Linguistic and Cultural Acquisition in a Migrant Community 9781847699909

Table of contents :
Contents
Contributors
Introduction
1. Emigration Versus Mobility. The Case of the Polish Community in France and Ireland
2. The Bookseller and the Basketball Player: Tales from the French Polonia
3. The Positional Distribution of Discourse like – A Case Study of Young Poles in Ireland
4. The Polish Diaspora in Austria: Mutual Attitudes, Processes of Integration and Identity Formation Among Immigrant Poles
5. Divergence, Convergence and Passing for a Native Speaker: Variations in the Use of English by Polish Migrants in Ireland
6. The Direction of Causality in the Relationship Between Phonological Short-term Memory and L2 Lexical Knowledge: The Case of Adult Polish Learners of English in Ireland
7. Segmental Acquisition in Polish Child and Adult Learners in Ireland
8. Variation in English Lexical Acquisition Among Polish Migrant Children in Ireland

Citation preview

Linguistic and Cultural Acquisition in a Migrant Community

SECOND LANGUAGE ACQUISITION Series Editor: Professor David Singleton, Trinity College, Dublin, Ireland This series brings together titles dealing with a variety of aspects of language acquisition and processing in situations where a language or languages other than the native language is involved. Second language is thus interpreted in its broadest possible sense. The volumes included in the series all offer in their different ways, on the one hand, exposition and discussion of empirical findings and, on the other, some degree of theoretical reflection. In this latter connection, no particular theoretical stance is privileged in the series; nor is any relevant perspective – sociolinguistic, psycholinguistic, neurolinguistic, etc. – deemed out of place. The intended readership of the series includes final-year undergraduates working on second language acquisition projects, postgraduate students involved in second language acquisition research, and researchers and teachers in general whose interests include a second language acquisition component. Full details of all the books in this series and of all our other publications can be found on http://www.multilingual-matters.com, or by writing to Multilingual Matters, St Nicholas House, 31–34 High Street, Bristol BS1 2AW, UK.

Linguistic and Cultural Acquisition in a Migrant Community

Edited by David Singleton, Vera Regan and Ewelina Debaene

MULTILINGUAL MATTERS Bristol • Buffalo • Toronto

Library of Congress Cataloging in Publication Data A catalog record for this book is available from the Library of Congress. Linguistic and Cultural Acquisition in a Migrant Community/Edited by David Singleton, Vera Regan and Ewelina Debaene. Second Language Acquisition: 69 Includes bibliographical references. 1. English language—Variation—English-speaking countries. 2. Second language acquisition—Ireland. 3. Second language acquisition—France. 4. Second language acquisition— Austria. 5. Language and culture—Ireland. 6. Language and culture—France. 7. Language and culture—Austria. 8. Polish people—Ireland—Social life and customs. 9. Polish people—France—Social life and customs. 10. Polish people—Austria—Social life and customs. I. Singleton, D. M. (David Michael) editor of compilation. PE2751.L56 2013 427–dc23 2013011820 British Library Cataloguing in Publication Data A catalogue entry for this book is available from the British Library. ISBN-13: 978-1-84769-989-3 (hbk) Multilingual Matters UK: St Nicholas House, 31–34 High Street, Bristol BS1 2AW, UK. USA: UTP, 2250 Military Road, Tonawanda, NY 14150, USA. Canada: UTP, 5201 Dufferin Street, North York, Ontario M3H 5T8, Canada. Copyright © 2013 David Singleton, Vera Regan, Ewelina Debaene and the authors of individual chapters. All rights reserved. No part of this work may be reproduced in any form or by any means without permission in writing from the publisher. The policy of Multilingual Matters/Channel View Publications is to use papers that are natural, renewable and recyclable products, made from wood grown in sustainable forests. In the manufacturing process of our books, and to further support our policy, preference is given to printers that have FSC and PEFC Chain of Custody certification. The FSC and/or PEFC logos will appear on those books where full certification has been granted to the printer concerned. Typeset by Techset Composition India (P) Ltd., Bangalore and Chennai, India. Printed and bound in Great Britain by Short Run Press Ltd.

Contents

Contributors Introduction David Singleton and Vera Regan

vii xi

1

Emigration Versus Mobility. The Case of the Polish Community in France and Ireland Ewelina Debaene

2

The Bookseller and the Basketball Player: Tales from the French Polonia Vera Regan

28

3

The Positional Distribution of Discourse like – A Case Study of Young Poles in Ireland Niamh Nestor

49

4

The Polish Diaspora in Austria: Mutual Attitudes, Processes of Integration and Identity Formation Among Immigrant Poles B. Bidzin´ska

73

5

Divergence, Convergence and Passing for a Native Speaker: Variations in the Use of English by Polish Migrants in Ireland Ewelina Debaene and John Harris

85

6

The Direction of Causality in the Relationship Between Phonological Short-term Memory and L2 Lexical Knowledge: The Case of Adult Polish Learners of English in Ireland Agnieszka Skrzypek

v

1

106

vi

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

7

Segmental Acquisition in Polish Child and Adult Learners in Ireland Romana Kopecˇková

134

8

Variation in English Lexical Acquisition Among Polish Migrant Children in Ireland Polly Walsh and David Singleton

152

Contributors

Barbara Ewa Bidzin′ ska was born in Warsaw, Poland. She holds a PhD in Applied Linguistics from Trinity College Dublin. Her research interests include language transfer in second/third language acquisition, language attrition, native language maintenance, and social psychology of inter-ethnic relations. She also translates from Polish into English and German. Ewelina Debaene holds a PhD in English Applied Linguistics from Warsaw University (2005). In 2005 Ewelina obtained teaching qualifications relative to Polish as second/foreign language and subsequently commenced her duties as a Lecturer in Polish Language and Culture at Trinity College Dublin. Between 2007 and 2010 she was involved as Associate Investigator in the Polish Diaspora Project run under the auspices of the Irish Council for the Humanities and Social Sciences. She is currently delivering a series of lectures at Marii Curie-Skłodowska University and the Medical University of Lublin. Ewelina’s research and teaching interests revolve around English language teaching methodology, English for Specific Purposes, the notion of cultural stereotypes and sociocultural aspects of migration. She is currently working on a book dealing with Polish migration. John Harris is a Fellow Emeritus of Trinity College Dublin where he has been Associate Professor of Psycholinguistics and Head of Department in the Centre for Language and Communication Studies. His research has focussed on the teaching and learning of Irish and modern languages, and Irish language revitalisation. In recent years he edited a special issue of the International Journal of Bilingual Education and Bilingualism on the Irish Language, and published papers in Language Teaching Research and AILA Review. He is Executive Editor of Language, Culture and Curriculum (Routledge Taylor and Francis). vii

