for multilinguals, languages are part of a repertoire that is accessed for their communicative purposes; languages are not discrete and separated, but form an integrated system for them; multilingual competence emerges out of local practices where multiple languages are negotiated for communication; competence doesnt consist of separate competencies for each language, but a multicompetence that functions symbiotically for the different languages in ones repertoire; and, for these reasons, prociency for multilinguals is focused on repertoire building i.e., developing abilities in the different functions served by different languages rather than total mastery of each and every language.

Suresh Canagarajah

As the notion receives increasing attention, scholars are documenting translanguaging in diverse social and educational contexts. The theorization of this practice is going on in different disciplines under different labels. The following are some of the terms used for translanguaging in different elds (identied according to the scholars who use them): Composition: codemeshing (Canagarajah 2006; Young 2004); transcultural literacy (Lu 2009); translingual writing (Horner et al. forthcoming) New literacy studies: multiliteracies (Cope and Kalantzis 2000), continua of biliteracy (Hornberger 2003), pluriliteracy (Garcia 2009), Applied linguistics: plurilingualism (Council of Europe 2000), third spaces (Guttierez 2008); metrolingualism (Pennycook 2010). Sociolinguistics: uid lects (Auer 1999); hetero-graphy (Blommaert 2008); poly-lingual languaging (Jorgenson 2008). The contexts of translanguaging range from academic reading and writing (Lu 2009), internet communication (Williams 2009), youth performative conversational interactions (Rampton 2008), hip hop (Pennycook 2007), childrens interactions (Jorgenson 2008), street signage (Gorter 2006), and indigenous literacy (Hornberger 2003), to mention a few. While the body of scholarship on social manifestations and classroom occurrences of translanguaging is increasing, teachers are also interested in the pedagogical implications of this practice. However, proactive teaching of translanguaging raises a difcult set of theoretical and practical questions that have not received adequate discussion so far. The purpose of this article is to adopt a critical orientation to the theorization and research of translanguaging in an effort to translate the ndings more effectively for classroom purposes. After outlining the issues for further research, I report some ndings from a classroom ethnography I conducted to illustrate how we might address the emerging questions in the eld.

2. Towards a balanced theoretical orientation In the context of a linguistics that theorizes competence and communication in terms of monolingual norms, it is appropriate that translanguaging is now being given a lot of attention in the academy. This is a matter of afrmative action. Many previous constructs arise from pitting one language against another, treating multilinguals as non-native and, therefore, lacking ownership in some languages. Multilinguals users linguistic variations are treated as marking their

Translanguaging in the classroom: Emerging issues for research and pedagogy

nonstandard or decient usage, resulting from interference from the other languages in their repertoire, and conditioned by their rst language or culture not to accommodate a second language effectively. All these terms (i.e., non-native, interference, conditioning, rst/second language and culture) are indications of what Vivian Cook (1999) calls a comparative fallacy. The tendency to adopt binary and hierarchical orientations to language has distorted the integrated nature of multilingual competence and communication. Translanguaging helps us adopt orientations specic to multilinguals and appreciate their competence in their own terms. However, any valorization of a suppressed communicative practice undergoes what feminist scholars Kirsch and Royster (2010: 647) have recently called the three Rs rescue, recovery, and (re)inscription. This preoccupation with salvage and afrmative representation leads to an uncritical orientation to marginalized rhetorical traditions. They can thus lead to a fourth R romanticization. It is not surprising that scholarship on translanguaging makes multilingual communication appear more diverse, dynamic and democratic than monolingual competence. We have to critique this new binary multilingual and monolingual and adopt a critical attitude towards the resources/limitations and prospects/challenges of translanguaging. As we address some critical questions on translanguaging, we have to examine certain misconceptions engendered by the four Rs. In many quarters, translanguaging is valorized as an urban and postmodern practice. For example, Rampton (1999: 425) considers crossing (i.e., a performative practice of translanguaging, in which one adopts the language of a community he/she is not typically associated with, for identication purposes) reective of life in late modernity when aestheticized multi-modal texts recruit people into life-style communities, into neo-tribes without socialisation where centres of authority are hard to nd and where entry is a matter of the consumers desire, personal taste, shopping skills and purchasing power (see also Rampton 2008). Such practices are also associated with urban life. Pennycook (2010) has recently coined the term metrolinguistics (to capture the uid and hybrid language practices of youth in the city). In both cases, the suggestion that translanguaging practices are postmodern and urban can give a misleading impression, and hide their vibrancy in other places and times. We do have evidence that translanguaging has been practiced in pre-colonial communities and in rural contexts. In South Asia, Africa and South America, rural life has featured considerable heterogeneity and multilingualism. Neighboring villages with different languages and tribal groups adopt translanguaging in contact situations. While villages in the West are homogeneous, villages in the southern hemisphere have always featured diversity. Khubchandani (1997)

Suresh Canagarajah

provides a book length treatment of the strategies adopted by villagers in South Asia to communicate to each other. Pollock (2009) has discussed how Sanskrit as a lingua franca in pre-colonial times cohabitated with local languages and communities, when Latin and English were imposed from above. The limitation in the scholarship on translanguaging in pre-colonial times is that it is not based on empirical observations. It is based on archival research. Even this archival documentation is thin as practices that deviated from the ideologies of modernity were suppressed or unrecorded in modern scholarship. Khubchandani has to painstakingly recover available evidence from present day life to theorize the translanguaging practices of pre-colonial times. As we get more evidence, we need to compare rural and urban, pre-colonial and postmodern practices in translanguaging to develop a more focused understanding of this mode of communication. Another overgeneralization is the broad constructs multilingual and monolingual. It appears as if the notion of multilingual competence borders on essentializing diverse communicative practices and attributing them as invariable essences of non-western communities. Furthermore, it is not clear whether being monolingual is an ontological reality. All of us have multilingual competence and adopt multilingual practices in our competence. Even the so-called monolinguals shuttle between codes, registers and discourses. Therefore, multilingual competence involves a massive generalization of practices in many regions, times and communities. Perhaps we can justify this construct as a case of strategic essentialism (see Spivak 1993). The construct is a useful heuristic to demystify the dominant construct of monolingual competence, articulate the features of alternate competences and demystify the language ideologies that motivate unitary models of language and competence. Perhaps we can adopt more diversied and localized constructs once the dominant paradigms have been deconstructed, and develop a more complex orientation to translanguaging. The theorization of translanguaging in terms of cognitive competence and individual prociency (see Franceschini 2010; Garcia 2009) has also led to some oversight. The overriding concern of translanguaging scholars has been with rectifying the Chomskyan orientation, which represents language competence as innate, monolingual and arising from a homogeneous environment. Scholars of translanguaging have invoked Vivian Cooks (1999) notion of multicompetence to combat the Chomskyan dominance. They rightly point out that multilingual competence is qualitatively different from the Chomskyan notion of monolingual competence, and that multilinguals parallel process the diverse languages in their repertoire even when they function in a relatively homogeneous language. However, what they overlook is that translanguaging is a social accomplishment. Translanguaging not only involves a person drawing from all

Translanguaging in the classroom: Emerging issues for research and pedagogy

the languages in his/her repertoire to communicate, it also involves shuttling between the languages brought by the other to co-construct meaning. Furthermore, translanguaging is performative. As Khubchandani (1997) demonstrates, translanguaging is not a case of applying a linguistic predisposition. It is a creative improvisation according to the needs of the context and local situation. It is an interactive achievement that depends on aligning ones language resources to the features of the ecology to construct meaning. We have to give greater importance to translanguaging as a form of social practice along the lines Pennycook (2010) has theorized language interactions recently.

3. Issues in current research While there are theoretical exaggerations of the sort reviewed above, there are also some limitations in the research carried out on translanguaging practices. In much of the research on crossing and related performative practices, scholars have focused on the production of difference, and not on the negotiation of difference. For example, instances of youth adopting the codes of out-group members is represented without considering how out-group members respond to these features or display uptake (see Rampton 2008). Pennycook (2010) also presents the ways in which his subjects mix codes in hip-hop to demonstrate new subjectivities and performances. We dont know how participants, other than the researcher, interpret and respond to these translanguaging displays. As Jane Hill (1999) observed some time ago, other communities may have mixed feelings about their codes being appropriated. For this reason, Blommaert (2005: 205 206) advises that we should consider the dialogical and, I might add, interactional dimension of translanguaging activity: Meaning including the attribution of identity categories is a dialogical practice in which the uptake of ones semiotic acts may be as consequential as the structure of the semiotic acts themselves. In other words, in order for an identity to be established, it has to be recognised by others. . . . people are not entirely free in semiotic work. If we keep this caution in mind, we will also note that translanguagers adopt certain calculated strategies to gain uptake. We have to go beyond studying the strategies of translanguaging production to studying strategies of negotiation. Studies on translanguaging have also been conducted in a product-oriented manner, leaving processes out of consideration. Blommaerts (2008) own Grassroots Literacy illustrates this approach. In studying the historical/biographical texts of two Congolese subjects, Blommaert interprets their writing to understand their choices. He gives complexity to their multilingual and multimodal writing, labeling it hetero-graphy. However, since he studies only the product

Suresh Canagarajah

and has no way of interviewing the writers themselves (who are deceased), he ends up with a decient view of translanguaging. His analysis suggests that the Congolese subjects adopt translanguaging as they dont have access to elite literacy. Blommaert also assumes that the authors are not adopting any proactive strategies to communicate to their intended recipients, failing to gain uptake and being silenced. To overcome such decient reading, we have to go beyond merely documenting instances of translanguaging and analyzing their linguistic and textual realizations. We have to ask further questions of process, such as: What strategies do translanguagers adopt to help readers/listeners interpret their language choices? What choices did they face in codes and conventions in their production? What considerations help them resolve their choices? What composing or cognitive stages characterize the production of translanguaging? These questions will enable us to gain an insider perspective on the processes that accompany translanguaging and help us teach it better to students. Current research has also limited translanguaging to multilingual interlocutors. In some instances, as in the corpus study of multilingual negotiations in English, the researchers have decided to leave out native English speaker (NES)1 subjects in order to preserve the validity of their data. Jenkins (2006: 160) notes, ELF [English as a Lingua Franca] researchers specically exclude mother tongue speakers from their data collection. Indeed, in its purest form, ELF is dened as a contact language used only among non-mother tongue speakers. However, this restriction leads to an articial communicative context. Contact situations sometimes include native and nonnative, monolingual and multilingual participants. In fact, classroom situations often involve NES or monolingual teachers. Furthermore, the exclusion of native speakers prevents us from understanding the agency of multilingual speakers. Multilinguals adopt creative strategies to negotiate meaning with NES and co-construct meaning. We have to study translanguaging in contexts where there is a mix of speakers in order to understand the strategies of communication of translanguagers. Furthermore, while a majority of existing studies on translanguaging are on face to face oral interactions, we havent studied how translanguaging works in other genres and modalities of communication. For example, youth mesh codes in computer-mediated discussions. They have been discussed as instances of hybridity rather than translanguaging (Lam 2004). However, we can compare such interactions with face to face interactions to understand if the translanguaging strategies are different. Similarly, we dont have enough studies on translanguaging in writing. In fact, there is a strong opinion among some scholars that translanguaging is not permitted in writing. Even those who accommodate diverse codes in speaking feel that such mixing are not appropriate for writing.

Translanguaging in the classroom: Emerging issues for research and pedagogy

Barbour (2002), for example, argues that since the rich paralinguistic cues of speaking are not available for interpreting writing, multilingual authors have to get the help of editors and translators to eliminate the localisms in their English. For him, the lack of gestures, tone and contextual details reduces the possibilities for guessing meaning. These assumptions lead him to argue that writing has to always adopt the standard language expected for that context. In the same vein, Peter Elbow (2002: 128) argues, Literacy as a culture or institution almost always implies just one dialect as the only proper one for writing: the grapholect. He argues that the accommodation of codes that deviate from standard language should be postponed for a time when those codes become the norms in writing. The prohibition of translanguaging in writing also results from the dominance of autonomous literacy (Street 1984). According to this ideology, texts are static products that contain self-evident meaning that can be extricated through detached reading. For this reason, while interlocutors negotiate meanings interactively in face to face conversations, they assume that such negotiations are not necessary for writing. This attitude leads to censoring translanguaging, and promoting standardized and discrete codes for literacy. To make matters further complicated, writing in educational contexts is a high stakes activity. While teachers permit translanguaging in face to face interactions for classroom interactions (i.e., group work, teacher/student conversations), they dont permit it in writing, which they consider a more formal activity where students performance is assessed. These assumptions can be easily challenged. Though writing lacks some of the communicative resources in speaking, it has other resources that can favour translanguaging. For example, writing has a materiality to it. One can use the successive drafts to prepare readers for interpreting translanguaging. Writing also has a visual dimension. One can use the visual resources to embed additional codes and modalities for meaning-making. Furthermore, the notion that writing can take only one code at a time and the assumptions of autonomous literacy are very recent European orientations to writing. Literacy in non-western communities has always been multilingual and multimodal (see Canagarajah 2006; de Souza 2002). For example, the well known manipralava writing in South Asia combined Tamil and Sanskrit (Viswanathan 1989). Such texts were actively negotiated for meaning by readers and writers. We are beginning to see scholarship that shows that even writing in the West favoured dynamic interactions for meaning-making before Enlightenment (see Ratcliffe 1999; Stark 2008;). Such literacies and textualities are making a comeback in the context of digital communication. Texts are multilingual and multimodal on the Web, and readers and writers are negotiating their hybrid codes more actively for meaning (Nicotra 2009; Williams 2009). Therefore, we have to be open to the possibil-

Suresh Canagarajah

ity that translanguaging will be actively practiced in literacy in the future. It is important for researchers to study translanguaging in writing.

4. Issues in pedagogical practice A further set of questions relate to the possibility of teaching translanguaging in classrooms. The pedagogical side is underdeveloped in general. While we have studied the practice of translanguaging in social life i.e., in urban youth encounters, linguistic landscapes, and the Internet we havent gured out how to develop such prociency among students in classrooms. In a recent study on translanguaging practices in bilingual classrooms, Creese and Blakledge (2010: 113) emphasize the need for further research to explore what teachable pedagogic resources are available in exible, concurrent approaches to learning and teaching languages bilingually. In making this call, they echo what other scholars like Lin and Martin (2005) have also considered important in order to move multilingual language acquisition forward. The studies we do have on school contexts show translanguaging to be a naturally occurring phenomenon. In a majority of these studies, acts of translanguaging are not elicited by teachers through conscious pedagogical strategies. They are produced unbidden. In fact, in many of these cases, translanguaging occurs surreptitiously behind the backs of the teachers in classes which proscribe language mixing (see the studies from diverse communities in Lin and Martin 2005; Heller and Martin-Jones 2001). In the more proactive situations, teachers have provided safe spaces for students to adopt their multilingual repertoire for learning purposes, and teachers have themselves collaborated with students in using the repertoire as a resource, as in the study by Creese and Blackledge. Pedagogical approaches such as the biliteracy workshop (Garcia 2009) and continua of biliteracy model (Hornberger 2003) theorize how students may shuttle between languages and modalities in their learning. However, we still have a long way to go in developing teaching strategies out of these broadly conceived models. What current classroom studies show is that translanguaging is a naturally occurring phenomenon for multilingual students. Translanguaging cannot be completely restrained by monolingual educational policies. It can occur with minimal pedagogical effort from teachers. However, such studies might give the impression that translanguaging doesnt have to be taught. If it occurs naturally in the most unbidden contexts, it might be argued, translanguaging is so fully developed among multilingual students in their home and community contexts that there is nothing further for the school to add, other than provide

Translanguaging in the classroom: Emerging issues for research and pedagogy

a context for it to be practiced. Such studies are bolstered by cognitive orientations to competence that posit that translanguaging is natural to multilinguals (Bhatia & Ritchie 2004: 794; see also Franceschini 2010). This line of thinking leads to the tendency to see translanguaging as an intuitive capacity for which multilinguals are naturally endowed. Such an assumption makes one wonder if translanguage needs to be taught at all. Treating translanguaging as a practice (not natural competence), I demonstrate from a course I taught that making opportunities for critical analysis will help students develop their translanguaging prociency further. As we develop teachable strategies of translanguaging, we have to consider some serious issues for assessing the effectiveness of this practice. An important consideration is if there is a place for error or mistake in translanguaging. If there is none, we dont have a concept of developmental translanguaging. We might go away with the assumption that translanguagers never err. Also, this might mean that translanguaged talk and texts are always perfect. The difculty in identifying errors is that translanguaging doesnt accommodate the notion of intereference from other languages. Mutual inuences from the languages in ones repertoire are treated as creative and enabling, not hindering, communication. Also, we dont know if there is a concept of normative translanguaging against which errors can be judged. We have to explore if we can move away from a normbased or form-based notion of error and adopt a practiced based orientation to developmental stages in translanguaging. In the same vein, we also need to adopt rhetorical considerations in assessing the effectiveness of translanguaging. In most studies on translanguaging, whether inside or outside the classroom, researchers have focused mostly on the information transfer, pragmatic meanings and implications for cognitive competence. They havent asked if the translanguaging is appropriate for that context in rhetorical terms. Could better choices have been adopted for more effective communication? To address such questions of rhetoric, we have to move from grammar to discourse, meaning transfer to aesthetic considerations. There are also questions related to the ways students negotiate power in the classroom when they translanguage. Since classrooms are often dominated by NES or monolingualist assumptions and autonomous literacy, how do students negotiate such constraints for successful communication? We have to also consider the implications of translanguaging prociency for social and educational success for multilingual students. Would this competence help or hinder them in their educational and professional prospects. The attitudes to translanguaging should also be studied in order to understand the possibilities for pedagogical success. How do students relate to translanguaging? Some minority communities have recently voiced concerns about this


Suresh Canagarajah

practice and its implications for community empowerment. Scott Lyons, a native American scholar, argues that this form of hybridity is a threat to heritage languages. He feels that translanguaging would make native American students complacent about language maintenance, and dilute the integrity of nativeAmerican languages. He states that translanguaging is hybridity and violates the elders rule of mutually assured separatism (Lyons 2009: 102). To address his charge we have to nd out how translanguagers perceive their relationship to the codes they mix in their utterance. Or, do they treat translanguaging as free of values? as Peter Auer (1999) suggests in his theorization of uid lects, treating them as a new code in their own right. I address these questions in a writing course I recently taught for graduate students at my university, using my ndings for the further development of translanguaging for classroom purposes.

5. A case study The data in this study came from a graduate-level course on the teaching of second language writing. I conducted a classroom ethnography on the development of teacher identities and literacy awareness. The class consisted of roughly half Anglo-American students, and half international students (from China, Taiwan, Japan, Korea, UAE and Saudi Arabia). A major assignment in the course was the writing of a literacy autobiography to critically reect on students own writing development and translate their insights into pedagogical practices. I interpret the translanguaging in this writing in the context of the activities and assignments produced throughout the course. The data sources are the following: successive drafts of essays (abbreviated as D1, D2, etc.); a weekly journal of students responses on readings and writings (J); assignments and activities (A); a peer commentary (PC); and surveys and interviews on writing development (I). In selected cases, I also conducted a stimulated recall procedure (SR) to query students on their rhetorical and linguistic choices. In this article, I report on the writing of a single student, Buthainah from Saudi Arabia. In her case, I gave her an early draft of my essay interpreting her translanguaging strategies for her response, as a form of member check procedure (MC). I adopted a dialogical pedagogy in this course. Students observed writing classes of other instructors or their own, and critically reected on their observations in relations to the readings, to develop more constructive pedagogies and professional identities. Also, the writing was collaborative. Peers and the instructor critiqued each draft for subsequent revisions. This activity of negotiating meaning with the writers helped me in many ways. In posing questions

Translanguaging in the classroom: Emerging issues for research and pedagogy


on their writing choices in my feedback, I helped students develop a critical awareness without having to impose my values on them. Often students challenged the assumptions behind my questions and helped me identify traces of dominant ideologies. Furthermore, this dialogical pedagogy helped me address an important methodological challenge in this kind of research. In their seminal article developing a social orientation to language acquisition, Firth and Wagner (1997: 285) argue that SLA research requires a signicantly enhanced awareness of the contextual and interactional dimensions of language use, an increased emic (i.e., participant-relevant) sensitivity towards fundamental concepts. If meaning is co-constructed by multilinguals in contact situations, researchers who stand outside the negotiation of meaning in the interaction cannot make valid observations on the communication. However, those who are part of the interaction may be considered as not having sufcient detachment to be analytical. My position as a teacher/researcher helped me resolve this dilemma to some extent. Though I had relative detachment as a teacher to adopting an analytical perspective, my dialogical teaching helped me actively negotiate meanings with the students and adopt an insider orientation to their writing process. I must acknowledge that I did give students samples of translanguaged writing (e.g., Smitherman 2003) and readings that complicate autonomous literacy (Canagarajah 2002) as part of the course syllabus. The students were also aware of examples of my own critical and translanguaged writing (e.g., Canagarajah 2006). Such a direct introduction to translanguaged writing may make critics to wonder if writing samples derived from such a course are shaped by my instructional approach. In other words, is my pedagogy biased towards producing such translanguaged texts? However, I dont think of any pedagogical context as value-free or neutral. All teaching contexts and teachers bring their own positions on literacy and multilingual communication. Furthermore, all teachers have to negotiate the dominant policies and ideologies of language and writing in their pedagogical context. As my questions in SR will reveal, I was often inuenced by the dominant expectations on writing in a university course. Therefore, the course was not totally shaped by a single ideology. Students too had to navigate through competing ideologies as they negotiated their footing toward the writing and the course. Even in courses where there is no explicit policy that favours translanguaging, those students who want to practice such writing nd affordances in their context. Thus the classroom is a site of competing ideologies, and students and teachers always have to negotiate their footing. For students to discern the features that are friendly to their interests and negotiate features that are unfavourable is part of their learning experience.


Suresh Canagarajah

5.1. Buthainahs negotiation strategies Buthainah opened her essay as follows:

I doont want to! was my response to my parents request of enrolling me in a nearby preschool. I did not like school. I feared it. I feared the aspect of departing my comfort zone, my home, to an unknown and unpredictable zone. My parents desired to enroll me in a private preschool. Due to my fear, I refused. My parents face discolored and the sense of disapproval appeared in their tone of speech. To encourage me, they recited a poetic line that I did not comprehend as a child but live by it as an adult. They said Who fears climbing the mountains Lives forever between the holes. As I grew up, knowledge became my key to freedom; freedom of thought, freedom of doing, and freedom of beliefs. In every decision I make relating to my academic world, a new wave full of challenges and unexpected turnouts unfolds. (D6)

The Arabic proverb with which she starts was uttered to her by her parents when she hesitated to go to school on the rst day. She translates this proverb later in the paragraph as Who fears climbing the mountains Lives forever between the holes. The Arabic proverb serves also as the title for the essay. In addition to such proverbs and verses in Arabic script, Buthainah also adopted other types of translanguaging in her esssay:
My experience learning English has interesting twists. In many different stages of my life, I had a different motivation. At the end of the road, however, knowledge became the key for freedom, ma sha Allah. In sixth grade, my instructor wanted to challenge the students, including moi, by having us write about Riyadh as the Arabian Capital for Culture. Our rst exposure to real English was at that airport. The man said beaucoup de choes that I could not understand. The motif used to divide her sections: At that time, my dear reader, I have not learned English in school yet since English was required to seventh graders and beyond; and I was in sixth grade . A ket-koot is a small chick in Arabic. At that time I had about seven chicks [P.S. couple of them died =(].

As we see, Buthainah transliterates Arabic expressions such as ma sha Allah. She uses French phrases in certain instances. There were also visual symbols that she considered part of her expressive repertoire: i.e., the motif she used to divide her sections, and other emoticons. Note also the elongation of doont in the paragraph quoted earlier to capture the auditory effect. Other idiosyncrasies such as the missing apostrophe in my parents request and the noni-

Translanguaging in the classroom: Emerging issues for research and pedagogy


diomatic expression a new wave full of challenges and unexpected turnouts unfolds will also be attributed to translanguaging by scholars and not treated as errors. Note that this is the sixth draft Buthainah wrote. She had ample opportunities to change any idiosyncratic features earlier. She retained these features as she considered them important for her voice. Asked why she chose to include so many Arabic verses, she mentioned:
My objectives for using thses poems are many. First, they are part of me. And this essay is about me. Thus, it seemed appropriate to include them in an essay on my literacey development. In addition, poetry is part of my Arabian culture because it is highly valued. . . . Why shouldnt I includ it? (SR)2

On the motif used to divide the sections, she explained:

It is a familiar shape that one may nd in Islamic art. Since I am a Muslim, and Islam inuenced me, it also inuenced my literacy experience. Thus, using this particular motif was a hint to the reader to my heritage (MC).

Even the spelling mistakes and unusual idiomatic and syntactic choices were explained by the fact that she focused more on rhetorical issues rather than form. She assumed that readers will negotiate them for meaning in context:
I did not see my essay as a one-way informative essay. It is a negotiated essay that seeks a better understanding from educators and future teachers to the multilingual experience. By addressing my readers, I am welcoming them to the discussion, which, in my perspective, [is] ongoing (emphasis added; MC).

As she expected, the idisyncracies in form didnt cause any problem for comprehension. No one complained about these issues. Her peers were willing to negotiate the essay for meaning with her. I like to focus rst on the strategies Buthainah adopts to encourage her readers to negotiate for meaning with her. I am especially concerned about the responses of NES subjects, as researchers havent adequately studied how they behave in contact situations. I call the rst set of strategies Recontextualization Strategies. What Buthainah does is change the dominant orientations to textual reception in western educational institutions, so that readers will adopt the orientation of what Buthainah calls above a negotiated essay. The very opening of the essay (quoted earlier) indicates the ethos of the text and indicates a different orientation to reading. As Buthainah opens with Arabic and doesnt provide translations, readers are compelled to work for meaning. They understand that meaning is not given to one on a plate. There is evidence that there was uptake of this strategy. Tim, an Anglo-American student, mentioned in a peer review:


Suresh Canagarajah By not translating you are excluding a wider audience, your non-Arabic speaking audience from being able to engage fully with the text. Perhaps you are challenging them to bridge that gap as readers. That if they want to gain access to your writing (to a piece of you, perhaps?) they have to meet you halfway somehow. Or, maybe these poems are a special treat you mean only for those able to read Arabic to experience (PC, 10/22).

Bridging the gap and meeting halfway are the attitudes multilinguals display when they interact with other language groups in contact situations (see Khubchandani, 1997). Tim adopts this strategy to gure out the meaning and engage with Buthainah to construct meaning. Thus he moves away from the dominant orientation of autonomous literacy to approach this essay. Buthainah also carefully gives clues about her preferred authorial identity to prevent readers from adopting native speaker norms. She shapes the context so that they will approach the text as a multilingual encounter. Buthainah distinguishes herself as a functional bilingual and not an ESL student throughout her essay. She says in an early paragraph in her essay:
I may sound and write like-or even better than- a native speaker, but I do not look like one. Many people would group all second language learners together not knowing that this grouping may be negative to the learner especially those who are considered functional bilinguals . . . language users who may have [a] few problems with English, but were beyond the realm of ESL (D6).

ESL status is stereotypically considered developmental and decient. However, Buthainah constructs a more empowering identity for her. She suggests that readers should go beyond issues of form to consider the rhetorical and meaning-making abilities of multilinguals. Readers did adopt such an orientation to their interpretation, not letting form bother their reading. Chrissie, an Anglo-American student, shows uptake. She interprets Buthainahs deviations from norms as not decient, but having rhetorical implications:
As a student at an American University, she knows that following the rules of academic writing will get her far, but I think she does so with a sense of who she is. She obviously respects her L1 and explicitly shows us, the reader, throughout the text with poems, and how expressions are beautiful in Arabic, but awkward in English (for example) (PC 10/28).

In certain places in the text, Buthainah seemed to confront some of the stereotypes of her readers more directly in order to facilitate a less biased reading. In one place she wrote:
Couple of years later, my father began a new journey by enrolling in a master program in United States of America. He applied and, later, the IECP at X Uni-

Translanguaging in the classroom: Emerging issues for research and pedagogy versity accepted him. When the paper works were complete, my family and I traveled from Saudi Arabia to United States by air plane [P.S. I wanted to travel on a camel, but they were all rented!]. (D6)


I wasnt sure if the parenthetical comment was necessary. Therefore, I asked Buthainah: This might be considered a digression by some readers. How would you respond to that criticism? Buthainah responded:
Yes, it could be to some readers. However, when someone writes about themselves, they have to consider the stereotypes and whats going on around them that may inuence the comprehension or the interpretation of the text. I wrote that sentence because there are, still to this day, people who think that I, as a Saudi, ride camels to school. It is a joke that tries to remove that stereotype. In addition, a joke was needed here because I may have readers who hold negative associations toward my ethnicity. And I tried to elevate that tension that the reader may have, and hopefully, it will never occur (SR).

It appears as if Buthainah is being proactive in bringing the biases into the open and forcing the readers to deal with them, so that they can read her essay with an open mind. In this and the other strategies above, Buthainah is engaged in negotiating the appropriate footing to read her essay. We nd from lingua franca English (LFE) research that multilinguals engage in considerable foot work to establish a favourable context for language negotiation. Planken (2005) nds that Scandinavian business professionals engage in considerable preparatory work to create a safe space to negotiate their differences before they actually talk business. One specic strategy is joking about their accent so that both sides can lay aside their inhibitions and biases. This strategy of constructing a favourable context for language negotiation is what Kramsch (2009) calls symbolic competence. While communicative competence encourages sticking to the conventional codes for a context for successful uptake, symbolic competence refers to the possibility of resisting conventions and renegotiating contexts for alternate identities and meanings. Multilinguals are adept at such strategies for meaning construction. I label another set of strategies Interactional Strategies, as they are calculated to engage the reader into co-constructing meaning. The direct address to the reader that we see in the last example was repeated in many other places in her essay. I was concerned that such addresses would violate the formality of academic writing. Buthainah responded to the criticism as follows:
I knew that I was taking a risk by addressing the reader. But, I wanted the reader to be included into my discussion. I did not see my essay as a one-way informative essay. It is a negotiated essay that seeks a better understanding from educators and


Suresh Canagarajah future teachers to the multilingual experience. By addressing my readers, I am welcoming them to the discussion, which, in my perspective, [is] ongoing (MC).

By addressing the readers directly, she intends to draw them more closely into her writing to negotiate meanings with her. Here again there was uptake. Chrissie mentioned: There are also times when she directly communicates to the reader, which shows how much she values the readers opinion (PC, 10/28). The feeling of being valued and encouraged to contribute their own reactions help readers to orientate to the text differently. They move away from the conventional detachment of autonomous literacy to engage with Buthainah in co-constructing meaning. Buthainahs strategy of delaying translations for her translanguaged items also encourages readers to engage in interpretive work. As we can see in the opening, she not only delays giving clues for interpretation, she also makes them less obvious. The English version of the Arabic proverb is not introduced as a translation. Readers have to align this with the other contextual cues to assume that both are connected. It is this imaginative reading that Buthainah had in mind when she translanguages. She said:
If I translated everything, then the readers would simply go through it. But, if I did not translate it or provide an immediate translation, then, I am encouraging the reader to question the relationship between the poem and the stories being told and promote critical thinking (MC).

Here, again, evidence from readers suggests that the intended effect was achieved. Mark, an Anglo-Canadian student, said:
To me, a non-Arabic speaker, this quote is a beautiful collection of alien writing, fascinating but incomprehensible. It is a statement to me that there is something Buthainah understands that I do not. It is a move that distances me from Buthainah but also leaves me intrigued and interest[ed] in reading more (PC, 10/28).

It appears that translanguaging helped him read closely for clues for interpretation. The alien codes create a curiosity that enables him to read with greater attention. In this sense, translanguaging probably creates a more engaged reading than monolingual writing. That interlocutors in a multilingual contact situation construct meaning through alignment has been well researched (Atkinson et al. 2007). Since multilinguals can never come ready with all the codes for the very divergent communicative contexts in contact situations, they have to align language, objects and other ecological resources for interpretation. Khubchandani (1997) calls this synergy. Readers of Buthainahs article also had to look for alignment to interpret her text.

Translanguaging in the classroom: Emerging issues for research and pedagogy


The most daring case of translanguaging and a pointer to a more radical strategy was Buthainahs refusal to provide any clues for interpretation on one occasion:

When I was in fourth grade, I became sincerely interested in enrolling in the Communication Club (CC). Students in the club have the opportunity to give a speech in front of all of the attendees at the school . . . (D6)

When I asked Buthainah what she had in mind, she said:

Translating this poem would take so much of its value and providing a two sentence explanation will not do any justice for these few lines. The message of these lines is that who desires the best, need to work for it. He/she needs to stay up late working for it just like how divers have to search for the natural pearls. And those who try to get to the top and not work for it, they will waste their life getting nothing. I feel that these few lines that I wrote above about this poem do not give it any justice. Leaving it stand alone is more powerful (SR).

That left me wondering what readers would do to negotiate these lines. Rita adopted what resembled Firths (1996) let it pass principle. She said: I decided not to worry about what I couldnt understand. I trusted my classmates to explain what was important. (I: 05/09). We know from Firths research that when multilinguals confront a language feature they dont understand, they go on with the communication, hoping that further occurrences of the item will provide more context to renegotiate it at a later point. Ritas statement also shows that she adopts an extended temporal and social context for meaning negotiation. She hopes that she will get more clues for interpretation with others in the class later on, and that her interpretation doesnt stop with her own reading at one point. Some other students responded to the lines at an aesthetic level. Eunja, a Korean student, exclaimed: Written Arabic How elegant language it is! (Im not quite familiar with spoken one) (PC, 10/22). Mark advised: I absolutely love the Arabic phrases in the text, please keep them, and I hope theyre Arabic


Suresh Canagarajah

otherwise I feel kind of foolish (PC: 10/22). It is important to realize that autonomous literacy has led to readers extricating meaning from texts and not responding to the visual and aesthetic dimension of the text. Blommaert (2008: 113) points out: The essential role of aesthetics in the production of meaning has, however, not been widely recognized in the study of language. . . . The idea that one of the essential functions of language is the poetic production of forms has not made it into the mainstream, and form and content are still rmly seen as separate domains of analysis. At another level, it appeared as if Buthainah was also turning the tables on native speaker students through this strategy. Just as mastering English involved struggle, failure, humility and great effort for her (which is the theme of her literacy autobiography), Buthainah is simulating this experience for her readers. When I mentioned this interpretation in my earlier draft, Buthainah wrote in her member check commentary that she was indeed giving a sample or a taste of the experience that language learners go through to those who never experienced it, which may help them understand these stories and experiences better (MC). It appeared as if NES students understood this strategy. Mark said:
Something can only be scene [sic] perhaps in the Arabic text. Perhaps Buthainah is willing to help the reader but at some point somethings can only be known to those who are willing to learn and become Arabic English bilinguals (PC, 10/28).

Chrissie was prepared to revise her attitude based on this recognition. She understood that the roles had shifted in this writing. It was her job to take steps to interpret the difcult lines:
Although, last week I wrote that she explain her Arabic poems, I now feel that they are a key part to her narrative. She is indirectly showing us, the reader, who she is through these poems . . . Perhaps it is up to us to gure out the signicance of these words?? (PC, 10/28).

From these evidences, it appears that this is a performative strategy. What Buthainah is doing through these lines is more important than what she is saying. The experience and process of making meaning is as important as grasping the content. What we nd from Buthainahs writing is that she adopts strategies that persuade NES students to negotiate meanings with her.3 To a considerable extent, the NES students negotiate translanguaging like the multilingual students in the class. In this sense, the assumption that native students should be kept out of studies on contact situations as they bring divergent assumptions should be questioned. NES subjects can step out of native speaker norms to adopt multilingual orientations when the situation warrants it. Multilingual students

Translanguaging in the classroom: Emerging issues for research and pedagogy


are able to anticipate such outcomes and write appropriately to help negotiate meaning. 5.2. Orientations to literacy I dont mean to exaggerate the extent of uptake and meaning construction in this literacy event. Some students didnt appreciate the strategies Buthainah adopted or didnt succeed in constructing meaning. An NES student, Stacey, observed tongue in cheek: I think that Buthainah has denitely gone above and beyond in letting her native voice ring through. At times though, I felt as if I was reading a discourse on the education system in Saudi Arabia. . . . we are interested in her version of it more so than a description with her voice appearing every so often (PC, 10/22). What we have to realize is that meaning is not preconstructed or unitary in translanguaging. In writing that is co-constructed, meaning varies according to the participants and contexts. Mark understands this possibility and is open to such outcomes: Buthainah may be making different meta-statements to different groups. . . . To me, a non-Arabic speaker, this quote is a beautiful collection of alien writing, fascinating but incomprehensible. It is a statement to me that there is something Buthainah understands that I do not . . . To Arabic speakers this quote could be an act of solidarity, an appropriative move perhaps bringing in an common quote and transforming it into something more (PC, 10/22). These realizations conrm that multilingual literacy is different from autonomous literacy. In autonomous literacy, readers believe that they can extricate a preconstructed content through objective and disciplined reading. In multilingual literacy, readers have to be ready for variable meanings from situated interpretation. There are other differences in the assumptions Buthainah and other students in this class bring to writing. The text has a social dimension i.e., it is collaboratively constructed, with readers writing and writers reading, their roles fused. The text has an expanded temporal dimension readers and writers are prepared to treat negotiation as ongoing (in Buthainahs words). They take meaning negotiation outside the text to include their interactions and talk around the text in their ongoing social interactions. The text has an expanded spatial dimension i.e., participants are sensitive to the visual and aesthetic dimensions of the text. In these senses, multilingual literacy is multimodal it accommodates oral and visual modalities that are kept out of autonomous literacy. Multilingual literacy is also performative what readers do in reading is as important as what they understand. Translanguaging works most effectively in this orientation to the text. To prepare students to translanguage in literacy, we have to also encourage them to change their orientations to the text. If they


Suresh Canagarajah

bring assumptions of autonomous literacy to this kind of writing, they wont appreciate multilingual strategies. In the context of Internet and digital communication, texts are indeed getting redened in these directions by scholars in rhetoric and composition (see Nicotra, 2009; Ratcliffe, 1999). In terms of the strategies adopted, it is evident that Buthainah is employing many of the strategies multilinguals adopt in face to face conversation, as indicated above. However, these strategies nd different realization, given the modality of writing. There are different resources available for writers to adopt these strategies. For instance, Buthainah is able to exploit the microecology of the text (Creese and Blackledge 2010) to give more visual clues to her readers (through punctuation, font size, script, symbols and emoticons) for alignment and interpretation. In terms of the negotiation of power, we see that Buthainah is very agentive. She is aware of power differences and adopts creative strategies to renegotiate footing and status differences with her readers. In some cases, as in the strategy of simulating the experience of multilinguals in learning English, she compels NES to look at communicative experience through the eyes of the multilinguals. Since such examples may give the impression that Buthainah is overly condent, I must also point out that she is very cautious and sensitive in other places. Buthainahs translanguaging becomes more bold, daring and creative in subsequent drafts. Compare the opening of her sixth draft (quoted earlier) with her second draft:
Oh God! Give me more knowledge My education dictum through the years is a verse in the Quran stating: (D2).

Here she starts with the English translation rst and follows up immediately with the Arabic translation. In her sixth draft, not only did she start with the Arabic rst, she also delayed her translation, and made the connection oblique, getting readers to do more interpretive work. Such examples also suggest that she is preparing her readers gradually for translanguaging, aware that not everyone is prepared for this kind of textuality. The question of power brings us to the criticism that translanguaging is a threat to heritage languages. We nd, however, from Buthainahs statements above that she considers translanguaging as a means for voice. She feels she is able to represent her values and identities more effectively through translanguaging. Hybridity can enable multilingual students to represent their complex subject positions that defy essentialization or stereotyping. I additionally discovered that even within hybridity, mutilinguals have ways of indicating their desired codes and identities. Since Buthainah included very elaborate, lengthy and elegant lines in Arabic, but used only occasional and brief phrases in French, I asked her: Most French words (unlike Arabic) are very simple words. Some

Translanguaging in the classroom: Emerging issues for research and pedagogy


readers may say they dont serve any signicant rhetorical functions in the essay (unlike the Arabic quotations). How would you respond? Buthainah said:
The reason that I did not include French poems or more French phrases is because French can not be compared to my Arabic language. The value of the Arabic language is much greater than that of French simply because it is the language of the Quran and the language of my heritage. To treat French the same way, it would be simply strange. In addition, when a reader is paying a close attention to my selection of French words and my selection of Arabic word, it would be apparent that what it was stated in the Arabic language contain signicant meaning while what was stated in French can be easily replaced by English or Arabic. Does this make sense? (SR).

What we nd is that Buthainah is able to indicate the different values attached to the different languages in her repertoire, conveying her greater investment in Arabic. This nding also goes against the view that translanguaging is a monolith, resembling Auers (1999) uid lects which dont have socio-pragmatic meanings. 5.3. Toward assessment How effective is Buthainahs translanguaging? Are there signs that she is developmental? As we saw earlier, Butainahs linguistic choices become more effective, daring and creative in successive drafts. She is exploring her footing and voice through several attempts. At times she displayed uncertainty about her choices. She gave different reasons for her parenthetical comments. First, in a stimulated recall interview she mentioned that she wanted to humor the reader. In a member check response, she mentioned that she wanted to give the readers more personal information through the asides. Later, she mentioned that she wanted to disarm readers of their stereotypes. Thus she claries her intentions and objectives through the writing process. She is also difdent about translanguaging. At the end of her third draft, she added a note to readers: p.s. There is something in this essay that I do not like. I am not sure what it is but I feel that this essay is different from everything I have done in the past (in an unpleasant way) (D3). Such attitudes show that teachers have some work to do in giving students constructive feedback, channeling their linguistic resources in appropriate directions, or afrming their choices. It appears from Buthainahs experience that some students may be looking for such guidance. Furthermore, we have signs that Buthainah is not fully in control of her text. There are inconsistencies in her usage. She transliterates the following


Suresh Canagarajah

expression in three different ways: ma sha allah; Masha Allah; ma sha allah. When I asked her about these differences, she mentioned that they were mistakes, and her preferred usage was Ma Sha Allah. Similarly, there were spelling errors in other cases. I pointedly asked her about the choice in one case:
You misspell verses as versus. Since you have been very careful with your choice of Arabic and other stylistic devices in this essay, I was wondering how you would explain these spelling mistakes. Did you think these issues were less important? Did you think the readers will easily understand your meaning and therefore you dont have to worry too much about editing problems?

Buthainah replied:
I am quite embarrassed about this error (and another mistake below). I had multiple drafts of this essay, but did not notice this error. Of course, if I noticed it, I would have corrected it. I could have misspelled it, and the word document auto-corrected it. I was so engaged in developing the content that I did not notice it (SR).

What is interesting here is that Buthainah herself makes a distinction between error and mistake, indicating that teachers may have to help students identify them. Though I tried to pursue this distinction with Buthainah, I didnt make much progress in understanding her denitions of them. I provide some tentative denitions here: mistakes appear to be unintentional and unsystematic choices, such as the realization of Ma Sha Allah. However, when choices that are intentional fail to gain uptake, we can consider them errors. They can fail for many reasons. They may not have much rhetorical purchase. They may not also achieve success in encouraging readers to co-construct meaning. Buthainahs use of emoticons and smiley faces are of this category. Though she considered them part of her expressive repertoire, they were not identied by readers as enriching their interpretive experience. They may also be too informal for a university level essay. There are certainly more complex visual signs that can be used for expressive purposes. From this perspective, there is a social dimension to error. Errors occur when certain translanguaging choices are not effectively negotiated for meaning. Thus we can arrive at a practice-based or performative orientation to error, different from a norm- or form-based denition. An example that touches on error but also relates to rhetorical effectiveness are cases of mixed metaphors and unidiomatic expressions. In her opening paragraph, Buthainah writes:
As I type each word in this literacy autobiography, storms of thoughts stampede to be considered and mentioned. Which experiences should I value, which shall I consider, and which should I ignore. . . . As I click the keys on the keyboard, an

Translanguaging in the classroom: Emerging issues for research and pedagogy illustration of my literacy development shunt me to continue my ongoing learning adventure from my academic communities, my home, and my life experiences. (D6)


I asked her: The phrases I have highlighted . . . in this paragraph will be considered unidiomatic by native speakers. Did you have any second thoughts about using such phrases? Buthainah replied: I do not see why only bulls stampede this verb can be used guratively as well. I do not think that this is an issue of native speakers of English, I think that it is a stylistic choice (SR). Though Buthainah insisted that she considered these choices creative, some readers have told me that they consider them inappropriate. There is a difference of opinion here and, thus, a failure of uptake. In such cases, it is useful that my questions help Buthainah reect on her choices. My dialogical teaching approach helps Buthainah develop a reective and critical awareness of her choices. This pedagogical approach is transferable to other teaching contexts as well. Though teachers may not always feel condent about correcting students choices, especially in cases where they dont have competence in the repertoires of their students, their dialogical engagement with the texts through questions, feedback and reviews (in addition to peer review) will help student writers develop their translanguaging prociency further.

6. Conclusion To conclude, I would like to discuss the implications of translanguaging for the academic and writing prospects of multilingual students. How does translanguaging help them? Scholars have pointed out that the promotion of one form of language standard written English (SWE) as the register for academic writing could alienate multilingual students and restrict their options for voice (Smitherman 2003). At the same time, we cannot say that anything goes, and allow students to adopt any registers and conventions they want in academic writing. A pragmatic resolution is to take the existing conventions seriously but nd ways of bringing in ones codes and values in a guarded and appropriate manner. This would be a strategy of resistance from within. It would gradually pluralize the academic text and facilitate more far- reaching changes over time. Some multilingual scholars are already adopting such a writing approach. I have published elsewhere the writing style of African American sociolinguist Geneva Smitherman (see Canagarajah 2006). She introduces African American Vernacular English for limited purposes in the overall framework of SWE. Her strategy is a form of translanguaging, and proves that academic conventions


Suresh Canagarajah

are open to negotiation. I show such examples to students to suggest that they too can negotiate academic conventions to bring in their repertoires for voice. I emphasize that this kind of resistant writing should be undertaken with caution. Their choices should be rhetorically and contextually well motivated. Though students will often gain uptake and succeed, I also tell them that this strategy is risky. However, for many multilingual scholars and students, such a risk is worth the price for the eventual pluralization of academic literacy and classroom discourse.

1. Though the notion of the native speaker has been deconstructed and rejected, I use the label here for want of a better term to distinguish students and subjects who speak dominant varieties of English. 2. Some of the typographical idiosyncrasies in the interview comments are explained by the fact that they were conducted through email. 3. For a more comprehensive list of Buthainahs strategies, see Canagarajah, forthcoming.

Atkinson, D., Churchill, E., Nishino, T., & Okada, H. 2007. Alignment and interaction in a sociocognitive approach in second language acquisition. Modern Language Journal 91. 169188. Auer, Peter. 1999. From codeswitching via language mixing to fused lects: Toward a dynamic typology of bilingual speech. International Journal of Bilingualism 3. 309 332. Barbour, Stephen. 2002. Language, nationalism, and globalism: Educational consequences of changing patterns of language use. In Paul Gubbins & Mike Holt (eds.), Beyond boundaries: Language and identity in contemporary Europe, 1118. Clevedon, UK: Multilingual Matters. Bhatia, T. K. & W. Ritchie. 2004. Bilingualism in South Asia. In T. K. Bhatia & W. C. Ritchie (eds.), The handbook of bilingualism, 780807. Oxford: Blackwell. Blommaert, Jan. 2008.Grassroots literacy: Writing, identity and voice in Central Africa. London: Routledge. Blommaert, Jan. 2005. Discourse: A critical introduction. Cambridge: Cambridge University Press. Canagarajah, A. Suresh. Forthcoming. Codemeshing in Academic Writing: Identifying Teachable Strategies of Translanguaging. Modern Language Journal.

Translanguaging in the classroom: Emerging issues for research and pedagogy


Canagarajah,A. Suresh. 2006.The place of World Englishes in composition: Pluralization continued. College Composition and Communication 57. 586619. Canagarajah, A. Suresh. 2002. Critical academic writing and multilingual students. Ann Arbor: University of Michigan Press. Cook, Vivian. 1999. Going beyond the native speaker in language teaching. TESOL Quarterly 33(2). 185209. Cope, Bill & Mary Kalantzis (eds.). 2000. Multiliteracies: Literacy learning and the design of social futures. London and New York: Routledge. Creese, Angela & Adrian Blackledge. 2010. Translanguaging in the bilingual classroom: A pedagogy for learning and teaching? Modern Language Journal 94 (1). 103115. Council of Europe. 2000. Common European framework of reference for languages: Learning, teaching, assessment. Language Policy Division. Strasbourg. http://www. EN.pdf (accessed 24 November 2010) De Souza, L. Mario. 2002. A case among cases, a world among worlds: The ecology of writing among the Kashinawa in Brazil. Journal of Language, Identity, and Education 1 (4). 261278. Elbow, Peter. 2002. Vernacular literacies in the writing classroom? Probing the culture of literacy. In Schroeder, Christopher, Helen Fox, & Patricia Bizzell (eds.), ALT DIS: Alternative discourses and the academy, 126138. Portsmouth, NH: Boynton/Cook. Firth, Alan. 1996. The discursive accomplishment of normality. On lingua franca English and conversation analysis. Journal of Pragmatics 26. 237259. Firth, Alan & Wagner, Johannes. 1997. On discourse, communication, and (some) fundamental concepts in SLA research. Modern Language Journal 81. 285300. Franceschini, Rita. 2010. Multilingualism: Theoretical and historical perspectives. Paper presented at American Association of Applied Linguistics, Atlanta, 6 March. Garcia, Ofelia. 2009. Bilingual education in the 21st century: A global perspective. Oxford: Wiley-Blackwell. Gorter, Durk. 2006. Linguistic landscape: A new approach to multilingualism. Clevedon: Multilingual Matters. Gutierrez, Kris. 2008. Developing a sociocritical literacy in the third space. Reading Research Quarterly 43. 148164. Heller, Monica & Marilyn Martin-Jones (eds). 2001. Voices of authority: Education and linguistic difference. Westport, CT, and London: Ablex Hill, Jane. 1999. Styling locally, styling globally: What does it mean? Journal of Sociolinguistics 3. 542556. Hornberger, Nancy (ed.). 2003. Continua of biliteracy: An ecological framework for educational policy, research, and practice in multilingual settings. Clevedon: Multilingual Matters. Horner, Bruce, Min-Zhan Lu, Jacquelyn Royster & John Trimbur. Forthcoming. Language Difference in Writing: Toward a Translingual Approach. College English. Jenkins, Jennifer. 2006. Current perspectives on teaching World Englishes and English as a Lingua Franca. TESOL Quarterly 40 (1). 157181.


Suresh Canagarajah

Jrgensen, J. N. 2008. Poly-Lingual languaging around and among children and adolescents. International Journal of Multilingualism 5 (3). 161176. Khubchandani, L. M. 1997. Revisualizing boundaries: A plurilingual ethos. New Delhi, India: Sage. Kirch, Gesa & Jacquelyn Royster. 2010. Feminist rhetorical practices: In search of excellence. College Composition and Communication 61 (4). 640672. Kramsch, Claire. 2009. The multilingual subject. Oxford: OUP. Lam, W. S. Eva. 2004. Second language socialization in a bilingual chat room: Global and local considerations. Language Learning & Technology 8(3). 4465. Lin, Angel & Peter Martin (eds.). 2005. Decolonisation, globalisation: Language-ineducation policy and practice. Clevedon: Multilingual Matters. Lu, Min-Zhan. 2009. Metaphors matter: Transcultural literacy. Journal of Advanced Composition 29. 285294. Lyons, S. (2009). The ne art of fencing: Nationalism, hybridity, and the search for a Native American writing pedagogy. Journal of Advanced Composition 29. 77106. Nicotra, Jodi. 2009. Folksonomy and the Restructuring of Writing Space. College Composition and Communication 61. 259276. Pennycook, Alastair. 2010. Language as a local practice. London: Routledge. Pennycook,Alastair. 2007. Global Englishes and transcultural ows. London: Routledge. Planken, B. 2005. Managing rapport in lingua franca sales negotiations: A comparison of professional and aspiring negotiators. English for Specic Purposes 24. 381400. Pollock, Sydney. 2009. The language of the gods in the world of men: Sanskrit, culture, and power in premodern India. Los Angeles: University of California Press. Rampton, Ben. 2008. Language in late modernity: Interaction in an urban school. Cambridge: CUP. Rampton, Ben. 1999. Styling the other: Introduction. Journal of Sociolinguistics 3. 421 427. Ratcliffe, Krista. 1999. Rhetorical Listening: A Trope for Interpretive Invention and a Code of Cross-Cultural Conduct. College Composition and Communication 51. 195224. Spivak, Gayatry. 1993. Outside in the teaching machine. New York: Routledge. Stark, Ryan. 2008. Some aspects of Christian mystical rhetoric, philosophy and poetry. Philosophy and Rhetoric 41. 260277. Street, Brian. 1984. Literacy in theory and practice. Cambridge University Press, Cambridge. Smitherman, Geneva. (2003). The historical struggle for language rights in CCCC. In Geneva Smitherman & Victor Villaneuava (eds.), From intention to practice: Considerations of language diversity in the classroom, 739. Carbondale: Southern Illinois University Press. Viswanathan, Gauri. 1989. Masks of conquest. New York: Columbia University Press. Williams, Bronwyn. 2009. Multilingual literacy strategies in online worlds. Journal of Advanced Composition 29. 255258.

Translanguaging in the classroom: Emerging issues for research and pedagogy


Young, Vershawn. 2004. Your average Nigga. College Composition and Communication 55. 693715. Suresh Canagarajah is the Erle Sparks Professor in the Departments of English and Applied Linguistics at the Pennsylvania State University. He taught earlier at the University of Jaffna, Sri Lanka, and the City University of New York. He teaches ethnographic research methods, World Englishes, and theories of rhetoric and composition. Email:

Modernist language ideologies, indexicalities and identities: Looking at the multilingual classroom through a post-Fishmanian lens

Abstract This paper focuses on the construction of immigrant minority pupilsidentities in a regular multicultural primary school classroom in the Netherlands. It presents three ethnographic data sets. The rst set features the evaluative discourse of a Dutch medium primary school teacher and it focuses on the ways in which this class teacher indexes pupils identities on an axis of (linguistic) disorder versus order on the basis of an attributed monolingual upbringing. The second set features the evaluative discourse of Moroccan girls of both Berber and Arabic-speaking origin. Although the Dutch language is a given in their lives, their identity belongings are strongly anchored on the axis of purity versus impurity, established on the basis of their language skills in the immigrant minority language. The third set features a sabotage move perpetrated by two pupils of Turkish background, who assert the validity of international languages as opposed to their home language and who use fakeArabic to escape the pressure exerted on them by the classroom researcher. The paper concludes by proposing a revisited understanding of multilingualism that can give justice to the complexity of the pupils own sociolinguistic repertoires and identity performances. This renewed understanding is based on a post-Fishmanian awareness that sees language use and identity construction as polycentric semiotic performances not necessarily bound to groups. Key words: Identity, multilingualism, Fishman, immigrant pupils, primary education, the Netherlands, interpretive ethnography

1. Introduction Cultural and linguistic diversity in Dutch mainstream society have been the object of heated public and political debates for decades, with increased attention being paid to the need for integration (more recently addressed as participa-


Massimiliano Spotti

tion) of immigrant minority group members within mainstream Dutch society. This integration is marketed and sold as something to be achieved through the means of Dutch alone (Extra and Spotti 2009). As a result of immigration, the Dutch educational system has also been confronted with complex new patterns of multilingualism and the identities of immigrants and their offspring, and primary education is one of the institutionalized environments in which monoglot policing has taken place. That is, it is one of the institutional environments in which government policy has enabled the Dutch language, rather than being side kicked by immigrant minority languages, to be the only language of instruction in the curriculum (Bezemer and Kroon 2007). This situation, however, is in sharp contrast with the ndings of ethnographic research that reconstruct a discontinuity between monoglot language policies and the heteroglot language repertoires of immigrant minority pupils (cf. Bezemer 2003; Spotti 2006). With the above as backdrop, this paper focuses on the construction of immigrant minority pupils identities in a regular multicultural primary school classroom in the Netherlands. It presents three ethnographic data sets. The rst set features the evaluative discourse of a Dutch medium primary school teacher and it focuses on the ways in which this class teacher indexes pupils identities on an axis of (linguistic) disorder versus order on the basis of an attributed monolingual upbringing. The second set features the evaluative discourse of Moroccan girls of both Berber and Arabic-speaking origin. Although the Dutch language is a given in their lives, their identity belongings are strongly anchored on the axis of purity versus impurity, established on the basis of their language skills in the immigrant minority language. The third set features a sabotage move perpetrated by two pupils of Turkish background, who assert the validity of international languages as opposed to their home language and who use fake Arabic to escape the pressure exerted on them by the classroom researcher. The paper concludes by proposing a revisited understanding of multilingualism that can give justice to the complexity of the pupils own polylingual sociolinguistic repertoires and identity performances. The central concepts here are modernist language ideologies, indexicality, and identity. Together, these three concepts help construct a viable conceptual pathway for the study of identity construction. It is therefore necessary to outline my understanding of these three concepts at the outset of this paper.

2. Modernist language ideologies, indexicality and identities Modernist language ideologies are belief systems that have served, and still serve, nation-states and their institutional ramications such as education in

Modernist language ideologies, indexicalities and identities


setting up and perpetrating national order (Baumann and Briggs 2003; see also Silversteins 1996, 1998 work on a culture of monoglot standard). Modernist language ideologies present languages as codied in specic artefactualised linguistic objects: grammars, dictionaries etc. (Blommaert 2008) that have a name (e.g. Dutch, Turkish, Arabic, etc.), and whose speakers have clearly denable ethnolinguistic identities, i.e., I am a speaker of language X and therefore I am a member of group Y. These ideologies, seen as contributing to the maintenance of national order, revolve around two tenets: the establishment of a standard or norm for language behaviour that is common to all inhabitants of a nation-state, and the rejection of hybridity and ambivalence in any form of linguistic behaviour. Of these two closely related tenets, the former is the goal towards which the latter is seen to contribute. That is, the rejection of hybridity is embedded in the search whether in writing or in pronunciation for a standard (see Agha 2003 for a comprehensive explanation of the emergence of Received Pronunciation of English [RP] as a product of characterological discourses). The standard is presented as the norm and, as such, is sold and marketed as the uncorrupted variety of the ofcial/national language and often associated with the righteous moral values of its users (see Agha 2003: 231273). Finally, given that languages are understood as nite entities bound by syntactical rules and grammars, their usage can be assessed. From this it follows, then, that there will be language users whose use of language can be evaluated as better than that of others. As for education, categorising pupils on the basis of how skilful they are in the usage of the standard variety, or even in the usage of the school variety of a certain language, holds deep implications for identity construction. This leads us to the concept of indexicality. Any bits of language that someone uses carry an ideological load in that, in addition to their referential meaning, they also carry either pragmatic or social meaning (i.e. have indexicality). In other words, any bits of a language that one uses are potentially subject to evaluation against the standard/norm from others who inhabit the same socialisation space. A poignant example of this indexicalisation process is the evaluation of accents, which can be embedded in peoples discourse on language use (e.g., he speaks like a farmer or he surely is from the capital), and that are drawn on grounds of often implicit shared complexities of indexicality within a certain centering institution (see Dong 2009: 7273). For instance, an accent can be evaluated as funny because it indexes distance from the authorised standard accent which in turn is an index of prestige and constructs the identity of those performing it as an identity of someone who is well schooled. Indexicality is therefore the connective cement that links language use to social meanings, and all this is done through institutionally authorised evaluative discourses. This means that in any act of


Massimiliano Spotti

language use, there is always identity work involved and that indexicality points to the grassroots displays of groupness. Consequently, every utterance, even when not explicitly about identity, is an act of identity performance. This leads us to the third and nal concept of the conceptual framework employed here: that of identity. Space constraints do not allow for a complete review of the concept of identity (see Block 2006; Dong 2010; Joseph 2003; Spotti 2007). For the present purpose, it should sufce to pin down three things. First, identity is not something that someone possesses. Rather, it is something that someone constructs in social practice within a space of socialisation. Second, identity is not monolithic. Instead, it consists of a series of performative acts that take place according to the socialisation space one occupies. We can therefore talk of identities instead of identity and identities are constructs that are built on the basis of semiotic resources at ones disposal within a certain socialisation space. Third, identities are inhabited as well as ascribed. Inhabited identities refer to self-performed identities through which people claim allegiance to a group. Conversely, ascribed identities are attributed to one by others on the basis of evaluative criteria that make one either well-tted or ill-tted for a socially circumscribed category (e.g., the good neighbour, the bad student, the college beauty, the nerd; see also Goffman 1981). How do modernist language ideologies, indexicalities, and identities work together, then? Borrowing from Bakhtin (1981: 293), in any stratied urban society, languages, the connections between language varieties and the identities of different groups are not as straightforward as modernist language ideologies would have us to believe. Varieties are indexes of diverse, often conicting, symbolic meanings of social, cultural and ethnic belongings. More simply put, the bits of language that someone uses are not only a means for the direct expression of someones intentions but they are also objects that index identity belonging both in ones own eyes (inhabited identity) and in the eyes of others (ascribed identity). Language(s) and their words therefore carry an ideological load (see Rampton 2005: 75) because they are subject to the values at play at the time and in the space in which they are uttered (Blommaert 2005: 222223). It is according to the centring institution that someone is either part of, or tries to gain access to, that ones identity is constructed as that of a good (insider) member or a bad (outsider) member. This is done on the basis of either how successfully, or unsuccessfully, one manages to embrace the complexity of indexicalities present within that specic socialisation space. The evaluative discourses that construct identities result from either the respect or trespass of situated language norms. These, in turn, revolve around the central values of the centring institution in which bits of language have been deployed. A case in point would be a pupil

Modernist language ideologies, indexicalities and identities


learning standard Dutch who, engaged in mapping graphemes onto phonemes in a primary school classroom, fails to use standard Latin script and, consequently, is assessed as having sloppy orthographic skills, (mis)recognised as having faulty literacy skills and, nally, ascribed the identity of an illiterate pupil (see Blommaert, Cleve and Willaert 2006). The phenomenology of super-diverse migration movements that makes up for a new form of diversity that impinges upon the group oriented diversity that has characterized (Western) Europe from the 1970s onwards, shows that mobility of people involves mobility of linguistic and semiotic resources. It follows, that a sedimented understanding of language and identity as nite entities, and an understanding of language use according to sedentary patterns is now complemented by translocal forms of language use (see Jacquement 2005). The combination of both holds unexpected sociolinguistic effects and it has outreaching consequences for peoples identity performances.

3. The study The study was part of a larger interpretive ethnographic inquiry on identity construction in one Dutch and one Flemish multicultural primary school classroom. Both case studies aimed at analysing evaluative discourse ethnographically to understand how immigrant minority pupils identities are constructed in the discourse authored by policy documents, school staff members and pupils, as well as in classroom interactions. The data consists of eld notes, long open-ended interviews with staff members, focused group discussions with the pupils and audio recordings of classroom interactions. Approximately 55 hours of classroom interactions were observed and audio taped. The observer never sought to actively participate in the classroom interactions. Interviews followed the plot of the long open-ended interview (McCracken 1988), and they were structured around general topics like the educational and professional background of the teacher, the knowledge that s/he held of the pupils sociolinguistic background, of their ethno-linguistic, cultural and religious belongings as well as about events that were observed during the unfolding of the classrooms daily life. The rst interview was conducted with the classroom teacher after a weeks visit in the classroom and another three interviews were conducted either to further elucidate the teachers evaluative discourse (indicated below with S02) or to gain the retrospective view of the teacher on the taped classroom episodes. Central to the analysis here are also the focused group discussions carried out with the pupils (below indicated with GD01). The groups were formed based on the quantity of contact that pupils


Massimiliano Spotti

had with each other. The discussions turned out to be friendly chats where pupils could express their views on topics that emerged from the questionnaires administered to them to gather some initial information on their ethnic, linguistic and religious belongings and from the eld notes drawn during the observation period. All discussions took place in the afternoon, mainly in the schools staff room and lasted between 30 to 45 minutes for each group. The discussions were all audio taped and the pupils were made aware that the audiotape recorder was on as the group chat started. Their discussions touched upon various topics. Starting from the pupils knowledge of their parental patterns of migration, the discussion moved to the exploration of the pupils own understanding of their identity belongings. As in the interviews with staff members, my position in all the discussions was limited to giving them prompts and asking them to either expand on or clarify their statements. I showed my curiosity in what they had to say and tried to limit, when needed, the intervention of the more talkative ones to allow each group member to participate. As for classroom interactions, a pool of the recordings was selected and transcribed1 from the synopsis drawn out of the eld notes and audiotapes. These were thought to have the potential to be selected as incidents that shed light on how identities of immigrant minority pupils are constructed in interaction and contribute to construct a description so that others may see what members of a social group need to know, understand, interpret, and produce to participate in appropriate ways (Green and Bloome 1997: 186). The transcriptions of the recordings are presented in English, with the Dutch text underneath in italics. These transcriptions were combined together with the eld notes gathered during the observation time for a tentative analysis and interpretation along the lines of the key incident approach (Erickson 1986; Kroon and Sturm 2000). As Erickson (1986: 108) points out in illustrating the meaning of the term key:
[a] key event is key in that the researcher assumes intuitively that the event chosen has the potential to make explicit a theoretical loading. A key event is key in that it brings to awareness latent, intuitive judgments the analyst has already made about salient patterns in the data. Once brought to awareness these judgments can be reected upon critically.

The reviewing of the pool of incidents initially selected on the basis of the researchers intuitive assumptions has given way to a rst tentative analysis that was then either discarded or taken further in a more coherent and deeper analysis and interpretation of the incident in hand.An incident is selected as keyin that it represents tangible instances of the working of the cultural ecology (the normal) of a certain sociocultural space and of its social organization. Its analysis has

Modernist language ideologies, indexicalities and identities


helped to shed light on how identities are performed in the discourses present in the classroom under investigation (cf. Guba and Lincoln 1989: 176; Polanyi 1989). 3.1. The school, the classroom and the pupils The data that I present here were collected in the school year 20042005 at St. Josephs Catholic Primary, a regular multicultural primary school in Duivenberg, a medium-sized city of approximately 200,000 inhabitants in the south of the Netherlands. At that time, the school had a high concentration of immigrant minority pupils and an exclusively Dutch-speaking teaching staff. On February 15th 2005, Form 8a at St. Josephs Catholic Primary amounted to eighteen pupils in total, eight boys and ten girls. The age of the pupils ranged from eleven to thirteen years due to some pupils having repeated one or more school years. None of the pupils had been enrolled during the ongoing school year; thirteen of them had attended St. Josephs Catholic Primary since Form 1. All pupils reported to be of immigrant minority background. According to the school register, all pupils but one had been assigned an educational weight of 1.902 i.e. they were all registered as pupils in need of additional educational support as a consequence of their parents low educational and socio-economic background (the norm for educational weight being 1.0). The exception is Walid, who has an educational weight of 1.0 and whose parents were both born in Morocco and are highly educated. All pupils reported to speak a language other than or another alongside Dutch at home. Concerning the country of birth of the pupils, thirteen out of the eighteen pupils were born in the Netherlands. Out of the remaining ve pupils, three were born in the Dutch Antilles, one in BosniaHerzegovina and one in Morocco. Half of Form 8a came from the Moroccan immigrant community and of these pupils only Walid and Khalid were born to parents of Moroccan Arabic-speaking background, while the rest was born to Berber parents. Among the latter, Hajar born in the Netherlands to a father of Arabic-speaking background and a mother of Berber background understands and speaks Berber. However, her network of classroom friendships claims to address her mostly in Moroccan Arabic. Affa was the only pupil born to a second-generation Moroccan Arabic-speaking father and a rst-generation Moroccan Arabic-speaking mother. In order to gather information on the home languages present in the classrooms under investigation, all pupils have been asked by their class teacher to ll in a home language survey (cf. Broeder and Extra 1998). Table 1 reports the home languages, gender and names of the pupils as gathered from the home language survey carried out in this class. All names of the pupils are ctive.


Massimiliano Spotti

Table 1. Gender, names and home languages of the pupils in Form 8a Boys Khalid Soan Roble Cemal Walid; Zakariya Joshwa Osman Girls Samira; Lemnja; Siham Rhonda Ozlem Hajar; Affa Lejla Meryem Micheline Home language(s) Dutch and Arabic Dutch and Berber Dutch and Papiamentu Dutch and Somali Dutch and Turkish Dutch, Arabic and Berber Dutch, Bosnian and Croatian Dutch, Papiamentu and English Dutch, Turkish and Arabic Dutch, Papiamentu, English and Spanish

The data gathered from Form 8a home language survey are not in agreement with the annotations made in the class register by the Form 8a class teacher. She, in fact, relied on her own well-educated guess about the pupils (supposed yet untapped) ethnic afliation and home languages as well as on the information given in the pupils enrollment forms. The gender, names and home languages of the pupils as they appeared in the class register are reported in Table 2.
Table 2. Gender, names and home language(s) of the pupils following Form 8a register Boys Osman Walid; Zakariya Joshwa Roble Cemal Girls Affa Samira; Lemnja; Siham Lejla Rhonda Meryem; Ozlem Micheline Home language(s) Dutch Dutch and Moroccan Dutch and Bosnian Dutch and Papiamentu Dutch and Somali Dutch and Turkish Dutch, Papiamentu and English

The class register does not report any information on Hajar, Khalid and Soan. It also indicates that Osman, born in the Netherlands to Turkish parents, and Affa, born in the Netherlands to a second-generation Moroccan father and a rst-generation Moroccan mother, only have Dutch as their home language. Further, while the home language survey indicates Berber as one of the home languages for eight pupils of Form 8a, in the class register the home language of these pupils is given under the umbrella term Moroccan. The class register also does not report the use of any language other than Turkish for the pupils coming from the Turkish group. In the home language survey, though, Arabic

Modernist language ideologies, indexicalities and identities


is also mentioned by half of the pupils with a Turkish background who attend Quran classes at the weekends.

4. Multilingualism through a modernist lens: the class teacher Miss Sanne, the class teacher of Form 8a, is 23 years old. She was born in Duivenberg to Dutch native parents, she holds Dutch nationality and she has lived in Duivenberg all her life. Sanne was brought up in a multicultural neighborhood. In Sannes view there is simply nothing special about foreign people; they are just, you know, they live here too (S03: 57) and she believes that her way of thinking about foreigners has been strongly inuenced by her upbringing as she learned that we all live here (. . . ) we live in the Netherlands and we have to do it all together with each other [. . . ] (S03: 59). In recalling her primary school experience that started in 1986, she states that there were indeed a few children from immigrant minority groups in her class, but not so many as at St. Josephs, and that they were all just able to get on with each other. Miss Sannes statement I have not a single Dutch child in my class is used as an explanation for why her pupils perform worse than those pupils at other schools in Duivenberg. The lack of parental qualications and these parents being non-native Dutch are the basis for Miss Sannes own reasoning in explaining St. Josephs extra investment in the Dutch language with a particular focus on vocabulary. We now move further in the analysis of Miss Sannes evaluative discourse and we encounter the cases of two pupils, i.e. Mohammed and Lejla, whose language attributions marked the opposite ends of the ascriptive category immigrant minority pupil. 4.1. Mohammed Miss Sanne starts with Mohammed, a thirteen-year-old Somali child who was in Miss Sannes class during the previous school year. At that time, Mohammed, who had been in the Netherlands since he was eight years old, was uent in the Somali language (S02: 314). However, in Sannes discourse, prociency in the Somali language turned out to be detrimental to Mohammeds Dutch language development because:
Sanne: So he had (. . . ) when he was eight so he had to learn a second language Dus die heeft (. . . ) toen ie acht was heeft ie dus een tweede taal moeten leren Max: (hmm) Sanne: and the Somali language has a different sentence structure (. . . ) en Somalische taal heeft een andere zinsopbouw (. . . )


Massimiliano Spotti Max: (hmm) Sanne: than the Dutch language so he always spoke in twisted sentences. dan de Nederlandse taal dus hij sprak altijd in kromme zinnen Max: (hmm) (S02: 316321)

At the age of eight, Mohammed was already uent in Somali his mother tongue and he had to learn Dutch as a second language. As Miss Sanne reports in the coordinate phrase that follows (318), the Somali language has a different sentence structure to Dutch. This has led Mohammed to use Somali syntax in Dutch and to always speak in twisted sentences, i.e., abnormal sentences compared to standard Dutch or, at least, the local variety of Dutch spoken in the city where the school is located. Mohammeds difculties in speaking Dutch properly are explained with the syntactical interference hypothesis where the second language learner inappropriately transfers structures of his rst language to the second. Miss Sanne adds:
Sanne: And if you get it also at home, because that mother, she, of course, was also having problems with that [Dutch language: MS] herself En als je dat ook van thuis uit, want die moeder, die was, natuurlijk, daar ook mee aan het stoeien Max: (hmm) Sanne: and that father too, he also spoke hardly any Dutch. en die vader ook die sprak ook nauwelijks Nederlands Max: (hmm) Sanne: so he could not hear it properly from home either so he (. . . ) yes he used lets say the Dutch language with the structure Dus hij kon het ook niet van thuis uit goed aanhoren dus hij (. . . ) ja hij gebruikte zeg maar de Nederlandse taal met de opbouw Max: (hmm) Sanne: from the Somali language. vanuit de Somalische taal. (S02: 323329)

Mohammed not only uses twisted sentences in Dutch because his language use is based on the structure of Somali, a language that uses SOV-order in its main clause in comparison with the Dutch SVO-order (cf. Saeed 1999). Also, as introduced by the causative conjunction so, both Mohammeds parents are responsible for the syntactical interference between the two nite linguistic entities that in the dichotomy presented by the teacher are part of Mohammeds repertoire, i.e., Somali and Dutch. The father, in fact, spoke no Dutch and the mother also suffered from Somali sentence structure in her use of Dutch. The

Modernist language ideologies, indexicalities and identities


parental lack of Dutch prociency has consequences for Mohammeds identity, as the lack of Dutch in the home is indexical of a pupil with a language disadvantage. 4.2. Lejla Miss Sannes discourse dealt also with Lejla, an eleven-year-old girl born in Bosnia- Herzegovina to Bosnian parents who came to the Netherlands when she was three years old. Miss Sanne explains:
Sanne: Lejla is also (. . .) lets see she has lived here ever since she was three or so, therefore also still really very young when she already . . . . . . ?? a new language (. . .) look and small children can pick up a (. . .) another language really easily that is simply, yeah, scientically proven. Lejla die is ook (. . . ) even kijken die woont hier al sinds dat ze drie is of zo dus ook nog heel erg jong dat ze al een nieuwe taal (. . . ) kijk en kleine kinderen kunnen heel makkelijk een andere taal oppikken dat is gewoon, ja, wetenschappelijk bewezen Max: (hmm) Sanne: And indeed she is also better at Dutch than other children and that is also because her parents have also just spoken Dutch at home from the beginning. En zij is ook inderdaad beter in het Nederlands dan andere kinderen en dat komt ook omdat haar ouders ook gewoon vanaf het begin af aan hier gewoon ook thuis Nederlands praten. (S02: 443445)

In the utterances above, Lejla is in an advantaged position in picking up a second language because she came to the Netherlands at a very young age. Further, Miss Sanne tries to obtain objectiveness for her claim. In the utterance look and small children can pick up a (. . .) another language really easily, she uses the imperative look to substantiate the evidence of her claim. Further, she calls upon the critical age hypothesis, implying that a putative language learning function is much more developed in younger children who approach the learning of a second language more easily than those who approach a second language at an older age (see Singleton 1994, 129 for a comprehensive discussion of the age factor in second language acquisition). Not only is the age at which Lejla came into contact with Dutch relevant; her parents language behaviour is also now regarded as a key element in Lejlas good language development. It is noteworthy that Lejlas parental language behaviour is accompanied by the adverb simply. The use of this adverb may indicate that the practice of


Massimiliano Spotti

speaking Dutch at home is regarded by Miss Sanne as nothing more than what parents should do by default with their children at a young age. However, at home, Lejla and her parents have a language repertoire that includes Croatian, English, Dutch and Bosnian, the latter being the language Lejla denotes as her language, and in which she claims to have both passive and active literacy skills that she reports to use for verbal exchanges with her younger siblings and with her parents. Miss Sanne therefore looks at Lejlas multilingual repertoire through a monolingual lens that sees one language in the home, that is Dutch, being key to Lejlas being good at Dutch. This is in contrast to Mohammed whose parents Dutch, or the lack thereof, caused him to use a twisted sentence structure and therefore deviate from the standardform of expression. However, in terms of their institutional identities, the sociolinguistic difference between the repertoires of these pupils is erased as they are both categorised as 1.90, i.e. as pupils who are almost twice as heavy to teach compared to a pupil with Dutch-native parents.

5. Multlingualism through a modernist lens: the Moroccan girls In spite of Miss Sannes general view about the poor results of her pupils in Dutch, the immigrant minority pupilslanguage use in Form 8a showed a marked preference for Dutch in daily language exchanges both within and outside the classroom. Arabic and Berber, being the most popular home languages in the classroom, of course also had their own space, although mostly outside formal instruction time. During a gym lesson, for instance, when the girls were playing handball, one could indeed nd a more prominent rate of home language use than in the classroom, and, as can be expected, Arabic and Berber were popular in derogatory exclamations and imperatives such as pass me the ball. The marked preference for Dutch was quite different when the girls were engaged in a focused group discussion concerning the diacritics of identity belonging, i.e. language use and ethnic afliation. The discussion I had with them unfolded as follows:
Lejla: Them live in the Netherlands also and are a hundred percent Moroccan too. Hun wonen toch ook in Nederland en zijn ook honderd procent Marokkaans. Is that so girls? Is dat zo meisjes?


[The other girls agree loudly]

Modernist language ideologies, indexicalities and identities Max: Lemnja: Yeah right, behave properly. Ja hallo, doe normaal. What do you mean a hundred, what do you mean a hundred, a thousand right? Wat honderd wat honderd, duizend of niet? A thousand percent Moroccan, what do you mean? Duizend procent Marokkaans, hoe bedoel je? Simply. Gewoon. A million. Miljoen. My whole life. Mn heel leven. Uncountable. [At the same time as Hajar] Ontelbaar. Uncountable Moroccan? Ontelbaar Marokkaans? Oh no, oh no, one percent is for Dutch as I speak I do talk that mostly. Oh nee, oh nee, n procent is voor Nederland want ik spree ik praat dat wel het meeste. Yes one little percent then. Ja n procentje dan. One comma zero zero percent. E n komma nul nul procent. e


Max: Hajar: Lemnja: Hajar: Lemnja: Max: Hajar:

Lemnja: Hajar:

(GD01: 05017)

Lejla compares her own being Bosnian with her fellow classmates, stating that even though they were born in the Netherlands, these girls are also a hundred per cent Moroccan. When the girls, who are all of Berber origin, react to Lejlas statement they all voice an outspoken, i.e., more than a hundred per cent, afliation to the umbrella term Moroccan: what do you mean a hundred, a thousand right?, all my life, uncountable. This unquantiable afliation to being Moroccan is tempered by the role that Dutch has in these girls everyday lives. Hajar, in fact, admits that Dutch is the language she speaks the most and this leads her to attribute one comma zero zero percent to Dutch; a point also raised by Lemnja who states one little percent then. When dealing with the role of Berber and Arabic in Form 8a, Lemnja and Hajar eagerly explained the difference between these two languages and how they came to be the languages of Morocco. After that, the discussion continued as follows:
Max: Yes and then the Arabs came there. So you speak Berber? Ja en dan de Arabieren zijn daar gekomen. Dus jullie spreken Berbers?


Massimiliano Spotti Siham: Yes. Ja. Lemnja: Pu[:]re Berber. Pu[:]re Berbers. Hajar: Pu[:]re Berber. Puu[:]r Berbers. Siham: They, they are half Arabic they are (. . .) [Looking at Samira and Lemnja] Zij, zij zijn half Arabisch zij zijn ( . . . ) Hajar: I, I, I (. . .) Ik, ik, ik ( . . . ) Samira: Half Berber. [Shouting out] Halve Berbers. Hajar: I, I, I am, I speak them [Berber and Arabic MS] both. Ik, ik, ik ben, ik spreek ze allebei. Max: Hey, hey, hey, hey, can I say something myself? Hey, hey, hey, hey, mag ik zelf iets zeggen? Lemnja: No. Nee. Hajar: I am one hundred percent Berber but I simply speak Arabic because I cannot speak Berber that well. Ik ben honderd procent Berber maar ik spreek gewoon Arabisch want ik kan niet zo goed Berbers. (GD01: 183193)

In the discussion above, Siham ascribed the identities of Lemnja and Samira, but not that of Hajar, as half Arabic. In response to Sihams identity ascription, Samira counteracts with half Berber. At the opposite end, Hajar who in the beginning also claimed to be pure Berber has to defend her Berber afliation. She explains the cause of her language use and disentangles a limited prociency in Berber, compensated with the use of Arabic, from being less Berber. Being pure Berber is therefore coupled with being a user of the Berber language alone, while the speaking of Arabic and/or a limited knowledge of Berber, as in Hajars case, is seen as not pure. Hence the category they are half Arabic. Contrary to what one could expect for immigrant minority group members, we see that the discourses proposed by these girls are constructed within a modernist ideology of language use and national/ethnic belonging. The macro language politics at play in the Maghreb world are thus being (re)proposed within the verbal micro-interactions of this super-diverse group of pupils giving way to processes of identity misrecognition, sanctioning and contestation. For instance, Samira, who was ascribed by Siham as halfArabic, objects to being constructed

Modernist language ideologies, indexicalities and identities


at the periphery of the Berber group and reiterates her Berber Arabic afliation where the use of Berber rst and Arabic second may be done with the purpose of contesting Sihams initial ascription of her as half Arabic. Hajar, instead, stresses with her upset remark that she has been ascribed as not Berber. In contrast to other research on the language use of Moroccan youngsters (Jaspers 2005: 287), the link between Arabic and Berber was strongly emphasised by the girls in question. The use of Berber and the link between monolingualism and being pure Berber is an act of identity performance that is assessed, weighed and measured socially. In this case, though, the social measurement does not happen on the basis of language hierarchies along a majority versus minority divide. The status of Arabic and Berber and the need to express competence in Berber might be a reaction product to the symbolic hierarchy that exists between these two codes in the Maghreb world (De Ruiter, Saidi and Spotti 2009) and within the local immigrant minority community in Duivenberg, to which these two girls belong. However, as one of the girls points out in the discussion, Berber is no match for Arabic in the existing religious and pan-Arabic political frameworks, and the fact that Arabic is usually spoken in cities as opposed to the rural areas where most of the parents of these pupils originate from can be a reason for the low appreciation that Arabic has and for the identity ascription as half-Arab of those girls who either do not know or have a limited prociency in Berber.

6. Escaping from the modernist lens of authority We now come to the third data set which features a conversation between myself and two boys, Osman and Cemal, both of Turkish immigrant minority background. The conversation focused on these pupilsuse of Turkish across different institutional environments and it unfolded as follows:
Max: And do you, do you do that often, Turkish, at school? En doe je, doe je dat vaak, Turks, op school? Osman: Yes. Ja. Max: Oh. Osman: At break. In de pauze. Max: Yes? [Turning to Cemal] So do you also speak Turkish with everyone because it makes you look very tough? Ja? Dus spreek je ook met iedereen het Turks want het maakt je heel stoer?


Massimiliano Spotti Cemal: It is not tough. Het is niet stoer. Max: No? Nee? Cemal: Not for me. Simply it [Turkish: MS] is a normal language. Voor mij niet. Gewoon is normale taal. Max: It is simply a normal language. Het is gewoon een normale taal. Cemal: If it was English it would have been a bit tough perhaps, it is a famous language. Als het Engels was een beetje stoer misschien, het is een beroemde taal. Max: Oh that is what you mean. Oh zo. Cemal: Or French, and Turkish is not so well-known so (uh) (. . .) yes (. . .) Of Frans, en Turks is niet zo beroemd dus (uh) (. . .) ja (. . .) (GD01:349362)

Community languages are often referred to as a we-code that (in this case) places the pupils who speak these languages in an advantaged position compared to those with a they-code (cf. Gumperz 1968), i.e., in this case, all the pupils who do not have a knowledge of Turkish. However, in Cemals case Turkish is addressed as simply it is a normal language because it is not so famous or well-known as English or French, both international languages. In Cemals discourse, the macro- politics of language are encapsulated in a micro-discursive practice of language ranking and of construction of order (Foucault 2007). In other words, Cemal is presenting a hierarchical ranking of the languages that he thinks do count there, and leaves Turkish in the neutral position as he addresses it as a normal language. This (implicit) denial of the we-code character of Turkish, though, surprises the researcher and is in conict with what happens in the following episode between Osman, Cemal and myself before the beginning of a physical education lesson. The conversation proceeded as follows:
Max: So then do you speak a bit of Moroccan? Zo dan spreek jullie een beetje Marokkaans? Osman: No, no a real Turkish man speaks only just Turkish. Nee, nee een echt Turks man spreekt alleen maar Turks. Cemal: Shouf shouf habibi. [Giggles] Shouf shouf habibi. Osman: Yes, shouf shouf habibi. [Giggles] Ja, shouf shouf habibi. [They both tie their shoelaces and run off to the gym hall laughing: MS]

Modernist language ideologies, indexicalities and identities


The sentence No, no, a real Turkish man speaks only just Turkishmay indicate a patrimonium-loaded connotation that, as reported in the home language survey, corroborated the fact that Osman only spoke Turkish with his father (cf. Pujolar 2001: 137). However, contrary to the language practice just claimed by Osman for a real Turkish man, Cemal utters the Arabic words shouf shouf habibi. These words, apart from being (incorrect) Arabic, have an emblematic function in that they call upon something like Arabic they are in fact the title of an, at the time, popular Dutch TV comedy on the lives of three young men of Moroccan ancestry in a city in the Netherlands who run into all kinds of adventurous situations that are mainly connected to their immigrant condition in a Dutch environment. Further, in agreement with Cemal, Osman utters the same Arabic sentence, after which they run off to the gym hall laughing. Cemal and Osmans linguistic practice therefore can be interpreted as nonauthentic (Jaspers 2005: 116) because it is incongruent with what, following Osman, a real Turkish man would do languagewise, i.e., speak Turkish only. The use of a semiotic artefact, i.e., an emblematic Arabicutterance that belongs to the youngsters knowledge of a specic piece of Dutch popular culture, may instead be interpreted as an example of language sabotage employed to take off the pressure exerted on the boys by the authority, i.e., the gullible researcher who is exploring their linguistic practices. At the same time the episode shows that, although there is no explicit policy regulating them, immigrant minority languages and hybrid forms of semiotic expression(s) that go beyond straightforward ethnic afliations do play a role in the institutionalized everyday life of these pupils. As such, shouf shouf habibi could be qualied as an authentic expression of late modern mixed language repertoires in which languages are no longer artefactualised matters and where languages are not exclusively bound to specic immigrant groups (see Rampton 2006) but can be and are actually used ad libitum.

7. Discussion and conclusions Multilingualism has been the agship of sociolinguistics for decades. This has entailed advancing the claim that, in general and in principle, multilingualism and the preservation of ethno-linguistic identities is a positive thing. Furthermore, sociolinguistics has addressed human beings as users of either one or more languages or language varieties, with language(s) being understood either as distinct denotational codes or linguistic systems and their users studied on the basis of the question who speaks (or writes, or signs) which language (or language varieties), to whom, when, where and possibly to which end (see Fish-


Massimiliano Spotti

man et al. 1968; Fishman 1972, 1989). This approach was valid in that it has given way to powerful notions such as language maintenance, language loyalty, language shift, language (repertoire) change, language death, etc. Furthermore, it has paved the way towards the study of language use and of identity construction as a matter of the internal organization of particular speech networks. Building on these notions, some sociolinguistic research has focused on the use of immigrant minority languages in the homes of immigrant minority pupils, advocating their use to be evidence for the ethnolinguistic vitality of groups (Extra and Ya mur 2004), or on the role that a language or one of its varieties plays g in a given nation. Certain varieties have also been seen as endangering the existence of other less powerful languages (i.e. indigenous minority languages), and ofcial languages as hindering the language rights and the display of ethnolinguistic identities of minorities (Das Gupta 1970; Huss 1999; Svonni 2008). The sociolinguistics of this era has produced an understanding of multilingualism and its identity byproducts as an agglomeration of fairly neat monolingualisms that can be linked to certain groups. In other words, it has addressed monolingualism as a majority thing and multilingualism as a minority thing. Moreover, this branch of sociolinguistics has tackled multilingualism as a problem that it should address for the benet of (immigrant) minorities and their identity maintenance, and with an eye to the maintenance of the national order. This kind of sociolinguistics seems to have failed to address the grassroots realities of multilingualism and identity construction in three ways. First, it has not managed to solve the tension between immigrant minorities own linguistically heterogeneous repertoires and the general tendency in education to see immigrants contributing positively to mainstream society by learning the language of the majority, preferably in its school variety. Second, it has missed noticing that institutional ideologies of homogeneity are also part and parcel of the discourses of immigrant minority group members themselves when dealing with their identity diacritics, i.e., language use and ethnic belongings. Third, it has overlooked the fact that crossing, hybridity and ambivalence in language use and identity construction are an endemic condition of pupils inhabiting (urban) globalised educational environments. What I hope to have illustrated here is that modernist language ideologies are not only top-down phenomena (i.e. appropriated only by the majority to attribute certain identities to minorities). Rather, micro-crystallisations of these ideologies can be found both in discourses of the majority (e.g. certain pupils Dutch is poor because at home their parents do not use Dutch only), as well as minorities (e.g. one is not a pure Berber because one does not have a good knowledge of Berber and, consequently, is ascribed the identity of a half-Arab). As the conversation between Osman, Cemal and myself illustrated, linguistic

Modernist language ideologies, indexicalities and identities


diversity is not to be seen as directly related to group membership alone and that an (ethnic) group and an (ethnic) identity now seem to be a great deal less clear and less relevant than what they have been in the past two and a half decades of sociolinguistic research (see also Rampton 2006: 17). These ndings therefore not only question the assumptions of institutional key gures about the identities of their immigrant minority pupils; more generally, they suggest that not all forms of multilingualism are productive and empowering. Some accordingly to the institutional space are unwanted, disqualied, and actively endangering to people. We should therefore look at language use and identities through the lens of a post-Fishmanian awareness, addressing the two as polycentric semiotic performances confronted with the demands of specic language and identity markets set in specic times and spaces where different, often conicting, orders of indexicalities are at play.

1. The transcription uses (. . . ) for a pause, (-) for an abrupt stop, [:] for emphasis, [xx] for inaudible fragment, and [text MS] for a comment. 2. In 2005, the Dutch educational system used to assign to each pupil an educational weight. A pupil born to Dutch parents and raised in the Netherlands was assigned an educational weight of 1.00. A pupil with at least one of his parents having been born abroad and where the household breadwinner would be a manual labourer, was assigned a 1.90. That is, this pupil is almost twice as heavy a burden on the educational system as a pupil of Dutch origin.

Agha,Asif. 2003. The social life of cultural value. Language & Communication 23. 231 273. Bakhtin, Mikhail. 1981. The dialogic imagination. Austin: University of Texas Press. Bauman, Richard & Charles Briggs. 2003. Voices of modernity. Cambridge: Cambridge University Press. Bezemer, Jeff. 2003. Dealing with multilingualism in education. Amsterdam: Aksant. Bezemer, Jeff & Sjaak Kroon. 2006. You dont need to know the Turkish word. Immigrant minority language teaching in policies and practice in the Netherlands. L1 Educational Studies in Language and Literature 6(1). 1329. Blommaert, Jan. 2005. Discourse. Cambridge: Cambridge University Press. Blommaert, Jan. 2008. Artefactual ideologies and the textual production of African languages. Language & Communication 28(02). 291307.


Massimiliano Spotti

Blommaert, Jan, Lies Creve & Evita Willaert. 2006. On being declared illiterate: Language-ideological disqualication in Dutch classes for immigrants in Belgium. Language & Communication 26(1). 3454. Block, David. 2006. Multilingual identities in a global city. London stories. Basingstoke: Palgrave MacMillan. Broeder, Peter & Guus Extra. 1998. Language, ethnicity and education. Case studies on immigrant minority groups and immigrant minority languages. Clevedon: Multilingual Matters. Das Gupta, Jyoritinda. 1970. Language conict and national development. Berkeley: University of California Press. De Ruiter, Jan Jaap, Saidi Redouan & Massimiliano Spotti. 2009. Teaching minority languages: The case of Arabic in Europe. Report for the DG Education and Culture. PLUSVALOR Project (144368-2008-IT-KA2-KA2MP; Dong, Jie 2009. The making of migrant identities in Beijing: Scale, discourse, and diversity. PhD Thesis defended at Tilburg University. Dong, Jie. 2010. The making of migrant identities in Beijing: Scale, discourse, diversity. Bristol: Multilingual Matters. Erickson, Frederic. 1986. Qualitative methods on research on teaching. In Merlin Wittrock (ed.), Handbook on research on teaching, 119161. Basingstoke: Palgrave MacMillan. Extra, Guus & Kutlay Ya mur. 2004. Urban multilingualism in Europe: Immigrant mig nority languages at home and school. Clevedon: Multilingual Matters. Extra, Guus & Massimiliano Spotti. 2009. Language, migration and citizenship: A case study on testing regimes in the Netherlands. In Gabrielle Hogan-Brun, Clare MarMolinero & Patrick Stevenson (eds.), Discourses on language and integration: Critical perspectives on language testing regimes in Europe, 6182. Amsterdam: John Benjamins. Fishman, Joshua. 1972. The sociology of language: An interdisciplinary social science approach to language in society. Rowley, Mass.: Newbury House Publishers. Fishman, Joshua. 1989. Language and ethnicity in minority sociolinguistic perspective. Clevedon: Multilingual Matters. Fishman, Joshua, Robert Ferguson & Jyoritinda Das Gupta (eds.). 1968. Language problems of developing nations. New York: Wiley. Foucault, Michel. 2007. Security, territory, population: Lectures at the College de France 19771978. Basingstoke: Palgrave MacMillan. Goffman, Erving. 1981. Forms of talk. Philadelphia: University of Pennsylvania Press. Green, Judith & David Bloome. 1997. Ethnography and ethnographers of and in education: A situated perspective. In James Flood, Shirley Brice Heath & Diane Lapp (eds.), Handbook of research on teaching literacy through the communicative and visual arts, 181202. New York: Macmillan. Guba, Egon &Yvonne Lincoln. 1989. Judging the quality of fourth generation evaluation. In Egon Guba and Yvonne Lincoln (eds.), Fourth generation evaluation, 228251. Newbury Park, CA: Sage.

Modernist language ideologies, indexicalities and identities


Gumperz, John. 1968. The speech community. International encyclopedia of the social sciences, 381386. London: MacMillan. Huss, Leena. 1999. Reversing language shift in the far north. Linguistic revitalization in Northern Scandinavia and Finland. Acta Universitatis Upsaliensis. Studia Uralia Upsaliensia 31. Uppsala. Jacquement, Marco 2005.. Transidiomatic practices: Language and power in the age of globalization. Language & Communication 25. 257277. Jaspers, Jrgen. 2005. Linguistic sabotage in a context of monolingualism and standardization. Language & Communication 25. 279297. Joseph, John. 2003. Language and Identity: National, Ethnic, Religious. Basingstoke: Palgrave Macmillan. Kroon, Sjaak & Jan Sturm. 2000. Comparative case study research in education: Methodological issues in an empirical-interpretive perspective. Zeitschrift f r Erziehungswisu senschaft 3(4). 559576. Polanyi, Livia. 1989. The American story: A cultural and structural analysis. Norwood: Ablex Publications. Pujolar, Joan. 2001. Gender, heteroglossia and power. A sociolinguistic study of youth culture. Berlin: Mouton de Gruyter. Rampton, Ben. 2005. Crossing: Language and ethnicity among adolescents. Manchester: St. Jerome. Rampton, Ben. 2006. Language in late modernity: Interaction in an urban school. Cambridge: Cambridge University Press. Saeed, John Ibrahim. 1999. Somali reference grammar. Wheaton, MD: Donwoody Press. Silverstein, Michael. 1996. Monoglot standard in America: Standardization and metaphors of linguistic hegemony. In Donald Brennels & Ronald Macaulay (eds.), The matrix of language, 284306. Boulder: Westview. Silverstein, Michael. 1998. Contemporary transformations of local linguistic communities. Annual Review of Anthropology 27. 401426. Singleton, David. 1994. Introduction: A Critical Look at the Critical Period Hypothesis in Second Language Acquisition Research. In David Singleton & Zsolt Lengyel (eds.), The Age Factor in Second Language Acquisition, 129. Bristol: Multilingual Matters. Spotti, Massimiliano. 2006. Constructing the other: Immigrant minority pupils identity construction in the discourse of a native Dutch teacher at a Dutch Islamic primary school. Journal of Multicultural Discourses 1(2). 121135. Spotti, Massimiliano. 2007. Developing identities. Amsterdam: Aksant. Svonni, Mikael. 2008. Smi in the Nordic Countries and Russia. In Durk Gorter & Guus Extra (eds.), Multilingual Europe: Facts and policies, 233249. Berlin: Mouton de Gruyter. Massimiliano Spotti,, has defended his doctoral thesis on the construction of immigrant minority pupils identities in multicultural primary school classrooms in the Netherlands and Flanders in 2007. At present, he is a researcher at Babylon, Centre for Studies of the Multicultural Society at Tilburg University. He also covers the post


Massimiliano Spotti

of researcher within the FiDiPro (Finland Distinguished Professor) scheme, at the Department of Languages, University of Jyv skyl . His current research activities include a a Dangerous Multilingualism, Language Testing Regimes (Extra, Spotti & Van Avermaet 2009) and Institutional Responses to Superdiversity.

Language policy and planning in Hong Kong: Colonial and post-colonial perspectives

Abstract Hong Kong ceased to be a colony of Britain on June 30th , 1997, thus entering a new stage of its development and evolution as a uniquely-constituted city state and urban metropolis. The Hong Kong Special Administrative Region of China (HKSAR) inherited a linguistic ecology that owed much to its previous existence as a British colony, where the Chinese language had had no de jure status until 1974. From 1995, the stated policy of government has been to promote a biliterate (Chinese and English) and trilingual (Cantonese, Putonghua and English) society, and various measures have also been taken to promote the use of Chinese as a medium of instruction in schools. Immediately after the change in sovereignty, Putonghua became a compulsory school subject for the rst time. This paper will examine the issue of language planning and policies partly from an historical perspective, but also through a consideration of current policies and practices across a range of domains, including government, law and education. One major conclusion that emerges from this discussion is that, from a language policy perspective, the relationship between Chinese and English in the Hong Kong context is potentially far less contentious than that between Cantonese and Putonghua.

1. Introduction Although questions of language policy and planning have received much attention in Hong Kong over the last two decades or so, these issues continue to engage both academic commentators and the wider population. Perhaps one major reason for this has been the speed of economic, political and social change in modern Hong Kong society, which, between the 1960s and 1990s, saw Hong Kong transform from a colonial backwater to a post-colonial global city. This


Kingsley Bolton

chapter attempts to review current issues in language planning from a number of perspectives. First, the chapter starts with a brief survey of language planning from the perspective of sociolinguistics, proceeds to a discussion of the history of language planning and policies in colonial Hong Kong, and then focuses on contemporary language policies in the Hong Kong Special Administrative Region (HKSAR) of China.

2. Language planning theories and their relevance to Hong Kong According to Mesthrie, Swann, Deumert and Leap (2000), the term language planning was originally coined by the Norwegian-American sociolinguist Einar Haugen (1959), and might be broadly dened as all conscious efforts that aim at changing the linguistic behaviour of a speech community (Mesthrie et al. 2000: 384). The same author also explains the term language policies, as referring to the more general linguistic, political and social goals underlying the actual language planning process (Mesthrie et al. 2000: 384). An obvious problem here, however, is the similarity of the two terms, language planning and language policies, which are frequently used in overlapping fashion. More explanation is offered by Spolsky and Lambert (2006) who attempt to disambiguate these two expressions. For them, language policy may be explained thus:
The language policy of a speech community [. . . ] consists of the commonly agreed set of choices of language items whether sounds or words or grammar or language varieties whether codes or dialects or named languages and the beliefs or ideologies associated with those choices. It can be found in language practices and beliefs or in formal policy decisions such as laws, constitutions, or regulations. (Spolsky & Lambert 2006: 561)

In contrast, language planning (or language management) may be dened as follows:

Language management, planning, engineering, cultivation and treatment are actions taken by formal authorities such as governments or other agencies or people who believe that they have authority, such as parents, teachers, or academies, to modify the language choices made by those they claim to have under their control [. . . ] Language management itself has three components: the development of explicit language plans and policies, their implementation (by rules or laws or resource allocation), and the evaluation of results and effects. (Spolsky & Lambert 2006: 561)

There seems to be agreement that the term language policy seems to have a broader application in referring to the more general beliefs, considerations

Language policy and planning in Hong Kong


and ideologies relating to the orientation of such decision-making bodies as governments and educational authorities. The term language planning is thus more specically applied to the actions of such decision makers. Spolsky and Lambert further note that most analyses of language policies and planning have been concerned with examining formal, governmentally backed activities at the national or regional level aimed at controlling language knowledge and use within a country or region (Spolsky & Lambert 2006: 562563). Following Kloss (1969), they also highlight the distinction between status planning versus corpus planning, while also accepting Coopers (1989) addition of a third-level of acquisition planning. For Kloss, status planning referred to the choice of ofcial and national languages, etc., while corpus planning concerned itself with the establishment and regulation of particular aspects of language, such as dictionaries, grammar and writing systems. Coopers (1989) notion of acquisition planning was crucially concerned with the determination of which languages should be taught to those who do not speak them and how (Spolsky & Lambert 2006: 563). Perhaps what is most typical of the kind of approach adopted by Spolsky and Lambert, however, is the underlying assumption that functional models can adequately account for, and usefully illuminate, the kinds of choices made by governments in language planning, in terms of principles, procedures and processes. Language policy/planning (LPP) studies, as an area of sociolinguistic research and practice, began to achieve prominence in the 1960s, and were often associated with language surveys in developing nations, particularly those in the early stages of post-colonial independence (Fishman, Ferguson & Gupta 1968). Other sociolinguists, including Cooper (1989), Neustupn (1970), and Kaplan y and Baldauf (1997), among others, followed Fishman in attempting to build rational models of language planning suitable for implementation in the developing world and other multilingual settings. In broad terms, these were often written as almost apolitical structural-functional explanations of the dynamics of decision-making, in those societies under discussion. From the 1990s onwards, however, new approaches to language planning were clearly visible, characterised by a strong interest in issues related to language and inequality and inuenced by critical and postcolonial theory. Thus, for such scholars:
Linguistic theories adopted by language planners, rather than being neutral, objective, scientic tools, were viewed [. . . ] as detrimental to the development of equitable language policies in complex multilingual settings. This realization led to a rather broad calling into question of received ideas about the nature of language itself, and of the degree to which scholars of language were perpetuating assumptions that had the effect of rationalizing the support of colonial languages,


Kingsley Bolton and concomitant economic interests, at the expense of indigenous languages and local economic development. (Ricento 2006)

In this context, scholars such as Tollefson (1991), Phillipson (1992), and Pennycook (1998) have done much to promote critical perspectives on language, greatly extending earlier sociolinguistic approaches by promoting an explicit awareness of issues relating to inequality, power and the politics of language.

3. Language planning in Hong Kong: the historical context For much of its recorded history, language planning in Hong Kong was a byproduct of the British colonial system, which governed Hong Kong between 1842 until 1997. During most of that period, English was not directly challenged as the language of government and law in Hong Kong society until the era of late colonialism from around 1970 until 1997, when issues of language planning and policies were brought into sharp focus through a number of debates on language issues, as well as through government legislation and interventions. Such debates and interventions have continued to the present, although the reunication of Hong Kong with China has added a number of new complexities to such issues, including, not least, the relationship between Hong Kong language planning and policies and those of mainland China. During the First Opium War, in January 1841, the island of Hong Kong was ceded by the Chinese government to the British. The following year, this was ratied by the Treaty of Nanking, and the British trading and missionary community that had previously taken residence in the Portuguese enclave of Macao soon transferred to the island. From 1842 to 1845 the population of the island grew remarkably from around 5,000 to more than 20,000, reaching 40,000 by 1853, and topping 120,000 by the early 1860s. The vast majority of such immigration into Hong Kong came from neighbouring Guangdong province, and were typically classied as belonging to four distinct groups: the Punti, locals, i.e. Cantonese; the Hakka; the Tanka, boat dwellers; and the Hoklo, from eastern Guangdong province (Munn 2001: 71). In 1860, the British annexed the Kowloon Peninsula and, in 1898, added an additional swathe of its hinterland known as the New Territories, so that, by the early twentieth century, British Hong Kong had come to include all three territories, which together constituted the territorial entity of colonial Hong Kong throughout most of the twentieth century, until the celebrated 1997 Handover, which nally returned sovereignty over Hong Kong to the Peoples Republic of China.

Language policy and planning in Hong Kong


British colonial language policy has been explained by a range of historical commentaries of varying accuracy and perspectives. In the post-colonial imaginary, British colonial language policy worldwide has been linked to a blatant linguistic imperialism which sought to impose the language of the imperial power on colonised communities in Africa, Asia and elsewhere (Phillipson 1992, 1994). Other, more nuanced, accounts have explained British colonial language policy in terms of strategies of divide and rule, pointing to the British predilection in many colonial territories for reserving English-medium education for an elite, and propagating vernacular education for the masses (Pennycook 1998). In its bluntest form, the Phillipson/Pennycook perspective on colonial language policy has seen the spread of English as a tool of colonial control and subjugation, either explicitly and overtly through the imposition of English, or, more subtly, through the establishment of systems of parallel languages in such societies. A related argument is that in many former colonies the legacy of such policies has extended to the present, through the deep and indissoluble links between the practices, theories and contexts of ELT [English Language Teaching] and the history of colonialism (Pennycook 1998: 19). The critical perspectives of both Phillipson and Pennycook have inuenced discussions of the history of language policies put forward by a number of local scholars in recent years, including Tsui (2004) and Hopkins (2006). A rather different picture has emerged from recent research by others, who have attempted to provide a much more detailed picture of the development of the colonial educational system, including the work of Sweeting and Vickers (2005, 2007) and Evans (2008a, 2008b), following Brutt-Grifers (2002) reconceptualisation of British colonial language policy as an historically contested terrain, where local elites often campaigned or negotiated for access to English-medium education. For example, what emerges from Evans careful (2008a) discussion of language policy in Hong Kong between 1855 and 1900 is the picture of a complex patchwork of government Chinese, government Anglo-Chinese, aided Chinese, aided English, and aided Anglo-Chinese schools. Within this system, moreover, by far the largest enrolments were in the aided Chinese schools. For their part, Sweeting and Vickers (2005) also emphasise the complexity of the nineteenth century school system in the colony, noting that for a number of decades there was no top-down imposition of a clear, consistent language policy. Later, after the governments Central School (which later became Queens College) was founded, both Chinese and English were used as instructional languages for a number of years, which also mirrored the situation in many missionary schools in the territory. Sweeting and Vickers also report that vernacular education expanded substantially in the rst decades of the twentieth century, and that educationalists and government spokesmen


Kingsley Bolton

repeatedly acknowledged the importance of both languages in the education system. The work of Evans and Sweeting in retrieving the historical record is crucial in providing some kind of balance, not only to the discussion of the history of language issues in Hong Kong, but also as a key to the present. So far, only part of this project has been completed. For example, whereas their work has now done much to illuminate the complexity of government policies on language education in the nineteenth century, very little detailed work (to my knowledge) has been carried out on the inuence of missionary and religious schools in Hong Kong, which have deployed both Chinese and English in various types of schools in the territory.1 Indeed, a detailed historical narrative of language education in Hong Kong remains to be written, although recent work by local historians such as John M. Carroll has emphasised the collaborative nature of colonialism, and the role played by local Chinese elites in policy-formation in many key areas of society (Carroll 2007). Despite the lack of a clear historical narrative of high colonial history in Hong Kong, a greater clarity arguably exists for the period from the 1960s until the 1990s, an era that could broadly be described as late British colonialism. During the 1970s, after the disturbances of the Cultural Revolution in China and social unrest and riots in Hong Kong in 1966 and 1967, the government began to give greater recognition to the Chinese language. In 1974, Chinese was recognised as a co-ofcial language in the territory, while, around the same time, the colonial authorities also established a system of free, compulsory primary and secondary education, as well as extensive systems of public housing and public health. Until the 1970s, the English language had been the sole ofcial language of government, the ofcial language of law and, de facto, the more prestigious medium of secondary and university education. The Ofcial Languages Ordinance of 1974 established that Chinese and English would thenceforth enjoy equality of use and, subsequently, measures were taken by the government to promote this policy. A decade or so later, after the negotiations between Beijing and London determined the arrangements for the 1997 Handover, the position of Chinese was further strengthened by the publication of The Basic Law of the Hong Kong Special Administrative Region, Article 9 of which stated that: In addition to the Chinese language, English may also be used as an ofcial language by the executive authorities, legislative and judicial organs of the Hong Kong Administrative Region (Chinese Government 1992: 7). In 1995, the Hong Kong Government announced that its new language policy would be to develop a civil service which is biliterate in English and Chinese and trilingualin English, Cantonese and Putonghua (Lau 1995: 19), an ofcial

Language policy and planning in Hong Kong


policy statement which is still in force. In the run-up to 1997, an increasing proportion of Cantonese was used in Legislative Council speeches, and since then the Legislative Council has almost exclusively used Cantonese to conduct its affairs. Since the early 1990s the government has been trying to establish training courses in Putonghua for Hong Kong civil servants but, at present, Cantonese, rather than Putonghua, is still the dominant variety. Similar changes have taken place in the legal system in Hong Kong and, from the late 1980s to the mid-1990s, amendments to the Ofcial Languages Ordinance have promoted legal bilingualism in the law courts. In December 1995, the rst civil High Court case was heard in Putonghua and, in August 1997, the rst criminal case was conducted in Cantonese in the High Court (Cheung 1997, 2000). Immediately before the 1997 change of sovereignty, there was widespread concern that the post-colonial period would see the immediate downgrading of English and the rise of Putonghua in key ofcial domains. As is discussed in the next section, an abrupt change in language policy and management has not occurred in Hong Kong, at least in most domains of society. However, one major change in policy that did occur immediately before July 1997 concerned education, when, shortly before the transition, the colonial Hong Kong government formulated a new policy on the medium of instruction for secondary schools. On March 22nd , 1997, it was announced that approximately only 100 secondary schools (some 22 per cent of the total of 460) would be allowed to use English as a teaching medium and that punitive measures (e.g., a maximum ne of $25,000 and two years in jail), might be used against school principals who did not follow the instructions of the government (Kwok 1997). This policy has been largely maintained since 1997, although, very recently, it has been amended to provide more opportunities for the use of English, not least in order to prepare secondary school students for what is a predominantly English-medium university system. Notwithstanding such recent changes, the adoption of a new rm policy in promoting Chinese was the most visible change in language policy at the end of the colonial period, although it might also be argued that the adoption of Putonghua as a compulsory school subject which occurred around the same time is likely to have even more lasting consequences.

4. Current language planning and policies in Hong Kong Problems of investigation and interpretation of language policies in Hong Kong are not only conned to the historical past but also extend to the present. In certain settings, including the European Union in recent years, language policies are explicitly articulated, set down and disseminated through public documents,


Kingsley Bolton

reports and regulations. In Hong Kong, however, this has rarely been the case and, in the contemporary HKSAR, there are few (if any) documents that set out an ofcial language policy for all the major domains of society in a cohesive and principled fashion. Instead, there are a number of diverse laws and policy statements that have been issued by government (and continue to be issued) in the colonial and post-colonial period that have combined to shape language planning practices since 1997. Such practices have evidently been moved by circumstances or by public opinion, at times leading the government to respond, in an immediate pragmatic fashion, to the political pressures of the day. 4.1. Language planning in the Hong Kong government In Hong Kong, ofcial language policies now regulate which languages are used in government ofces in the HKSAR, although these policies changed signicantly during the decades before 1997, in the immediate period before the transition from British to Chinese rule. The denition of what the ofcial languages are, and decisions about ofcial language policies in Hong Kong are determined in part by the Basic Law governing Hong Kong, which provided a mini-constitution for the territory before the change of government. The Basic Law of the Hong Kong Special Administrative Region of the Peoples Republic of China has two articles that deal specically with language policies and planning, Articles 9 and 136:
Article 9 In addition to the Chinese language, English may also be used as an ofcial language by the executive authorities, legislature and judiciary of the Hong Kong Special Administrative Region. (Chinese Government 1992) Article 136 On the basis of the previous educational system, the Government of the Hong Kong Special Administrative Region shall, on its own, formulate policies on the development and improvement of education, including policies regarding the educational system and its administration, the language of instruction, the allocation of funds, the examination system, the system of academic awards and the recognition of educational qualications. (Chinese Government 1992)

The Decision of the Standing Committee of the National Peoples Congress on the English Text of the Basic Law of the Hong Kong Special Administrative Region of the Peoples Republic of China is also relevant (Adopted on 28 June 1990):
The 14th sitting of the Standing Committee of the Seventh National Peoples Congress hereby decides that the English translation of the Basic Law of the Hong

Language policy and planning in Hong Kong Kong Special Administrative Region of the Peoples Republic of China which has been nalized upon examination under the auspices of the Law Committee of the National Peoples Congress shall be the ofcial English text and shall be used in parallel with the Chinese text. In case of discrepancy between the two texts in the implication of any word used, the Chinese text shall prevail (cited in Ghai 1999: 570).


In addition to the legal provisions made for the use of Chinese and English in the Basic Law, the government has also, in a number of statements, summarized its policy as that of trilingualism and biliteracy, which refers to the promotion of trilingualism in Cantonese, English and Putonghua, and biliteracy in written Chinese and English. Interestingly, this policy began in the 1990s, with particular reference to the use of languages in the civil service in the 1990s, but, by 2002, this had been extended to the general population, so that an ofcial report of the Education Committee in 2002 states that [t]he language policy of the HKSAR Government is to enable students and the working population to be biliterate (in Chinese and English) and trilingual (in Cantonese, Putonghua and English) (HKSAR Government 2002). Indeed, throughout the last thirteen years, the formula of trilingualism and biliteracy has received support in numerous government and educational pronouncements, although, to my knowledge, no fully detailed report or fully comprehensive rationale of this policy has been ofcially published. The promotion of the policy of trilingualism and biliteracy involves the government in accepting the use of all three languages at the spoken level, as well as a great deal of translation of documents and ofcial papers. This work is carried out by the Ofcial Languages Division of the Civil Service Bureau of the HKSAR government, whose website states its responsibilities as providing translation, interpretation and editing services to Government bureaux and departments as well as:
developing the institutional arrangements for the use of ofcial languages in the Civil Service, including setting guidelines, reviewing Civil Service language practices, and providing language advisory services to bureaux and departments; promoting the effective use of the ofcial languages, in particular Chinese and Putonghua, in the Civil Service [. . . ]; monitoring the use of the ofcial languages and the implementation of the language policy in bureaux and departments (HKSAR Government 2010).

Within the Government, one of the most important institutions is the Legislative Council (Legco), the body that is responsible for discussing and approving the laws of Hong Kong. The Legislative Council occupies a place at the heart of the political process, although, in theory, it is independent of government and


Kingsley Bolton

the civil service. In compliance with the ofcial language policy of the Hong Kong government, all proceedings in Legco are translated into a variety of languages.At present, the vast majority of debates and discussions in committees take place in Cantonese but, at the spoken level, are routinely translated from Cantonese into English by simultaneous interpreters so that members of Legco and members of the public listening to the debates have access to translation. Today, only very occasionally is English or Putonghua heard in the Legco chamber, and on almost all occasions Cantonese is the sole language of spoken communication (Government translator, personal communication, May 18 2010). The vast majority of linguistic work that takes place in Legco involves not spoken communication, but written communication, as all debates and subsequent reports of debates (the Hansard or proceedings) are recorded in two versions, a Chinese version and an English version. The body that has the responsibility of carrying out these translations is the Translation and Interpretation Division of the Legislative Council Secretariat, which is responsible for the production of the Hong Kong Hansard, and is also responsible for translating questions, motions, papers, minutes and other Council papers (Legislative Council 2010). According to an interview with one of the Chief Translators of the Secretariat, around 80% of the documents written for the Legislative Council are currently drafted in English, and then translated into Chinese, so that English very much remains the default written language of government (Government translator, personal communication, May 18 2010). 4.2. The languages of the legal system The governments language policy also regulates the languages used in written statutes and in the law courts. Prior to the 1980s, English was the sole and dominant language of the legal system in Hong Kong. Now the government is ofcially committed to a bilingual legal system, and in 1998, a Committee on Bilingual Legal System was set up to advise the Government on bilingualism in the legal domain and how the goal of a bilingual legal system could be achieved. On the Department of Justice website, it is stated that, in light of Article 9 of the Basic Law, both Chinese and English therefore have a part to play in the language of the law (HKSAR Government 2010b). Specic policies are stipulated for the Common Law, Statute Law and the law courts as follows:
The common law The principles of the common law are to be found in the judgments of the courts, both in Hong Kong and in other common law jurisdictions around the world.

Language policy and planning in Hong Kong The language in which those judgments have been delivered over the years is almost exclusively English. There are hundreds of thousands of reported cases which form the basis of the common law, and it would obviously be impractical to attempt to translate these into Chinese. While in future there is likely to be an increasing number of judgments in Hong Kong delivered in Chinese, English will continue to be the only medium in which the majority of judgments from overseas is reported. Statute law In keeping with the Basic Laws provisions on bilingualism, all legislation in Hong Kong is enacted in both Chinese and English, and both versions are accorded equal status. Thanks to the bilingual legislation programme begun in 1989, authentic Chinese texts have been completed of all pre-existing legislation which had been enacted in the English language only, and Hong Kongs statute book is now entirely bilingual [. . . ]. The courts In July 1995, the Ofcial Languages Ordinance (Cap 5) was amended to enable any court to use either or both of the ofcial languages in any proceedings before it as it thinks t; to enable a party or his legal representatives or a witness in proceedings in a court to use either or both the ofcial languages, or such other language as the court may permit; to provide that the decision of a court to use one of the ofcial languages in any proceedings before it, is nal; and to empower the Chief Justice to make rules and issue practice directions to regulate the use of Chinese language in the courts. Efforts are being made on various fronts to improve the use of Chinese in the higher courts. A Practice Direction on the use of Chinese in the Court of First Instance has been prepared by the Judiciary. Training for bilingual judges has also been introduced, including the provision of courses on Chinese judgment writing skills. No matter whether English or Chinese is used in the proceedings, everyone has a right to use the language of his choice to give evidence. The court will arrange interpretation facilities. (HKSAR Government 2010b)


From the above, we can see that three main concerns are articulated. The rst concerns the role of English in the common law system, which is the norm in the UK, and other English common law systems, such asAustralia, New Zealand, etc. The second concern is with the written laws (statute law) of the HKSAR, where it is stated that all legislation in Hong Kong is enacted in both Chinese and English, and both versions are accorded equal status. The third concern relates to the language of the courts and the expressed desire to improve the use of Chinese in the higher courts. However, despite the aim of providing legal proceedings in either or both of the ofcial languages, at present, Cantonese is the dominant language of the lower courts, while English still remains the major language of the higher courts.


Kingsley Bolton

A recent study by Ng (2009) has examined the use of Cantonese and English in Hong Kong courts in great detail. Ng notes that it was only in 1987 that the Ofcial Languages Ordinance was amended to state that all ordinances shall be enacted and published in both ofcial languages (cited in Ng 2009: 71), and that it was not until 1989 that the rst Bilingual Ordinance was enacted. Since the early 1990s, a massive project to translate the statutes of Hong Kong has taken place, a remarkable process involving the translation of more than 19,000 pages of legislation (Ng 2009: 72). Despite this, even today, the vast majority of court documents and judgements are written in English, as are most legal reference books and case law records. At the spoken level, in theory, Chinese enjoys equal status with English, but in practice, there is a clear hierarchy in Hong Kong courts. This hierarchy goes from the lowest courts (the Magistracies), through the District Courts (Criminal and Civil), to the Courts of First Instance (Criminal and Civil), to the Courts of Appeal (Criminal and Civil), and nally to the Courts of Final Appeal (Criminal and Civil). Ngs analysis indicates a clear asymmetry in legal bilingualism in the Hong Kong law courts:
Cantonese is used with decreasing frequency as one moves up the court hierarchy, and it is totally absent in the Court of Final Appeal. In a parallel manner, the presence of English exhibits an inverted pyramidal distribution. English is used less frequently in the lower courts but retains its strong presence in the higher courts. Horizontally, there is a growing practice of mixed-language trials in the Court of First Instance and the District Court. Cantonese is used when witnesses are examined, but English remains the language of choice when law is debated (Ng 2009: 253).

4.3. The languages of education As has been the case in many other multilingual societies, issues related to the choice of languages to be used in schools have been controversial and sensitive, as it is here that ordinary citizens are most likely to perceive their lives directly affected by language policy. In Hong Kong, the history of the medium of instruction issue in the modern era dates back most immediately to the early 1970s, when the British colonial administration attempted to introduce a policy of using Chinese as the medium of instruction. In 1973, the government published a Green Paper, or policy proposal, on language education, which asserted that:
The medium of instruction bears signicantly upon the quality of education offered at post-primary level. Pupils coming from primary schools where they have been taught in the medium of Cantonese have a grievous burden put on them when

Language policy and planning in Hong Kong required to absorb new subjects through the medium of English. We recommend that Chinese become the usual language of instruction in the lower forms of secondary schools, and that English should be studied as the second language (cited in Gibbons 1982: 117).


After the publication of the 1973 Green Paper, the government met strong opposition from parents and schools about such plans to introduce Chinese Medium Instruction (CMI). Following these protests, the government backed down from pressing ahead with Chinese-medium instruction, and issued a 1974 White Paper which decided on a laissez-faire approach to the medium of instruction issue. The school system that then evolved from the 1970s until the early 1990s was one with around 90 per cent of primary schools teaching through Cantonese, and a similar percentage of secondary schools claiming to be English medium (or Anglo-Chinese). In reality this meant that school textbooks in most schools were overwhelmingly written in English, while at the spoken level the amount of English used varied greatly according to the type of school and the abilities of staff and students. This so-called laissez-faire approach generally continued until March 1997, when the government introduced a policy of rm guidance to schools. As reported above, the new policy then established a system where around 75% of secondary schools were required to teach through Chinese (CMI schools), and some 25% were permitted to teach through English (EMI schools), if they could demonstrate the feasibility of so doing. This policy was consistently promoted by the Education Bureau of the HKSAR government for around ten years after 1997, and was even re-afrmed by a government report of December 2005, despite frequent challenges by parents and schools who felt disadvantaged. In 2008 and 2009, however, a signicant shift in government policy began, which resulted in a new report on Fine-tuning the Medium of Instruction for Secondary Schools (HKSAR Government 2009). Essentially, the report moves away from the rm guidance policy of 1997, to provide for greater exibility about how language management will take place in individual schools. According to the recommendations of the report, schools will no longer be classied into CMI [Chinese medium instruction] schools and EMI [English medium instruction] schools and [t]heir [. . . ] teaching modes will become more diversied, including all CMI, CMI/EMI in different subjects and total EMI immersion. This, it is claimed, allows schools more exibility in using EMI for one or more subjects for different classes. Thus, it is expected that the choice and number of subjects taught in EMI would likely vary between classes within individual schools as well as among schools, and that MOI arrangements in schools will become more diversied (HKSAR Government 2009: 5).


Kingsley Bolton

This ne tuning policy goes into effect in autumn 2010, and the media have already reported that a number of schools are planning to switch from Chinese- to English-medium instruction (Yau 2009a). This loosening of government policy seems to have been motivated by a number of factors, including the desire of the business community to maintain Hong Kongs competitiveness as a centre for international commerce and nance, the pragmatic need to prepare students for a university education (in a society where most of Hong Kongs universities are ofcially English-medium),2 as well as a groundswell of public opinion from many local parents. Despite this, given the complicated dynamics of education in the HKSAR, whatever policy is implemented is likely to encounter resistance from some quarter, and this change of direction has already attracted criticism from a number of leading educators (Yau 2009b). 4.4. Cantonese and Putonghua One key aspect of language planning in Hong Kong has a direct bearing on the educational debates discussed above and is, indeed, at the heart of many sociolinguistic issues in Hong Kong society. That is the relationship between Cantonese and Putonghua. The status and functions of the Cantonese language are unique to Hong Kong, which has been described as the greatest Cantonese city that the world has ever seen (Harrison & So 1997: 12). Many local linguists have been less concerned about the tension between English and Chinese than a potential conict between Cantonese and Putonghua. The widespread use of Cantonese in Hong Kong society, in high domains as well as low, is obviously at odds with the ofcial policy in China, which promotes Putonghua, together with simplied Chinese characters instead of the full characters used in Hong Kong and Taiwan. Cantonese may be a mere regional dialect in the PRC, but Hong Kong is the Cantonese-speaking capital of the world. In many senses, Cantonese is the essential community language. Given the wide use of the language in education, religion, the print and broadcast media and government, the status of Cantonese is much higher than is normally thought and cannot be simply brushed aside as the vernacular (Sin & Roebuck 1996: 252). In stark contrast, the PRCs ofcial language policy since 1956 has included the unication of the Chinese language, the promotion of Putonghua, the removal of illiteracy, the propagation of simplied characters, and the promotion of the ofcial romanization system of pinyin (Bolton & Lam, 2006: 350). In October 2000, the national government published a new law, entitled The Law of the National Commonly Used Language and Script of the Peoples Republic of China, which stipulated that: Schools and other educational organizations

Language policy and planning in Hong Kong


will take Putonghua and standard Chinese characters as the basic language and characters to be used in teaching and study (Rohsenow 2004: 41; Zhang &Yang 2004: 154). Such laws have taken effect in most other Chinese cities, including Guangzhou, where an estimated 56 million people out of a population of 12 million have in-migrated from other parts of China, thus adding to a process of language shift away from Cantonese in the community (Lai 2009). Ironically, many of the Hong Kong governments pronouncements in support of Chinese-medium instruction in the period immediately after 1997 emphasized the benets of mother tongue education along the same lines as the renowned 1953 UNESCO report that valorized mother tongue education in schools. However, one issue regularly occluded in language debates is that mother tongue education (in its vernacular European sense), simply does not exist in many regions of China. For many of Chinas students, Putonghua is a second language, and, in many regions, children learn a home dialect before going to school, which is often very different from Putonghua, for example, Cantonese, Fukienese, or Shanghainese. But national language policy is quite clear, and has only a limited acceptance of regional dialects and minority languages in educational and other ofcial domains. In the context of the Hong Kong Special Administrative Region (HKSAR), whose systems of government and administration are expected to converge with those of mainland China in another 40 years, the question may be not whether Putonghua will be introduced as an ofcial teaching medium, but when. Debates on the desirability of increasing the use of Putonghua have already surfaced among local educationalists, and such discussions are regularly reported in the press. Even more important is the fact that, in 1998, Putonghua became a compulsory subject in all Hong Kong schools, and today more and more children than ever before are now learning Putonghua as a second (or third) language. Whether that means that eventually Putonghua will displace Cantonese in the more formal domains of language use is a question of a good deal of speculation and, in this context, the domain of education is particularly sensitive. In 2003, the government-backed Standing Committee on Language Education and Research (SCOLAR) published the Action Plan to Raise Language Standards in Hong Kong, where it stated that its policy on the teaching of Chinese in schools included endorsing the Curriculum Development Councils long-term goal of teaching Chinese Language in Putonghua, and encouraging schools to try using Putonghua to teach Chinese Language (SCOLAR 2003). However, the action plan stopped short of urging a rm policy to promote the language, conceding that further studies on the conditions required to ensure a successful switch and prevent negative outcomes (SCOLAR 2003). At present, attempts to extend Putonghua to teach Chinese within the public school system have


Kingsley Bolton

achieved only limited success, but there is always the possibility that in an altered political climate the government may attempt to implement this policy more strenuously. At present, however, in spite of some initiatives to promote Putonghua in schools and discussions concerning the use of the national language in other domains, it is evident that the majority of Hong Kong people have a strong attachment to Cantonese, and Putonghua still has only a limited range of functions in present-day Hong Kong (Zhang & Yang 2004). The extent to which the status and functions of Cantonese in Hong Kong may become highly contentious in the coming years became clear in the summer of 2010 when language debates surfaced in both Hong Kong and across the border in Guangzhou. In early July, reports began to surface in the South China Morning Post of a push by the authorities in Guangzhou to switch the language of a number of prime-time TV programmes in Guangzhou from Cantonese to Putonghua before the Asian Games were hosted there in November 2010. It was at that time that the Guangzhous Peoples Political Consultative Conference submitted a proposal to the local government on Monday to order the citys most popular television station, Guangzhou Television (GZTV), to replace the use of Cantonese by Putonghua on its main channels (He 2010). The same report also noted that:
With 110 million people, Guangdong has rapidly become the most populous province. But most of the recent increase has been migrant job-seekers, and now half its residents do not speak Cantonese. [. . . ] Guangzhou, the provincial capital, once spearheaded the mainlands economic reform. But rivals such as Shanghai and Beijing have caught up and even surpassed it. The dialect seems strange to outsiders. [. . . ] There is a two-pronged attack on Cantonese internal migration on the one hand, and the government policy of a common language for a unied country and harmonious society on the other. (He 2010)

On Sunday, July 25, 2010 matters came to a dramatic head in Guangzhou when several hundred Guangzhou residents took part in a (reportedly amiable and lowkey) street demonstration at the exit of Guangzhou metros Jiangnanxi station against the proposed language switch. According to reports, most of those who participated were young people under the age of thirty (Zhai andYu 2010). Such a street demonstration, however amiable and good-natured, was dramatic in the sense that any unauthorized gathering of people in China is typically viewed as a cause for alarm, and the fact that few if any similar demonstrations in support of regional dialects in China have ever taken place previously. Immediately following the Sunday demonstration, a representative of the Guangzhou city government responded by asserting that it had no plan to marginalise the use of Cantonese or replace it with Putonghua. The spokesman, a Mr Ouyang,

Language policy and planning in Hong Kong


then went on to claim that the campaign in support of Cantonese had been orchestrated by people with ulterior motives (Zhai, 2010). One week later, a second demonstration took place in support of Cantonese at Peoples Park in Guangzhou and a nearby shopping area on August 1, involving hundreds of people. This time the authorities deemed the gathering illegal, and arrested at least 20 people for legal questioning and simultaneously took measures against Internet blogs and discussion groups (Lau, 2010). On the same day, a small demonstration involving some two hundred people took place in Hong Kong in order to express solidarity with pro-Cantonese activists across the border. Following this, however, the situation was defused and at least temporarily resolved when the Head of the Communist Party in Guangzhou announced an apparent climb-down from the proposed restrictions on Cantonese television on August 4. Referring to a report from the ofcially-sanctioned China News Service, the South China Morning Post announced that: Communist Party Chief Wang Yang told a meeting on Wednesday to mark the 100-day countdown to the Asian Games that there was no question of Cantonese being banned, suggesting that there had been a misunderstanding which had been exploited by people who have ulterior motives (Yu 2010). Following this apparent climbdown, it appears that the Guangzhou authorities backed away from their plans to restrict broadcasting in Cantonese and, in the event, the Asian Games took place without further linguistic controversy throughout November 2010. However, this remains an issue that is likely to re-surface, not only in Guangzhou but also in Hong Kong. Indeed, for many Hong Kong people, ofcial attitudes to linguistic diversity reect attitudes to a much wider range of social and political issues. Commenting on these events, the Hong Kong journalist, Stephen Vines, relates the recent controversy about Cantonese to the authoritarianism of the Beijing government and issues of Chinese identity, arguing that China has demonstrated [the governments] determination to curb or even destroy the linguistic diversity that exists in the nation, and that [i]n Guangdong, home of Cantonese, the language battle is accelerating (Vines 2010). He goes on to argue that:
Cantonese enhances a sense of identity. It is this that scares the rulers in Beijing; ofcials across the border are already accusing the defenders of Cantonese of having ulterior motives. [. . . ] Authoritarian governments have great difculty with diversity; they see it as undermining their authority and sowing the seeds of discontent. Even quite innocent manifestations of local pride and regional identication are frowned upon unless ofcially instigated and approved. [. . . ] Ofcials, like the born-again patriots who run Hong Kong, strive to demonstrate their prociency in the national language, wearing it as a badge of loyalty. And there has been a constant battle against the development of local languages in


Kingsley Bolton literature and the mass media. [. . . ] Anyone challenging this process is quickly labelled a splittist in the wonderful language of Maoism. This is a serious charge and is thrown about indiscriminately at both those who genuinely desire to split from Beijing, such as Tibetans, and at others who are happy to be in the Chinese state but seek a stronger sense of local identity [. . . ] there is a sullen suspicion of anyone showing the smallest sign of asserting anything other than ofcially approved forms of identication with the state. This makes the nation innitely poorer, not more united. (Vines 2010)

5. Conclusion In many other contexts for language policy, the crucial questions have concerned the choice and cultivation of national and ofcial languages in the post-colonial context. As is evident from the preceding discussion, the situation in Hong Kong is rather different. Modern Hong Kong was essentially founded by refugees eeing from the control of a Communist regime and the horrors of the Cultural Revolution, and the identication of Hong Kong people with the national language of the PRC is tempered by the experience of the last six decades. For orthodox language planning, Hong Kong may appear to be an exceptional case, as Tsui notes:
With Chinas resumption of sovereignty over Hong Kong, the Chinese language, being the national language, should enjoy a much higher status. The people of Hong Kong should have a stronger awareness of Chinese identity. The adoption of Chinese as the medium of instruction in the majority of the schools should enhance the status of the Chinese language, as well as strengthen the national identity of Hong Kong people and their patriotic sentiments. Consequently, the community should be less resistant to this policy [. . . ] this has not proved to be the case. Whereas most former colonies have been eager to establish their national identity upon decolonization [. . . ] this does not seem to have happened in Hong Kong. (Tsui 2004: 108109)

The essential reason for this, as Tsui concedes, is that language policy in the HKSAR has been decided by the political agenda set for Hong Kong in the transition to its reunication with mainland China. In this context, neither traditional language policy theorizations nor the standard critical responses appear to offer easy solutions. Some thirteen years into the post-colonial experience, little on the surface of Hong Kong society seems to have changed. English still enjoys high prestige as a co-ofcial language of government and law, and as the dominant language of higher education and the business community. Cantonese enjoys an unequalled

Language policy and planning in Hong Kong


status in many domains of high and not-so-high use, including the Legislative Council, the mass media, popular culture, and much else. Putonghua has yet to be heavy-handedly imposed as the language of national and ofcial power on Chinas most dynamic and prosperous southern city, which is still enjoying the benets of the one country, two systems policy, devised by Deng Xiaoping. In this context, however, for many Hong Kong people the notion of mother tongue education may evoke fear of the imposition of the big brother tongue, as there is widespread suspicion that Cantonese-medium education may segue into Putonghua-medium education at some point in the not-too-distant future. Post-colonialism in the Hong Kong context has its own specic characteristics, and as Carroll has commented, [a]lthough Hong Kong has returned to China, it has not been de-colonized. Instead, he argues, it has been re-colonized with the metropole simply shifting from London to Beijing (2007: 192). That may be true, but as yet the full weight of metropolitan and national policies, including language policy, has not been felt in Hong Kong. Viewed from this perspective, the vitality of Cantonese as a community language (layered with a measure of English) is a touchstone for continued lifestyle of a city-state whose identity combines a unique blend of colonial modernity, global capitalism and diverse contacts with Asia, Europe, North America and the world. The Draft Agreement on the future of Hong Kong signed by the Chinese government in the 1980s promised the territory a high degree of autonomy, according to a policy of one country, two systems. It remains to be seen how long this autonomy will survive in many spheres of society, including key societal settings, such as government, law, education, and media, which are so crucially linked to the cultural and linguistic identity of the Cantonese-speaking Hong Kong people.

1. The role of religious schools in Hong Kong continues to be important to the present day, and a number of Catholic, Protestant, and other religious organizations still run signicant numbers of primary and secondary schools in the society. This is a state of affairs greatly in contrast with neighbouring Guangdong and other parts of China, where missionary schools operated in large numbers throughout the Republican period, but closed their doors or were re-organised after the Communist Party came to power in 1949. 2. Hong Kong has seven government-funded universities: City University of Hong Kong, Hong Kong Baptist University, Hong Kong University of Science and Technology, Lingnan University, The Chinese University of Hong Kong, The Hong Kong Polytechnic University, The University of Hong Kong. In addition, there is a tertiary-


Kingsley Bolton level educational institute, The Hong Kong Institute of Education. The Chinese University of Hong Kong is ofcially a bilingual institution, while the language policy of The Hong Kong Institute of Education stipulates the promotion of trilingualism and biliteracy. All other institutions are ofcially English-medium.

Bolton, Kingsley & Agnes S. L. Lam. 2006. Applied linguistics in China. In Keith Brown (ed.), Encyclopedia of language and linguistics [2nd edition, Vol. 1], 350356. Oxford: Elsevier. Brutt-Grifer, Janina. 2002. World English: A study of its development. Clevedon: Multilingual Matters. Carroll, John M. 2007. Edge of empires: Chinese elites and British colonials in Hong Kong. Hong Kong: Hong Kong University Press. Cheung, Anne. 1997. Language rights and the Hong Kong courts. Hong Kong Journal of Applied Linguistics 2. 4975. Chinese Government. 1992. The basic law of the Hong Kong Special Administrative Region of the Peoples Republic of China. Hong Kong: One Country, Two Systems Economic Research Institute. Cooper, Robert L. 1989. Language planning and social change. Cambridge: Cambridge University Press. Evans, Stephen. 2008a. Disputes and deliberations over language policy: the case of early colonial Hong Kong. Language Policy 7. 4765. . 2008b. The introduction of English-language education in early colonial Hong Kong. History of Education 37(3). 383408. Fishman, Joshua A., Charles A. Ferguson & Jyotirindra Das Gupta. 1968. Language problems of developing nations. New York: Wiley. Ghai, Yash P. 1999. Hong Kongs new constitutional order: The resumption of Chinese sovereignty [2nd edition]. Hong Kong: Hong Kong University Press. Gibbons, John. 1982. The issue of the medium of instruction in the lower forms of Hong Kong secondary schools. Journal of Multilingual and Multicultural Development 3. 11728. Government translator. 2010. Interview with translator, the Legislative Council, Hong Kong. Interview. May 18, 2010. Harrison, Godfrey & Lydia K. H. So. 1997. The background to language change in Hong Kong. In Sue Wright & Helen Kelly-Holmes (eds.), One country, two systems, three languages: A survey of changing language use in Hong Kong, 817. Clevedon: Multilingual Matters Ltd. Haugen, Einar. 1959. Planning for a standard language in Norway. Anthropological Linguistics 1. 821. He, Huifeng. 2010. Cantonese faces fresh threat in its birthplace. South China Morning Post. July 7, 2010: 1.

Language policy and planning in Hong Kong


HKSAR Government. 2002. Progress Report on the Education Reform (1): Learning for Life, Learning Through Life. < html.html> (15 May 2010). . 2009. Education Bureau Circular No. 6/2009: Fine-tuning the Medium of Instruction for Secondary Schools. < 7372 edbc09006e.pdf> (15 May 2010). . 2010a. Ofcial Languages Division website. < only/eng lish/aboutus/org/scsd/1470.html> (15 May 2010). . 2010b. Department of Justice website. < index.htm #6> (15 May 2010). Hopkins, Mark. 2006. Policies without planning? The medium of instruction issue in Hong Kong. Language and Education 20(4). 270286. Kaplan, Robert B. & Richard B. Baldauf. 1997. Language planning from practice to theory. Clevedon: Multilingual Matters. Kloss, Heinz. 1969. Research possibilities on group bilingualism: A report. Quebec: International Center for Research on Bilingualism. Kwok, Shirley. 1997. New rule will halve schools using English. South China Morning Post. March 22, 1997: 7. Lai, Chloe. 2009. Linguistic heritage in peril. South China Morning Post. October 11, 2009: 12. Lau, Chi-kuen. 1995. Language of the future. South China Morning Post. September 18, 1995: 19. Lau, Mimi. 2010. Legal questioning of journalists defended. South China Morning Post. August 4, 2010: 4. Law, Sau Wah 2002. Language planning in Hong Kong. Unpublished BA dissertation, The University of Hong Kong. Legislative Council. 2010. <> (15 May 2010). Mesthrie, Rajend, Joan Swann, Andrea Deumert & William L. Leap. 2000. Introducing sociolinguistics. Edinburgh: Edinburgh University Press. Munn, Christopher. 2001. Anglo-China: Chinese people and British rule in Hong Kong, 18411880. Surrey: Curzon Press. Neustupn , Jir V. 1970. Basic types of treatment of language problems. Linguistic Comy munications 1. 7798. Ng, Kwai Hang. 2009. The common law in two voices: Language, law, and the postcolonial dilemma in Hong Kong. Stanford: Stanford University Press. Pennycook, Alastair. 1998. English and the discourses of colonialism. London: Routledge. Phillipson, Robert. 1992. Linguistic imperialism. Oxford: Oxford University Press. Ricento, Thomas. 2006. Language policy: Theory and practice: An introduction. In Thomas Ricento (ed.), An introduction to language policy: Theory and method, 10 23. Oxford: Blackwell.


Kingsley Bolton

Rohsenow, John S. 2004. Fifty years of script and written language reform in the P. R. C. In Minglang Zhou (ed.), Language policy in the Peoples Republic of China: Theory and practice Since 1949, 2143. Boston: Kluwer Academic. SCOLAR. 2003. Action Plan to Raise Language Standards in Hong Kong. <http://www. ActionPlan-Final Report(E).pdf> (15 May 2010). Sin, King-Kui & Derek Roebuck. 1996. Language engineering for legal transplantation: Conceptual problems in creating common law Chinese. Language and Communication 16. 235254. Spolsky, Bernard & Richard D. Lambert. 2006. Language planning and policy: models. In Keith Brown (ed.), Encyclopedia of language and linguistics [2nd edition, Vol. 6], 561575. Oxford: Elsevier. Sweeting,Anthony & EdwardVickers. 2005. On colonizing colonialism: the discourses of the history of English in Hong. World Englishes 24(2). 113130. . 2007. Language and the history of colonial education: The case of Hong Kong. Modern Asian Studies 41(1). 140. Tollefson, James W. 1991. Planning Language, Planning Inequality. New York: Longman. Tsui, Amy B. M. 2004. Medium of Instruction in Hong Kong: one country, two systems, whose language? In James W. Tollefson & Amy B. M. Tsui (eds.), Medium of instruction policies: Which agenda? Whose agenda?, 97106. Mahwah, New Jersey: Lawrence Erlbaum. Vines, Stephen. 2010 Why Cantonese threatens Beijings language of power. South China Morning Post. July 31, 2010. [Accessed from South China Morning Post,]. Yau, Elaine. 2009a. Parents ock to schools that will switch to English. South China Morning Post. January 5, 2009: 1. . 2009b. Rethink policy on MOI, says expert. South China Morning Post. 28 March 2009. Yu, Verna. 2010. Guangdong chief speaks up for Cantonese. South China Morning Post. August 6, 2010: 4. Zhai, Ivan. 2010. Guangzhou blames people with ulterior motives for pro-Cantonese campaign. South China Morning Post. July 29, 2010: 6. Zhai, Ivan & VernaYu. 2010. Pro-Cantonese rally to hop the border into HK. South China Morning Post. July 27, 2010: 1. Zhang, Bennan & Robin R. Yang. 2004. Putonghua education and language policy in postcolonial Hong Kong. In Minglang Zhou (ed.), Language policy in the Peoples Republic of China: Theory and practice since 1949, 143162. Boston: Kluwer Academic. Kingsley Bolton,, is Chair Professor of English in the Department of English at City University of Hong Kong and Professor English Linguistics at Stockholm University, Sweden. He has delivered plenary papers and invited lectures

Language policy and planning in Hong Kong


at conferences in Asia, Europe and the US, and has published widely on language and society, English across Asia, and language and globalization. His publications include a monograph on the history of English in China, Chinese Englishes: A Sociolinguistic History (Cambridge University Press, 2003), and World Englishes: Critical Concepts in Linguistics (six volumes, co-edited with Braj B. Kachru, Routledge, 2006). From 2003 04, he served as Elected President of the International Association for World Englishes (IAWE), and, from January 2011, he will be co-editor of the Wiley-Blackwell journal, World Englishes.

English language ideologies (ELI) in Olympic Beijing


Abstract Academic positions vary greatly as regards attitudes towards and expectations of global English around the world. In this paper, I focus on the English language learning context in China and use data collected during the period when Beijing was preparing for and hosting the 2008 Olympic Games (20062008) to examine how English and the study of English are perceived and accessed by learners engaged in teaching and learning English outside formal educational institutions, namely, Olympic community English classes and English corners. In particular, I will explore, via interviews and group discussions, the reasons underlying learners voluntary choice of English learning; their perceptions of the status of English relative to Chinese in a globalizing context, the potential inuence English imposes on Chinese language and culture, and the signicance of English to Chinese society. My research ndings show that my informants associate English (learning) with multiple benets to life and career; they express a strong condence that English will not be a threat to the Chinese language and culture and they claim that English is useful to the development of China both now and in the long run. In this paper, I posit that my informants opinions reect a view of the social world from a particular historical, economic or political perspective or several perspectives combined and I relate their positions on the global spread of English in China to issues of language ideologies. I believe that the nonlinguistic socio-economic, social-historical and socio-political factors exert a crucial role in the emergence of dominant ideologies. Hence, besides presenting and interpreting the prevailing English language ideologies as captured in the discourse of my research respondents, in the last part of the paper, I will explore the possible social, cultural and political factors which caused the ideologies of language to emerge and the implications borne out by these ideologies in the context of Chinas ongoing globalization.


Lin Pan

Key words: English language ideologies, globalization, civilian discourse, 2008 Beijing Olympics, China

1. Introduction Academic positions manifest profound divergence as regards attitudes towards and expectations of global English. On the one hand, there are informed assertions that the global spread of English is natural, neutral and benecial (Brutt-Grifer, 2002; Crystal, 1997; Graddol, 2006). On the other hand there are claims that global English is detrimental in that it brings about social stratication, exclusion and problems associated with education, literacy, and language rights (Bamgbose, 2000; Brock-Utne, 2001; Phillipson, 1992; Phillipson and Skutnabb-Kangas, 1996; Skutnabb-Kangas, 2000; Skutnabb-Kangas and Phillipson, 1995). This issue of the spread of English is particularly relevant in the context of China. As a society that has been aiming to open up and modernize itself since the end of the 1970s, China has witnessed a meteoric rise in the popularity of English in recent decades, and the learning of English outside compulsory educational settings has ourished to the point where it has become a phenomenon worthy of scholarly attention. While the sociolinguistics of language identity and English language ideology in China has attracted the attention of some researchers in recent years (Du, 2001; Du, 2005; Jin and Cortazzi, 2003; Lo Bianco, Orton and Yihong, 2009; Xu, 2005), my research concentrates on the period when China was preparing for and hosting the 2008 Beijing Olympics. I anchor my research in this moment because I view this period (20062008) in China, especially in Beijing, as one that experienced accelerated social change and my hypothesis is that social change transforms and generates new discourses and ideologies. Hence, in this paper, I use data collected during this period to examine the ways in which English and the study of English are perceived and accessed by those engaged in teaching and learning English outside formal educational institutions, namely, Olympic community English classes1 and English corners2 which are independent of ofcial educational regulations. This paper features group discussions and interviews. The research respondents include working professionals as well as retired senior citizens.3 Though they come from distinct backgrounds, they all voluntarily chose to learn English in their spare time. My research questions centre upon such themes as the meaning of English and English learning for Chinese people, the inuence which English has brought about and will bring to Chinese language and culture and the signicance of English to Chinese society in general. To get a

English language ideologies (ELI) in Olympic Beijing


view on the above questions, I spent three months (July to September 2008) at English corners and Olympic English classes when Beijing was preparing for and hosting the Olympic Games and talked to the English learners there. I then conducted in-depth group discussions and interviews with four learners who were interested in sharing their stories and opinions and helping me with my research. To choose interview extracts, I identied thematic patterns, concepts, and shifting meanings arising from the interviews and group discussions, then categorized them according to their recurring themes. I then selected extracts from different categories based on my research questions and adopted thematic and content analysis (Braun and Clarke, 2006; Burr, 1995). In analyzing each interview extract, I followed Gees (1999) principles of dissecting discourses data and arranged all extracts into stanzas and lines (see extract 1 to 4 in section 3 for example). Each line is supposed to carry a piece of new and salient information and is numbered separately. There is often a pause, slight hesitation, or slight break in tempo after each line in speech (Gee, 1999). The letters a and b after some line numbers indicate that they contain the same information and belong to the same line. However, as the information embraced within a single line of speech is very often too small and limited to capture all that a speaker wants to say, I grouped a set of lines into a stanza. According to several scholars (Gee, 1986; 1991; Hymes, 1996; Scollon and Scollon, 1981), a stanza is a set of lines devoted to a single topic, event, image, perspective or theme. Therefore, whenever there are more than two topics or issues addressed in one extract, it is divided into stanzas and each stanza is named according to its theme. In this way, each specic theme which arises in the interview is foregrounded to facilitate thematic and content analysis.

2. English Language Ideologies Gramsci, Althusser, and Bourdieu used concepts such as hegemony (Gramsci, 1971; 1988), symbolic power, misrecognition (Bourdieu, 1991) and ideological state apparatus (Althusser, 1971; 2004) to elaborate upon certain ideologies and especially language ideologies, recognizing ideology as common sense or a form of invisible power (Bourdieu, 1991, p. 23) which routinely manipulates social life. I see English language ideologies as sets of ideas, sometimes manifested as discourse, that either support or critically question the spread of English as a global language. The sets of ideas usually present views of the social world from a particular historical, economic or political perspective and demonstrate a critical and political awareness of the effects of the global spread of English. In other words, different from the individualistic nature of attitudes towards the


Lin Pan

global spread of English, English language ideologies represent perspectives which have become accepted modes of knowledge and thought by a particular group to such an extent that they have lost much of their ideological appearance by having become naturalized, or simply common sense. With regard to English language ideologies, rst there are views which see the spread of English as inherently good for the whole world. These views usually trumpet the benets of English over other languages, suggesting that English has both intrinsic (the nature of the language) and extrinsic (the functions of the language) qualities superior to other languages (McCrum, MacNeil and Cran, 2002). This type of ideology is categorized as colonial-celebration. Second is instrumentalism that regards English language as a gatekeeper to the modernization of a state and the acquisition of social and economic prestige for individuals. As the worlds foremost auxiliary language by now, English is viewed as a linguistic capital (Bourdieu and Passeron, 1977) that is easily convertible to other forms of capital, such as educational qualications and higher education. This view is quite often voiced from what Kachru denes (2006) the Outer and Expanding Circles, where people regard English as a window to the world and a tool that empowers them, after having been previously oppressed by western imperialism and hegemony. Another set of ideas which is believed to form the dominant framework in TESOL is that the spread of English is natural, neutral and benecial, as long as it can coexist in a complementary relationship with other languages. Pennycook (2000) uses the term laissez-faire liberalism to summarize this general stance. Laissez-faire liberalism suggests that everyone should be free to do what they like with English, to use English in benecial ways and to use other languages for other purposes. It claims that we should neither engage in ideological nor political discussions of language and that we should make freedom of choice our central mode of understanding. On the other hand, there are ideologies critical of the global spread of English. One of the earliest discussions of this is by Cooke (1988) who has described English as a Trojan horse, arguing that it is a language of imperialism and of particular class interests. Phillipson and Skutnabb-Kangas have further developed this idea by introducing two concepts: linguicide and linguicism. They dene linguicide as the extermination of languages, an analogous concept to (physical) genocide (1995, p. 83) and linguicism is used by Skutnabb-Kangas (1988; 1998) to express her view towards the inequitable allocation of language rights. She argues that it represents a sort of linguistic racism (1988, p. 13; 1998, p. 16) and ideologies, structures and practices are used to legitimate, effectuate and reproduce an unequal division of power and resources (both material and immaterial) between groups on the basis of the language they speak (1995, p. 83). Based on this major concern, three subcategories of ideas are fur-

English language ideologies (ELI) in Olympic Beijing


ther illustrated: language ecology, linguistic imperialism and language rights. From the language ecology point of view, the problem with the spread of English is a complex disruption of the ecology of languages. The view of linguistic imperialism (Phillipson, 1992) argues that English penetrates or invades local spaces and disturbs the ecological balance that exists between people, language, culture and their environment, as it supports a global system of world trade which advantages the centre (rich and powerful) countries and disadvantages the peripheries (poorer countries). In addition, Tsuda (1994) and Phillipson and Skutnabb-Kangas (1996) insist that a linguistic human rights perspective should form the cornerstone of an ecology of a language paradigm whereby it is proposed that the right to identify with, to maintain and to fully develop ones mother tongue(s) should be acknowledged as a self-evident, fundamental individual linguistic human right (Skutnabb-Kangas, 1998, p. 20). The above review examines the varied views of the spread of English exhibited in different societies around the world and puts them into different ideological categories. Although they all hold water from their own particular angles, neither is sufcient to elucidate the complexity of ideological ramications of the spread of English in a particular locality. This paper takes the stance that the divergent English language ideologies are a result of the different perspectives taken which could be attributed to the unique social contexts and historical experiences of a particular locality. The paper also posits that quite often different layers of ideologies are intertwined simultaneously and the ideologies may evolve and become transformed through time in different localities. In this spirit, in the following sections I will investigate the language ideologies and implications present when China was preparing for and hosting the Olympic Games, 2008.

3. The Data As part of my research, I was rst interested in nding out why both retired citizens and professionals still working devote their spare time to learning English voluntarily and what English means to them. My research ndings showed that my respondents endowed English and the learning of English with multiple values and benets, relating English and English learning to the improvement of the quality of their lives, self-fullment, facilitation of communication, attainment of mutual understanding and career development. The extracts from interviews and group discussions reproduced below are what I was told when I engaged my informants in conversations about their perception of the multiple values embedded in learning English.


Lin Pan

English improves my life quality of life and keeps me up to date. Extract 1 is part of a discussion with two retired individuals at the English corner at Chaoyang Library. This English corner was set up in 2004, three years after the successful bid for the 2008 Olympics. Since then, Chaoyang library has been organizing whole-day English teaching and practice programmes every Saturday (the Olympic English class on Saturday morning and the English corner on Saturday afternoon, both open to everyone and with learning English for the Olympics as a goal). The discussion took place on a Saturday afternoon when a number of people were practising their oral English. The two major interviewees involved were a man, Ma, in his late 40s and a woman, Wu, in her 70s. Both of them had been attending this English corner for one year and three years respectively. The interview questions centred on why Chinese people, especially retired people, want to learn English. Learning English was positively labelled by the two interviewees as a means to improve their quality of life, to understand and learn from the outside world and a way to keep up with the present times.
Extract 1: 1a 1b 2 3 4 5 6 7a 7b 8a 8b 9 Stanza 1: English is for pastime and job hunting Researcher: I want to nd out why we Chinese put so much effort into learning English. I am interested in this question and at the same time confused by it. Dont you think we are kind of crazy about the language? Ma: Yeah, yeah, yeah Researcher: Why is it so? Is it really useful to people? Ma: I feel many people treat it as a pastime, especially when they are old. But for you young people, I guess it is for job hunting, employment and so on. Here, I feel many old people want to improve their quality of life. Stanza 2: Learning English to understand the world Researcher:(surprised) Wow, to improve the quality of life? Ma: Yeah, that is what I think. Researcher: Can we improve the quality of our lives by learning English? Ma: Actually, it is to enrich their spare time. Now, English is everywhere, even in (Chinese) songs, so if we do not learn English, I feel we cannot understand the outside world. Through learning English, you can learn from their (foreign countries) strong points/strengths to

10 11 12 13 14a 14b 15a 15b

English language ideologies (ELI) in Olympic Beijing 15c 16 overcome our own weaknesses. There are many aspects in foreign countries that you can learn from. Stanza 3: using English to promote Chinese culture Researcher: Well, do you think we also have some good aspects that they can learn from? Ma: Yes, of course. Our traditions can also be promoted by using English. You should use English to communicate with them. I feel if we understand English, our life quality of life will be different. You know, you get to know something different, such as foreign culture, ways of life, they are all different (from ours). Stanza 4: Learning English to understand the world (a senior lady joined the group) Researcher (turning to the senior lady): So can you tell me why you want to learn English? Whats the use of English in your life and in your work? Wu: Now, in China, in Beijing, we are opening up to the world on the street, many words are in English, we are like illiterates. Like illiterates who come to the city (if we dont understand English). Others: Yes, true. (nodding and agreeing). Wu: I am like a blind man. Ma: In the 50s or 60s, if you did not know Chinese characters, you were illiterate, now you are illiterate if you do not know English. Just now we said English is to improve our quality of life, now it is Wu: When the country had just been liberated, we learnt Chinese characters by posting the word, say our on the our jar, posting the word rice on the rice jar, we are now using the same method in learning English.