viii

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Romana Kopečková is a visiting DAAD scholar at the English Department of Westfälische Wilhelms-Universität Münster, Germany, where she is conducting post-doctoral research into cross-linguistic influence in third/additional phonological development. She has published on age-related differences in second language speech perception and production. Her most recent research interest includes the role of prosody in spoken academic discourse. Niamh Nestor is a PhD student in Sociolinguistics at University College Dublin, Dublin, Ireland. Her research explores language variation, identity, and the acquisition of sociolinguistic competence among a group of Polish children and teenagers living in Ireland. Recent publications include: (2011) The new kid on the block. A case study of young Poles, language and identity. In M. Darmody, N. Tyrrell & S. Song (eds) The Changing Faces of Ireland: Exploring the Lives of Immigrant and Ethnic Minority Children (pp. 35–52) (with V. Regan); and (2009) Out of the Communist Frying Pan and into the EU Fire? Exploring the Case of Kashubian. Language, Culture and Curriculum 22 (2), 95–119 (with T. Hickey). For five years Niamh lived and worked in Poland as a TEFL teacher in both public and private schools. She was also a full-time teacher at a teacher training college. Vera Regan has a PhD in linguistics from the Université d’Aix en Provence and is Associate Professor of Sociolinguistics at University College Dublin. She is Chevalier de l’ordre des Palmes Académiques and has served as President of the European Second Language, President of the Association for French Language Association and President of the Association of Canadian Studies in Ireland. She has held Visiting Professorships at the University of Pennsylvania and the University of Ottawa. She has published widely on variation theory and second language acquisition. David Singleton took his BA at Trinity College, University of Dublin, and his PhD at the University of Cambridge. He is a Fellow of Trinity College Dublin, where he is Professor of Applied Linguistics. He also holds the title of Research Professor at the University of Pannonia. He has served as President of the Irish Association for Applied Linguistics, as Secretary General of the International Association of Applied Linguistics and as President of the European Second Language Association. His publications range across a wide spectrum of topics – including syllabus design and pedagogical grammar – but currently focus mainly on cross-linguistic influence, the second language lexicon, the age factor in language acquisition and multilingualism.

Contr ibutors

ix

Agnieszka Skrzypek holds a PhD in Psycholinguistics from Trinity College Dublin. She is currently relocating from the University of Limerick to the University of Aberdeen to take up the post of a Lecturer in Linguistics. She has taught on Applied Linguistics and TESOL programmes in a number of third-level institutions in Ireland and Poland. Additionally, she has considerable expertise in statistical research methods. Her research interests include, inter alia, the role of short-term memory in L2 learning and L2 processing, computerized working memory training, cognitive linguistics and educational psychology. She has published papers in international journals and edited volumes. Polly Walsh is a Speech and Language Therapist and a graduate of Trinity College Dublin. She has over ten years’ experience working with children and adults with intellectual disability. Since 2008, she has worked for St. John of God Community Services. Her research interests include early childhood intervention and intellectual disability, alternative and augmentative communication, and bilingualism and second language acquisition. Her most recent work has involved developing accessible policy documents for people with intellectual disability.

Introduction David Singleton and Vera Regan

The chapters in this book are all based on research conducted under the broad auspices of a major project funded during the period 2007–2010 by the Irish Research Council for the Humanities under the title ‘Second language acquisition and native language maintenance in the Polish diaspora in Ireland and France’. The influx of Polish migrants into Ireland had not, before the commencement of the project, been the subject of substantive research from a linguistico-cultural perspective. Our multi-disciplinary, comparative project sought to make good this deficit with respect to the acquisition and use of the languages of the host community and to the transmission of the first language of migrants to their children. The project was designed to yield both sociolinguistic and psycholinguistic findings and, also, to incorporate sociocultural and educational dimensions. A parallel investigation of the Polish community in France was carried out to identify language similarities and differences between the two communities. The point of this comparison was that Polish people had been migrating to France over a very long period. Notwithstanding the divergences between circumstances under which the two groups migrated, the language behaviour patterns of Poles who settled in France would, in our view, provide indicators regarding the linguistic future of Poles in Ireland. Given there had been no attempts in linguistic studies so far to compare Polish migrant communities in two different European receiving countries, we saw our project as constituting an original contribution to the area of sociolinguistic research and migration studies. The project proceeded under the general supervision of the three editors of this volume: David Singleton, Vera Regan and Ewelina Debaene. A further colleague, John Harris, informally participated in the project on a voluntary basis. Three postgraduate researchers were funded by the project, Romana Kopečková, Niamh Nestor and Agnieszka Skrzypek, and Barbara Bidz´inska and Polly xi

xii

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Walsh, who were not financially supported by the project worked within its general ambit. Barbara Bidzin´ska worked with Polish migrants in Austria, thus extending the geographical range of our investigation, and Polly Walsh worked with very young children, extending in similar fashion the range of ages covered by our activities. All of the above figure in the pages that follow. It is perhaps worth emphasizing the fact that we endeavoured throughout to render our data-collection procedures as user-friendly as possible, offering participants, for instance, the possibility of responding either in Polish or in a language of their host country (English/French). Participants were consistently thanked – verbally and/or in writing – for their contribution to the project and were made aware of the project website. Moreover, in connection with our recruitment and data-collection activities in Dublin, we offered a much-appreciated free six-month course in English to a large number of our adult participants. A further point worth recalling is that all the non-Polish researchers working on the project acquired some knowledge of the Polish language, have participated in Polish cultural life in Ireland, and visited Poland. We believe that our efforts in all of the above respects had a very positive impact on the way in which our project was looked upon and responded to by the participants. Sociolinguistic interviews were conducted in both France and Ireland from the start of the project until the end of its funding period, by, and under the supervision of, the two Associate Investigators, Vera Regan and Ewelina Debaene. Data collection involving a general socio-biographical profiling also took place in the context of the other individual projects. Agnieszka Skrzypek designed the language elicitation associated with the above-mentioned English language course. Language elicitation was also carried out by Romana Kopečková. This particular strand of language data collection project involved testing Polish participants on their perception and production of Irish-English vowel sounds. The contributions to the present volume divide largely into those which proceed from a sociolinguistic perspective and those from a psycholinguistic perspective. The first five contributions all deal with social aspects of second language (L2) acquisition and maintenance, of which the first two deal with the acquisition of French in France by Polish migrants, two address the acquisition of English in Ireland by Polish migrants, and the remaining chapter is concerned with the experiential and acquisitional profile of Polish migrants in Austria. The last three papers all have a more psycholinguistic orientation and all deal with Poles’ acquisition of English as an L2 in Ireland. The more sociolinguistically oriented chapters by Debaene, Debaene and Harris, Nestor, and Regan all treat of the area of SLA and identity, which has recently become an extremely productive area of inquiry (see, for