17a 17b 18 19 20 21 22a 22b 23

24a 24b 25 26 27 28a 28b 29 30 31 32 33 34 35 36 37

The above extract is subdivided into 4 stanzas according to its themes. In stanza 1, Ma identies the difference in attitudes towards English between the young people and the senior ones. He portrays English learning in a positive way since the status of English is associated with the improvement of his quality of life. Such an enthusiastic attitude is in sharp contrast to the preceding eras when learning English was frequently labelled with negative connotations and access to English was often denied by the State (see section 4 for details). It seems, with globalization and accompanying social changes taking place in China, this newly available language choice becomes for him a key to a new, enriched


Lin Pan

life, an empowering addition to his linguistic repertoire and indispensable for accessing all the indexicalities of the language. I would argue that his gracious embrace of English language, therefore, is not an individualistic free choice but an act aligned with the dynamics within the broader social discourse. Later in stanza 2, Ma further comments on the importance of English: Now, English is everywhere, even in (Chinese) songs, so if we do not learn English, I feel we cannot understand the outside world. The interviewee rst points out that English facilitates communication with the world and learning English helps towards a mutual understanding between people and countries. In addition, his words reveal that the cultural inuence of English, which is fast making inroads into Chinese society, has become a driving force for him and possibly other older people to opt for English courses. Also worth noticing is the perspective Ma adopts. Confronted with the cultural inuence of English, instead of raising protests against what some have called a cultural or ecological threat (Mhlhusler, 1996; Phillipson and Skutnabb-Kangas, 1996; Skutnabb-Kangas, 2000; Skutnabb-Kangas and Phillipson, 1995), he takes a relatively neutral or even passive stance with his somehow apologetic confession I feel we cannot understand the outside world. He dichotomizes the world into China and the outside world, and indicates that he feels English is a tool helping him link the two. Similar to Mas comments, in stanza 4 (line 2437), the female interviewee Wu also admits the inuence of English in China and says that now China is opening up to the world, and English words are seen everywhere on the streets. She then compares herself and her group of learners to illiterates and a blind man, and confesses that they use the same method and put in the same amount of effort to learn English as they learnt Chinese when they were young. This point of view can be better understood in relation to globalization and Chinas efforts towards internationalization as indicated by China is opening up to the world. Here again, research respondents knowingly anchor their individualistic choices within the macro, all-encompassing context of globalization, which has determined English learning as an imperative both necessary and inevitable. Through their self-depiction of becoming illiterates and a blind man, they take on an implicitly subordinate role by exalting the signicance of English over that of Chinese. Here, it is clear that the power of the language is often determined by its advantages of communication or practical values; the speaker is attracted to the power of English and is oriented towards it (Calvet, 2006; De Swaan, 2001). In line 15, stanza 2, Mas utterance through learning English, you can learn from their strong points to overcome our own weaknesses reveals that English is regarded as an instrument for self-improvement. In a similar way, this instrumental view is voiced again in line 17 of stanza 3, by his comment: our

English language ideologies (ELI) in Olympic Beijing


traditions should also be promoted by using English and you should use English to communicate with them. The instrumental nature of his words echo , a slogan used learning from the barbarians to check them ( , for English a century ago) and using foreign things to serve China ( advocated by Mao Zedong, former Chairman of P. R. C., in the 1960s). It seems that in the long evolution of the ideological history of English in China, the nature of instrumental ideologies still remains unchanged in essence (this point will be further addressed in section 4). I like English and it will be benecial for my job and career. Extract 2 is from an interview with Ni, a young professional. She works for a well-known architectural design company in Beijing. She had already graduated from university but she was very enthusiastic about English learning and kept attending part time English courses and English corners. I was very curious about the reasons behind her enthusiasm and wanted to explore what learning English meant for her as a young and enterprising professional.
Extract 2: Researcher: You have continued learning English. Is this for your career or Stanza 1: the signicance of English learning Ni: Out of interest, actually. Studying English in itself can bring many advantages. You can make a lot of friends, and it opens a lot of windows for you. Also, once you know the language, and have a good command of it, it is convenient for communicating. No matter whoever you are with, if you have enough vocabulary, you will have better comprehension and better ability for expressing yourself. I also feel that youll have a special skill which nobody has except you, like CCTV9 news anchors look at their bearing, their ease, I envy them very much. I feel that if I could speak as uently, then it would be my strength. I really want to have such a skill, but I feel I stopped. I discontinued so my English is no longer uent now. Stanza 2 English brings benets to my career But I sometimes think that if my English was good, it could offer me the best opportunities. You know, if your English, especially your spoken English I have no problem with reading or writing English, but if I could speak good English,

1 2 3a 3b 4a 4b 5a 5b 6a 6b 7a 7b 8 9

10a 10b 11a 11b 11c

11d 12a 12b 13 14 15a 15b 16a 16b 17a 17b 17c 17d 18a 18b 18c 18d

Lin Pan then I . . . . I will be appreciated by my company. Researcher: you mean promotion?! Ni: Yes, yes. I will have a special skill. The boss would say emm, you are special and he will notice you. But I feel I havent shown myself . . . . Now I am looking after an American design company. Everytime they come, I go to receive them, and they have very good relations with our general manager. Our company is a hierarchical one one layer on top of another, its very conservative, very much like a state-owned enterprise. The general manager wont know Well, he will know that I am the project manager for design, but he will not know that you have good English. He does not understand you. But suppose my English is good, When they come, I dont need a translator, they speak English and I speak English, we can communicate smoothly isnt that benecial to me?! Stanza 3: English helps me to make friends Ni: Actually my English was good When I went with you to the class in 2002, I felt that my spoken English was good at the time I could speak with foreigners. I was an architects assistant, an assistant to an American architect. Then I could speak good English, My colleagues in the design company were all very kind to me. Emm, they might have felt I was a new-comer and they might have felt I was very special, so every time they were doing something, they would say, Ni, come here, they would ask me to join them. At the time I was very young, and I felt very happy with them. Stanza 4 English is a passport I feel people should be special, because when you have such special qualities, it is like having a passport. It brings a lot of opportunities and a lot of friends. Otherwise, if we are all the same, it makes no difference whether we make friends with you or with others. Besides, I feel it is a skill. If you have invested so much already and you have spent so much time,

19a 19b 20a 20b 20c 20d 20e 20f 21a 21b 21c 21d 22

23a 23b 23c 24a 24b 25 26a

English language ideologies (ELI) in Olympic Beijing 26b 26c 27 28a 28b then you should continue doing it. And for you, it is an accomplishment. Thats what I think. Im not sure where this will take me but I hope I can gradually improve.


In the above extract, Ni rst illustrates the importance that English holds for her, reecting on how it could be of value for her career and for enhancing her social network. She expresses her unconditional longing for a good command of English which is evident by her repeated use of positive adjectives and phrases throughout the four stanzas. These include it opens a lot of windows, it is convenient for your communication, better comprehension, better ability, special gift, ease, strength in stanza 1; best opportunity, be appreciated, special and good in stanza 2; special in stanza 3 and special, a passport, a skill, self-attainmentand elevate yourself in stanza 4. As previously stated, the particular use of the word special is very prominent and is repeated ve times in all four stanzas. Accordingly, Nis generous admiration of English TV news anchors in stanza 1, her frequent use of positive words and phrases in addition to the constant repetition of the word special points to her earnest yearning for a good command of English. The way she portrays the benets of English naturally link to Bourdieus concept (1992) of symbolic capital. Bourdieu sees symbolic capital as a crucial source of power, the value of which is derived from its scarcity. The acquisition of symbolic capital not only results in possession of the capital itself, but also in the accumulation of prestige, recognition and authority. In the same way, Ni regards English competence as a special skill- a symbolic capital the acquisition of which would bring her not only the skill itself but also the derivatives such as job promotion, friends, colleagues recognition, together with such valuable intangibles as elegance and gracefulness. In short, for Ni, the possession of good English communication competence will certainly lead to improvement of both her social and economic positions, difcult to obtain otherwise. The above two extracts have illustrated the signicance of English learning from the perspectives of retired citizens and a young professional. Given the fact that my research respondents associate multiple benets with English and English learning for both themselves and the country as a whole, I started to wonder about the inuence of English on Chinese language and culture, that is, how much faith do they have that learning English will affect Chinese language and culture? And if there is an inuence, is it going to be positive or negative? This research question surfaced as I recalled the 1980s when China put into effect the Reform and Opening-up policy, when English was still regarded as a


Lin Pan

source of spiritual pollution (jingshen wuran in Pinyin or ) and the social-cultural values which English represented were avoided to a great extent in learning and teaching (An, 1984;Yue, 1983). I wanted to explore whether or how much the linguicide and linguicism ideologies (Phillipson and SkutnabbKangas, 1995; Skutnabb-Kangas, 1988; 1998) hold true for the English learners in the Olympic classes. And when I enquired in my interviews whether English is perceived as a threat to Chinese language and culture, most respondents gave negative answers. In most cases, my interviewees well-dened condence with regard to Chinese language and culture was clearly unmistakable. The following group discussion with Ma and Wu provides such an example. Your root is Chinese.
Extract 3: 1 2 Stanza 1: English is simpler and advanced Ma: There are many road signs, many are in English, English is simpler than Chinese. Researcher: Many road signs are in English. Will this be a threat to our Chinese language? Wu: If they are advanced, we should learn together, improve together. You dont know it so you should learn it. The problem is that we are too old. Stanza 2: It is impossible for English to become a threat Ma: This issue was once discussed in a debating contest. A teacher said we Chinese have ve thousand years of civilization, so it is impossible (for English to become a threat). Wu: For example, Hong Kong and Taiwan, These places wanted to reduce the Chinese elements for many years but failed. Your root is Chinese. Our ve thousand years of civilization are respected by foreigners. Stanza 3: English and Chinese are complementary Wu: Also, while there may be 16 meanings for one English word. There may be more than 40 meanings for a similar word in Chinese. If youve reached a certain level and need to translate them into Chinese, then it depends on your Chinese language skills, not your English skills. Ma: Also I feel studying English is not a case that one diminishes the other, but rather that they are complementary. Sometimes, English is not as rich as Chinese, and Chinese can complement English.

3a 3b 4

5a 5b 5c 6a 6b 6c 7a 7b

8a 8b 9a 9b 10a 10b 11a 11b

English language ideologies (ELI) in Olympic Beijing 12 13 14 But Chinese is not as simple as English, not as concise as English. So they are complementary. But not a threat to each other.


Extract 3 is divided into three themes in three stanzas. In the rst stanza, Ma and Wu note their positive opinion about English language by saying that English is simpler than Chinese in line 2 and English is seen as an advanced thing in line 3a. The words simpler and advanced demonstrate an appreciative attitude towards English language as the celebrators (McCrum, MacNeil and Cran, 2002) view the spread of English to be inherently good, in that it both has intrinsic advantages in terms of the nature of the language (suggested by the word simpler) and extrinsic benets in terms of functions (by advanced things) that are superior to other languages. Then in stanza 2, with regard to the inuence of English on Chinese culture, Ma declares that English could never become a threat as the Chinese language and culture have ve thousand years of civilization. Wu also supports Mas opinion by quoting the situation in Hong Kong and Taiwan as further examples. According to her, the roots of a person or a nation cannot be changed (line 6c) and she reinforces her argument on the solidity of Chinese roots by commenting that our ve thousand years of civilization is respected by foreigners. By presenting foreigners respect towards Chinese culture as a counterargument, she makes the point that there is no reason for the Chinese to despise or abandon their own culture. These points are further elaborated and strengthened in stanza 3. First, for Wu, translation skills signify a persons language competence, and her prioritizing the mother tongue over English in achieving good translation ts well with her argument that your root is Chinese and this will not change. Ma in lines 10a and 10b also dismisses the idea that English could be a threat to Chinese and makes it clear that studying another language does not necessarily lead to the decline of the mother tongue, as languages could complement each other. The ideology reected is that though English is important, the Chinese national language is the primordial foundation of national culture and the matrice of the national mind (Hobsbawm, 1990, p. 51). Thus, the correlation between national language and national identity is well depicted. The above extract elucidates the point that instead of regarding English as a threat to Chinese language and culture, my respondents tend to think about English in a more constructive way. This made me reect on the role of Olympic English classes. Why did my respondents, more often than not, generate positive discourse on English and learning English? What are their objectives for learning English, outside the mainstream education system and not compulsory? And how do they see the signicance of English to Chinese society? The extract that


Lin Pan

follows is selected from the interview with Zhao, one of the advocates of the English Speaking Association of Chaoyang District, in which she explains to me the objectives of the training in the Olympic class and voices her opinions about the signicance of English in Chinese society in the long run. The Olympics as a goal was not enough. (We) did not want to stop learning English after the Olympics.
Extract 4: 1 Stanza 1: Studying English for the Olympics Zhao: In fact the objectives of our classes are very clear: they are held for the Olympics. Researcher: Then how about after the Olympics? Zhao: At the very beginning, we thought that the environment for learning English wasnt there, so we wanted to create such an environment. When we created the environment for learning, we needed to have an important objective, aimed at the Olympics. But when we were working towards this goal, we felt that the Olympics as a goal was not enough. Stanza 2: Studying English to go beyond the Olympics Zhao: So we needed to revise our aim. Our original class did not want to stop learning English after the Olympics. What could we do? We should train volunteers; we should meet our social need for volunteers, for example, in museums, exhibition centres, hospitals, schools, etc. they perhaps need people with such aspirations. Schools might need the experience of such people. Museums and exhibition centres need translators. For example, foreigners come to China and become very interested in Chinese things but if there is no one to explain these things to them, there are many who will not understand them. Therefore it is only by talking to them that these things will become clear. So that is our next training target. Researcher: Oh, so it was like that in the beginning. The Olympics served as an opportunity to implement these training classes. But after the training got started, you discovered a long-term goal. Zhao: Yes, when we rst started training, we did not have such a long-term goal. Now as the training goes on, the goal grows bigger and bigger. Researcher: Actually it is [

2a 2b 3a 3b 4a 4b

5 6 7 8 9 10 11 12a 12b 12c 12d 13

14 15

English language ideologies (ELI) in Olympic Beijing 16 Zhao: [Step by step Researcher: Therefore we can say that the Olympics is an opportunity, a platform to enable us to cast our vision beyond this.


Zhao, along with several other retired citizens, advocated the learning of English after the successful bid for the Olympics. With the help of their neighbourhood committees, they set up an English speaking association in Chaoyang District and opened more than 20 Olympic English classes afliated to it; as Zhao explains in line 1, all of the classes were opened for the Olympics. In fact, according to the ofce of the Beijing Speaks Foreign Languages Programme (BSFLP), Olympic English classes sprouted up across Beijing to ensure that by 2008, 5 million inhabitants in Beijing, that is, 35% of Beijing regular inhabitants, are prepared to help foreigners when Beijing hosts the Olympic Games. As revealed in stanza 1 of the above conversation with Zhao, the Vice-Chair of the Chaoyang English Speaking Association, claries that although all of the classes were for the Olympics, their initial idea to open them was to create an English environment. This means that the Olympic English classes were originally conceptualized as only a bridge to create an environment for English learning as their idea of learning and practising English came before the goal of learning for the Olympics. Moreover, in stanza 2, Zhao says that they realized that the Olympics could only be a short term goal and they dont want to stop learning English after the Olympics. In order to sustain their English learning environment, they re-targeted other social needs. This view of English beyond the 2008 Olympics as a long term goal, in a similar manner, was ofcially voiced by Chen Lin, a renowned professor of ELT in China and an adviser for the ofce of the Beijing Speaks Foreign Languages Programme (BSFLP). He explained, We learnt English for 2008 but we should not stop at 2008. We should go beyond 2008, as Beijing will be more open and active in international communication after the Olympics.4 He also said that I feel a greater signicance of learning English is that it changes the way people think, that the Olympic English mania has made Beijingers more open-minded. In other words, the Beijingers no longer remain defensive about their own language but they get to know another language and culture and communicate with another culture with good intentions. Worth noticing is the convergence of thoughts between Chen Lin, an ofcial representative, and Zhao, a civilian representative, in that both attribute the learning of English to be of instrumental value for having international communication and possibly going international. In addition, Chens labelling of the signicance of learning English goes explicitly against the ideologies reected in some other parts of the world, such as Africa (e.g. Zambia, Tanzania), some parts of Asia (e.g. India, Sri Lanka)


Lin Pan

(Kachru, 1990) and/or some European countries (e.g. France) (Flaitz, 1988), where voices proliferate to protect their national language against the invasion of English, and to minimize the cultural and ideological inuence that English will exert on peoples minds. The positive signicance attributed to learning English from both ofcial and civilian levels and the overlapping tone between the civilian and the ofcial discourses are worthy of further examination, which I will attempt in the following section.

4. Multi-layered ideologies and their contextual implications As stated earlier in section 2, English language ideologies are sets of ideas, sometimes taking the form of discourse, that either support or critically question the spread of English as a global language. In the above section, by examining the discourses of the language learners, I have explored the different layers of English language ideologies in contemporary China when the Olympic Games were due to take place in Beijing. The matrix of opinions expressed about learning English t into several ideological categories as we distinguished earlier in this paper. First, there is a wide array of instrumental views, exemplied by words such as English improves the quality of my life and keeps me up to date. In addition, young professionals see English as a passport to their job and career and English language competence is regarded as a gatekeeper towards a better quality of life, easier and better communication with the world and more promising career prospects a linguistic capital to use Bourdieus term (1977). Furthermore, it is not uncommon to nd a celebrative attitude among respondents/interviewees, who tend to view the spread of English as inherently good for the whole world, and that English has both intrinsic (the nature of the language) and extrinsic (the functions of the language) qualities superior to other languages (McCrum, MacNeil and Cran, 2002). In parallel, the view of laissez-faire liberalism, which sees the spread of English as natural, neutral and benecial, is also somewhat ubiquitous among my interviewees, exemplied by remarks such as your root is Chinese, it is impossible (for English to become a threat), and they see studying English, not that one diminishes the other, but rather that they are complementary. Additionally, the data shows that some impact of English in China is clearly felt by people, as in now, English is everywhere, even in (Chinese) songs and China is opening up to the world, and English words are everywhere on the streets . However, many interviewees showed great condence in the integrity and prosperity of Chinese language and culture. In contrast to the movement of remedying the wrongs of language hege-

English language ideologies (ELI) in Olympic Beijing


mony, to advocate the right of all people to speak the language of their choice, to ght language imperialism abroad and linguicism at home, and to strengthen language rights in international law as proposed by some scholars (Cooke, 1988; Day, 1985; Skutnabb-Kangas and Phillipson, 1995), the condence of my research respondents is exhibited by their remarks on the impossibility of English becoming a threat when faced with a civilization of 5,000 years and the roots of a person or a nation will not be changed. It could be argued that the English language ideologies discussed and summarized in the previous analysis echo Kachrus (2006) notion of the Outer and Expanding Circles of English speaking and learning countries. However, I would counter-argue that despite apparent similarities, the social, cultural and historical factors which inform China or other countries vary and are often distinctively unique. That is, peoples ideologies of English in China may be attributed to its unique social context and historical experiences. This is because, as made clear earlier, language ideologies are contextualised sets of beliefs about languages. In other words, they are cultural and political systems of ideas about social and linguistic relationships (Gal and Irvine, 1995; Irvine, 1989). Following this line of thinking, the ways in which languages are used and thought about by a particular person are never just about languages but also about community and society (Woolard, 2004) and the world as a whole (Calvet, 2006; De Swaan, 2001; Wallerstein, 2000; Wallerstein, 2004). As non-linguistic socio-economic, socio-historical and socio-political factors play a crucial role in the emergence of dominant language ideologies, a context-specic discussion and interpretation seems to be in order. Therefore, at the end of the paper, I will relate the English language ideologies I summarized above to Chinas specic social-historical background and will locate the right time and moment for the ideologies of language to emerge while explaining their implications. Firstly, in general, my informants expressed open-minded attitudes toward learning English, which can be categorized as either instrumental, celebratory or laissez-faire. At the national level, English is perceived as a necessary means for helping the nation open up further, a valuable resource for realising its modernisation programme, and an important cornerstone of international competition. On a personal level, prociency in English is seen as a key to a host of opportunities: to enter and graduate from university, to go abroad for further education, to secure desirable jobs in public and private sectors, and to be eligible for promotion to higher professional ranks. Consequently, English prociency is associated with superior national, social and economic prestige. These layers of meaning of English learning indisputably relate themselves to the impact of globalization and Chinas determined drive for internationalization in recent decades, as peoples discourse is a way to show how they perceive


Lin Pan

globalization and their discourse itself becomes a facet of globalization (McKay and Bokhorst-Heng, 2008). English gained momentum in China from the 1970s and since then has enjoyed a high level of prestige into the 21st century, mostly arising out of Chinas need for modernization and economic development. The states full embrace of English in various sectors has also generated ideological changes in people as they become increasingly convinced of the indispensability of English to Chinas development. The positive discourse about English made by my research respondents to a great extent bespeaks the swift socio-economic changes China has gone through. The value package of English places itself in sharp contrast to the previous eras when English was constantly associated with barbarism, imperialism and colonialism in the late 19th century and early 20th century, and a language of our enemy when the Peoples Republic of China had just been founded in the 1950s and 1960s (Adamson, 2004; Lam, 2005; Pan, 2010). Therefore, learning English is not only encouraged by the State as a means of accelerating modernization and internationalization, but it is also seen at a more personal and individualistic level as symbolic capital to attain brighter career prospects and a better life style. All the above serves as an explanation for the heavy emphasis placed on the benets English brings in peoples discourse. Secondly, with regard to the impact of English on Chinese language and culture, the responses are, more often than not, centred upon condence towards the integrity of Chinese language and culture, the adaptive learning attitude and the desire to improve the Chinese economy and promote Chinese culture via English. As previously stated, this response appears to be in conict with the well-founded view of linguicide and linguicism, which is used to explain inter-lingual relationships in many parts of the world. Nevertheless, a close examination of the history of Chinas educational interactions with foreign countries in general (Lam, 2005), and foreign-language teaching in particular (Jin and Cortazzi, 2003; Pan, 2007) reveal constant swings between selective introduction and prohibition of foreign culture. A period of selective exposure was always followed by a swing back to the prohibition of foreign values and beliefs or the substitution of foreign culture with Chinese culture, when the state was concerned that fundamental Chinese values and beliefs were being jeopardised by the inux of foreign ones. Hence, the cultural and ideological inuence of English has always been reduced to a minimum. As summarized by Adamson (2002, p. 231), China has had a strategy to mitigate undesirable cultural transfer in place since the mid-nineteenth century: a policy of controlled and selective appropriation, to use English for the purposes of state building, while maintaining cultural integrity. This selective appropriation tradition, in addition to Chinas efforts in recent decades at internationalization, offers a way

English language ideologies (ELI) in Olympic Beijing


to interpret why my informants, no matter whether young or old, are condent, optimistic and are willing to adapt when learning English. Last but not the least, as already noted, the State identies the mother tongue languages as being the languages of cultural identity and ethnic cohesion, and English as the language of commerce, global connectedness and modernization. With the nations accession to the World Trade Organization (WTO) and the successful bid for the 2008 Beijing Olympic Games in 2001, the popularity of English has reached new heights amongst government policy makers, educationalists and the Chinese public. Nevertheless, both civilian and ofcial discourses overlap in that the signicance of English learning is going beyond the Olympics and helping to equip society for a more modernized China. Such instances of overlapping may be better understood within the power and cultural power framework of China today. According to Blommaert (2008), when the cultural power exercised by the ruling class is no longer perceived as power, but has become the normal state of things and has begun to organise peoples lives in a way that is no longer perceived as oppressive, irrational or questionable, it becomes hegemonic power. Blommaerts idea can be traced to Bourdieus (1991, p. 23) elaboration of symbolic power and misrecognition and Gramscis (1971, p. 377) concepts of consent and hegemony. For Bourdieu, the term symbolic power refers not so much to a specic type of power, rather to an aspect of most forms of power as they are routinely deployed in social life. In addition, he points out that symbolic power is invisible power which is misrecognized as such and thereby recognized as legitimate (Bourdieu, 1991, p. 23). To interpret the concept of power in Gramscis terms (1971), it is a form of hegemony exercised by the ruling class to make their ideas the most natural and dominant ones for that society and develop and maintain consent. Hence, cultural power is not exercised coercively, but routinely. People consent to particular formations of power because the dominant cultural groups generating the discourse represent them as natural and legitimate (Gramsci, 1971, p. 377; Mayr, 2008, p. 13). By generating consent (legitimacy) among the population, the more legitimacy the dominant groups have, the less coercion the ruling class need to apply. The same applies to the way the State has been propagating the use of English in Chinese society in recent decades, as the functional usefulness of English to the nation as a whole is placed at the forefront, overshadowing its less instrumental implications. Therefore, individuals, as products of power, perceive it as legitimate and accept the language as a thing that is for their own good (Blommaert, 2008). We thus see that cultural power is functioning in its least overt physical form, that is, in the routine ow of day-to-day life and day-to-day discourse.


Lin Pan

1. Olympic English classes were gradually set up in different residential communities in Beijing after Chinas successful bid for the Olympic Games in 2001. They were usually advocated and organized by enthusiastic (retired) residents, sponsored by local resident committees as free venues and where English was taught by volunteer English teachers. The goal of Olympic English classes was to prepare the local citizens to become Olympic language volunteers and to otherwise improve the quality of residents language ability. There were no compulsory textbooks. 2. English corners were locations where people from various walks of life gathered together, usually once a week, to practise English. They were usually loosely organized in that no local authority was involved and people were free to join and leave. The only principle unanimously observed is that English, not Chinese, should be used in conversations. 3. Consent forms were signed before each interview and group discussions. All research participants are kept anonymous and pseudo names are used in this paper. 4. July 31st 2008, Xinhua New Net, last traced on 08 January 2009 8869032 1.htm

Adamson, Bob. 2002. Barbarian as a foreign language: English in Chinas Schools. World Englishes, 21 (2), 231243. Adamson, Bob. 2004. Chinas English: a history of English in Chinese education. Hong Kong: Hong Kong University Press. Althusser, Louis. 1971. Lenin and philosophy, and other essays. London: New Left Books. Althusser, Louis. 2004. Ideology and ideological state apparatuses. In Julie Rivkin and Michael Ryan (eds), Literary theory: an anthology. Oxford: Blackwell. An, Zhiguo. 1984. Ideological contamination claried. Beijing review, 27 (4). Bamgbose, Ayo. 2000. Language and Exclusion. Muenster, Hamburg, London: LIT Verlag. Blommaert, Jan. 2008. Notes on Power. Working paper in language diversity: University of Jyv skyl . a a Bourdieu, Pierre. 1991. Language and symbolic power. Cambridge: Polity Press. Bourdieu, Pierre. and Passeron, Jean-Claude. 1977. Reproduction in Education, Society and Culture. California: Sage Publications. Bourdieu, Pierre. and Wacquant, Loic. 1992. An Invitation to Reexive Sociology. Chicago: University of Chicago Press and Polity. Braun, Virginia. and Clarke, Victoria. 2006. Using thematic analysis in Psychology. Qualitative Research in Psychology, 3 (2), 77101.

English language ideologies (ELI) in Olympic Beijing


Brock-Utne, Birgit. 2001. Education for all in whose language. Oxford Review of Education, 21 (1), 11534. Brutt-Grifer, Janina. 2002. World English: A Study of Its Development. Clevedon Buffalo Toronto: Multilingual Matters Ltd. Burr, Vivien. 1995. An introduction to social constructionism. London: Routledge. Calvet, Louis-Jean. 2006. Towards an ecology of world languages (Andrew Brown, Trans.). Cambridge: Polity Press. Cooke, David. 1988. Ties that constrict: English as a Trojan horse. In Alister Cumming, Antoinette Gagn & Janet Dawson (eds.), Awareness: Proceedings of the 1987 TESL e Ontario Conference, 5662. Ontario: TESL Ontario. Crystal, David. 1997. English as a global language. Cambridge: Cambridge University Press. Day, Richard.R. 1985. The ultimate inequality: Linguistic genocide. In Nessa Wolfson and Joan Manes (eds.), The language of inequality, 163181. Amsterdam: Mouton de gruyter. De Swaan, Abram. 2001. Words of the world: the global language system. Cambridge: Polity. Du, Hui. 2001. The globalisation of the English language: Reections on the teaching of English in China. International Education Journal, 2 (4), 126133. Du, Hui. 2005. False alarm or real warning? Implications for China of teaching English. Journal of Educational Enquiry, 6 (1), 90109. Flaitz, Jeffra. 1988. The ideology of English: French perception of English as a world language. Berlin, New York and Amsterdam: Mouton de gruyter. Gal, Susan. and Irvine, Judith. 1995. The boundaries of languages and disciplines: how ideologies construct difference. Social Research, 62, 9671001. Gee, James. P. 1986. Units in the production of discourse. Discourse Processes, (9), 391422. Gee, James. P. 1991, A linguistic approach to narrative. Journal of narrative and Life History, (1), 1539. Gee, James. Paul. 1999. An introduction to discourse analysis: theory and method. London: Routledge. Graddol, David. 2006. English Next: why global English may mean the end of English as a Foreign Language. Plymouth: British Council. Gramsci, Antonio. 1971. Selections from the prison notebooks of Antonio Gramsci. London: Lawrence and Wishart. Gramsci, Antonio. 1988. An Antonio Gramsci reader: Selected writings, 1916 - 1935. In David Forgacs (ed.). New York: Scholcken Books. Hobsbawm, Eric. J. 1990. Nations and Nationalism Since 1780. Cambridge: Cambridge University Press. Hymes, Dell. 1996. Ethnography, linguistics, narrative inequality: towards an understanding of voice. London: Taylor & Francis. Irvine, Judith. T. 1989. When talk isnt cheap: Language and political economy. American Ethnologist, 16, 248267.


Lin Pan

Jin, Lixian. and Cortazzi, Martin. 2003. English language teaching in China: A bridge to the future. In Wah Kam. Ho and Ruth Y. L. Wong (eds), English language teaching in East Asia today. 131145. Singapore: times academic press. Kachru, Braj. B. 1990. The alchemy of English: the spread, function, and models of non-native Englishes. Urbana and Chicago: University of Illinois Press. Kachru, Braj B., Kachru, Yamuna, and Nelson, Cecil L. 2006. The handbook of world Englishes. Malden, MA: Blackwell Publication. Kachru, Yamuna. and Nelson, Cecil. L. 2006. World Englishes in Asian contexts. Hong Kong: Hong Kong University Press. Lam, Agnes. 2005. Language education in China: policy and experience from 1949. Hong Kong: Hong Kong University Press. Lo Bianco, Joseph, Orton, Jane. and Gao, Yihong. 2009 (eds). China and English: globalisation and the dilemmas of identity. Bristol: Multilingual Matters. Mayr, Andrea. 2008. Language and power: an introduction to institutional discourse. London: Continuum. McCrum, Robert, MacNeil, Robert and Cran, William. 2002. The story of English. (New and rev. / ed.). London: Faber and Faber and BBC Books. McKay, Sandra L. and Bokhorst-Heng, Wendy D. 2008. International English in its sociolinguistic contexts: towards a socially sensitive EIL pedagogy. New York and London: Routledge M hlh usler, Peter. 1996. Linguistic ecology: language change and linguistic imperialism u a in the pacic region. London: Routledge. Pan, Lin. 2007. Book review: Language education in China: policy and experience from 1949. Language Learning Journal, 35 (1), 131133. Pan, Lin. 2010. English Language Ideologies in China: a synchronic and diachronic perspective. London: Institute of Education, University of London PhD thesis. Pennycook, Alastair. 2000. English, Politics, Ideology: From Colonial Celebration to Postcolonial Performativity. In T. Ricento (ed.), Ideology, politics, and language policies: focus on English (Vol. 6, 107120). Amsterdam; Philadelphia: John Benjamins Publication. Phillipson, Robert. 1992. Linguistic imperialism. Oxford: Oxford University Press. Phillipson, Robert & Skutnabb-Kangas, Tove 1995. Papers in European language policy. Roskilde: University of Roskilde ROLIG-papir 53, 1995. Phillipson, Robert. and Skutnabb-Kangas, Tove. 1996. English only worldwide or language ecology? TESOL Quarterly 30 (3), 429452. Scollon, Ron. and Scollon, Suzie. 1981. Narrative, Literacy, and Face in Interethnic Communication. Norwood: Ablex. Skutnabb-Kangas, Tove. 1986. Multilingualism and the Education of Minority Children. Linguicism rules in education, 4272. Skutnabb-Kangas, Tove. 1998. Human rights and language wrongs a future for diversity? Language Sciences. 20 (1), 528. Skutnabb-Kangas, Tove. 2000. Linguistic genocide in education or worldwide diversity and human rights? Mahwah, N. J.; London: Lawrence Erlbaum.

English language ideologies (ELI) in Olympic Beijing


Skutnabb-Kangas, Tove. and Phillipson, Robert. 1995. Linguicide and Linguicism. in Robert Phillipson and Tove Skutnabb-Kangas (eds), Papers in European language Policy, 8391. Roskilde: Roskilde Universitetscenter, Lingvistgruppen. Tsuda, Yukio. 1994. The diffusion of English: its impact on culture and communication. Keio Communication Review, (16), 4961. Wallerstein, Immanuel M. 2000. The essential Wallerstein. New York: New Press. Wallerstein, Immanuel M. 2004. World-systems analysis: an introduction. Durham: Duke University Press. Woolard, Kathryn A. 2004. Is the past a foreign country?: Time, language origins, and the nation in early modern Spain. Journal of Linguistic Anthropology, 14 (1), 5780. Xu, Zhichang. 2005. Chinese English. What is it and is it to become a regional variety of English? Perth, Australia: Curtin University of Technology PhD thesis. Yue, Ping. 1983. Preventing ideological pollution. Beijing review. 24 (22). Lin Pan completed her PhD at the Institute of Education, London in 2010. She is interested in language and globalization, multilingualism, and English language teaching in China. She now teaches at Beijing Language and Culture University.

English as a medium of instruction in Asian education (from primary to tertiary): Implications for local languages and local scholarship

Abstract This article will review and critique the general trends towards the signicant increase in the teaching of English and the use of English as a medium of instruction in education throughout East and Southeast Asia. I shall focus on two levels. First I shall discuss the situation as regards the role of English in primary schools and then consider its role in tertiary education. I shall argue that the trend towards the ever earlier introduction of English into the primary curriculum, along with the push for the relevant national language, is not only pedagogically ill-advised for the great majority of primary school children in the region, but also represents a serious threat to local languages and, perhaps most importantly, to childrens sense of identity. The increasing trend towards Englishmedium programmes at the tertiary level also threatens local languages and the status and value of knowledge and scholarship written and disseminated in languages other than English. At the same time, indigenous knowledge disseminated through English, while it may reach an international audience, may be essentially reframed through being translated into English. I shall conclude with recommendations designed to encourage multilingualism in local languages at the primary level and the implementation of bilingual policies at the tertiary level. Keywords: language education; multilingualism; English as a medium of instruction; local scholarship

1. Introduction This article will review and critique the general trends towards the signicant increase in the teaching of English and the use of English as a medium of instruction in education throughout East and Southeast Asia.


Andy Kirkpatrick

In almost all countries of Asia, English has now become a core course in primary schools. Indonesia is an exception, where English remains an optional subject at primary level, but even there parental demand means that primary schools are required to offer English. Indonesia is also currently experimenting with international standard schools, known as SBIs from their Indonesian acronym sekolah bertaraf internasional. These SBIs use English as a medium of instruction for maths and science, theoretically from primary 4 but, in effect, often from primary 1. Throughout the rest of Asia, English is a compulsory subject in primary school and is gradually being introduced earlier and earlier into the curriculum. For example, in China it is now ofcially introduced at primary 3, but many schools, especially those in urban areas, teach English from primary 1. In some countries it is even a medium of instruction in primary school. This is currently the case in Malaysia and the Philippines, for example, where maths and science are taught through English from primary 1, although both countries have recently announced changes in this regard, which will be discussed below. Brunei primary schools will teach maths and science through English from 2011. In Singapore schools, English is the medium of instruction for all subjects, other than foreign languages (Kirkpatrick 2010). At the tertiary level, the number of programmes offered through English is increasing across Asia, especially, but not exclusively, at the postgraduate level. This is partly explained by the desire of universities to internationalize, and, as I shall argue below, internationalization can often mean Englishization. In the next section, the situation with regard to language education and the role(s) of English in primary schools and the consequences of this will be considered. I shall then consider the role of English as a medium of instruction at the tertiary level and conclude with recommendations designed to help maintain multilingualism at both the primary and tertiary levels.

2. English in Primary Schools I shall use, as a starting point, the situation in the ten countries which make up the Association of Southeast Asian Nations (ASEAN). These are, in alphabetical order, Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, the Philippines, Singapore, Thailand and Vietnam. Of these countries, ve (Brunei, Malaysia, Myanmar, The Philippines and Singapore) were once colonies of English speaking nations, three (Cambodia, Laos and Vietnam) were part of the French colonial enterprise, Indonesia was a Dutch colony and only Thailand escaped colonisation. These varied colonial histories have had an effect upon

English as a medium of instruction in Asian education (from primary to tertiary)


the role of English in each, with the former colonies of Great Britain and the United States retaining a strong institutional role for English (with the exception of Myanmar). However, English has replaced French and Dutch as the rst foreign language in the school systems of the other ex-colonial nations and English is also the rst foreign language taught in the Thai school system. I have described the role of English in ASEAN in detail elsewhere (Kirkpatrick 2010), so here briey mention that the ASEAN Charter, which was signed in February 2009, gives English a privileged status, making it the sole ofcial working language of the group. This is one explanation why countries such as Cambodia and Vietnam have adopted English in place of French as their rst foreign language. Other reasons for the privileging of English are familiar and include the desire of nations to participate in and benet from internationalization and the knowledge economy (Rappa and Wee 2006). In addition to seeking to participate in internationalization, nine of the ten countries of ASEAN achieved independence only relatively recently and are thus also each seeking to establish a distinct national identity. The adoption and promotion of a national language is a natural and effective way of doing this. This explains why, in Indonesia and the Philippines to take just two examples, the respective governments have expended so many resources into establishing their national languages, Bahasa Indonesia and Filipino. Both countries are richly multilingual. Some seven hundred languages are spoken in Indonesia (Hadisantosa 2010) and more than one hundred in the Philippines (Galang 2000). Both have national language institutes (the Pusat Pembinaan dan Pengembanan Bahasa in Indonesia and the Komyson sa Wakang Filipino in the Philippines), one of whose tasks has been to promote and modernize the national languages (Alisjahbana 1976; Gonzalez 2007). In this, Indonesia has probably been more successful than the Philippines. Reasons for the relative success of the Indonesian peoples acceptance of Bahasa Indonesia include the fact that it is based on Malay, a language which was spoken by only a tiny (3%) minority of the population, and thus was not seen to privilege a powerful group. Indeed this was one reason why Javanese, the language of the powerful and spoken by 75 million people, was not chosen to be the national language. In contrast, Filipino is based on Tagalog, the language spoken in and around the capital, Manila. This explains why many Filipinos are still not comfortable with the idea of a Tagalog-based language as the national language. For example, many Cebuano speakers who themselves outnumber Tagalog speakers prefer to speak English rather than Filipino as a language of national communication. This may be more a question of attitude than linguistic prociency. As the Filipino scholar Tupas, himself a speaker of two Visayan languages, Aklanon and Ilongo, points out, Filipino is the national language and it is now more of an issue of atti-


Andy Kirkpatrick

tudes, not inability, to speak the national language (2007:32). One measure of the success of the national language is based on the number of speakers. The table below (adapted from Montolalu and Suryadinata 2007: 48) shows, over a period of twenty years, the shifting percentages of Indonesians reporting that they are rst language speakers of particular languages.
Table 1. Vernacular Language Speakers in 1980, 1990, 2000 Language Javanese Sundanese Madurese Batak Minangkabau Balinese Buginese Indonesian (BI ) Others 1980 40.44% 15.06% 4.71% 2.12% 2.42% 1.69% 2.26% 11.93% 17.48% 1990 38.08% 15.26% 4.29% 1.97% 2.23% 1.64% 2.04% 17.11% 17.11% 2000 34.70% 13.86% 3.78% 1.91% 2.06% 1.42% 1.91% 34.00% 4.57%

The gures for the seven local languages listed here appear relatively stable. What is remarkable is the increase in those who reported that they consider themselves rst language speakers of Bahasa Indonesia (from 11.93 percent in 1980 to 34 percent in 2000) and the drop from 17.48 percent in 1980 to 4.57 percent of those who consider themselves speakers of vernacular languages other than the ones listed. This would suggest that many of the 700 or more languages of Indonesia are under increasing threat. We can probably safely assume that the relative increase in rst language speakers of BI and decrease of speakers of vernaculars will have been maintained over the most recent decade. Note that the gures for BI represent those who see themselves as rst language speakers. This increase in rst language speakers of BI can be attributed to social and geographical mobility, and the increase in inter-ethnic marriage associated with this. Children of mixed marriages who nd themselves living in urban areas will tend to develop BI as a rst language, especially as this is the language of education. The success of BI as a national language has done much to establish a sense of pan-Indonesian identity among the population. On the other hand, its very success as a national language has had a negative inuence upon many local languages. This negative inuence is greatly exacerbated by the adoption of English as the rst second language taught in the school system. Apart from anything else, the inclusion of English into the primary curriculum is always at the expense of another subject. A local language is the most commonly sacri-

English as a medium of instruction in Asian education (from primary to tertiary)


ced subject. As Hadisantosa reports in her evaluation of the new SBI schools . . . with (the) emerging and mushrooming demand for English, schools then drop the local language in order to give more time to the English teaching. As a result, in the long run, children and the younger generation can no longer speak the local language. This is culturally and linguistically pitiful (2010: 31). Remembering that Indonesia remains the sole nation which does not make English a compulsory subject, this pattern is replicated in even starker relief throughout Asia. For example, Vietnam has recently embarked on its National 2020 English Initiatives under which all primary school graduates are to achieve a level of English equivalent to the Common European Framework of Reference (CEFR) scales of A1 or A2 and all college graduates should have a working knowledge of English. (Hung and Duzdik 2010). While it is encouraging to see some move away from the more traditional Anglo-based testing tools, such as TOEFL or IELTS in the adoption of the CEFR scales (the plan is to adapt these to the Vietnamese context, but this has not yet taken place), the programme remains extremely ambitious. It will require the re-training of 62,000 secondary English teachers alone. An indication of the challenge lies in the fact that of the 250 EFL teachers tested in a trial project (and where the test used was the TOEFL) only 28 scored more than 500 (the very minimum required by some Englishmedium universities) and more than 50 scored less than 350 (a level equating to basic literacy only). One reason, but not the only one, for the recommendation below that the primary school should focus on local languages is that there simply are not enough primary level teachers with adequate English prociency to be able to teach English. Consider, for example, the scale of the problem in China, where English is a compulsory subject from primary 3. It is impossible to estimate the number of primary schools in China, but the assumption that there are enough qualied and linguistically procient English teachers to be able to provide quality English language teaching in all of them is unrealistic. If there are insufcient teachers to teach English as a subject in primary schools, there is an even more serious shortage of subject-specialist teachers who are expected to teach maths and science through the medium of English. At present, in addition to Singapore, both Malaysia and the Philippines expect primary 1 children to learn maths and science through English. The problems associated with this have recently been recognized by both governments. Malaysia has announced a radical change of policy and that maths and science will be taught through the national language, Malay, from 2012. Reasons for this change include the realization that many children from the lower socio-economic classes and rural areas were failing these subjects and that there were not enough qualied teachers with adequate prociency in English to teach them. The then Malaysian Deputy Prime Minister is reported in the Star newspaper of July 7


Andy Kirkpatrick

2009 saying that Only 19.2 percent of secondary teachers and 9.96 percent of primary teachers were sufciently procient in English. Two quotes from the debate over the medium of instruction controversy, recorded in the New Straits Times of 21 March 2009, help illustrate the different points of view. The rst comes from a Chinese Malaysian, who remarks:
They (the students) cant even understand English; how can you make them study Science and Maths in English?

The second quote comes from an opponent to the rst speaker, an ethnic Malay who nevertheless defends the use of English over Malay, arguing that, unlike Mandarin and English, Malay (Bahasa) is not an international language, which is why the policy of using English as a medium of instruction (PPSMI) needs to stay:
Mandarin is an international language, thats where the Malays are at a disadvantage. Because Bahasa is not an international language, thats why we are ghting and want PPSMI to be retained, because its an advantage to the Malays.

The argument that, as English is the language associated with modernization, it needs to be used to teach subjects like maths and science, is common. But, as I think most readers will agree, science and maths are cognitively demanding subjects and are therefore best taught in the learners rst language. As Bernardo has eloquently argued in proposing the learning of maths and science through local languages in the Philippines, there seems to be no theoretical or empirical basis. . . to obligate the use of English in teaching mathematics and that, there are clear and consistent advantages to using the students rst language. . . at the stage of learning where the student is acquiring the basic understanding of the various mathematical concepts and procedures (2000:13). Many researchers support this position in arguing that cognitively complex subjects should not be taught through a foreign language until the learner has achieved a certain level of prociency in the language. Some three decades ago, Cummins (1981) showed that migrant children to Canada required between 5 and 7 years learning English before they could use it successfully as an academic language. It is worth noting that, in the Canadian context, the children had much more access to English than students in Asian schools. Benson has similarly suggested that being taught an academic content through the L2 represents a multiple burden for the learner (2008:2) Many Filipino children are faced with this multiple burden. The Philippines has, since 1975, implemented a bilingual education policy (BEP), through which children learn maths and science through English, and other subjects through Filipino. But as reported above, Filipino is, in effect, Tagalog, the language of

English as a medium of instruction in Asian education (from primary to tertiary)


Manila, by another name. This means that a child from the Cebu region, whose rst language might be Bohol and whose second might be Cebuano, arrives in primary 1 having to learn in two new languages. This, along with research that showed the efcacy of mother tongue education (e.g., Walker and Dekker 2008), is why the Philippines government has recently announced that the use of the vernacular as a medium of instruction will be allowed from primary 1 to primary 3 in certain circumstances. This is to be welcomed, although most experts would argue that three years is not enough at least ve years is required (cf SkutnabbKangas et al. 2009) and the new policy is too new for any evaluation of it to be made. The only place in East and Southeast Asia where it is ofcial policy to use the local language rather than the national language as the medium of instruction in primary education is Hong Kong. This is because, in contrast to the rest of China, Hong Kong can decide on the medium of instruction. In China itself, the national language law is clear in prescribing the national language, Putonghua, as the sole medium of instruction, although national minorities are allowed to use their respective languages as an MoI in primary schools (Kirkpatrick and Xu 2001). This has meant that Hong Kong has adopted Cantonese, the mother tongue of the overwhelming majority of the population, as the medium of instruction in primary schools. However, the increasing inuence of China coupled with an increase in migration from China to Hong Kong, has meant that there is now pressure for Putonghua to become a medium of instruction, and some schools are currently trialling the teaching of Chinese subjects through Putonghua rather than Cantonese. English is also exerting increasing pressure in this regard. I shall consider the inuence on this of the majority of Hong Kongs universities only offering English medium education in the next section, but, not unnaturally, the universities language policies see parents demanding English medium education, especially at the secondary school level. This has led to the current government ne-tuning the existing secondary school policy which only allowed English medium education in about 25 percent of all secondary schools, the remainder being Chinese medium. This was an unfortunate policy as, apart form anything else, it led to Chinese medium schools being seen by parents as less prestigious than English medium ones. However, instead of insisting that all schools become multilingual sites, the governments new ne tuning policy means that the current Chinese medium schools which can meet certain criteria connected with student and staff quality, their linguistic prociency and the availability of resources, will be allowed to teach subjects and classes through English. Our research (Kan et al. 2010) shows that the immediate result of this will be an increase from September 2010 in the number of lower secondary maths and science classes to the tune of six hundred being taught through


Andy Kirkpatrick

the medium of English. This seems short-sighted in the extreme. Students who had previously been successful in their learning of maths and science through Chinese will now have to learn these subjects through English. Teachers who successfully taught these subjects through Chinese will now have to teach them through English. And why? To satisfy parental demand for English medium classes. In this section I have provided a necessarily selective review of the situation with regard to the teaching of English in primary schools across East and Southeast Asia. What is common across all systems is that the perceived need for a nations citizens to learn the national language and English as the international language has led to an increasingly early introduction of English into the primary curriculum, sometimes as a medium of instruction, and almost always at the expense of local languages. This has resulted in a number of unfortunate consequences. Not only is the early introduction of English likely to fail, as there simply arent enough qualied and linguistically procient teachers available either to teach it as a subject or to teach subjects through it, it also means that there is no place for the childrens mother tongue or other local languages in the primary curriculum. It is hard to overestimate the damage that this seemingly far-sighted policy of modernization and internationalization is inicting upon the lives of children and upon local languages. The drop-out rates of children, especially around primary 5, are alarming. The table below is adapted from the UNESCO report entitled Education for All by 2015. It identies a number of challenges facing certain ASEAN countries. It is noteworthy that the recommendations made by UNESCO to help overcome these challenges all require an increase in multilingual or multicultural education.

Table 2. Education for All: Challenges and Measures Country Cambodia Indonesia Laos Philippines Vietnam Main challenges reduce low levels of develop bilingual education reduce low levels of survival to grade 5 reduce low levels of survival to grade 5 reduce low levels of survival to grade 5 decrease number of children out of school Measures to help survival to grade 5 add mother tongue education for early grades develop multilingual materials make curriculum exible to allow for cultural diversity develop bilingual education for ethnic minorities

English as a medium of instruction in Asian education (from primary to tertiary)


With regard to the problems surrounding the maintenance of local languages inherent in language policies that focus on the national language and English, a recent UNESCO report, Safeguarding Endangered Languages, predicts that half of the worlds languages are under threat ( ich/index.php?lg=en&pg=00136). These consequences are all the more tragic as there is now a large amount of research which shows that not only does mother tongue and multilingual education (MLE) benet the children, including, most importantly in this context, their ability to learn other languages such as English later, but the adoption of MLE naturally revitalizes local languages (Cenoz 2009, Skutnabb-Kangas et al 2009, Benson 2008). In other words, using local languages in the primary school provides a win-win result. Children do better in all subjects and local languages are maintained. I return to this in the nal section of this article, but here turn to consider the role(s) of local languages and of English and their consequences for tertiary education in East and Southeast Asia.

3. English in Tertiary Education 3.1. The European Experience As Asia has adopted trends rst seen in Europe, this section begins with a brief review of the current situation in Europe. The aim of a study commissioned by the Academic Cooperation Association (ACA) was to identify the number of programmes which were being taught in English at universities across continental Europe and to compare the results with an earlier study undertaken by the same authors (Wachter and Maiworm 2008; Maiworm and Wachter 2002). The 2008 study identied 2,400 English-medium programmes, with programmes being dened as full degree courses at both Batchelors and Masters levels which were taught entirely in English. Subject specialist courses in English were excluded. The study also found that the majority of these programmes were being offered in northern Europe, with 774 in The Netherlands, 415 in Germany, 235 in Finland and 123 in Sweden, with the majority of the programmes being postgraduate. The Swedish gures have been challenged by Phillipson (2009) who points out that data from the Swedish Ministry of Education show that as many as 480 of the 680 MA degrees available in Sweden are taught in English. The numbers provided in the 2008 study dont give the overall percentage of programmes that are being taught in English, but they do indicate a substantial increase in the number of programmes taught in English identied in the earlier 2002 study which found only 700 of these. This increase from 700 to 2400 has led


Andy Kirkpatrick

one scholar to suggest that, it seems inevitable that English, in some form, will denitely become the language of education (Coleman 2006:11). This increase in English-medium programmes is one consequence of the Bologna Declaration of 1999 which established a European Higher Education Area (EHEA) in which the forty six member countries would undertake to reform the structures of their higher education systems in a convergent way ( policies/educ/bologna/bologna.pdf; accessed 9 October 2010). This convergence and the move to the internationalization of higher education has encouraged the use of English as a common medium, leading Phillipson to argue that, What emerges unambiguously is that in the Bologna process, internationalization means English-medium higher education (2009:37). The possibility of being swamped by English-medium programmes has not gone unnoticed and many countries have put in place language policies designed to ensure that local languages remain as the media of instruction. For example, the 2006 Declaration on a Nordic Language Policy was signed by the ve Nordic countries with the aim of implementing bilingual language polices in higher education ( 2007-746; accessed 9 October 2010). It is also important to point out that the number of English-medium programmes on offer in southern Europe, while increasing, remains relatively small. I now turn to consider the situation in East and Southeast Asia. 3.2. East and Southeast Asia In Asia, the internationalization of higher education has been primarily characterized by Asian students travelling to western countries to obtain degrees. Over 90 percent of the 2 million international students are shared between the United States, Britain, Australia, Germany and France (Howe 2009:384). The majority of these students come from Asia. This has proved an extremely lucrative business for the providing institutions. At the same time, the countries from which these students come experience two forms of loss. First, they lose hard currency, as the international students need to pay fees and for their cost of living overseas. Second, they also experience a brain drain, as many international students choose to remain in their host countries after graduation. It is not surprising then that many Asian countries are now trying to set up local education hubs in order to attract fee-paying students on the one hand and to offer an alternative to their own students on the other. Malaysia has been relatively successful in this (Gill 2004). There are now over one hundred colleges which have some form of twinning arrangements with overseas universities. Typically these offer the rst two or three years of a degree course in Malaysia, with the

English as a medium of instruction in Asian education (from primary to tertiary)


student only having to travel to the host country for the nal one or two years of their programme. Some overseas universities have established campuses in Malaysia, where the students can undertake their entire degree. Malaysia itself, however, has set up its own private universities there are now twelve. These include those set up by public companies, such as the Petronas University of Technology, established in 1997 by the national oil company to offer science and engineering courses. The medium of instruction is English and the universitys journal, Platform, publishes articles in English. Another is Tunku Abdul Rahman University, which offers an English-medium education to students who can afford the fees. An indication of how widespread the increase in English-medium education is comes from the fact that some Malaysian medical students are now choosing to do their studies in a number of Russian universities (including the Moscow Medical Academy, the Volgograd State Medical Academy and Nizhny Novgorod State Medical Academy) who now offer these degrees through the medium of English, but at around a quarter of the cost of comparable degrees offered by Australian and British universities. The competition is no longer solely restricted to the traditional providers. Japan has something of a reputation for being reluctant to diversify and internationalize, other than to try to explain to the world what it means to be Japanese. It remains a highly ethnocentric and gendered society (Howe 2009:386) and ranks last in the list of OECD nations with regard to the number of foreigners (and women) in higher education. Yet, the President of the nations top-ranked university, the University of Tokyo, has announced that, universities have to internationalize for the sake of diversity . . . People who are part of the same culture and language can no longer really develop intellectually (McNeill 2007 cited in Howe 2009:387). The extent to which cultural and intellectual diversity is the overriding motivation in the push to internationalize can be questioned, however. Although Knight (2008) has listed four potential motivations for universities to internationalize namely, political, economic, cultural and academic as we have seen, the nancial motivation is very strong and, in the case of private universities in Japan, international students are seen as a way of lling the places left empty by a falling Japanese population of the relevant age group. Waseda, a prestigious private university in Tokyo, will, beginning in the academic year 20102011, offer nine English-medium degrees, some at BA and others at MA levels, in subjects such as politics, economics, science and engineering. Waseda is one of the thirty Japanese universities participating in a nationwide internationalization project ( index.html, accessed 11 October 2010).


Andy Kirkpatrick

China is the major provider of international students, and it is not surprising that it is also moving to increase the number of English-medium courses. As mentioned earlier, the National Language Law of China limits the medium of instruction in schools to the national language, Putonghua (Mandarin), with the exception of certain national minority languages. Parental demand for Englishmedium courses is high however, and some secondary schools, especially those in wealthy urban settings such as Shanghai and Xian, are experimenting with teaching content subjects through the medium of English. They have to refer to these as some form of special English courses in order not to be seen to be breaking the law. In these classes, content subjects are taught by English teachers, with less than satisfactory results (Hoare 2010). At tertiary level, on the other hand, no such restriction exists. Indeed universities have been encouraged by the leadership to consider offering English-medium programmes. Zhu Rongji, the then Premier of China, visited the School of Economics and Management of Beijings prestigious Tsinghua University in 2001 and announced that, I hope all classes will be taught in English. I dont worship foreign languages, but we need to exchange our ideas with the rest of the world (China Varsities to Teach in English). Chinas University and College Admission System (CUCAS) (http://www., accessed October 11, 2010) provides a list of Chinese universities which offer degree programmes through English. These include more than thirty which offer some form of English-medium business degree and some forty which offer medical degrees in western medicine through English. Some also offer courses in aspects of traditional Chinese medicine through English, an issue to which I will return. I will now discuss the situation in Hong Kong in some detail, as it provides an example of where English-medium education dominates the tertiary sector, at the expense of Chinese and at the expense of the governments own language policy. The Hong Kong governments language policy aims to see its citizens trilingual (in Cantonese, English and Putonghua) and biliterate (in Chinese and English). In order to achieve this laudable aim, Cantonese is the medium of instruction in most government primary schools. As pointed out earlier, this makes Hong Kong the only place in Asia where the national language or English is not a main medium of instruction at the primary level. However, the governments trilingual-biliterate policy is undermined by the language policies adopted by the university sector. Six of Hong Kongs government-funded universities are English-medium. Only one, the Chinese University of Hong Kong, is ofcially bilingual and only one, The Institute of Education, is ofcially trilingual. Not surprisingly, therefore, parents are desperately keen to ensure that their children study through English at the secondary level. Those that can afford it, send their

English as a medium of instruction in Asian education (from primary to tertiary)


children to the schools of the English Schools Foundation (ESF), which were originally established to cater for the children of expatriates. This tendency to enrol ones children in private English-medium schools can be seen throughout Asia. To actually forsake the public school system that teaches in your own language for the private one that teaches in English is an increasingly common phenomenon (Wang 2007: xiv). Before the handover of Hong Kong back to China, the secondary school language policy had been left to the schools themselves. As a result, almost all offered or claimed to offer English medium education (Johnson 1994). In effect, however, many of these schools taught through what has been called a textual explanation approach (Luke and Richards 1982:50) by which subject teachers spent the majority of class-time translating the English language textbooks into Cantonese. I myself observed classes of this sort in the late 1970s, where teachers would translate the textbooks and children scrawl down the Chinese characters in the margins. The policy of allowing secondary schools to choose their own medium of education was altered just before the 1997 handover. Under the new policy, only schools which met a set of criteria based on student ability, the linguistic prociency of teachers and the availability of certain resources could set themselves up as English medium schools. This led to only 114 out of a total of 460 secondary schools being English-medium. The remainder were to teach through Chinese. This new policy was not popular with parents. As indicated earlier, the fact that the policy led parents to perceive that the Chinese-medium schools were academically inferior to the English-medium ones was one cause. A second was that parents still wanted an English-medium education, as the universities taught through English (Kirkpatrick and Chau 2008). After years of campaigning, the government relented and agreed to ne-tune the policy. The ne-tuning meant that Chinese-medium schools which could satisfy the criteria outlined above could offer classes through English. The immediate result of this, as reported earlier, has been a signicant increase in the number of subject courses, especially in maths and science, that are now being taught in English. This is potentially damaging for these students, as recent research ndings indicate that students in Chinese-medium schools who take their nal school exams in Chinese fare better in all subjects, with the exception of English, than those who have switched to study these subjects in English (Hong Kong Institute of Educational Research, 2005). In concluding this section, the adoption of English-medium education by six of Hong Kongs eight government-funded universities appears short-sighted in the extreme. Hong Kong has the opportunity, personnel and resources to internationalize education in a truly bilateral way. Hong Kong, despite the current


Andy Kirkpatrick

governments attempt to market it as Asias World City is, in essence, a Chinese city. It has the opportunity to showcase Chinese scholarship, culture and languages. We should move beyond the so-called established order dominated by the Anglo-Saxon paradigms and instead develop systems and standards that could preserve national heritage and promote rich cultural traditions (Mok 2007:447). Hong Kong could be leading the way in developing multilateral international higher education. This would include promoting local scholarship and knowledge and it is this issue to which I now turn.

4. The Anglicization of Scholarship As well as increasingly becoming the international language of instruction and assessment in higher education, English has long been dominant as the language of international scholarship. English is by far the most important language of scientic and scholarly conferences (Ammon 1996:26). Ammon (1996) also reports that the European Science Foundations working language is English as is the language of the Foundations journal, Communication. More than 90% of the information in the Science Citation Index (SCI) is extracted from journals in English taken mostly from English language journals (Truchot 2002:10). To show just how widespread publication in English has now become, even the journal Association International de Linguistique Appliqu has, despite its e French title, only published articles in English since 2003. The move towards publications in English, has already reduced multilingualism in the eld, and may eliminate the status of any other language as an international language of science (Hamel 2007:66). The need to publish in English is exacerbated by the way universities in Asia reward and recognize the publications of their academic staff. It is common to nd universities in the region insisting that staff publish only in journals listed in the Science Citation Index (SCI) or the Social Science Citation Index (SSCI). This almost inevitably requires publication in English. It also undermines the status and role of local journals and publishing in local languages. Generally speaking, the reward for publishing in English in approved journals is far greater than publishing in local languages in local journals. For example, Chinese academics earn much more from publishing in English in SCI journals than they do for publishing in Chinese in local languages, even though the local language journal might be a far more appropriate place for their publication and where it is likely to be read by far more people than in the SCI journal. This is a point which needs further elaboration. It is commonly assumed that publishing in English in international journals assures the author of a wide readership. This is

English as a medium of instruction in Asian education (from primary to tertiary)


simply not the case. Countless articles in countless journals go unread, as many academics checking their citation indexes can attest! In contrast, however, an article published in Chinese in a local journal can be assured of a readership of thousands, if not hundreds of thousands. It is also much more likely to have real impact. The privileging of English in this way may also be undermining the status and prestige attached to knowledge written in languages other than English. For example, the requirement that doctoral theses include primary source material seems to have been discarded. I have noted that, in my own eld of applied linguistics, students from Mainland China writing PhD theses in universities in Australia, the UK or the US seldom make reference to scholarship written in Chinese, even though there may be a rich tradition of Chinese language scholarship on the topic. Whether this is because their supervisors or possible examiners are not able to read Chinese or whether they are not advised or encouraged to make use of the scholarship written in their own language, I am not sure. But either way, it is an alarming trend and one which appears to be caused by the privileged position of English as a language of scholarship. The requirement to disseminate knowledge through English may also radically alter the nature of the knowledge being disseminated. For example, a study on the transmission of traditional Chinese medicine (TCM) through English found that TCM should more appropriately be called modern Chinese medicine (MCM) (Kirkpatrick 2009). TCM is characterized by ambiguity and diversity (Hsu 1999, Scheid 2002), characteristics that have been excised from what now passes as TCM. A brief look at the problems involved in the translation of an apparently simple but fundamental concept of TCM, qi, also illustrates the problems associated with translating fundamental concepts of indigenous knowledge into English. Translations of qi include:
that which makes things happen in stuff; stuff in which things happen; energetic conguration; a conguration of energy; (nest matter) inuences; emanations; vapours

(Scheid 2002:48).

The increasing role of English in higher education is not, therefore, simply limited to its use as a medium of instruction and assessment. It is also increasingly used as the international language of scholarship and the dissemination of knowledge. This further threatens the status and role of local languages and knowledge written in languages other than English. In the conclusion, I draw


Andy Kirkpatrick

some recommendations that may help revitalize the place of local languages and knowledge.

5. Conclusion and Recommendations This article has described and illustrated the growth of the role and use of English in education in Asia, with the situation at primary and tertiary level being given prominence. It has shown that throughout Asia, English is being introduced earlier and earlier into the primary curriculum, sometimes as a medium of instruction and always at the expense of local languages. This, combined with the emphasis placed on the need for children to learn their respective national language, means that children are often unable to learn through their mother tongue or other local languages at great sacrice to their own personal development and the vitality of local languages. One cause of the increasing demand for English at the primary level, is its perceived role as the language of internationalization and modernizaton and its associated role as the lingua franca of the internationalization of higher education. This, in turn, threatens the role of languages other than English as languages of education, scholarship and the dissemination of knowledge. In this nal section, I shall make some recommendations which might help the revitalization of local languages at both primary and tertiary levels. 5.1. Recommendations for the Primary Level The major recommendation for primary education in Asia is that the primary school should focus on local languages. This will, no doubt, raise an outcry among parents who feel that they want their children to learn English and that the earlier they learn it the better. All things considered, it is true that young children nd it easy to learn languages. But the conditions and resources need to be adequate, the teachers need to be qualied and linguistically procient and the language to be learned needs to play an identiable role in the speech community. For the vast majority of Asian primary school children, few, if any, of these conditions are met. Instead they are taught English by teachers whose own level of English is low, who may not be trained and in schools that are inadequately resourced and where there is a dire shortage of appropriate language teaching materials. The result is that these children are condemned to failure. The fact that English has usually displaced the local language from the classroom further means that they are deprived of the opportunity of learning their own mother tongue, again at great sacrice to their personal and cogni-

English as a medium of instruction in Asian education (from primary to tertiary)


tive development. This is why I argue that it is so essential for the primary school to invest in local languages. As many scholars of multilingual education have demonstrated (cf, for example Cummins 1981; Benson 2008; Cenoz 2009 Garcia 2009; Skutnabb-Kangas et al 2009), providing primary school children education in their own mother tongue sees them perform well in all subjects. Fluency in the mother tongue also allows them to build bridges to uency and literacy in other languages. That is to say, uency in the mother tongue and the national language will help the children in their later learning of English. This can be left until secondary school, where the likelihood of there being linguistically procient qualied teachers, adequate resources and appropriate teaching materials is far greater. In this context, it is important to point out that, in Europe, the current policy is that children should learn their rst language + two. While one of the two is almost always English, the policy is aimed to ensure that all children grow up trilingual. No such policy exists in Asia. 5.2. Recommendations for the Tertiary Level The rst recommendation for universities in Asia is that they develop and implement bilingual language policies which are designed to revitalize the local language as a language of education. The Nordic countries offer possible models. For example, Sweden and Denmark have set targets of parallel competence in English and Swedish/Danish (Phillipson, in press). Preisler has argued for a system of complementary languages in tertiary education whereby the two languages will be functionally distributed within the individual programme according to the nature of its components (2009:26). In the context of Hong Kong, this means that the six universities which currently operate an English-medium only language policy would be required to introduce bilingual policies. The second recommendation is for universities in Asia to work together to establish new centres for journal publishing. There is no reason why Asian universities could not combine resources to set up bilingual journals, publishing in the respective national language and English. At the same time, they need to promote local scholarship and knowledge by publishing in local languages and, if need be, to provide translations of these publications in English. In summary, the recommendations can be very simply stated in the following way:
that the primary school focus on local languages; that the tertiary sector implement bilingual policies for teaching and publishing.

In this way, local languages can be revitalized at the primary level, while allowing children the opportunity to learn through the local languages to their great


Andy Kirkpatrick

benet, sense of identity and cognitive development. Local languages can be used to disseminate local knowledge andAsian universities can create conditions for higher education to become multilateral and multilingual. This would allow local languages and English to play complementary roles in education.

Alisjahbana, Takdir S. 1976. Language planning and modernisation. The case of Indonesian and Malaysian. The Hague: Mouton. Ammon, Ulrich. 1996. The European Union. Status Change During the Last 50 Years, In Joshua Fishman, Andrew Conrad & Alma Rubal-Lopez (eds.) Post-imperial English, 241267. Berlin: Mouton de Gruyter.. Benson, Carol. 2008. Summary overview. Mother tongue-based education in multilingual contexts. In Caroline Haddad (ed.), Improving the quality of mother tonguebased literacy and learning. Case studies from Asia, Africa and South America, 211. Bangkok: UNESCO. Bernardo,Allan B. I. 2000. The multifarious effects of language on mathematical learning and performance among bilingual/bilinguals?: a cognitive science perspective. In Ma Lourdes Bautista, Teodoro A. Llamzon & Bonifacio P. Sibayan (eds.), Parangalcang Brother Andrew: a festschrift for Andrew Gonzalez on his sixtieth birthday, 303316. Manila: Linguistic Society of the Philippines. Cenoz, Jasone. 2009. Towards Multilingual Education. Bristol: Multilingual Matters. China varsities to teach in English. 2001, September 20. Reported in the South China Morning Post and cited in Web Pages of the Straits Times Interactive, http://straits (accessed August 15 2009). Coleman, James A. 2006. English-medium teaching in European higher education, Language Teaching, 39.114. Cummins, Jim. 1981. Age on arrival and second language learning in Canada: a reassessment. Applied Linguistics 1.13249. Galang, Rosita G. 2000. Language planning in Philippine education in the 21st century: toward language-as-resource orientation. In Ma Lourdes Bautista, Teodoro A. Llamzon & Bonifacio P. Sibayan (eds.), Parangalcang Brother Andrew: a festschrift for Andrew Gonzalez on his sixtieth birthday, 267276. Manila: Linguistic Society of the Philippines. Garcia, Ofelia. 2009. Bilingual education in the 21st century: A global perspective. Malden, MA: Wiley-Blackwell. Gill, Saran Kaur. 2004. Medium-of-instruction policy in higher education in Malaysia: nationalism versus internationalization, in James Tollefson and Amy Tsui (eds.), Medium of instruction policies Which agenda? Whose agenda? 135152. Mahwah, New Jersey: Lawrence Erlbaum.

English as a medium of instruction in Asian education (from primary to tertiary)


Gonzalez, Andrew. 2007 Language, nation and development in the Philippines. In Lee Hock Guan & Leo Suryadinata (eds.), Language nation and development, 716. Singapore: Institute of Southeast Asian Studies. Hadisantosa, Nilawati. 2010. Insights from Indonesia. In Richard Johnstone (ed.), Learning through English: policies, challenges and prospects, 2446. London: British Council. Hamel, Rainer Enrique. 2007. The dominance of English in the international scientic periodical literature and the future of language use in science.AILA Review 20. 5371. Hoare, Philip. 2010. Content-based language teaching in China. Contextual inuences on implementation. Journal of Multilingual and Multicultural Development 31,1. 6986. Hong Kong Institute of Educational Research, The Chinese University of Hong Kong. 2005. Further evaluation on the implementation of the medium of instruction guidance for secondary schools (20022004): executive summary. FileManager/EN/Content 1914/es14 further%20eval%20on%20the%20imp%20of %20the%20moi%20(20022004).pdf (accessed October 11 2010). Howe, Edward R. 2009. The internationalization of higher education in East Asia: a comparative ethnographic narrative of Japanese universities, Research in Comparative an International Education, 4, 4.384392. Hsu, Elisabeth. 1999. The transmission of Chinese medicine. Cambridge: Cambridge University Press. Hung, Nguyen Ngoc & Duzdik, Diana L. 2010. A call for collaboration: Vietnams National English 2020 Initiatives. Paper given at the International Conference on Innovations in ELT, SEAMEO RETRAC, Ho Chi Minh City, 2324 September. Johnson, Keith. 1994. Language policy and planning in Hong Kong. Annual Review of Applied Linguistics 14. 177199. Kan, Vincent, Lai, K.C., Law, Agnes and Kirkpatrick, Andy. (forthcoming). Fine-tuning Hong Kongs medium of instruction policy. Hong Kong: Research Centre for Language Education and Acquisition in Multilingual Societies, Hong Kong Institute of Education. Kirkpatrick,Andy. 2009. English as the International Language of Scholarship, in F. Sharian (ed.) English as an international language, 254270. Bristol: Multilingual Matters. Kirkpatrick, Andy. 2010. English as a lingua franca in ASEAN: a multilingual model. Hong Kong: Hong Kong University Press. Kirkpatrick, Andy and Chau, Michael. 2008. One Country, two systems, three languages: a proposal for combining Cantonese, Putonghua and English in Hong Kongs schools, Asian Englishes,11, 2. 3245. Kirkpatrick, Andy & Xu, Zhichang. 2001. The new language law of the Peoples Republic of China. Australian Language Matters 9(2). 1418. Maiworm, Friedhelm and Wachter, Bernd. 2002. English-language-taught degree programmes in European higher education. Bonn: Lemmens.


Andy Kirkpatrick

Mok, Ka Ho. 2007. Questing for internationalization of universities in Asia: critical reections, Journal of Studies in International Education, 11, 3/4.433454. Montolalu, Lucy & Suryadinata, Leo 2007. National language and nation-building: the case of Bahasa Indonesia. In Lee Hock Guan & Leo Suryadinata (eds), Language nation and development, 3950. Singapore: Institute of Southeast Asian Studies. Phillipson, Robert. 2009. English in higher education. Panacea or pandemic? Angles on the English-Speaking World, 9. 2957. Phillipson, Robert. In press. Language policy and education in the European Union. In Nancy Hornberger (ed.) Encylopedia of language and education, In Stephen May (ed.) Vol 1 Language policy and political issues in education. Dordrecht: Springer. Preisler, Bent. 2009. Complementary languages: The national language and English as working languages in European universities. Angles on the English-Speaking World, 9.1028. Rappa, Antonio & Wee, Lionel. 2006. Language policy and modernity in Southeast Asia: Malaysia, Singapore the Philippines and Thailand. New York: Springer. Scheid, Volker. 2002. Chinese medicine in contemporary China. Durham: Duke University Press. Skutnabb-Kangas, Tove., Phillipson, Robert., Mohanty Ajit & Panda Minati. 2009. Multilingual education: concepts, goals, needs and expense: English for all or achieving justice. In Tove Skutnabb-Kangas, Robert Phillipson, Ajit Mohanty & Minati Panda (eds.) Social Justice through Multilingual Education, 320344. Bristol: Multilingual Matters. Truchot, Claude. 2002. Key Aspects of the Use of English in Europe. Strasbourg: Council of Europe. Tupas, Ruanni. 2007. Go back to class: the medium of instruction debate in the Philippines. In Lee Hock Guan & Leo Suryadinata (eds.), Language nation and development, 1738. Singapore: Institute of Southeast Asian Studies. UNESCO. 2007. Education for all by 2015. Will we make it? Oxford: Oxford University Press ( (accessed August 15 2010). Wachter, Bernd and Maiworm, Friedhelm. 2008. English-taught programmes in European higher education. Bonn: Lemmens Walker, Stephen and Dekker, Diane. 2008. The Lubuagan mother tongue education experiment (FLC). A report of comparative test results. Manila: Summer Institute of Linguistics International. Wang Gungwu. 2007. Keynote Address. In Lee Hock Guan & Leo Suryadinata (eds.), Language nation and development, ix-xvii. Singapore: Institute of Southeast Asian Studies. Andy Kirkpatrick is Chair Professor of English as an International Language at the Hong Kong Institute of Education (HKIEd) and Director of the Institutes Research Centre into Language Education and Acquisition in Multilingual Societies. His research interests include the development of regional varieties of English, with a particular focus on

English as a medium of instruction in Asian education (from primary to tertiary)


Asian Englishes. He is editor of the Routledge Handbook of World Englishes (2010), and author of English as a Lingua Franca in ASEAN: the Multilingual Model (Hong Kong University Press 2010). Email:

Universities in the Anglophone centre: Sites of multilingualism1


Abstract This article sets out to examine the increasingly complex linguistic ecology of universities in countries in the Anglophone centre. As universities in these settings have responded to operating in a globalised world, recruitment of students and staff who are multilingual and/ or bi-dialects has signicantly increased. However, the diverse and rich linguistic resources that have been brought into the sector are largely ignored or treated as problematic. My intention is to raise linguistic diversity as an issue that needs greater debate and research in these universities, to problematise the monolingual ethos and practices of much of the sector, and to make the case for imagining universities in these settings as sites of multilingualism. This is in the interests of maintaining discourses that represent higher education as in the public good, in which universities have a vital role to play in contributing to the development of pluralistic, multicultural and multilingual societies at national, regional and global levels, in educating critical citizens of the world (Giroux 2004: 17), and in promoting an ethos of personal growth that better represents what humanity might become (Gibbs et al. 2004: 191).

1. Introduction During the past few decades a mass system of postsecondary education has developed practically across the globe. UNESCO (2009) reports dramatic increases in the student population, with numbers rising from 28.5 million in 1970 to 150.6 million in 2007. A key driver in this expansion has been globalisation, that is the observable ongoing process of the increasing and ever-more intensive interconnectedness of communications, events, activities and relationships taking place at the local, national or international level (Block 2006: 3).


Sin Preece

Altbach et al (2009: ii) observe the interconnectedness of globalisation in relation to higher education, commenting on how the sector is being shaped by the increasing integration of national economies, information technology and knowledge networks together with the spread of English and other forces beyond the control of academic institutions. In universities in Anglophone centre settings, this interconnectedness is also shaping the language or linguistic ecology, which I use in this paper to refer to the cohabitation of an array of languages and dialects in the institution, by fostering the conditions for the growth in recruitment of staff and students from culturally diverse populations around the globe. This is particularly the case in countries that have traditionally been regarded as forming the Anglophone centre, here taken to be the English-dominant nations of the United States, the United Kingdom, Anglophone Canada, Australia and New Zealand. As universities in these countries have expanded, their linguistic ecology has evolved to include the linguistic repertoires of both socalled native speakers of English and bi- and multilingual users of English. The linguistic ecology spans a heterogeneous staff-student population from a range of linguistic minority communities in the Anglophone centre as well as a multitude of English-knowing bi- and multilinguals from around the globe who have been attracted to study and work in Anglophone centre institutions. The linguistic repertoires of this population reect diverse social, cultural and educational traditions and, in many cases, the superdiversity (Vertovec 2006) of urbanised communities around the globe. However, universities in the Anglophone centre appear to have taken a largely myopic stance to these developments. There is often a mismatch between the monolingual ethos and ideology of institutions and the linguistic diversity of their staff and students. As Kaplan and Baldauf (2009: 43) point out, the attention of educational institutions is commonly riveted on the national/ ofcial language and perhaps, one or two larger minority languages in the polity. In the case of Anglophone institutions, this riveting of attention has meant a xation on English and a monoglot ideology (Silverstein 1996), in that universities make efforts to preserve the use of standardised varieties of Anglo-American English and use these prestigious varieties along with disciplinary literacy practices as a critical tool (Blommaert 2010: 173) for positioning students and staff as in or outside normalcy (ibid: 165). This has contributed to the linguistic repertoires of bi- and multilingual students being problematised (Ruiz 1984), rather than being treated as multilingual capital (Eversley et al. 2010), and solutions being sought for xing what are deemed to be language deciencies. This problem-solution stance adopts an atomistic approach to linguistic diversity, in which the language and literacy practices of the academic community are compartmentalised and taught discretely, separated both from the subject of

Universities in the Anglophone centre: Sites of multilingualism


study and from the linguistic repertoires of bi- and multilingual students and staff. Research into language and identity in higher education in Anglophone centre countries illustrates that this problematising approach to linguistic diversity often results in bi- and multilingual students being positioned as in need of language remediation and being required to take programmes of English language and academic writing (see e.g. Marshall 2010; Martin 2010; Preece 2009, 2010; Simpson and Cooke 2010). This research illustrates ways in which institutional othering of multilingual students often exacerbates feelings of stigmatisation and encourages an oversimplied view of how to address the language needs of a diverse body of students. While there is clearly a role for specialised English language teaching in universities in the Anglophone centre, studies such as the ones above suggest that provision needs to be underpinned by robust institutional policies on linguistic diversity that are sensitive to the complexities of the linguistic ecology of the institution and informed by bi- and multilingual education. Shohamy (2006: 173) contends that monolingualism is a myth detached from reality that must be recognized as such by educational systems. This paper aims to contribute to the debunking of this myth in higher education by examining ways in which the linguistic ecology of universities in Anglophone centre countries has become more diverse as a result of globalisation. I nd the ecology metaphor generative for this discussion. According to Mhlhusler and Fill (2001: 3), ecological approaches are concerned with examining the diversity of inhabitants of an ecology and nding ways of sustaining this diversity. In this paper, I am concerned with making linguistic diversity visible in Anglophone centre universities and also with increasing its status. Creese and Martin (2003: 1) also observe that ecological perspectives require an exploration of the relationship of languages to each other and to the society in which these languages exist [including] the geographical, socio-economic and cultural conditions in which the speakers of a given language exist, as well as the wider linguistic environment. Here then, I am concerned with taking a more holistic approach to linguistic diversity than is usually the case in the sector, and treating it as a resource to be maintained and utilised, rather than a problem to be eradicated. I start by outlining how the global higher education stage has not only facilitated the spread of English in the higher education sector, but also created the conditions for the diversication of the linguistic ecology in universities in the Anglophone centre. Using the UK as an example, I then look at the linguistic ecology at the macro level by examining how the internationalisation agenda, that is the policies and practices that universities implement in response to oper-


Sin Preece

ating in a global environment (Altbach et al 2009), has facilitated the increased movement of students and staff across national boundaries. Following this, I draw on data from my study of language and identity in higher education in the UK (Preece 2009) to illustrate how the access agenda, that is the policies and practices of universities to increase the recruitment and retention of students from underrepresented groups of the domestic population, has affected the linguistic ecology at the micro level of the classroom. I approach bi- and multilingualism as socially constructed concepts that, as Blackledge and Creese (2010: 6) assert, have different meanings across different spaces and times. There are various denitions of bi- and multilingualism that reect different research traditions (see Kemp 2009 for an overview). In this paper, I follow Grosjeans (2010: 4) denition of bilinguals as those who use two or more languages (or dialects) in their everyday lives to refer to university students who routinely experience their lives in more than one language, and, in most instances, in more than one variety of English. Viewed within the context of universities in the Anglophone centre, this usually involves a prestigious variety of standardised English and a non-prestigious vernacular variety or World English.

2. The Global Higher Education Stage Altbach et al (2009) comment on the ways in which universities are now operating on a global higher education stage. Within a system of global higher education, English has become an exceptionally powerful language. Altbach et al (2009: ii) depict its spread as unprecedented since Latin dominated the academy in medieval Europe. In a similar vein, Crystal (2004: 37) argues that English has become the normal medium of instruction in higher education in many countries including several where the language has no ofcial status. Others take a much more critical stance to the spread of English in the sector. Phillipson (2009: 201) asks whether there is now an assumption that English is the default medium of higher education, hegemonically projected as being normal?. In Phillipsons view, the rise of English as the dominant language of academia is anything but normal. Rather it has been engineered to serve the wider social, political and economic forces of capitalism, in which universities, particularly those in the Anglophone centre, are increasingly run as corporate concerns in the global marketplace. Altbach et al (2009) argue that globalisation has increased the tension between the strongest universities, regarded as centres of higher education due to their international reputation for research and excellence, and those on the

Universities in the Anglophone centre: Sites of multilingualism


periphery. Institutions regarded as centres are mainly located in the Englishdominant countries of the United States, the United Kingdom, Australia, Canada and New Zealand, and some non-Anglophone industrialised nations, such as France and Germany. According to Altbach et al, the centre-periphery binary has been exacerbated by the ways in which universities are competitively ranked on an international scale. Those institutions at the top of the international ranking generally use English for teaching and research, have signicant levels of research funding and offer a wide range of programmes in different disciplines. A look at the world university rankings in 2009, for example, reveals that out of the top 200 universities in the world, 119 were located in the Anglophone centre, with 71 in the United States, 28 in the United Kingdom, 9 in Australia and 8 in Canada (Times Higher Education 2010). This system of ranking allows those universities at the top to occupy powerful positions on the global higher education stage and to exert signicant inuence on attitudes to language and literary practices in the sector. At the same time as globalisation has facilitated the spread of English in higher education, it has also created the conditions for the linguistic ecology in universities in the Anglophone centre to diversify. These conditions have been fostered by the sectors response to operating in a globalised world, in which policies have been implemented to increase both the numbers of international students and the numbers of domestic students from under-represented groups in the national population. These policies are often referred to as the internationalisation and access (widening participation) agendas respectively. As will be discussed, these agendas have impacted on the linguistic ecology in Anglophone centre universities by facilitating the increase in numbers of bi-, multilingual and/ or bi-dialectal users of English. 2.1. The internationalisation agenda Altbach and Knight (2007) refer to the policies and practices that higher education institutions have set in place to deal with operating in a globalised world as internationalisation and comment on how the implementation of the internationalisation agenda is increasing the mobility of academic staff, students, higher education programmes and institutions across the borders of nation states. They argue that internationalisation is compounding existing centre-periphery inequalities between universities (see also Altbach 2004, Friedman 2005, Odin and Mancias 2004, Phillipson 2009). They point out that while internationalisation creates a two-way street (p. 291) between universities in different parts of the world, the process is largely regulated by Anglophone countries, and, to a smaller degree, other major industrialised nations in the EU, such as France and


Sin Preece

Germany. One consequence is that the vast majority of the 2.5 million students who are currently studying overseas have migrated northwards, from countries in Asia, Latin America and Africa, to institutions located in the Anglophone centre; this trend is predicted to continue with the number of international students rising from 2.5 million to around 7 million by 2020 (Altbach et al 2009). Universities in the Anglophone centre largely control this process by regulating admissions to their home institutions and exporting their own programmes of study southwards (via distance learning and off-shore campuses located on the periphery). This state of affairs seems set to continue given the nancial rewards for universities and the current crisis in public funding. The growth in student mobility has impacted on the linguistic ecology in many universities in the Anglophone centre. Taking the UK as an example, in 2008/9, the number of non-UK students in UK higher education institutions rose by 8.2 per cent from 341,790 to 369,970. Non-UK students represented just over 15 per cent of the total student population and this proportion is expected to rise. Nearly 250,000 students came from countries outside the EU, while just over 100,000 were from EU countries other than the UK (Higher Education Statistics Agency 2010). Table 1 illustrates the region of domicile for non-UK students in UK universities and demonstrates that there have been increases in student numbers in UK universities from all regions of the world, with the exception of South America and unknown EU countries. Drawing on Kachrus (1992; 1985) model of World Englishes, many of these students are from the Outer and Expanding Circles. For example, in 20089 Indian students formed the largest cohort in UK universities

Table 1. Region of domicile of non-UK domicile students 2007/8 and 2008/09 in UK Higher Education Institutions (Higher Education Statistics Agency 2010) Region of domicile Other European Union Other Europe Africa Asia Australasia Middle East North America South America Non-European Union unknown Total non-UK domicile 2007/8 112,150 12,070 32,295 137,485 2,285 16,690 22,810 4160 1845 341,790 2008/9 117,660 13,745 35,180 150,755 2,310 19,325 24,610 3590 1800 369,970 % change +4.9% +13.9% +8.9% +9.6% +1% +15.8% +7.9% 13.7% 2.4% + 8.2%

Universities in the Anglophone centre: Sites of multilingualism


from an Outer Circle country, with 34,065 students2 , whereas Chinese students formed the largest contingent of students from a country in the Expanding Circle with 47,035 students3 (Higher Education Statistics Agency 2010). Students from the Outer Circle in UK universities primarily come from postcolonial settings in which English is generally regarded as a second and/ or ofcial language and is used in a variety of domains. It is highly likely that many will have received some or all of their compulsory education in English-medium schooling. In these settings, indigenous varieties of English have evolved that have de-Anglicised Standard British English by incorporating localised or nativised grammatical, phonological, lexical/ idiomatic and discourse features of their own (Jenkins 2003; Platt et al. 1984). On the other hand, students from countries in the Expanding Circle come from settings in which English is more likely to be perceived of as a foreign language that more exactly conforms to the norms for standardised varieties of Anglo-American English. The use of English in this Circle is often portrayed as restricted to international, rather than intranational, domains. Given the global spread of English, however, it is questionable whether the differentiation between the roles of English in the Outer and Expanding Circle countries is still fully viable. Over a decade ago, Crystal (1997: 56) commented that the distinction between English as a second and foreign language had less contemporary relevance than previously and that English was in greater use in some places in the Expanding Circle than in some of its traditional bases in the Outer Circle. This point has been widely taken up and there have been a range of studies into the nativisation of English in countries in the Expanding Circle (see e.g. Berns 2005; Bolton 2002; Lowenberg 2002; Rista-Dema 2008; Seargeant 2005; Velez-Rendon 2003). Lowenberg (2002) also points out that students from countries in the Expanding Circle are increasingly studying English in Outer Circle settings where they are learning the norms of Outer Circle standardised varieties of English. He observes how these Outer Circle norms are inuencing the development of nativised varieties of English in the Expanding Circle. The debate on World Englishes suggests that students from various countries in the Outer and Expanding Circles may well be bringing a diversity of English dialects into universities in the Anglophone centre as part of their bi- and multilingual linguistic repertoires. As shall be discussed, however, localised varieties of English rarely receive a warm welcome in universities in Anglophone centre settings. A similar picture of linguistic diversity emerges from the other EU students (i.e. non-UK) within UK institutions. Many of these students can also be regarded as bi- or multilingual users of English. Many come from European countries with two or more ofcial languages either at the state or regional level.


Sin Preece

Similarly to the literature on the Expanding Circle, there is a body of literature documenting the spread of English in member states of the European Union outside the UK and Ireland (see e.g. Berns 1995; Berns 2005; Lowenberg 2002). This literature illustrates how different national and regional varieties of European English are evolving. This suggests that the cohort of European students in UK universities are bringing diverse linguistic repertoires into the sector that include European varieties of English, the 22 ofcial languages of the EU other than English, the indigenous regional and minority European languages and the heritage languages of immigrant communities in Europe. Finally, it is questionable whether international students in UK universities who are from other Anglophone centre countries are monolingual native speakers of English. Philippson (2009: 202) warns against the discourse of historical amnesia that hides the multilingualism and diversity of English-dominant countries, such as the UK and the USA. Given the high levels of cultural and ethnic diversity in settings in the Anglophone centre, it is perfectly possible that a sizeable proportion of students migrating between Anglophone nations are also bi- or multilinguals. So far, I have discussed how the linguistic ecology in universities in the Anglophone centre has been affected by the increasing recruitment of international students (and EU students in the case of the UK) who are bi- or multilingual users of English. This has been driven by internationalisation policies that are encouraging cross border movement of students, largely from periphery to centre institutions, the majority of which are located in the Anglophone centre. In the following section, I look at the access agenda and discuss how this is affecting the linguistic ecology at the micro level of the classroom. 2.2. The access agenda Allen et al (2005) comment on the attempts that governments in OECD countries have made over the last few decades to improve access to higher education for non-traditional students, that is individuals from underrepresented groups in the domestic population. In the Anglophone centre this agenda has contributed to signicant increases in the numbers of domestic students who are multilingual and/ or bi-dialectal users of English of some sort. This body of students is composed of the rst, second, third and subsequent generations of minority ethnic groups, the African American and Latino communities in the United States and, to a much more limited degree, the indigenous peoples of Australia, the United States, Canada and New Zealand. These students span a range of different socioeconomic backgrounds, from working class families to those from more socially elite and afuent professional backgrounds.

Universities in the Anglophone centre: Sites of multilingualism


While the access agenda has its roots in social justice, globalisation has reframed the debate on access to higher education in neoliberal economic discourses. In this scenario, access to higher education is framed in terms of meeting the needs of industries in the global marketplace by providing a highly educated and skilled workforce. Allen et al (2005: 2) comment that, [h]igh-modernity or late- or post-modernity means that the industrial heartlands of countries such as the UK and USA have been devastated and, in order to compete in an increasingly tough global market, knowledge and professional skill development are important to the future of our societies. In the UK, this economizing of universities (McLean 2006: 45) can be seen by placing universities in the Department for Business, Innovation and Skills (BIS). This Department is composed of ten management groups,4 which not only compete for funding but have the collective mission to [build] a dynamic and competitive UK economy by: creating the conditions for business success; promoting innovation, enterprise and science; and giving everyone the skills and opportunities to succeed. To achieve this it will foster world-class universities and promote an open global economy (Department for Business Innovation and Skills 2010). Universities are expected to [create] a highly-skilled workforce that is capable of establishing an enterprise culture where everyone with talent is inspired to turn ideas into successful enterprises in the free market (ibid). Within the UK, the drive towards skilling the workforce has impacted on the linguistic ecology in universities by encouraging a rise in the numbers of students from non-traditional backgrounds, many of whom are from working class linguistic minority communities. To examine how this is affecting the linguistic ecology at the micro level of the classroom, in the following section I turn to data from my study of language and identity in higher education (Preece 2009).

3. Multilingual classrooms I conducted my study with 93 rst-year undergraduate students from a range of linguistic minority groups while they were taking an academic writing programme at a post-1992 university in London. This programme, on which I was a lecturer, had been established to improve the retention of students from widening participation backgrounds. In this case, the majority came from backgrounds that have traditionally been categorised as working class on the basis of occupation. There was also very little history of higher education among the students parents and elders. Students were enrolled on the programme based on the results of an academic literacy test taken during induction. Most were required to attend the programme rather than opting into it as a free


Sin Preece

choice. The data were gathered over a period of two years and consist of audiorecordings of group work interaction in the classroom and interviews of key participants, eld notes from classroom observations and reections on classroom events, an open-ended questionnaire and information from the student record system. The linguistic repertoires of the participants in the study included English along with a variety of Creoles and heritage languages in use among members of the Caribbean, South Asian and African communities in London. Some participants were migrants who had been born and educated in Outer Circle settings and had arrived in the UK as young adults. A few came from countries in the Expanding Circle where they had studied English as a foreign language at school. These participants had been in Britain for a short period of time, had come with the intention of studying, following which they were likely to return to their home country. However, most were the children of settled minority ethnic communities in the UK, many of South Asian heritage. Generally these participants had received all or the majority of their compulsory schooling in the British state system and were aged 1821. While the institution did not acknowledge the participants diverse linguistic repertoires in any formal or systematic manner, the participants drew attention to these when given opportunities to do so during the research. An example of this comes in the following extract that was audio-recorded in the classroom, in which four participants, Kavi (aged 21), Sita (aged 19), Tano (aged 23) and Hibba (aged 22) are discussing what languages they use at home and with their friends.
K = Kavi, S = Sita, T = Tano, H = Hibba5 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. K: S: K: T: K: T: K: H: K: T: H: K: H: I spe- I can speak Tamil/ sort of yeah right <laughs> can you speak any language? I can speak Tamil wha- what language is that? thats a Sri Lankan language obviously yeah/ I can speak my Ghanaian [language so yeah [Ghanaian language/ he can speak THREE langua[ges [I speak FOUR languages [FOUR languages [oh::: <ironic tone> [Arabic er Urdu/ Hindi right and Punjabi

Universities in the Anglophone centre: Sites of multilingualism 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. K: H: K: H: K: S: H: K: S: T: H: K: can you write/ can you/ can you write them? yeah yeah youre writing them all? yeah I can WRITE in Tamil but not fully uent Im uent in that [I can/ I can [you can read? yeah I can read Im [uent in that [proper proper reading? like if I gave you a newspaper would you read it from cover to cover? yeah/ I can read some articles (xxx) NOT how I read English (classroom data)


The interaction takes on the tone of a sociable sparring contest in which the participants compete with each other for who speaks the most languages and who is the most multiliterate. The interaction opens with Kavi presenting himself as a Tamil speaker, although he mitigates this with the comment sort of (turn 1). Sitas utterance and laughter (turn 2) have a teasing quality. Her ability to tease Kavi about his level of expertise in Tamil comes from their shared heritage as part of the Tamil diaspora in London. Both of their families had sought asylum in the UK as a result of the civil war in Sri Lanka with Kavi arriving in the UK at the age of 11 and Sita being British born. A series of exchanges then follow between Tano and Kavi (turns 37) in which Tano, a mature Ghanaian student who had been the UK for two years at this point, establishes Kavis heritage culture and represents himself as speaking my Ghanaian language. His decision not to name the languages in his linguistic repertoire may have been motivated by his perception of his interlocutors knowledge of Ghana and its linguistic complexities. Nonetheless, Kavi has ascertained something of the linguistic diversity that Tano brings to their group as he announces that Tano speaks three languages (turn 7). The raised volume of three and the tone of this utterance suggest that Tanos multilingualism is to be admired. At this point, Hibba, a British Asian, interrupts (turn 8) in what appears to be a competitive move, to position herself as speaking more languages than Tano, which she then goes on to list (turns 11, 13). The tone of Tanos oh (turn 10) sounds ironic and similarly to Sitas utterance (turn 2) may have been an attempt at playfulness. This is followed by a series of exchanges (turns 1425) in which levels of literacy are compared. The interaction is largely controlled by Kavi and Hibba as they weigh up which of them has the greatest level of expertise in their heritage languages. While Hibba claims to be literate in four languages, Kavi concedes


Sin Preece

that he has a greater command of reading in English than in Tamil. While it seems likely that Hibbas greatest expertise is also in English, her claim to have mastery of four languages may have been a strategy for jockeying for position among her peers. Within this group, this appears to have been successful, as her claims went on to earn her the admiration of the group. Throughout the interaction, the participants adopt a multilingual positioning. Their shared experiences as multilinguals appears to facilitate group sociability and construct linguistic diversity as a natural part of daily life. The portrayal of linguistic diversity as a normal part of everyday life was evident throughout the data. However, a variety of positions were adopted towards this diversity that illustrate the participants afliation towards the languages and dialects in their linguistic repertoires within the setting of the academic writing programme and the institution more widely. These positions were uid, with participants sometimes adopting seemingly contradictory positions. I have discussed elsewhere (Preece 2010) how these positions can be expressed along a continuum of weaker afliation to heritage languages/stronger afliation to English and weaker afliation to English/stronger afliation to heritage languages. In this paper, I will focus on a number of positions that the students adopted in relation to their linguistic repertoires that appear helpful for considering the needs and identities of heterogeneous groups of multilingual students in universities in the Anglophone centre. These positions are English as L2, separate bilingualism, exible bilingualism, separate bi-dialectalism and exible bi-dialectalism. 3.1. English as L2 A key position for several of the participants was English as L2, which can be placed towards the weaker afliation to English/ stronger afliation to heritage languages end of the continuum. As might be expected, migrants commonly adopted this position, particularly if they had arrived in the UK in their late teens and early adulthood. Akinwole illustrates this well. After growing up in Nigeria, where he attended English-medium schools, he migrated to the UK at the age of 19 to join his mother. On arrival, he spent ve years working and studying, before he was able to progress into university. In interactions about his linguistic repertoire, Akinwole displayed strong afliations to Yoruba, depicting this as his local language, and much weaker afliations to English. As Eversley et al (2010) list Yoruba as one of the top-ten languages in London, it is possible that local is a reference to Akinwoles everyday life in areas of London populated by the Yoruba community rather than a reference to his life back home in Nigeria. Akinwole expressed the view that:

Universities in the Anglophone centre: Sites of multilingualism


I never feel comfortable speaking in English with my family and friends because I always feel that I am in a formal setting and environment (Akinwole, questionnaire).

Akinwole presents himself as ambivalent to using English at home and with his Yoruba-speaking friends. His association of English with formality suggests that he primarily associates it with institutional settings, perhaps learned during his schooling in Nigeria. A similar picture emerges from students who can be classied as 1.5 generation. These students migrated to the UK part way through their compulsory schooling and as Marshall (2010: 43) contends, they often inhabit liminal spaces, living in two worlds, somewhere between rst- and second-generation immigrant. One student who exemplies this is Geet. Geet (aged 19) spent his early childhood in Kenya, using Gujarati at home and English and Swahili at school. When he was eight, his family sought asylum in the UK and Geet made the transition from Kenyan schooling into the British education system. After arriving in the UK, Geet also attended a Gujarati complementary school to maintain and improve his oral and written literacy. On comparing Gujarati and English, he claims that:
I have noticed . . . when I am . . . with my friends, say at work, . . . if they speak Gujarati then I will speak Gujarati as well. When I am talking [in] that, Im more calm and Im actually more easy going . . . When I speak with people just with English, I sometimes stutter . . . I might not . . . really have the words to say, they are difcult (Geet, Interview 1).

Despite living in the UK from a young age, Geet positions English as his L2. He presents himself as having difculties with expressing himself in English and claims that spoken interaction that has to be conducted in English only remains as a source of anxiety. Geet goes on to claim that his main motivation for speaking English is to improve his prociency:
I have Gujarati friends [at university] but they dont usually speak Gujarati that much so . . . Im always speaking English . . . so I can improve my English . . . I mean . . . if I always speak to people [in] English, Im able to improve my English. But at home . . . I have my Gujarati mother tongue language (Geet, Interview 1).

Classroom observations indicated that Geet was often marginalised during group work, particularly when interaction involved witty repartee. As Geet found it almost impossible to participate in the peer group banter of his British-born peers, his silence became the subject for jokes at his expense. As I have discussed (Preece 2006, 2009), adopting the position of English as L2 may have assisted Geet in negotiating social relations with his peers. By adopting Gu-


Sin Preece

jarati as his mother tongue and English as his second language, Geet was able to position himself as possessing communicative competence in Gujarati and a person worthy of respect among his peers. This position also enabled him to mitigate the problems that he was having with his academic work by allowing him to explain these in relation to being a language learner. 3.2. Separate bilingualism Separate bilingualism (Creese et al. 2008) was a commonly adopted position for the majority of the participants in the study. This can be placed midway along the continuum of weaker afliation to heritage languages/ stronger afliation to English and stronger afliation to heritage languages/weaker afliation to English and gives the impression of equality and balance between languages. Separate bilingualism indicates a view of languages as discrete and bounded entities that are used in different domains and need to be kept apart. This separation has been described in various ways, such as the two solitudes (Cummins 2005) and parallel monolingualism (Heller 1999, 2006). Drawing on Gravelle (1996: 11), Creese and Blackledge (2010: 105) observe how separate bilingualism represents a view of the multilingual/bilingual student/teacher as two monolinguals in one body. An example of this comes in the following extract of interview data with Saba (aged 18), a British Pakistani, who uses English, Urdu and Punjabi, in which we are discussing the associations that she makes with these languages. Saba is typical of the British born participants in that she had received very little sustained schooling in her heritage languages.
To tell you the truth the language that you speak at home . . . obviously there is more tradition comes into it, culture comes into it, religion comes into it, which is good cause obviously you need to like . . . keep a hold of your tradition and your culture and stuff like that, which is good, so your language Punjabi, Urdu comes into that. English . . . obviously you are in the country everyone speaks it you have to know it . . . the outside world is more based on . . . English . . . and I think that our language our mother tongue is for like home and your family and keeping the traditions, stuff like that (Saba, Interview 1).

In this interaction, Saba makes a clear distinction between the languages in her linguistic repertoire, demarcating Punjabi and Urdu for the domain of her family and English for use in the outside world. She constructs Punjabi and Urdu as maintaining the cultural and religious traditions of her heritage culture as well as interactions among family members. The private realm is constructed as excluding English in opposition to the public domain, which is portrayed as more or less English only. This example is typical of the ways in which many

Universities in the Anglophone centre: Sites of multilingualism


of the participants adopted a separate bilingual positioning. This positioning bounded heritage languages to the domestic and private space of the family and heritage community and English to the public domains of life beyond the ancestral community. 3.3. Flexible bilingualism Another position marking the midway point on the continuum is exible bilingualism (Creese et al. 2008). Unlike separate bilingualism, in which efforts are made to keep languages apart, exible bilingualism indicates the use of codeswitching, in which the speaker makes a complete shift to another language for a word, phrase, or sentence (Grosjean 2010: 512). This was characterised in the data as mixing of English and heritage language(s). This position is illustrated in the interview interaction below with Ling (aged 20), a British Chinese student, in which we are discussing her use of code switching.
Ling: Cantonese is mainly for like my parents and my grandma but um when I am communicating with my brothers and sisters we kind of mix up Cantonese and English. SP: Can you think of any situations when you would really mix up the Cantonese with the English when you are talking to your friends? What particular subjects do you think you talk about? Ling: Um anything really cause it is like when I am talking to you it is like you wouldnt know Chinese so Ill have to stick to English but when you know someone that is Chinese and they know English as well you kind of mix it up just naturally its not like when you change a subject (Ling, Interview 1).

Lings linguistic repertoire encompassed English, Cantonese, Vietnamese, British Sign Language and Mandarin. In this interaction, she differentiates her use of exible bilingualism from separate bilingualism. In interactions with others whose linguistic repertoire encompasses Cantonese and English, such as her siblings and friends, Ling adopts the position of a exible bilingual by reporting that she routinely juxtaposes Cantonese and English regardless of the topic of conversation. However, when her interlocutors linguistic repertoire differs (her elders and me), she is obliged to use one language at a time and one language only (Li and Wu 2009). This resonates with a variety of studies, such as Creese et al (2006: 38) and Li Wei and Wu (2009) whose study of complementary schools in the UK found that second generation minority ethnic teenagers valued being able to code-switch between their heritage languages and English. One use of code-switching was to construct a distinctive identity from rst generation migrants. As Creese et


Sin Preece

al (2006: 38) comment, exible bilingualism is a way in which British born bilinguals can differentiate themselves from newly arrived migrants and adopt positions as sophisticated young people in relation to the migrant freshie who is not procient in the linguistic practices of the peer group. Unlike separate bilingualism, exible bilingualism blurs the boundaries of the private and public worlds that second generation minority ethnic students inhabit, suggesting that they routinely use English in the home context alongside heritage languages and that heritage languages are used alongside English outside the home. Grosjean (2010) observes that code-switching is commonplace among bilinguals who share the same languages. He identies a variety of reasons for codeswitching, including the attractiveness of one language compared to another for expressing particular ideas, imitation of others and emphasis of social role. Grosjean also comments on the ways in which code-switching is often viewed negatively as creating an unpleasant mixture of languages, produced by people who are careless about the way they speak (p. 52). This perception was prevalent among the participants in my study who frequently associated code-switching with linguistic decit. This view is typied by Awino (aged 32), a Kenyan migrant who had been living in the UK for ten years at the time of the research:
When I speak Swahili with my friends I nd I break it with English sometimes/ speak English as well as Swahili/ some kind of mixes/ which is not really good (Awino, classroom data).

This utterance comes from interaction between Awino, Biba (aged 22) and Leela (aged 19), a British Moroccan and British Asian respectively, about the languages that they use at home and with friends. Despite portraying code switching as a habitual practice, they also conformed to ideological norms that associate it with decit and semilingualism. These negative connotations often came to the fore when intergenerational relations were the subject of discussion:
S = Seema, M = Maya 1. S: I speak English [mixed like (.) I dont know 2. M: [talked with ((xx))/ and I was like (.) half and half= 3. S: =half yeah: (1) but most of the time its in English okay/ now theyve like/ changed (1) 4. M: they know it 5. S: they think its bad though/ you know how parents are 6. M: its very bad we dont know the language <laughs>

In the extract above two British Asians of Gujarati heritage, Seema (aged 19) and Maya (aged 19), are discussing their parents negative response to their habitual

Universities in the Anglophone centre: Sites of multilingualism


use of code switching. The face-threatening nature of this exchange, in which they are sharing the experience of not only disagreeing with their parents but also acting against their wishes, is suggested through the covert references to their parents and elders in the ambiguous use of they (turns 35). The use of an impersonal pronoun in combination with the pauses in turn 3 is suggestive of shared experiences of family relationships. Mayas utterance (turn 6) may be enacting a parental disapproving voice with the accompanying laughter signalling that this is a subject for humour. However, the laughter may also mask anxieties about not fullling parental expectations and of not conforming to dominant discourses on bilingualism that assume that languages must be kept separate (see e.g. Blackledge and Creese 2010; Cummins 2005; Heller 2006). 3.4. Separate and exible bi-dialectalism The notion of separation and exibility is also helpful for reecting on the participantsrelationship with English. As mentioned, all were bi-dialectal users of English. The majority were habitual users of a vernacular variety, in particular London English (Harris 2006), while some were users of World Englishes. They routinely adopted bi-dialectal positions that were articulated through a posh/slang binary to differentiate between the prestigious variety of Standard English that they were required to use in higher education and the non-prestigious varieties of English with which they strongly identied. In interactions about their use of posh and slang, the participants adopted the positions of separate and exible bi-dialectalism. These mirror separate and exible bilingualism in that the rst involves the separation of standard and vernacular varieties of English while the latter involves their juxtaposition. An example of separate bi-dialectalism is illustrated in the following interaction between Awino, Biba and Leela, in which they are discussing their attempts at making friends with their undergraduate peers.
1. L: when we came ere/ I mean/ if I saw a posh person I actually spoke posh with them/ but if I saw somebody who was happy with their slang/ I spoke slang with them/ and I think thats how you socialise with them/ 2. A: yeah/ 3. B: its how you adapt to different people [thats what adapting is about/ 4. A: [yeah/ youve got to adapt/ yeah/ 5. L: yeah/ 6. B: adapting to different people an their cultures yknow/ an their backgrounds (classroom data).


Sin Preece

In the interaction, Biba, Leela and Awino co-construct boundaries between posh English and slang, which enables them to categorise the people that they are encountering in university into people like themselves, who are happy with their slang and posh others. They portray themselves as keeping posh and slang apart by using posh with one group of potential friends and slang with another. The regular reference to posh and slang in reference to the student body suggests that the participants have constructed an understanding based on shared experiences of social class norms for appearance, language and behaviour. It also suggests that the participants paid attention to the soundscape (Harris 2006), accommodating to the accents and conversational styles of their fellow peers as a way of tting in and making friends. It may be that separate bi-dialectalism enables working class students to navigate alien social terrain that includes many more people from middle class and professional backgrounds than they have previously encountered during schooling. In this instance, this positioning also facilitates a narrative of success in establishing new social relationships in higher education. Similarly to exible bilingualism, the participants also adopted the position of exible bi-dialectalism. An example of this comes from interview data with Tahir (aged 22), a British Pakistani, in which we are discussing how he makes sense of the academic work in his discipline:
Theres ve of us so we . . . all sit down and were trying to [explain] . . . the good thing is . . . whoever knows in that circle how to do it, he explains to all of us and because we all know each other well, know slang and stuff, wed explain it in a way we will understand . . . whilst if the teacher explains I might not get [it] . . . So Id explain it in my terms to make sure they understand it . . . showing the thinking, how I know how to do it (Tahir, Interview 2).

In this interaction, Tahir reports that he has formed a study group with his peers. This is represented as a circle, suggesting collaboration and cooperative learning. Flexible bi-dialectalism is used as a tool for constructing understanding of disciplinary knowledge by juxtaposing slang and stuff with the language of the academic community. This position also appears to be empowering in that it enables Tahir to speak from a position of expertise in which he can scaffold the intellectual development of his male peers. This is in sharp contrast to the institutional positioning that he occupies as a remedial English language user on the English language programme. While vernacular English has little legitimacy within the institution, the participants representations suggest that far from being a hindrance, they found vernacular English a valuable resource for establishing peer group relations and constructing a bridge into their academic work. This resonates with Ramptons

Universities in the Anglophone centre: Sites of multilingualism


(2006: 316) vernacularisation of school knowledge, in which adolescents use posh and Cockney stylisations as a way of mediating school work and getting down to the task at hand. Given the closeness in background and educational experience of Ramptons participants to those in my study, it seems likely that this strategy has been transferred from London schools into higher education.

4. Discussion The statistics for non-UK students in UK institutions and data from research conducted on the academic writing programme illustrate some of the ways in which the linguistic ecology of universities in theAnglophone centre has become more complex. However, despite the linguistic diversity in their midst, universities in Anglophone settings still seem to be operating with the assumption that their students are monolingual native speakers of English who naturally use a standardised variety of Anglo-American English and who enter university with considerable expertise in the literacy practices common in Anglo-American academic traditions. Although universities in the Anglophone centre claim to celebrate the cultural diversity of the student body, there is often less than a warm welcome for the linguistic diversity that accompanies cultural and social diversication. While ethnic monitoring in Anglophone settings is commonplace, there appears to be little systematic documentation of the linguistic repertories of students and staff. There is also a lack of institutional wide language policies that could inform curricula design and delivery. Curricula still largely adopt a monolingual stance that encourages uncritical use of one language only (OLON) (Li and Wu 2009) and by extension one dialect only (ODON) for most of the time. Universities in the Anglophone centre have not kept pace with the changing student demographic and need to devise institutional language policies that take pluricentric, rather than monocentric, perspectives to linguistic diversity. These could then be used to inform curricula design, delivery and assessment that are appropriate at the local level and more sensitive to the needs and identities of a linguistically diverse student population. In order to support the development of language policies there is a need for further research into linguistic diversity in Anglophone centre universities. Quantitative studies that map the linguistic repertoires of the staff-student population at institutional level would be helpful in developing the bigger picture, in raising awareness of linguistic diversity in the sector and for examining claims that multilingualism is the norm for signicant numbers of staff and/ or students in these settings. Further qualitative work exploring identity and pedagogical issues is also required as multilingual students do not form a homogeneous


Sin Preece

group and are likely to exhibit differing levels of expertise and afliation to the languages in their repertoires. They are also likely to have differing investment in the academic language and literacy practices of the Anglophone centre. For international students who have incurred very sizeable costs of studying in an Anglophone country, they may well be invested in the standardised English of the Anglophone centre, viewing this as carrying status and prestige. This may fuel resistance to using L1 in the university context. For domestic students, however, and for some from postcolonial settings, they may experience ambivalent feelings about the academic linguistic and literacy practices to which they are expected to conform. There are also emerging patterns from research about the ways in which multilingual students both conform to and resist the dominant norms for compartmentalising the languages and dialects in their linguistic repertoires. It would be helpful to draw on the research into separate and exible bilingualism in complementary schooling in the HE sector and to extend the notion of separation and exibility to varieties of English. This would enable further examination of the salience of separate and exible bilingualism and bi-dialectalism within the context of higher education. Kachrus model could provide a starting point for recognising linguistic diversity and hybridity of English use. While this model clearly has its limitations for describing the current situation in relation to English in different parts of the world, it serves the purpose of highlighting the pluralism, heterogeneity, and diversity of English world-wide (Berns, 1995: 10). It can be used to draw attention to the diverse linguistic repertoires of students in universities in the Anglophone centre and provide a way of critiquing attitudes in these universities to the linguistic diversity in their midst. Within this context, it also helps to re-position the remedial English user as an English-knowing bilingual (Pakir 1999) or bilingual user of English (McKay 2002) and to call attention to Lowenbergs (2002: 433) plea to nd ways of differentiating between deciencies in the second language acquisition of English . . . [and] varietal differences in the speakers usage resulting from their having previously learned and used . . . non-native normative features [italics in original]. At present, there is little differentiation between the two as is illustrated in Simpson and Cookes (2010: 70) study that illustrates how the sectors negative reaction to non-prestigious varieties of English of migrant students can contribute not only to these students experiencing downwards educational trajectories, but also to their longer terms prospects of educational success being damaged. This resonates with Blommaerts (2010) argument that the linguistic resources of those coming from the periphery are evaluated against a mainstream that reects the national order of the Anglophone centre (p. 173). Describing the asylum application of Joseph, a Rwandan refugee in the UK, Blommaert critiques the

Universities in the Anglophone centre: Sites of multilingualism


ways in which linguistic resources of those from the periphery not only [lose] weight and value during the journey [to the UK], but are also used by those in authority in the Anglophone centre to create arguments for doing things to people, such as refusing asylum applications or enforcing immersion education. Altbach et al (2009: viii) comment that one of the biggest challenges facing higher education in the global world is making opportunities available on an equitable basis. As they argue if current trends of internationalization continue, the distribution of the worlds wealth and talent will be further skewed. In the very unequal world in which we live and work, it is my contention that universities in the Anglophone centre need to exercise particular care not to increase the gap between the haves and have nots through language practices that privilege elite groups, whether these be monolingual or bilingual users of English, who already hold a disproportionate share of power and wealth. The call to imagine universities as multilingual spaces is intended as a way of valuing the knowledge and linguistic resources of bi- and multilingual students from a wide range of backgrounds, of creating spaces for other knowledge and perspectives to be critically evaluated and of nding ways of using linguistic diversity as a bridge into academic studies and a resource in the design, delivery and assessment of the curriculum. In conclusion, the mass expansion of tertiary education has resulted in an increasingly complex linguistic ecology in universities in theAnglophone centre. The linguistic ecology has been shaped by policies that universities have put into place to deal with globalisation, in particular the internationalisation and access agendas. While these universities are sites of multilingualism, this is barely recognised. When language appears on the agenda, it is generally with a problem label that frequently results in bi- and multilingual students being labelled as in need of remedial English tuition. It is my contention that we need to recognise linguistic diversity in the sector and consider ways of treating it as an asset. The rst step in this direction is to imagine universities in Anglophone settings as multilingual spaces and to acknowledge that multilingualism needs to be given space beyond the heads of the students or of . . . teaching . . . staff who are bi- or multilingual (Phillipson 2009: 210).

1. This paper is based on a presentation at the 2010 Bloomsbury Round Table 2. representing a rise of 31.5% on 2007/8 gures 3. representing a rise of 3.7% on 2007/8 gures


Sin Preece

4. Universities and Skills are one management group in BIS. This group is situated with nine other management groups, namely: Business; Economic and Policy Analysis; Fair Markets; Finance and Commercial; Innovation and Enterprise (including the Better Regulation Executive); Legal, People and Communications; Science and Research; Shareholder Executive and UK Trade & Investment. 5. All participant names are pseudonyms.

Allen, Liz, Lucy Solomon, John Storan and Liz Thomas. 2005. International comparators of widening participation in higher educaction policy and practice: Higher education in the USA, student fees, nancial aid and access. Action on Access. Altbach, Philip, Liz Reisberg and Laura Rumbley. 2009. Trends in global higher education: tracking an academic revolution. Executive Summary. A Report prepared for the UNESCO 2009 World Conference on Higher Education. Paris: United Nations Educational, Scientic and Cultural Organisation. Berns, Margie. 1995. English in the European Union. English Today 4. 311. Berns, Margie. 2005. Expanding on the expanding circle: where do WE go from here? World Englishes 24. 8593. Blackledge, Adrian and Angela Creese. 2010. Multilingualism: a critical perspective. London: Continuum. Block, David. 2006. Multilingual identities in a global city: London stories. London: Palgrave Macmillan. Blommaert, Jan. 2010. The Sociolinguistics of globalization. Cambridge: Cambridge University Press. Bolton, Kingsley. 2002. Chinese Englishes: from Canton jargon to global English. Special Issue of World Englishes 21. 181200. Creese, Angela, Arvind Bhatt, Nirmala Bhojani and Peter Martin. 2006. Multicultural, heritage and learner identities in complementary schools. Language and Education 20. 2343. Creese, Angela and Adrian Blackledge. 2010. Translanguaging in the bilingual classroom: a pedagogy for learning and teaching? The Modern Language Journal 9. 103 15. Creese, Angela and Peter Martin. 2003. Multilingual classroom ecologies: inter-relationships, interactions and ideologies. In Creese, Angela and Peter Martin (eds.) Multilingual classroom ecologies: inter-relationships, interactions and ideologies Clevedon: Multilingual Matters. 17. Creese, Angela, Barac Taskin, Arvind Bhatt, Adrian Blackledge, Shahela Hamid, Li Wei, Vally Lytra, Peter Martin, Chao-Jung Wu and Dilek Yagcioglu-Ali. 2008. Investigating multilingualism in complementary schools in four communities. Birmingham: University of Birmingham.

Universities in the Anglophone centre: Sites of multilingualism


Crystal, David. 1997. English as a global language. Cambridge: Cambridge University Press. Crystal, David. 2004. The past, present and future of World English. In Gardt, Andreas and Bernd Huppauf (eds.) Globalization and the future of German Berlin: Mouton de Gruyter. 2745. Cummins, Jim. 2005. A proposal for action: strategies for recognizing heritage language competence as a learning resource within the mainstream classroom. Modern Language Journal 89. 58592. Department for Business Innovation and Skills. 2010. About BIS. about (accessed 10 December 2010). Eversley, John, Dina Mehmedbegovic, Antony Sanderson, Teresa Tinsley, Michelle vonAhn and Richard Wiggins. 2010. Language capital: mapping the languages of Londons school children London: CILT. Gibbs, Paul, Panayiotis Angelides and Pavlos Michaelides. 2004. Preliminary thoughts on a praxis of higher education teaching. Teaching in Higher Education 9. 18394. Giroux, Henry. 2004. Betraying the intellectual traditional: public intellectuals and the crisis of youth. In Phipps, Alison and Manuela Guilherme (eds.) Critical pedagogy: political approaches to language and intercultural communication Clevedon: Multilingual Matters. 721. Gravelle, M. 1996. Supporting bilingual learners in schools. Stoke-on-Trent: Trentham Books. Grosjean, Francois. 2010. Bilingual: life and reality. Cambridge Massachusetts: Harvard University Press. Harris, Roxy. 2006. New ethnicities and language use. Basingstoke: Palgrave Macmillan. Heller, Monica. 1999. Linguistic minorities and modernity. London: Continuum. Heller, Monica. 2006. Linguistic minorities and modernity. 2nd ed. London: Continuum. Higher Education Statistics Agency (2010). Statistics students and qualiers at UK HE institutions. Jenkins, Jennifer. 2003. World Englishes: a resource book for students. London: Routledge. Kachru, B. 1992. Models for non-native Englishes. In Kachru, B (ed.) The Other tongue. Englishes across cultures 2nd. Urbana: University of Illinois Press. Kachru, Braj B. 1985. Standards, codication and sociolinguistic realism: the English language in the outer circle. In Quirk, Randolf and H. G. Widdowson (eds.) English in the world: teaching and learning the language and literatures. Cambridge: Cambridge University Press. 1130. Kaplan, Robert and Richard Baldauf. 2009. An Ecological perspective on language planning. In Creese, Angela, Peter Martin and Nancy Hornberger (eds.) Encyclopedia of Language and Education, vol. 9 2nd. New York: Springer. Kemp, Charlotte. 2009. Dening multilingualism. In Aronin, Larissa and Britta Hufeisen (eds.) The exploration of multilingualism: development of research on L3, multilingualism, and multiple language acquisition Amsterdam: John Benjamins. 1126. Li, Wei and Chao-Jung Wu. 2009. Polite Chinese children revisited: creativity and the


Sin Preece

use of codeswitching in the Chinese complementary school classroom. International Journal of Bilingual Education and Bilingualism 12: 193211. Li,Wei and Chao-Jung Wu. 2009. Polite Chinese children revisited: creativity and the use of codeswitching in the Chinese complementary school classroom. International Journal of Bilingual Education and Bilingualism 12: 193211. Lowenberg, Peter H. 2002. Assessing English prociency in the Expanding Circle. World Englishes 21. 4315. Marshall, Steve. 2010. Re-becoming ESL: multilingual university students and a decit identity. Language and Education 24. 4156. Martin, Peter. 2010. They have lost their identity but not gained a British one: nontraditional multilingual students in higher education in the United Kingdom. Language and Education 24. 920. McKay, Sandra. 2002. Teaching English as an international language: rethinking goals and approaches. Oxford: Oxford University Press. McLean, Monica. 2006. Pegagogy and the university: critical theory and practice. London: Continuum. Mhlhusler, Peter and Alwin Fill (eds.). 2001. The Ecolinguistics reader: language, ecology and environment. London: Continuum. Pakir, Anne. 1999. Connecting with English in the context of internationalism. TESOL Quarterly 33. 10312. Phillipson, Robert. 2009. Linguistic imperalism continued. Hyderabad: Orient Blackswan. Platt, John Talbot, Heidi Weber and Ho Mian Lian. 1984. The New Englishes. London: Routledge. Preece, Sin (2006). British Asian undergraduate students in London. In Block, David, Multilingual Identities in a Global City: London Stories. London: Palgrave Macmillan. 17199. Preece, Sin. 2009. Posh talk: language and identity in higher education. London: Palgrave Macmillan. Preece, Sin. 2010. Multilingual identities in higher education: negotiating the mother tongue, posh and slang. Language and Education 24: 2140. Rista-Dema, Mimoza. 2008. Verb-subject word order in Albanian English. World Englishes 27. 41933. Ruiz, Richard. 1984. Orientations in language planning. NABE: The Journal for the National Association for Bilingual Education .1234. Seargeant, Philip. 2005. Globalisation and recongured English in Japan. World Englishes 24. 30920. Shohamy, E. 2006. Imagined multilingual schools: How come we dont deliver? In Garcia, O., Tove Skutnabb-Kangas and M. Torres-Guzman (eds.) Imagining multilingual schools. Languages in education and globalization Clevedon: Multilingual Matters. 17183.

Universities in the Anglophone centre: Sites of multilingualism


Silverstein, Michael. 1996. Monoglot standard in America: standardization and metaphors of linguistic hegemony. In Brenneis, D and R Macaulay (eds.) The Matrix of language: contemporary linguistic anthropology Boulder: Westview Press. 284306. Simpson, James and Melanie Cooke. 2010. Movement and loss: progression in tertiary education for migrant students. Language and Education 24. 5774. Times Higher Education (2010). Times Higher Education-QS World University Rankings 2009: Top 200 world universities. UNESCO Institute for Statistics (2009).Trends in tertiary education: Sub-SarahanAfrica. UIS Fact Sheet, July 2009, No. 01. Velez-Rendon, Gloria. 2003. English in Colombia: a sociolingustic prole. Special Issue of World Englishes 2. 18598. Vertovec, Steven. 2006. The emergence of super-diversity in Britain. http://www.compas. (accessed 09 November 2010). Sin Preece is a Lecturer in TESOL education at the Institute of Education, University of London. She is interested in applied linguistics in educational settings and her research interests include linguistic and cultural diversity, language learning and widening participation and language and gender. She is the author of Posh Talk: Language and Identity in Higher Education and has published several articles on multilingual and gender identities.

Risking or boosting masculinity? Mens language choices in multilingual settings


Abstract In this paper I draw upon recent insights and advances in the study of gender in multilingual settings to analyse and discuss the constitutive role of language in gender performativity, more specically in the performance of masculinities. A focus on mens rather than womens communicative behaviour and language choices in bilingual settings is motivated by a continuing dearth of studies of men as gendered beings. In this paper I present case studies of the linguistic choices and, to some extent, language practices of three Australian-born men whose parents had migrated to an urban setting in Australia. These case studies provide an insight into the complex interrelationship between linguistic choices and the performance of their masculinities in various settings and across different stages of life. A key nding emanating from this research concerns the centrality of the (male) peer group in shaping the linguistic choices these men make and how these may inuence current and future bilingual practices, ultimately impacting on language maintenance efforts.

1. Gender research and multilingual settings In the past 2 decades signicant advances have been made in the exploration and study of gender, gendered behaviour and gendered practices in bilingual settings and multilingual contexts (e.g., Burton et al. 1992; Gal 1991; Goldstein 2001; Pavlenko et al. 2001; Pujolar 2001; Rampton 1995; Trechter 2003; Winter & Pauwels 2005 to name but a few). This led to the observation by Piller and Pavlenko (2001:13) in their introduction to the arguably rst comprehensive discussion of gender in the context of multilingualism and second language learning, that the gender-blindness from which much research into biand multilingualism suffered, is starting to disappear. Indeed until the 1990s


Anne Pauwels

focus and attention on the role of gender and sex in shaping, inuencing or reecting linguistic practices and choices in multilingual settings were minimal (with some notable exceptions, e.g., Callan and Gallois 1982; Gal 1978) and tended to operate with a concept of gender as a set of traits (based on biology) differentiating men and women. Differences in bilingual practices and choices, differential rates of language maintenance or shift were explained in terms of the social and cultural consequences of biological roles (e.g., for details Holmes 1993; Pauwels 1995, 1997). For example, immigrant womens greater degree of language maintenance in a range of communities was linked to their (biology-based) role of primary carers for children and the aligned cultural role of primary bearers and transmitters of cultural and linguistic traditions. More recently, explanations of language practices based on essentialist concepts of gender, ethnicity, race and sexuality among others have been subjected to considerable scrutiny given their (signicant) limitations. In relation to gender (and sexuality) the eld of language and gender has been the main force in upping the level of theorising around these concepts (e.g., Butler 1990, 1993; Cameron 1997a; Eckert&McConnell-Ginet 1992), although it took some time before they were applied to multilingual settings due to a monolingual focus (monolingual bias according to Piller & Pavlenko 2001: 1) prevailing in language and gender research. Current work on gender and language favours a social constructionist or performative view of gender (e.g., Holmes & Meyerhoff 2003; McElhinny 2003) stressing the importance of a more context-sensitive approach which treats gender as a system of social relations and discursive practices whose meaning varies across speech communities (Piller &Pavlenko 2001: 3). Moving beyond gender as an essential category allows for a capturing of the uidity, dynamics and permeability of gender across time, contexts and cultures:If gender is viewed as a social, historical and cultural construct, then it comes as no surprise that normative masculinities and femininities, as well as beliefs and ideas about relations between sexes, may vary across cultures as well as over time within a culture (Piller & Pavlenko 2001:22). To this description I would add that variation not only occurs over time within a culture but also within a persons lifespan. What is considered to be normative masculinity/ies or femininity/ies in adolescence may well be (very) different from those affecting an adult or an elderly person. Furthermore the consequences of transgressing normative masculinities or femininities tend to be different, not only according to culture but also depending on life stage.

Risking or boosting masculinity? Mens language choices in multilingual settings


2. Gender, adolescence and linguistic identity Much sociolinguistic research on language has revealed the centrality of adolescent speech behaviour in effecting language change (e.g., Labov 1972; Eckert 1989) as well as pinpointed the period of adolescence and young adulthood as a particularly crucial and salient stage in the formation of an individuals linguistic identity (e.g, Eckert 2000, 2004; Hewitt 2003; Rampton 1995, 2006). For many young people across cultures and in all parts of the world this period marks the rst major opportunity to shape, form and negotiate expressions of identity away from parental control and to seek alignments of various kinds with ones peer group. Often it is not only a period of rebellion or of resistance against norms, traditions and what may be considered hegemonic practices but also a period in which acceptance by ones peers through afliation with a variety of subcultures is of great importance. In multilingual settings choosing to speak one language or code over the other, or engaging in innovative, non-conformist linguistic practices vis a vis the languages in ones community are typical ways in which ` immigrant, bilingual, ethnic and minority adolescents as well as members of certain youth subcultures express and negotiate their identities linguistically (e.g., Bailey 2000; Cheshire 2008; Maher 2005). Of particular importance during this period of transition from childhood to adulthood is the shaping of a gendered identity which involves a constant negotiation between desirable, desiring and normative performances of femininity and masculinity. Whilst these negotiations around gender affect adolescents in any kind of setting, those adolescents operating in multilingual and multicultural settings face the additional task of negotiating different cultural normativities associated with masculinity and femininity (Piller and Pavlenko 2001) often leading to very complex situations. For example, transgressing normative masculinity or femininity may be harshly sanctioned in one of their cultural spheres, yet seen as acceptable in their other cultural sphere. Thus attempts to resist normative femininity/masculinity as part of adolescent rebellion in one culture may be simply perceived as normative gendered behaviour in the other culture. The case studies presented in this paper will show how three men have dealt with this complex transcultural negotiation of their masculinities and how this has impacted on their linguistic practices and language choices beyond adolescence.

3. Mens language choices in multilingual settings Given the prominence of feminist approaches in the study of gender and language, it is understandable that the linguistic treatment of women has occupied


Anne Pauwels

centre stage for many years, both in monolingual and multilingual settings. After all, for many centuries womens linguistic behaviour was seen as either aberrant and hence in need of explanation or as unworthy of attention. Since the 1990s increasing attention has been paid to studying and analysing mens behaviour from a feminist perspective in an attempt to put gender back into men, i.e. to examine men as gendered beings. A collection of papers edited by Sally Johnson and Ulrike Meinhof (1997) has been instrumental in exploring mens communicative practices and linguistic choices from this perspective. Central to this exploration is the need to recognise the dialectal nature of the relationship between femininity and masculinity and thus extend the social constructionist or performative notion of gender to the latter. This gives rise to the notion of variable, contextualized masculinities (Johnson 1997:20) because There is no underlying reason why men as a group should be linguistically homogeneous. And there is no inherent reason why the discursive strategies used by individual men should be consistent - either from one situation, or even one utterance to the next Johnson (1997: 21). This stance is not without its critics (both feminist and non-feminist). From a feminist perspective the criticism has focused on the impact that the recognition of heterogeneity in mens behaviour, of multiple masculinities and the acceptance of a more complex view of masculinities have had on the issue of masculine hegemony (e.g., Grifn 1989). Taking this criticism into account Johnson (1997: 21) draws upon Robert Connells (1995) views that studying the multiplicity and diversity in masculine performances is not so much a question of learning to appreciate the diversity but for understanding the way in which such variation plays a part in the overall construction of hegemonic masculinity. Although there has been a recent increase in studies examining mens communicative behaviour in multilingual settings (e.g., Pujolar 2001, Teutsch-Dwyer 2001) the majority of work examining the nexus between language practices and masculinities continues to be located in predominantly monolingual or monocultural settings (e.g., Cameron 1997b; Coates 2003; Kiesling 1989). With this paper I wish to make a further contribution to exploring the dynamics of language and masculinity in multilingual settings.

4. Gender and language dynamics in transitional bilingual settings In the following sections I present and discuss three case studies that explore and illustrate the interplay between language choice and gender (masculinities in particular) and how this may impact on questions of language maintenance. Within a framework of critical sociolinguistic studies of multilingualism (e.g.,

Risking or boosting masculinity? Mens language choices in multilingual settings


Heller 2007) they provide scenarios for more detailed and localised discussions of language identities assisting in identifying and rening language maintenance opportunities and restrictions. The issue of language maintenance arises foremost in societies where there is no tradition of stable bilingualism, either in practice or through policy. In many such societies, often characterised by a considerable migrant intake, the maintenance of the immigrant language is not considered a responsibility of the state but as that of the community in question (e.g., Ozolins 1993). Of course there is considerable variation amongst such societies in terms of their attitudes towards ongoing bilingualism and the amount of support or opposition they have vis a vis language maintenance efforts (e.g., Lo ` Bianco 2007). Furthermore I have chosen to focus the case studies on members of the so-called second generation, i.e. children born and raised in the country to which their parent(s) migrated. Although the label second generation to denote the locally-born children of migrants is increasingly subjected to debate (e.g., Winter & Pauwels 2007), there is a strong consensus among language scholars of the pivotal role this generation plays in shaping the language practices of future generations. Unlike their parents they are immersed from birth into the new linguistic environment and learn the dominant language of this new environment with relative ease. Within domains linked to family life, members of this second generation are usually exposed to the parental language(s) in which they gain varying levels of competence. Consequently, many children of migrants grow up bilingually. Of crucial interest in the study of language maintenance and shift is the question whether and how this second generation continues its bilingual practices and thus enables subsequent generations to access the language(s) of their ancestors. Not surprisingly, the examination of the language dynamics, linguistic choices and attitudes of this second generation has become a major research focus. Furthermore, though the case studies are not strictly speaking longitudinal, they do provide insights into these mens linguistic practices from childhood through adolescence to (young) adulthood.

5. The Case studies 5.1. Data The data for these case studies have come from two large-scale projects investigating language practices in bilingual settings. The rst project was a three year study of the language dynamics and the language maintenance patterns in two generations of three ethnolinguistic groups residing in Australia: German-


Anne Pauwels

Australians, Greek-Australians and Vietnamese-Australians. Data on their language use was collected mainly through questionnaires, interviews and participant observation. The second project (which is still in its pilot phase) plans to examine the language practices of Australians who have migrated or left Australia on a long-term basis. Although this is not a new phenomenon, to date there has been little systematic attention paid to this group of Australians or on the impact of increased outward migration and transnational mobility on Australian society. In a recent study by Hugo et al. (2003) these Australians are referred to as Australias diaspora. Within this group, my particular focus is on the Australian-born children of immigrants who move to live long-term in their parental home country. The modes of data collection include in-depth interviews, analysis of diaries, memoirs, blogs, etc., as well as questionnaires about various aspects of language use. To date in depth interviews with 4 participants have been undertaken and their questionnaire data have been analysed. Two case studies Max and Con have been drawn from the rst project and the third case study Roelof comes from the second project on the language practices of the Australian diaspora. 5.2. Proles of the three men: Max, Con, and Roelof Max is a German-Australian who was 32 at the time of the investigation (1993). He was born in Melbourne from German-born parents who had migrated in the early 1960s. Max has one sister who is three years younger. He was exposed to German in the home although his parents did not insist on him speaking German. In lower secondary school he learned German for three years but did not continue with it. Max is tertiary educated and works as an engineer in an automotive company which has quite a lot of contact with Germany.At university he met his partner who does not have a German background but who is keen for Max to teach her some German. According to Max she encourages him to use more German to his German friends and family. At the time of the interviews they had been together almost ten years. They do not have children yet. Max has regular contact with his parents, relatives and friends who are German-born or like him are children of German immigrants. His mother died seven years before the interview with Max. Con is a Greek-Australian who was 28 years old at the time of the interview (1993). He has one sister who is three years younger and he is engaged to a second generation Greek-Australian woman who is 25 years old. He still lives with his parents. He went to Greek school as a little boy and later attended Greek language classes run by the state school system. He sat a nal year exam in Greek. He went to university where he studied law and is now working as a

Risking or boosting masculinity? Mens language choices in multilingual settings


solicitor in a rm that counts many Greek-Australians amongst its clientele. His friendship circle is a mixture of Greek-Australians and predominantly AngloAustralians. He has regular contact with his extended family including many cousins who are also second generation Greek-Australians. Roelof is a Dutch-Australian who was almost 45 at the time of the interview (2006). He was born in Melbourne, the youngest of 5 children two of whom were born in Australia. He grew up in a household in which a lot of Dutch was spoken and heard. Like Con, Roelof attended Dutch language classes run by the state school system between 12 and 18 and sat a nal year exam in Dutch. He went to university to study agriculture. When he was 27 years old he went to live in the Netherlands to undertake a traineeship in the tulip industry as he was keen to set up a bulb-growing business in Australia. Initially he lived with his aunt and uncle whose two sons were Roelof s age and who were quite uent in English. After his traineeship Roelof was offered a position in the export branch of the company and soon thereafter he moved into his own accommodation. After three years in the Netherlands he married a colleague. They stayed in the Netherlands for another 9 years and had two children. During his twelve years in the Netherlands Roelof maintained regular contact with his family in Australia including two return visits. He also had occasional visits from his Australian-based family and relatives. In 2000 Roelof decided to move his family to Australia, where they have been living ever since. 5.3. Linguistic skills and prociency In an investigation about language choice in bilingual settings it is important to gain an understanding of the linguistic skills the participants have as well as an insight into how they rate their language skills. Although there is no direct relationship between language prociency and language use, a participants actual and/or self-rated prociency may inuence the language choices they (can) make and the language practices they can engage in. Hence Roelof, Con and Max were asked to rate their prociency in their home languages Dutch, Greek and German respectively. In addition the interviewer rated their prociency in speaking, listening and reading based on their performance in the interview which also included some language tasks. The International Second Language Prociency Scales (originally known as the Australian Second Language Prociency Ratings) developed by David Ingram and Elaine Wylie (Ingram & Wylie 1979) was used to undertake the ratings. It uses a 12-point scale ranging between 0 and 5 to rate each skill with 0 standing for no prociency to 5 referring to native-like prociency. The participants English language skills were not assessed given their educational trajectories (all had completed university-level


Anne Pauwels

education). The Table below summarises their self-rated and interviewer-rated language prociency in three skills.
Table 1. Self-rated and interviewer-rated language prociency skills of 3 interviewees
Participant Con Max Roelof Self-rating: Speaking 2+ 1+ 2+ Self-rating: Listening 4 4 4 Self-rating: Reading 3 2+ 3 Rating: Speaking 2+ 2 2+ Rating: Listening 4 3 4 Rating: Reading 2+ 1 3

Overall, there is considerable consistency between the self-assessment and the assessment undertaken by the interviewer. Only Maxs self-ratings tend to be different from the interviewer ratings: he underrates his spoken prociency whilst slightly overrating his listening and reading skills. Con slightly overrates his reading skills. Their ratings for speaking and listening, which are the key skills for making language choices, are between 2 and 2+ for speaking and between 3 and 4 for listening. In the ISLPR scale, ratings of 2 and 2+ for speaking entail basic social prociency with the following description for 2.
I speak X well enough to take part in face-to-face conversations with most background speakers and in telephone conversations describing familiar things and relating familiar events, and conveying my opinions fairly precisely off the cuff. I use a range of complex sentences. I often have trouble coming up with the vocabulary I need. I use a variety of constructions with clauses but I make mistakes in grammar, particularly when I am trying to express more complex ideas. Beyond basic courtesy forms I have limited ability to tailor my language.

The description for a listening scale of 3 states:

I understand almost everything when I am participating in social conversations with background speakers of Indonesian on fairly complex and abstract topics (e.g. the extent to which a government should subsidise sporting activities). I can generally follow a conversation I overhear between background speakers (e.g. on a bus) even though I cant understand some things that they say. I can use the telephone for most purposes and I understand most TV and radio news stories.

For a rating of 4 it is:

I understand most things in the language, even things as difcult as complex radio documentaries with fast speech. However I tend to miss subtle plays on words or references to deep aspects of the culture. I have difculty with some accents. (ISLPR scale)

These ratings suggest that the participants have sufcient levels of prociency in the home/community language to operate relatively comfortably.

Risking or boosting masculinity? Mens language choices in multilingual settings


5.4. Language practices in the family All three men grew up in a family where the community language (i.e. Dutch, German or Greek) was the main, if not sole language used for family interactions. In Cons and Roelof s case, parents were quite insistent that they speak Greek/Dutch in the family. They were encouraged as well as reprimanded if they replied in English. Con describes his home situation as completely Greek, all the time Greek at home whereas Roelof recalls that both his father and mother would feign lack of comprehension until he answered in Dutch. Max thinks he only spoke German before he went to school: No, I dont remember it myself but I was told it was German. As soon as he went to school his parents no longer insisted on him speaking German with them English and German was denitely spoken in the family. Whereas parental lack of prociency in English played a role in the language choices Con made, neither Roelof s and Max parents experienced (major) difculties with English. Almost from the start English was the preferred language for communication with siblings, especially younger ones or those close in age to the participants, although the presence of parents could change that. In response to the interviewers question whether he would ever speak German with his sister, Max gives a categorical No I dont. Con claims to mainly use English with his sister except if his parents are around, then you speak Greek. Three of Roelof s siblings are much older than him (10 to 15 years older) and he said that he would follow their lead with regard to language use when he was in Australia. However, with Joost who is two years older than him and who was also born in Australia, he would always speak English except when his parents insisted on them speaking Dutch. By the time Max was 17 he spoke mostly English with his parents, out of habit and German had become a language he heard rather than spoke in the home. For Roelof and Con, adolescence did not change the language dynamics within the family much with Dutch/Greek continuing to dominate the interaction with parents and English with siblings. For Max, a change came when his father became more dependent on him and his sister after his mother died he was 25. His father had always spoken more German than his mother and now reverted even more to German. Max felt that he needed to speak more German with his father, yet he struggled as he was not feeling condent enough to do so. This lack of condence in speaking German seemed to be exacerbated when his younger sister was around. He felt that she seemed more uent in German than him. He also remarked that his sister now wanted to speak German with him when his father was around although they had never spoken German to each other when younger. His reaction to this apparent


Anne Pauwels

change was to remain relatively silent in such interactions. Language choices with his father and sister have not changed dramatically since he moved out of the parental home and started sharing a home with his partner, a non-German speaker who is interested in learning German. In fact, when his sister visits them at home, his partner uses the occasion to practice some German with her, something his sister is happy to do. Sometimes Max joins in with a few German words but does not switch into German. During a participant observation session around an extended family get together (christening of a nephew) Max could be heard to speak in German with his father and some older relatives and older friends of the family when he was on his own with them. If they were joined by visiting relatives and friends from Germany (both young and old) he remained silent or responded in English. His sister on the other hand seemed to relish the opportunity to practice her German with them. Roelof moved out of the parental home when he was 19 and went to university. He recalls that little changed in terms of language use when he visited his parents but that his telephone conversations with his parents were usually bilingual: he spoke English and they spoke Dutch. Language use with his siblings remained the same when he met them in the parental home but changed almost completely to English outside of this location. His move to the Netherlands meant that there was almost no face-to-face contact with parents and siblings. When he called his parents he greeted them in Dutch and then usually switched to English. Upon arrival in the Netherlands his interactions with his aunt, uncle and other relatives irrespective of their age were predominantly in Dutch. Over time, despite being immersed in a Dutch-speaking environment, English became Roelof s main language of interaction with his cousins and other peer group contacts (see below). 5.5. Language practices and the peer group For many second generation Australians the use of the community language is associated primarily with interactions involving members of the rst generation: parents, older relatives and family friends. In other words, it is the intergenerational language. In some families and communities, this practice continues to be inuenced by the older generations limited competencies in English. In most cases though, it is linked to the older generations preference for the use of the community language in familial settings for a plethora of reasons. For peer group interactions involving friends, relatives and other members of the same ethnolinguistic group the preferred, if not unmarked language choice tends to be English, as most investigations into the language practices of the second generation in Australia have shown (e.g., Bennett 1990; Cavallaro 1997; Rubino

Risking or boosting masculinity? Mens language choices in multilingual settings


2006; Tsokalidou 1994). Peer group interactions and settings, especially during adolescence, provide opportunities to play and experiment with emergent linguistic and cultural identities, to explore questions of hybridity and in some cases escape the linguistic and cultural constraints of the home environment. These contexts are critical sites in the performances of gender which inuence and shape linguistic practices. In bilingual settings the linguistic choices made in such sites can shape the future of the minority language in the community, especially after the passing of the rst generation. Con, Roelof and Max all have access to peer group interactions involving members of their ethnolinguistic group including relatives, friends and acquaintances. Cons peer group networks often involve friends, relatives and acquaintances of Greek-Australian background. This is not surprising as the Greek community in Australia is both expansive and close-knit (Clyne 1991). Cons friendship groups and networks have always included Greek-Australian peers because this is kind of unavoidable in Melbourne according to Con. He recalls that when he rst went to school he would sometimes seek out Greek kids so that he could speak Greek as he got tired from concentrating on speaking English. However, this changed rather rapidly and English became the default language among his Greek-Australian friends and peers, even whilst attending Greek school. Greek was spoken in the classroom but outside of the formal lessons, the students, including Con, reverted to English. Con commented that he started using more Greek with his friends, especially male friends, in high school, we kinda used English with Greek words thrown in. Con enjoyed this way of speaking with his friends, i.e. code-switching, and he engaged in it frequently even in the presence of non-Greek speaking friends. Con describes his language patterns in these settings as follows: it depends on the ow of the conversation sometime I I in that regard I found myself to tend to drift from both from both la. . . to from one language to the other; it just depends on the ow of the conversation or the topic of the conversation whether youre discussing Greek politics or youre discussing [incomprehensible]. and, around friends, mostly English with a bit of Greek thrown in between. Although Con felt condent about his prociency in Greek, he did comment that he sought out opportunities to improve his Greek, especially by speaking with same age (male and female) visitors from Greece: I tend to enjoy the conversation in Greek, ah and mainly for the ah only . . . probably for a selsh reason that is I want to improve my own Greek and I


Anne Pauwels

enjoy having a uent conversation in Greek with someone who is actually also uent in Greek, and that tends to improve my own Greek; if there is someone from Greece I think because . . . even though they speak English, you . . . I personally feel that for my own sort of, at least its just, I enjoy the conversation more because they are so uent in Greek that I enjoy the conversation and it allows my own development of the Greek language.

Max also continues to have peer group networks that include German-Australians and German speakers. Yet he claims that interactions with such peers are almost entirely conducted in English. This was conrmed in the participant observation sessions comprising Max and his German-Australian friends and cousins. There was almost no evidence of any form of code-switching or codemixing to the extent that German placenames and names were pronounced in their Anglophone forms. Although Maxs friendship networks are ethnically more diverse than Cons, he nevertheless continues to have a select group of close German-Australian friends and family with whom he meets regularly. Unlike Con, Max does not seize opportunities to improve his German when he meets up with more uent language users. In fact he mentions that he consciously avoids speaking German with his peers, especially if he perceives them as being more competent in German. Interestingly, one of his best friends whom Max sees regularly, is perceived as a uent speaker of German I have a friend who speaks it uently yet he never speaks German with him: The reason is I really just dont feel condent enough. Later on in the interview he elaborates on his reluctance to speak German to his peer group: I suppose I dont make a well maybe I do make a bit of a conscious effort in the sheer fact that I dont have enough condence to speak it so I suppose it would be and also Id be very conscious of them judging the way I would be speaking German so therefore I dont do it I guess. Although Maxs friendship networks give him regular opportunities to speak German he does not take advantage of these, apparently for fear of having his German competence judged as inadequate. His partner conrmed this and commented that it was particularly obvious with his male friends whom Max perceived as being more competent speakers. Only in situations where he is condent that his prociency is better than that of his interlocutors is he willing to use German. For example, he mentions that he has done some basic translating for English-speaking monolingual colleagues at work: To try and interpret for someone else just you know um just to sort of to basically yeah try and get a message across. Max restricts his use of German to situations where he is expected to speak German (i.e. with the older rst generation) or where he feels that his prociency outranks or is at least on a par with that of his interlocutors. As these situations occur less and less

Risking or boosting masculinity? Mens language choices in multilingual settings


frequently his use of German is likely to diminish considerably, possibly leading to complete language shift. Roelof s situation is more complex due to his move to the Netherlands where he is immersed in a Dutch-speaking environment. Yet his language choices are more akin to Maxs than to Cons. Roelof recalled that he was not too apprehensive about having to speak Dutch all the time when he moved to the Netherlands. He felt condent that he would manage the transition quite well and was looking forward to being immersed in a Dutch language environment. However, this changed quite dramatically after a few weeks living with his relatives in the Netherlands. His aunt and uncle did not speak English very well and he was comfortable speaking Dutch with them. In addition, they regularly praised him for having maintained Dutch so well and being so good at it. Interactions with his cousins, two brothers who were slightly younger than him (22 and 25 respectively) were initially in Dutch but soon switched to English. This seemed to be a combination of their wish to practice English with me and Roelof feeling intimidated by their apparent uency in English. He mentions that he felt quite awkward when they made remarks like just speak in English cause we understand and its easier for you when he was trying to explain something complex in Dutch. Although they did not correct his mistakes, they hurried me up by switching to English. Roelof mentions that their English wasnt that crash hot but that they did not seem to care about it and that they feel condent using it. As he got on well with the two brothers, he joined their network of friends and the same language pattern arose when interacting with these friends. He joked that I was like an English language guru to them but also admitted that I am getting less condent about my Dutch. Despite feeling that his Dutch language abilities had improved signicantly since his arrival in the Netherlands, he grew less condent using Dutch with his friends and colleagues, especially if they were male, and with his cousins. This growing lack of condence in his Dutch language skills triggered in Roelof the same reaction that Max displayed withdrawal from the use of the community language. Whilst he did not withdraw from social situations or peer group interactions with Dutch speakers he seldom, if ever, used Dutch with them. He noted that in such encounters they were happy for me to keep speaking English as long as I could understand what they were saying, which was not too difcult except if they spoke dialect. Asked why he seemed more reluctant to use Dutch with male friends, he commented: Somehow it feels that they are more critical, you know more . . . that you are you know less competent or something because they look up almost admire you when you speak uent English, it gives you status. When he married a Dutch speaker, his use of Dutch increased in the home, although a fair bit of English was spoken. During this time he also started using more Dutch with his friends


Anne Pauwels

when they made home visits. However, he continued speaking English when he had nights out with the boys. 5.6. Language practices with the next generation Neither Con nor Max had children, nieces or nephews at the time of the interviews. It was therefore not possible to examine their language practices in relation to the next generation at the time. However, both Con and Max were contacted in 2007 to ascertain their current family situation. Neither was available for further interviews but Con reported that he had two young children and that Greek continued to be spoken in the family. Max reported that he had separated from his partner and that there were no children. On the other hand, Roelof at the time of the interview had two children allowing an insight into the question of language maintenance in this family. Roelof and his wife, Hanna mentioned that code-switching was their normal way of interaction before they had children. Roelof mentioned that he felt more at ease speaking Dutch with Hanna, especially in the home: Speaking Dutch with her was fun and normal, no stress about getting it wrong. This changed again when they had their rst child as both parents decided it would be good to raise the child bilingually. Roelof would speak English to the child and Hanna would speak Dutch. Both Roelof and Hanna commented that they were pretty consistent in this pattern even when we had our second one. The presence of two young children who spoke Dutch with their mother, relatives and friends did impact on the language for family communication: Dutch started dominating as the language of the family although Roelof continued speaking mainly English to his children. After moving his family to Australia, Roelof s use of Dutch diminished further with few situations calling for the use of Dutch. His children have become Englishdominant although they continue speaking some Dutch with their mother and also with their Netherlands-based relatives. The language of the family is predominantly English with little evidence of code-switching. Roelof s main use of Dutch is in communication (mainly via phone and video calls) with his Dutch in-laws. Roelof s parents have died and his contact with his siblings is almost entirely in English.

6. Men and language choice in multilingual settings These admittedly brief descriptions of these mens reported language practices in the context of family and friendship settings offer an opportunity to gain insight into how their masculinities shape and are performed through the lan-

Risking or boosting masculinity? Mens language choices in multilingual settings


guage choices they make in multilingual settings. The men in these three case studies display some similarities: they are second generation men, they are well-educated, they had some formal instruction in the community language and their self-rated and tested language skills are comparable. Yet they are also markedly different in terms of ethnolinguistic background as they are part of very different ethnolinguistic communities in Australia. The Greek-Australian community is one of the largest ethnic communities in Australia with large concentrations and a relatively dense network along ethnolinguistic lines in major capital cities, especially Melbourne (e.g., Clyne 1991). Census data and research results point towards a community with high levels of language maintenance even in the second generation in comparison to many other groups. The Dutch and German communities in Australia are almost at the other extreme of the language maintenance spectrum. The Dutch community regularly tops the list of ethnolinguistic communities with the highest rate of language shift in both rst and second generation. The German-speaking community also shows high rates of language shift in both generations but less so than the Dutch community (e.g., Clyne 1991). The language proles of these three men are hence not at odds with those characteristic of their ethnolinguistic communities, although Roelof s case is probably the most atypical. However, the focus in this paper is not on how representative they are of their ethnolinguistic communities or on drawing comparisons between the men but on exploring and shedding light on how their performances of masculinity shape as well as are constitutive of their language choices. 6.1. Men and language use in the home: doing lial masculinity The three men grew up in families where the community language had a central place. It is very likely that their pre-school language experience was predominantly in that language. Even Roelof, who has four older siblings, is likely to have heard and spoken mainly Dutch. Upon school age English makes its presence felt in the lives of these men, not only in the school and outside the home but they also bring it home. For Con and Roelof, English is a lingua non grata in the home, at least when speaking to the parents, whereas Maxs parents seem to accept the entry of English in the home. Within the home environment the lial side of their masculinity takes prominence and that side requires choosing and using the language preferred by the parents. Both Con and Roelof, even Max but to a lesser extent, accept this as part of their role and duty as sons of immigrant parents. Furthermore, this environment is also relatively secluded and hence a safe one in language terms: speaking Greek, German or Dutch in the home amongst family members does not pose risks to the preferred linguistic perfor-


Anne Pauwels

mances associated with projected masculinities outside the home. For Max who repeatedly mentions that he does not feel condent about using German, the home seems to be the only environment in which he dares use German without the threat of being exposed as not competent. Cons home environment is probably the most conducive to the use of the community language not only because of his parents limited English language competence but also because the language associated with his lial duty enhances his masculinity within a wider network of Greek friends and relatives including his peers. For Roelof, speaking Dutch with his parents at least when he was living in Australia is an expression of being dutiful. This duty seems to fade when he is no longer in Australia and his contact is via the telephone. Doing lial masculinity involves the use of the community language to ones parents, at least in the home. Although their preference may be to speak English to them for whatever reason, they tend to comply with the wishes of their parents and speak Dutch, Greek or German respectively. When doing lial masculinity does not intersect or interfere with other performances of masculinity that demand other language choices, these three men abide by the language associated with the home. 6.2. Mens language choices in the peer group: risking masculinity or boosting masculinity? All three men have siblings as well as friends/relatives who belong to their peer group and with whom they share an ethnolinguistic background. Doing masculinity in these contexts is not bound by lial duties that demand the use of the community language. Yet these peers share common or similar experiences of language, culture, traditions and values associated with their migrant parents. It is foremost in interactions with these peers that questions of identity are explored: for many members of the second generation language choices are central to the performance and expression of hybrid gendered identities or hyphenated belongings (e.g., Greek-Australian, Dutch-Australian, GermanAustralian). Outside the home the Australiancomponent of their hybrid identity takes precedence, linguistically marked by the use of English as the preferred language including for communication among friends and other peers sharing the same ethnolinguistic background. Using the community language within these contexts is a marked choice that problematises the complexity of their gendered hybrid identities. For Roelof and Max the use of Dutch/German in these encounters poses a risk to their preferred projection of masculinity, which entails displays of competence including linguistic ability in the community language. For example, Max not only avoids speaking German with his friends

Risking or boosting masculinity? Mens language choices in multilingual settings


because he does not feel condent, but he even makes a conscious effort not to speak German with his best friend who is a uent speaker. Despite the closeness of these two friends, the alleged linguistic superiority of his friend silences or mutes Max in speaking German. Although he does not avoid situations involving peers who are German speakers, he ensures his masculinity is not diminished by choosing English as his language of choice. This is also conrmed in his later interactions with his sister, who wishes to use more German to him but he refuses. Linguistic competence also marks Roelof s preferred masculine projection in peer groups. Despite receiving positive feedback about his Dutch language prociency from his relatives (aunt, uncle and initially his cousins), he seems increasingly intimidated by the linguistic condence his cousins display in their use of English. As a result of this Roelof starts losing condence in his ability to speak Dutch to his peers and retreats from the use of Dutch in these encounters. His withdrawal from Dutch is motivated by a perceived loss of masculine status that is intertwined with notions of linguistic competence. His retreat into the linguistic comfort zone of English is a way of restoring a positive masculine identity. The restoration is aided by the fact that his Dutch cousins and friends like to improve their English. His status as a native speaker of English imbues him with linguistic authority leading to a positive masculine identity. In both Maxs and Roelof s case the muting or silencing is more pronounced in the context of male peers, suggesting that demonstrating or displaying linguistic competence in the community language is a critical aspect of performing masculinity in the company of other bilingual men, especially during adolescence and young adulthood. Cons language choices and practices in peer group interactions are in marked contrast to those of Max and Roelof. Although English also occupies the status of default, even preferred language for communicating with Greek-Australian peers, Con clearly relishes opportunities in which he can improve his Greek. Unlike Roelof and Max, he is not intimidated by (male as well as female) friends, visitors and colleagues superior linguistic uency in Greek and welcomes engaging with them in Greek to improve his own prociency. Cons preferred projected masculinity in these settings also contains an element of competence, yet it is constructed as a desire to improve prociency rather than as a display of superior prociency. Hence using Greek, even if there is no perfection, and seeking ways to improve and practice is not a threat to his projected masculinity in these settings. In fact, these linguistic practices may boost his status as a Greek-Australian male, not only in his own eyes but also in those of his peer group as well as parents. Conversely, they do not undermine his status as an Australian-Greek male as his competence in English is beyond doubt.


Anne Pauwels

7. Men, masculinities and bilingualism Case studies are designed to allow for more in-depth, localised and individualised insights into particular forms of behaviour, practices and actions. These case studies have indeed revealed how Con, Roelof and Max live their bilingual lives and how language choices and practices shape and are shaped by their contextualised masculinities. Although all three men were judged to have sufcient prociency and uency in the community language to allow for meaningful interactions with community language speakers, they choose to manage this linguistic resource in different ways. Con engages in practices which are likely to boost his knowledge of the community language. Max and Roelof impose severe restrictions on their use of the community language. This is especially marked in the case of Roelof who moves to the Netherlands and is thus immersed in a Dutch language environment. Unlike Max and to some extent Con, Roelof has plenty of opportunities to boost this linguistic resource, yet curbs his use signicantly, restricting it to interactions with older relatives and occasionally his wife. Underlying this behaviour by Roelof and Max seems to be a fear of exposing linguistic limitations in the company of (male) peers whose linguistic competencies are considered superior. I contend that for Max and Roelof a display of linguistic competence in German and Dutch respectively is a critical component of their masculine performance in such peer group settings. As exposing limitations in their linguistic competency is perceived as damaging their desired projection of masculinity, they retreat from the use of the community language to the safety of English. This was clearly voiced by Roelof when he mentions being granted guru status by his cousins and their friends as a native speaker of English. Cons behaviour is in marked contrast to that of Roelof and Max, as he relishes opportunities to display his knowledge of Greek in the peer group and seeks out ways of improving his competence in Greek through interactions with much more procient speakers. It seems that for Con, using Greek (however imperfect) with ones peers not only expresses the recognition of a shared identity but positively boosts his masculinity in the peer group. Of particular relevance to further study of gendered behaviour in bilingual settings is the pivotal role the peer group plays in inuencing linguistic choices and practices. Although all three men have now reached middle age, their current linguistic behaviour has been shaped by the choices they made in adolescence and young adulthood. The latter in turn were signicantly inuenced by the role of language and language choice in the performance of their masculinities. In terms of language maintenance, it is likely that Roelof and Max may nd it increasingly difcult to nd safe settings for their use of Dutch/German, especially with the passing of the rst generation. With time this lack of use is

Risking or boosting masculinity? Mens language choices in multilingual settings


likely to have an impact on prociency, thus further reducing the chances of language maintenance. The performance of their desired masculinities have had a negative effect on their bilingual practices. Conversely, Cons willingness to use Greek in a wide variety of settings and to seek improvement from more competent speakers undoubtedly puts him in a much better position to pass on the language to the next generation.