Introduc t ion

xiii

example, Regan, 2013). Debaene’s chapter is a comparative, descriptive look at the Polish diaspora which formed in France in the 1980s and the recent, post-EU-accession Polish community in Ireland. It examines the two migration patterns: (1) migration for settlement and (2) mobility within the EU. Various circumstances under which the two groups migrated, their respective concerns at various stages of their stay abroad, the coping strategies used to minimize stressful situations, the language-learning strategies deployed, as well as degree of maintenance of native language and culture are reported. The research methods described include interviews and media coverage analyses, conducted both in France and Ireland. The factors which emerge from these preliminary results as most influential on the process of integration, have to do with, inter alia: reasons for migrating, distribution and number of migrants in the host country, socio-economic status, length of residence, frequency of visits to the home country, and institutional representation. Regan’s contribution uses a mixed method approach to investigate the use of vernacular French by Polish speakers living in France. It combines a quantitative, variationist analysis of a frequently occurring variable in French, the deletion and retention of ne, the first particle of negation, with a qualitative analysis which further explores the implications of the quantitative results. The computer multivariate analysis calculates the probability of multiple factors on the use of the vernacular variant. It was hypothesized that those who lived longer in France were likely to use more vernacular variants. The factors which were found to be most significant included length of residence (LOR) in France. Whereas, for the group as a whole, the longest LOR correlates with the highest deletion rates, the figures for two individuals run counter to this direction. At the two ends of the continuum, we see that one participant (the ‘bookseller’, Mariusz), a high-proficiency speaker with the longest LOR, has, surprisingly, the lowest deletion rate; on the other hand, the ‘basketball player’ (Anna), with a lower LOR, deletes at a very high rate. Anna’s high deletion rate is all the more interesting as women tend to produce lower rates of vernacular variants. The intriguing case of these two speakers prompted a further investigation into the trajectories of their lives through qualitative analysis of these lives and life choices. The two speakers are positioned in opposite directions in space and time. Mariusz orients himself towards Poland and its cultural and literary past. Anna orients herself towards contemporary France and the future. Combining quantitative and qualitative data shows how individuals can contrast with the background of broader social structural categories, depending on the stances they take. These two contrasting speakers demonstrate that, while overall categories are a useful predictor of patterns of L2 acquisition, individuals are more

xiv

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

than simply exemplars of such overall categories. Their personal biographies and current relationships with others in their society lead to significant variations from overall patterns. Nestor’s chapter focuses on the use of discourse markers in the speech of adolescent Poles in Ireland. The impact of Ireland’s recent population change has been particularly striking in the pupil profiles of Irish schools. At present, approximately 12% of primary school pupils (Naughton et al., 2010) and 9% of post-primary school students were born outside of Ireland (Department of Education and Skills and the Office of the Minister for Integration, 2010). There continues to be a need for improved policymaking to be better informed by more evidence-based academic research, with particular reference to enhanced social cohesion and inclusion policies. This is particularly necessary given the fact that non-Irish-born children represent a significant percentage of the school population. Equally, the question of integration into the host community continues to be pertinent. One indicator of integration is language. Sankoff et al. (1997) claim that the fluency with which a speaker uses discourse markers is an indication of how integrated he/she is into the local speech community. This chapter analyses the discourse marker like and its use by young Polish L2 speakers of English. The analysis shows that this discourse marker is a salient feature of the speech of many of the speakers observed and that such speakers follow Irish English patterns, but that there is a high degree of interspeaker variation. A more detailed analysis elucidates notable differences between speakers and indicates that gender and location may have a potentially significant impact. The Debaene and Harris chapter explores the relationship between attitudes and identity on the one hand and situationally related variations in language use by non-native speakers on the other. This joins the current body of work on identity and Polish L2 speakers of Irish English, see (Nestor & Regan 2011; Regan, in press; Regan et al., 2013). The subjects of the study in this volume are 25 Polish nationals who came to Ireland after the accession of Poland to the EU in May 2004 and have lived in their adopted country for at least three years. Specific attention is given to the phenomena of convergence (Pardo, 2006) and accommodation (Giles et al., 1991). The authors explore the extent to which Polish non-native speakers of English in Ireland aim for, or believe they approximate, native speaker performance in English in different settings (or aspire ultimately to pass for native speakers of English). The speakers’ own views on these issues are provided and their assessment of the importance of such issues in relation to identity, the maintenance of solidarity with Polish friends and groups, and their integration into Irish society generally. The mixed-method

Introduc t ion

xv

approach deployed includes the use of questionnaires, face-to-face interviews, and focus group meetings. Issues include: •

• •

How determined are immigrant Polish people’s attempts to acquire native (Irish) English speaker norms and performance in relation to accent, linguistic structure and content, non-linguistic interaction, register, idiom and communication generally? How are interactions in English different in terms of convergence and accommodation depending on whether the participants consist of nativenative, native/non-native or non-native/non-native groups or pairs? How Polish speakers’ accentedness in English limits their ability to adapt their speech to native pronunciation? How are convergence and accommodation related to issues of identity?