Bailey, B. 2000. Language and negotiation of ethnic/racial identity among Dominican Americans. Language in Society 29. 555582. Bennett, E. Jane. 1990. Attitudes of the second Dutch to language maintenance and ethnic identity. PhD Thesis, Monash University, Australia. Burton, Pauline, Ketaki Kushari Dyson & Shirley Ardener (eds.) 1994. Bilingual women: Anthropological approaches to second language use. Oxford/Providence: Berg. Butler, Judith. 1990. Gender trouble: Feminism and the subversion of identity. NewYork: Routledge. Butler, Judith. 1993. Bodies that matter. New York: Routledge. Callan, Victor & Cynthia Gallois 1982.Language attitudes of Italo-Australian and GreekAustralian bilinguals. International Journal of Psychology 17. 345358. Cameron, Deborah. 1997a. Theoretical debates in feminist linguistics: Questions of sex and gender. In Ruth Wodak (ed.), Gender and discourse, 2136. London: Sage. Cameron, Deborah. 1997b. Performing gender identity: Young mens talk and the construction of heterosexual masculinity. In Sally Johnson & Ulrike H. Meinhof (eds.), Language and masculinity, 4764. Oxford: Blackwell. Carr, Jo & Anne Pauwels 2005. Boys and language learning. Real boys dont do languages. London: Palgrave Macmillan. Cavallaro, Francesco. 1997. Language dynamics of the Italian community in Australia. PhD Thesis, Monash University. Cheshire, Jenny, S. Fox, P. Kerswill & E. Torgersen 2008. Ethnicity, friendship network and social practices as the motor of dialect change: Linguistic innovation in London. Sociolinguistica 22.123. Clyne, Michael. 1991. Community languages. The Australian experience. Cambridge: Cambridge University Press. Coates, Jennifer. 2003. Men talk. Oxford: Blackwell. Connell, Robert W. 1995. Masculinities. Oxford: Polity Press. Eckert, Penelope. 1989. Jocks and burnouts: Social categories and identities in the high school. New York: Teachers College Press. Eckert, Penelope. 2000. Linguistic variation as social practice: The linguistic construction of identity in Belten High. Oxford: Blackwell.


Anne Pauwels

Eckert, Penelope. 2004. Adolescent language. In Edward Finegan & John R. Rickford (eds.), Language in the USA: Themes for the twenty-rst century, 360375. New York: Cambridge University Press. Eckert, Penelope &Sally McConnell-Ginet. 1992. Think practically and look locally: Language and gender as a community-based practice.Annual Review ofAnthropology 21. 461490. Gal, Susan. 1978. Peasant men cant get wives: Language and sex roles in a bilingual community. Language in Society 7.1. 117. Gal, Susan. 1991. Between speech and silence: The problematics of research on language and gender. In Micaela Di Leonardo (ed.), Gender at the crossroads of knowledge, 175203. Berkeley: University of California Press. Goldstein, Tara. 2001. Researching womens language practices in multilingual workplaces. In Aneta Pavlenko, Adrian Blackledge, Ingrid Piller & Marya Teutsch-Dwyer (eds.), Multilingualism, second language learning and gender, 77102. Berlin: Mouton-De Gruyter. Grifn, Christine 1989. Review of Brod, Kimmel and Connell. Feminist Review 33. Autumn. 103105. Heller, Monica (ed.) 2007. Bilingualism: a social approach. Houndsmills, Basingstoke: Palgrave MacMillan. Hewitt, Roger. 2003. Language, youth and the destabilisation of ethnicity. In Roger Harris & Ben Rampton (eds.), The language, ethnicity and race reader, 188198. London: Routledge. Holmes, Janet. 1993. Immmigrant women and language maintenance in Australia and New Zealand. International Journal of Applied Linguistics 3.2. 159179. Holmes, Janet & Miriam Meyerhoff (eds.) 2003. The handbook of language and gender. Oxford: Blackwell. Hugo, Graeme, D. Rudd & K.Harris 2003. Australias diaspora: Its size, nature and policy implications. Canberra: CEDA Information Paper No 80. Ingram, David & Elaine Wylie 1979. Australian Second Language Prociency Ratings. Mimeograph. Johnson, Sally. 1997. Theorizing language and masculinity: a feminist perspective. In Sally Johnson & Ulrike H. Meinhof (eds.), Language and masculinity, 826. Oxford: Blackwell. Johnson, Sally & Ulrike H. Meinhof (eds.)1997. Language and masculinity. Oxford: Blackwell. Kiesling, Scott F. 1998. Variation and mens identity in a fraternity. Journal of Sociolinguistics 2:1. 69100. Labov, William. 1972. Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press. Lo Bianco Joseph. 2007. Contrasting and comparing minority language policy: Europe and Australia. In Anne Pauwels, Joanne Winter & Joseph Lo Bianco (eds.), Maintaining minority languages in transnational contexts, 78104. Houndmills, Basingstoke: Palgrave Macmillan.

Risking or boosting masculinity? Mens language choices in multilingual settings


McElhinny, Bonnie. 2003. Theorizing gender in sociolinguistics and linguistic anthropology. In Janet Holmes & Miriam Meyerhoff (eds.), The handbook of language and gender, 2142. Oxford: Blackwell. Maher, John C. 2005. Metroethnicity, language and the principle of Cool. International Journal of the Sociology of Language 175176. 83102. Ozolins, Uldis. 1993. The politics of language in Australia. Cambridge: Cambridge University Press. Pauwels, Anne. 1995. Linguistic practices and language maintenance among bilingual women and men in Australia. Nordlyd 23. 2150. Pauwels,Anne. 1997. The role of gender in immigrant language maintenance inAustralia. In Wolfgang W lck & Annick De Houwer (eds.), Recent studies in contact linguistics, o 276286. Bonn: D mmler Verlag. u Pavlenko, Aneta, Adrian Blackledge, Ingrid Piller & Marya Teutsch-Dwyer (eds.), Multilingualism, second language learning and gender. Berlin: Mouton-De Gruyter. Piller, Ingrid & Aneta Pavlenko 2001. Introduction: Multilingualism, second language learning, and gender. In Aneta Pavlenko, Adrian Blackledge, Ingrid Piller & Marya Teutsch-Dwyer (eds.), Multilingualism, second language learning and gender, 116. Berlin: Mouton-De Gruyter. Pujolar, Joan. 2001. Gender, heteroglossia and power. Berlin: Mouton-De Gruyter. Rampton, Ben. 1995. Crossing:language and ethnicity among adolescents. London: Longman. Rampton, Ben. 2006. Language in late modernity: Interaction in an urban school. Cambridge: Cambridge University Press. Rubino, Antonia. 2006. Linguistic practices and language attitudes of second generation Italo-Australians. International Journal of the Sociology of Language 180. 7188. Teutsch-Dwyer, Marya. 2001. (Re)constructing masculinity in a new linguistic reality. In Aneta Pavlenko, Adrian Blackledge, Ingrid Piller & Marya Teutsch-Dwyer (eds.), Multilingualism, second language learning and gender, 175198. Berlin: Mouton-De Gruyter. Tsokalidou, Roula. 1994. Cracking the code: An insight into code-switching and gender among second generation Greek-Australians. PhD Thesis, Monash University, Australia. Trechter, Sara. 2003. A marked man: The contexts of gender and ethnicity. In Janet Holmes & Miriam Meyerhoff (eds.), The handbook of language and gender, 423 443. Oxford: Blackwell. Winter, Joanne &Anne Pauwels 2005. Gender in the construction and transmission of ethnolinguistic identities and language maintenance in immigrant Australia. Australian Journal of Linguistics 25.1.153168. Winter, Joanne & Anne Pauwels 2006. Language maintenance in friendships: Second Generation German, Greek and Vietnamese migrants. International Journal of the Sociology of Language 180. 123139. Winter, Joanne & Anne Pauwels 2007. Language maintenance and the second generation: Policies and practices. In Anne Pauwels, Joanne Winter & Joseph Lo Bianco (eds.),


Anne Pauwels

Maintaining minority languages in transnational contexts, 180200. Houndmills, Basingstoke: Palgrave Macmillan. Anne Pauwels is Professor of Sociolinguistics and Dean of the Faculty of Languages and Cultures at the School of Oriental and African Studies, London. Her research deals with the social and sociolinguistic aspects of language and communication, with particular attention to multilingual and transnational settings. Her main research foci include multilingualism, language maintenance/shift, language policy in relation to language learning in schools and universities as well as various aspects of the relationship between gender and language. Email:

The multiple faces of multilingualism: Language nominations


Abstract Language nominations are appellations commonly assigned by language users to their languages, such as mother tongue, foreign language, heritage language, languages of new minorities. They are the markers of the diversity of multilingualism, indicating both the linguistic and societal characteristics of languages and the perspectives of their users. The changeability of nominations over time reects the evolution of sociolinguistic situations within local and global communities, including attitudinal development, regarding the range of languages sharing particular environments (language ecology). Cumulatively, language nominations can be regarded as a naturally evolving classication reecting the diversity of the functions and status of languages in the contemporary world a classication that provides useful tools for monitoring the dynamics of multilingualism in particular communities and within a specic timeframe. This paper introduces and describes the notion of language nominations and shows the roles that specic language appellations may play for multilingual individuals and multilingual societies. It also demonstrates how language nominations can be used in research on multilingualism. Key-words: Multilingualism; language nominations; liminality, identity, language and timespace, linguistic diversity, mother tongue

1. Introduction This paper has two principle aims: one is to present and describe the notion of language nominations and demonstrate the role which specic language appellations play for multilingual individuals and multilingual societies. The other aim is to show how language nominations can be used in research on multilingualism and in multilingual practices.


Larissa Aronin, Muiris O Laoire and David Singleton

The new linguistic dispensation (Aronin and Singleton 2008) is characterised by a vast diversity of populations and multiple languages in space and time. Diversity and identity have become especially active foci of investigation in the study of languages and society. The recognition and celebration of diversity go hand in hand with what Riley called identity rush, (Riley 2010: 376). Multiple and diverse identities are recongured, recovered and rejected (Riley 2010: 376) against the background of a myriad of language repertoire congurations and patterns of language use. Each language in a repertoire plays a particular role unique to the individual involved. Language roles for individuals only very loosely coincide with the allocation of roles which a community may give to its languages. What in certain formal contexts is referred to as ofcial language may be the mother tongue for some and a third language for others.The great range of language role allocations contributes to what can be called multiple faces of multilingualism.1 The face metaphor includes many aspects which are important in multilingualism.2 The meaning of the word face includes nuances such as appearance, facade, side, dimension, and also self-respect and respect, reputation, dignity, as well as honour, prestige, image status, position. These we easily recognize as being hot spots of current research on language and society. The metaphor of multiple faces appositely captures the nature of multilingualism in that it represents its manifold nature, the possibility of outward and inward contents and the necessity to reveal its heretofore undiscovered depths. Some faces are not easily seen, and they may take considerable time and scholarship to discern. Some faces of multilingualism come to the surface later than others. Elsewhere we have pointed out that many language-related processes and phenomena have, of late, become especially visible owing to recent societal shifts and changes, in particular to those in the domain of language use. In other words, under the current sociolinguistic dispensation, issues which previously were impossible to single out, are now becoming apparent (Aronin and Singleton 2008; Aronin and Hufeisen 2009). We have called this quality of contemporary multilingualism liminality. Figure 1 presents our proposed outline of current developments and properties of multilingualism, among the three basic properties of which we place liminality. In what follows we suggest that language nominations reect a range of perceptions regarding specic languages on the part of those who speak them and on the part of those who may appraise the languages and their speakers. First, we dene and describe the concept of language nominations. Then we show how language nominations function de facto for individuals and communities. After suggesting a classication of language nominations, we outline some possibilities relative to the use of this concept in research. The nal section serves

The multiple faces of multilingualism: Language nominations


Figure 1. The Properties and Developments of the Current Global Linguistic Dispensation (Aronin and Singleton 2008; Aronin and Hufeisen 2009: 105)

to illustrate such a use, drawing on an empirical study by Laoire and Aronin (2005) as an example.

2. Language nominations The concept of language nominations refers to language appellations such as mother tongue, second language, foreign language, ofcial language, majority language, minority language, etc., which are assigned to various languages according to their perceived role and value for an individual or a community ( Laoire and Aronin 2005). Language nominations are terms that have been traditionally attached to languages used in society which necessarily accompany or replace the proper names of particular languages (e.g. English, Turkish, Urdu, etc which, of course, are also language nominations).


Larissa Aronin, Muiris O Laoire and David Singleton

Language nominations, however, are more than just labels. They clearly carry emotional and functional connotations and they point to the political and societal status of given languages. In addition to carrying strong subjective connotations, language nominations advert to the value and role currently assigned to a given language by a particular society/community. For example, mother tongue carries the connotations of origin, permanency and inseverable emotional ties, while second language and foreign language, on the other hand, connote distance between the inner world of the speaker and his/her languages. Such terms may also convey subjective evaluation and/or information about the place that a society allocates to a specic language in a specic place and period of time. The genesis of language nominations is twofold. In part, their creation is the work of ordinary language users i.e. is bottom-up in nature, but their origins may also be characterized by a degree of top-down imposition. Appellations in common usage often (but not always) coincide with the terminology of the language sciences, e.g. native language, foreign language. Some such nominations come into circulation naturally by way of popular coinage, on the basis of a shared popular understanding of the designated phenomenon rather than on the basis of the decisions of some academic or political body. Other language nominations are consciously and carefully coined by professionals involved in language research, language policy, language teaching, etc. Thus, in the USA language professionals observing the upsurge of interest by the grandchildren of immigrants in languages of their forebears felt compelled to construct and dene the term heritage language (Kagan 2005; Wiley 2001; Vald s 2000). e Naturally evolving language nominations are vital in individual language acquisition.Very telling evidence of the signicance of language nominations for individual language acquirers comes from Barron-Hauwaert (Barron-Hauwaert 2004). She researched the use of languages in trilingual families, her sample coming mostly from Switzerland and Belgium. Barron-Hauwaert writes:
With three languages, labeling each language is necessary. Half the sample of children used the proper name (i.e.French or English) while the other half used personalized labels such as Mummys/Daddys {language, Mama spricht Deutsch [Mama speaks German] or Papa dit ca en fran ais [papa says it in c French]. Some parents reminded the child of the location of a language: A lcole on parle le fran ais [at school we speak French], or that English is Australias c language (as they previously lived in Australia). The author further notes that she assumed There was a correlation with age, [. . .] as a younger child would say Mummys language and an older one English. (Barron-Hauwaert 2000: 7)

The multiple faces of multilingualism: Language nominations


Families may use a personalized label when the child begins to talk, such as Mummysays . . . to help the child label the phenomenon of the mother and the father speaking differently. Later, when the child has mastered two languages, use of the proper name is more appropriate (Barron-Hauwaert 2000: 7). This is an example of natural evolving language nominations, but more importantly, it shows an important role for nominations in individual early language acquisition. It can be said that language nominations have a reective function since they reect and mirror the current state of the attitude to and role of each language in the picture. Languages have symbolic value in society and very often a language nomination is an expression of that value. In the days of the Soviet Union, for example, the Russian language was the language of interethnic communication. After the dissolution of the USSR it is still used as a lingua franca by speakers and organizations in business and political negotiations. In many of these states Russian is a minority language while in others it has been sidelined and nominated as a foreign language, and as such now competes for a place in the school curriculum with English, German or French. Being aware of language nominations also allows for thinking about these roles in a particular time and in a particular discourse. In modern times we increasingly meet the nomination additional language(s) used where previously foreign language, second language, third language, etc. would be used. The term additional situates the language within a set of values where many languages are seen as enjoying more or less equal status. Using the rather neutral term additional perhaps implies that the situation of using more than one language is not unusual or strange. It can be argued that there exist some individuals and communities who do not know which language they use. Such comments usually refer to the cases when a given set of speakers and those outside their particular community of speakers apply different proper names to the same language, as in the case described in Mesthrie et al. (2000: 10) in connection with the fuzzy boundaries between languages in Papua New Guinea.
The language spoken in Bolo village is also from a linguists point of view identical to Aria, but Aria speakers from other villages say it is not Aria. They say Bolo speakers really speak Mouk. However, the people of Salkei village, who speak Mouk, say that Bolo people speak Aria. As for the Bolo speakers themselves, they claim to be An m speakers (Romaine 1994: citing Thurston 1987). If this e were not complicated enough, the An m people of another village do not think e that the Bolo speak acceptable An m any more e (Mesthrie et al. 2000: 10)


Larissa Aronin, Muiris O Laoire and David Singleton

The above-described situation demonstrates how languages identical from the linguists point of view are perceived as self-evidently different by different communities speaking them and acquainted with them. The difference in nominations (or mis-nomination from someone elses point of view) demonstrates the social processes of inclusion-exclusion, defying or accepting ethnical links. Janson (2002) notes that the naming of a language is a crucial step in its emerging identity: [a] language without a name does not exist (Janson 2002: 230). This is, then, a further example of language nominations functioning to reect language status and attitudes. When we name a language with any of its appellations, frequently we also overtly or covertly evaluate it, whether or not there is any intention of doing so. The nomination of international language is perceived as more weighty than that of lesser-used language, although the latter appellation is a perfectly worthy one. Language nominations, thus, at least to a degree, perform a measuring/rating function. But, mainly, the assessment is not good or bad, rather for orientation on the global-local, close-far scale of languages. New language nominations are continually cropping up in the relevant literature, arising out of research, individual coinage and varying degrees of consensus; recent examples include: pluricentric languages (Clyne 1992), migrant languages, lesser used languages, and LOTE languages other than English. Even if a language nomination is created in a top-down manner, the impetus for its coinage is the societal signicance of the function and role of the phenomenon it designates and the terminological needs of social practice in respect of this phenomenon. The European Charter for Regional or Minority languages illustrates this point, abounding in terms and denitions which are seen to be required by the situation it addresses: a. regional or minority languages means languages that are: i. traditionally used within a given territory of a State by nationals of that State who form a group numerically smaller than the rest of the States population; and ii. different from the ofcial language(s) of that State; it does not include either dialects of the ofcial language(s) of the State or the languages of migrants; b. territory in which the regional or minority language is used means the geographical area in which the said language is the mode of expression of a number of people justifying the adoption of the various protective and promotional measures provided for in this Charter; c. non-territorial languages means languages used by nationals of the State which differ from the language or languages used by the rest of the States pop-

The multiple faces of multilingualism: Language nominations


ulation but which, although traditionally used within the territory of the State, cannot be identied with a particular area thereof. (European Charter for regional or minority languages 1992 Treaties/Html/148.htm.) The interest in minority languages, which has become especially intense in recent decades, brings about the need to be more specic and to acknowledge and consider ner distinctions in the roles and vitality of minority languages in different circumstances. A number of terms have been proposed, discussed and used (sometimes in very specic contexts) by researchers thus: regional languages, local languages; unique languages, autochthonous languages. In a very different context the nomination titular languages is increasingly entering the general discourse following the increase of information and research on the post-Soviet countries. It is often the case that a more nuanced status of a language in a community is signalled by being associated with new labels and by a discussion of the relative merits of, and the differences between the nominations in question. An example of a scholarly attempt to further explore already numerous roles and appellations in respect of English is the study recently undertaken by Cook (2008) who analyses two language nominations of English English as a Lingua Franca (ELF) and English as a Second Language (ESL). Any meaningful change or particular need in language-related activity such as, for example education, can trigger a discussion of nominations for the languages involved. In fact, clarifying and specifying the nominations of languages involved may be the rst necessary step in language planning. Many specic sociolinguistic situations, be they in education, culture or politics, require deliberation of nominations or around nominations as a pre-requisite for planning or understanding the situation. In education, for example, it helps to establish clear goals and to design effective curricula suitable for every kind of learner. We offer here two examples where any investigation of the educational systems in question inevitably requires a pre-determination of the pertinent nominations as a point of departure. The rst comes from the Basque Autonomous Community (BAC) in Spain, and the second refers to teaching Irish in Ireland. In her examination of education through Basque in the BAC, Cenoz (2009: xiii) begins with what can be interpreted as an attempt to nd the proper nomination for the languages used in the Basque Autonomous Community by weighing the specic circumstances of the Basque, Spanish and English languages against the denitions of heritage, second and foreign languages. Such a tting and matching process, actually trying out the proper nominations for languages,


Larissa Aronin, Muiris O Laoire and David Singleton

turns out to be indispensable for working with the realities of the Basque educational system and is in fact a common prerequisite for dening policies and designing curricula in respect of any educational system.
Language learning in the Basque educational system is linked to research on heritage, second and foreign languages as dened by Kramsch (2007:5). In fact Basque can be considered a heritage language learned by members of an ethnic group desirous to reconnect with the culture of their ancestors because it is learned by Basque L1 speakers and by Basque students who no longer speak Basque and learn it at school. At the same time, speakers of Basque as a rst language learn Spanish as a second language dened as a language other than the mother tongue learned in an environment in which that language is the dominant language or where the language is an international language of commerce and industry. All students in the Basque Autonomous community also learn a foreign language understood as a language that is learned in an instructional environment or during a temporary sojourn abroad as part of general education or for professional purposes. (Cenoz 2009: xiii).

Further, Cenoz is aware of the fact that the existing nominations do not exactly cover the current sociolinguistic situation and that it is difcult to t educational and sociolinguistic realities within strict boundaries and there is overlap between the scope of heritage, second and foreign languages (Cenoz 2009: xiii). She explains that:
. . . the spread of Basque as the language of instruction and the new multilingual and multicultural situation in the Basque Country have developed into situations which do not t completely into the three categories of heritage, second and foreign language. It is very common for speakers of Spanish as a rst language and for immigrants who speak other languages to be instructed through Basque even if this was not the language used by their ancestors. These situations cannot be considered heritage language learning in a strict way but it is difcult to consider them second language because Basque is not the dominant language or an international language and it is not a foreign language either because it [is] an ofcial language in some parts of the Basque Country. (Cenoz 2009: xiiixiv)

Laoire (2010: 231249) discussing the approaches to teaching Irish in the past, shows that nominations are instrumental in dening approaches to teaching the language. Irish, an autochthonous language, is now taught communicatively using methods adopted from foreign language pedagogy. Irish is a native language for some (three percent of the population), an ofcial language for all, a second language for the majority of learners, and a foreign language for learners from the immigrant and migrant community. But it is not the kind of heritage lan-

The multiple faces of multilingualism: Language nominations


guage that is taught, say in the USA, in the case of Chinese or Spanish speakers, because in Ireland Irish is taught to all students. In order to assess the teaching of Irish and draw up future plans in an informed way, the discussion must inevitably be centred on the situation, role and status of the language for its various sets of learners. Each language nomination for Irish thus necessitates a differentiated curriculum and pedagogical approach. Thus, language nominations are indispensable and need to be used when dealing with (planning, evaluating, analysing) education through specic languages or language education. In addition to the reective and assessment functions, language nominations perform another important function: they can modify or exert some inuence on identities and groups of language users. We have already discussed above the formative/ important role of language nominations in the process of acquiring two or more languages by a child in a family. At times of major political events, ethnic turbulence, or any disquiet or commotion, for instance, the public appellation (nomination) for a certain language can produce a tangible impact on the outcomes of events and the social situation. A language nomination can lead to greater respect and use of the language, or it may have an opposite diminishing effect. A language nomination, in fact, can guide people towards dealing in a particular manner with a given language or towards a particular set of attitudes in respect of a particular language and its users. For both individuals and groups of people language nominations are instrumental in negotiating their identity and status. Since nominations indicate the status of a language in a society, they are very often the subject of struggle and negotiation activities or a long period of political lobbying. A well known example is the case of the Modovan-Romanian language (see for example in Alpatov 2000: 185). The language act of 1989 nominated the state language of Moldova as Romanian. Then a struggle took place between radical nationalists oriented to Romania and the more moderately inclined parties who preferred the term Moldavian language. In 1994 the then president Snegur managed to include an article in the new Moldavian constitution using the nomination Moldovan language. In 1995 a student demonstration ensued in which one of the main demands was a return to the previous Romanian language nomination. About a year later the Speaker of the parliament, Luchinsky, suggested a compromise: Moldovan (Romanian) language which was also rejected. Language nominations are instrumental in negotiating individual identity. Two opposite scenarios are known to be used to ascertain ones identity: either subscribing to, or defying certain language nominations. In the rst case individuals link themselves or demonstrate their links with a language which is prestigious or important to them in some way. Many such examples are reported in discussions of census results, where immigrants nominate the ofcial/national


Larissa Aronin, Muiris O Laoire and David Singleton

language of a host country as their native language. People who wish to assimilate or not be different from the general population of a country or region report themselves as the native speakers of the majority language. The other side of the coin is that some language speakers for a variety of reasons do not admit their rst language as their mother tongue. We have seen in the previous section that language nominations are used de facto, whether consciously or unconsciously, by individuals in order to negotiate, and stabilize their identities as well as by societal agencies in order to process languages in education, politics and culture.

3. Language Nominations: the potential of the use of this concept We argue here that the concept of language nominations can be fruitfully used in research and in practice (see also Aronin and Singleton 2010). Language nominations are numerous and diverse, and they perform a variety of functions. We propose here a classication of language nominations into four main categories: 1. Language proper names, such as French, Tamil, Armenian. 2. The continuum of close far, covering the range between the poles of mother tongue and foreign language. 3. The complex hierarchical continuum of social-geographic dimensions. e.g. international (implies simultaneously of world-wide spread and importance) and regional languages (less geographical spread and of less importance than international, but more than lesser-used, tribal languages). 4. Emotional-identity scale, including such nominations for languages as heritage, home, ethnic and community. Each of these nominations hints at the proximity of a given language to the speaker revolving around his/her permanent and temporary identity characteristics, such as his/her past, emotions and attitudes, interests and predilections, needs and career. It can be easily seen that these four axes along which one can situate language nominations overlap. A language may be seen from the angle of career needs, while simultaneously referring to the identity situation (for example, for a German-speaking wife of an Irishman living in Ireland, English is probably the language she works in, and the language in which she communicates with close friends and colleagues) (axes 2 and 4); Alternatively, a language can be seen from the perspective of a particular hierarchical continuum for example, language appellations of various weights/importance between international and lesser used languages. A language and its importance can be categorised and studied from the point of view of any one of these axes, ignoring the other

The multiple faces of multilingualism: Language nominations


categories for a particular situation and particular research needs. Usually researchers take one axis at a time, when considering the specic needs of their analysis. It is also clear that one and the same language can receive more than one nomination from its speakers or outside evaluators. Languages differ in the number of nominations that are attached to them. For example, the French language, French labelled by a proper name nomination, (axis 1) is also an international language (axis 3), since it is one of the most widely used for business, culture and tourism communication along with English, German and Spanish; it is a foreign language for pupils and students in Ireland, Russia, the UK, and many other countries (axis 2), a heritage language, a colonial language for Algeria (axis 2), a minority language among the Lebanese immigrant community in London and a majority language in France. Unlike French, the Chuvash language (a Turkic language spoken in central Russia primarily in the Republic of Chuvashia and adjacent areas) does not have so many nominations. It is a minority language in some regions of Russia. The number and the exact number of nominations involved may be used as a kind of rating of a particular language. In this way, for example, we can say that the more nominations a language has, the more functions it performs, the more active it is in the global or local scene, depending on whether it is associated with more global nominations or with locally oriented nominations. Deliberating over given data on how people nominate their languages realizes only part of the potential of the concept of language nomination. We claim here that more targeted use of the phenomenon of language nominations can be highly benecial in research contexts. Language nominations do not always remain static in the face of time and space changes. With regard to the geographic spread of languages, particular appellations are not permanently attached to particular languages (Aronin and Singleton 2010). For example, German changes its nomination of ofcial language to that of minority language as one travels across the border from Germany to Italy, and Hungarian undergoes the same shift in appellation in the transition from Hungary to Slovenia (see Cenoz and Gorter 2008). Similarly, the nomination lingua franca is aptly attached to Hindi in some states of India, but this language could not be appropriately labelled as a lingua franca throughout the entirety of the country (Kachru et al. 2008). The changeability of nominations over time is remarkable. It reects the evolution of language situations within particular and global communities. In former times the terms native language, mother tongue and rst language denoted the rst language spoken as a child, the language of ones family and ones home and the language of the wider community in which one lived. These days,


Larissa Aronin, Muiris O Laoire and David Singleton

we notice that it is by no means unusual for a person to take some time and to think very hard before responding to the question what is your native language? Increasingly, popular usage applies such terms to a persons strongest language or the language considered by the individual as his/her most important language, whether or not this was the language originally encountered as the parental or home language. Moreover, whatever the order of acquisition, the language perceived as the rst language may or may not be the language of the wider community. In addition to further clarifying and adding new language nominations, researchers replace nominations which, in their opinion, are not suitable any more, in the sense that that they do name the phenomenon adequately. In introducing a new approach to multilingualism, Herdina and Jessner (2002) have chosen the less conventional nominations of primary, secondary and tertiary languages rather than the more commonly used rst language, second language and third language. As time passes by, and the liminal aspects of multilingualism become ever more manifest, the need to change language nominations is felt by some researchers to be increasingly urgent. In 2009 Hammarberg devoted his paper in the 6th International Conference on Third Language Acquisition and Multilingualism in Bolzano, Italy entirely to the issue of the basic language nominations. In a later article (Hammarberg 2010), Hammarberg questions traditional appellations of non-rst language learners in favour of other nominations which he suggested. His essential argument is that the problems with the expressions rst, second and third language have become more apparent with the emergence of research on L3 acquisition. The recent focus on L3 acquisition means that one has begun taking the complexity of multilingual language learners language background into account. He than continues: This gives rise to reection about some of the currently used basic terminology in the eld, in particular how the concepts rst, second and third language are understood. (2010: 91). Hammarberg stresses the need for reconsideration and clarication of the concepts L1, L2 and L3 from the point of view of multilingual language users and learners. Hammarberg proposes to change these established terms and suggests primary, secondary and tertiary language as possible replacements. (2010: 91) The increase in the number of nominations and the evident increase in their use in social and scientic discourse, as well as the tendency to make strenuous efforts to clarify and disambiguate the language nominations in question speak volumes about the increasing importance of language nominations in the current language dispensation (Aronin & Singleton 2008). Language nominations can be considered to be sensitive and dynamic diversity markers which enable us to keep a nger on the pulse of the changes in functions and status of any given language in a given society. They may also provide a starting point for various

The multiple faces of multilingualism: Language nominations


Figure 2. Acquisition of languages. The ordering hierarchy (Hammarberg 2010: 101).

kinds of comparisons which emerge from the perceptions of actual language communities, and which are thus less likely to be open to charges of researcher bias. Language nominations can also be used to explore the success or otherwise of a language policy, or to gauge the distance between the desired status of a certain language and its de facto role in a particular society. For example, a language may be granted ofcial status in a given community, but members of that community who speak it as a rst language may not nominate it as their native language if asked to ascribe language nominations to their repertoire of languages.


Larissa Aronin, Muiris O Laoire and David Singleton

We have seen so far that analysing the number and character of nominations in relation to specic languages, comparing nominations across languages, and tracing the changes in language nominations over space and time can be fruitful in sociolinguistic research. Looking closely into a particular sociolinguistic situation through the lens of language nominations can also bring fresh insights. In the next section we show how new ndings can be obtained using language nominations as a tool for exploring the identities of multilinguals.

4. An empirical study of the identity of multilinguals using the language nomination concept This section aims to show the potential of using language nominations as a point of departure in studies especially interested in the linguistic identity of speakers. Language nominations were used as a tool in the research by Laoire and Aronin carried out in Ireland and Israel in 2005.3 The study focused on how multilinguals themselves nominate their languages, with a view to revealing and describing the characteristics of multilinguals. The purpose of the study also included that of understanding the subjective value of each language as opposed to other languages in a particular persons language constellation. The study sample comprised 37 multilinguals, citizens of Israel from autochthonous (n = 5) and immigrant (n = 32) populations and 29 EU citizens temporarily living/studying/working (n = 24)/ in Ireland and immigrants (n = 4). The countries of origin were many: Israel, Argentina, Azerbaijan, Bulgaria, Chile, Israel, France, Morocco, Switzerland, Russia, Ukraine, Moldova, Uzbekistan, Poland, USA, Scotland (UK), and the USSR; in Ireland, France, Finland, Germany, Spain and Sweden. The informants reported using 20 languages in Israel and 19 in Ireland, even though Ireland was not at the time considered to be multilingual. Both vibrant and spreading, pluricentric and non-pluricentric languages were involved, LWC (English, French, Spanish, German) and autochthonous (Hungarian, Moldavian, Bulgarian, Catalan, Valenciano, Berber), modern and heritage (Yiddish), Slavic, Semitic, FinnoUgric, nonIndo-European (Japanese), with corresponding various kinds of scripts and with different script direction from right to left (Arabic and Hebrew) and from left to right (Cyrillic and Roman scripts used in the rest of the languages). This diversity allowed the authors to draw conclusions as to multilinguals (in general) in these particular multilingual settings in Israel and Ireland, and such ndings indeed may be applicable elsewhere. A questionnaire was designed aimed at eliciting information on multilinguals thoughts, opinions, and world outlook in regard to the languages in their

The multiple faces of multilingualism: Language nominations


particular constellations. The data were analysed according to the nominations of languages used by the participants (mother tongue, native language, L1, L2, L3 etc.). To ponder the collected data on language nominations fractal- type images were constructed consonant with the newer frameworks put in place by fast unfolding complexity science. The founder of fractal geometry Benoit Mandelbrot (1989: 7) called it a new geometric language, which is geared towards the study of diverse aspects of diverse objects, either mathematical or natural, that are not smooth, but rough and fragmented to the same degree at all scales (for more on the use of complexity in multilingualism research see Aronin & Singleton 2008a). The canon of research tools has shifted in recent times, moving from traditional methods of data presentation and analysis and now admitting into research methodology images once considered unscientic and irrelevant. Mathematicians once believed that using a visual image for scientic speculation was entirely inappropriate. For this reason Fractal Geometry was initially rejected by the majority. Garcia (1991: 30) comments on this as follows:
The problem does not always lie in the image itself, but with the interpretation of the image. True, those conjectures made by visual observation should ultimately nd support in proofs. However, the amount of research sparked by the visual pondering of such images demonstrates that intuition should initially run unchecked. Though a general distrust of images in the mathematical community has historically conned such benets to the arts, Fractal Geometry is a eld where visual exploration is essential.

The study used fractal-like images with the aim of locating certain ndings which might be overlooked and not come to the fore in a solely quantitative/qualitative approach. There were 20 languages mentioned as being mastered at various levels by the 37 Israeli respondents and 19 languages by respondents in Ireland.The frequency of inclusion of each of these languages in personal language constellations varied (see Figure 3 and 4). Although the 2005 study yielded a variety of results, only the nding directly referring to language nominations is discussed in this article. In Figure 5 the languages of multilingual participants are arranged according to language nominations. The participants were not guided by any predetermined denition of what constituted native language, mother tongue or second language. Analysis focused on understanding the subjective value of each language as opposed to other languages in a particular informants constellation. The Figure illustrates the use of languages by the participants of the study from Ireland and Israel in correspondence with the nominations (marked on the


Larissa Aronin, Muiris O Laoire and David Singleton

Figure 3. Fractal- like images of languages known by multilinguals in Israel. Vertical numbers correspond to the respondents; different shades in grey scale represent languages

horizontal axis) applied to each of their languages (e.g., mother tongue, second language, foreign language) by each multilingual (vertical axis). In the Figure languages are indicated by specic colour, e.g., English orange; German blue; French green; Russian violet, etc. It was found that one language was nominated as often as three times, i.e. one and the same language, e.g., French, German, or Spanish. In Israel, for example, in 24% of cases and in Ireland in 69.9% of cases, one language was given three nominations by its speakers (see Table 1). In both countries when triple nominations were supplied they most frequently included what was considered to be the closest, most emotional nominations (mother tongue, native tongue and L1). It was noted that the image of the languages represented on the left side, those which were nominated in the MNL1 triple nominations (TN) happen to be mostly LWC and/or pluricentric languages. This distribution is very clear in the case of Israel and, while less clear in the case of Ireland, the emergent pattern would tend to support a similar trend. In

The multiple faces of multilingualism: Language nominations


Figure 4. Fractal- like images of languages known by multilinguals in Ireland. Vertical numbers correspond to the respondents; different shades in grey scale represent languages

Ireland with relatively recent newcomers (average 6.4 months) this trend still exists with the German, Spanish and French languages. Other non-widely used languages feature prominently e.g. Finnish and Swedish. In looking at some of the differences between the data of the two countries, one immediately notices that triple nominations are sometimes allocated to languages other than languages of wider communication (Finnish, Swedish). This may be due to a short period of stay in Ireland and respondents intention to leave on completion of their studies. Alternatively, this is perhaps due to participants emotional attachment to either or all M/N/L1 (M- mother tongue; N- Native language; L1). There were cases when one and the same language was given all these three nominations (see Figure 4 and 5) what we called triple nominations. In this context it is worthy of note that two respondents from Finland classied Swedish as L3 even though authorities would conventionally regard Swedish in Finland as either L2 or L1 (Bj rklund 2005), thus underscoring participants distancing o from Swedish.

Larissa Aronin, Muiris O Laoire and David Singleton

Figure 5. Fractal- like images of language representation in multilinguals in Ireland and Israel. Legend for Image 1: Mmother tongue; N- Native Language; L1 First language; L2, L3, Ln successive languages. Vertical numbers correspond to the respondents; different shades in grey scale represent languages.

The multiple faces of multilingualism: Language nominations Table 1. Cases of triple nominations Israel 22 cases 43, 24% 11 cases 69.75% Russian 2 cases 9.09% English 1 case 4.54% Spanish 1 case 4.54% French 1 case 4.54% Bulgarian Ireland 20 cases 69.9% 9 cases 45% German 2 cases 10% Spanish 3 cases 15% Swedish 3 cases 15% French 3 cases 15% Finnish


Pondering the coloured fractal-like representations of the samples in these two sociolinguistic situations yielded the following ndings of regularity: triple nomination, when one language was given triple nominations by all its speakers; most frequent triple nominations include what we consider to be the closest, most emotional nominations; the languages which were nominated in the prime rank nominations are mostly LWC and/or pluricentric languages. In this study multilinguals emerged as being autonomous and self-guided, selfsustained and exible in the way they assign functions and values to their languages. Allocating nominations to their languages, in other words choosing from a range of different possible permutations, they match each language with particular values and practical functions or combine them in accordance with their current perceived needs and current environment and not in any generally pre-expected or pre-determined way.

5. Brief Summary In this article we have introduced the concept of language nominations which is dened as language appellations that are assigned to various languages according to their perceived role and value for an individual or a community. Language nominations deal with the liminality of multilingualism by serving as indicators of the important aspects of use of multiple languages by individuals in society. Language nominations carry out at least four overlapping functions: a reective function, a modifying function, a measuring function (to the extent, of course, available at the current state of scientic development) and a means of studying people and communities using multiple languages.


Larissa Aronin, Muiris O Laoire and David Singleton

Language nominations are markers of diversity, capturing in one naming unit linguistic, social and personal dimensions of a language. The dynamics of language nominations can be seen as marking the changes relative to language in society and to individual language users. In this article we have emphasized the need to monitor the vibrant and dynamic reality of multilingualism in a more informed and attentive manner. As well as shedding light on the role that any named language may play for an individual or speech community, examining language nominations may provide valuable authentic data in future research studies and accounts of multilingualism.

1. This new metaphor was used as a title for a recent workshop held in London: The Multiple Faces of Multilingualism 2425 June, 2010. 2. For more about metaphors in investigating multilingualism see Aronin & Hufeisen 2009 chapter 6: 103120. 3. The study and the ndings were originally presented at the L3 conference in Fribourg, 2005.

Alpatov, Vladimir. 2000. 150 Languages and politics: Socio-linguistic problems of the USSR and the post-soviet space. ( p . 2000. 150 : 19172000. : + ). Moscow: Institute Vostokovedeniia, 1997. Aronin, Larissa & Britta Hufeisen (eds.). 2009. The exploration of multilingualism: development of research on L3, multilingualism and multiple language acquisition. Amsterdam: John Benjamins. Aronin, Larissa & David Singleton. 2008a. Multilingualism as a new linguistic dispensation. International journal of multilingualism 5(1). 116. Aronin, Larissa & David Singleton. 2008b. The complexity of multilingual contact and language use in times of globalization. Conversarii. Studi linguistici 2. 3347. Aronin, Larissa. & David Singleton. 2010. Affordances and the diversity of multilingualism. In Larissa Aronin & David Singleton (guest eds.), [Special Issue The Diversity of Multilingualism]. International journal of the sociology of language (Mouton de Gruyter). 105129. Barron-Hauwaert, Suzanne. 2000. Issues surrounding trilingual families: Children with simultaneous exposure to three languages. In Jasone Cenoz, Britta Hufeisen, Ulrike Jessner (eds.), Trilingualism-Tertiary languages German in a multilingual world

The multiple faces of multilingualism: Language nominations


[special issue]. Zeitschrift f r Interculturellen Fremdsprachenunterricht (Journal of u Intercultural learning in the German language) 5(1). project ejournak/jg 051/beitrag/barron.htm (accessed 1 May 2000). BarronHauwaert, Suzanne. 2004. Language strategies for bilingual families: The-oneparent-one-language approach. Clevedon: Multilingual Matters. Bjrklund, Siv. 2005. Toward Trilingual education in Vaasa/Vasa, Finland. The International Journal of the Sociology of Language 171. 2340. Cenoz, Jasone. 2009. Towards multilingual education: Basque educational research from an international perspective. Bristol: Multilingual Matters. Cenoz, Jasone & Gorter Durk. (eds.). 2008. Multilingualism and minority languages: Achievements and challenges in education. AILA Review 21. Amsterdam: John Benjamins. Cook, Vivian. 2008. ELF: Central or atypical second language acquisition? Paper presented at AILA (International Association of Applied Linguistics) Congress, Essen, Germany, 2429 August. European Charter for regional or minority languages. 1992. Treaty/EN/Treaties/Html/148.htm (accessed 1 May 2000). Garcia, Linda. 1991. The Fractal Explorer. Santa Cruz: Dynamic Press. Hammarberg, Bjrn. 2010. The languages of the multilingual: Some conceptual and terminological issues. In C. Bardel & C. Lindqvist (eds.), Approaches to Third Language Acquisition. IRAL 48/23. [Special Issue]. 91104. Herdina, Philip & Ulrike Jessner. 2002.A dynamic model of multilingualism: perspectives of change in psycholinguistics. Clevedon: Multilingual Matters. Janson, Tore. 2002. Speak: A short history of languages. Oxford: Oxford University Press. Kachru, Braj, B., Yamuna Kachru & S. N. Sridhar. (eds.). 2008. Language in South Asia. Cambridge: Cambridge University Press. Kagan, Olga. 2005. In support of a prociency-based denition of heritage language earners: the case of Russian. International journal of bilingual education and bilingualism 8(23). 213221. Mandelbrot, Benoit. 1989. Multifractal measures, especially for the geophysicist. In Benoit B. Mandelbort & Christopher H. Scholz (eds.), Fractals in geophysics, 542. Basel: Birkhauser Verlag. Mesthrie, Rajend, Joan Swann, Andrea Deumert & William L. Leap. 2000. Introducing sociolinguistics. Philadelphia: John Benjamins. Laoire, Muiris. 2010. Change and challenge in the teaching of Irish. In Breffni ORourke & Lorna Carson (eds.), Language learner autonomy: Policy, curriculum, classroom (A festschrift in honour of David Little), 231249. Oxford: Peter Lang. Laoire, Muiris & Larissa Aronin. 2005. Thinking of multilinguality my self or my various selves?An exploration of the identity of multilinguals. Paper delivered at the Fourth International Conference onThird LanguageAcquisition and Multilingualism, Fribourg/Freiburg and Biel/Bienne, Switzerland, 8th 10th September.


Larissa Aronin, Muiris O Laoire and David Singleton

Riley, Philip. 2010. Reections on identity, modernity and the European Language Portfolio. In Breffni ORourke & Lorna Carson (eds.), Language learner autonomy: Policy, curriculum, classroom (A festschrift in honour of David Little), 373385. Oxford: Peter Lang. Valds, Guadelupe. 2000. Introduction. Spanish for native speakers, Volume 1. AATSP professional development series handbook for teachers K-16. New York: Harcourt College Publishers. Wiley, Terrence. 2001. On dening heritage learners and their speakers. In Joyce Kreeft Peyton, Donald A. Ranard & Scott McGinnis (eds.), Heritage languages in America: Preserving a national resource, 109142. McHenry: Delta Systems. Dr. Larissa Aronin is a senior lecturer at the Oranim Academic College of Education, Israel and is a research associate at Trinity College, Dublin. She has published in a range of international journals on a wide array of topics connected with multilingualism such as The international Journal of the Sociology of Language, The International Journal of Multilingualism, Language Teaching. She contributed to The Encyclopedia of Applied Linguistics Wiley-Blackwell (to be published in 2012) and co-edited The exploration of multilingualism: Development of research on L3, multilingualism and multiple language acquisition (John Benjamins, 2009). She serves as a secretary of the International Association of Multilingualism, and is an Advisory Board Member of Language Teaching (CUP). Email address: Dr. Muiris Laoire is a senior lecturer at the Institute of Technology Tralee, Ireland. He is author of textbooks, academic books and several articles on sociolinguistics, multilingualism language regeneration and language pedagogy. His most recent international publications appear in the Language Policy (Kluwer) and in Current Issues in Language and Society (Multilingual Matters), Language Awareness (123&4 2003), International Journal of the Sociology of Language, (Mouton) The Encyclopaedia of Language and Linguistics (Elsevier Science) and The International Journal of Bilingualism and Bilingual Education and The International Journal of Multilingualism. Email: Professor David Singleton took his B.A. at Trinity College Dublin and his Ph.D. at the University of Cambridge. He is a Fellow of Trinity College Dublin, where he is Head of the Centre for Language and Communication Studies. He has published across a wide range of topics, but his principal areas of interest in recent times have been cross-linguistic inuence, age-related factors in language acquisition, the lexicon and multilingualism. He is contactable at

On abduction in receptive multilingualism. Evidence from cognate guessing tasks


Abstract Most researchers agree on the idea that multilinguals have specic advantages over monolinguals in learning more languages and, more generally, develop specic competencies for expanding and managing their multilingual repertoire. Of particular interest, most notably in the context of the European strive for the upward revaluation of smaller, less used languages, is the development of receptive competences in several genetically closely related languages (cf. e.g. the EuroCom-projects). This contribution presents empirical evidence that, on the one hand, seems to support the claim that multilinguals are more efcient in developing receptive competences in new, previously unlearnt languages. The article focuses on the question how multilinguals use their languages in order to guess the meaning of cognates in unlearnt but genealogically close languages. A series of studies is discussed whose aim is to tap into this process of interlingual inferencing. Different measures for phonological and graphematic distances across languages are established and correlated with the rates of successful cognate recognition in the search for a threshold of string similarity beyond which recognition becomes unlikely. The role of different types of the participants multilingual repertoires is assessed, and other factors inuencing good performance in cognate recognition are identied. The process of interlingual inferencing is discussed as a form of abductive reasoning, and quantitative and qualitative data are analyzed to support the idea that this type of abduction is an essential driving force in receptive multilingualism and language comprehension in general.


Raphael Berthele

1. Inferencing and abductive reasoning in language learning and use There is converging evidence for the claim that bi- and multilinguals have, at least in some respects, advantages over monolinguals in learning additional languages (Cenoz 2003: 82, Cummins 2000: 35). The explanations offered for these advantages in third language acquisition (TLA) refer to advantages in cognitive development or to enhanced language awareness (Jessner 1999), sometimes also termed the M-factor (Herdina and Jessner 2002: 131). Learning and using two languages is hypothesized to lead to enhancements in particular cognitive skills and/or language awareness, which in turn leads to measurable advantages at least regarding general efciency in TLA (cf. also de Angelis 2007, Le Pichon Vorstman et al. 2009). The empirical studies presented in this chapter focus on a narrowly dened aspect of linguistic competence, the recognition of cognates in closely related but unknown languages. More concretely, this research attempts to tap into a cognitive mechanism which is likely to contribute to the advantages in TLA within genealogically related languages observed by scholarly research in the past. The goal of the deliberately reductionist approach discussed below is to shed new light on a pivotal resource in multiple language learning: interlingual correspondences. The research questions that will be addressed in this chapter are the following: 1) Are there multilingual proles (regarding the languages in the individual multilingual repertoire) that favour the rapid recognition of familiar words in unknown target languages? 2) Are there other characteristics of the participants that inuence the quality of the interlingual recognition process? 3) Which linguistic properties of target items (words) allow for spontaneous interlingual recognition, and which limit the spontaneous inferability of items? These three questions will be addressed using data elicitation methods described in Section 2. Section 1 provides a brief overview of the context of this research and the most relevant literature, and Section 3 offers some concluding remarks. The main point of this chapter is theoretical. It will be argued that one of the key inferencing procedures that underlie the linguistic tasks in the scope of our studies is of the abductive type. The particular nature of abduction, then, can be used as an explanation as to why bi-/multilinguals do better at the particular tasks we used and, as a consequence of this rst conclusion, why many bi/multilinguals might be better at language learning in general.

On abduction in receptive multilingualism. Evidence from cognate guessing tasks


1.1. M-Factor; bi-/plurilingual competence A focus on cross-linguistic similarities is becoming increasingly important as a means of promoting the usage value of less commonly used languages (LOTE, languages other than English) in light of an over-emphasis on teaching English. As one of many examples, the EuroCom framework (Hufeisen and Marx 2007, McCann et al. 2003) aims at the rapid development of comprehension skills in reading and listening in one family of languages at a time (Romance or Germanic). Picking up on an idea already present in Lados (1957) contrastive analysis hypothesis, these frameworks pursue a double focus: on the one hand, the potential of interlingually transparent words and structures is documented, on the other, particular features that are unique for the particular language are listed and presented. Although the general idea of this enterprise seems very promising, one of its main weaknesses is its orientation towards a detailed, lengthy, list-based presentation of interlingual correspondences that draws on the philological tradition. The usability of these lists remains at best unclear. As an example, sound correspondences across Germanic languages are presented in EuroComGerm (Hufeisen and Marx 2007) on no fewer than 44 pages. Given the density of these contents, and despite the fact that the authors have made great efforts to reduce the philological complexity, it seems that there should be better, more frugal ways of preparing multilinguals for the task of reading in an unknown but genealogically familiar language.

1.2. Inferencing, transfer and hypothesis testing According to Carton (1971), one of the fundamental principles in language learning is the ability to draw inferences. Studer (2008) deems inferencing a central process in almost any receptive activity of language users. Inferences can be drawn based on co-textual or contextual information, based on knowledge of the target language, of languages in general or based on knowledge that links different languages in the multilingual system together. It is these latter inferences, the interlingual ones, which are at the very core of this chapter. Furthermore, such interlingual inferences are closely related to central questions on the potential (but also the danger) of language transfer, as discussed e.g. in Odlin (1989). For the remainder of this contribution, we consider interlingual inferencing as a mechanism that operates on potential transfer bases, i.e. on acquired and learnt items and structures in any language pertaining to the multilingual repertoire. Based on inferentially emerging interlingual identication, multilinguals can be said to transfer items and structures from one language into another. Both inferencing and transfer therefore participate in


Raphael Berthele

the process of the dynamic construction of interlanguages in the form of often short-lived, spontaneous, hypothetical grammars, a process that is considered central to most modern theories of second/foreign language acquisition (see e.g. Gass and Selinker 1994: 67) as well as in multilingual language didactics (see e.g. Meissner 2001). However, the exact nature of these inferencing processes often remains surprisingly unclear. Most often, particularly in so called constructivist teaching approaches, the emphasis lies on discovery, on inductive procedures. These procedures are argued to be more efcient and more cognitively adequate for language learning than the traditional deductive or instructivist methods (cf. Wolff 1994).1 In this regard, however, authors tend to forget that there is a third type of inferencing, namely abduction.2 Abduction was introduced into semiotic theory by Peirce (1931) who was interested in the mental operation of guessing and viewed abduction as a process of forming an explanatory hypothesis (CP 5.171). In the view of Eco (1984), abduction is the daring attempt to speculate about potential rules that might explain the meaning of a signier (42).

Figure 1. Three types of inferencing according to Eco (1984: 42). Bold lines refer to given knowledge, dotted lines refer to inferenced knowledge.

Let us consider a token from the data collected in the studies discussed below and how it relates to the three types. One of the tasks involved the spontaneous guessing of the meaning of cognate verbs in Danish, a language none of the informants has learnt. As an example (1), the participants had to infer the meaning of the item blive, an item that was presented either aurally or in its written form (but not both). (1a) Aural-oral variant: [bli:ve] What does this Danish verb mean? Participant: From French plier [plie:] or from English believe, [pli5] Interviewer: Whats that in English?

On abduction in receptive multilingualism. Evidence from cognate guessing tasks


Participant: believe, but this doesnt t in Interviewer: Why? [. . . ] (1b) Written task: What does the Danish verb <blive> mean? bleiben; in Spanish v and b are often pronounced the same, [. . . ] thus BLIBE German bleiben (,to stay)

Both examples are actual responses drawn from the study that will be described in more detail below. Example 1b is a vivid example of interlingual abduction: The participant is confronted with a stimulus, she has a gut feeling that it could be related to the German bleiben, and she looks for potential phonological rules that might support and even explain this spontaneous analysis, using her knowledge of Spanish grapheme-phoneme correspondences. Based on Ecos schematic representation of induction, deduction and abduction, this particular token can be rephrased as in Figure 2.

Figure 2. Deduction, induction and abduction exemplied with Germanic sound correspondences.

As Figure 2 shows, there are other ways of linking together the rule, the case and the result. In deduction, the result is an unavoidable entailment of the rule and the case, and in induction results can allow for the inferential generalization of a rule. The gure should make it quite clear that both induction and abduction by nature involve uncertainty regarding the validity of the inference as a whole to at least some degree. This uncertainty is particularly important in abduction, since here the empirical basis is rather weak (often only 1 result). A second difference between abduction and induction needs to be explicitly mentioned


Raphael Berthele

here: the former presupposes the knowledge of rules, as the one given in example (1b), whereas the latter is about nding the rule based on a corpus of data. This difference, as will be argued in this article, is a crucial one from an acquisitional and from a teaching point of view.

2. Empirical Investigations 2.1. Lexical inferencing across Germanic and Romance languages Two preliminary studies were carried out that will be summarized very briey here (see Berthele 2008, Berthele and Lambelet 2009 for more details). Both studies were inspired by the methodology developed in Mller-Lanc (2003). The studies make use of paper and pencil tasks targeting unlearnt languages that are genealogically related to the main languages of the participants. More importantly, the target languages and the learnt languages share a great deal of vocabulary. These interlingually shared words, commonly referred to as cognates, differ to substantial degrees with respect to their similarity. It seems thus important to deconstruct the lay category of cognates and to reconstruct it based on scientic criteria (cf. Section 2.4.). All participants had to ll in a detailed language biography questionnaire in order to self-assess their prociency in their languages (including dialects). The tasks were carried out under time pressure (Task A: 15, Task B: 5, Task C: 10). The translation attempt for each item in an unknown Lx involved writing down the (guessed) meaning of the item as well as the interlingual (or contextual) transfer bases, words or rules, which lead to the guess (cf. example 1b).

2.2. Results Studies 1a and 1b Firstly, a general correlation between the number of languages spoken by the participants and the number of successful translation attempts was calculated. This analysis reveals a positive but relatively weak effect for task C only (inferencing of cognates without context; Romance targets: r = 0.392, n = 135, p < 0.001; Germanic targets: r = 0.249, n = 179, p < 0.001). Secondly, the question was addressed as to whether there are particular multilingual proles that stick out with respect to the ability to draw interlingual inferences. Different groups were formed, e.g. a group of bilinguals has been dened as participants who speak at least two languages/dialects on a level of at least 4 (maximum: 6; labeled (almost) balanced bilingual in Figure 3). Another group of intra-family bilinguals was dened as speaking at least two

On abduction in receptive multilingualism. Evidence from cognate guessing tasks


Table 1. Two structurally analogous studies on Romance and Germanic target varieties Study 1a 183 Students of German at the Universities of Zrich, Marburg, Fribourg/Freiburg German (dialect or standard) 4.25 17 Dutch words (from a reading text, authentic article, cognates and non-cognates) Dutch reading comprehension (7 questions) 29 Danish/Swedish verbs, all high frequency, all (near) cognates Study 1b 150 Students of Psychology at the University of Fribourg/Freiburg French or Italian 4.29 17 Romontsch Sursilvan words (from a reading text, translated newspaper article, cognates and non-cognates) Romontsch Sursilvan reading comprehension (7 questions) 29 Romanian and Romontsch Sursilvan verbs, all high frequency, all (near) cognates

N (# of participants) sample

participants L1 mean # of languages in the repertoires Task A: Infer meaning of words (Word list with context)

Task B: Reading comprehension Task C Infer meaning of cognates (Word list without context)

Germanic (study 1a, 2 Germanic > 4 in Figure 3) or Romance (study 1b, 2 Romance+> 4 in Figure 3) languages on a level of at least 4. Individuals not falling into these categories are labelled (normal) multilingual in Figure 3. There is considerable evidence on both the Romance and the Germanic side that the most skillful interlingual inferers are those participants who master two languages in their repertoire that are related to the target languages, i.e. two Romance or two Germanic languages respectively (cf. Table 2). On the Germanic side this shows the multilingual potential that lies in bilingualism with dialects and a standard language, since a large number of the participants who claim to have high prociency in two Germanic systems are speakers of a German dialect (Alemannic, Platt, etc.) and the German standard language (cf. the discussion in Berthele 2008). To sum up the result of the search for the good interlingual inferer we can conclude that:


Raphael Berthele

Figure 3. Multilinguals proles and respective success in tasks A, B, and C.

a) the more multilingual a participant is, the more likely she is to correctly infer word meaning (all three tasks in the case of the Germanic targets, tasks A and C in the case of the Romance targets); b) multilinguals who assess themselves as having a high prociency in at least two of their languages do better than multilinguals without this particular type of self-assessment; c) multilinguals with high prociency in two languages that are close to the target languages perform better than all other groups. In order to be able to increase the amount of variance explained, we designed a follow-up study described in the remainder of this chapter. This study draws on the method used in task C. 2.3. Study 2: Listening and reading comprehension in Germanic languages The follow-up study only focuses on Germanic verb targets without context. A total of 163 adolescents and young adults (between 13 and 35 years old) participated, all native speakers of a Swiss German dialect (or bilinguals with heritage languages or Romansh). They had to ll in a similar language prole questionnaire as in study 1. Additionally, the informants took 3 modules of Mearas (2005) language aptitude tests (called Llama B, D and E: word learning, sound recognition, sound-symbol correspondence). The stimuli were 28 Danish and

On abduction in receptive multilingualism. Evidence from cognate guessing tasks


Swedish verbs, all high frequency and (near) cognates, i.e. they are potentially detectable with counterparts in English and/or German (cf. below, Section 2.4.). The informants were presented with half of the stimuli in written form and half in aural form. Two conditions were created in order to have all verbs presented in both forms. A subsample of the informants performed the same task in a think-aloud protocol, i.e. not as a paper and pencil task but face to face with a eldworker who recorded the verbal protocols and wrote down the responses. Example (1a) above is taken from this data. The data have been analyzed in two ways: rstly, with respect to the quest for the prole of the ideal interlingual inferer, and secondly, with respect to linguistic features that interact in statistically relevant ways with the empirical difculty of the stimulus items. In order to answer the rst question, a regression analysis was carried out. The dependent variable was the standardized score of the proportion of correct inferences an individual managed to draw. As independent variables, supposedly important factors were entered stepwise: vocabulary learning ability (Llama B), sound recognition (Llama D), sound-symbol correspondence (Llama E), number of languages in the repertoire, self-assessed prociency in L1, L2, . . . , Lx, age, school/educational level. The stepwise regression reveals that there are four independent variables that contribute in a statistically meaningful way to the variance of the target variable (cf. Table 2): 1. age (the older, the better); 2. vocabulary learning ability; 3. English prociency; 4. # of languages in the repertoire. These four variables account for 62% of the variance.
Table 2. Stepwise regression coefcients for the number of successful inferencing attempts adjusted R2 = .615 Model: 1 Age Age 2 Llama B vocabulary learning Age 3 Llama B vocabulary learning English prociency Age Llama B vocabulary learning 4 English prociency # of languages in the repertoire B .196 .167 .018 .110 .019 .280 .089 .016 .310 -.236 Std. Error .044 .041 .006 .045 .005 .116 .044 .005 .111 .111 Beta .587 .501 .393 .330 .420 .318 .267 .356 .352 -.244 R2 change .344 .147 Sig. .000 .000 .003 .019 .001 .021 .050 .004 .009 .042




Raphael Berthele

The analysis shows that the ability to draw inferences increases with age, that there is at least one component of language aptitude that seems to interact with this ability, and that the multilingual repertoire, and most prominently the prociency in English, contributes signicantly to successful inferencing. If we consider these results in the light of the presupposition that we are dealing with abduction rather than induction, the results make perfect sense: multilinguals with high prociency in many languages, and above all in languages that are relatively similar to the target language, can make use of their repertoire in the sense that it provides not only important lexical transfer bases, but also potential rules and regularities, i.e. conceptual knowledge and strategic know-how that ts into the top box in Figure 1. It is this knowledge and these skills that allow the interlingual guessing game to become increasingly accurate, as we will see in much more detail in Sections 2.5 and 2.6. Before discussing these results and their potential consequences for the acquisition and learning of third/additional languages, another question should be briey addressed: What is the inuence of particular linguistic characteristics in the target words on successful interlingual comprehension? 2.4. What exactly is a cognate? Analysis of individual items The term cognate is well established in psycholinguistic investigations of the multilingual mental lexicon (e.g. in Dijkstra 2003), and the recognition of interlingual cognates is taken for granted and thus part of an implicit methodology in comparative historical linguistics (for an example see Campbell 2004: 126). Despite its being taken for granted, cognate is a rather problematic category. There are clear cases of obvious code similarity (homography or homophony) across languages, such as the Danish and the English spelling of the verb <give>. In other cases, however, minimal or even major differences in pronunciation occur, and where cognateness ends often remains unquestioned in the literature. Both in spelling and in pronunciation, even within what is commonly construed as a language, there can be substantial allophonic and allographic variation. I would suggest conceiving the category of cognate as a radial category with prototypical, i.e. almost or completely identical examples at its center and increasingly different examples on a graded scale. Moreover, the category has fuzzy boundaries since it remains unclear where interlingual cognateness ends. As shown in Figure 2 above, knowledge of the history of sound changes in Germanic and synchronic sound correspondences allow for the identication of graphematically and phonologically quite different items. The construal of cognateness as a radial category avoids the pitfall of assuming the interlingual indentiability of items in an aprioristic way. In order

On abduction in receptive multilingualism. Evidence from cognate guessing tasks


Figure 4. The radial category of synchronic cognateness

to account for the gradedness of the cognate category, the interlingual distances need to be measured. Once these measures are established, it is possible to correlate them with the empirically measured ease of inferencing in our experiments. Measuring string distance is possible by applying a variant of the algorithm proposed by Levenshtein (1966). Whereas this algorithm, in its simplest form, only counts the smallest number of insertions or deletions that have to be carried out in order to transform one string into the other, there are elaborations that also take into account the phonological distance between the respective segments (Heeringa, et al. 2006). For the purposes of this study, I have used two variants of the Levenshtein algorithm. Firstly, a featureless comparison of the graphematic strings has been carried out (cf. columns 2 and 4 in Table 3). This produces a rough measure for the graphematic similarity of the two written forms of the cognates. Secondly, the aural stimuli of the words have been transcribed phonetically, these transcripts have been coded using the X-Sampa3 standard and fed into the L044 software, which is a tool that calculates feature-based Levenshtein distances for word pairs. Very generally, it seems reasonable to hypothesize that the items empirical difculty in the interlingual inferencing task correlates negatively with the Levenshtein distances. In order to test this hypothesis, the proportion of correct answers per item per task variant has been calculated (cf. last row in Table 3).


Raphael Berthele

Table 3. Example items with Levenshtein distances kunne (Danish) Graphematic level Target Cognate German Cognate English English: distances German: distances Empirical difculty <kunne> <knnen> <can> 0.805 0.33 0.453 Phonological level [kh un@] [kh oen:@n] [kh :n] 0.24 0.14 0.161 mena (Swedish) Graphematic level <mena> <meinen> <mean> 0.40 0.50 0.557 Phonological level [miEnA] [maIn@n] [mi:n] 0.29 0.14 0.081

A correlation analysis of these values with the Levenshtein distances shows a negative interaction with the distances between the target item and the English cognate for both the written and the audio stimuli (written: r = 0.416, n = 28, p = 0.014; aural: 0.349, n = 25, p = 0.026). No such interaction can be found correlating the target and the German values (written: r = 0.14, n = 28, p = 0.471; aural: r = 0.002, n = 28, p = 0.497). 2.5. Listening comprehension: Analysis of individual Items The data from the listening comprehension condition will rst be discussed. As Figure 5 shows, there seems to be a threshold at around 0.22 for the distance measure in the listening comprehension condition.Above this threshold, no items were inferred by more than 15% of the participants. Levenshtein distances with respect to English thus seem to be a modestly useful predictor for the inferability of the target items. However, as Figure 5 illustrates, there is no simply linear relationship between the two variables. Among the items with relatively small phonological distances we still nd considerable variation. Some items are comparatively distant but still easy to infer, e.g. komma. In some cases, this can be interpreted due to the high similarity to German (kommen), in others, however, this explanation does not seem to be equally manifest (ville wollen). The lack of clear correlations and the unexplained patterns in Figure 5 raise the question whether it is not mere distance/difference, but rather particular types of differences that are the key to the empirical differences across the items in this interlingual guessing puzzle. A further explorative analysis was thus carried out that analyzes phonological and graphematic features of particular segments of the pairs of cognates.

On abduction in receptive multilingualism. Evidence from cognate guessing tasks


Figure 5. Inferability of verbs and Levenshtein distances to English

Firstly, since the data points in Figure 5 suggest that there is a meaningful difference with respect to interlingual inferencing between items with a Levensthein distance of more than 0.218, the target items have been categorized in two groups (low Levenshtein distance D < 0.218; high Levenshtein distance D > 0.218). Secondly, a more elaborate item-map has been constructed for each item in the list. The simplied schematic structure of the target items is C-V-C-V pattern (cf. kunne, mena). In some cases, the consonant positions are lled with consonant clusters, as in skrive. For each of these four positions, phonetic features have been coded based on the phonetic transcriptions of the items. The item-map thus contains variables which code the main articulatory differences between the cognates for each pair of words: Consonants: place/manner of articulation; phonation Vowels: aperture; backness; roundedness; diphthong/monophthong; Vowels and Consonants: segment differences (e.g. affricates vs. fricatives; insertions or deletions of sounds); quantity


Raphael Berthele

If there is more than one segment per idealized C or V position, as in the case of consonant clusters, the differences between these clusters have been coded as a whole. Based on this item-map with an impressive number of variables (54 for the aural data) it is possible to run a cross-tabulation analysis which indicates whether each particular feature coincides with higher or lower success in interlingual inferencing. An exploratory analysis has been carried out using classication trees with the dependent variable correct inference (a nominal variable) and the whole phonological item-map as independent variables. The goal was to nd out which of these numerous variables discriminate best with respect to the dependent variable.6 In other words, the question here is which linguistic feature clearly brings about less success in inferencing when two cognates differ with respect to it. This analysis was carried out for two groups of participants separately: for the good inferers and for the poor inferers. These groups were determined by dividing the sample according to the criterion of whether a participant scored below or above the mean of correct inferences (36%) for the whole sample. This analysis produces the picture shown in Figure 6 and Figure 7. The exploration of these categorization trees shows which phonological features discriminate within our data with respect to successful interlingual inferencing. In order to lay out the logic of these trees, two selected paths that lead through the tree in Figure 7 will be analyzed as an example. Firstly and most importantly, if an item has a Levenshtein distance that is higher than the threshold discussed above, it is very unlikely for the good inferers to nd the cognate. Only 61 (18.2%) of the items are successfully translated, 275 (81.8%) remain unrecognized. However, if the Levenshtein distance is below the threshold, the good inferers succeed in 363 attempts (56.5%). The second criterion that plays a statistically signicant role is the comparison of phonation of the second consonant in English and in the target item. If there is a difference in phonation, or if one of the forms lacks a second consonant, it is less likely that the inferences will be correct (43.1% of correct inferences). Within these cases, the rst vowel criterion is important: if the rst vowel shows a difference in roundedness when comparing English and the target item, the inferences are more likely to be successful than if not. Node 8 thus stands for verbs such as ha a a [h6:], g ["go:@], and st ["st@P], that are relatively well translated (69.5%).7 Let us now turn to the items that do not present any difference in phonation regarding the second consonant, i.e. to all items that belong to daughters of node 3. The next criterion that plays an important role is the comparison of the aperture of the rst consonant in the German and the target cognates. If there is no difference at all or merely a minor one (node 5), 77.4% (178 tokens) are

On abduction in receptive multilingualism. Evidence from cognate guessing tasks


Figure 6. Listening comprehension (bad inferers)

correctly inferred. If there is a difference larger than 1 step, the majority of the items is not inferred correctly (53.3%). Analyzing the daughters of the former category (node 5 in Figure 7), the last criterion is the manner of articulation comparing the rst English consonant and the target words. If the words present the same manner of articulation, there is an 88.5% likelihood that they will be inferred correctly (items such as komma, lgge, ligge). Even if this feature differs (node 10), the majority of the tokens are correctly inferred (tale, tnka; 54.1% correct inferences).


Raphael Berthele

Figure 7. Listening comprehension (good inferers)

All nodes in the two gures above could be analyzed in this manner. For the sake of brevity we will not go into more details here but the following table shows a synopsis of the hierarchy of the features that the categorization tree analysis produces as being statistically signicant for the good and bad interlingual inferers. Table 4 provides a hierarchically structured account of the features that interact in a statistically signicant way with the probability that participants carry out correct inferences. The table can be used for a tentative exploration of structural constraints governing the interlingual inferencing task.

Table 4. Important structural characteristics in listening comprehension poor inferers Feature expected or unexpected relationship expected phonation C2 aperture roundedness manner of articulation # of segments C1 V1 expected V2 expected L EN GE GE GE backness manner of articulation backness unexpected C2 GE expected V1 GE EN manner of articulation monophthong/ diphthong place of articulation expected V1 EN quantity C2 EN quantity expected unexpected expected unexpected expected expected unexpected expected unexpected Consonant Vowel Levenshtein EN GE Feature expected or unexpected relationship

good inferers

rank of constraint

consonant vowel Levenshtein




















On abduction in receptive multilingualism. Evidence from cognate guessing tasks




Raphael Berthele

1. The most important factor for the good inferers is the Levenshtein criterion, i.e. whether an item lies below or above the threshold visible in Figure 7 above. Good inferers are thus relatively likely to infer items that are below this threshold but highly unlikely to infer items that lie beyond this threshold. This criterion only shows up in fourth position for the group of poor inferers. This means that these participants do a relatively poor job even if the Levenshtein distance is below this threshold. 2. Of the 6 features listed in Table 4 that relate to vowels, 5 have an unexpected relationship regarding the phonological difference between the cognates and the number of correct inferences; in other words, difference here coincides with more correct inferences. 3. All consonant differences between the cognates listed in Table 4 stand in the expected relationship with the number of correct inferences. 4. Of the 15 features listed in Table 4, 7 relate to consonants, 2 to the Levenshtein criterion, and 6 to vowels. One could tentatively hypothesize that consonants seem to play a slightly more important role, but only when taking into account the unexpected relationship between vowels and correct inferences discussed in (2) and the invariably expected effect direction between consonants and correct inferences (3). 5. Of the 15 features listed in Table 4, 9 relate to English and 6 to German. Moreover, the rst two levels in the respective hierarchies are always occupied by feature comparisons to English. 6. Good inferers infer correctly if consonants are the same (3 C variables vs. 2 V variables, one of which behaves inversely). Poor inferers, at rst glance, show a less clear picture 4 variables are related to vowels, 4 to consonants. Delving deeper, however, it turns out that all four vowel variables show unexpected patterns, i.e. the vowel differences coincide with higher levels of correct inferences compared to word pairs with the same vowels. This suggests an even stronger role of consonants: if they are the same, bad inferers have a relatively high chance of inferring the words correctly, but differences in vowels do not seem to trigger worse performances. 2.6. Reading comprehension: Analysis of individual Items Before commenting further on these exploratory results, lets briey turn to the analogous analysis of the written task. For the reading comprehension data, the coding was simply based on the question of whether the particular segment is the same, similar (v-w, b-p, etc.) or different (t-s). Since we have little access to the participantsimagined phonology when reading the words, the vowels were strictly coded in a nominal and rather

On abduction in receptive multilingualism. Evidence from cognate guessing tasks


Figure 8. Reading comprehension (bad inferers)

approximate way for sameness or difference, erasing all diacritics and length markers in all languages (i.e. German <ie> is treated as <i>, <> as <a>, Swedish < > as <a>, etc.). a Table 5 can again be used for a tentative exploration of structural constraints governing the interlingual inferencing task in the reading condition. 1. Levenshtein distance above or below the mean does not predict correct inferencing in a measurable way. 2. Of the 16 statistically signicant variables that enter the model and that are listed in Table 5, 9 relate to vowels and 7 to consonants. 3. Half of the variables relate to German, half to English. 4. Of the 9 vowel variables, 4 (or 5, if 2a is included) have an unexpected relationship with the changes in correct inferences, i.e. difference coincides frequently with more correct inferences.


Raphael Berthele

Figure 9. Reading comprehension (good inferers)

5. Of the 7 consonant variables, 2 have an unexpected relationship with the changes in correct inferences, i.e. difference coincides in two cases with more correct inferences. 6. The good inferers seem to be particularly efcient if the rst part of the item is identical to English (V1 and C1 on levels 2 and 3 of the tree). On the other hand, if the English onset (C and V) are not absolutely identical, as in kunne, or if there is no English cognate at all, the German onset plays a crucial role: If it is identical, then the word can be recognized (blive), if not, the word is hardly recognized (skulle).

On abduction in receptive multilingualism. Evidence from cognate guessing tasks Table 5. Important structural characteristics in reading comprehensio good inferers rank of constraint 1 2a consonant vowel Levenshtein V2 V2 EN GE expected or unexpected relationship unexpected (vowel never identical) expected expected bad inferers Consonant Vowel Levenshtein V1 V2 EN GE


expected or unexpected relationship expected expected



2b 3a 3b 3c 4a 4b 4c 4d 5

V1 C1


C2 V2 V2 V1


unexpected unexpected unexpected expected unexpected expected expected unexpected expected




V2 C2 C2 C2 C1

2.7. Discussion and support from qualitative data The results of these tree analyses are not absolutely clear, and some of the patterns uncovered are rather puzzling, in particular those which stand for higher comprehension proportions coinciding with phonological/graphematic differences. Nevertheless there are some preliminary conclusions that can be drawn from the analyses of the listening and the reading comprehension task. In the listening comprehension condition, consonantal contrasts (or their absence) seem to be a more important predictor for successful inferencing than vowels. If vowels are concerned, comprehension can even be better in cases where they are different from the English or German counterpart. The same unexpected pattern in the vowel category can be found in the reading comprehension data. However, if consonants often play a measurable role, it is (with the exception of C2) in the expected direction, i.e. difference coincides with less successful inferencing. Table 4 and Table 5 show a slightly higher importance of the rst segments (onset), since differences in the second CV segment are less frequently listed and often in an unexpected relationship. This conrms Mller and Zeevaerts


Raphael Berthele

(2010: 9) observation that similarity or sameness of the onset is more important than in the rhyme in cognate recognition. In the experimental data discussed in this contribution, we are not actually dealing with cases of automatic bi- or multilingual word recognition as investigated in other psycholinguistic paradigms (cf. Dijkstra (2003), Grosjean (2008)). Although there may well be an automatic component in cognate recognition, in this study we are rather interested in higher-level cognition applied by multilinguals to resolve the interlingual puzzles. Nevertheless, it seems useful to take into account the notions put forward in theoretical and empirical studies of the bi- or multilingual lexicon (cf. Cenoz, et al. 2003, Pavlenko 2009): following the majority view in the eld, we can assume that the multilingual lexicon is an integrated lexicon that consists of a mix of words (Dijkstra 2003: 17), although some of the models differ with respect to the architectural details regarding how words or features are associated with languages or organized in language specic ways, and the conguration of activation and inhibition across or within layers vary across the different models. Most of these models, however, distinguish among different structural layers (feature, letter, word, language in BIA (Bilingual Interactive Activation), feature, phoneme, word in BIMOLA (Bilingual Interactive Model of Lexical Access)) which are incrementally activated depending on the visual/aural input, and it is generally assumed that there does not seem to be such a thing as a language switch. In our case, even if there were one, the participants are put in a situation where they are explicitly asked for a multilingual search across their lexicon. Based on the explorations of the data presented above, it seems that not all of the structural elements (letters, phonological features, phonemes, classes of phonemes such as consonants) participate in the same way in this search: in the auditory condition the results discussed above suggest that informants rst rely on similarities of consonantal features. If these similarities are sufciently strong or salient, differences in vowels are disregarded and chances are good that the correct cognate candidates will be selected. If consonantal patterns differ, particularly with respect to manner and place of articulation (cf. Table 4), cognate recognition becomes unlikely. Although the picture is not totally unambiguous in the case of the reading comprehension data, the tendency seems to be similar: the good inferers in particular rely on consonants. The overall difference between the cognate strings as measured by the Levenshtein algorithms only plays a signicant role in the listening condition, and here only with respect to English. We can thus hypothesize that the pattern comparison that interlingual inferers apply is selective in nature and does not simply compare the forms in a holistic way. The following passage from a think-aloud protocol in the reading comprehension condition illustrates this main focus on selected features:

On abduction in receptive multilingualism. Evidence from cognate guessing tasks


(3) (think-aloud task, P: Swiss German native participant, F: Field worker; computer screen displays the form <skulle>, the task is carried out in Swiss German) P: skulle oh dasch aber es herzigs Wort skulle this is a cute word ii tnt e chli nach Tootechopf oder so sounds like skull or something [long silence, 17s] skulle dasch sicher es Prolwort, wome gar nd cha ablite this surely is a prole word that cannot be derived [12s] kchi Aanig, chani au sge i heb kchi Aanig? no idea, can I also say that I dont have a clue? skulle schifaare ni skulle to ski no [3s] skulle F: was teichsch, was cht das isch . . . what do you think what this . . . P: rollen, aber nur wils zwi -ll- ht, aber weiss au nd to roll, but only because there are two -ll- but I dont know aber da macht aso das isch absoluut . . . but this is totally . . . skulle skill, ah villicht fhig sein oder so oder wissen skill, ah, maybe capable of or so, or to know skulle ja sgemer wissen yes, lets say,to know vo irgendwie skill, Fhigkeit oder so from skill somehow, ability or so The example illustrates how the participant varies the vowels (/u/, /i/, /o/) while explicitly mentioning associations between forms that share consonantal elements (e.g. roll because of the shared <ll> graphemes). The item is also an example that illustrates the difculty of linking cognates that differ with respect to (onset) consonants. However, it is by no means impossible to overcome consonantal impediments (see example 1b): In retrospect, it seems that the results discussed based on Table 4 (consonant identity being more important for bad inferers than for


Raphael Berthele

good inferers) are in accordance with the abductive strategies that can be observed with particularly gifted multilingual subjects: the search for cognates does not stop after a rst maybe too simple and too direct string analogy has been identied: (bliveto believe). Good inferers thus consider not only modications of vowels, but they also search for potential rules in the consonantal domain, as the Spanish grapheme-phoneme rule discussed in example 1b. As Mller (in preparation) has shown, there are consistent patterns of interlingual plausibility that multilinguals display when they are asked to provide potential cognate forms within the Germanic languages.

3. Conclusions Based on the analyses in Section 2, we can identify at least two partial cognitive processes that contribute to good guessing capacities based on the intra- and interlingual competences: 1) A exible and selective comparison of features and patterns, focusing on consonants and neglecting or systematically varying the vowels. 2) Good hunches regarding when to continue and when to stop searching. The rst process can be seen as a form of perceptual tolerance in pattern recognition, i.e. the patterns are aligned in a exible way allowing for more or less systematic variation. Coming back to the results reported in Section 2., we can now try to understand the differences between groups with different multilingual proles: as we have seen, the most efcient interlingual inferers are those multilinguals who master two languages that are closely related to the unknown target language. This nding as such is not overwhelmingly surprising, since one could argue that these multilinguals simply have more potential transfer bases that nurture the inferencing task. In the light of the results discussed in this Section, however, I hypothesize that there is more to this than the simple number of potentially transferrable forms. As schematically represented in Figure 10, inspired by the revised hierarchical model (Kroll and Stewart 1994), I would like to suggest that multilinguals with such highly developed dia-systems be they due to prociency in two Romance languages or to prociency in a Germanic dialect and a standard language develop a meta-system that is a form of abstraction over the bilingual mental corpus. A cognate in an unknown language will rst be associated with one or more than one word forms in the multilingual lexicon, and at least if our hypothesis is correct in the case of a bilingual Romance lexicon, as in Figure 10, the form corresponds to an abstracted meta-form that combines the features that are similar or the same

On abduction in receptive multilingualism. Evidence from cognate guessing tasks




interlingual Inference: potential cognate recognition

c?r(r)?r Linter


Figure 10. Schematic abstraction over French and Italian cognates of the verb to run in multilinguals. <?> stands for a phoneme/grapheme placeholder. Cuorer is the Romontsch Sursilvan cognate.

in the two Romance languages and has placeholders where the two languages differ. The relative invariability of consonants and the unreliable nature of vowels is a feature well known from dialectology: Schmellers (1872) dictionary of Bavarian dialects has introduced a consonant skeleton organizing principle, since the synchronic vowel instability (caused by diachronic changes) makes an alphabetic order of a multi-dialect dictionary impossible. The nding reported in Section 2.2., i.e. that it is mainly speakers of dialect (and Standard German) who are good at interlingual inferencing, is therefore not very surprising. This abstract schematic entry in the multilingual lexicon is the basis for the enhanced potential lexicon as measured by the interlingual inferencing task: the multilinguals develop quick and efcient interlingual heuristics within a language family regarding potential interlingual differences and similarities. These schematic forms and the heuristics that at the same time allow for their emergence and cognitive entrenchment enable multilinguals to be particularly efcient in recognizing cognates in unknown languages. The second process listed above, at rst sight, is more mysterious. How do our subjects decide when to stop and when to continue searching? Good stopping rules, according to Gigerenzer (2007), are important in simple heuristics in general just as much as in interlingual inferencing: as shown in the transcript above, sometimes searches are merely a waste of time. In other cases, however, the search is aborted too quickly since a supposed cognate has been identied that satises the interlingual comparison (e.g. bliveto believe). The data dis-


Raphael Berthele

cussed here do not provide evidence that permits tapping into this particular part of the interlingual inferencing process with sufcient ecological validity. In naturalistic situations, there would be additional, contextual information available that contributes in signicant ways to this decision, most notably the semantic t of the item within the co- and context. The articial and decontextualized nature of the word list data investigated here deliberately neglects these important constraints, in order to focus on the purely formal, linguistic side of the process. The latter certainly plays a role in naturalistic exolingual contexts, where multilinguals are confronted with items or texts in poorly mastered or unknown languages or varieties, and we can hypothesize that the better the interlingual heuristics on the linguistic level, the better the chances to successfully understand utterances and texts in related unlearnt languages. The challenge of the purely linguistically based process of abduction as the one described in Section 2. is therefore a dialectic one: if the subject is confronted with a cognate form, it is good to speculate based on the linguistic knowledge that is at hand. However, speculation can go too far or be a waste of time, so it is good to know what a likely candidate is, or, if none can be found that satises the expectations, it is important to know when to abort the search (and, in naturalistic contexts again, when it is appropriate and important to consult a dictionary or ask a procient speaker). As the quite fuzzy description of these two processes makes clear, we are dealing with speculative and probabilistic processes that are not amenable to direct scientic observation. The main result of the empirical studies presented in this contribution is that the quality of these interlingual guessing procedures depends on many factors, both concerning the multilingual subjects and the linguistic contrasts involved. Quality increases with increasingly procient multilingual systems, and I have argued that the main process that enables multilinguals to be efcient in the tasks is a form of multilingual abduction, an inferencing process that exploits multlilinguals knowledge of what are likely and of what are unlikely correspondences across and within languages. And nally, if the tendencies discussed in Section 2.5. and 2.6. turn out to be validated by other studies, the sound correspondence section of the EuroCom materials could and should be simplied and shortened by at least 50% to the most important consonant correspondences.

1. Radically inductive approaches to foreign language teaching have been criticized for quite some time now, and the pendulum is swinging back to approaches that

On abduction in receptive multilingualism. Evidence from cognate guessing tasks



3. 4. 5.



include focus on form or to mixed post-method era practices. However, especially in the German-speaking world, the idea of radically constructivist foreign language teaching is still surprisingly alive and kicking. Abduction has been mentioned before in linguistics: rstly as a mechanism in language change (Mc Mahon 1994), and secondly by Chomsky (1968) as a principle in rst language acquisition. cf. [July 28, 2010] cf. [July 28, 2010] The feature-based and featureless distances have both been normalized using the longest length among the set of least cost alignment solutions, as recommended in Beijering et al. (2008). Additionally, this method acts as a control for the total number of alpha errors when running multiple statistical tests. For nominal dependent variables, the QUEST algorithm is the recommended procedure. These examples illustrate an important aw in the method applied, a difculty that at this point has not been satisfactorily resolved: all of the participants are not only procient in standard German, but also in one or several Alemannic Swiss German dialects as their respective L1s, many of which have cognate forms that are quite close to the target items (e.g. Zurich German /h6:/ have, /St6:/ stand). However, since the different Alemannic dialects spoken by the different informants are quite different in phonology, it turned out to be impossible to construct an analogous item-map for the dialectal cognates. The study discussed here thus, in future variants, should control systematically for the native dialects and include them in the analyses.

Beijering, Karin, Charlotte Gooskens & Wilbert Heeringa. 2008. Predicting intelligibility and perceived linguistic distance by means of the Levenshtein algorithm. In Marjo van Koppen & Bert Botma (eds.), Linguistics in the Netherlands. 1324. Amsterdam: John Benjamins. Berthele, Raphael. 2008. Dialekt-Standard Situationen als embryonale Mehrsprachigkeit. Erkenntnisse zum interlingualen Potenzial des Provinzlerdaseins. In Klaus J. Mattheier & Alexandra Lenz (eds.), Dialektsoziologie/Dialect Sociology/Sociologie du Dialecte. Sociolinguistica, Volume 22. 87107. Tbingen: Niemeyer. Berthele, Raphael & Amelia Lambelet. 2009. Approche empirique de lintercomprhension: rpertoires, processus et rsultats. LIDIL. 15162. Campbell, Lyle. 2004. Historical linguistics: an introduction, 2nd edn. Edinburgh: Edinburgh University Press. Carton, Aaron S. 1971. Inferencing: a process in using and learning language. In Paul Pimsleur & Terence Quinn (eds.), The psychology of second language learning. Papers from the second international congress of applied linguistics, Cambridge, 812 September 1969. 4558. Cambridge: University Press.


Raphael Berthele

Cenoz, Jasone. 2003. The additive effect of bilingualism on third language acquisition: A review. International Journal of Bilingualism 7/1. 7187. Cenoz, Jasone, Ulrike Jessner & Britta Hufeisen. 2003. The multilingual lexicon. Dordrecht: Kluwer Academic Publishers Chomsky, Noam. 1968. Language and mind. New York ; Chicago [etc.]: Harcourt Brace & World. Cummins, Jim. 2000. Language, power and pedagogy : bilingual children in the crossre. de Angelis, Gessica. 2007. Third or additional language acquisition. Clevedon/Buffalo/ Toronto: Multilingual Matters. Dijkstra, Ton. 2003. Lexical processing in bilinguals and multilinguals. In Jasone Cenoz, Ulrike Jessner & Britta Hufeisen (eds.), The multilingual lexicon, 1126. Dordrecht: Kluwer Academic Publishers. Eco, Umberto. 1984. Semiotica e losoa del linguaggio. Torino: Einaudi. Gass, Susan & Larry Selinker. 1994. Language transfer in language learning. Amsterdam: John Benjamins. Gigerenzer, Gerd. 2007. Gut feelings: the intelligence of the unconscious. New York: Viking. Grosjean, Fran ois. 2008. Studying bilinguals. Oxford: Oxford University Press. c Heeringa, Wilbert, Peter Kleiweg, Charlotte Gooskens & John Nerbonne. 2006. Evaluation of String Distance Algorithms for Dialectology. Proceedings of the Workshop on Linguistic Distances. Sidney, Australia Association for Computational Linguistics, 5162. Herdina, Philip & Ulrike Jessner. 2002.A dynamic model of multilingualism. Perspectives of change in psycholinguistics. Clevedon et al.: Multilingual Matters. Hufeisen, Britta & Nicole Marx. 2007. EuroComGerm Die sieben Siebe. Germanische Sprachen lesen lernen. Aachen: Shaker Verlag. Jessner, Ulrike. 1999. Metalinguistic awareness in multilinguals. Cognitive aspects of third language learning. Language Awareness Vol. 8: 3&4. 2019. Kroll, J. F. & E. Stewart. 1994. Category interference in translation and picture naming: Evidence for asymmetric connections between bilingual memory representations. Journal of Memory and Language 33. 14974. Lado, Robert. 1957. Linguistics across cultures: Applied linguistics for language teachers. Ann Arbor: University of Michigan. Le Pichon Vorstman, Emmanuelle, Henriette De Swart, Viktorija Ceginskas & Huub Van Den Bergh. 2009. Language learning experience in school context and metacognitive awareness of multilingual children. International Journal of Multilingualism 6. 258 80. Levenshtein, Vladimir I. 1966. Binary codes capable of correcting deletions, insertions, and reversals. Soviet Physics Doklady, 10(8). 70710. Mc Mahon, April M. S. 1994. Understanding language change. Cambridge: McCann, William J., Horst G. Klein & Tilbert D. Stegmann. 2003. EuroComRom How to read all the Romance languages right away. 2nd revised edition. Aachen: Shaker.

On abduction in receptive multilingualism. Evidence from cognate guessing tasks


Meiner, Franz-Joseph. 2001. Vom induktiven zum konstruktiven Lehr- und Lernparadigma. Methodische Folgerungen aus der Mehrsprachigkeitsdidaktischen Forschung. In Franz-Joseph Meissner & Manfred Reinfried (eds.), Bausteine fr einen neukommunikativen Franzsischunterricht. 2150. Tbingen: Narr. Mller, Robert. in preparation. hnlichkeits-Intuitionen bei der Erkennung von germanischen Kognaten. Mller, Robert & Ludger Zeevaert. 2010. Da denke ich spontan an tafel Zur Worterkennung in verwandten germanischen Sprachen. Zeitschrift fr Fremdsprachenforschung, 21, 2: pp. 217248. Mller-Lanc, Johannes. 2003. Der Wortschatz romanischer Sprachen im Tertirsprachenerwerb. Lernerstrategien am Beispiel des Spanischen, Italienischen und Katalanischen. Tbingen: Stauffenburg. Odlin, Terence. 1989. LanguageTransfer: cross-linguistic inuence in language learning. Cambridge: CUP. Pavlenko, Aneta. 2009. The bilingual mental lexicon: interdisciplinary approaches (Bilingual education & bilingualism [70]). Buffalo (N.Y.): Multilingual Matters. Peirce, Charles Sanders. 1931. The collected papers Vol. V: Pragmatism and pragmaticism. Cambridge, MA: Harvard University Press. Schmeller, Johann Andreas. 1872. Bayerisches Wrterbuch von J. Andreas Schmeller, 2. edn. Mnchen,: R. Oldenbourg. Studer, Thomas. 2008. Inferenzen als Prinzip des Sprachverstehens (218 Bl.). Zrich: Universitt Zrich. Wolff, Dieter. 1994. Der Konstruktivismus: Ein neues Paradigma in der Fremdsprachendidaktik? Die neuen sprachen, 93/5. 40729. Raphael Berthele is professor of multilingualism at Fribourg University. He has worked and studied in Fribourg, Tbingen, Berkeley (CA) and Berne. His research interests cover cognitive, psycholinguistic, and variationist approaches to language usage. Recently, he has been working on the empirical investigation of multilinguals comprehension skills in foreign languages. Email:

Language input environments and language development in bilingual acquisition


Abstract A striking feature of young bilingual childrens language use is that there can be so much difference between the ways in which they use either of their languages and the levels of language skill in each. As I will argue, many of the differences between individual bilingual childrens use of their two languages can be attributed to differences in the language input environments for each of the languages. These language input environments concern, amongst others, language use patterns in the parent pair, age of rst regular exposure to each language, relative and absolute frequencies of input for each language, and interaction strategies. Evidence for my claims will be based on a survey study of 3,390 bilingual children, a more in-depth study of 31 bilingual families, and ndings from the literature. If indeed it is the case that differences in language input environments can explain much of the variation between one particular childs use of two languages (and, by extension, inter-individual variation between bilingual children), this has important repercussions for the assessment of bilingual children in school and elsewhere.

1. Introduction It is a common observation that children under age 6 who started to learn a second language at age 4 or 5 make mistakes in that language, dont know many words in that language, and speak that language much worse than children who are monolingual in that language. Usually, people outside the small circle of specialists who study young childrens language learning attribute this difference between these young new bilinguals and monolinguals to the very fact that the bilinguals are dealing with two languages. Typically, the learning opportunities


Annick De Houwer

that young new bilinguals had for learning their second language are not taken into account. This fact became very clear to me when I happened to speak to a speech therapist at a university hospital clinic in the mid 90s who explained that they were treating a four-year-old child from Ghana for a language learning disorder since the child did not speak any Dutch (the majority language). The speech therapist was well aware that the boy had only arrived in a Dutch-speaking environment two months previously. This speech therapists blatant blindness to the notion that childrens amount of experience with a second language may be an important explanatory factor in early second language acquisition (ESLA) reminded me of a similar sort of blindness in an American father of a girl (Lauren) who had been raised with English and Dutch from birth. However, Lauren heard English only about three hours a week (her father worked hard, and only spent a few hours time with Lauren on week-ends). When Lauren was three years old, the only words she said in English were yes and no. This enraged her father, who would not even consider the possibility that the amount of time he spent speaking English to her had anything to do with her own use of English (De Houwer 2009). Instead, he accused Lauren of rejecting him.1 Yet another example of blatant blindness to the idea that the language input environment may affect young childrens language learning is the case of Sven. Svens mother talked to me at a conference in Brussels. She was very worried about Sven. He was 4.5 years old and very unhappy. He hardly spoke. Yet when he was younger he had spoken Arabic uently. This was when the family lived in Cairo (where Sven was born). His diplomat parents spoke Swedish to him from birth, but often were out so there was an Arabic-speaking nanny for Sven from the very start. At age two and a half, Sven was a happy little boy who spoke Arabic uently, and who understood Swedish but hardly spoke it. Then the family moved to Brussels, where Sven was put in a French-English school. The nanny stayed in Cairo. As Svens mother explained to me, she and Svens father thought that adding two more languages would be great for Sven, so he would speak Arabic, Swedish, French and English. But in Brussels, Sven became very unhappy, and did not speak English or French, and hardly any Swedish. At the beginning his parents thought he just needed some adjustment, and they had heard of the silent period when children rst come in contact with a second language. But when Sven continued to be pretty much silent Svens mother started to worry. Svens parents thought that Sven perhaps missed his nanny, so they invited the nanny to come to Brussels when Sven was 4.5 years old. They expected him to speak Arabic to the nanny and were very surprised that he could no longer even understand her, let alone speak Arabic. The nanny went back to Cairo. Svens mother spoke to me shortly after this disastrous visit

Language input environments and language development in bilingual acquisition


and didnt know what to do. She was highly surprised to hear me suggest that Sven needed much more time to hear Swedish she had interpreted Svens very low speaking skills in Swedish as the result of him being tired and unhappy at school, and neither she nor her husband had ever considered that the very small amount of time they spent with Sven speaking Swedish might have anything to do with his low Swedish speaking abilities. I also advised changing Svens school. Luckily there was a school in Brussels that offered Swedish (in addition to French). Through a mutual acquaintance I heard about a year later that Sven had changed considerably and was now speaking Swedish and French much better, and that he appeared much happier. So far, few systematic studies have been carried out to investigate the specic role played by the overall amount of language input in two languages to young children. However, there are a few studies that have looked at other global aspects of childrens language input environments in relation to bilingual development. Also, a comparison of different studies can yield important insights. It is the aim of this article to offer a brief review of some recent studies that have either implicitly or explicitly addressed the relation between language input environments and language development in bilingual acquisition. Language input environment is the overall term I use to refer to a number of different aspects of the language input that children hear. These include, but are not limited to, the number of utterances children hear in a language, the length of time children have heard a language, the way languages are distributed among parents in a bilingual family, and the way parents respond to childrens choice of language. The rst three all relate to input frequency (e.g., the length of time that a child will have heard a language is in fact a frequency measure), which is most likely the most important environmental factor in bilingual acquisition. Bilingual acquisition as understood here encompasses both Early Second Language Acquisition and Bilingual First Language Acquisition (both explained below). I focus only on language production, not on comprehension. Also, I limit my discussion to children who have not received any reading or writing instruction at school. This effectively limits the discussion to children up until around the age of 6.

2. Early Second Language Acquisition Monolingual children may become bilingual once they start to regularly hear a second language. Such children, if they are under age 6, are in the process of Early Second Language Acquisition (ESLA; De Houwer 1990, 1995, 2009). Recently, some studies have looked at the amount of time since the rst day


Annick De Houwer

of exposure to a second language that it takes for pre-school age children to actually say more in the new language than just a few words or set formulas. Suzanne Schlyter reported that many Swedish-speaking preschoolers in a French-speaking school needed more than a years input in French before they started to say enough sentences in French for Schlyter and her colleagues to have enough material to analyse the childrens sentence structures (Schlyter & Granfeldt 2008). Likewise, 15 out of 35 L1 German children between the ages of 2;11 and 6;8 with daily input at a French-speaking school for 3 months to just over a year said either nothing in French, or just single word utterances (Meisel 2008). For those children who did produce French sentences Meisel (2008) found qualitative differences on the morphosyntactic level between the children in relation to the length of time they had heard French (data were analyzed for 10 children; see also Meisel 2009). Meisel (2009) suggests there is a large degree of interindividual variation between ESLA children in how fast they learn to speak a second language. This is most clearly shown in a study of three Turkish-speaking children who started to acquire German in an all-German-speaking preschool (Rothweiler 2006). After 15 months of hearing German four hours a day from the teacher and other children (excluding vacations and week-ends, presumably), one child on average said 2.9 words per sentence (the boy Furkan), another child 2.4 (the girl Ece), and yet another child only 2 (the girl Melisa). For childrens early language acquisition, the mean length of their utterances is an important indication of the structural complexity of their speech (Brown 1973). We can thus see that a similar amount of exposure time led to quite different levels of structural complexity amongst these three children (my interpretation of the Rothweiler data).2 If we compare the ndings for the Turkish-German ESLA children reported on by Rothweiler (2006) to ndings from children who have had exposure to German from birth, however, we see that at the same age, children with exposure to German from birth usually produce longer sentences than age-matched ESLA children with 15 months of exposure only. When Furkan produced an average of 2.9 words per sentence, he was 4 years old. At age 4, the children Marta, Aurelio, Jan, Lukas and Carlotta all had an average mean length of utterance in German between 3.7 and 5.5 (data from Cantone 2007; my comparison). When Ece said 2.5 German words per utterance on average, she was 4 years and 3 months old. At that age, the children Jan, Lukas and Carlotta had an average mean length of utterance between 3.8 and 4.8 (data from Cantone 2007; my comparison). Thus, at the same age, preschool children who had input in German from birth said much longer utterances than children who had heard German for only about the last third of their lifetime. This little comparison is only suggestive, of course,

Language input environments and language development in bilingual acquisition


and much more data from many more children need to be collected in order to make proper comparisons between ESLA childrens learning rate of their second language and the rate at which this language develops in children who heard it from birth. Also, other and more sophisticated measures than mean length of utterance need to be considered. Another point on which ESLA children may differ from children who have heard these ESLA childrens second language from birth is the types of language structures they use. ESLA children under age 6 are regularly reported to use morphosyntactic structures in their L2 that do not appear in the speech of children who have heard that L2 from birth (Ekmek i 1994; Ervin-Tripp 1974; c Fantini 1985; Li Wei 2011; Meisel 2008; Pfaff 1994; Tabors 1987; Wong Fillmore 1979; Zdorenko and Paradis 2007). While some of these structures show inuence from ESLA childrens L1, other structures do not (e.g., Granfeldt et al. 2007). These qualitative differences between L2 structural use by children who have had much less time to learn this L2 than age-matched peers acquiring this language from birth are added evidence for the fact that the length of time that children under 6 have had to learn a language is of importance in bilingual development. Findings regarding the expressive vocabulary development of 19 young children acquiring English as a second language further suggest that the interaction between the length of time that children have had to hear their L2 and their chronological age is an important variable to take into account (Golberg et al. 2008). After 9 months exposure to English, the L2 learners studied by Golberg et al. used fewer different word types in production than age-matched learners who had heard English from birth. However, after 3 years exposure to English, the same L2 learners used more different word types in production than younger native speakers who had had the same amount of exposure in English (Golberg et al. 2008: 58). This suggests that older L2 learners can learn vocabulary faster. The different ndings for different domains of language use show that the relation between language input and bilingual development can be different depending on the aspect of language use considered, but especially with children aged 4.5 or younger, the length of time that children have heard their new L2 is paramount, regardless of which domain is considered. This is of particular importance for applied settings such as preschools and the speech clinic. This point will be taken up again in the conclusion.


Annick De Houwer

3. Bilingual First Language Acquisition The children Marta, Aurelio, Jan, Lukas and Carlotta mentioned in section 2 all heard German from birth, indeed, but they had also all heard Italian from birth (Cantone 2007). As such, they were all going through a process of Bilingual First Language Acquisition (BFLA, De Houwer 1990, 2009; Meisel 1989). In BFLA children, the period of time they heard each of their two languages is the same for both languages, i.e., it coincides with their chronological age. If indeed the length of exposure to a language is so important, as suggested in the previous section, then one should expect BFLA children to develop their two languages at the same rate. There might well, however, be differences between children, just as there is individual variation among ESLA children (see earlier), but one would not expect much variation within one particular child. However, the facts are quite different. There may be quite a large difference between a BFLA childs two languages.At one extreme, a BFLA child may speak only one language. Or a BFLA child may speak two languages, but may speak them at highly different levels of prociency. Then there are BFLA children who, as one might in fact expect, speak both languages equally well (De Houwer 2009). The evidence to date suggests that much of this variation between a single BFLA childs use of two languages nds its roots in childrens language input environments. 3.1. Speaking only one language, in spite of hearing two languages at home I rst consider those BFLA children who speak only one language, in spite of the fact that they hear two languages at home from their parents. A survey study I carried out in Flanders, Belgium, was able to show that the precise pattern of parental bilingual language use plays an important role in deciding whether a BFLA child will speak two languages or just one. Amongst others, the survey looked at the language use of 3,390 BFLA children in dual-parent families between the ages of 6 and 9 and their siblings. These children were growing up in a home in Flanders where the parents spoke Dutch (the majority language used also at school) and another (minority) language X, which could be any of the 73 languages reported in the survey. Only 75% of the children spoke both Dutch and language X. A full quarter of the 3,390 children spoke just Dutch. I found that the language distribution within the parent pair plays a big role in explaining why so many children spoke only Dutch (De Houwer 2007, 2009). Those bilingual families where children ended up speaking just the majority language, Dutch, tended to be families where both parents used the majority

Language input environments and language development in bilingual acquisition Table 1. Language distribution in the parent pair and childrens bilingual use Parent 1 speaks NL + X NL + X NL NL + X Parent 2 speaks X NL + X X NL


the children speak two languages 97% 79% 73% 34%

NL = the majority language, Dutch; X = any other language (the minority language)

language at home, and only one parent used the minority language (Table 1). Parents who both spoke the majority as well as the minority language at home and parents who both spoke the minority language at home, with one parent in addition using the majority language, had much better chances of having children who spoke both the majority and the minority language. Most likely there are several contributing elements that can explain the ndings from the survey but the most important one may be related to how often children hear each language, that is to frequency of input: if only one parent speaks the minority language, but in addition that same parent also speaks the majority language, there may be less input in the minority language than in families where both parents speak the minority language (in addition to one or both also speaking the majority language). Secondly, families where both parents speak the minority language at home may create much more of an environment conducive to using that minority language. After all, the family already has two members speaking that language rather than just one. Parents may also be more inclined to speak the minority language amongst themselves. These are certainly issues to be further explored. The fact that a much smaller but also much more detailed survey in another country with another majority language (Japanese) found very similar results (Yamamoto 2001) strongly suggests that the specic majority-minority language constellation involved is not a major factor that explains why many children raised bilingually speak only one language. Another reason why BFLA children may speak only one language while hearing two at home is that there is no particular pressure for them to actually speak two languages. Parents may answer a childs question in Spanish with a response in German, if that is the language they usually use with their toddler. These parents are using what Lanza (1997) has termed bilingual discourse strategies, i.e., conversational patterns that allow the use of two languages within a conversation. Other bilingual parents may be much less tolerant of the use of two languages within a conversation, and socialize the child into using mainly one language within a stretch of discourse. These bilingual parents use what is termed monolingual discourse strategies (Lanza 1997).


Annick De Houwer

Although so far there have not been enough studies documenting these discourse strategies and their possible effect on childrens language use, it is very likely that if parents use mainly bilingual discourse strategies, and allow the childs use of the other language Alpha, children have no need for speaking language A, and will become part of that large group of children raised bilingually who understand two languages but speak only one. Finally, it should be emphasized that the language that wins in a BFLA setting where children speak only one language tends to be the majority language (see also, e.g., Yamamoto 2001). I have so far not heard any reports of BFLA children who only spoke the minority language, in spite of hearing a majority and a minority language at home, and the same majority language at school. The survey study I discussed in this section also shows quite clearly that the so-called one person, one language strategy (1P/1L) does not necessarily lead to success, that is, to children who actually speak two languages: in 27% of the families where one parent spoke Dutch, and the other parent another language, the children did not speak that other language (Table 1). Earlier, Patterson (1999) had already investigated the role of parental input strategy in Spanish-English bilingual families in the U.S.A. She looked at the relation between the use of mixed utterances by a group of 102 bilingual toddlers and parental input strategies. She found no difference between the use of mixed utterances by children from homes where parents used the 1P/1L approach and by children from homes where both parents adressed children in both languages (1P/2L). Patterson also examined the role of language presentation in function of childrens rst use of word combinations, an important milestone in child language development. Again, she found no differences between children from homes where both parents adressed them in both languages and children from 1P/1L homes in whether children had reached this important milestone or not. It would appear, then, that the supposed superiority of the 1P/1L parental input strategy is not supported by the actual empirical evidence. 3.2. BFLA children who speak one language markedly better or more than the other one Recently, studies of BFLA have started to pay more attention to BFLA children who speak one language markedly better than the other one. Leopold (1970, but rst published during WWII) already reported on such a child, German-English Hildegard, but it wasnt until the late 90s that studies started to appear that paid more attention to this uneven development of a BFLA childs two languages (De Houwer 2002).

Language input environments and language development in bilingual acquisition


Pearson et al. (1997) were the rst to explicitly investigate why some BFLA children seem to develop at the same pace in both languages while others do not. They tried to explain differences between the number of different words that Spanish-English BFLA children produced in either language by referring to whether there was a difference in the overall proportion of input in each language or not. This proportion was obtained by asking parents what proportion of the time their child heard each language. Using this input measure of relative frequency that Pearson et al. themselves called crude, the authors were able to show that the more children heard Spanish rather than English, the more Spanish words they knew, and the other way round. Children who heard Spanish about equally often as English did not show major differences between the number of Spanish and English words in their production repertoire. In a similar vein, Chan and Nicoladis (2010) report a larger proportion of Mandarin words in the rst 50 word production repertoires of two BFLA boys who heard markedly more Mandarin than their other language, English. In a study that looked only at one of bilingual childrens two languages, viz., Basque, Barre a et al. (2008) found similar evidence. They found that n children with less than 60% input in Basque produced fewer words in Basque than children with more than 60% input in Basque. Looking at another area than lexis, Parra et al. (2011) found evidence that relative frequency of input to young Spanish-English bilingual children could explain some of the variance in the childrens phonological memory skills in each language. Although these studies are highly suggestive, relative frequency is ultimately a by-product of absolute frequency of input, i.e., of how often children hear utterances and words in each of their languages. It is not unlikely that the relative frequency effects found so far are in fact by-products of underlying differences in absolute frequency. The absolute frequency with which children hear a language was found to be of utmost importance for young childrens lexical development in a monolingual setting. Hart and Risleys (1995) longitudinal study of 42 English-speaking monolingual families found that children in families where there was a lot of adult talk knew many more words than children in families that talked much less overall. The inuence of the sheer number of words that children heard per time unit was much more important, for instance, than the manner in which adults addressed children (note that, as recently summarized by Hoff (2006), the frequency with which people talk to children will have various repercussions for many other important aspects of child directed speech). For bilingual acquisition there to my knowledge so far has been only one study that has published conversational input data on several bilingual families


Annick De Houwer

and their childrens language development. The data are based on longitudinal recordings of natural conversations in the home for each of ve child-caregiver pairs and comprise a total of 24,755 transcribed utterances (Allen et al. 2002; this study is summarized in Allen 2007). The families speak English and Inuktitut. Allen et al.s (2002) study was not focused on nding links between the input data and the childrens speech. However, their study contained sufcient detail for me to be able to attempt such an analysis (De Houwer 2009: 124). This analysis revealed that the more the caregivers in each family talked, the more their children talked. This conrms Hart and Risleys suggestion that verbose caregivers will generally have verbose children. More interesting from a bilingual perspective, however, is that the rank orders for the two languages separately show great similarities between the adult caregivers and the children in each family. In Table 2, I show a new analysis based on Allen (2007). I computed the absolute numbers of utterances in English and Inuktitut based on the percentages in these tables (henceforth absolute frequencies). I did this for both the adult caregivers and the children in each of the 5 families.
Table 2. Absolute frequencies of adult caregiver and child utterances in English and Inuktitut in 5 families (based on Tables 2 and 3 in Allen 2007) family AI SR PN SA AW child 1231 (1) 1015 (2) 721 (3) 707 (4) 324 (5) English adults 1584 (2) 1717 (1) 1091 (3) 721 (4) 628 (5) Inuktitut child 1102 (2) 135 (5) 192 (4) 800 (3) 1427 (1) adults 2087 (3) 558 (5) 889 (4) 2499 (2) 3998 (1)

Note: the numbers between brackets refer to the rank orders

For English, Table 2 shows a near perfect match between the rank orders for the children and the adults. There just is a reversal for two rank orders that differ in only one point: children AI and SR have rank orders (1) and (2), whereas their respective caregivers have rank orders (2) and (1). Similarly, for Inuktitut there also is a near perfect match, with again a reversal for two rank orders that differ in only one point: children AI and SA have rank orders (2) and (3), whereas their respective caregivers have rank orders (3) and (2). Thus, caregivers who spoke more English compared to others had children who spoke more English, and caregivers who spoke more Inuktitut compared to others had children who spoke more Inuktitut.

Language input environments and language development in bilingual acquisition


Admittedly, the analysis here is quite crude, but it does suggest that it is worthwhile to start to look for absolute frequency effects for the two languages that bilingual children are acquiring. Looking at only 5 families is not enough, however. The chance is too high that what one nds is a pure coincidence. One needs to look at larger groups of children in order to make reliable comparisons. In an ongoing study that I am carrying out in cooperation with Dr. Marc Bornstein, Child and Family Research, National Institute of Child Health and Human Development, USA, we developed a language diary that allowed us to obtain reliable estimates of the number of waking hours per day that a group of 31 BFLA Dutch-French children between the ages of 5 and 20 months heard in Dutch and French. We also obtained data for times when both languages were being used in the same conversational setting. As we reported earlier (De Houwer and Bornstein 2003), we found great differences between families in the amounts of time that children heard just one of the languages, or both of them in a single conversational setting. We also found that in this period of 15 months there were great uctuations in the amount of time individual children heard their two languages. This uidity appears to be quite normal in a bilingual setting: based on an overview of various reports on individual children in several bilingual families in different countries and speaking different languages and different language combinations, I showed that a temporary increase or decrease in the contact that bilingual children have with each of their languages is quite common (De Houwer 2009). Whats more, such changes appear to have quite an immediate effect on childrens language use. Children may suddenly start to speak one of their languages much more, and at a higher level of complexity, or, conversely, they may start to speak a language much less, and less well. Place (2009) undertook a similar study to that of De Houwer and Bornstein (2003) for 31 English-Spanish bilingual families in Florida, USA. Using the language diary developed for the Dutch-French study, Place asked caregivers to record language exposure information for one week when the children in the bilingual families were 25 months of age. Place found that the presence of English was overwhelming in these families, in spite of parents commitment to bilingualism.3 For the Dutch-French language diary study, we also obtained measures of the number of different words that the 31 BFLA children produced in Dutch and French (for further information on this group of children and on the lexical measures used, see De Houwer et al. 2006 and De Houwer 2010). Preliminary (hitherto unpublished) results for 15 of the 31 bilingual toddlers at age 20 months show strong trends suggesting that the more often children hear a particular


Annick De Houwer

language spoken, the more words they will produce in that language (De Houwer 2006b). Table 3 presents the relations between the number of contact hours for either Dutch or French and the number of words produced in either language. I rst express this relation in terms of the number of children who, compared to the others, were (a) above or (b) below the median as far as the number of hours they heard French and Dutch was concerned. Then within each group (a and b), I plotted the number of children who were above or below the median for word production.
Table 3. The relation between absolute input frequency in a language and the number of words that children produce at age 20 months
French contact hours words produced number of children Dutch contact hours words produced number of children (a) above the median 6 2 (b) below the median 2 5 above the median below the median above the median below the median (a) above the median 7 1 (b) below the median 1 6 above the median below the median above the median below the median

The shaded cells in Table 3 show how most of the children with the higher number of hours of input in French also produced the most French words (7 out of 8 children), and how most of the children with the lower number of hours of input in French also produced the least French words (6 out of 7 children). The same picture applies to Dutch. These preliminary results are encouraging, and we plan to continue working with these data to see whether the trend we see so far is conrmed for the entire group of children.

4. Absolute and relative frequency of input in bilingual settings Ever since Pearson et al. (1997) showed that the relative proportion of input in two languages to bilingual children may have an effect on the number of words they produce, scholars reporting on bilingual children have started to describe childrens language environments in terms of the estimated proportion of input in each language. This has denitely been an improvement over the situation earlier, when often bilingual learning environments were not described at all (see De Houwer 1990 for a critique). Giving a rough estimate of proportions of use of

Language input environments and language development in bilingual acquisition


each language is perhaps easier than giving an indication of the average number of hours per day that a child hears each language. Nevertheless, describing bilingual childrens language learning environments in terms of absolute amounts of input is most likely more informative. Compare, for instance, the language input information in the English-Inuktitut study by Allen et al. (2002) and Allen (2007). In Table 4 I have plotted the amount of input in the 5 families studied in these publications in terms of both absolute and relative frequency.
Table 4. The same input information expressed in both absolute and relative terms (based on Tables 2 and 3 in Allen 2007) family AW SA AI PN SR caregivers English absolute relative 628 (5) 13.3 721 (4) 20.9 1584 (2) 39.7 1091 (3) 50.9 1717 (1) 72.7 caregivers Inuktitut absolute relative 3998 (1) 84.6 2499 (2) 72.4 2087 (3) 52.3 889 (4) 41.5 558 (5) 23.6

Note: the percentiles do not add up to 100% because mixed utterances were not taken into account; the numbers in brackets refer to rank orders in a vertical comparison

One of the assumptions when scholars refer to relative proportions of language input in two languages is that a 5050 distribution is ideal. In Table 4 we see one family that comes close to this ideal: family PN. For this family, one would expect the child to develop the two languages in a balanced way, and one would not expect large differences between the childs two languages. As I show in Table 5, there is no such balance in evidence for the child in family PN (with a 3/4 to 1/5 distribution for both languages; mixed utterances are not included). Furthermore, we see a balance in child SA that comes as a surprise compared to the very unbalanced input distribution of 2172% for the caregivers in family SA. Table 6 shows the numerical point differences between the relative use of either language by caregivers and children. A large point difference reects a much stronger use of one language over the other; a small point difference indicates much more of a balance. Compared to each other, again a rank order can be established amongst the 5 families. Table 6 shows that this rank order is very different for the caregivers and the children. Thus, the relative frequency of use of either language by the caregivers in the 5 English-Inuktitut families does not predict the relative frequency of use of each language by the children


Annick De Houwer

Table 5. Childrens language use expressed in absolute and relative terms (based on Tables 2 and 3 in Allen 2007) family AW SA AI PN SR childrens English absolute relative 324 (5) 18.3 707 (4) 45.0 1231 (1) 47.6 721 (3) 76.2 1015 (2) 84.2 childrens Inuktitut absolute relative 1427 (1) 80.6 800 (3) 50.9 1102 (2) 42.6 192 (4) 20.3 135 (5) 11.2

Note: the percentiles do not add up to 100% because mixed utterances were not taken into account; the numbers in brackets refer to rank orders in a vertical comparison Table 6. Point differences between the relative frequency of use of English and Inuktitut (based on data in Tables 2 and 3 in Allen 2007) family AW SA AI PN SR caregivers 71.3 (1) 51.5 (2) 12.6 (4) 9.4 (5) 49.1 (3) children 62.3 (2) 5.9 (4) 5.0 (5) 55.9 (3) 73.0 (1)

Note: the numbers in brackets refer to rank orders in a vertical comparison

in these families. This stands in stark contrast to the rank orders for the absolute frequency of input as discussed earlier. These were virtually identical. Of course, the above comparison again is quite crude, but is does suggest that using absolute measures of input frequency might have more explanatory power than relative measures of input frequency. It should also be noted that the comparisons here are based on very different kinds of data than the ones examined by Pearson et al. (1997). The latter referred to reported data only and to the number of different words children knew, whereas the English-Inuktitut data are based on transcriptions of recorded conversations and refer to the number of utterances said in a particular language.

Language input environments and language development in bilingual acquisition


5. Conclusion When I spoke to Laurens father in the early 80s, to the speech therapist in the 90s, and to Svens mother not even 10 years ago, I could only respond to them based on my conviction and common sense that in a bilingual setting childrens opportunities to learn each language matter a great deal. I am not sure that the totality of the studies so far that have addressed links between bilingual childrens language learning environments and the bilingual development process offer enough empirical support for this as yet. However, the results of the various studies so far are quite suggestive. The ndings from both ESLA and BFLA with regard to the opportunities that children have had to learn their two languages, either as a function of the amount of time they have had to hear a particular language or as a function of the language choice patterns they have encountered, strongly suggest that frequency of input is indeed a highly important factor. Simply put, the more young children have heard a particular language, the more chance there is they will learn to speak a particular language, and to speak it well. This last point is important for bilingual families to realize: if they are keen on raising children who speak two languages uently and have a large vocabulary in each, children must be given the opportunity to learn. This means that parents should speak to their children a whole lot. Frequent book reading is also an excellent tool to boost bilingual development. In clinical practice, whenever there is concern about bilingual childrens level of language development, care should be taken to assess childrens past language learning opportunities. Insufcient input in one of the two input languages may be part of the cause for what may look like delayed development in one of them (De Houwer 2006a). In ESLA situations, children must be given the opportunity to hear their second language very frequently. For the organization of nursery schools with second language learners it thus becomes important to create a multitude of different settings in which individual children hear the new language. This includes creating many situations where young second language learners interact with children who speak the second language already. This is not easy, especially if there are children in the classroom who share the same L1. Yet it is these peer interactions through which children can learn a great deal. Hearing the new language just through a few remarks from the teacher and a short story every day is not enough. Obviously, there are different factors besides amount of input that play a role. It is my rm belief, however, that if bilingual children have had a lot of opportunities to hear and engage in each of their two languages, they will soon become uent speakers of both.


Annick De Houwer

1. When I asked the father why he thought that the amount of time a child hears a language should not matter, he replied that this is what Noam Chomsky said (the father had an M.A. in Translation Studies and had taken Linguistics courses); I spoke to Laurens father in the early 80s. 2. The individual differences between these three children are most likely a result of differences in their personalities, ability to perceive new sounds, their L1 prociencies and much more; however, this issue goes beyond the aim of the present paper. 3. In a follow-up study (Place & Hoff to appear), various aspects of the childrens language exposure are related to aspects of their language development. However, at the time of preparation of the present article the text was not yet available.

Allen, Shanley. 2007. The future of Inuktitut in the face of majority languages: Bilingualism or language shift? Applied Psycholinguistics 28. 515536. Allen, Shanley, Fred Genesee, Sarah Fish & Martha Crago. 2002. Patterns of code mixing in English-Inuktitut bilinguals. In M. Andronis, C. Ball, H. Elston & S. Neuvel (eds.), Proceedings of the 37th Annual Meeting of the Chicago Linguistic Society (Vol. 2), 171188). Chicago, IL: Chicago Linguistic Society. Barre a, Andoni, Maria-Jos Ezeizabarrena & I aki Garca. 2008. Inuence of the linn n guistic environment on the development of the lexicon and grammar of Basque bilingual children. In Carmen Prez-Vidal, Mara Juan-Garau & Aurora Bel (eds.), A portrait of the young in the new multilingual Spain. Issues in the acquisition of two or more languages in multilingual environments, 86110. Clevedon: Multilingual Matters. Brown, Roger. 1973. A rst language. The early stages. Cambridge: Harvard University Press. Cantone, Katja. 2007. Code-switching in bilingual children. Dordrecht: Springer. Chan, Wai Han & Elena Nicoladis. 2010. Predicting two Mandarin-English bilingual childrens rst 50 words: Effects of frequency and relative exposure in the input. International Journal of Bilingualism 14(2). 237270. De Houwer, Annick. 1990. The acquisition of two languages from birth: a case study. Cambridge: Cambridge University Press. De Houwer, Annick. 1995. Bilingual language acquisition. In Paul Fletcher & Brian MacWhinney (eds.), Handbook of child language, 219250. London: Blackwell. De Houwer, Annick. 2002. Uneven development in bilingual acquisition. Keynote lecture presented at the Joint meeting of the IASCL and the SRCLD, Madison, U.S.A., 1621 July.

Language input environments and language development in bilingual acquisition


De Houwer, Annick. 2006a. Die Bedeutung des Inputs fr die Evaluation der Sprachkompetenzen bilingualer Kinder. Keynote lecture presented at the Jahrestag, Deutscher Bundesverband fr Logopdie, Berlin, Germany, 1314 September. De Houwer, Annick. 2006b. Input factors affecting bilingual development. Invited plenary lecture, Language Acquisition and Bilingualism Conference, Toronto, Canada, 47 May. De Houwer, Annick. 2007. Parental language input patterns and childrens bilingual use. Applied Psycholinguistics 28(3). 411424. De Houwer, Annick. 2009. Bilingual First Language Acquisition. Clevedon: Multilingual Matters. De Houwer, Annick. 2010. Assessing lexical development in Bilingual First Language Acquisition: What can we learn from monolingual norms? In Madalena Cruz-Ferreira (ed.), Multilingual norms, 279322. Frankfurt: Peter Lang. De Houwer, Annick & Marc H. Bornstein. 2003. Balancing on the tightrope: language use patterns in bilingual families with young children. Fourth International Symposium on Bilingualism, Tempe, Arizona, USA, April 30May 3. De Houwer,Annick, Marc H. Bornstein & Sandrine De Coster. 2006. Early understanding of two words for the same thing: A CDI study of lexical comprehension in infant bilinguals. International Journal of Bilingualism 10(3). 331347. Ekmek i, zden. 1994. Bilingual development of English preschool children in Turkey. c In Guus Extra & Ludo Verhoeven (eds.), The cross-linguistic study of bilingualism, 99112. Amsterdam: North-Holland. Ervin-Tripp, Susan. 1974. Is second language learning like the rst? TESOL Quarterly. 111127. Fantini, Alvino. 1985. Language acquisition of a bilingual child: a sociolinguistic perspective (to age ten). Clevedon: Multilingual Matters. Golberg, Heather, Johanne Paradis & Martha Crago. 2008. Lexical acquisition over time in minority rst language children learning English as a second language. Applied Psycholinguistics 29. 4165. Granfeldt, Jonas, Suzanne Schlyter & Maria Kihlstedt. 2007. French as cL2, 2L1 and L1 in pre-school children. In Jonas Granfeldt (ed.), Studies in Romance bilingual acquisition Age of onset and development of French and Spanish, 742. Lund, Sweden: Lunds Universitets Spr k- och Litteraturcentrum. a Hart, Betty & Todd Risley. 1995. Meaningful differences in the everyday experiences of young American children. Baltimore: Paul Brookes. Hoff, Erika. 2006. How social contexts support and shape language development. Developmental Review 26. 5588. Lanza, Elizabeth. 1997. Language mixing in infant bilingualism. A sociolinguistic perspective. Oxford: Clarendon Press. Leopold, Werner. 1970. Speech development of a bilingual child. A linguists record. New York: AMS Press. (Original work published 19391949) Li Wei. 2011. The early acquisition of English as a second language: The case of young Chinese learners of English in Britain. In Annick De Houwer & Antje Wilton (eds.),


Annick De Houwer

English in Europe today: Educational sociocultural perspectives, 105122. Amsterdam/Philadelphia: John Benjamins. Meisel, Jrgen. 1989. Early differentiation of languages in bilingual children. In Kenneth Hyltenstam & Loraine Obler (eds.), Bilingualism across the lifespan. Aspects of acquisition, maturity and loss, 1340. Cambridge: Cambridge University Press. Meisel, Jrgen. 2008. Child second language acquisition or successive rst language acquisition? In Belma Haznedar & E. Gavruseva (eds.). Current trends in child second language acquisition: A generative perspective, 5580. Amsterdam/ Philadelphia: John Benjamins. Meisel, Jrgen. 2009. Second language acquisition in early childhood. Zeitschrift fr Sprachwissenschaft 28. 534. Parra, Marisol, Erika Hoff & Cynthia Core. 2011. Relations among language exposure, phonological memory, and language development in Spanish-English bilingually developing two year olds. Journal of Experimental Child Psychology 108. 113125. Patterson, Janet L. 1999. What bilingual toddlers hear and say: Language input and word combinations. Communication Disorders Quaterly 21(1). 3238. Pearson, Barbara Zuerer, Sylvia Fern ndez, Vanessa Lewedeg & Kim D. Oller. 1997. The a relation of input factors to lexical learning by bilingual infants. Applied Psycholinguistics 18. 4158. Pfaff, Carol. 1994. Early bilingual development of Turkish children in Berlin. In Guus Extra & Ludo Verhoeven (eds.), The cross-linguistic study of bilingualism, 7597. Amsterdam: North-Holland. Place, Sylvia. 2009. Inuences on toddlers dual language exposure in bilingual homes. Poster presented at the 2009 Biennial Meeting of the Society for Research in Child Development, Denver, Colorado, USA, 24 April. Place, Sylvia & Erika Hoff. to appear. Properties of dual language exposure that inuence two-year-olds bilingual prociency. Child Development. Rothweiler, Monika. 2006. The acquisition of V2 and subordinate clauses in early successive acquisition of German. In Conxita Lle (ed.), Interfaces in multilingualism, o 91113. Amsterdam/ Philadelphia: John Benjamins Schlyter, Suzanne & Jonas Granfeldt. 2008. Is child L2 French like 2L1 or like adult L2? Paper presented at IASCL 2008, The XI Congress of the International Association for the Study of Child Language, Edinburgh, United Kingdom, July 28-August 1. Tabors, Patton. 1987. The development of communicative competence by second language learners in a nursery school classroom: An ethnolinguistic study. Unpublished doctoral dissertation. Boston: Harvard University. Wong Fillmore, Lily. 1979. Individual differences in second language acquisition. In Charles Fillmore, D. Kempler & W. Wang (eds.), Individual differences in language ability and language behavior, 203228. New York: Academic Press. Yamamoto, Masayo. 2001. Language use in interlingual families: A Japanese-English sociolinguistic study. Clevedon: Multilingual Matters. Zdorenko, Tatiana & Johanne Paradis. 2007. The role of the rst language in child second language acquisition of articles. In A. Belikova, L. Meroni & M. Umeda (eds.),

Language input environments and language development in bilingual acquisition


Galana 2: Proceedings of the Conference on Generative Approaches to Language Acquisition North America 2, 483490. Somerville, MA: Cascadilla Proceedings Project. Annick De Houwer (PhD, Free U. of Brussels) is Professor of Language Acquisition and Teaching at the University of Erfurt in Germany. She is also Director of the Language Center there and of the European Bilingual Studies Group, ERBIS. Her research specialty is early bilingual acquisition but she has also published on Dutch child language, teen language, and intralingual subtitling. Her most recent publications include Bilingual First Language Acquisition (2009, Multilingual Matters). Email:

The role of phonemic awareness in early L2 reading for adult English language learners: Pedagogical implications

Abstract This paper reports on part of a larger study investigating the role of phonemic awareness (PA) and decoding ability in second language (L2) reading development among adult English language learners who have not mastered reading in their native language. The data reported in this paper were part of a larger longitudinal study that investigated a variety of factors related to initial reading development for adult L2 learners of English. A group of 26 participants enrolled in adult ESL education classes in the United States were followed for one year and assessed on measures of phonemic/phonological awareness and decoding ability. The results of this study document the relationship between phonemic/phonological awareness and L2 beginning reading in English for adult learners and provide evidence for similar relationships existing between phonemic/phonological awareness and early reading for adult learners and children learning to read in English. While reading and literacy development are complicated, the ndings of this study suggest the importance of the decoding process for adult learners in the beginning stages of ESL reading. Key Words: Adult Education, English as a Second Language, Literacy Development, Phonemic Awareness, Phonological Awareness, Reading Research, Second Language Reading.

1. Introduction Growing numbers of adult learners whose native language is not English are in need of reading instruction at the adult education level. The U.S. Department of Education, Ofce of Vocational and Adult Education (2005) reported that


Margo DelliCarpini

43.8% of the participants in federally funded adult education during the 2003 2004 school year were English language learners (ELLs). Of this number, 50% participated in either beginning literacy or beginning level ESL classes during that program year (U.S. Department of Education, 2006).1 According to the most recent statistics of the National Assessment of Adult Literacy2 (NAAL, United States Department of Education, 2005) in the year 2003 there were eleven million adults not literate3 in English4 in the United States. These individuals comprised two subgroups: seven million who were unable to answer even basic questions on the assessment, and four million who were unable to participate due to a language barrier, which meant that they neither spoke nor understood Spanish or English, the two languages used in the interviews. Of the former group, four million were Hispanic, and combined with the four million in the latter group, the results from the survey indicate that between four and eight million adults in the non-literate in English category are non-native speakers of English.5 These current data represent no change in the trend reported in the 1992 National Adult Literacy Survey (NALS) which illustrated that second language (L2) adults in general fared poorly compared to other groups in terms of their ability to successfully interpret tasks in the prose literacy, document literacy and quantitative literacy categories.6 Based on the aforementioned demographic data increased knowledge of the L2 reading process in general, and specically, the early reading process for adult ELLs who are beginning readers, will inform instruction and strengthen existing programs for adult ELLs. While a number of studies have investigated L2 literacy (for a comprehensive review see Burt, Peyton & Adams 2005) the focus has been primarily on top down strategies such as schema, context clues and the transfer of comprehension strategies from the rst language (L1) (Eskey 1988; Hinkel 2006). In an annotated bibliography on adult L2 reading research (Adams & Burt 2002) between the years1980 and 2002, there were only ve studies identied investigating beginning reading in adult ESL learners (Griffen 1990; Klassen & Burnaby 1993; Robson 1982; Strucker 1997). In order to effectively meet the educational needs of pre-readers and beginning readers additional research of the beginning adult L2 reading process is necessary. What is known is that beginning reading encompasses a variety of factors: a persons background experiences, his or her language, culture, and the need to participate in certain types of literacy events. In addition, initial reading in an alphabetic system like English includes mastery of word level skills. The initial process of reading in monolingual children, and more recently, bilingual children, has been studied extensively (see Armbruster, Lehr & Osborn, 2003; Barone & Morrow,2003; Droop &Verhoeven 2003; Slavin & Cheung 2005). One

Phonemic awareness and beginning adult L2 reading


important factor of beginning reading in alphabetic systems that has emerged from this research has been the role that phonemic awareness plays in the early reading process. It is this component of reading, considered a word level skill, or a bottom up skill, that this study investigates for adult ESL learners.

2. Phonological and phonemic awareness: A review of the literature Phonological awareness (PA) is an awareness of the sounds of speech and is considered a metalinguistic ability. Phonemic awareness is the ability to perceive, identify and to manipulate the sounds in spoken language (National Reading Panel 2000). Phonological and phonemic awareness deal with units of speech, and there is frequently a distinction made between phonemic awareness (the smallest units, phonemes) and phonological awareness (dealing with larger units such as onsets and rhymes) (Cunningham, Hoffman,& Yopp 1998). For the purposes of this research phonemic and phonological awareness will be conated and referred to as PA. Numerous studies have found PA to be a necessary element in the beginning reading process in an alphabetic system, and PA has been found to be a predictor of future reading ability in children at the Pre-K and elementary level (Abouzeid 1992; Bentin 1992; Ehri, Wilce & Taylor 1988; Lundberg 1989; Maclean, Bryant & Bradley 1988; Vellutino & Scanlon 1988). Research has provided evidence for the importance of the ability to phonologically decode words as a skill necessary to developing word recognition (Vellutino & Scanlon 1991). Bentin (1992) claims that the ability to decipher phonology from writing is a prerequisite for reading and understanding written words at the rst encounter, and needs to be mastered before efcient reading can occur (p. 204). In addition, decits in phonological processing have emerged as a leading cause of reading disabilities (Torgesen 1991). In fact, perhaps the most important single conclusion about reading disabilities is that they are most commonly caused by weaknesses in the ability to process the phonological features of language(Torgensen, Wagner, Rashotte, Rose, Lindamood, Conway & Garvan 1999: 579). Decits in phonological processing ability can negatively impact a readers ability to develop uent decoding ability, therefore negatively affecting efcient word recognition (Sabatini, Venezky, Kharik & Jain 2000). Task complexity has also been the focus of research in order to better dene PA as a construct, to determine what the measures tell us regarding learner ability, and as a diagnostic tool. Researchers have worked to dene PA in terms of . . . complexity of the units on which the operations are performed. . . cognitive demands of the task. . . and complexity of the syllable structure of items that are presented in each task . . . (Jim nez & Venegas 2004: 798). e


Margo DelliCarpini

Yopp (1988) investigated the types of PA skills that were most predictive of decoding ability and that should be included in both research and practice. She found that each of the tests of phonemic ability that were assessed7 had a high level of predictive validity, and discussed breaking PA into Simple Phonemic Awareness and Compound Phonemic Awareness. She identied tasks like blending (where sounds are blended together to form words or nonsense words: [k--t] [cat]) to be an easier, or a Simple PA task, which requires one operation, and tasks like deletion (say the word black . . . delete the [b] and say the word that remains [lack]) to be a more difcult or Compound PA task, which requires the performance of one operation, then holding that operation in shortterm memory while performing another operation. Children in Yopps study had an easier time blending than deleting phonemes and identication tasks such as rhyme and phoneme isolation were the easiest tasks for children to perform. English is not the only language that has been investigated in terms of the importance of PA to the early reading process. Studies conducted with children whose native languages include Spanish, Korean, Arabic and Latvian have further contributed to the large body of knowledge on the important role of PA and early reading (Abu-Rabia, Share & Mansour 2003; Carrillo 1994; Kim & Davis 2004; Sprugevica & Hoien 2003). Researchers have hypothesized that PA is (a) a prerequisite for learning to read, (b) inuenced by reading instruction and practice, and (c) both a cause and a consequence of reading acquisition (i.e., reciprocal) (Smith, Simmons, Kameenui 1995). When researchers rst began to investigate the relationship between PA and reading ability, they focused on the phoneme level since it was argued that letters in alphabetic systems (usually) represent individual phonemes. In order to learn to read, children need to be aware of the individual phonemes and phonetic segments in spoken words before they would be able to learn about the correspondence between phonemes (sounds) and graphemes (letters) (Castles & Coltheart 2004; Gough & Hillinger 1980). In addition, researchers have argued that an awareness of units such as rhymes will facilitate the mapping of phonemes to letter sequences (Castles & Coltheart 2004; Goswami 1993; Goswami & Bryant 1990). While the research into PA takes on a variety of formats and comes from different disciplines, the studies that inform the current research deal with the correlation between PA and beginning reading, PA in bilingual children learning to read in English, and the nature and role of PA in adult monolingual learners who are beginning readers.

Phonemic awareness and beginning adult L2 reading


2.1. Correlations between PA and decoding ability Over the past three decades research documenting the causal relationship between PA and reading has grown and strengthened. Recent reviews of the research have concluded that the data provide strong enough evidence to establish this causal relationship between PA and early reading for monolingual learners (Adams 1990; Rack, Snowling, Olson 1992; Spector 1995). A large number of correlational studies have consistently shown signicant and predictive relationships between PA and early reading success among monolingual children (Mann 1984; Vellutino & Scanlon 1987; Wagner, Torgesen, Laughon, Simmons & Rashotte 1993; Wagner, Torgensen, Rashotte, Hecht, Barker, Burgess, Donahue & Garon 1997), as well as correlations between PA and childrens subsequent reading and spelling abilities (Adams 1990; Bradley & Bryant 1985; Mann 1993; Share & Stanovich 1995). Moreover, evidence for the causal relationship between PA and reading exists in the empirical studies that have investigated the positive effect of explicit instruction in PA on childrens spelling and reading achievement (Ball & Blachman 1991; Castiglioni-Spalten & Ehri 2003; Ehri, Nunes, Willows, Schuster,Yaghoub-Zadeh, Shanahan 2001; McCutchen, Abbot, Green, Beretvas, Cox, Poller, Quiroga & Gray 2002). In essence, An awareness of phonemes is necessary to grasp the alphabetic principle that underlies our system of written language (Chard & Dickson 1999: 263). 2.2. PA in bilingual children learning English Research conducted with bilingual children learning to read in English has provided evidence for correlations between PA in the L2 (English) and literacy in the L2 (Stuart-Smith & Martin 1997).In addition, evidence has emerged for the cross-linguistic transfer of PA skills and has highlighted the positive ways that L1 PA abilities can be applied to the L2 reading (Cisero & Royer 1995; Comeau, Cormier, Grandmaison & Lacroix 1999; Dickinson, McCabe, ClarkChiarelli & Wolf 2004; Durgunoglu, Nagy & Hancin-Bhatt 1993; Durgunoglu & Oney 1999; Gottardo 2002; Gottardo, Yan, Siegel & Wade-Woolley 2001; Walpole 2001). Moreover, L1 PA scores in kindergarten have been shown to be predictive of L2 PA scores in rst grade for children learning English as a second language (Cisero & Royer 1995). 2.3. PA and adult monolingual beginning readers Researchers have also investigated PA and its relationship to reading in adult learners who are non-readers. Cross-linguistic research (English, Portuguese,


Margo DelliCarpini

Serbo-Croatian, Spanish, Turkish) suggests that older learners who cannot read an alphabetic system have difculty manipulating phonemes. Non-literate adult learners perform very poorly on traditional PA tasks, like phoneme deletion or substitution (Lukatela, Carello, Shankweiler & Liberman 1995; Morias, Bertelson, Cary & Alegria 1986). Moreover, recent work by Durgunoglu and Oney (2002) suggests that, at least in terms of PA, similar cognitive processes exist between children and non-literate adults acquiring reading skills in Turkey (in their native language) and that adults who received explicit letter, sound and syllable instruction showed more signicant gains in word-recognition and spelling performance (Durgunoglu & Oney 2002:261) than adults enrolled in traditional adult literacy programs in Turkey. To summarize, there is strong evidence for the correlation between PA and early reading in monolingual and bilingual children, the transfer of PA in the L1 to L2 reading, and the positive effect of explicit instruction of PA skills for both children and adults who may have decits in their PA abilities, either due to lack of formal education or a reading disability. Research has also provided evidence for the similarity between early literacy acquisition between bilingual adults and children (DelliCarpini 2006). With the increase of adult ELLs who have not yet learned to read in either their native language or in English as their second language, it would be an obvious extension of the L1 research to investigate the relationship of PA and learning to read in a second language for the rst time for these older learners. To address this, in part, the present study sought to answer the following questions: Does PA in a second language in adult learners predict initial decoding ability in that language (in the same way that it does for monolingual children learning to read for the rst time)? Are some types of PA (as tested by the various subtests) more predictive of initial decoding ability than others for adult L2 learners? If so, which types of PA seem to be important to beginning reading for these learners?

3. The study 3.1. Participants The study involved 26 adult ESL students (native speakers of Spanish) attending adult education programs in a suburban county in New York State. Participants were followed for a one year period and assessed on measures of PA and decoding ability. Participants age ranged from 1846 years old with an average age of 22.7 years. 65% of the participants were male (17) and 35% were female (9).

Phonemic awareness and beginning adult L2 reading


Participants were randomly identied from course rosters and class placement in either Beginning ESL Literacy or Beginning ESL levels.8 Program placement was determined by trained intake counselors employed by the respective programs. Once students were identied as possible participants, an interview was conducted and students with self-reported levels of prior education totaling less than 4 years were included in the study. In fact, the average overall level of prior formal education was very low for these students and all reported being unable to read in English and not reading well or at all in Spanish. Data were collected during four collection sessions over the school year (October, December, February and April). 3.2. Classroom context Participants were enrolled in three related adult education. The rst program was a traditional adult education program that had a large ELL population across several sites in the county and data was collected at two of the sites. The second program was an English Literacy Civics Program (EL/Civics) that focused on the development of English literacy skills through the content area of civics education and citizenship preparation. The third program was an Even Start Family Literacy Program that integrated adult education, parenting education, early childhood education, and interactive literacy activities between caregivers and their children. 3.3. Tasks A battery of commercially produced tasks designed to investigate the level of a subjects PA and decoding ability in English were employed (see appendix A for the instruments used). These tasks were administered for research purposes and were not part of the programs assessment systems. A variety of existing measures were used to ensure that the assessment tools were eld tested and to address a full range of PA skills, options that are not always available when using one instrument. For example, some of the English PA measures were assessed using the Comprehensive Test of Phonological Processing Version II9 /CTOPP (Wagner,Torgesen & Rashotte 1999) which has internal consistency that exceeds .80 and has very limited error potential. In addition, some measures were also taken from the Woodcock Reading Mastery Tests-Revised (1998) which also has reliabilities at the .80 level, so the reliability of assessment tools was ensured in this way. The assessment measures selected are commonly used in both educational settings and research settings to determine the level of PA, the effect of inter-


Margo DelliCarpini

vention on PA ability, and the correlations between PA and decoding, as well as the predictive nature of PA to later decoding ability. Based on the research in the eld, the selected tasks have been shown to be related to each other and to reading. By using commercially produced assessments, similar to those which have been used in numerous other PA research studies, the ndings of this study have the ability, in part, replicate the studies that inform this research. Each of the tasks was administered individually and each testing session took approximately 45 minutes per person. Tasks were administered and scored as per the published directions, and for the purposes of this study, both raw scores and percent correct scores were established (see results section). All tasks were modeled and 35 practice tokens were administered before the scored part of the assessment began to ensure that participants understood the directions (given in both Spanish and English) and task. A brief description of the tasks follows: Segmentation: Two segmentation tasks from the CTOPP were administered. In the rst, phoneme segmentation, the investigator said a word out loud and then asked the subjects to push pennies towards the researcher to indicate the number of constituent phonemes they heard in each word. The instructions10 for this task specically asked participants to push a penny forward for each sound they hear. The second segmentation task, phoneme blending, requires subjects to say a word that the researcher had stretched out. This required subjects to recompose, or blend the constituent phonemes of a word. For example, the researcher would say [k--t] and the correct response would be [cat]. Isolation: In the phoneme isolation task, taken from the Phonological Awareness Test /PAT (Robinson & Slater 1997) participants were asked to identify either beginning or ending phonemes of the word provided (referred to as sounds by the investigator). Deletion: In the phoneme deletion task (CTOPP), participants had to say the word that remains after they delete a phoneme. For example, the investigator gave the directions, take away the rst sound in the word and say what is left. If the stimulus is [black] and the instructed phoneme [b] is deleted, the remaining word is [lack]. Substitution: The phoneme substitution task (PAT) required participants to substitute one phoneme for another and say the new word. For example, if the original word was [black] and participants were instructed to take away the [b] and add [f]. The remaining word was [ack].

Phonemic awareness and beginning adult L2 reading


Decoding tasks: Real and invented word decoding tasks were administered. These tasks were from the Basic Skills cluster of the Woodcock Reading Mastery tests (Woodcock Johnson). The Word Identication Test involves real words and the Word Attack Test involves pseudo words. 3.4. Statistical Analysis Both descriptive and inferential statistical analyses were conducted. Initially, means, standard deviations, ranges, (and percent correct and converted means /CM for PA tasks)11 were computed for all study variables including the PA measures (i.e. phoneme segmentation, phoneme blending, phonological isolation, phonological substitution, and phonological deletion) and the reading acquisition measures (i.e. real word decoding and invented word decoding) at all four time points. Correlations among the ve measures of PA and among the two measures of reading acquisition were then computed at each time point. The next step was to compute total scores for PA and for reading acquisition. This was done by converting each of the raw scores into z-scores (with means of 0 and standard deviations of 1) and then summing the scores separately for each time point. That is, z-scores for the ve measures of PA were summed to arrive at total scores at each time point for each language (hereafter referred to as the PA scores), and z-scores on the two measures of decoding ability were summed to arrive at total scores at each time point for each language. In this manner, each of the ve PA tasks and both of the decoding tasks had equal inuence on the respective composite scores. Two-tailed tests and an alpha ( ) level of .05 were used for all hypothesis tests. At each time point, two multiple regression analyses were performed with the ve PA tasks as the predictors and either (a) decoding invented word scores or (b) decoding real word scores as the outcome variable, for a total of eight multiple regression analyses. To determine which PA tasks were most strongly related to the decoding tasks, both the statistical signicance of each PA task and partial correlations were examined.

4. Results Table 1 illustrates (1) the general task performance at all data collection points and (2) the development of skills over time (means and standard deviations). In addition, the minimum score and maximum score on each task is provided. All twenty-six participants were included in each task. The number of task items,


Margo DelliCarpini

which represents the maximum score possible, is as follows: PA segmentation, 15; PA Blending, 15; PA Isolation, 15; PA Substitution, 15; PA Deletion, 10.
Table 1. Descriptive Statistics for PA Tasks Minimum and maximum performance on each task at each time point with means and standard deviations. Task PA Segmentation English Time 1 PA Segmentation English Time 2 PA Segmentation English Time 3 PA Segmentation English Time 4 Task PA Blending English Time 1 PA Blending English Time 2 PA Blending English Time 3 PA Blending English Time 4 Task PA Isolation English Time 1 PA Isolation English Time 2 PA Isolation English Time 3 PA Isolation English Time 4 Task PA Substitution English Time 1 PA Substitution English Time 2 PA Substitution English Time 3 Min. 0 1 2 2 Min. 0 3 5 8 Min. 0 2 5 2 Min. 0 0 0 Max. 9 9 11 15 Max. 6 7 10 15 Max. 10 12 15 15 Max. 3 7 8 Mean 2.54 2.94 4.15 7.04 Mean 2.66 3.88 4.04 12.38 Mean 3.73 5.46 7.23 9.27 Mean .65 1.31 2.27 SD 2.80 2.80 3.25 3.64 SD 1.79 2.25 2.68 2.40 SD 3.27 3.56 4.08 3.72 SD 1.65 2.02 2.16

Phonemic awareness and beginning adult L2 reading


PA Substitution English Time 4 Task PA Deletion English Time 1 PA Deletion English Time 2 PA Deletion English Time 3 PA Deletion English Time 4

0 Min. 0 0 0 0

10 Max. 10 10 10 10

3.54 Mean 1.35 3.19 4.42 4.42

2.53 SD 3.11 3.42 3.43 3.43

Table 1.1 and 1.2 further illustrate PA task performance. For the purposes of analysis, the rst PA assessment interval was termed the pre-test and the nal assessment interval was termed the post-test. The number of items on each subtest differed, so in order to compare difculty among measures, a converted means (CM) was used (Yopp, 1988). The CM was arrived at by averaging the percent correct for each subject on each of the subtests.

Table 1.1. Pre-test performance Pre-test PA Task Rhyme Isolation (15) Blending (15) Segmentation(15) Deletion (10) Substitution (15) Table 1.2. Post-test PA Task Blending (15) Isolation (15) Segmentation(15) Deletion (10) Substitution (15) Min. 8 2 2 0 0 Max. 15 15 15 10 10 Mean 12.38 9.27 7.04 4.42 3.54 SD 2.4 3.72 3.64 3.43 2.53 % correct 82.5% 61.8% 46.9% 44.20% 23.6% CM 0.838 0.666 0.473 0.469 0.266 Min. Max. Mean SD % correct CM

0 0 0 0 0

10 6 9 10 3

3.73 2.66 2.54 1.35 0.65

3.27 1.79 2.8 3.11 1.65

24.8% 17.0% 16.9% 13.5% 4.0%

0.266 0.186 0.171 0.094 0.066


Margo DelliCarpini

Participants began with very low levels of PA in English. This nding is consistent with ndings from other work investigating the PA skills of nonliterate mono-lingual adults. In addition to development of PA over time for these learners, evidence for a hierarchy of task difculty emerges between time 1 and time 4. Phoneme substitution is most difcult for these participants, and isolation is the easiest for them to perform at all time points. As mentioned earlier, task complexity has been a focus of PA research with children and monolingual adults (Jim nez & e Venegas 2004; Stahl & Murray 1994; Yopp 1988). The results here suggest that PA for this group of students does fall into a pattern very similar to the relative of difculty Yopp (1988) found in her study with kindergarten children.12 In the Yopp study, rhyme was found to be easiest for the children to perform, followed by isolation, and phoneme deletion the most difcult. In the present study, the hierarchy of difculty for the participants, from least difcult to most difcult, was as follows: isolation/blending, segmentation, deletion and substitution. In terms of the relationship between PA and decoding ability, which researchers investigating beginning reading in children learning an alphabetic system have shown to be a critical factor in the process, as mentioned earlier in this paper, the results of this research provide evidence for a high level of correlation between PA and decoding ability at each time point (Table 2). In other words, phonological awareness skills are related to successful decoding for these learners.
Table 2. Correlations between English PA scores and English decoding ability at each time point English PA Time 1 .87 .90 .85 .79

English Decoding ability Time 1 Time 2 Time 3 Time 4 Note. p < .0005.

Time 2

Time 3

Time 4

The correlations were statistically signicant (ranging from .79 to .90 with ps < .0005). The data suggest that PA in English is correlated to decoding ability in English for adult second language learners acquiring initial literacy skills in English and therefore, the presence of PA for ESL literacy level learners is a critical component in learning to read in English.

Phonemic awareness and beginning adult L2 reading


To determine which measures of PA in English are more predictive of decoding ability in English the scores on the individual PA and decoding ability tasks were used, rather than the composite scores. The correlational analyses were followed by two multiple regression analyses at each time point with the ve PA as the predictors and either (a) invented word decoding scores or (b) real word decoding scores as the outcome variable. The rst regression analyses were performed at Time 1. Overall, 85.3% of the variance in real word decoding (RWD) was explained by the ve PA tasks (F(5, 20) = 23.13, p < .0005) and 87.5% of the variance in invented word decoding (IWD) was explained by the ve PA tasks (F(5, 20) = 22.95, p < .0005). In terms of the specic PA tasks, blending ( = .32, t = 2.58, p = .018), isolation ( = .38, t = 2.55, p = .019), and deletion ( = .66, t = 3.74, p = .001) were statistically signicant for RWD. The partial correlation between deletion and real word decoding (.64) was higher than any of the other partial correlations. In terms of IWD, both deletion ( = .72, t = 4.42, p < .0005) and isolation ( = .36, t = 2.61, p = .017) were statistically signicant. The partial correlation between deletion and invented word decoding (.70) was higher than that between isolation and invented decoding (.50), indicating that deletion was the best predictor of invented word decoding. The second regression analyses were performed at Time 2 and at this point in time 71.9% of the variance in RWD was explained by the ve PA tasks (F(5, 20) = 10.26, p < .0005). Additionally, 85.2% of the variance in IWD was explained by the ve PA tasks (F(5, 20) = 28.02, p < .0005). In terms of RWD, none of the PA tasks were statistically signicant individually, but the partial correlation between blending and RWD (.38) was higher than any of the other partial correlations. When looking at the IWD task, only deletion ( = .51, t = 2.58, p = .018) was statistically signicant. The partial correlation between deletion and IWD (.50) was also higher than any of the other partial correlations. The third regression analyses were performed at Time 3 and 61.2% of the variance in RWD was explained by the ve PA tasks (F(5, 20) = 6.32, p = .001). The ve PA tasks explained. 71.4% of the variance in IWD (F(5, 20) = 10.00, p < .0005) at this point in time. In terms of the specic PA tasks, none were statistically signicant but isolation (.40) had the highest partial correlations to RWD. In terms of the correlations between the ve PA tasks and IWD, as with RWD at this time point, none were statistically signicant, although deletion had the highest partial correlation (.32). The fourth regression analysis was performed at Time 4. At this point in time 57.9% of the variance in RWD was explained by the ve PA tasks (F(5, 20) = 5.50, p = .002). In terms of the specic PA tasks, only isolation ( = .44, t = 2.08,


Margo DelliCarpini

p = .0503) approached statistical signicance. The partial correlation between isolation and RWD (.42) was higher than any of the other partial correlations for RWD. In terms of IWD, 60.3% of the variance was explained by the ve PA tasks (F(5, 20) = 6.08, p = .001). In terms of the specic PA tasks, only substitution ( = .56, t = 2.42, p = .025) was statistically signicant. The partial correlation between substitution and IWD (.48) was also higher than any of the other partial correlations.

5. Discussion The present study establishes a relationship between PA and initial decoding ability for adult second language learners who have low or no L1 literacy and who share an L1 (Spanish). There is evidence for some tasks being more highly correlated to decoding in this population, but further investigation will be required to rmly establish these tasks as the best predictors of L2 early reading. The results indicate that PA in the L2 is correlated with initial L2 decoding ability for these adult ESL students. This supports the general claim that there is a causal relationship between L2 PA and L2 decoding ability in literacy level ESL adult students. In terms of correlations, there were signicant correlations between specic tasks and decoding ability. Deletion was most highly correlated with invented word decoding ability, and for real word decoding ability, deletion and isolation tasks were statistically signicant. In research on PA with children, isolation has been shown to be crucial to reading and evidence emerges for its importance in real word decoding in this research with adult L2 adult learners as well. What emerges from the above data is that PA in general is correlated with decoding ability for adult L2 learners of English, but no one individual task emerges as the PA task which could be used as a diagnostic. At different times, different tasks seem to be more highly correlated individually than others; however, the important conclusion is that PA in general is correlated with beginning decoding ability for adult L2 populations and as such, with further investigation into the role of task type could be used as a diagnostic tool as well as an instructional device. Although this research did not specically address the issue of task difference in measuring PA, research that has addressed this issue for children and adults (Jim nez & Venegas 2004; Lukatela, Carello, Shankweiler & Liberman e 1995; Morias, Cary, Alegria & Bertelson 1979; Stahl & Murry 1994) has found phoneme isolation to be the easiest task for learners. An examination of the means (Table 1.1) shows that isolation was the easiest for this group as well,

Phonemic awareness and beginning adult L2 reading


extending the earlier nding to older learners and the adult ELL population. In related research, Jim nez and Venegas (2004) found that for low literacy adult e native speakers of Spanish found phoneme isolation to be the easiest PA task to perform, which supported the ndings from research with children (Stahl & Murray 1994). The results from the present study support the ndings from both of these studies. The question of the path to initial reading ability for older ELLs with little or no L1 reading skills is the primary focus of this research. Despite the current educational focus on reading and literacy skills for learners in all age groups, there has been little to no systematic research on word level strategies and initial literacy development for older ELLs who are learning to read for the rst time in their second language. In order to learn to read an alphabetic system, learners must rst acquire a word level awareness (Calfee, Lindamood & Lindamood 1973; Fox & Routh 1975; Treiman & Baron 1981; Tunmer & Nesdale 1985). For early readers comprehension depends on successful decoding ability (Gough & Juel 1991; Stanovich 1982). If the relationship between PA and decoding ability is similar for adults and children learning to read for the rst time, then, with further research, it may be established that subsequent reading comprehension ability is dependant on successful decoding ability for these adult L2 beginning readers as well. Such a nding would have implications for practice in terms of what pedagogically sound approaches to teaching reading would best meet the needs of literacy level adult ELLs. 5.1. Pedagogical implications The nding that PA is an important factor in decoding success, a critical stage of the alphabetic reading process has implications for practice. It would follow that a routine diagnostic should include a base-line measure of PA using one of many commercially and valid assessment measures. For learners who are found to have decits in PA abilities, pedagogical approaches that recognize this importance and include activities that promote awareness of the sound structure of words may facilitate the development of PA, enhance success during the decoding stage, and put learners on a path to advanced literacy development. The role of word level skills has been under- focused at the adult level for a variety of reasons, but one notable reason is the prevalent use of oral assessment measures in United States federally funded adult ESL programs. When programs and funding agencies rely on an instrument that focuses one skill over others for placement, student assessment, and, most importantly program evaluation and program funding, the types of skills that are included in this assessment may become the skills that are over emphasized in the programs curriculum.


Margo DelliCarpini

An oral assessment measure would certainly contribute to this if there is no counterbalance in the form of reading and writing assessment. For the students who have not mastered reading in their native language, but nd their way to a new country which places a great emphasis on the ability to read and write for communication, social interaction, critical analysis, to acquire information and, perhaps most importantly, to achieve economic selfsufciency, it is critical that they be put on a path of literacy development that draws on their rich schema and gives them access to participation in the community. These students need to develop the basic skills necessary to decode text and then progress through the learning to read stage to one where they are reading to learn. 5.2. Directions for future research The sample size in this study was small, and replication with larger populations is a needed direction for future research. In addition, participants shared the same native language (Spanish) and investigation into populations with different L1 backgrounds, including non-alphabetic backgrounds, is necessary to fully understand the relationship between PA and decoding in adults who are learning to read for the rst time or who are developing emergent literacy skills in their second language. Finally, the nature of instruction was not the focus of the present study, and future research that investigates different instructional approaches and intervention protocols can be conducted to further unpack the role of PA in early literacy development for adult ESL literacy level students.

6. Conclusions This study was designed to investigate the role of PA in initial decoding ability for beginning adult L2 readers of English. Evidence for the correlation of PA to decoding ability has emerged in the population under study. Like the child and monolingual adult studies that have informed this research, the data in this study show that the correlations between English PA and English decoding ability are statistically signicant. Based on the inferential statistical analysis the conclusion can be drawn, as in the child studies, that PA in English is predictive of decoding ability in English for this population. In terms of which tasks were most highly correlated with decoding ability, for both real and invented word decoding, deletion and isolation tasks were the best predictors of decoding ability. This supports the child studies that discuss isolation as being a critical factor in learning to read an alphabetic system. In addition, the data suggest

Phonemic awareness and beginning adult L2 reading


that literacy level adult ESL students and monolingual children have similar underlying cognitive processes, as evidenced for the similarity in the hierarchy of task difculty and the correlation between phoneme isolation and decoding in English and general PA ability in English and English decoding ability. While the results of this study extend the importance of PA skills and initial literacy development in an alphabetic system to an additional population, the results are a small contribution to the area of second language reading research. Further investigation into the development of initial reading in the second language for adults who have not acquired these skills in their L1 and are learning for the rst time in their L2 will provide valuable information for researchers and practioners alike, and will help to inform practice and develop strategies that best meet the needs of students such as the population under investigation in this work.

Appendix Assessment Instruments Employed

Tastk Poneme Isolation Phoneme Substitution Phoneme Segmentation Phoneme Blending Phoneme Deletion Instument Ponological Awareness Test (PAT) PAT Comprehensive Test of Phonological Processing (CTOPP) CTOPP CTOPP

1. National Reporting System (NRS) denitions of adult ESL levels which are used at the U.S. Federal level to determine funding and program placement for adult education programs receiving federal funding; Beginning ESL Literacy: Individual has no or minimal reading or writing skills in any language. May have little or no comprehension of how print corresponds to spoken language and may have difculty using a writing instrument. Beginning ESL: Individual can recognize, read and write numbers and letters, but has a limited understanding of connected prose and may need frequent re-reading; can write a limited number of basic sight words and familiar words and phrases; may also be able to write simple sentences or phrases, including very simple messages. Can write basic personal information. Narrative writing is disorganized and unclear; inconsistently uses simple punctuation (e.g., periods, commas, question marks); contains frequent errors in spelling. 2. Formerly the National Adult Literacy Survey (NALS).