Bidzin´ska’s chapter analyses the situation of the Polish diaspora in Austria including mutual attitudes, processes of integration and identity formation among immigrant Poles. The study analyses the complex immigrant experiences of Poles in Austria and the impact of those experiences on identity formation of members of the Polish diaspora. The study starts from the premise that one cannot fully understand the complexities of integration without prior knowledge of historical background and relations between the receiving society and particular ethnic groups. Therefore, the author familiarizes the reader with the various contexts of the Polish immigration to Austria. The article has been divided into two parts: theoretical considerations, and empirical study. First, the reader is encouraged to focus on the concept of identity and to consider factors which play an active role in its formation – both on the individual and societal level. Second, the reader’s attention is drawn to the concept of Polish identity: what does it mean to be Polish? In this section of the article, different directions in Polish political thought, attachment to the Slavic heritage, strong links to the Catholic Church, and the concept of Polish martyrdom. Third, immigrant policies in Austria are discussed. However, the author warns that the reader should remember that the mentioned regulations were effective at the time of data collection – some laws may have already changed or be amended in the future. The second part of the article presents the participants of the study. This part has also been divided into subsections and includes the research questions which guided the study, tabular presentation of data, and a short discussion. In the last section of the article some directions for future research in the field are suggested. Psycholinguistic issues are the focus of the chapters by Skrzypek, Kopečková, and Walsh and Singleton. The topics of phonological short-term

xvi

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

memory, phonological approximation, the age factor and lexical acquisition are leitmotivs in this part of the book, resonating with recent and forthcoming publications by the authors in question such as Muñoz and Singleton (2011), Singleton and Kopečková (in press) and Skrzypek and Singleton (2013). Skrzypek’s contribution deals with phonological short-term memory and the L2 learner, particularly in relation to vocabulary acquisition and in relation to adult learners. In the last two decades the relationship between the rate of vocabulary learning and the ability to repeat nonwords (phonotactically licit made-up words that do not exist in the language on which they are based) has been the focus of intense research. Significant positive correlations between scores representing vocabulary knowledge and scores representing the ability to repeat nonwords or lists of nonwords have been reported (for an overview see e.g. Baddeley, 2007; Gathercole, 2006a, 2006b). As correlations are not in themselves indicative of the direction of causality, this relationship can be interpreted as either supporting the claim that phonological short-term memory (conceptualized as immediate memory for verbal wordlike material tapped by nonword tasks) causally affects vocabulary learning or that the reverse holds true. Skrzypek’s paper examines this issue in the context of adult L2 learning using the cross-lagged correlational paradigm. In Kopečková’s chapter, second language speech learning among Polish child and adult migrants living in Ireland is examined. Polish children participating in her study perceived similarities between native and non-native vowel sounds differently from the manner in which Polish adults did, which, in turn, predicted and explained their superior L2 production abilities. The author, therefore, argues that perception of cross-language phonetic similarity may play an important role in how accurately non-native sounds are acquired. Other factors, such as native language phonology and quality and quantity of L2 experience, are also discussed in the study as determining the extent to which non-native sounds can be produced with native-like accuracy by learners of diverse age groups. Walsh and Singleton report an investigation of the vocabulary development of Polish children and Irish children of the same age over a six-month period. The Polish children’s English vocabulary levels and patterns of lexical development were compared to those of their Irish peers and also across the Polish group. The gap which emerged between the Polish and Irish children in terms of English vocabulary size was found to be rapidly diminishing - to the extent that on the second testing occasion some of the Polish children’s scores were within the native-speaker range. The results also suggest, unsurprisingly, that the Polish children know less about the English words than

Introduc t ion

xvii

their Irish peers. The results indicate, too, wide variation within the Polish group, and point to length of residence and of schooling in Ireland, amount of everyday exposure to English and parental expertise in English as sources of individual differences. Altogether the volume provides a linguistic and cultural profile of the Polish diasporic communities in Ireland, France and Austria. The eight studies present original research on the acquisition and use of the languages as well as aspects of cultural acquisition in those host communities. The multiple research approaches used have provided an unusually rich and comprehensive picture of human migrant experience and can suggest further research in the area of migration and language.

References Baddeley, A.D. (2007) Working memory, thought, and action. Oxford: Oxford University Press. Department of Education and Skills and the Office of the Minister for Integration. (2010) Intercultural Education Strategy, 2010–2015. Online document: http://www.education. ie/servlet/blobservlet/mig_intercultural_education_strategy.pdf. Department of Social Protection, Ireland. Online document. Gathercole, S.E. (2006a) Nonword repetition and word learning: The nature of the relationship. Applied Psycholinguistics, 27 (4), 513–543. Gathercole, S.E. (2006b) Complexities and constraints in nonword repetition and word learning. Applied Psycholinguistics 27 (4), 599–613. Giles, H., Taylor, D.M. & Bourhis, R.Y. (1973) Toward a theory of interpersonal accommodation through speech: Some Canadian data. Language in Society 2, 177–192. Muñoz, C. and Singleton, S. (2011) A critical review of age-related research on L2 ultimate attainment. Language Teaching 44, 1–35. Nestor, N. and V. Regan (2011) New kids on the block: A case study of young Poles, language, and identity. In M. Darmody, N. Tyrrell and S. Song (eds) The Changing Faces of Ireland: Exploring the Lives of Immigrant and Ethnic Minority Children (pp. 35–52). Rotterdam/Boston/Taipei: Sense Publishers. Naughton, B., O’Mahony, A. and Miley, S. (2011) How to combat primary school segregation in a multicultural world: Ireland’s experience. In J. Bakker, E. Denessen, D. Peters and G. Walraven (eds) International Perspectives on Countering School Segregation (pp. 33–49). Antwerp, Apeldoorn: Garant. Pardo, J.S. (2006) On phonetic convergence during conversationalinteraction. Journal of the Acoustical Society of America 119, 2382–2393. Regan, V. (2013) Variation and second language acquisition. In J. Herschenson and M. Young-Scholten (eds) The Cambridge Handbook of Second Language Acquisition. Cambridge: Cambridge University Press. Regan, V. (in press) Tales from the Celtic Tiger: Polish migrants, language and identity. In M. Kelleher and M. Kenneally (eds) Irishness and Québécité: New Perspectives. Montreal: McGill University Press. Regan, V. et al. (2013) Discourse like and social identity – a case study of Poles in Ireland. In B. Migge and M. Ni Chiosain (eds) New Perspectives on Irish English. Amsterdam: John Benjamins.

xviii

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Sankoff, G., Thibault, P., Nagy, N., Blondeau, H., Fonollosa, M-O and Gagnon, L. (1997) Variation in the use of discourse markers in a language contact situation. Language Variation and Change 9, 191–218. Singleton, D. and Kopečková, R. (in press) Second language phonology: a critical perspective on critical period perspectives. In A. Berndt (ed.) Fremdsprachen in der Perspektive lebenslangen Lernens. Bern: Peter Lang, in press. Skrzypek A. and Singleton, D. (2013) Controlled versus free production of English collocations by adult Polish learners: The role of short-term memory. Vigo International Journal of Applied Linguistics 10, 105–129.