Margo DelliCarpini

3. The NAAL denes literacy from a functional perspective as the ability to use printed and written information to function in society, to achieve ones goals, and to develop ones knowledge and potential. (NAAL Question and Answers, p. 1, USDOE, 2005). 4. The category Non-literate in English is a newly established category that identies all adults who lack minimum basic reading skills necessary to participate in the main National Assessment of Adult Literacy (NAAL). This category was established to answer the outstanding issue of the broadness of category 1 in the 1992 National Adult Literacy Survey (NALS). The non-literate in English category included an assessment for the least literate adults completing the survey (NAAL, USDOE, 2005). 5. The identication of a respondent as Hispanic does not assume that the person is a non-native speaker (NNS) of English, but it can be interpreted from the report that a sub-set of those identied as Hispanic are NNS. 6. The NALS has divided literacy ability into three sections: Prose Literacy, Document Literacy and Quantitative Literacy. In addition, the NALS includes a range of skills that focus on not only decoding and comprehension, but on the range of literacy skills and practices that an adult needs to function in society. Questions used authentic formats and involved test items such as using newspapers (to measure prose literacy) using a transportation schedule (to measure document literacy) and using a loan advertisement that required participants to calculate interest (to measure quantitative literacy). 7. Yopp specically investigated: rhyme, auditory discrimination, phoneme blending, phoneme counting, phoneme deletion, phoneme segmentation, sound isolation and word- to word matching. 8. National Reporting System (NRS) Guidelines: %20Table%204-4-06.doc 9. for ages 724 10. All instructions were given in both Spanish and English. 11. The number of items on each sub-test differed, so in order to compare difculty among measures, a converted means (CM) was used (Yopp, 1988). The CM was arrived at by averaging the percent correct for each subject on each of the subtests. 12. Yopp (1988) did not use phoneme substitution as an assessment.

Abouzeid, Mary. 1992. Stages of word knowledge in reading disabled children. In S. Templeton& D. R. Bear (Eds.), Development of orthographic knowledge and the foundations of literacy, 279306. Hillsdale, NJ: Lawrence Erlbaum Associates. Abu-Rabia, Salim, David Share and Maysaloon Said Mansour. 2003. Word recognition and basic cognitive processes among reading-disabled and normal readers in the Arabic language. Reading and Writing: An Interdisciplinary Journal, 16. 423442.

Phonemic awareness and beginning adult L2 reading


Adams, Marilyn. 1990. Beginning to read:Thinking and learning about print. Cambridge, MA: MIT Press. Adams, Rebecca and Miriam Burt. 2002. Research on reading development of adult English language learners: An annotated bibliography. Center for Applied Linguistics. Retrieved from: ( resources/bibliographies/readingbib. html accessed 23 August 2010). Armbruster, Bonnie, Fran, Lehr and Jean Osborn. 2003. A child becomes a reader: Birth through preschool. Washington, D.C.: National Institute for Literacy. Ball, Ellen and Benita Blachman. 1991. Does phoneme segmentation training in kindergarten make a difference in early word recognition and development of spelling? Reading Research Quarterly, 26(1). 4966. Barone, Diane and Lesley Mandel Morrow. 2003. Literacy and young children: Researchbased practice. New York: Guilford Press. Bentin, Shlomo. 1992. Phonological awareness, reading, and reading acquisition: A survey and appraisal of current knowledge. In R. Frost & L. Katz (Eds.), Orthography, phonology, morphology, and meaning, 193210. Amsterdam, The Netherlands: North-Holland. Barton, David and Mary Hamilton. 1998. Local literacies: A study of reading and writing in one community. London: Routledge. Bradley, Lynette., and Peter Bryant. 1983. Categorizing sounds and learning to read A causal connection. Nature, 301. 419421. Bradley, Lynette, and Peter Bryant. 1985. Rhyme and reason in reading and spelling. Ann Arbor: University of Michigan Press. Burt, Miriam., Peyton, Joy Kreeft, and Rebecca Adams. 2003. Reading and adult English language learners: A review of the research. Washington, DC: National Center for ESL Literacy Education. Burt, Miriam, Peyton, Joy Kreeft, and CarolVan Duzer. 2005. How should adult ESL reading instruction differ from ABE reading instruction? Washington, DC: Center for English Language Acquisition. ( resources/briefs/reading dif.html accessed 23 August 2010). Calfee, Robert.C., Patricia Lindamood & Charles Lindamood. 1973. Acoustic-phonetic skills and reading- Kindergarten through 12 grade. Journal of Educational Psychology, 64(3). 293298. Carrillo, Marisol. 1994. Development of phonological awareness and reading acquisition. A study in Spanish language. Reading and Writing, 6. 279298 Castiglioni-Spalton, Maria and Linnea Ehri. 2003. Phonemic awareness instruction: Contribution of articulatory segmentation to novice beginners reading and writing. Scientic Studies of Reading, 7. 2552. Castles, Anne. & Max Coltheart. 2004. Is there a causal link from phonological awareness to success in learning to read? Cognition, 91(1), 77111. Center for Applied Linguistics. 2003. Best Plus. Washington DC: CAL. Chard, David., and Shirley Dickson. 1999. Phonological awareness: Instructional and assessment guidelines. Intervention in School and Clinic, 34(5. 261270.


Margo DelliCarpini

Cisero, Cheryl, and James Royer. 1995. The development and cross-language transfer of phonological awareness. Contemporary Educational Psychology 20(3., 275303. Comeau, Liane, Pierre Cormier, Eric Grandmaison and Diane Lacroix. 1999. A longitudinal study of phonological processing skills in children learning to read in a second language. Journal of Educational Psychology, 91(1). 2943. Cunningham, James, Patricia Cunningham, James Hoffman and Hallie Yopp. 1998. Phonemic awareness and the teaching of reading. Newark, NJ: International Reading Association. DelliCarpini, Margo. 2006. Early reading development in adultELLs. The Academic Exchange Quarterly, 10(2). 192196 Dickinson, David., Allyssa McCabe, Nancy Clark-Chiarelli and Anne Wolf. 2004. Cross-language transfer of phonological awareness in low-income Spanish and English bilingual preschool children. Applied Psycholinguistics, 25. 323347. Droop, Mienke., and Ludo Verhoeven. 2003. Language prociency and reading ability in rst- and second-language learners. Reading Research Quarterly 38 (1). 78103. Durgunoglu, Aydin Ycesan and Banu ney. 2002. Phonological awareness in literacy acquisition: Its not only for children. Scientic Studies of Reading, 6(3). 245266. Durgunoglu, Aydin Y cesan, William Nagy & Barbara Hancin-Bhatt. 1993. Cross Lanu guage Transfer of Phonological Awareness. Journal of Educational Psychology, 85 (3). 453465. Ehri, Linnea. C., Simone Nunes, Dale Willows. Barbara Valeska Schuster, Zohreh Yaghoub-Zadeh and Timothy Shanahan. 2001. Phonemic awareness instruction helps children learn to read: Evidence from the National Reading Panels meta-analysis. Reading Research Quarterly, 36. 250287. Ehri, Linnea, Lee Wilce and B. Taylor. 1988. Childrens categorization of short vowels in words and the inuence of spellings. In K. E. Stanovich (Ed.), Childrens reading and the development of phonological awareness, 149177. Detroit, MI: Wayne State University Press. Eskey, David. 1988. Holding in the bottom: An interactive approach to the language problems of second language readers. In P. Carrell, J. Devine, & D. Eskey (Eds.), Interactive Approaches to Second Language Reading. 93100. Cambridge: Cambridge University Press. Fox, Barbara, & Donald Routh. 1984. Phonemic analysis and synthesis as word attack skills: Revisited. Journal of Educational Psychology, 76. 10591064. Gee, James Paul. (1992). The social mind: Language, ideology, and social practice. New York: Bergin & Garvey. Goswami, Usha. 1993. Towards an interactive analogy model of reading development decoding graphemes in beginning readers. Journal of Experimental Child Psychology, 56. 443475. Goswami, Usha., & Peter Bryant. 1990, Phonological skills and learning to read. London: Erlbaum Gottardo, Alexandra. 2002. The relationship between language and reading skills in bilingual Spanish-English speakers. Topics in Language Disorders, 22(5). 4670.

Phonemic awareness and beginning adult L2 reading


Gottardo, Alexandra., Yan, Bernice., Linda Siegel and Lesley Wade-Woolley.2001. Factors related to English reading performance in students with Chinese as a rst language: More evidence of cross-language transfer of phonological processing. Journal of Educational Psychology, 93. 530542. Gough, Philip, and M. L. Hillinger. 1980. Learning to read: An unnatural act. Bulletin of the Orton Society, 30. 179196. Gough, Philip, and Connie Juel,. 1991. The rst stages of word recognition. In L. Rieben & C. A. Perfetti (Eds.), Learning to Read, 4756. Hillsdale, NJ: Erlbaum. Hinkel, Eli. 2006. Current perspectives on teaching the four skills. TESOL Quarterly, 40 (1). 109131. Jim nez, Juan and Enrique Venegas. 2004. Dening phonological awareness and its relae tionship to reading skills in low-literacy adults. Journal of Educational Psychology, 96 (4). 798810. Kim, Jeesun. & Chris Davis. 2004. Characteristics of poor readers of Korean hangul: Auditory, visual and phonological processing. Reading and Writing, 17(1/2). 153 185. Klassen, Cecil. & Barbara Burnaby. 1993. Those who know: Views on literacy among adult immigrants in Canada. TESOL Quarterly, 27 (3). 77397. Liberman, Isabelle Yoffe, Donald Shankweiler and Alvin Liberman. M. 1989. The alphabetic principle and learning to read. In Shankweiler, D. and I.Y Liberman, (Eds.) Phonology and reading disability: solving the reading puzzle, 133. International Academy of Research on Learning Disabilities Monograph Series. Ann Arbor: University of Michigan Press. Lundberg, Ingvar. 1989. Lack of phonological awareness: A critical factor in dyslexia. In C. V. Euler, I. Lundberg, & G. Lennerstrand (Eds.), Brain and reading, 221232. London: Macmillan Press Limited. Mann, Virginia. 1993. Phoneme Awareness and Future Reading Ability. Journal of Learning Disabilities, 26 (4). 259269. Mann, Virgina.. 1984. Longitudinal prediction and prevention of early reading difculty. Annals of Dyslexia, 34. 117136. McCutchen, Deborah., Robert Abbott, Laura Green, S. Natasha Beretvas, Susanne Cox, Nina Potte, Teresa Quiroga and Aaudra Gray. 2002. Beginning literacy: Links among teacher knowledge, teacher practice, and student learning. Journal of Learning Disabilities, 3(1). 6986. National Assessment of Adult Literacy. 1992 & 2003. National Institute for Literacy 2006. Frequently asked questions: What is literacy? Washington, DC: NIFL. ( accessed 23 August 2010). Rack, John., Margaret Snowling and Richard Olson. 1992. The nonword reading decit in developmental dyslexia: a review. Reading Research Quarterly, 27(1). 2853. Robson, Barbara. 1982. Hmong literacy formal education, and their effects on performance in an ESL class. In B. T. Downing & D. P. Olney, (Eds.), The Hmong in the west: Observations and reports, 201225. Minneapolis, MN: University of Minnesota.


Margo DelliCarpini

Sabatini, John., Richard Venezky, Polina Kharik and Richa Jain. 2000. Cognitive Reading Assessment for Low Literate Adults: An Analytic Review and New Framework. Technical Report TR00-01. Sponsored by Ofce of Educational Research & Improvement, Washington DC Philadelphia, PA. National Center on Adult Literacy (NCAL) Slavin, Robert., & Alan Cheung. 2005. A synthesis of research on language of reading instruction for English language learners. Review of Educational Research, 75 (2., 447284. Smith, Sylvia, Deborah C.Simmons & Edward Kameenui. 1995. Synthesis of Research on Phonological Awareness: Principles and Implications for Reading Acquisition. (Technical Report No. 21). Eugene: National Center to Improve the Tools of Educators, University of Oregon. Spector, Janet. 1995.Phonemic Awareness Training: Application of Principles of Direct Instruction. Reading and Writing Quarterly: Overcoming Learning Difculties, 11(1). 3752. Sprugevica, Ieva. & Torleiv Hoien. 2003. Enabling skills in early reading acquisition: A study of children in Latvian kindergartens. Reading and Writing16(3). 159177. Stahl, Steven and Bruce Murray. 1994. Dening phonological awareness and its relationship to early reading. Journal of Educational Psychology, 8. 221234. Stanovich, Keith. 1982. Individual differences in the cognitive process of reading. Journal of Learning Disabilities, 15. 485493 Strucker, John. 1997. What silent reading tests alone cant tell you. Focus on Basics, 1 (B). 1317. Stuart-Smith, Jane and Deirdre Martin. 1997. Investigating literacy and pre- literacy skills in Panjabi/English schoolchildren. Educational Review, 49(2). 181197. Torgesen, Joseph., Richard. K., Wagner, Carol. A. Rashotte, Elaine Rose, Patricia Lindamood, Tim Conwayand Cyndi Garvan. 1999. Preventing reading failure in children with phonological processing difculties: group and individual responses to instruction. Journal of Educational Psychology, 91 (4). 579593. Treiman, Rebecca and Jonathon Barron. 1983. Phonemic analysis training helps children benet from spelling-sound rules. Memory & Cognition, 11. 382389. Tunmer, William, and Andrew Nesdale.1985. Phonemic segmentation skill and beginning Reading. Journal of Educational Psychology, 77 (4). 417527. U.S. Census Bureau: Census 2000: U.S. Department of Education, Ofce of Vocational and Adult Education. 2006. Adult education and family literacy act: Report to congress on state performance. Washington, DC. U.S. Department of Education, Ofce of Vocational and Adult Education, Division of Adult Education and Literacy 2005. State administered adult education program, program year 20032004 enrollment. Washington, DC. U.S. Department of Education. 1992. National adult literacy survey. Washington, DC: National Center for Education Statistics. ( accessed 11 July 2006).

Phonemic awareness and beginning adult L2 reading


Vellutino, F. R., & D. M. Scanlon. 1988. Phonological coding, phonological awareness, and reading ability: Evidence from a longitudinal and experimental study. In K. E. Stanovich (Eds.), Childrens reading and the development of phonological awareness, 77120. Detroit, Michigan: Wayne State University Press. Vellutino, Frank and Donna Scanlon. 1987. Phonological Coding, Phonological Awareness, and Reading Ability: Evidence from a Longitudinal and Experimental Study. Merrill-Palmer Quarterly, 33 (3). 321363. Vellutino, Frank and Donna Scanlon. 1991. The effects of instructional bias on word identication. In L. Rieben and C. A. Perfetti (Eds.), Learning to Rea, 189204. Hillsdale, NJ: Erlbaum. Wagner, Richard., Joseph Torgesen, Pamela Laughon, Karen Simmons & Carol Rashotte. 1993. Development of Young Readers Phonological Processing Abilities. Journal of Educational Psychology, 85 (1). 83103. Wagner, Richard, Joseph Torgesen and Carol Rashotte. 1994. Development of reading related phonological processing abilities: New evidence of bi-directional causality from a latent variable longitudinal study. Developmental Psychology 30. 7387. Wagner, Richard., Joseph Torgesen and Carol Rashotte. 1999. Comprehensive test of phonological Processing/CTOPP Austin, Texas: PRO-ED. . Wagner, Richard, Joseph Torgesen, Carol Rashotte, Steve Hecht, Theodore Barker, Stephen Burgess, John Donahue and Tamara Garon. 1997. Changing relations between phonological processing abilities and word level reading as children develop from beginning to skilled readers: A 5-year longitudinal study. Developmental Psychology, 33 (3.) 468479. Yopp, Hallie. 1988. The validity and reliability of phonemic awareness tests. Reading Research Quarterly 23, 159177. Margo DelliCarpini received her Ph.D. in Linguistics and Masters degree in TESOL from Stony Brook University. She is a professor of TESOL at Lehman College, CUNY where she works with TESOL teacher candidates and conducts research on second language literacy development and teacher collaboration. Email her at margo.dellicarpini@lehman.

Identity repertoires on the Internet: Opportunities and constraints


Abstract Over the past two decades, the rise of the Internet has enabled new forms of social relationships, new forms of interaction and community building, and new opportunities for creating, articulating and ratifying identities. These new developments have both positive and negative aspects. The positive ones are the new patterns and practices which enable people to create something new, and the Internet indeed offers quite a range of such opportunities for expanding, altering or developing identity repertoires. The negative ones, which are just as important as the opportunities, are the constraints on such forms of creativity. Through two examples from emerging Chinese Internet (sub)cultures, this paper explores the dynamic between opportunities for and constraints on the creative deployment of identity repertoires on the Internet. Keywords: Identity, identity repertoire, the Internet, China

1. Introduction I Twitter, therefore I am this is the title of a symposium held at Tilburg University in the summer of 2010.1 The title is not so surprising, or at least not at present. A mere ve years ago it would have been unthinkable, and the speed and degree to which we have been habituated (or, as some would say, socialized) to new media formats as tools for being or becoming someone is something that seems to often escape the attention of observers. Things move fast, and research only catches up with new developments at a relatively slow pace.2 The effect of that is that papers such as this one face challenges, the main one being that they have to summarize and comment on an ever-broadening range of recent fast developments in other words, shoot a moving target. Another one is that they risk becoming dated in no time: for instance, while Facebook at the time


Piia Varis, Xuan Wang and Caixia Du

of writing has become one of the biggest changes in society and culture ever seen its reported 500 million users and the 700 billion minutes these people spend on it per month testify to this (Facebook 2010) it is not beyond the bounds of possibility that no one will even remember Facebook a decade from now. The rise of the Internet has over the past two decades generated a whole range of new patterns of human interaction, enabled new forms of human relationships and, consequently, also offered new opportunities for creating, articulating and ratifying new identities. The mechanisms of these new developments are being explored, even if research on it is still quite fragmentary. However, we believe that we do see some general lines developing here, positive ones as well as negative ones. The positive ones are the new patterns and practices which enable people to create something new, and the Internet indeed offers quite a range of such opportunities for expanding, altering or developing identity repertoires. The negative ones are the constraints on such forms of creativity, and we want to emphasize from the outset that the constraints are as important as the opportunities, even if constraints perhaps have not yet drawn sufcient academic attention. We will attend to these issues rst by means of a broad sketch of the main issues surrounding the theme of this paper. We shall see that virtual reality is quite a misnomer, because what is virtual is very real in the lives of many people, especially when identity processes are concerned. We shall also see that throughout these identity processes, issues of authenticity and authentication, ratication and legitimation are salient, and that we encounter Internet ideologies in this eld dominant ideologies of freedom and liberty that guide people towards new developments in their identity repertoires. Such ideologies need to be set off against the background of the panoptic and regulated nature of the Internet a virtual space of real control and power. This is where constraints come into play. Doing identity work on the Internet has to follow certain rules and norms: in addition it is open to scrutiny and assessment by often unknown and/or unsuspected others. The two cases presented here, taken from emerging Chinese Internet subcultures, can illustrate this dynamic between opportunity and constraint. This is the general plot of the paper, but before embarking on our discussion we need to introduce two perhaps rather obvious but nevertheless fundamental introductory points. First, what happens on the Internet can naturally not be dissociated from the wider patterns of social transformation we have witnessed over the past decades globalization processes entailing a complex of social, cultural, political and economic features. People, capital and goods move in new ways; the patterns of circulation of popular culture have altered, and we have

Identity repertoires on the Internet: Opportunities and constraints


new truly global art forms such as hiphop; the meaning of key concepts in the social sciences such as community, diaspora, culture, and identity have been altered, and the speed and scope of these processes have indeed often been attributed to the Internet. It goes without saying that what follows needs to be set against that background (see e.g. Nederveen Pieterse 1995; Appadurai 1996; Castells 1996; Maffesoli 1996; Block 2004; Androutsopoulos 2006; Georgakopoulou 2006; Karaganis 2007 for different perspectives on these developments). Second, when we talk about identity the specic focus here we have to take on board the achievements of decades of advanced scholarship (here we can think of such diverse works as, for instance, Goffman 1981; Butler 1990; Hall and du Gay 1996; Gumperz and Cook-Gumperz 1997 [1982]; Blommaert 2005). That is, we cannot see identity as an essential feature of individuals, but as a bundle of processes and practices. We also need to pluralize identity, and start from the assumption that people do not have one identity, but perform a repertoire of identities by means of resources they have acquired and have at their disposal for such purposes. Language, forms and norms of communication, genres and styles all belong to that complex of resources, and while none of those resources alone is enough to generate a particular identity, none of them can be overlooked in explaining patterns of identity production. Third, we have to understand that identities are dynamic and changeable, and that people invest considerable amounts of energy in modifying and changing their identities in relation to other people and different contexts. Fourth, from this it follows that we should not think of identity just as something that is produced by someone as an isolated self-sufcient entity; we also need to consider its uptake and response from others, for identity is dialogical. And, nally, although the performative, plural, dynamic and dialogical character of identity may be at odds with lay discourses and self-perceptions which stress singularity and stability (I havent been myself lately), we need to accept that all of this is normal, and indeed a core characteristic of the social processes we observe and examine. With these general and also widely subscribed to remarks in mind, we can now engage with the particular eld we want to address here: the late modern jungle of Internet identities.

2. Virtually real identities Internet identities and online self-presentation have recently attracted more and more academic attention (see e.g. Turkle 1995; Bargh et al. 2002; Ellison et al. 2006; Lepp nen 2007; Zhao et al. 2008). In online environments, we have to a


Piia Varis, Xuan Wang and Caixia Du

write ourselves into being (Sund n 2003) or, as danah boyd3 (2009: 145) e puts it: One cannot simply be online; one must make ones presence visible through explicit and structured actions. In practice, having to write oneself into being means that on many forums, one can start from scratch, and write into being the kind of being one wants to be. Here we of course encounter differences between anonymous and nonymous sites for identity construction: for instance a social network site such as Facebook is a nonymous site; users present themselves there, in many if not most cases, with their real name, with a picture of themselves attached to that name to further authenticate their real identity. On anonymous sites we perhaps see more room for manoeuvring and identity play we are for instance able to present ourselves with a self-invented user name. Nonymous sites, however, should by no means be seen as limiting the creative mobilization of different identity repertoires, for apart from selecting what information one wants to present of oneself, which aspects of ones identity to highlight and which ones to conceal, one is also to a certain extent free to choose ones audience, for instance in the case of Facebook, ones Friends (see also Gershon 2010: 174179). The audience one constructs for ones identity performance also has an effect on what is presented and what is not, for the kind of identity one wants to perform has to be authenticated and ratied by those observing the performance. Nevertheless, what having to write oneself into being means is that because of this have to this obligation we must experiment with different identity repertoires. We have to write ourselves into being, and for this purpose we mobilize different resources genres, styles, discourses, and so on that we have at our disposal (see e.g. Lepp nen et al. a 2009). An upshot of this, of course, is that in online environments we encounter what Naomi S. Baron (2008) has referred to as the on my best day phenomenon. In the (at least to some extent) disembodied world of virtual identities, the identity written into being does not need to correspond to the corporeal self and its material conditions. The specic environment of the Internet compels us towards the use of certain types of identity resources. It is naturally clear that selective self-presentation and identity work also takes place ofine we need to manoeuvre different audiences and different contexts in our everyday lives all the time (and it is noteworthy that for instance Goffmans [1959] work has been applied to studying online environments too). However, in online environments the scarcity of clues ones audience often has for determining the authenticity of ones identity claims i.e. whether the claims to certain identities correspond to the corporeal self behind the computer screen and its material conditions further facilitates selective self-presentation and creativity. In contrast to ofine social relationships, which derive much of their meaning from their history and

Identity repertoires on the Internet: Opportunities and constraints


context, people often enter Internet platforms with no history, or they perform to an audience they have specically found or summoned for a certain performance to take place. Internet identities written into being can also be conceptualized on the axis of visibility-invisibility. One can choose the arenas on which one wishes to display oneself and write oneself into being: where to be seen, with whom (who are my Friends, and what does that suggest about me and my identity; see e.g. Walther et al. 2008), and with what sort of identity claims, both implicit and explicit. One can again make oneself invisible by choosing to stop engagement with these arenas, and move to (an)other one(s) perhaps those which allow for more freedom in terms of identity construction. If nonymity becomes a burden, there is also the possibility of writing oneself into being with an alias, or an avatar. In the latter case, one gets both visibility and invisibility; an alias or an avatar will do the identity work, be visible, while ones own name and face can remain invisible. That is, the substance of what is ones identity can be crafted using all kinds of resources for making things either appear, come into being, or disappear. All this of course comes in a convenient wrapping of an ideology of individual freedom and liberty. Given its nature, and the opportunities it provides for social action, the Internet can be seen as a space in which people (and not only young ones) can experiment with new and different forms of self-presentation, engagement with others, and forms of community in what is often experienced as a space of unrestricted creativity. While this range of new opportunities can give the impression of a chaotic and limitless universe of human social creativity, it is useful to draw attention to the enormous degree of stability, and even predictability, that we see through all this. Stability online takes many forms; there are explicit and implicit rules of conduct that we have to follow at least if we want to make our identity endeavours successful in the sense of being accepted and authorized by others inhabiting the same online world and architectural constraints to be negotiated. Acknowledging the stabilities and normativities is not to deny the fact that the range of new genres that the Internet offers us means increasing possibilities for identity and community formation here we only have to think of the emergence of blogging as a format for the creation of all sorts of worlds that were previously unimaginable. However, identities are performed within these genres, and genres themselves are normative not a case of anything goes, at least if one wants to make oneself understood and accepted. The emergence of new environments and new genres may give us the illusion of increased freedom, for in new environments norms emerge and those appropriating these environments and employing the new genres bring about the stabilisation of


Piia Varis, Xuan Wang and Caixia Du

certain practices and behaviours and thus participate in the creation of norms. Yet, users themselves despite what they may think are not the (only) norm setters: communicative acts, self-presentation and social relationships need to be managed and manoeuvred within certain parameters (see boyd 2001; Killoran 2003; Thurlow et al. 2004; Herring and Paolillo 2006; Lanier 2010 for different perspectives). All of this is also very consequential: careless self-presentations may come back to haunt one later in life youthful misjudgements in self-presentation may still be accessible years afterwards to for instance future employers searching for information on a potential employee. With online environments, we also run the risk of being taken out of context (boyd 2008), and things said and done for certain communicative and identity purposes can be recontextualized elsewhere, for completely different purposes. Identity work carried out in online environments can be stored and used for other (identity) purposes later and, consequently, as an effect of this archive function of the Internet, we have archontic power (Derrida 1996): that is, there are (invisible) others who are in a position to collect, classify and interpret information provided by us, hence exercising power on us and the online archive we construct for and of ourselves (see also e.g. Richardson and Hessey 2009 for a discussion on Facebook as a way of archiving the self; also OHara et al. 2009; Weisbuch et al. 2009). This is also when the Internet becomes a late-modern panopticon (Foucault 1995 [1975]), allowing for observation, surveillance and information-gathering without those who are being observed and of whom information is being gathered necessarily being aware of it. A further complication in online identity work is that in many ways we have to comply with ready-made identity formats. In many virtual environments, becoming someone means writing oneself into being according to pre-selected identity markers such as the familiar A/S/L (age-sex-location) grid (e.g. boyd 2001). This means that part of our identity work is already laid out for us, and that the more subtle means we have for identity construction ofine are not available for us. Further, with such identity markers we often also become attached to a certain physical body in time and space which is in apparent contradiction with the idea of freedom and liberty online to be who one wants to be in the way one wants to be. Although it can be suggested that such ready-made identity formats facilitate communication and interaction, the motivation for them is of course in many cases economic. That is, our identity work can be used by (invisible) others for marketing and other purposes which may have nothing to do with the original purpose for which the online identity was constructed (boyd 2001). Facebook users, for instance, are not only making identity statements to their visible audience (Friends), but also to invisible parties who want to sell their products and

Identity repertoires on the Internet: Opportunities and constraints


services hence for example the phenomenon of gender-specic advertising that those presenting their gender in their prole get (see also Jespers 2010). Thus, while our rst impression of virtual identity processes may be that of bewildering diversity and freedom, a closer look yields a somewhat different picture: rules are designed and followed, people follow explicit and implicit norms and nd themselves faced with non-negotiable platforms, often with unforeseen consequences. This is also when the virtualbecomes real: a part of real economic, social, cultural and political processes of evaluation, stratication and control. Ideologies of freedom and creativity, revolving around the presumed virtual nature of Internet spaces, obscure this fundamental reality of what goes on there. Although Tsuis (2003: 66) view that rather than being a technology of freedom, the internet is well on its way to becoming a technology of control is not, as we have seen, the whole story, we have to be mindful of the constraints that are included in the neat package of freedom. We shall next illustrate this point with two examples from emerging Internet (sub)cultures in China.

3. Creativity within limits: Two cases from China We should start our exploration of the limitations with a more general point concerning constraints. What we wish to illustrate here is the kinds of restraints people are confronted with within different online environments. But we should start a bit further back, and consider who exactly are these people with the privilege to engage in the new virtual identity practices. According to a report by the state-run China Internet Network Information Center (2010), the number of Internet users in China by June, 2010 had reached 420 million. The speed with which the number of Chinese Internet users (or netizens, as they are called ) has increased is quite remarkable: between the end of 2009 and June, 2010, there has been a reported increase of 36 million users (China Internet Network Information Center 2010). In June 2010, the Chinese Government reported its determination by the year 2015 to increase the number of those with Internet access to 45 per cent of the population (Information Ofce of the State Council of the Peoples Republic of China 2010). The Internet is also becoming an important tool for personal expression and public participation in China, and its potential in bringing about social change and enabling and facilitating organised forms of political action should not be overlooked (e.g. Yang 2003a; Wang and Hong 2010). It seems that more and more people in China do have Internet access, but the technologies of freedom are not available for everyone. For instance, geographical distribution of access is uneven: those living in urban spaces are more


Piia Varis, Xuan Wang and Caixia Du

likely to enjoy the availability of the new forms of identity and community building than those living in rural areas, and those living in western China are less privileged in terms of Internet access than their fellow citizens in eastern parts of the country (Information Ofce of the State Council of the Peoples Republic of China 2010). Thus, when it comes to the opportunities for new technologically mediated identities, obviously there are haves and have-nots already in terms of access to the resources for realizing these opportunities. However, this does not mean that these opportunities are totally out of reach for people living on the margins of societies. This becomes clear with our rst case4 a rapper from Enshi, China (a remote rural area in Central-Western China) who can be considered as someone absolutely marginal in multiple senses of the word and yet has access to opportunities provided by the Internet. His personal trajectory is easily summarized. He is an educational underachiever and a school dropout who rebelled against the educational system. Something of a juvenile delinquent, he has been involved in street crime, and has been in prison several times. Apart from a small circle of friends, from a similar disadvantaged background, he does not have an extensive social network. With his record, he is also basically unemployable. What we can conclude from this brief biography is that we are looking at a case of someone who is short of almost every form of capital (Bourdieu 1986): the economic, cultural and social resources he has at his disposal for building meaningful group belongings and social networks, for improving his mobility and, consequently, for accumulating further resources, seem scarce. The material reality of this rapper thus seems to be that of someone socially very invisible, but this is only one part of his lived reality. He does live with his parents, in a small bedroom with hardly any comforts. But he has a computer connected to the Internet, and this is where the other part of his existence starts. This underachiever may not appear to be in possession of a lot of social capital, but through his musical endeavours he has been able to generate social networks and visibility for himself. Music itself of course functions as a kind of a (sub)cultural glue in the lives of many people around the world, and our underachiever has been able to get some status for himself by engaging in the global phenomenon that is hiphop. However, what we also know about our rapper is that he does not perform or tour much his visibility is conned to the Internet. Online, we see him appear in many different forums, most notably www.51. com and, which can be characterized as a social network site, and a MySpace-kind-of-platform, or a music-DIY site, respectively. On these forums he promotes his music, builds social networks, is part of communities of likeminded people and, even though it may be surprising in the case of such an

Identity repertoires on the Internet: Opportunities and constraints


underachiever, he even has followers (and even became a topic for research in a Dutch university). On different online forums, through the different multimodal resources available, he is able to write another kind of self into being. For this rapper from Enshi, China, the Internet enables the deployment of identity repertoires otherwise not within his reach. Here we could apply the idea of the online disembodiment thesis (e.g. Campbell 2004), according to which online and ofine worlds and experiences can be different and separated, and corporeal, ofine bodies and realities do not necessarily matter online. In a sense this is what happens here: an underachiever can become someone with status; someone with a legitimate identity within an essentially global hiphop network in spite of being locally marginalized; someone whose existence as a certain kind of self is authenticated by his followers and others with whom he interacts online. However, that is not the whole story and, unfortunately, our rapper does not have unlimited freedom in establishing himself as a certain kind of self. His identity repertoires may have expanded with the Internet and he is able to escape the material locality he is in, but these new opportunities also come with constraints. The rst constraint we can observe is quite a practical and also a major one: his carefully crafted identity work may disappear anytime. Websites may be shut down. That is the case at least with a site that has more than 100 million users, therefore by no means a minor social tool which was shut down for undeclared reasons earlier in 2010. If the site is blocked, its more than 100 million users are not able to download music, interact with each other and work on their social networks and identities. The freedom of our rapper and these other users is limited. China proclaims itself to be a harmonious i.e. homogeneous and highly normative society, and does not approve of the dissemination of harmful information online (see e.g. Zittrain and Edelman 2003; MacKinnon 2008, 2009; Zhang and Wang 2010). Hence the limitations for someone like the Enshi rapper, who on the surface seems quite harmless. He is, after all, only broadening his identity horizons and building a community of music enthusiasts. However, he is part of a subculture, and his specic art form (hiphop) can also be viewed negatively as too Western and not an authentic part of Chinese culture. He also engages in asking provocative questions online, for instance concerning the freedom of speech in China, and some of the lyrics for his songs that he posts online feature profanities. In no time at all, these things which are clearly identity statements, and help construct an image of someone who is not only part of a subculture, but also outspoken, bold and critical may disappear. We can safely say that at least some of the rules and norms the users of have to comply with are by no means implicit. On the contrary,


Piia Varis, Xuan Wang and Caixia Du

Figure 1. The presence of Cyberpolice at

identity and community formation are literally policed in a very explicit way, and this is clear from the presence of the Chinese Cyberpolice on the site (see Figure 1). The Cyberpolice monitor online behaviour and the posting of harmful information wrong kinds of identity statements, wrong forms of interaction and relationship formation have undesired consequences (see e.g. Qiu 1999; MacKinnon 2008). Posts, messages and whole sites, along with certain selves written into being and made visible, may disappear.5 What the Enshi rapper is forced to do, then, is to migrate from one site of engagement to another, and whichever site happens to be fully functioning will be his main forum. So essentially he needs to be able to do two things: be mindful of what he says mindful of the kinds of repertoires and resources he mobilizes and be ready to be a nomadic user, and move from one forum to another. The rapper clearly benets from the identity opportunities the Internet can provide. It allows him to become more widely recognized and to break out of Enshi. But this he does within a policed Internet environment, in which his opportunities need to be carefully checked against the risks of this enterprise. An additional facet of this is that he is also transgressing rules he may not even know to exist, and in this he is not alone. We can witness a similar dialectics of opportunities and constraints within the budding and in many ways already blooming Chinese blogging culture, which is our second example.6 Blogging in general has become hugely popular and also increasingly inuential during the past decade or so. Many efforts have been made to characterize this new genre (e.g. Herring et al. 2005), but it seems to escape denition. As Miller and Shepherd (2004: n.p.) put it, Blogs can be both public and intensely personal in possibly contradictory ways. They are

Identity repertoires on the Internet: Opportunities and constraints


addressed to everyone and at the same time to no one. That is, blogs can be characterized as highly personal as a late-modern form of diary (e.g. McNeill 2003; Rak 2005) while at the same time they can be extremely public in nature. Audiences can also be unexpected and unsuspected. There might be a familiar circle of readers, consisting of, say, friends, family, or colleagues, but a blogger may also have (and all this unassumingly) a very wide audience, and members of this audience can naturally be physically located anywhere in the world provided with Internet access. Blogging can also be viewed within the framework of the new knowledge culture (Jenkins 2006; see also L vy 1997), where new types of community, e notably in the cyberspace, are formed. To quote Jenkins (2006: 27),
(. . . ) these new communities are dened through voluntary, temporary and tactical afliations, reafrmed through common intellectual enterprises and emotional investments. Members may shift from one group to another as their interests and needs change, and they may belong to more than one community at the same time. These communities, however, are held together through the mutual production and reciprocal exchange of knowledge.

These communities, with shared objectives (and here we may think of a range of objectives, be they purely intellectual or concerned with knowledge sharing, or ones with affective functions as well), appear online in many different forms, and a community of bloggers can be viewed as a collective knowledge community. The knowledge exchanged, shared and built upon may be from any sphere of life; even a very supercial exploration of the blogosphere allows one to discover individuals and communities posting on cooking, gardening, sexuality, fashion, eating disorders, and so on. These communities come together through, for instance, blogrolls (i.e. a list of blogs that a blogger may recommend to others by providing links to them) and comments people post on blogs. The opportunities offered by this new online genre for information, identity and identication are mind-boggling. To take the specic location of China as an example, it is reasonable to suggest that such opportunities used to be centrally managed identities ascribed to one were more difcult to escape, and the ones introduced as forbidden ones more difcult to perform. However, with the emergence of the Internet and blogging as part of it, the expansion of ones identity repertoires is easier, not to mention the implications for collective work and community building (see e.g. Yang 2003b for musings on the cyberpotential for these purposes). As Jenkins (2006: 29, drawing on L vys 1997 work) indeed notes, the new e communities emerging may also serve political functions, and be instrumental in processes of democratization. To take the case of the Chinese blogging scene


Piia Varis, Xuan Wang and Caixia Du

as an example (and to risk over-simplication by referring to it as a single entity), the political functions may of course be either the main goal of bloggers, or the side product of blogging on something with no directly observable political meaning. The case we present here is an example of the latter. To reiterate our point of departure, we do see that blogging means increased opportunities for identity construction and community building. However, as we saw with the case of the Enshi rapper, these opportunities come with restraints. What is seen as harmonyin Chinese society entails not saying and doing certain things, not even in the land of freedom that the Internet is supposed to be. As the case of the rapper already illustrated, there are consequences for making certain identity statements and trying to employ certain identity repertoires. Blogging makes no exception in this. This was perhaps most obvious prior to the 2008 Beijing Olympics, when a lot of cleaning both virtual and material had to be done to present to the rest of the world a certain kind of China. A harmonious virtual China does not include vulgar language or swear words, and politically dubious statements are naturally seen as threatening. Prior to the Olympics, an Internet forum that had been running for about ve years was shut down after someone posted a critical comment on their blog on the building of expensive stadiums while great numbers of Chinese were living in poverty. The blog in question was not created for overtly political purposes; it was one on literature a (sub)cultural forum for likeminded people to share and create knowledge and establish identities. In this case, as in many others, someones (virtual) life and identity was erased, and the blogger forced to migrate and display his/her identity elsewhere on the Internet. In other cases, we witness increasing individual and collective creativity to manoeuvre through the screening of identities and practices and, ultimately, to avoid erasure. There are CD-roms available for unblocking websites that have been closed down, and we are also able to observe astonishing degrees of linguistic creativity from bloggers and other Chinese Internet users to make sure that they can avoid control and censorship. In practice this means metaphorical talk, indirect and obscure expressions and the use of homophones for instance to be able to say what one wants for certain identity and communal purposes as has been the case recently with online discussions on Liu Xiaobo, the 2010 Nobel Peace Prize winner. Quickly these roundabout ways of expression become shared knowledge among Internet users and communities and identity work can remain under construction. Here we see the precariousness of identities and knowledge communities that exist in the blogosphere: with blogging, audiences can be unexpected, unsuspected and undesired, and the Chinese Cyberpolice is potentially part of the

Identity repertoires on the Internet: Opportunities and constraints


audience. This was the case with our example here: a wrong kind of an identity was displayed by someone, consequently implying the potential formation of a wrong kind of a knowledge community. Mobilizing a political repertoire was a move that required intervention from unsuspected members of the audience monitoring the blogosphere. Identity correction is not always as drastic as it was in this case. Sometimes identity policing is more in the details: single words in blogs become omitted, or more accurately replaced with two stars, which tells the blogger him- or herself as well as his or her audience that a mistaken kind of an identity had been assumed, and this needed to be rectied. The Chinese Cyberpolice is effective in its work, for whole blogs can be blocked within hours of posting something that is considered harmful or sensitive, and be replaced with the text Cherish your life, keep away from blog7 ). The message sent to the blogger as well as to ging ( anyone wishing to access the blog is clear, in the sense that an identity has been displayed employing certain repertoires, and perhaps even been authenticated and approved by other Internet users, possibly members of the same knowledge community. What remains unclear with this type of intervention from the Cyberpolice and the identity correction they make, is what is addressed here as life. Is it the virtual life of the blogger that is being referred to, or the material life, so to speak, and can we or should we in fact make such a distinction? What seems virtual here is in fact very real, and the imprisonment of Internet activists should be one of the clearest examples of the real nature of all of this. What is obvious is that the opportunities provided by the Internet may be unprecedented, but so are the constraints. What the Chinese cases show us is that Internet users are not only monitored from the outside, in what can be described as a late-modern panopticon (Foucault 1995 [1975]). In addition, and as a consequence, they also embark on self-policing, and the identities that are displayed become not only monitored by others but also by the users themselves. So we witness two kinds of creativity: the rst kind is brought about because of the new socio-technical tools available, and it is a very positive development that people can engage in these forms of creativity, identity work and community building. The second kind of creativity, however, is forced creativity to be able to inhabit online environments, certain moves and manoeuvres are necessary, as identities and whole communities need to be legitimized and approved of not only by immediate participants but also by invisible others.


Piia Varis, Xuan Wang and Caixia Du

4. Conclusions By way of conclusion, we might want to remind ourselves of the famous 1993 cartoon in The New Yorker which has become a kind of a symbolic item in discussions on Internet and identity. The cartoon, featuring two dogs, one sitting in front of a computer screen, and the other one on the oor next to it, included the caption On the Internet, nobody knows youre a dog. That is how we tend to think about online environments: if you are a man, you may present yourself as a woman; if you are poor, you may present yourself as rich; if you are unhappy, you may still present yourself as happy. We may have the feeling of having another reality to live in an alternative reality that we can use to complement or replace the corporeal one, if it happens to be unpleasant or insufcient for us. Indeed, sometimes no one needs to know that you are a dog. However, as we have seen in our discussion here, our online behaviours can be monitored by (often invisible) others, and these others may be interested in knowing whether you are indeed a dog or not for different economic or political purposes for example. In different online environments we need to, or at least are encouraged to, provide all sorts of demographic details about ourselves, to display our cultural interests and our networks. Scholz (2008: n.p.), in his discussion on what he labels as the Web 2.0 ideology, i.e. popular public discourse on the Internet and its recent developments, describes the current situation as follows: Today, marketers can even learn about the stories that we tell ourselves about ourselves; they are represented in the proles of our social networking sites. As part of a sea change toward the feasibility and importance of keeping private things private, American youth cares much less about their privacy than users of the Web a decade ago. Today, young people dont mind so much that they share their friends lists, conversations, and navigational habits not only with their acquaintances but also the companies who interpret much of this data. With these rms (and possibly government bodies) as daily condantes, latent possibilities for total control have opened up. The online actions Scholz refers to can of course be done for identity purposes, and the Internet provides opportunities for broadening our identity repertoires (and, to add an important point to Scholzs characterisation, not just for American young people). And all this is at present very much part of our lived realities. We now live in an era where imagination has become a social practice (Appadurai 1996); indeed, it is now central to all forms of agency, is itself a social fact, and is the key component of the new global order (Appadurai 1996: 31). The new socio-technical tools we have at our disposal for imagining identities and communities enable us to conjure up selves that satisfy our desires

Identity repertoires on the Internet: Opportunities and constraints


to be what we want to be, within the kinds of communities we wish to be part of as statements of our identities. That is, the opportunities for imagining and acting on our imaginings are like nothing we have been able to imagine before, and the scope and signicance of opportunities we have for extending our immediate corporeal realities should not be undermined. However, the new technologies of self (Foucault 1988) come with restrictions of the kind we also have not been able to imagine before. Here we have to remind ourselves of some of the features that come with the package that is presented to us in popular media ideologies (i.e. beliefs about and attitudes towards the possible and appropriate uses and functions of different media; see Gershon 2010: 3) as one of unlimited freedom: control, archiving, surveillance and censorship. The new freedoms for imagining and creating identities are undeniable, but as Lo (2009: 384) puts it: For the nation-state, the emergence of new forms of self in the society requires further social surveillance, risk management, and regulations. Creativity involves transgression, and in societies obsessed with security and stability, transgression is a risk. The cases presented here are not only anecdotal evidence but they are indicative of larger patterns. There is no need for doomsday scenarios, but the Chinese cases presented here clearly illustrate a point. Identities and communities online are not all about freedom and liberty, and this is not only about what is legal or illegal as such norms and rules for communication and identity and community building are also obeyed without explicit legal restrictions. As Internet users, we navigate and manoeuvre in different environments, and our navigation and manoeuvring can be complicated by the fact that these environments often come with no histories and sometimes with very little, emergent, contextual cues. The explicit censorship we come across in the case of China is not the core point here. It is widely known that China limits the use of the Internet, and the scope and nature of control may seem like something foreign for Internet users elsewhere. But China is not alone in this and by no means the only place where control, screening and proling happens in China, we witness overt policing of the Internet, elsewhere we see it in more covert forms. However, with Internet identities we can indeed talk about repertoires in the plural, for there is room for manoeuvring, and if one online environment does not satisfy our desires in imagining who we want to be, then there will always be another one that will allow that to happen. The two cases from China presented here are cases of nomadic Internet users, of nomadic subjects (to put Braidottis 1994 term into literal use): users who migrate from one forum to another in search of a place for identity construction that can be conrmed by other subjects, both visible and invisible. What needs to be remembered here is that these late-modern


Piia Varis, Xuan Wang and Caixia Du

nomadic subjects leave traces of themselves and are controlled by both visible and invisible others, and this also makes identity work more precarious. All this needs to be taken into account in our endeavours to explain the new identity phenomena we encounter: what we now have is new forms of subjectivity and new kinds of life projects new opportunities for identity repertoires, and new forms of constraint.

1. This paper has been written in the context of the research project Transformations of the Public Space (TRAPS) at the Department of Culture Studies, Tilburg University. 2. By way of illustration, a now leading journal in the eld, Global Media and Communication, was only launched in 2005. 3. danah boyd does not use capitals in writing her name. We adopt this preference throughout the paper. 4. This case is based on eldwork by Xuan Wang. See also Wang (2011). 5. A very recent case being the Internet censorship following the announcement of the dissident Liu Xiaobo as winner of the Nobel Peace Prize. 6. This case is based on eldwork by Caixia Du. 7. Interestingly, this Cyberpolice message borrows from the real world trafc police genre trafc signs in China would use a similar expression when asking people to be cautious and mindful of their actions.

Androutsopoulos, Jannis. 2006. Multilingualism, diaspora, and the Internet: Codes and identities on German-based diaspora websites. Journal of Sociolinguistics 10(4). 520547. Appadurai, Arjun. 1996. Modernity at large. Cultural dimensions of globalization. Minneapolis: University of Minnesota Press. Bargh, John A., Katelyn Y. A. McKenna & Grainne M. Fitzsimmons. 2002. Can you see the real me? Activation and expression of the true self on the Internet. Journal of Social Issues 58(1). 3348. Baron, Naomi S. 2008. Always on. Language in an online and mobile world. Oxford: Oxford University Press. Block, David. 2004. Globalization, transnational communication and the Internet. International Journal on Multicultural Societies 6. 1328. Blommaert, Jan. 2005. Discourse. A critical introduction. Cambridge: Cambridge University Press.

Identity repertoires on the Internet: Opportunities and constraints


Bourdieu, Pierre. 1986. The forms of capital. In John G. Richardson (ed.), Handbook of theory and research for the sociology of education, 241258. New York: Greenwood Press. boyd, danah. 2001. Sexing the Internet: Reections on the role of identication in online communities. Paper presented at Sexualities, Medias and Technologies: Theorizing Old and New Practices. University of Surrey, 2122 June. papers/ (accessed 6 October 2010). boyd, danah. 2008. Taken out of context. American teen sociality in networked publics. Berkeley, CA: University of California dissertation. boyd, danah. 2009. None of this is real: Identity and participation in Friendster. In Joe Karaganis (ed.), Structures of participation in digital culture, 132157. New York: Social Science Research Council. Braidotti, Rosi. 1994. Nomadic subjects: Embodiment and sexual difference in contemporary feminist theory. New York: Columbia University Press. Butler, Judith. 1990. Gender trouble. New York: Routledge. Campbell, John Edward. 2004. Getting it on online: Cyberspace, gay male sexuality, and embodied identity. Binghampton, NY: Harrington Park Press. Castells, Manuel. 1996. The rise of the network society. Cambridge, MA: Blackwell. China Internet Network Information Center. 2010. Statistical report on Internet development in China. The 26th survey report. index.htm (accessed 24 November 2010). Derrida, Jacques. 1996. Archive fever: A Freudian impression. Chicago: University of Chicago Press. Ellison, Nicole, Rebecca Heino & Jennifer Gibbs. 2006. Managing impressions online: Self-presentation processes in the online dating environment. Journal of ComputerMediated Communication 11(2). (accessed 6 October 2010). Facebook. 2010. Statistics. (accessed 15 November 2010). Foucault, Michel. 1988. Technologies of the self. In Luther H. Martin, Huck Gutman, & Patrick H. Hutton (eds.), Technologies of the self, 1649. Amherst: University of Massachusetts Press. Foucault, Michel. 1995 [1975]. Discipline and punish: The birth of the prison. NewYork: Vintage. Georgakopoulou, Alexandra. 2006. Postscript: Computer-mediated communication in sociolinguistics. Journal of Sociolinguistics 10(4). 548557. Gershon, Ilana. 2010. The breakup 2.0. Disconnecting over new media. Ithaca: Cornell University Press. Goffman, Erving. 1959. The presentation of self in everyday life. Edinburgh: University of Edinburgh Social Sciences Research Centre. Goffman, Erving. 1981. Forms of talk. Philadelphia: University of Pennsylvania Press.


Piia Varis, Xuan Wang and Caixia Du

Gumperz, John J. & Jenny Cook-Gumperz. 1997 [1982]. Introduction: Language and the communication of social identity. In John J. Gumperz (ed.), Language and social identity, 121. Cambridge: Cambridge University Press. Hall, Stuart & Paul du Gay (eds.). 1996. Questions of cultural identity. London: Sage. Herring, Susan C. & John C. Paolillo. 2006. Gender and genre variation in weblogs. Journal of Sociolinguistics 10(4). 439459. Herring, Susan C., Lois Anna Scheidt, Elijah Wright & Sabrina Bonus. 2005. Weblogs as a bridging genre. Information Technology & People 18(2). 142171. Information Ofce of the State Council of the Peoples Republic of China. 2010. The Internet in China. 1622956.htm (accessed 24 November 2010). Jenkins, Henry. 2006. Convergence culture. Where old and new media collide. NewYork: New York University Press. Jespers, Raf. 2010. Big Brother in Europe. Berchem: Epo. Karaganis, Joe (ed.). 2007. Structures of participation in digital culture. NewYork: Social Science Research Council. Killoran, John B. 2003. The gnome in the front yard and other public gurations: Genres of self-presentation on personal home pages. Biography 26(1). 6683. Lanier, Jaron. 2010. You are not a gadget. London: Allen Lane. Lepp nen, Sirpa. 2007. Cybergirls in trouble: Fan ction as a discursive space for interroa gating gender and sexuality. In Carmen-Rosa Caldas-Coulthard & Rick Iedema (eds.), Identity trouble: Discursive constructions, 156179. London: Palgrave-Macmillan. Lepp nen, Sirpa, Anne Pitk nen-Huhta, Arja Piirainen-Marsh, Tarja Nikula & Saija Peua a ronen. 2009. Young peoples translocal new media uses: A multiperspective analysis of language choice and heteroglossia. Journal of Computer-Mediated Communication 14. 10801107. L vy, Pierre. 1997. Collective intelligence: Mankinds emerging world in cyberspace. e Cambridge, Mass.: Perseus Books. Lo, Kwai-Cheung. 2009. The web marriage game, the gendered self, and Chinese modernity. Cultural Studies 23(3). 381402. MacKinnon, Rebecca. 2008. Flatter world and thicker walls? Blogs, censorship and civic discourse in China. Public Choice134. 3146. MacKinnon, Rebecca. 2009. Chinas censorship 2.0: How companies censor bloggers. First Monday 14(2). fm/article/view/2378/2089 (accessed 23 November 2010). McNeill, Laurie. 2003. Teaching an old genre new tricks: The diary on the Internet. Biography 26(1). 2447. Maffesoli, Michel. 1996. The time of the tribes. London: Sage. Miller, Carolyn R. & Dawn Shepherd. 2004. Blogging as social action: A genre analysis of the weblog. In Laura Gurak, Smiljana Antonijevic, Laurie Johnson, Clancy Ratliff & Jessica Reyman (eds.), Into the blogosphere. Rhetoric, community, and the culture of weblogs. as social action a genre analysis of the weblog.html (accessed 21 August 2010).

Identity repertoires on the Internet: Opportunities and constraints


Nederveen Pieterse, Jan. 1995. Globalization as hybridization. In Mike Featherstone, Scott Lash & Roland Robertson (eds.), Global modernities, 4568. London: Sage. OHara, Kieron, Mischa M. Tufeld & Nigel Shadbolt. 2009. Lifelogging: Privacy and empowerment with memories for life. Identity in the Information Society 1(1). 155 172. Qiu, Jack Linchuan. 1999. Virtual censorship in China: Keeping the gate between cyberspaces. International Journal of Communications Law and Policy 4. 125. Rak, Julie. 2005. The digital queer: Weblogs and Internet identity. Biography 28(1). 166182. Richardson, Kathleen & Sue Hessey. 2009. Archiving the self? Facebook as biography of social and relational memory. Journal of Information, Communication & Ethics in Society 7(1). 2538. Scholz, Trebor. 2008. Market ideology and the myths of Web 2.0. First Monday 13(3). 1945 (accessed 22 November 2010). Sund n, Jenny. 2003. Material virtualities. New York: Lang. e Thurlow, Crispin, Laura Lengel & Alice Tomic. 2004. Computer mediated communication. Social interaction and the Internet. London: Sage. Tsui, Lokman. 2003. The panopticon as the antithesis of a space of freedom: Control and regulation of the internet in China. China Information XVII(2). 6582. Turkle, Sherry. 1995. Life on the screen: Identity in the age of the Internet. New York: Simon & Schuster. Walther, Joseph B., Brandon Van Der Heide, Sang-Yeon Kim, David Westerman & Stephanie Tom Tong. 2008. The role of Friends appearance and behavior on evaluations of individuals on Facebook: Are we known by the company we keep? Human Communication Research 34. 2849. Wang, Shaojung Sharon & Junhao Hong. 2010. Discourse behind the forbidden realm: Internet surveillance and its implications on Chinas blogosphere. Telematics and Informatics 27. 6778. Wang, Xuan. 2011. I am not a qualied dialect rapper: Genre innovation and authenticity. Sociolinguistic Studies. In press. Weisbuch, Max, Zorana Ivcevic & Nalini Ambady. 2009. On being liked on the web and in the real world: Consistency in rst impressions across personal webpages and spontaneous behavior. Journal of Experimental Social Psychology 45. 573576. Yang, Guobin. 2003a. The co-evolution of the Internet and civil society in China. Asian Survey 43(3). 405422. Yang, Guobin. 2003b. The Internet and the rise of a transnational Chinese cultural sphere. Media, Culture & Society 25(4). 469490. Zhang, Weiyu & Rong Wang. 2010. Interest-oriented versus relationship-oriented social network sites in China. First Monday 15(8). bin/ojs/index.php/fm/article/viewArticle/2836/2582 (accessed 23 November 2010).


Piia Varis, Xuan Wang and Caixia Du

Zhao, Shanyang, Sherri Grasmuck & Jason Martin. 2008. Identity construction on Facebook: Digital empowerment in anchored relationships. Computers in Human Behavior 24. 18161836. Zittrain, Jonathan & Benjamin Edelman. 2003. Internet ltering in China. IEEE Internet Computing 7(2). 7077. Piia Varis is a post-doctoral researcher within the research group Transformations of the Public Space (TRAPS) at the Department of Culture Studies, Tilburg University, the Netherlands. Caixia Du is a PhD student in TRAPS. Xuan Wang is a PhD student in the Department of Culture Studies, Tilburg University. Email:

Tourism discourse: Languages and banal globalization


Abstract Described as the one of the greatest population movements of all time, tourism is rmly established as one of the worlds largest international trades. And it is not just people who are on tour; language too is on the move. In this paper we examine some of the ways that our research has shown language commonly being taken up in tourisms search for exoticity and authenticity. Specically, we present a series of different touristic genres (broadcast media, guidebook glossaries, guided tours) where local languages are stylized, recontextualized and commodied in the service of tourist identities and of tourisms cosmopolitan mythology. It is in this way that the globalizing habitus (Jaworski and Thurlow 2010) of tourism privileges or elevates those who choose to travel, containing linguistic/cultural difference under a guise of celebration and respect. These playful, seemingly innocuous textualizations of language/s are also exemplary enactments of banal globalization (Thurlow and Jaworski 2010), the everyday, micro-level ways in which the social meanings and material effects of globalization are realized. Keywords: language, local languages, tourism, commodication, cosmopolitanism, difference, banal globalization, globalizing habitus

Touristic culture is more than the physical travel, it is the preparation of people to see other places as objects of tourism . . . the touristic gaze and imaginary shape and mediate our knowledge of and desires about the rest of the planet. (Franklin and Crang 2001: 10) Symbolic capital, a transformed and thereby disguised form of physical economic capital, produces its proper effect inasmuch, and only inasmuch, as it conceals the fact that it originates in material forms of capital which are also, in the last analysis, the source of its effects. (Bourdieu 1977: 183)


Crispin Thurlow and Adam Jaworski

Figure 1. The promise of contact

1. Why tourism matters (for language scholars) Academics working in the interdisciplinary eld of critical tourism studies often have to justify their scholarly interest to those unable to see beyond their own personal experience of tourism as a frivolous, recreational activity. And yet, as a truly global industry perhaps even the worlds largest, single international trade there are few people whose lives remain unaffected by tourism, be it people privileged enough to tour or people who are toured. For anthropologist Ed Bruner (2005: 10), tourism is simply one of the greatest population movements of all time. It is precisely because of its scale and inuence that anthropologists, sociologists, geographers and others have looked to examine the

Tourism discourse: Languages and banal globalization


social and cultural practices by which tourism is organized and experienced (for an overview, see Thurlow and Jaworski 2010). As a whole, this growing body of research clearly demonstrates tourisms powerful role in reshaping cultural practices, establishing ideologies of difference, and perpetuating unequal relations of power (Favero 2007). Tourism seldom merely represents cultural difference or reects existing socioeconomic relations within and between countries; instead, it is instrumental in producing the very culture that tourists set out to know, and in (re)organizing relations between groups, communities and entire nations (Lash and Urry 1994; Bauman 1998; Kirshenblatt-Gimblett 1998). As Adrian Franklin and Mike Crang note (quote above), tourism serves as a very inuential, privileged lens through which many people make sense of not only a particular destination or culture but of the world at large. In seeking to contribute a uniquely sociolinguistic or discourse analytic perspective to the eld of tourism studies, our own work focuses on the role of language and communication in tourism. More importantly, however, we are keen to learn what tourism tells us about language and languages nowadays. We therefore share with other colleagues a broader interest in understanding the life of language/s under globalization or, more correctly, postindustrial capitalism (see, for example, Heller 2003; Blommaert 2005; papers in Coupland 2010). Just as tourism has proved to be an obvious research domain/topic for anthropologists and sociologists, it is a key site for the study of human communicative processes most obviously with regards intercultural contact and exchange, but also in terms of the circulation of linguistic material (e.g. genres, discourses and styles). For all its economic weight and political consequence, tourism is an intensely social and communicative business. In many respects it is the ideal industry for global capitalism because it is highly exible, constantly reexive and deeply semiotic As a service industry, tourism is fundamentally and, at times, solely semiotic in nature because, like advertising and marketing, a key part of what is actually produced and consumed is the semiotic context of the service. Not only does tourism involve face-to-face (or more mediated) forms of visitorhost interaction, like in many other types of service interactions, but the ultimate goods purchased by tourists during their travels are images, lifestyles, memories, tastes, encounters, and so on. Much of the signicance the cultural capital of tourism lies also in the tourist haze created as tourists return home with their own travel stories about well-trodden destinations, the souvenirs they bring for the folks back home, and indeed the photos of themselves in exotic locations. More than this, however, the tourist imagination and tourist practices are always heavily (in)formed by and pregured in the heavily mediatized representations of television holiday programmes, travel brochures, newspaper


Crispin Thurlow and Adam Jaworski

travelogues, postcards, guidebooks and so on (see Thurlow and Jaworski 2010). In this way, as Mike Crang (1999: 361) explains, a structure of expectation is created, where the pictures circulating around sights are more important than the sites themselves . . . The signs that mark out what is to be looked at become as, or more, important than the sites themselves. With particular reference to photography, John Urry (2002) calls this search for the already seen the hermeneutic cycle:
What is sought for in a holiday is a set of photography images, which have already been seen in tour company brochures or on TV programmes. While the tourist is away, this then moves on to a tracking down and capturing of these images for oneself. And it ends up with travellers demonstrating that they really have been there by showing their version of the images that they had seen before they set off. (p. 129)

What we mean to show in this paper is that language, languages and, in particular, local languages often feature no less heavily than photos/images when it comes to tourists search for authenticity and difference. Along with material goods such as photographs and souvenirs, snippets of local languages too are (re-)packaged and promoted as useful props (or trinkets) in the enactment of tourisms performances of exoticity. The tourist linguascape (Jaworski et al. 2003) thus serves as an extension of the tourist gaze (Urry 2002, after Foucault 1976), the socially organized, systematized and disciplining ways in which tourism is structured and learned. In the case of language, tourists are also drawn into a regime of truth about the nature of language and linguaculture (Agar 1994), as well as the relative value of local languages in the global linguistic marketplace.

2. Tourism as a language market Until not so long ago, tourism research tended to focus almost exclusively on the role of visuality, spurred in large part by John Urrys commonly misunderstood notion of the tourist gaze just mentioned. To be sure, the visual representation and production of sights/sites continues to be a dominant mode for visitors and hosts. It is, however, by no means the only sense through which travel is experienced (see Franklin and Crang 2001).To some extent, any lack of scholarly attention to language is explained by the nature of the tourist experience itself; in Michael Cronins (2000:82) words, sightseeing is the world with the sound turned off . As David Dunn (2005, 2006) further explains, tourists usually end up gazing simply because they cannot understand the languages spoken by the

Tourism discourse: Languages and banal globalization


objects of their gaze. And yet, our own research suggests that language (along with other communicative modes) is everywhere in tourism; in fact, language and languages sit at the very heart of the tourist experience, its representation and its realization, its enculturation and its enactment. The expansion of tourism as a dominant cultural industry is one of the major areas of economic activity under globalization which has highlighted the signicance of language commodication in the study of shifting identities, interpersonal relations and group structures. Of course, the political economy of language has long been recognized (e.g. Bourdieu 1991; Irvine 1989), and so have the general processes of commodication and appropriation of language in the new economic order of exible accumulation and of timespace compression. In her work on bilingual areas of francophone Canada, for example, Monica Heller (e.g. 2003) demonstrates how the collapse of traditional industries (cod-shing, mining, logging, etc.) in the second half of the twentieth century, and their substitution with new information and service-based industries (most notably, call centres and tourism), have led to the commodication of both language (understood as a measurable skill) and identity (especially in relation to other forms of cultural practice such as dance and music in tourism). In these domains of economic practice based on contact between different linguistic markets through advances in communication technology (call centres) or mobility (tourism), linguistic and other symbolic resources become highly marketable commodities. Due to the new conditions for its commodication, language, together with other forms of cultural practice, is arguably more easily detached from identity and used as a strategic styling resource (cf. Bell 2009; Cameron 2000; Coupland 2007). One place where this happens often is in tourism discourse where local languages may be marketed and traded as metonyms of place (Urry 2007) and/or markers of cultural authenticity to be consumed by tourists. It is this that we mean to show now by turning to some data which illustrates just some of the ways that language/s is/are commonly and, we suggest, problematically taken up in tourism discourse. To this end, we want to look at three examples: travel shows, guidebook glossaries and what we like to call the greeting game. We end with a more critical/social theoretical reection on the politics of representation and the implications of tourisms use of local languages. The issue for us is one of language ideology as it dovetails with the ideologies (or mythologies) of both tourism and globalization; in each case, and as Bourdieu (quote above) reminds us, the symbolic market however insubstantial and playful it appears is grounded in the material inequalities of the global political economy.


Crispin Thurlow and Adam Jaworski

3. Case study #1: The language play of travel shows The broadcast media offer some of the quintessential ways in which tourism is pre-gured as the preparation of people to see other places as objects (quote by Franklin & Crang above). In other words, people learn what the value of tourism is, what it means to be a tourist and the rules of engagement with local places, local people and, indeed, local languages. Take a look at the following extract from the start of one episode of the popular (in the USA, at least) radio show Travel with Rick Steves. (We will show shortly that this singular moment of tourism discourse is by no means limited to Rick Steves, to radio or to the USA.) This episode was rst aired in May 2008 and was repeated in January 2010; it is Steves Rome City Guide. After a short preamble (lines 1 to 5, edited from the original), he takes a moment to introduce two local guides/friends, starting in line 6.1
Extract 1: From Travel with Rick Steves (10 May 2008) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 RS: as you peel through its fascinating and jumbled layers youll nd Romes buildings cats all that laundry crazy trafc and two and a half million Romans endlessly entertaining (.) and of course the thing about Rome is it has so much history(.) Not only does it have a lot of history but its a vibrant opportunity to connect with todays Italy (.) today I have joining me two Roman guides friends of mine who have helped me with my tours and my guidebook research and today are joining us(.) Susanna Perucchini is here in our studio and Francesca Caruso joins us by telephone from Rome (.) do I say benvenuti? benvenute si benvenuti (laughs) Francesca come va? (laughs) molto bene e tu? ciao bella (laughs) (laughs) ciao (laughs) (laughs) I gotta say ciao bella si yeah you said it well yeah that was perfect ciao bella because I wouldnt want to say ciao bello (laughs) no you wouldnt Ive learned (.) no


Tourism discourse: Languages and banal globalization 26 27 FC: RS: denitely not (laughs) OK thank you (.) well its great to have you both here


Steves performance of Italian (and Italianicity more generally, cf. Barthes 1977 [1964]) conrms his relatively limited grasp of the language; for example, his incorrect gendered inection of benvenuti (line 9), his inability to recognize the correction offered (lines 10 and 11), the arguably pragmatic inappropriateness of ciao bella, and, by the same token, his slightly awkward (if not also heteronormative) game with bello. This is not to say that he has no Italian. Nor is it to deny credit where credit is due: at least, some might argue, hes having a go. This is hardly, however, a serious or committed attempt to take up or to move into Italian. Indeed, the framing of this local language as a largely playful resource for use by Steves is keyed as endlessly entertaining (line 3) in his preamble and by the laughter throughout (lines 12, 14, 15, 16, 17, 18, 24 and 26). Its a low-stakes game, all part of a pleasurable encounter with otherness. Steves rests comfortably in the knowledge that after his dabble with Italian he may return safely to English after all, both his Italian guests are uent English speakers. Note also the promise of contact (see opportunity to connect, line 5) which sits at the heart of so much tourism discourse. In a world (or interactional frame) marked as play, grammatical accuracy is clearly less important if at all and the consequences of pragmatic failure the need for serious conversational repair mitigated by largely one-sided relationships of power. At least this is how these encounters are presented to listeners or, as we will now show, viewers. Across a lot of our research, we have found much the same playful framing of local languages in different domains of tourism discourse (see Jaworski et al. 2003; Jaworski and Thurlow 2009; Thurlow and Jaworski 2010). For now, however, we present the following short extracts from once hugely popular British television holidays shows; these extracts are just four instances where a presenter is seen to use a local language. To be clear, the depiction of local languages does not happen often in shows like these which are for all intents and purposes infotainment targeted at predominantly English-speaking audiences. We start with Extract 2 as it happens, another performance of Italianicity to exemplify the deployment of a local language as an ideal marker of authenticity and/or as an exotic backdrop.
Extract 2: Vera, the chief pasta maker and TV presenter Mary Nightingale 1 2 3 MN: (voiceover) I found all the hotels very comfortable and whats nice is theyre all so individual and they feel so (.) Italian (1.0) this farmhouse has been in the family for generations (.) Vera is the boss (.) and the

4 5 6 7 8 9 10

Crispin Thurlow and Adam Jaworski chief pasta maker (cut to Veras kitchen where she is making pasta) queste la pasta queste this is the pasta these sono (.) limpasto (.) mangiala cruda mangiala cruda are (.) the mixture (.) eat it raw eat it raw (picks up a single strand of raw pasta, turns away from Vera, raises the piece of pasta to the camera) theres a piece of Veras tagliatelle (.) isnt that absolutely beautiful (.) its perfect (realising that MN is no longer listening to her, turns and looks bafed at MNs interest in the piece of pasta)




In this extract, Italian is spoken only by the local expert (Vera) who is cast for her stereotypical expertise as the chief pasta maker whereby pasta once again metonymically condenses all things Italian (it all feels so Italian, line 2). In turning to the camera midway through Veras instructions, the presenter conrms the primary interactional alignment (with the audience back home) and proves that Veras talk is really intended as little more than another scene-setting resource. It is not what Vera is saying that is important but rather that she is simply saying something in Italian while making pasta. The person, the food and the language all serve as metonymic markers of place and of difference. The use of local languages as exotic backdrop or soundscape is one of the most common ways in which the tourist linguascape is produced (again, see Jaworski et al. 2003; Jaworski and Thurlow 2009; Thurlow and Jaworski 2010). In Extract 3, we nd the relegation of a local language extended through its use as a ludic resource, as an object of play. In this case, the tourist-presenter (CD) is interacting with larger than life store-keeper Massimo.
Extract 3: Ciche ciche cento Massimo and TV presenter Craig Doyle 1 2 3 4 5 CD: M: CD: M: (voiceover) the Italians are passionate about food and no one more so than Massimo one of Sienas larger than life grocers grande grande= large large =grande large questo e pesto pesto genovese (.) guarda (picks up a packet of pasta) ` this is pesto pesto from Genoa (.) look [ si si pesto fresh tomatoes si yes yes pesto so I need these as well yeah (picks up a tray of blackberries) avanti

6 7 9



Tourism discourse: Languages and banal globalization


10 11 12 13 14 15 16 17 18

CD: M: CD:

M: CD: M: CD: M:

19 20


lets go on oh yeah tomatoes dried si chiamano ciliegini (shows CD some dried tomatoes) theyre called cherry tomatoes (points over Ms shoulder) look at that out there (pops a tomato in his mouth with a look of mock guilt-cum-innocence) [ (looks away briey) grazie grazie (.) (CD pays) thank you thank you a posto cosi anything else? ciche ciche cento no (.) ciche ciche ciu (.) ciao a presto (shakes hands with CD) bye see you soon [ ciao grazie bye thank you (walks out of shop; voiceover) thank you Massimo

It is worth noting that this last extract (like Extract 4 below) is also structured by a musical score in this case, a lively swing-jazz tune, ideal for the kind of comedic sketch being staged. Rather like Rick Steves above, the presenter here clearly has only a basic grasp of Italian; while Massimo presses on with explaining his produce, the presenter throws about one or two familiar snatches of phrasebook Italian (grande line 4, si line 6, grazie lines 15 and 19, and ciao line 19) and nishes with the largely nonsensical ciche ciche cento (line 19) a phrase which possibly, for a British audience at least, echoes the Cinquecento from Italian car manufacturer Fiat (which was also being heavily advertised at the time this show was rst broadcast). For the presenter, it is all just a game for the benet of the folks back home; both Massimo and Italian are just props in his skit. To match his somewhat condescending interaction with the local person, the presenter also condescends into the local language in precisely the way that Pierre Bourdieu (1991: 69) explains it:
One can see in passing that strategies for the subversion of objective hierarchies in the sphere of language, as in the sphere of culture, are also likely to be strategies of condescension, reserved for those who are sufciently condent of their position in the objective hierarchies to be able to deny them without appearing to be ignorant or incapable of satisfying their demands.

None of this is to deny the fun of trying out snippets of the local language, although this enjoyment should not conceal the privilege of not really having to


Crispin Thurlow and Adam Jaworski

depend on the language. The pleasurable value of local languages for tourists (as presenters or otherwise) is commented upon explicitly in Extract 4, where the presenter (LR) excitedly tries out her one line of Spanish in a service encounter with a barman. Pragmatic matters such as the arguably inappropriate use of the familiar t-form (puedes) are again less important than the pleasure of learning and having a go.
Extract 4: Been learning that all day! (with TV presenter Lisa Riley) 1 2 3 4 5 6 LR: (voiceover) its well worth taking a wander up the side streets off the square (camera on LR and friend) where you can nd traditional tapas bars just like this one (points to bar) ((shall we take a look)) (LR walks to bar; to barman) hola me puedes dar la carta por favor? hello can you give me the menu please? Barman: (hands over menu) ((unclear)) LR: gracias (to camera, cheerfully) been learning that all day (giggles) thank you

Throughout these types of mediatized instances of local language, what is shown is only the most supercial form of engagement with local people (almost always in service ecounters see below) as an arch performance of contact. In one of our favourite moments and the last one we will present here Extract 5 points again to the playful but very limited use of the local language (the one-word greeting bula in line 3) but is also a clear reminder that the use of local languages is invariably a theatrical prop for staging hospitality and reassuringly friendly local people. In this particular case, all we want to show is how the highly scripted/staged nature of these types of televised encounters with hosts is accidently revealed: note the easily missed use of the TV presenters name by the vendor in Line 5.
Extract 5: Bula, bula, John! (with TV presenter John Savident) 1 2 3 4 5 6 (JS apparently wandering through a market place) away from the hotel the town of Nandi [sic.] is just ten minutes away (.) Fiji is such a friendly place and youre always greeted with a big smile (cut to a woman smiling) and a call of ((BULA)) the local greeting (to a street vendor) bula Vendor: bula bula John (JS continues walking past her stall, laughs to her) how are you bula bula la la la JS:

In her study of the use of mock Spanish by Anglo-Americans in the United States, for example, Jane Hill (2001[1998]) demonstrates how apparently jocular incorporations and ungrammatical approximations of other languages are

Tourism discourse: Languages and banal globalization


employed by non-native speakers as an important identity resource. Hill argues that playful, ippant snatches of, in her case, Spanish-language materials, serve to elevate the identities (or Whiteness) of Anglo-Americans. To our mind, much the same argument may be made for the use of phrasebook (see next section) expressions by presenter-tourists and the general linguascaping of tourist destinations; in this case, however, it is the elevation and constitution ofAmericanness by Rick Steves and Britishness by the TV presenters which is at stake. In fact, we argue that the main aim in both cases (radio and TV) is to create for the listeners/viewers a pleasing sense of belonging to an imagined community of tourists and cosmopolitan global citizens. This is achieved largely through the presenters specic exploitation of the sociolinguistic resource known as crossing (see Rampton 1995) which is the use of a language (or variety) of a group of which the speaker cannot legitimately claim membership. It is through their playful, transient crossings into local languages that these presenters position themselves as cosmopolitans not in the sense of their being culturally engaged with or embracing local people (cf. Hannerz 1996), but rather with respect to their appeals to the elite cachet of global citizenship. These are people freely traversing national boundaries but staying rmly rooted in their mutual identication as (American and British) nationals. Ultimately, it is in this way also that, as powerful ideological mediators, the shows and their style-setting presenters promote a regime of touristic and intercultural truth: this is what it means to be a tourist, this is where the value of local languages lies.

4. Case study #2: Language learning in guidebook glossaries We turn now from our broadcast data to a different set of data: the glossaries of travel guidebooks as a quintessential genre in tourism and of what we like to call a discourse on the move (Thurlow and Jaworski 2010). These highly condensed crash courses tucked away at the end of most guidebooks are also in the business of stylizing tourists and of commodifying local languages; they often go a step further, however, by purporting to offer workable scripts for intercultural contact between tourists and hosts. In this case, as with the advertisement in Figure 2, the exaggerated promise of contact produces a number of questionable ideas about the nature of language, of language learning and of human interaction in general. The representation of local languages in guidebooks indeed, the decision itself to include a section on the local language/s is structured by a desire to produce authenticating markers of exoticity and cultural difference. In this sense, the glossaries are no different from the kinds of playful linguascaping


Crispin Thurlow and Adam Jaworski

Figure 2. The promise of contact #2

we showed above. Not surprisingly, therefore, many guidebooks begin with a metalinguistic commentary on the perceived oddity of the local language; in doing so, they rehearse a number of common language myths about, for example, how hard or different the language is, or how beautiful or ugly it is (cf. Bauer and Trudgil 1998). From an expert point of view, of course, these are always only ever matters of subjective or relative taste and of ideology. This explains why Polish should be fearsome (Extract 6a) and Australian Strine is laconic, poetic and prolic (Extract 6b). It is a matter of ideology (not linguistics) that the inclusion of a Strine glossary thereby also renders it equivalent to Polish or any other major (national) language typically covered in guidebooks.
Extracts 6a to 6f a. . . . Polish is pretty fearsome for people outside the Slavonic circle . . . b. The lingo has a laconic, poetic originality and a prolic profanity. c. Chinese languages are rich in homonyms and much of their superstitious beliefs, poetry and humour is based on this wealth. The Cantonese word for silk sounds the same as the words for lion, private, poem, corpse and teacher. d. Fortunately, staff at most tourist ofces and hotels are uent English speakers . . . e. . . . Mexicans are delighted with foreigners who try to speak the language . . . f. Mistakes made by visitors are kindly tolerated, and even your most bumbling attempts [at Finnish] will be warmly appreciated.