1 Emigration Versus Mobility. The Case of the Polish Community in France and Ireland1 Ewelina Debaene

A comparative study of the Polish Diaspora, which formed in France in the 1980s, and the recent post-accession Polish community in Ireland has been undertaken to examine the two migration patterns: (a) migration for settlement and (b) mobility within the EU. Various circumstances under which the two groups migrated, their respective concerns at various stages of their stay abroad, the coping strategies used to minimise stressful situations, language-learning strategies, as well as maintenance of native language and culture have been analysed. The research methods to date have included interviews and media coverage analyses, conducted both in France and Ireland. The preliminary results have indicated factors that are most influential on the process of integration, among others: reasons to migrate, distribution and number of migrants in the host country, socioeconomic status, length of residence, visits to the home country and institutional representation.

Introduction The point of departure for this comparative study is to examine the experience of a well-established Polish Community in France with a view to gaining a better understanding of the processes involved in the current migration of Polish people to Ireland. The two groups being compared are (a) those 1

2

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Poles who came to France in the 1980s, as part of the ‘Solidarity generation’ and chose to leave Poland during a time of political upheaval, economic depression and Communist rule and (b) Poles who came to Ireland after the accession of Poland to the European Union in May 2004. In the following parts of this paper, primary attention will be given to the findings of an interview-based study undertaken among members of the Polish Community, which formed in France during the period of Solidarity (the 1980s). Insights from interviews with representatives of various Polish organisations in France and Polish people who have lived in France for 20 years or more, worked, and established families there, will enable us to examine the process of integration. Against a backdrop of the history of Polish migration to France, major factors which influenced the policy and strategies for migration implemented in France in the 1980s will be discussed. In its final parts, the paper will discuss insights from interviews conducted over the last two years among members of the Polish Community in Ireland. Based on this comparison, factors which are most influential on the routes of integration and decisive in the choice of strategy for migration will be delineated.

Polish Migration to France: Historical Background Emigration has been a staple feature of the Polish societal landscape for at least 200 years. Poland was virtually wiped off the map of Europe for a period of 123 years, from 1795 when it was partitioned between three neighbouring countries until the year 1918, after which it finally regained its independence. It is not an overstatement to say that the most valuable contribution to Polish cultural heritage (artistic, educational, literary) was made outside the Polish borders. As a result of long-lasting bonds between Poland and France, Poles have long been migrating to France and establishing political and cultural centres there, most of which are still operating today. Numerous places, buildings and statues in France still bear witness to the Polish legacy and Cité Nationale de l’Histoire de l’Immigration (Immigration Museum opened in Paris in October 2007) devoted a substantial gallery space to materials related to the history of Polish immigration in France. It is estimated that French ‘Polonia’, as the Polish Community outside the Polish borders is referred to, numbers 800,000 or even 1 million, and in Europe it is outnumbered only by the Polish Community in Germany (Judycki and Judycka, 1996: 18; Nowa Encyklopedia Powszechna PWN, 1995: 430). Five main periods of Polish migration to France can be distinguished.

Emigrat ion Versus Mobilit y

3

(1) The Great Emigration – the wave of the Polish refugees between 1831 and 1835. The Poles were migrating then, fleeting repression associated with the partition and exacerbated after the collapse of the November Rising in 1831. The name ‘great’ refers rather to the fact that this was considered an emigration of political elites, and is not reflected in numbers, estimated at approximately 5000–10,000 (e.g. Dybkowska et al., 2002: 179; Geremek & Frybes, 2004: 115). (2) The interbellum period, called otherwise the Second Great Emigration – the influx of Polish workers after the immigration convention, which regulated the employment of foreign workers in industry and agriculture, was signed in 1919. The Poles, coming primarily from poverty-stricken rural and mining areas in Poland, established a large community in France, mainly in its northern, industrial regions, which in 1931 numbered 507,811 (Jendrowiak, 2002: 45; Schnapper et al., 2003: 19; Ponty, 2004: 126). Push and pull mechanisms were primarily economic in this period. The majority of newcomers found employment in mining and agriculture. Slag heaps still bear witness to the long-lasting traditions of the profession dominated by the Polish community there. (3) The post-1968 emigration, caused by political upheavals and the antiSemitic policy in Poland. The total number of Polish citizens (mainly of Jewish origins) who left Poland in result of anti-Semitic policy is estimated 12,000–20,000 (without breakdown into destination countries) (cf. Dybkowska, 1996). (4) The post-1980 (between 1980–1989) emigration – the main focus of our research –frequently referred to as ‘Solidarity migration’ (see following section on the Polish Community in France). (5) The year 1989 and after – emigration which took place after the Round Table talks and the collapse of the Iron Curtain. Free movement of people across the borders facilitated migration, which was primarily economically motivated. Poles who came to France after 1989 in a majority of cases intended to stay for a couple of years, save up enough to invest in Poland and then to return (cf. Hładkiewicz, 2002). Following Poland’s accession to the EU in 2004, the perspective of the influx of Polish migrant workers into France was subject to a heated debate, popularly associated with the ‘Polish plumbers’. In fact, the numbers were much less significant than predicted, and the European Economic Community (EEC) workforce now represents less than 1% of the EU’s active population (Michel, 2007). As our research so far has focused on the Polish Community which formed in France in the 1980s, it will be treated in more detail below.

4

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

The Polish Community in France Formed After 1980 The so-called ‘Solidarity Emigration’ reached a peak during and in the aftermath of the declaration of Martial Law (1981–1983), when the ruling Communist party in Poland allowed one-way cross-border movement in an attempt to eliminate the opposition leaders. In terms of factors which influenced the decision to emigrate, this outflow of Polish people is a complex phenomenon and by no means easy to classify. Despite the common notion that it was made up of the anti-communist activists, fleeing political repression (e.g. Z˙ alnin´ski, 2005), economic factors also played a non-negligible role, as many historians agree (e.g. Habielski, 1995: 37). Polish people who came to France before 1989 usually intended to stay permanently. Table 1.1 indicates the number Polish nationals who were granted stay permits in France between 1 January 1980 and 31 December 1989. Data are obtained from the French Ministry of Internal Affairs. Poles who arrived in France before the collapse of communism in 1989 were effectively granted political asylum, work permits and, in a matter of 3 years, French citizenship. The procedure for granting political refugee status was handled by OFPRA, L’Office français de protection des réfugiés et apatrides. Especially during the times of Martial Law, the majority of applications were approved, as follows: in 1981: 68,4%, of approved applications in 1982: 93.1%; in 1983: 82.3% (source: OFPRA). An overview of literature of the subject, further substantiated in the course of interviews with employees of the Polish institutions in France, suggests that no reliable data exist regarding distribution of Polish people in France, e.g. in urban and rural areas (Representative of the Polish Consulate in Paris, personal communication, May, 2007). In the same vein, Schnapper et al. remark that foreigners often disappear from the French statistics as ‘the French pattern of integration is based on the idea of the legal transformation Table 1.1 Polish nationals granted permits to stay in France 1980–1989 Year:

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

No. of Polish 73,628 74,173 72,947 71,456 70,205 68,817 Data 61,528 57,399 56,505 nationals unavailable granted stay permit in France: Source: AGDREF: Infocentre

Emigrat ion Versus Mobilit y

5

of immigrants and their children into French citizens’ and it is a very difficult task to obtain statistical data, especially as concerns second-generation migrants (Schnapper et al., 2003: 18). According to our informants, the majority of Polish people who came to France in the 1980s settled in the Parisian region. Paris offered the best chances for accommodation and employment, whilereas Polish organisations and support centres, operating mainly in the capital, facilitated the initial period in France (Representative of the Polish Catholic Mission, personal communication, May, 2007). Unlike the early decades of the 20th century, when miners constituted almost 80% of the Polish Community in France, it is impossible to talk of dominant professional sectors in the Polish Community formed in the period of Solidarity. It is a common notion, however, that this group of migrants consisted of members of well-educated groups (‘inteligencja’). After the initial stage when, hampered by limited language skills, they often had to accept jobs below their qualifications (as au-pairs or manual workers), the newcomers were later employed in a vast range of professions.

French Policy Towards Immigration Even though the French debate on immigration has for the most part focused on ‘non-European’ migrants, some of its implications have also impacted on the integration path pursued by the Polish Community. In France, an immigration country for over a century and a half, there is a strong tradition of an ‘assimilation’ policy. As Kastroyano observes, ‘[t]he [. . .] French model, based on republican individualism, implies and entails the assimilation of individuals who have become citizens by choice’ (Kastroyano, 2002: 2). Transforming migrants into French citizens, as opposed to promoting a regional identity, has been the consistent integration objective and pattern (cf. Schnapper et al., 2003: 15) Assimilation in France leads to ‘the disappearance of foreignness’, according to Costa-Lascoux (1989: 10). In consequence, ethnic communities would receive little recognition in the public domain, and be most likely associated with the private sphere. The migrants would therefore be expected to internalise values and socialise in French society, while, only if they desired to do so, preserving their culture of origin by means of a ‘cultural DIY’ process (cf. Schnapper et al., 2003: 16). As was seen in the ‘nationality laws’, otherwise referred to as ‘Frenchification’, education policy and laws binding in the workplace, a strong tendency would be to secure equal rights for migrant communities and their speedy assimilation (Schnapper et al., 2003: 21–26).

6

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

In the process of translating ideology into practice, despite the universalistic principles put in place, France has been long experiencing difficulties with suburban ghettos (banlieus), ethnic enclaves and inner-city slums, places that combine foreignness and poverty and are seen as sites of conflict between cultures, and ‘between the communities and the nation’ (Kastroyano, 2002: 3). Despite the existence of special programs for groups that are excluded from the process of assimilation (Schnapper et al., 2003: 28-37), the terms ‘exclusion’, ‘ghettoisation’, and ‘ethnicisation’ underline the French debate on immigration (Kastroyano, 2002: 28). It has to be emphasised, however, that such issues would not typically emerge with regard to Polish migrants (Catholic, white and enjoying a general reputation as a ‘hard-working’ community) who would typically be viewed as ‘exemplary, model migrants’. For a long time multiculturalism, understood as presence and state of both cultural and ethnic diversity within the demographics of a particular social space (Parekh, 2006: 3), has been in conflict with ‘political tradition’ and ‘national myth’, whereas the minority-based approach was supposed to weaken the social fabric and cohesion (cf. Schnapper et al., 2003: 42). In the last decades, however, there has been much debate in France, notably inspired by controversies such as l’affaire des foulards, starting in 1989, about the French assimilationist model. A number of key commissions and reports (e.g. the Haut Conseil à l’intégration and the Stasi Commission) have considered such matters. The substantial financial support for such agencies as the Le Fonds d’Action et de Soutien pour l’Intégration et la Lutte Contre les Discriminations also suggests that there has been some development in French thinking. It would be fair to say that French approaches to integration have been subtly modified in important respects although the underlying perspective remains one of assimilation (Overview of integration policy in France 2002–2005, p. 9). Below the experience of the Polish Community members in France in the 1980s will be presented, based upon preliminary research findings, in the wider context of French immigration policy presented above.

Research Design, Preliminary Findings Sampling and timing Subject recruitment in France started in May 2007 in the region of Lille, Dunkirk and Paris. Our Informants, interviewed to date, can be divided into two groups. Group 1 encompassed 10 Polish married couples with children who migrated to France between 1982 and 1990. Subject recruitment started in

Emigrat ion Versus Mobilit y

7

May 2007 in the region ‘Le Nord’ (four couples) and Paris (six couples). The subjects were aged 46–62 and were sampled from various professional settings. Most of the subjects in this group were found thanks to the network of family and personal contacts. Four out of 20 subjects were first met via Polish organisations in France (e.g. The Polish Catholic Mission). It needs to be strongly emphasised, however, that an attempt was made in the course of subject recruitment procedure to avoid a self-selecting sampling. Hence, there was only partial reliance upon Polish institutions as a source of contacts, and random selection was preferred. Our subjects varied extensively with regard to their level of education, family situation (i.e. marital status, number of children), reasons to emigrate, socio-economic status, level of language proficiency in French, exposure to and reliance on the Polish language, degree of contact with the host country, and degree of contact with Polish centres in France. In parallel, employees of various Polish organisations (educational, diplomatic, cultural and others) were interviewed (group 2). In the absence of written sources or reliable statistical data, personal communication was an all the more valuable method of inquiry. Fifteen employees of Polish institutions and cultural centres in Lille, Dunkirk, Lens and Paris have been interviewed to date. These participants come from the following centres: • • • • • • • •

The Polish Consulate, Polish schools, Polish libraries, Polish bookshops, Polish resource centres, the Polish Catholic Mission, Polish Literary Institute in Maisons-Laffitte, Centre de Civilization Polonaise, Sorbonne, Paris IV.

The organisations were selected from ‘French Polonia’ official websites (among others: http://www.polonianet.pl) and were also reached by personal recommendation from other subjects participating in the study. The interviews with this group were meant: • • • • • •

to obtain insights into the activity of Polish organisations in France; to obtain quantitative data concerning Polish migration to France; to find out about French immigration strategies; to better understand the situation of Poles in France after 1980; to find out about areas viewed as problematic; to create a database of consultants and informants.

8

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Insights from interviews in this group will serve as sources for the sociohistorical background and migration strategies formulated and implemented by French policy-makers regarding the Polish Community in France, and cited throughout this paper. Below the interview procedures will be described.

Interview procedure: Areas of investigation All the interviews in group 1 which are drawn upon in the following parts of this paper were conducted in May and June 2007. In some cases, where further clarification was needed, telephone conversations followed the face-to-face interviews. Each representative family participating in the study gave an interview of at least two hours. All the interviews with subjects in group 1 were conducted in Polish. The co-ethnic research was an advantage in building good relationships with the interviewees, based on mutual comprehension and shared background knowledge. This is especially true when one thinks of the era when our subjects chose to emigrate, that is the Solidarity period, when socio-historical context, national symbols and values were especially acute. On a general note it can be said about the subjects that they tended to open up and provide well thought-out, genuine insights into their particular respective situations. They enjoyed a retrospective outlook on their decision to migrate and found it pleasurable to lend their expertise. As a result, many anecdotal comments enriched the data with the subjects’ individual perspectives. Even though an attempt was made to structure the conversations (see below) and to make them cohesive, at this initial stage most emphasis was laid on the heterogeneity of the respective stories rather than making an attempt to produce generalisations and look for homogeneity.2 The interviews were not recorded, but the catalogue of questions served as a framework for these conversations.

Areas of investigation All our participants were asked about the following topics. (1) (2) (3) (4) (5) (6) (7) (8)

Date of arrival in France Age Origins (region of Poland they come from) Education Professional career Reason/s to emigrate Intended length of stay in France (on arrival) Self-assessed level of French as it is now

Emigrat ion Versus Mobilit y

(9) (10) (11) (12) (13) (14) (15)

9

Self-assessed level of French on arrival Language learning strategies Attitudes towards the host culture Attitudes towards the home culture Domains where French and Polish language are used Visits to Poland/returns Questions concerning children: number of children, languages proficiency in children (both as regards Polish and French), educational experience of Polish children in French mainstream schools.

Below, the subjects’ insights provided within the above topical areas will be presented against the backdrop of the results of interviews with employees of Polish cultural organisations, in so far as possible substantiated in the overview of literature of the subject.

Integration of the Polish Community in France after 1980: Attitudes of French Society towards Polish immigrants Whatever motives lay behind their decisions to emigrate, the Poles came with an intention, often based on necessity (cf. section on Polish community in France), to stay in France and to start their lives from scratch in their new country. Their determination to master the language and to interact with the French nationals was therefore very strong. The majority of new arrivals spoke no or very little French when they first came. It is all the more significant that no major Polish Community centres were established to assist, through the medium of Polish, the newcomers in their initial period in France. The Solidarity Committee was established in Paris in 1982 with a view to coordinating supportive actions for Poland (financial help, sending parcels with goods unobtainable in Poland during the Martial Law, publishing informative texts, undertaking attempts to influence Western governments and calling for amnesty for political prisoners). Supportive actions in the world of culture were undertaken by ‘Kultura’, The Polish Literary Institute and its editor-in-chief, Jerzy Giedroyć. These, however, targeted mainly the Polish state, not the Polish Diaspora.3 On the whole, the French were very supportive of Polish anti-communist actions and joined anti-communist protests (cf. Kubasiewicz-Houée, 2007: 151) such as the one held to express disapproval of Wojciech Jaruzelski’s official visit to France in 1985. They also facilitated the immigrants’ first steps in the new country by providing simplified procedures for obtaining legal social status, accommodation (HLM – council housing), access to requalification professional programmes and free language courses. The Polish

10

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Catholic Mission made a major contribution and many interviewees underscored they had been provided with assistance in their search for employment and accommodation there. It is worth of mention that The Polish Church maintained a strong presence in France during the Communist period, notably through the maintenance of a seminary for Polish priests in the building of the Irish College in Paris (now returned to Irish management). The respondents usually underline the positive attitude and understanding offered by the French, including with language learning support and integration: I was lucky the lady I was au-pairing for strongly encouraged me to speak French. Each time I was to do shopping she didn’t allow me to go the supermarket but to the ‘le marche’ in the street where I had to communicate with the vendors, ask for prices, negotiate if I had to.’ (Female, mid 50s, Paris)4 My husband’s colleagues ‘took care of him’. They slowed their speech, made him repeat the most needed phrases, they bought him a dictionary, at the beginning they assisted him in his duties (delivering goods to shops). (Female, mid 50s, Lille) None of the subjects complained of factors that would impede their integration into French society. Only a few subjects expressed the opinion that many French people lacked basic knowledge about Poland and demonstrated a superior attitude toward the non-nationals. This, however, is not indicative of real problems but rather a personal sentiment. Most subjects would commonly stress they had been accepted by the French and felt respected as hard-working and reliable workers. Even though French attitudes were highly assimilationist and allowed little space for expression of alternative cultural identities and languages, at least in the public domain, most of our subjects did not expect to be treated otherwise. They were rather willing to internalise French values. We always accepted our position of a guest in France. We were happy to learn to live like the French. Even when inviting French people to our house parties, we made sure the meal was at French eating times and also French national dishes were served. Polish traditional food was only an addition, a sample. (Female, mid-50s, Paris) It is significant that most of our subjects strongly emphasised they had not looked for Polish circles in France, and would much more willingly network

Emigrat ion Versus Mobilit y

11

with French nationals. Reasons for such behaviour included the subjects’ determination to master the French language and provide themselves with ample opportunities to converse with native speakers. Second, many subjects were afraid of entering into relationships with other Polish people as it was feared that newly met compatriots might turn out to be intelligence agents working for the communist party. As reported by our subjects, in the times in question a decision to emigrate could bear heavily on the emigrants’ relatives who stayed in Poland. Dziwniel, in his émigré’s memoir, recollects his wife following an ‘acculturation apprenticeship’: ‘She was invited by a French family to stay with them. She was learning how to live “à la française”. She was accompanying them everywhere, went shopping with them, prepared meals together so as to fully immerse herself in the French lifestyle.’ (Dziwniel, 2007: 97. Author’s translation). The subjects generated a great sense of belonging, when French people accepted them on equal terms: When our younger daughter was born my husband’s superior gave him a huge financial reward. He was a highly valued employee. (Female, mid 50s, Paris) At the same time, the informants often retrospectively observe integration often had to be initiated by them process and they had to make the first move: I remember, at the beginning I was not invited to ‘their’ parties. And one day I invited my colleagues to my son’s communion. They came, and then, the following month, my boss invited me and my family to his boy’s christening. I was so happy! (Female, mid 50s, Lille)