Tourism discourse: Languages and banal globalization


In Extract 6a, popular stereotypes about Chinese/Cantonese are the focus of guidebook commentary. While it is true that Cantonese is rich in homonyms, that language play and verbal taboo are frequent, it is also true that beyond the simple citation forms, all the words listed in the second sentence of the extract are easily disambiguated in spoken Cantonese due to their bi-morphemic (two character) structure, presence of classiers, and other morpho-syntactic features. What this description also overlooks or downplays is that the very same principles are true of most languages; for example, an easy equivalent might be English homonyms like pore, poor, pour or rain, reign, rein or raise, rays, raze, which in Spanish, say, would be rendered differentially as poro, pobre, verter, lluvia, reinado, reinda, and aumento, rayos, arrasar, respectively. (And this is not to mention the relative challenge of the idiosyncrasies of English spelling). Of course, the real value in pointing to grammatical and phonetic characteristics such as these does not lie in their linguistic validity or signicance, but rather in their perceived oddity and, therefore, implied exoticity. Once again, in the context of tourism, language is given value (and attention) because of its symbolic rather than representational or interpersonal function. This is why local languages are seldom presented as any real obstacle or necessity. For the most part, guidebooks assume English as the default language of exchange, often reassuring readers that locals can manage English (e.g. Extract 6d) or even that locals are delighted or warmly appreciative of any attempts by the tourist to speak the local language (Extracts 6e and 6f). Ultimately, of course, interactions between hosts and tourists are structured by the realities of economic-political exchange whereby the burden of communication (Lippi-Green 1997) always falls squarely with the host. This relationship of inequality is articulated throughout the design and content of guidebook glossaries. As part of their wider Small Talk series of guides, the Lonely Planet promises to help tourists Chat, eat, shop and celebrate your way through the wonders of Northern Europe. Meanwhile, DK Eyewitness 15-Minute Spanish claims to teach tourists how to Order a meal. Book a room. Buy a ticket. Ask directions. Make conversation. The highly optimistic promise of a more conversational exchange (or chat) is intriguing and begs a closer look at the kind of conversational material on offer. What is striking is not only the limited prescription of communicative topics but also assumptions made about the nature of the relationship between visitors and local people. The vocabularies on offer are clearly restricted to the functional requirements of service transactions (that is, how much is . . . ?, where can I nd . . . ? and so on); alternatively, they encourage a level of relational engagement that seldom ventures beyond the supercial


Crispin Thurlow and Adam Jaworski

courtesies of greeting rituals (for example, hello, good morning, my name is . . . , I am from . . . ). Almost never is there a vocabulary made available which might otherwise help facilitate a more substantial, extended exchange or conversation even at the most basic level, say, of I like my job because . . . , This is the rst time Ive . . . , I believe that . . . and so on. There is certainly little in these glossaries that might move visitors towards the kind of intercultural, cross-lingual encounter with the Other which sits at the heart of tourism mythology and which the guidebooks themselves consistently imply. Instead, glossaries leave tourists stranded in a permanent state of greeting, introduction and purchase perhaps appropriate to the liminal, eeting nature of most hosttourist encounters. On close inspection, the glossaries suggest to us that apart from the most phatic exchanges (greetings, leave-takings, expressions of thanks, and so on) and the tourist-centred needs of getting to a specic location, getting a good nights sleep (possibly in a room with a view), nding a meal and a toilet, getting a good bargain in a shop or at a market, the most likely situation in which the tourist will want to speak to a host is in an emergency. To some extent, this goes some way towards contradicting the mythology of travel as being always safe and pleasurable. The preponderance of words and phrases related to accidents, illness and all sorts of other mishaps could provide a useful script for any travel insurance company advert. The following is a relatively unordered list of the English language phrases to be used in case of an emergency.
Help! Watch out! Thief! Fire! Stop! Call a doctor! Call the police! Call an ambulance! Call the re department! Where is the nearest hospital? I want to contact my embassy Could I use the telephone? Could you help me, please? Im ill Im sick Im injured I do not feel well I feel ill I have a headache I have a stomach ache I need to rest I have a fever Im allergic to penicillin Im allergic to antibiotics The child is/the children are sick We need a doctor I need a prescription for cold cough cut u hayfever headache pills hospital nausea sore throat

Tourism discourse: Languages and banal globalization dentist doctor thermometer drug store medicine pills accident ambulance emergency police policeman foreign affairs police pickpocket rapist Ive been robbed Ive/weve been mugged They stole my . . . Im lost


For the most part, the language instruction in glossaries centres on the transactional demands of service encounters rather than the interactional demands of a conversational relationship. And yet, a close examination of the language information provided in tourist guidebooks shows that there is rarely sufcient information to conduct anything but the most rudimentary of conversations. Although tourists might be encouraged to believe they are becoming global communicators and acquiring a global linguistic repertoire of tourism by using the language sections, the focus on the practicalities of travel and transactional language contrasts with the common myth of travel broadening the mind. Take a look at our hypothetical compilation of a typical range of conversational phrases offered in the guidebooks we sampled:
Welcome! Hello Glad to meet you. How are you? Very well, thanks. Whats your name? My name is . . . Where are you from? Im from . . . Im a tourist/student Im from Europe. How old are you? Im 25 Are you married? How do you say . . . No (not so) Yes I want . . . No. I dont want it. Do you like . . . ? I like it very much I dont like . . . May I? It doesnt matter. Can you please help me take a photo? Is it ok to take a photo? Goodbye.

Any conversation based on this vocabulary would unavoidably be something of a one-way street; this is clearly not a vocabulary of exchange but merely of encounter whereby the local person (imagined in bold) remains, for the most part, unspoken and unknown. Certainly, the DK Eyewitness promise made in Figure 2 of becoming somehow indigenous seems highly improbable if this is all the small talk one can muster. Guidebook glossaries are prime examples of what we have elsewhere characterised as codied, xed regimes of translated truth which . . . promote the literal and denotative, the formulaic and reductive, at the expense of the subtle, the complex, the messy, the lived (Thurlow 2004: 83). In fact, the very raison d tre of guidebooks is, somewhat ironically in the e case of glossaries, to minimize or at least mitigate contact with local people. It is in much the same way that Daniel Boorstin (1964: 91) commented some time ago on the effect travel agencies had in insulating the tourist from the


Crispin Thurlow and Adam Jaworski

travel world (cf. Bhattacharyya 1997; Jack and Phipps 2003). And herein lies the central contradiction of guidebooks. It is clear from the inclusion of language glossaries in the guidebooks that publishers feel that they need to cater for the eventuality of hosttourist interaction. On the one hand, therefore, tourists carry glossaries to aid with hosttourist communication, but on the other, the need for tourists to interact with hosts is reduced if they use a guidebook. Glossaries function primarily to full the ludic and identicational needs of tourists, which we talked about in more detail above, and in this regard they are more than adequate. To be fair, tourists who want to go further can always turn to a phrase book or a language course. As a whole, however, guidebooks usually play with the backstage frisson, the mythologized desire of tourists to seek the otherwise illusive authenticity of a real connection or a genuine encounter with the true Other. In this way, language glossaries promote themselves as resources for cracking the code of the local and for crossing into alterity. There is inevitably a satisfying and enjoyable sense of mastery in both these processes; it is an almost narcissistic which is not to say necessarily inconsiderate or malecent delight. As generic practices in themselves, the word/phrase listings included towards the end of virtually all guidebooks are fairly unique, with nothing of the attempted scope of dictionaries or language coursebooks, little of the relative detail in phrasebooks, they typically share the supercial, incidental quality of glossaries. In the rote learning tradition of the audio-lingual language learning methods and with the pretence of the real-life notional-functional method (see Pennycook 1989, for a critical review), this is language instruction that stops well short of communicative competence and that seldom goes much further than a foreign-language translation of Do you speak English? (cf. Phipps 2007). Language is, in the process, abstracted and rendered simultaneously representative of and autonomous from its cultural context. As such, it is not only the local language that is reduced, packaged and glossed but also, of course, the local culture more generally. These are, after all, the quintessential texts of the quintessential culture industry of otherness (Favero 2007). Ultimately, the tourist linguascape presented in guidebook glossaries is one of language/s almost totally disembedded from the cultural context of any plausible, extended hosttourist relationship even though the expectation (or promise) is that these linguacultural snippets might eventually be deployed in the service of intercultural exchange. Even then, however, these presumed relationships (or interactions) are clearly predicated on (and work to reinscribe) an asymmetrical model of communication and thus the the host-tourist relations of power by which tourism is typically organized.

Tourism discourse: Languages and banal globalization


5. Case study #3: The greeting game in guided tours We turn now to our nal case example of a touristic genre where local languages are stylized, recontextualized and commodied. Perhaps the quintessential marker of the hospitality underpinning the touristhost contract, greetings are everywhere in tourism, whether it is the cover of South African Airways inight magazine Sawubona, postcard greetings from the eponymous Aloha State or the bore da from Wales, a personalized online greeting from Californias Governor and First Lady, or the performance of the Maori hongi in New Zealands ofcial branding of itself (see Figures 3 through 7). In each case, greetings such as these are recontextualized and commodied in ways that violate their normal felicity conditions (Duranti 1997). The hongi, for example, has become one of the key resources for packaging Maori heritage across New Zealands tourist landscape. In Tamaki Maori Village in Rotorua, tourists can have a picture taken in a makeshift photographic studio of doing a hongi with an actor and images of Maori people performing a hongi can also be bought on postcards and posters (see Jaworski 2009; Jaworski and Thurlow 2010). In reality, it is virtually impossible for a tourist to perform a hongi with a Maori person other than in the context of a paid performance. As with the kinds of scripted exchanges in guidebook glossaries, any attempt by a tourist to initiate a hongi would be pragmatically fraught. Once again, therefore, a linguacultural snippet is taken up as a largely metacultural token or curiosity for authenticating a touristic spectacle of difference. Our nal example is in fact an instance of a commodied greeting exchange (involving a hongi) between a Maori guide/coach driver and an American tourist en route for a nights entertainment at Tamaki Maori Village at Rotorua, New Zealand. The tourists are collected by several coaches from hotels in the area. Once all the tourists are on board, their designated bus (waka, boat, vessel), the guide/ driver welcomes everyone and announces that the tourists will not only experience Maori song, dance, food, etc., but will also become Maori for the night. Each busload of tourists is branded as a tribe (iwi) with a chief (rangatira) (elected from among the tourists; the chief seems to usually end up being white, male, American), and the driver offers to teach the tourists some Maori language typically just one phrase/ greeting formula Kia Ora, Hello/ Good Luck/ Good Health/ Thank You (cf. Auger 2002), which may be emically more signicant than the category greeting usually implies, though we surmise that most international tourists remain oblivious to these pragmatic subtleties. Kia Ora is, then, to be repeated in unison by the tourists following a prompt from the guides and other performers. The chiefs become privileged participants in representing their tribes in the Vil-


Crispin Thurlow and Adam Jaworski

Figures 37. The ubiquitous welcome of tourism

Tourism discourse: Languages and banal globalization


lage Welcome, gift presentation, various speeches throughout the night, and so on. One of such privileges includes the performance of a hongi with the guide/ driver in front of all the other tourists on the bus before its departure for Tamaki Maori Village, as represented in the following extract:
Extract 6: Didnt he do well? G = Guide (Driver), K= Kenny, the chief tourist, T = Unidentied tourists 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 G: for all you people from different tribes (.) this is how we the Maori people will usually greet each other (.) grab my right hand Kenny (.) (off mike) stand up stand up (Kenny stands; he towers over the driver) (light laughter) (1) now go down on the step (Kenny goes one step down) (laughter) (2) ok we- (.) put your left hand on my shoulder Kenny (1) ok (.) now what we do, we press our noses together twice (.) and then we say kia ora ok (.) nice and gentle (.) dont go (thrusts his head forward quickly towards Kennys face; Kenny tilts his head backwards in a reex) (light laughter) (2) and whatever you do: (.) dont kiss me. [ (inaudible speech to Guide) (continued laughter) (2) (Guide and Kenny perform a hongi, cameras ash) kia ora: didnt he do well (.) how-bout a big round= (applause 4 sec.; loud female voice) yeeeahh =of applause for Kenny (.) my people interpret the hongi like this when the two noses come together (.) its the sharing of common breath creating a legion of friendship (1) as a point of interest for you, we the Maori tribe here in Te Arawa are familiar to all this area of Rotorua and Bay of Plenty (.) we are the only Maori tribe in New Zealand that hongi twice (.) all other tribes do it once (.) thats our trade mark. (.) were now gonna pull out rangatira our big kahuna, the big chief Ken here to the entrance way (.) Ill make the ofcial welcome the challenge youre gonna have a wonderful evening (.) kia ora (loud voices) kia ora

T: G: T: G:

T: G: K: T: G: T: G:



Crispin Thurlow and Adam Jaworski

Although the hongi is framed as a typical Maori greeting (lines 12) and as a way of establishing a legion of friendship (line 22) between two people, the guides display and lesson in Maori etiquette has an undercurrent of cultural subversion and resistance to dominant ideologies of tourism. The guide does not unambiguously adopt a stance of a friendly, deferential and subservient host. Under the guise of humour reminiscent of genres where mock-aggression and mild humiliation are part of the participation ritual (e.g. TV quizshows), he positions Kenny, the archetypal powerful and wealthy Westerner about to be exposed to Pre-European lifestyle experience of customs and traditions ( Maori%20Village/Home), as a relatively powerless and ignorant foreigner. In order to teach Kenny the hongi ritual, the guide instructs him to adopt appropriate body posture. When Kenny comes to the front of the coach and faces the guide, the latter unceremoniously orders Kenny to go one step down, to reduce the difference in their height having their faces at the same level is more amenable to hongi and symbolically maintains a proxemic equilibrium between the two men. The driver uses unmitigated directives, stand up stand up (line 3), now go down on the step (line 6), reminiscent of an adult disciplining a child, and this bossing Kenny around elicits outbursts of laughter from the onlookers on the coach. In lines 1011, the guide teases Kenny, implying that he is likely to hongi inappropriately nice and gentle (.) dont go (thrusts his head forward. . . ). The guides hyperbolic head-butt is clearly an exaggeration for comic effect as he cannot realistically expect Kenny to act in such a foolish manner. The guide also seems to intentionally frighten Kenny with his mock head-but only to elicit a reaction of slight panic from Kenny and more laughter from the other tourists. Another ridiculing turn at Kennys expense is the guides teasing, heteronormative joke, and whatever you do: dont kiss me (line 14). The guide then proceeds with the hongi (line 17) and again positions Kenny as a childlike gure who deserves a round of applause as a reward for his performance (another game show-like feature). The guide then appears in total control of the situation, a knowledgeable expert, as well as a mocking directorchoreographer of the scene, blatantly Othering (Jaworski and Coupland 2005) Kenny by adopting the key of teasing and ridicule. Indeed, the guides control over this intercultural exchange is also manifested in his artful management of tourists crossing into Maori by his constant pronominal ingrouping/outgrouping; visitors are reminded that they (you people) are outsiders merely playing at being Maori (we).

Tourism discourse: Languages and banal globalization


Tourism discourse and/as banal globalization

If the system is to work, the agents must not be entirely unaware of the truth of their exchanges, while at the same time they must refuse to know and above all to recognize it. In short, everything takes place as if agents practice were organized exclusively with a view to concealing from themselves and from others the truth of their practice . . . (Bourdieu 1977: 6)

We started this paper with one of Pierre Bourdieus well-known observations about the nature of symbolic power/privilege; we start our conclusion with his extension of this idea. In both cases, we are reminded of the fundamental power of language in the contexts of everyday life; how the value of symbolic resources like language is inextricably tied to the political/economic privilege (or not) of speakers, and how easily readily, even those who benet most from the inequalities of symbolic markets misunderstand (or misrecognize, to use Bourdieus own terms) their privilege. In the context of tourism as a global cultural industry we nd the role of local languages exemplifying a number of ways that Bourdieus critique rings true. Following Bourdieu (also 1991), all linguistic exchanges are also economic exchanges; however, under the new economic conditions of globalization, existing language forms and congurations (e.g. bilingualism) are put to new uses, gain new value and become objects of intense scrutiny, as well as vehicles and sites of ideological struggle, contestation, legitimation and authentication of ethnic, national and other subject positions. In the context of tourism, this is especially clear in the proliferation of theme parks, open-air museums, festivals and spectacles laying out displays of ethnicity, nationality, culture, urban or industrial heritage through the (re-)invented narratives of group origins, history and present-day lives (e.g. Kirshenbatt-Gimblett 1998; Bruner 2005; Heller 2011). These are also the most obvious areas of tourism-driven activity, where language (and other semiotic codes) become vehicles of explicit staging (Edensor 2001) or high performance (N. Coupland 2007) in which gathered (rather than simply co-present) participants overtly orient to the formal properties of code through metapragmatic commentary and the evaluation, translation and labelling of linguistic items. Such performances are heavily marked by claims to ownership, belonging and authenticity, or, conversely, by pragmatic instrumentalism, playfulness and appropriation, and not infrequently by a mixture of all these positions dynamically and dialectically negotiated in the process of staged, ritualized enactments and interactions. The role of language in identity formation is crucial, then, but not as straightforward and clear-cut as might be assumed there is no one-to-one correspondence between linguistic units and ethnic, social or cultural formations (Le Page and Tabouret-Keller 1985;


Crispin Thurlow and Adam Jaworski

for discussion see N. Coupland 2006). Sociolinguistic items, be they language codes or subtle phonological variants, may be strategically deployed as indexes of specic identities, but their projection and interpretations are always ltered through a plethora of objective and subjective dimensions of self- and otherperception, uptake, interpretive frames and communicative goals, uptake and the political economy of difference (Heller 2003). For example, as observed by Rampton (1995), traditional conceptions of what it is to be a native speaker break down when instrumental language use is separated from its symbolic value as a means of manifesting and asserting ones ethnic or national allegiances or loyalties, or when language inheritance is separated from language allegiance and the degree of linguistic expertise. Likewise, in the context of tourism, we see speakers deploying old linguistic resources in novel forms, styling self and other in new, often surprising ways, playing with social norms and establishing new regimes of truth, and unexpectedly conating instrumental and emotive uses of language, or shifting between use-value and exchange-value. Even though tourists may experience their encounters with hosts as singular, most touristhost interactions are ritualized (Jaworski and Thurlow 2011), ranging from low-level play with relatively little attention to the accountability to an audience (Bauman 2001 [1975]), through to mundane and high or artful performances (cf. reference to N. Coupland, above). The performative nature of the exchanges in our Extracts 1, 3, 4, and 5, for example, is certainly heightened by their mediated nature which accords the status of ratied audience to the radio listeners and TV viewers. All the same, the insertion of foreign languages into tourists talk, together with laughter, jokey metapragmatic comments, intentional and unintentional mangling of the target forms re-keys (Goffman 1974) these greeting sequences, introductions and service encounters into humorous, make-believe greeting sequences, introductions and service encounters. In Extract 5, Ken enacts a ritual hongi greeting with the guide in which he is symbolically stripped of his status as an adult, middle-class, White American and assumes a performed, as-if status (Turner 1974) as a Maori novice in an initiation ritual. All of these strips of activity then are ritualized, performative, self-reexive (self-positioning) and dynamic (reframing), involving verbal and nonverbal elements from two or more recognizable linguistic codes that cannot be easily juxtaposed as pre-gured, separate and discrete languages. Associated predominantly with urban, ludic and playful language use, such texts have been variously referred to as heteroglossic (after Bakhtin 1981), transidiomatic (Jacquemet 2005), crossed (Rampton 1995), polylingual (Mller 2008), or metrolingual (Otsuji and Pennycook 2010, following Mahers 2005, 2010 notion of metroethnicity). In Pennycooks characterization of metrolinguistic practice the idea of

Tourism discourse: Languages and banal globalization


playfulness resurfaces as one in which people of different and mixed backgrounds use, play with and negotiate identities through language; it does not assume connections between language, culture, ethnicity, nationality and geography, but rather seeks to explore the contingencies of these categories; its focus is not on language systems but on languages as emergent from contexts of interaction (2010: 85). No doubt, tourist multilingual practices demonstrated by our data (or models for tourist multilingualism as offered by the guidebook glossaries) draw on their situated, local knowledge of what constitutes (or not) the target language. For this reason, we may think of their multilingual competencies as symbolic (Kramsch 2006; Kramsch and Whiteside 2008), i.e. only appropriating or approximating someone elses language, although also, possibly, enriching their own communicative competence (Hymes 1972) with additional resources freeing up new embodied experiences, emotional resonances, and moral imaginings as part of the exchange of symbolic goods under globalization (Kramsch 2006: 251). However, as Kramsch observes, symbolic competence is not so much a set of identiable linguistic skills.
Rather, it is a mindset that can create relationships of possibility or affordances (van Lier 2004: 105), but only if the individual learns to see him/herself through his/her own embodied history and subjectivity and through the history and subjectivity of others. Our symbolic survival is contingent on framing reality in the way required by the moment, and on being able to enter the game with both full involvement and full detachment. In this sense, the notion of symbolic competence is a late modern way of conceiving of both communicative and intercultural competence in multilingual settings. (Kramsch and Whiteside 2008: 668; our emphasis).

Trying out snippets of local languages by tourists may indeed be greeted by hosts with pleasure as acts of touristsmaking an effort to engage, show appreciation and involvement with the above), but rituals and performances may be as much a source of enjoyable enhancement of experience as a fearsome act with the potential to subvert the status quo (cf. Bauman 2001 [1975]). All communicative acts laden with a high dose of meta-cultural commentary or staging are rich sites of ideological work, including manifestation of linguistic ideologies, which are signicant for social as well as linguistic analysis because they are not only about language. Rather, such ideologies envision and enact links of language to group and personal identity, to aesthetics, to morality, and to epistemology (Woolard and Schieffelin 1994: 5556). No doubt, it is easy to interpret our data merely as allowing the tourists to step outside their everyday identities and slip into non-committal, innocent and playful role-play of a cultural Other.


Crispin Thurlow and Adam Jaworski

However, as noted above, the seemingly innocent here-and-now of touristhost interactions is rooted in broader historical trajectories of travel, colonization, global inequalities and privilege, which reminds us of Blommaerts (2005: 131) argument that [t]he synchronicity of discourse is an illusion that masks the densely layered historicity of discourse. And as we have noted above with reference to Jane Hills (2008) work, taking on tourist languages and identities as material for humorous transformations of tourist identities has a more profound, language ideological effect of re-asserting the tourist as having the upper hand in their dealings with hosts. Unless, that is, hosts use precisely the same strips of activity (cf. Extract 6, above) to turn the tables on the tourists and humorously but pointedly reverse the power dynamic through play and deploy their cultural and linguistic heritage as an instrument of power/knowledge. At a more general level, however, for us, the ideological force realized in or generated by the collective actions of tourists orienting to or displaying their symbolic competencies in the languages of their travel destinations lies in what we have been calling banal globalization (Thurlow and Jaworski 2010, 2011) in parallel with Szerszynski and Urrys (2002, 2006) term banal globalism, and following the ideas of Mike Billig (1995, on banal nationalism) and Ulrich Beck (2006, on banal cosmopolitanism). We choose to invoke the notion of banality for framing and understanding tourism discourse as being rooted in everyday communicative actions and textual practices including those mediatized moments in popular culture (e.g. the broadcast media). By everyday we do not mean to say that these actions/practices are either foolish or inconsequential: on the contrary. It is, we suggest, at the level of innocent texts and harmless (inter)actions that globalization is actually realized. For example, Szerszynski and Urry (2002) nd examples of globalizing imagery in everyday, recurring TV imagery which includes globes, birds-eye-views of generic global environments, images of the exotic Others consuming global brands and products, children standing for the globe in charitable appeals, and so on. These discursive practices may well be trite but they are far from trivial. Just as small talk is always pragmatically speaking big talk (cf. J. Coupland 2000) and just as reiterative performances of gender solidify and naturalize the heteronormative matrix (Butler 1990), so too do the mundane practices embodied and mediated of tourism turn out to be global in their reach and possibly also in their impact. Acknowledgements We are grateful to Giorgia Aiello for her help with thinking through the pragmalinguistic implications of Rick Steves use of Italian in Extract 1 and also the confusing (even to a native speaker) Italian in Extract 3. We also thank Panama

Tourism discourse: Languages and banal globalization


Authority of Tourism and the New Zealand Tourism Board for permission to reproduce their materials. Every reasonable effort was made to contact copyright holders for images which are otherwise reproduced under fair use standards.

1. The orginal broadcast of the Rome City Guide can be downloaded from Rick Steves own archives at <>.

Agar, Michael. 1994. Language shock: Understanding the culture of conversation. New York: William Morrow and Company. Bakhtin, Mikhail M. 1981.The dialogic imagination: Four essays [ed. Michael Holquist]. Austin: University of Texas Press. Barthes, Roland. 1977[1964]. Rhetoric of the image. In Image, Music, Text, 3251. New York: Hill and Wang. Bauer, Laurie and Peter Trudgill. (eds.). 1998. Language myths. New York: Penguin Books. Bauman, Richard. 2001 [1975]. Verbal art as performance. In Alessandro Duranti (ed.), Linguistic anthropology: A reader, 165188. Malden, MA: Blackwell. Bauman, Zygmunt. 1998. Globalization: The human consequences. Cambridge: Polity. Beck, Ulrich. 2006. The cosmopolitan vision. Cambridge: Polity. Bell, Allan. 2009. Language style as audience design. In Nikolas Coupland and Adam Jaworski (eds.), The new sociolinguistics reader, 265275. Basingstoke: Palgrave Macmillan. Bhattacharyya, Deborah P. 1997. Mediating India: An analysis of a guidebook. Annals of Tourism Research, 24. 371389. Billig, Michael. 1995. Banal nationalism. London: Sage. Blommaert, Jan. 2005. Discourse: A critical introduction. Cambridge: Cambridge University Press. Boorstin, Daniel. 1964. The image: A guide to psuedo-events in America. New York: Harper. Bourdieu, Pierre. 1977. Outline of a theory of practice. Cambridge: Cambridge University Press. Bourdieu, Pierre. 1991. Language and symbolic power [ed. J. B. Thompson; trans. G. Raymond and M. Adamson]. Cambridge: Polity Press. Bruner, Edward M. 2005. Culture on tour: Ethnographies of travel. Chicago: University of Chicago Press. Butler, Judith. 1990. Gender trouble: Feminism and the subversion of identity. NewYork: Routledge.


Crispin Thurlow and Adam Jaworski

Cameron, Deborah. 2000. Styling the worker: Gender and the commodication of language in the globalized service economy. Journal of Sociolinguistics, 4. 323347. Coupland, Justine. (ed.). 2000. Small talk. London: Longman. Coupland, Nikolas. 2006. The discursive framing of phonological acts of identity: Welshness through English. In Janina Brutt-Grifer and C. Evans Davies (eds), English and ethnicity, 1948. Basingstoke: Palgrave Macmillan. Coupland, Nikolas. 2007. Style: Language variation and identity. Cambridge: Cambridge University Press. Coupland, Nikolas (ed.). 2010. The handbook of globalization. Oxford: Wiley-Blackwell. Crang, Mike. 1999. Knowing, tourism and practices of vision. In David Crouch (ed.), Leisure/tourism geographies: Practices and geographical knowledge, 238256. London: Routledge. Cronin, Michael. 2000. Across the lines: Travel, language, translation. Cork: Cork University Press. Dunn, David. 2005. We are not here to make a lm about Italy, we are here to make a lm about me . . . British television holiday programmes representations of the tourist destination. In David Crouch, Rhona Jackson and Felix Thompson (eds.), The media and the tourist imagination: Converging cultures, 154169. London: Routledge. Dunn, David. 2006. Singular encounters: Mediating the tourist destination in British television holiday programmes. Tourist Studies, 6. 3758. Duranti,Alessandro. 1997. Universal and culture-specic properties of greetings. Journal of Linguistic Anthropology,7. 6397. Edensor, Tim. 2001. Performing tourism, staging tourism: (Re)producing tourist space and practice. Tourist Studies, 1(1). 5981. Favero, P. 2007. What a wonderful world!: On the touristic ways of seeing, the knowledge and the politics of the culture industries of otherness. Tourist Studies, 7(1). 5181. Foucault, Michel. 1976. The birth of the clinic. London: Tavistock. Franklin, Adrian and Mike Crang. 2001. The trouble with tourism and travel theory. Tourist Studies, 1. 522. Goffman, Erving. 1974. Frame analysis: An essay on the organization of experience. New York: Harper & Row. Hannerz, Ulf. 1996. Transnational connections: Culture, people, places. London: Routledge. Heller, Monica. 2003. Globalization, the new economy and the commodication of language and identity. Journal of Sociolinguistics, 7. 473498. Heller, Monica. 2011. Paths to post-nationalism. New York: Oxford University Press. Hill, Jane. 2001 [1998]. Language, race, and white public space. In Alessandro Duranti (ed.), Linguistic anthropology: A reader, 450464. Malden, MA: Blackwell. Hill, Jane. 2008. The everyday language of White racism. Oxford: Blackwell. Hymes, Dell. 1972. On communicative competence. In John Pride and Janet Holmes (eds.), Sociolinguistics, 269283. Harmondsworth: Penguin.

Tourism discourse: Languages and banal globalization


Irvine, Judith. 1997. When talk isnt cheap: Language and political economy. In Donald Brenneis and Ronald Macaulay (eds.), The matrix of language: Contemporary linguistic anthropology, 258283. Boulder, CO: Westview. Jack, Gavin and Alison Phipps. 2003. On the uses of travel guides in the context of German tourism to Scotland. Tourist Studies, 3. 281300. Jacquemet, Marco. 2005. Transidiomatic practices: Language and power in the age of globalization. Language and Communication, 25. 257277. Jaworski, Adam. 2009. Greetings in touristhost encounters. In Nikolas Coupland and Adam Jaworski (eds.). The new sociolinguistics reader, 662679. Basingstoke: Palgrave Macmillan. Jaworski, Adam and Justine Coupland. 2005. Othering in gossip: you go out you have a laugh and you can pull yeah okay but like. . . . Language in Society, 34. 667694. Jaworski, Adam and Crispin Thurlow. 2009. Talking an elitist stance: Ideology and the discursive production of social distinction. In Alexandra Jaffee (ed.), Stance: Sociolinguistic perspectives, 195226. New York: Oxford University Press. Jaworski, Adam and Crispin Thurlow. 2010. Language and the globalizing habitus of tourism: A sociolinguistics of eeting relationships. In Nikolas Coupland (ed.), The Handbook of Language and Globalisation, 256286. Oxford: Wiley-Blackwell. Jaworski, Adam and Crispin Thurlow. 2011. Making contact: Language, tourism and globalization. London: Routledge Kirshenblatt-Gimblett, Barbara. 1998. Destination culture: Tourism, museums, and heritage. Berkeley: University of California Press. Kramsch, Claire. 2006. From communicative competence to symbolic competence. The Modern Language Journal, 90. 249252. Kramsch, Claire and Anne Whiteside. 2008. Language ecology in multilingual settings: Toward a theory of symbolic competence. Applied Linguistics, 29. 645671. Le Page, Robert B. and Andre Tabouret-Keller. 1985. Acts of identity: Creole-based approaches to language and ethnicity. Cambridge: Cambridge University Press. Lash, Scott and John Urry. 1994. Economies of signs and spaces. London: Sage. Lippi-Green, Rosina. 1997. English with an accent: Language, ideology, and discrimination in the United States. New York: Routledge. Maher, John C. 2005. Metroethnicity, language, and the principle of Cool. International Journal of the Sociology of Language, 11. 83102. Maher, John C. 2010. Metroethnicities and metrolanguages. In N. Coupland (ed.), The handbook of language and globalization. Malden, MA: Wiley-Blackwell. pp. 575 591. Mller, Janus S. (2008). Polylingual performance among Turkish-Danes in late-modern Copenhagen. International Journal of Multilingualism, 5. 217236. Otsuji, Emi and Alastair Pennycook. 2010. Metrolingualism: Fixity, uidity and language in ux. International Journal of Multilingualism, 7. 240254. Pennycook, Alastair. 1989. The concept of method, interested knowledge, and the politics of language learning. TESOL Quarterly, 23. 589618. Pennycook, Alastair. 2010. Language as a local practice. London: Routledge.


Crispin Thurlow and Adam Jaworski

Phipps, Alison. 2007. Learning the arts of linguistic survival: Languaging, tourism, life. Clevedon: Channel View. Rampton, Ben. 1995. Crossing: Language and ethnicity among adolescents. London: Longman. Szerszynski, Bronislaw and John Urry. 2006. Visuality, mobility and the cosmopolitan: Inhabiting the world from afar. The British Journal of Sociology, 57. 113131. Thurlow, Crispin 2004. Relating to our work, accounting for our selves: The autobiographical imperative in teaching about difference. Language and Intercultural Communication, 4. 209228. Thurlow, Crispin and Adam Jaworski. 2010. Silence is golden: Elitism, linguascaping and anti-communication in luxury tourism. In Adam Jaworski and Crispin Thurlow (eds.), Semiotic Landscapes: Language, Image, Space, 187218. London: Continuum. Thurlow, Crispin and Adam Jaworski. 2010. Tourism discourse: Language and global mobility. London: Palgrave MacMillan. Thurlow, Crispin and Adam Jaworski. 2011. Banal globalization? Embodied actions and mediated practices in tourists online photo-sharing. In Crispin Thurlow and Kristine Mroczek (eds.), Digital discourse: Language in the new media. New York: Oxford University Press. Turner, Victor. 1974. Dramas, elds, and metaphors: Symbolic action in human society. Ithaca: Cornell University Press. Urry, John. 2002. The Tourist Gaze [2nd ed.]. London: Sage. Urry, John. 2007. Mobilities. Cambridge: Polity. Van Lier, Leo. 2004. The ecology and semiotics of language learning. A sociocultural perspective. Dordrecht: Kluwer Academic. Woolard, Kathryn A. and Bambi B. Schieffelin. 1994. Language ideology. Annual Review of Anthropology, 23. 5582. Crispin Thurlow,, is Associate Professor of Interdisciplinary Arts & Sciences at the University of Washington (USA) where he also holds adjunct positions in Linguistics, Anthropology and Communication. His recent books include Tourism Discourse: Language and Global Mobility (with Adam Jaworski, 2010, Palgrave Macmillan) and Digital Discourse: Language in the New Media (with Kristine Mroczek, 2011, Oxford University Press). Adam Jaworski,, is Professor of Language and Communication at Cardiff University (Wales, UK). His books include Semiotic Landscapes: Language, Image, Space (with Crispin Thurlow, 2010, Continuum) and Discourse, Communication and Tourism (with Annette Pritchard, 2005, Channel View). Adam is also co-editor of the book series Oxford Studies in Sociolinguistics.

Do you have another Johan? Negotiating meaning in the operating theatre*


Abstract This paper discusses language use at a workplace in a context of instability and diversity. Its focus is on the operating theatre, where communication is an integral part of complex, collaborative tasks, impacting on patient-safety, staff well-being and overall quality of health care. In the operating theatre health care professionals gather to work on the recurring task of surgical operations, in teams that exist only for the duration or parts of the task. Not only do the members of these unstable teams have different professional backgrounds, such as surgery and nursing, they also draw on different, social, cultural and linguistic resources. The paper shows how this instability and diversity which is so characteristic of contemporary society plays out in the moment-by-moment use of language at the operating table. On the basis of prolonged eldwork in a London hospital and a unique set of audio- and video-recordings we show how surgeons formulate requests and how nurses and surgical trainees disambiguate these requests on the basis of their prior experiences with surgical instruments and equipment, the surgical procedure, and, crucially, the surgeons idiolect. We analyze instances where this process of disambiguation is highly successful, as well as examples where it is not. We tease out the strategies that nurses and surgeons deploy to deal with this ambiguity and explore ways to deal with instability and diversity in professional communication. Keywords: professional interaction; intercultural communication; medical discourse; linguistic ethnography

1. Introduction A key challenge for the National Health Service (NHS) in the United Kingdom (UK) and other major organizations across the world in the years to come is to


Jeff Bezemer, Alexandra Cope, Gunther Kress and Roger Kneebone

deal with professional communication in unstable and diverse teams that is, communication between people with diverse professional, social, cultural and linguistic backgrounds who work on recurring tasks, such as surgical operations, in teams that exist only for the duration of the task, or even only for parts of it. Thus these professionals step in and out of newly formed teams, participating in the performance of complex tasks, often with colleagues whom they have never met before. Health care services are increasingly delivered by such transient teams (Finn and Waring 2006). Contingencies inherent in health care (e.g., emergencies), high workloads and reliance on temporary staff, regulatory caps on working hours (e.g. the European Working Time Directive) and demands of the workforce (e.g., working part-time) all contribute to the promotion of exible and transient team work. The NHS faces an annual turnover of clinical staff of over 20 per cent (Hutt and Buchan 2005), and there are limited opportunities for developing shared frames of reference. Signicant parts of this workforce have been trained overseas. About 25 per cent of London based nurses are trained overseas (Buchan et al. 2005), while 35 per cent of all NHS medical staff did their undergraduate training outside the UK (Hutt and Buchan 2005). The paper discusses how these changing social and economic contexts impact on professional communication and, ultimately, on public service delivery. Previously members of an operating team developed, over the years, a shared language for, e.g., naming instruments, articulating requests and announcing the next step in an operation. Now, they bring different, socially and culturally shaped professional experiences to the team, raising the need for opportunities to produce and retain shared understandings. Thus while there is more diversity in human resources, there are far fewer opportunities to develop a shared language and pass on essential knowledge and expertise to new employees. The paper explores this tension through detailed analysis of audio and video recorded operations at a major teaching hospital in London, seizing a unique opportunity to study communication in a site where it has real consequences for patient-safety, staff well-being, and the overall quality of healthcare (Lingard et al. 2004; Nestel and Kidd 2006; Williams et al. 2007). We begin by reviewing the research that has been done in this area, both in the social and the medical sciences, and set out our own, linguistic-ethnographic approach. Following that we discuss the key features of communication in the operating theatre, focusing on how surgeons formulate requests for instruments and how nurses and surgical trainees disambiguate these requests on the basis of their prior experience with surgical instruments and equipment, the surgical procedure, and, crucially, the surgeons idiolect. We discuss examples of instances where this process of disambiguation is highly successful, as well as examples where it is not. We tease out the strategies that nurses and surgeons deploy to deal with this am-

Do you have another Johan? Negotiating meaning in the operating theatre


biguity and explore ways to deal with instability and diversity in professional communication.

2. Researching communication in the operating theatre Communication at the workplace is a key concern of applied linguistics (Gumperz 1982; Clyne 1994; Bremer et al. 1996; Di Luzio et al. 2001; Cicourel 2003; House et al. 2003; Stubbe et al. 2003), yet few applied linguists and conversation analysts have gained access to the operating theatre. Catherine Pettinari (1988) researched text and talk in operating theatres. She looked at how surgical trainees learn to write operation reports, having observed the operations that a range of trainees reported on over time. A small number of CA studies are based on audio and video recordings of communication in the operating theatre (Mondada 2007; Svensson et al. 2007; Koschmann et al. forthcoming), yet only in Svenssons work has the interaction between surgeon and nurse been explored. He analyzes how the timely exchange of instruments between nurse and surgeon is achieved in verbal and non-verbal communication. The majority of studies on communication in the operating theatre published in medical journals are based not on observations and audio and video recordings but on what nurses and surgeons themselves report, in interviews (Keddy et al. 1986; Svensson 1996; Snelgrove and Hughes 2000; Gjerberg and Kjlsrd 2001; Nestel and Kidd 2006; Williams et al. 2007) or in surveys (Bourhis et al. 1989; MacKay et al. 1991; Gjerberg and Kjlsrd 2001; Hojat et al. 2001; Manojlovich and DeCicco 2007). Only Lorelei Lingard c.s. did observational work on communication in the operating theatre and published her ndings in journals for medics and nurses (Espin and Lingard 2001; Lingard et al. 2002; Lingard et al. 2004). These ndings are based on structured observations of a large number of operations to allow for descriptive statistics of the occurrence of communication failures. Indeed, there has been very little cross-over between medical and social/ linguistic studies of the operating theatre. Linguists and conversation analysts have only had limited access to the operating theatre as a research site, while medical researchers have had limited access to (and appreciation of) the more ne grained analytical apparatus offered by applied linguistics and conversation analysis to study communication. In all medically oriented studies problems of communication were reported, yet the data collection techniques did not capture any of the details of the interaction within which the problems arose. Indeed many such problems initially go unnoticed, especially by the surgeons and nurses themselves. Much of what nurses and surgeons do is instantiated in the subtle and ne grained detail of body movements such as the positioning


Jeff Bezemer, Alexandra Cope, Gunther Kress and Roger Kneebone

of a retractor, or a shift in gaze from operative eld to scrub nurse. Thus video analysis produces a much richer and nuanced account of communication than what can be captured on-the-spot and in eld notes by researchers, or what can be recollected and re-articulated in interviews with the participants after the observed event. Our study adopts a linguistic-ethnographic approach, bringing together close analysis of multimodal communication with ethnographically informed analysis of the wider context. Access was gained to a major teaching hospital in London. Fieldwork took place between June 2009 and July 2010. We observed 40 operations, involving 5 consultant-surgeons, 5 surgical registrars, 5 (senior) house ofcers, 10 medical students, 25 nurses and operation department practitioners and 5 anaesthetists. The operations cover different general surgical procedures, lasting between 45 minutes and 6 hours, totalling approximately 70 hours. The overall time spent observing in theatres exceeds that, covering not only the actual operations but also the preparations and cleaning up in the operating theatre and its adjacent rooms: the prep room, where nurses sort the instruments, and the anaesthetic room, where the anaesthetist puts the patient asleep. Many hours were spent between cases, when often opportunities arose to talk to staff and students. Staff were also seen in coffee rooms and departmental meetings. The study is based on close collaboration between clinicians and ethnographers. In the rst phase of data collection observations were carried out by an ethnographic researcher (Bezemer). In the second phase data was collected jointly by the ethnographic researcher and a surgeon (Cope). We hold regular data sessions with one more surgeon (Kneebone) and a semiotician (Kress), discussing small clips of video recordings of teaching and learning. We have collected audio and video recordings of 10 cases, using a wireless microphone worn by one of the surgeons, and in-built video cameras in the handle of the operating lamp to capture the operative eld. We also record the view that is created by the laparoscope (a camera that is inserted into body cavities). We keep eld notes of all operations observed, including the 10 cases which were recorded and jointly observed, producing two different sets of eld notes. All staff in theatre and all patients involved have given informed consent to collect these data. Ethical approval was granted by the NHS Research Ethics Committee. The paper uses descriptive and analytic procedures from applied linguistics, social semiotics and ethnography for investigating the temporal unfolding of multimodal communication and its associated social processes in situated encounters (Roberts et al. 2000; Roberts et al. 2003; Wass et al. 2003; Iedema et al. 2006; Kissmann 2009; Bezemer and Jewitt 2009; Heath et al. 2010). The analysis is focused on a) the participants of situated encounters in the operating

Do you have another Johan? Negotiating meaning in the operating theatre


theatre, that is, nurses and surgeons, and their socially and culturally shaped repertoires of communicative resources, their habitual practices, expectations, and identities; b) the types of activities in which they engage, their embodied interaction, the objects they use, and the physical surroundings; c) the institutional context of the operating theatre and the policies, discourses, and ideologies that shape it (Rampton 2007). The analysis proceeded through a sequence of steps. Field notes were selected in which observations were reported on communication between nurses and surgeons. Through iterative viewing of the audio and video data instances of communication were selected for close analysis which exemplied emerging themes. Data sessions provided opportunities to examine video clips with a multidisciplinary research team, develop preliminary analyses and identify phenomena worthy of more detailed analysis (Heath et al. 2010). We then proceeded to describe key incidents (Erickson 1977) and place them in some relations to the wider social context, using the key incident as a concrete instance of the workings of abstract principles of social organization (p. 61). The key incident we reconstruct in this paper is based on detailed analysis of an audio clip capturing what was said by nurses and surgeons during an operation, a video clip of communication inside the patients body, involving hand movements and handling of instruments by the consultant and his assistants; eld notes and photographs of what happened around the operating table; documents circulating in the operating theatre, such as forms and reports; interviews with the nurses involved on the following day; and a short interview with the consultant, held several months after the operation, in which we checked key aspects of the analysis presented here.

3. Communication in the operating theatre The operating theatre is a complex site of communication (Lingard et al. 2002; Lingard et al. 2004; Bezemer et al. forthcoming). Most operations involve a team of surgical trainees led by a consultant-surgeon, and a team of specialized theatre nurses usually led by the scrub nurse, who stands at the operating table to pass instruments from a trolley to the surgeons (see Figure 1). The scrub nurse also communicates with circulating nurses, who bring materials from stock rooms and set up technical equipment around the operating table. Typically, communication between nurses and surgeons involves a consultantsurgeon making a request to the scrub nurse, e.g., Clip please. The scrub nurse then responds to this request by passing the requested instrument, and does not use speech in this interactional exchange at all. Often the request is not


Jeff Bezemer, Alexandra Cope, Gunther Kress and Roger Kneebone

Figure 1. Inside the operating theatre: the scrub nurse (far left), rst assistant, second assistant and consultant-surgeon (far right) at work.

articulated in speech either. Instead the scrub nurse relies on her intercorporeal knowing (Hindmarsh and Pilnick 2007; Bezemer et al. forthcoming), i.e., her ability to read and place subtle cues in the surgeons bodily expressions in the context of the unfolding operation, thus anticipating upcoming requests. As the senior theatre nurse at the research site told us, scrub nurses are expected to watch the operation carefully. Requests can also come from the consultant-surgeons trainees, in particular when they are given the chance to lead parts of the operation under the consultants supervision. The consultant is usually assisted by two assistants. Assistants can be medical students, house ofcers, senior house ofcers (SHOs) or specialist registrars. Typically the registrar acts as rst assistant, while an SHO acts as second assistant. Requests by surgeons usually take the form of imperatives and seldom identify a specic addressee (Green lights off please). The few exceptions we recorded (Have you got some local anaesthetic Miranda?) typically happened at the beginning or end of the operation, when topics of conversation, participant status, turn-taking, et cetera mark a signicant change in the denition of the situation. During the operation, however, nomination of addressees is usually implicit. The object of a request implies if the request is directed towards the scrub nurse or someone else in operating theatre. Like the participating surgeons, the scrub nurse is scrubbed up, that is, s/he can only touch sterile materials such as the surgical instruments. Material or equip-

Do you have another Johan? Negotiating meaning in the operating theatre


ment which is not sterile, such as suction machines, or light switches, cannot be touched by the scrub nurse, and therefore a surgeons request for changing the set-up of such equipment is, by implication, directed at the circulating nurse(s). Early announcements of upcoming requests are also common. For instance, when a consultant stated, I need a rectal washout in a minute circulating nurses started wheeling in trolleys from the prep room with the necessary kit and a stool for the consultant to sit on. Rather than asking for specic instruments or equipment the surgeon identies an activity; the nurses are expected to know which instruments are required for carrying out this activity; and they are expected to take the announcement of the activity as an (indirect) request, not as a statement. The request stands for a preferred sequence of coordinated activities, aimed at making available, at the surgeons preferred time and place, the tools that this consultant prefers to work with. All of these activities, as well as the formulation that triggers them, vary from surgeon to surgeon. The formulation, I need a rectal washout, is a metonym, part of the surgeons idiolect which nurses are expected to learn. Requests do not always result in the scrub nurse providing that which was requested (Williams et al. 2007). Screens, lights, suction machines, gas dispensers, diathermy machines all equipment fails to work from time to time, and it is a recurring source of irritation. Typically, when a surgeon realizes that equipment is failing, the surgeon states it rst, by saying something like, Theres no suction. This declarative is not only a request to x it, but also a face-threatening acknowledgement of accountability. Indeed a nurse may have failed to plug in the machine, but equally surgeons may be responsible for gas leaks or even obstructions: The surgeon was actually standing on the tube when he stated Theres no suction. Circulating nurses respond to such requests checking the equipment, usually without saying anything. Unwanted responses from scrub nurses are equally face-threatening, and also potentially harmful. A scrub nurse passing the wrong instrument, or passing the right instrument in the wrong way, may lose face, and runs the risk of being positioned as incompetent; if unnoticed by the surgeon, the wrong instrument may harm the patient. These implications have an important psychological knock-on effect: they cause stress among nurses and surgeons, which impacts negatively on performance and ultimately compromises patient safety (Arora et al. 2010; Nagpal et al. 2010). An important source of unsuccessful requests, i.e., when surgeons do not get what they asked for, is the name of instruments. There are an innite number of different instruments, and an innite number of different names are used by surgeons and nurses to refer to them. The meaning of any of the names used can only be understood in the context in which they appear. There are hundreds of


Jeff Bezemer, Alexandra Cope, Gunther Kress and Roger Kneebone

different graspers, for instance, so that when surgeons ask for one nurses will need to disambiguate the request, drawing on their knowledge of the procedure which is being performed, of the theatre they are in, and of the surgeons idiolect. For instance, one consultant-surgeon asked for a middle blade during an operation. When the scrub nurse subsequently stared at the instrument trolley, the consultant pointed at one of the retractors and said, Its there, look. The scrub nurse then picked up the retractor and handed it to the consultant. While not entirely idiosyncratic the use of the term middle blade is not widespread. There are many different kinds of blades, including scalpel blades, where middle would refer to the size of the blade. In this context, however, the surgeon referred to a particular type of abdominal retractor blade, where middle refers to the relative position of the blade. The scrub nurses hesitation suggests that she had difculty disambiguating the request. The consultant then verbally and non-verbally claried which retractor he wanted. The variety in naming instruments is higher among surgeons than among nurses. A study which asked surgeons and theatre staff (i.e. nurses and operation department practitioners) to name instruments depicted on photographs showed that
as a group, surgeons results were more widely distributed [. . .]. Theatre staff members were more consistent and scored signicantly higher marks than surgeons, reecting the nding that 75% of theatre staff underwent specic teaching as opposed to 22.6% of surgeons.This knowledge is reinforced daily with repeated instrument counting after each operation. Theatre staff work with a variety of consultants and have greater exposure to different surgical instruments. Conversely, individual consultants and consequently their surgical trainees may operate in a specic subspecialty and may be familiar with a limited range of instruments. (Yeung, Cope et al. 2008:n.p)

One may add to this analysis that as surgeons are in a much more powerful position than nurses they are also in a position to subvert certain linguistic norms. In other settings such as classrooms, for instance, students who have been observed to openly question linguistic norms were the most procient students in the class. Following Bernstein (1971) and Bourdieu (1977), Jaspaert and Ramaut (2000) argue that those students, in particular, who do not belong to the dominant group (in their case the dominant language community) will increase their chances of symbolic gain and thus improve their position in the class when they accommodate to the norms of the linguistic market. In the following sections we will analyze these naming and disambiguation practices in more detail. Focusing on one operation, we will describe, rst, what the operation entailed, who was involved, and which instruments were used.

Do you have another Johan? Negotiating meaning in the operating theatre


Second, we will explore examples of successful and unsuccessful disambiguation.

4. A surgical operation The patient has problems swallowing food. That is symptomatic of achalasia, a condition leading to insufcient or uncoordinated relaxation of muscles in the lower oesophagus. The operation, a Hellers cardiomyotomy, involves splitting the ring of muscles where the oesophagus joins the stomach (the so-called GEjunction). The splitting should decrease the muscle contraction and allow food to pass through the lower end of the oesophagus into the stomach. This is done by a laparoscopic or key hole procedure. That means that only small incisions are made to reach into the abdominal cavity, using a camera or laparoscope to create a magnied view on screens around the operating table. The surgeons look up at these screens as they are performing the operation, and others, such as the scrub nurse, can follow the same view on one of the screens within their sight (see Figure 1). The operation is led by a consultant-surgeon. He is assisted by a rst assistant, who is a specialist registrar (who is at least 5 years into his postgraduate surgical training), and a second assistant, who is a senior house ofcer (who is at least 2 years into his postgraduate surgical training). The second assistant stands on the left side of the operating table, and the consultant and rst assistant are on the right side, with the consultant closest to the head of the patient. The operation proceeds through ve key holes, providing openings for ve different instruments. A camera is used to create a view of the abdominal cavity. The camera is held by the rst assistant, who constantly moves the camera around in order to provide the view required by the consultant. Some of these movements are initiated by the rst assistant himself, based on his anticipation of what the consultant wants to see, while others follow from verbal or non-verbal instructions by the consultant. A liver retractor is used to lift the liver. Once the consultant has placed the liver retractor he hands it over to the second assistant, who holds it in this position with his left hand, for the duration of the operation. A grasping forceps or bowel clamp is used to create tension on the muscle which is separated. This instrument is also placed by the consultant, and then handed over to the second assistant, who holds it in his right hand. He needs to hold the instrument in such a way that the right tension is created: too little tension will make the consultants job of splitting the muscle difcult, while too much tension may cause bleeding. Throughout the operation the consultant repeatedly repositions this instrument to provide the right retraction, and verbally corrects the second assistants instrument handling.


Jeff Bezemer, Alexandra Cope, Gunther Kress and Roger Kneebone

The other two key holes are used to insert the two instruments that are actually used to split the muscle bre. These two instruments are controlled by consultant. In his left hand, he holds a grasping forceps to explore the focal area. Of the many different grasping forceps available on the surgical instruments market this surgeon prefers to use the Johan. The jaws of a Johan are fenestrated, and only one of the two sides opens up (see Figure 2). In his right hand the consultant either uses another grasping forceps, allowing him to tear muscle bre apart; or he uses a type of instrument that is connected to the diathermy machine, allowing him to cut tissue and seal off blood vessels using electricity. He alternately uses three instruments for this task: the so-called diathermy hook, the Maryland and the Ligasure.

Figure 2. A Johan grasping forceps

5. Formulating instrument requests Table 1 lists all the surgeons requests for instruments from the scrub nurse during the rst hour of the operation. It shows that the diathermy hook is alternately referred to as hook, diathermy hook, and diathermy, while grasping forceps are referred to with generic names (grasper), anaphora (the other), and proper names (Maryland, Johan). Indeed the meaning of these names is ambiguous. While diathermy hook is sufciently specic to rule out any of the other instruments on the trolley diathermy is not: there are at least three different instruments available that can be connected to the diathermy machine. Maryland and Johan too are names of a particular kind of grasping forceps and rule out any other grasper, but grasper is a generic name that can refer to a range of different graspers, including the Maryland and the Johan. In spite of the ambiguity the scrub nurse provides the expected instrument in most cases. For instance, 8 minutes into the operation, the consultant asked for a grasper. In response, the scrub nurse provided a grasping forceps. Only seven minutes later the consultant asks for a grasper again, yet this time the

Do you have another Johan? Negotiating meaning in the operating theatre


Table 1. Consultant-surgeons requests for laparoscopic instruments during the rst hour of the operation Time (hh:mm) 8.00 8.12 8.46 9.25 14.25 17.19 25.22 30.25 39.55 40.00 50.38 52.32 53.33 54.45 55.58 56.30 57.31 59.13 1.11.54 Request liver retractor grasper hook Johan any grasper Ligasure hook Maryland the grasper ? the other hook another Johan bowel clamp bowel clamp a Johan diathermy the other Johan hook diathermy Provision liver retractor grasping forceps diathermy hook grasping forceps Maryland Ligasure diathermy hook Maryland ? Ligasure grasping forceps diathermy hook not available ? bowel clamp grasping forceps diathermy hook grasping forceps diathermy hook

scrub nurse provides a Maryland. Her knowledge of the context of the requests allowed her to disambiguate them. The rst request was made just after the liver retractor had been placed, and before any other instrument had been inserted. The second request was made just after a request for a tonsil swab, a small cloth used to absorb blood, and just before the consultant started dissecting. The Maryland allows the surgeon to move the tonsil swab around and also to dissect using diathermy. Indeed he requests for the Maryland to be connected to the diathermy machine immediately after he has used the tonsil swab. The scrub nurse anticipated that he needed to dissect using the Maryland very soon, so instead of providing a grasping forceps such as a Johan, which cannot be used for dissection, she provided a Maryland, even though it was suggested that any grasper would do (Grasper please. Any grasper.). Other instances of disambiguation however were not successful, as in the episode we will focus on now. In the transcript below CSstands for consultantsurgeon; SN for scrub nurse; CN for circulating nurse; XN for one of


Jeff Bezemer, Alexandra Cope, Gunther Kress and Roger Kneebone

the nurses (unclear who); and SR for specialist registrar. x means we were unable to hear what was said. Otherwise standard orthography is used.
Extract 1: Do you have another Johan? 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. Do you have another Johan? Holds up an instrument? No an clears throat x We asked you about half an hour ago. I did. I did mention it. So why did nobody get it? Maybe nobody answer. We can open the stack, we can use the stack for second case, for the next case. CS And what are we going to do with the next case? XN x. CS No these are bowel clamps. CN Two is not enough? CS No is enough but the handles are different. So I cant work in the same way. And if you, if I ask for anything. And you dont have it. Can you just tell me. Because I assume that you get them. If I ask about them and you dont answer. SR Can you use the erm endo grasper? CS No. I cant. Okay. Gimme the bowel clamp. SN Passes bowel clamp. CS Thank you. CS Can you clean the camera? SR Hot water please. Fresh hot water. SR Youve got two yeah? CS Yeah but two different handles. SR Is it the right length? SN Offers jug with water, SR dips camera in water. CS Okay. Okay. Lets show it me. SR Puts camera back in. CS Okay. bowel clamp now. SN Offers Johan CS No is not bow- what is Do you know what is a bowel clamp and what is a Johan? Show me the bowel clamps and show me the Johans. SN Holds instruments up. CS Yeah. This is bowel clamp. Okay? [silence of 21sec] CS Okay. Gimme a- give me a Johan. Johan please. CS SN CS SN CS SN CS CN

16. 17. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 33. 34. 35. 36.

Do you have another Johan? Negotiating meaning in the operating theatre 37. 38. SN CS Offers Johan. Thank you.


The episode starts and ends with the consultant-surgeon asking for a Johan. In between, a sequence of exchanges unfolds between the consultant, the scrub nurse and one of the circulating nurses, and also between the consultant and the specialist registrar. Some follow the typical pattern of a) the consultant requesting something, the scrub nurse or registrar providing what was requested, and the consultant acknowledging receipt. There are three such exchanges. First, when the consultant asks for a bowel clamp in Line 18, he gets a bowel clamp. Second, when he asks the registrar to clean the camera the registrar cleans the camera (after asking for and receiving hot water from the scrub nurse allowing him to rinse the camera). Third, when he asks for a Johan in Line 37 he gets a Johan. But other exchanges do not follow this pattern. When he asks for a Johan in Line 1 he doesnt get what he wants. When one of the nurses offer bowel clamps he rejects them in Line 11. When he asks for a bowel clamp in Line 29 he doesnt get it. A different type of exchange is also used, consisting of a) a request for clarication or conrmation, and b) providing clarication. Three such exchanges are initiated by the circulating nurse and the registrar. First, in Line 12, the circulating nurse asks two is not enough?, and the consultant replies, No is enough but the handles are different. So I cant work in the same way. Second, in Line 16, the registrar asks, Can you use erm the endo grasper?, to which the consultant replies, No I cant. Third, in Line 23, the registrar asks, youve got two yeah, and the consultant replies, Yeah but two different handles. Fourth, in Line 25, the registrar asks, Is it the right length?, to which the consultant does not reply. Two clarication sequences are initiated by the consultant. First, in Line 7, he asks, So why did nobody go and get it? to which no one replies. Second, he asks the scrub nurse if she knows the difference between a bowel clamp and a Johan, and asks her to show him one of each (and she appears to have passed the test). So in a time span of one minute and ten seconds three requests for instruments/actions are successful, three others are unsuccessful, and four requests for clarication/conrmation are made by the circulating nurse and registrar, all of which are negated by the consultant. The unsuccessful exchanges cause increasing agitation on the side of the consultant, as evidenced, e.g. in Line 27 where he intervenes in the registrars rinsing of the camera, requesting him to put the camera back in so he can proceed with the operation, saying Okay Okay. But the unsuccessful exchanges and negated requests for clarication also suggest that the scrub nurse, the circulating nurse and the registrar are all


Jeff Bezemer, Alexandra Cope, Gunther Kress and Roger Kneebone

having difculty understanding what the consultant is up to. So exactly what is the consultant up to, and how much of that is understood by his colleagues? At the start of the episode the surgeon is about to start tearing muscle apart. He needs two graspers for that. He prefers to use two Johans. From the laparoscopic camera record we know that when he is asking for another Johan he is indeed already holding one Johan is his left hand; so the other Johan must be for his right hand. The camera record also shows that at the end of the episode, after hes asked for a Johan again and acknowledged its receipt (thank you, Line 38) he indeed holds two Johans in his hands. That means that there must have been at least two Johans from the start. So why does he not get another Johan when he rst asks for one in Line 1? We do not know what instrument the scrub nurse offers in response to that rst request for another Johan, but we do know that the consultant rejects what is offered and he subsequently appears to start repeating his original request (No. An-[other Johan]). The following turns all seem to suggest that the Johan he wants is not available: The scrub nurse starts accounting for it not being available (I did. I did mention it), relaying responsibility to the circulating nurses to whom she forwarded the consultants announcement right at the start of the operation that he needs two Johans for this procedure. So while there must have been at least two Johans ready to be used, the consultant-surgeon and the scrub nurse are talking about how it could happen that they ended up with one Johan short despite the consultants early announcement. That discrepancy causes confusion among the circulating nurse and the registrar. First the circulating nurse asks, Two is not enough? and second the registrar asks Youve got two yeah?. And in both cases the consultant explains what the problem really is: the handles of the two available Johans are different, and So I cant work in the same way. There are many different types of handles. For the task at hand the handles without locking ratchets seem most apt, as it requires constant opening and closing of the jaw to grasp different bits of the muscle. Handles are delivered separately, and nurses attach them to the various graspers prior to the operation. Thus the confusion about the availability of Johans is based on the ambiguity of the name Johan: as a name for a grasper, and a name for a grasper + handle. It is not clear if the scrub nurse was unable to disambiguate the request for another Johan at the start, but her accounting for why it is not available does suggest that she understood that the consultant didnt just need two Johans, but two Johans with the same handles. When we asked her the following day, her recount of the event did indeed suggest that she is aware of this surgeons preferences for that. However the circulating nurse, and also the registrar, are unable to disambiguate the request, and are left wondering why the consultant does not

Do you have another Johan? Negotiating meaning in the operating theatre


use the two Johans that are available. They also come up with alternatives. The circulating nurse suggests to open the stack for the next case, which happens to be the same procedure (Line 8), and pull out the Johan with the right handle in that pack. The consultant turns that suggestion down as it would create even more problems during the next case. The registrar suggests using an endo grasper apparently missing the point of why the consultant needs two Johans with the same handles. Ultimately the solution to the problem comes from the circulating nurse, who goes to get another Johan with the right handle from the private wing of the hospital. It arrives about 15 minutes after the beginning of the episode. Meanwhile, the consultants agitation has increased as he is struggling to perform the operation without his preferred instruments.

6. Codifying surgical preferences So how could this have happened? Why were there not two Johans with the right handle right at the disposal of the consultant at the start of the operation? The key to this question lies in the way surgeons preferences are codied on their so-called cards and how these codications get translated when instruments are ordered through the hospitals booking system. The theatre nurses in this hospital (and indeed in many other hospitals) keep cards for every consultantsurgeon they work with, detailing, for every procedure which the consultant performs, which instruments he or she needs. The card for the procedure carried out in the episode includes a standard laparoscopic set, which contains a range of instruments commonly used in key hole operations. Additional items are listed separately, such as an item called Frenchie frocep (YOHAN) (sic; see Figure 3). On the back of the typed-out card is a handwritten list, which is probably an updated version. Here it says frenchie, listed under the heading extras. So on the surgeons card the name Johanbecomes frenchie, frenchie forcep (the r in frocep must have been a typo), and Yohan. On neither side is it mentioned that two Johans are required, nor are the handles specied. A senior scrub nurse at our research site explained to us that indeed frenchie and Johan refer to the same forceps. Surgeons cards are indicative of the variety in the way operations are performed. As surgeons work with different trainers in different hospitals, often in different social and cultural contexts, they develop their own professional preferences, from the draping of the patient to the closing of the skin. Out of the innite variety of surgical instruments they will choose to work with specic instruments, and they will choose to refer to those instruments using one of a range of possible names (Johan). Whilst operating, surgeons expect that


Jeff Bezemer, Alexandra Cope, Gunther Kress and Roger Kneebone

Figure 3. Excerpt of the consultant-surgeons preference card

nurses are familiar with these preferences and that they will facilitate their work in line with those preferences. Hence nurses are to make sure that the surgeons preferred instruments will be at their disposal. They also need to know how surgeons actually name their preferred instruments, so that they can pass on the right instrument upon surgeons requests. Nurses are well aware of the variation in surgical practices among the many surgeons they work with. The cards they keep is one way in which local knowledge of surgical practices in a hospital is maintained and passed on to newcomers. Even when these newcomers are highly experienced nurses, they will need to learn about those specics: that Mr Peterson wants an extra Johan grasping forceps when he does a cardiomyotomy, for instance; and that Mr James calls a certain retractor blade a middle blade. They also need to learn the specics of the preferred instruments: how to assemble them or dismantle this particular make of this type of instrument, for instance. The episode shows that this history of collaboration affects not only the moment-by-moment communication between scrub nurse and consultant-surgeon, but all communication in the operating theatre. The scrub nurse in the episode had not worked with the consultant for long; the consultant asked her for her name at the start of the operation. The circulating nurse has a longer history of collaboration with the consultant-surgeon, but the procedure (a cardiomyotomy) is a relatively uncommon one; this consultant probably does no more than 5 such procedures a year. Thus there is very little opportunity for this team of nurses and surgeons to develop a shared language for instruments. This is a major constraint for these teams, and one that breaks with the widespread expectation of surgeons epitomized in the slogan well-know in the surgical community, Give me what I want, not what I asked for. By that they mean that scrub nurses should know exactly what surgeons need, so much so that they should know when to ignore surgeons incidental requests for the wrong instrument. The episode shows that while it is increasingly difcult for nurses and others

Do you have another Johan? Negotiating meaning in the operating theatre


to disambiguate surgeons requests as a result of limited opportunities to work together on the entire range of procedures, the expectation that nurses are aware of all the surgeons idiolects has remained unchanged.

7. Dealing with instability and uncertainty So what can nurses and surgeons do to prevent the type of situation exemplied by the episode, whereby surgeons do not have at their disposal the instruments they prefer to work with and nurses and registrars do not understand exactly what the surgeon is requesting? What can be done to improve the communication between nurses and surgeons? Just before the start of operations staff are now required to jointly go through the Surgical Safety Checklist. The checklist was designed by the World Health Organization and was introduced in England in 2009. One of the questions on the checklist is, Are there any specic equipment requirements or special investigations? And indeed, we have observed consultants asking in that context about the availability of instruments (Youve got the pelvic set?). Better still, when special instruments are required, consultant-surgeons sometimes check if they are available some time (say, one hour) before the operation starts, or they ask their registrar to check this for them. On several occasions we have observed them popping their head around the operating theatre when nurses are preparing for the operation on the patient who is being anaesthetized in the anaesthetic room next door and asking if certain instruments are available. Nurses cannot always conrm that the instruments are available, but they then have at least some time before the operation starts to get the instruments from somewhere. Nurses may also ask surgeons about the stuff they need in advance of the operation. One nurse asked a registrar questions about how he wants things done (Peter, are you going to use. . . ? Peter, how do you want it . . . Now Peter you have your way of . . . how do you . . . ). Another nurse went into the prep room with a consultant asking questions about some of the instruments, which indeed are cutting-edge and not yet widely used. The Surgical Safety Checklist is an institutional response to instability and diversity based on a notion of standardization and homogeneity, like so many of societys responses to multilingualism (cf. Bezemer and Kroon 2008). A similar response is often heard from clinicians when the names of instruments are discussed. As early as 1899, Mr Truax, a US based surgeon, writes,
One object sought in this work is to assist in securing a standard nomenclature for surgical instruments. The custom of calling the same instrument by various names is annoying and confusing. For instance, a periosteal elevator is often referred to or


Jeff Bezemer, Alexandra Cope, Gunther Kress and Roger Kneebone

described as a levator, raspatory, elevator, dry dissector or periosteotome; a plain spring dressing forceps, may be called a thumb forceps, a dissecting forceps, a plain artery forceps, a tissue forceps etc., and even standard text-books sometimes refer to forceps for hemostatic purposes as nippers. (Truax 1899: 8)

More than a 100 years later the clinicians we talked to are still inclined to argue for more standardization to prevent confusion about the names of instruments. Such a language policy seems neither feasible nor desirable. Choosing a standard has, of course, power implications, but it also creates the suggestion that a name for an instrument is meaningful outside its context of use. That was also the underlying assumption of the study we quoted earlier suggesting that theatre staff are better and more consistent in naming instruments than surgeons. Our observations show that, rather, the meaning of instrument names is entirely dependent on shared understandings of the situation and of each other. Instead of aiming for more standardization, which would breach Grices conversational maxim of quantity (imagine the consultant in our example asking for a laparoscopic Johan grasping forceps, 330mm long and 3mm diameter), we would argue for ensuring that surgeons and nurses share a denition of the situation. It should be the responsibility of both the scrub nurse and the consultant to ensure both parties know what is going on, so that the scrub nurse can accurately disambiguate requests. Acknowledgement of diversity and negotiation of meaning including seeking and providing clarication- is key to communication in the operating theatre, while standardization is not workable nor desirable.

8. Conclusion Our analysis shows how changes affecting the communicational landscape throughout society are dealt with at a clinical workplace. Social and economic changes are clearly visible in the operating theatre: continuity in teamwork is increasingly being replaced by ephemerality; social and cultural homogeneity by diversity, division of professions by cross-disciplinary collaboration; and less so in the operating theatre- hierarchical power structures by open, participatory power structures. All of these shifts imply a move away from stability and predictability to instability and provisionality (Kress 2010). The cases we discussed are indicative of attempts to come to grips with these changes. Standardization and codication are responses to diversity which have been well-documented in applied linguistics. Old expectations which characterized an era of stability and continuity are now challengedthat colleagues know each other through

Do you have another Johan? Negotiating meaning in the operating theatre


and through; that they have learned to say and do things in exactly the same way; that the dominating power sets the standard. Now, in an era of instability and diversity, meaning needs to be negotiated in situ, and a shift towards more open, participatory power structures may facilitate that. We hope that our paper offers a sociolinguistically informed analysis of some of these issues which is not only of interest to applied linguists but which is also useful to the clinicians themselves.


We are grateful to all NHS staff who welcomed us in their operating theatres. We also like to thank the funders of our project, Mapping Educational Activity in the Operating Theatre. They are the Royal College of Surgeons, who funded a research fellowship (20092010), and the London Deanery, who granted an award under the Simulation and Technology-Enhanced Learning Initiative (20092011). Several people have, directly or indirectly, provided invaluable feedback on the analysis presented in this paper, including Celia Roberts, Debra Nestel and Gerald Murtagh, and some of the clinicians featuring in the paper.

Arora, S., L. Hull, N. Sevdalis, T. Tierney, D. Nestel, M. Woloshynowych, A. Darzi & R. Kneebone. 2010. Factors compromising safety in surgery: stressful events in the operating room. The American Journal of Surgery 199(1). 6065. Bernstein, Basil. 1971. On the classication and framing of educational knowledge. In M. Youngs (ed.), Knowledge and Control. New Directions for the Sociology of Education, 4769. London: Collier Macmillan. Bezemer, Jeff & Carey Jewitt. 2009. Multimodal Analysis: Key Issues. In Lia Litosseliti (ed.), Research Methods in Linguistics, 180197. London: Continuum. Bezemer, Jeff & Sjaak Kroon. 2008. Teachers Practical Knowledge, Standard Language and Multicultural Classrooms. In Marilyn Martin-Jones, Anne-Marie de Meja and N. H. Hornberger (eds.), Encyclopedia of Language and Education, 2nd Edition, Volume 3: Discourse and Education, 225236. New York: Springer Science + Business Media LCC. Bezemer, Jeff, Gerald Murtagh, Alexandra Cope, Gunther Kress & Roger Kneebone. in press. Scissors, Please: The practical accomplishment of surgical work in the operating theatre. Symbolic Interaction. Bourdieu, Pierre. 1977. The Economics of Linguistic Exchanges. Social Sciences Information 16. 645668.


Jeff Bezemer, Alexandra Cope, Gunther Kress and Roger Kneebone

Bourhis, R.Y., S. Roth & G. MacQueen. 1989. Communication in the hospital setting: A survey of medical and everyday language use amongst patients, nurses and doctors. Social Science & Medicine 28(4). 339346. Bremer, Katharina, Celia Roberts, Marie-Thrse Vasseur, Margaret Simonot & Peter Broeder. 1996. Achieving Understanding: Discourse in Intercultural Encounters. London: Longman. Buchan, J., R. Jobanputra, P. Gough & R. Hutt. 2005. Internationally Recruited Nurses in London. Prole and Implications for Policy. London: Kings Fund. Cicourel, Aaron V. 2003. On contextualizing applied linguistic research in the workplace. Journal of Applied Linguistics 24(3). 360373. Clyne, Michael. 1994. Intercultural Communication at Work. New York: Cambridge University Press. Di Luzio, Aldo, Susanne G nthner & Franca Orletti. 2001. Culture in Communication: u Analyses of Intercultural Situations. Amsterdam: Benjamins. Erickson, Frederick. 1977. Some Approaches to Inquiry in School-Community Ethnography. Anthropology and Education Quarterly 8(3). 5869. Espin, S. L. & L. A. Lingard. 2001. Time as a catalyst for tension in nurse-surgeon communication. Association of Perioperative Registered Nurses Journal 74(5). 672 682. Finn, R. & J. Waring. 2006. Organizational Barriers to Architectural Knowledge and Teamwork in Operating Theatres. Public Money & Management 26(2). 117124. Gjerberg, E. & L. Kjlsrd. 2001. The doctor-nurse relationship: how easy is it to be a female doctor co-operating with a female nurse? Social Science & Medicine 52(2). 189202. Gumperz, John J. 1982. Discourse Strategies. Cambridge: Cambridge University Press. Heath, Christian, Jonathan Hindmarsh & Paul Luff. 2010. Video in Qualitative Research. Analysing Social Interaction in Everyday Life. LA: Sage. Hindmarsh, Jonathan & Alison Pilnick. 2007. Knowing Bodies at Work: Embodiment and Ephemeral Teamwork in Anaesthesia. Organization Studies 28(9). 13951416. Hojat, M., T. J. Nasca, T. J. Nasca, M. J. M. Cohen, S. K. Fields, S. L. Rattner, M. Grifths, D. Ibarra, A. A.-G. de Gonzalez, A. Torres-Ruiz, G. Ibarra & A. Garcia. 2001. Attitudes Toward Physician-Nurse Collaboration: A Cross-Cultural Study of Male and Female Physicians and Nurses in the United States and Mexico. Nursing Research 50(2). 123128. House, Juliane, Gabrielle Kasper & Steven Ross (eds.). 2003. Misunderstanding in Social Life. Discourse Approaches to Problematic Talk. London: Longman. Hutt, R. & J. Buchan. 2005. Trends in Londons NHS Workforce. An Updated Analysis of Key Data. London: Kings Fund. Iedema, Rick, Debbi Long, Rowena Forsyth & Bonne Bonsan Lee. 2006. Visibilising clinical work: Video ethnography in the contemporary hospital. Health Sociology Review 15(2). Jaspaert, Koen & Gert Ramaut. 2000. Dont Use English Words in Dutch. Portrait of a Multilingual Classroom in Flanders. In I. Gogolin and S. Kroon (eds.), Man

Do you have another Johan? Negotiating meaning in the operating theatre


schreibt wie man spricht. Ergebnisse einer international vergleichenden Fallstudie uber Unterricht in vielsprachigen Klassen, 2740. M nster: Waxmann Verlag. u Keddy, B., M. Jones, H. Burton & M. Rogers. 1986. The doctor-nurse relationship: an historical perspective. Journal of Advanced Nursing 11(6). 745753. Kissmann, Ulrike Tikvah (ed.). 2009. Video Interaction Analysis: Methods and Methodology. Frankfurt am Main: Peter Lang. Koschmann, Timothy, Curtis LeBaron, Charles Goodwin & Paul Feltovich. Forthcoming. Can you see the cystic artery yet? A simple matter of trust. Journal of Pragmatics. Kress, Gunther. 2010. Multimodality. A Social Semiotic Approach to Contemporary Communication. London: Routledge. Lingard, L., S. Espin, S. Whyte, G. Regehr, G. R. Baker, R. Reznick, J. Bohnen, B. Orser, D. Doran & E. Grober. 2004. Communication failures in the operating room: an observational classication of recurrent types and effects. Qualilty and Safety in Health Care 13(5). 330334. Lingard, L., R. Reznick, S. Espin, G. Regehr & I. DeVito. 2002. Team communications in the operating room: talk patterns, sites of tension, and implications for novices. Academic Medicine 77(3). 232237. MacKay, R. C., K. Matsuno & J. Mulligan. 1991. Communication problems between doctors and nurses. International Journal of Quality in Health Care 3(1). 1119. Manojlovich, M. & B. DeCicco. 2007. Healthy work environments, nurse-physician communication, and patients outcomes. American Journal of Critcal Care 16(6). 536543. Mondada, Lorenza. 2007. Operating together through videoconference: Members procedures for accomplishing a common space of action. In S. Hester and D. Francis (eds.), Orders of Ordinary Action, 5167. Ashgate: Aldershot. Nagpal, K., A. Vats, K. Ahmed, A. B. Smith, N. Sevdalis, H. Johannsson, C. Vincent & K. Moorthy. 2010. A Systematic Quantitative Assessment of Risks Associated With Poor Communication in Surgical Care. Archives of Surgery 145(6). 582588. Nestel, Debra & Jane Kidd. 2006. Nursesperceptions and experiences of communication in the operating theatre: a focus group interview. BMC Nursing 5(1). Pettinari, Catherine Johnson. 1988. Task, talk, and text in the operating room: a study in medical discourse. Norwood: Ablex Pub. Corp. Rampton, Ben. 2007. Linguistic Ethnography:What andWhy? Unpublished paper, Kings College London. Roberts, Celia, Srikant Sarangi, Lesley Southgate, Richard Wakeford & Vall Wass 2000. Oral examinationsequal opportunities, ethnicity, and fairness in the MRCGP. British Medical Journal (320). 370375. Roberts, Celia, Vall Wass, Roger Jones, Srikant Saranghi & Annie Gillet. 2003. A discourse analysis study of good and poor communication in an OSCE: a proposed new framework for teaching students. Journal of Medical Education (37). 192201. Snelgrove, S. & D. Hughes. 2000. Interprofessional relations between doctors and nurses: perspectives from South Wales. Journal of Advanced Nursing (31). 661667.


Jeff Bezemer, Alexandra Cope, Gunther Kress and Roger Kneebone

Stubbe, Maria, Chris Lane, Jo Hilder, Elaine Vine, Bernadette Vine, Janet Holmes, Meredith Marra & Ann Weatherall. 2003. Multiple Discourse Analyses of a Workplace Interaction. Discourse Studies 5(3). 351388. Svensson, Marcus Sanchez, Christian Heath & Paul Luff. 2007. Instrumental action: the timely exchange of implements during surgical operations. In Liam Bannon, Ina Wagner, Carl Gutwin, Richard Harper, Kjeld Schmidt. (eds.). European Conference on Computer-Supported Cooperative Work. Limerick: Springer. Svensson, Roland. 1996. The interplay between doctors and nurses: a negotiated order perspective. Sociology of Health & Illness 18(3). 379398. Truax, C. 1899. The Mechanics of Surgery. Chicago: Chicago Press. Wass, Val, Celia Roberts, Ron Hoogenbaum, Richard Jones and Cees Van der Vleuten. 2003. Effect of ethnicity on performance in a nal OSCE examination: qualitative and quantitative study. British Medical Journal (326). 800803. Williams, R. G., R. Silverman, C. Schwind, J. B. Fortune, J. Sutyak, K. D. Horvath, E. G. Van Eaton, G. Azzie, J. R. I. Potts, M. Boehler & G. L. Dunnington. 2007. Surgeon Information Transfer and Communication: Factors Affecting Quality and Efciency of Inpatient Care. Annals of Surgery 245(2). 159169. Yeung, T. M., A. C. Cope & S. Appleton. 2008. Thingamajig please sister . . . you know the one I mean! A comparison of knowledge of surgical instrument nomenclature between surgeons and scrub nurses and implications for patient safety. Paper presented at The Society of Academic and Research Surgery Annual Conference, Botanical Gardens, Birmingham, 911 January. Dr Jeff Bezemer <> is a Research Fellow at Imperial College London and Convenor of the UK Linguistic Ethnography Forum. He is interested in learning, professional practice multimodal communication and ethnographic and discourseanalytic research methods. Alexandra Cope <> is a Clinical Research Fellow at Imperial College London and a Specialist Registrar in General Surgery in the Oxford Deanery. She is interested in postgraduate surgical education. Gunther Kress <> is Professor of Semiotics and Education and Director of the Centre for Multimodal Research at the Institute of Education, University of London. He is interested in questions of meaning and its semiotic realizations in interrelation with social and cultural organization. Dr Roger Kneebone <> is Reader of Surgical Education at Imperial College London. His research is focused on simulation and the contextualisation of clinical learning. He has developed innovative approaches to learning and assessing clinical procedures, using hybrid combinations of models and simulated patients.