Language acquisition and learning Language issues were the most acute problem facing the Poles who came to France in the 1980s. The majority of the respondents came to France with no French at all or with only a couple of years of school instruction. They admit it took them one to two years to acquire communicative skills and learn to effectively function in the working environment, whereas another two years had to pass before they could feel communicatively efficient and comfortable. The respondents stress the difficulty they personally experienced or mention anecdotes concerning their acquaintances related to the lack of linguistic skills on arrival in France. Even though there were more stressful factors coming into play, it was repeatedly emphasised by the

12

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

subjects that linguistic problems often led to frustrations and in some cases even put at risk their sense of well-being and family life (marital break-ups were not infrequent). At the beginning I felt mute and deaf. I desperately missed my family and friends. I had no one to talk to. (Female, mid 50s, Paris) My wife had an emotional breakdown when we first arrived in France. She said she wanted to go back to her family, friends and the main problem was that she was slow at learning French, couldn’t speak to people. (Male, late 50s, Paris) The Principle of the Polish School in Paris points to problems suffered by teenagers who on arrival were immediately introduced to French mainstream schools even though unable to communicate with their French teachers or peers. Such a boy (girl) is psychologically devastated. They can’t speak, can’t operate in peer group and this leads to strong personality crises (Polish School Principle, personal communication, May 2007). Cf. My son (14) suffered enormously when he went to the French school. It was one of the most difficult times we’ve gone through since our arrival in France. (Male, early 60s Paris) A vast majority of respondents strongly emphasise the positive enforcement they managed to draw from their ‘overcoming obstacles’ experience and the sense of achievement they derived from successful mastery of a new language. Emigration is for the fittest! I was strong. I went through hard times. But difficulties strengthen our characters, don’t they? (Female, mid 50s, Paris) In the 1980s the state offered free language courses to the unemployed members of foreign communities and such courses were usually held in town halls (Les mairies). The subjects vary substantially in their evaluation of the efficiency of instruction. A lady who was enrolled in a free French programme in 1989, for a period of 12 months, says it was a real breakthrough in her linguistic progress. Two other subjects complain the courses were not

Emigrat ion Versus Mobilit y

13

well matched to their level and addressed mainly to an Arabic-speaking community with little prior language learning experience, which slowed down the rate of instruction. Those subjects who were able to afford them would sign up for paid language courses offered by Alliance Française, and – even though they acknowledge their benefits - remember the costs as a significant expenditure in their budget. The general sentiment would be that more free-language initiatives would have been valued both as linguistic support and a boost towards integration. I wish I could have afforded a language course. My wife, as an unemployed person was entitled to one for free and I could see she was making a fast progress. (Male, mid 50, Paris) When my employer saw I was a good worker, he paid for two language courses during my first two years in France. I really appreciate that. (Male, early 60s, Paris.)

Polish language maintenance and loss Whereas French was and still is critical for the majority of subjects at work and for social contacts, Polish would most typically be used at home, in telephone conversations with family and friends in Poland, less frequently in conversation with other Polish people in France. Two out of ten Polish couples interviewed insisted on the exclusive use of Polish at home, whereas the remaining subjects spoke Polish to their spouses and to their children, but allowed children to answer in French. A most powerful in reversing language shift (cf. Fishman cited in Garcia et al., 2006: 21) were stays in Poland: an opportunity for children to be around and to talk with their Polish grandparents, relatives and friends: My children spoke Polish at home for three months following their holidays in Poland, usually until Christmas, then switched back to French. (Female, mid 50s Paris) Another efficient way to maintain and transmit the native language, as observed by the informants, was sending the children to the Polish School (funded by the Polish state) which, affiliated to the Embassy, implemented the Polish primary and secondary level mainstream education programmes in areas absent from the French curricula (e.g. Polish language, history and geography). As a form of complementary education, classes take place on Wednesday and Saturday, two days free of school in France. Even though it

14

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

could be expected that this additional duty may be a burden for Polish children, they usually enjoy their time at school, as the Polish School Principle informs. Located in central Paris, the Polish School is not easy to reach for families from the outskirts of Paris. Sometimes the whole extended family is involved in bringing the children to school. The parents drive them in the morning and the grandparents collect them after the classes. (Polish School Principle, personal communication, May 2007) The school is highly valued by parents as it helps sustain Polish language and culture abroad and on successful completion gives extra points in school leaving exams (baccalauréat). Two out of ten families interviewed sent children to the Polish schools, one in Lille and one in Paris. The majority of subjects have managed to maintain and transmit spoken Polish language to their children. Only three couples, out of the ten interviewed, however, succeeded in teaching their children how to write in Polish. The parents interviewed take pride in their children’s ability to speak Polish and admit it required much effort and discipline on their part to transmit their native language: It was not at all easy to teach our children Polish when they go to French schools have French friends, speak French everywhere except home. (Female, mid 50s, Paris) By contrast, factors that may lead to a loss of native tongue (the case of one family interviewed), as anecdotally reported by the subjects with reference to their Polish acquaintances, are: (a) A belief, held by parents, that Polish spoken at home will slow down the integration of their children into the French society and create too much confusion (b) Attempts by parents to improve their French and with this aim in mind to make French the only language spoken at home and to model their speech on their children. I have to admit we made a mistake of speaking French at home at one stage. The children will learn French at school anyway, whereas we were unable to express ourselves properly and it wasn’t good either for us or for the children. We quickly abandoned the idea. (Male, mid 40s, Paris)

Emigrat ion Versus Mobilit y

15

(c) Attempts by parents to take on a French identity at the expense of a Polish one. This would most likely be the case in those Polish people who act under the conviction that Polish language and culture are inferior to the French and do not want to be associated with the former. (d) Sheer lack of time, long working hours, lack of Polish centres around. My daughter who came to France aged five, lost Polish at the age of ten. She always recounted her daily events from school in French and spoke French to us even if we spoke back in Polish. (Female, mid 50s, Lille) In the absence of a wider sample to substantiate this hypothesis, it may only be speculated that those children without Polish (now aged 15 or more) may now try by themselves to rediscover their roots. This observation would be in accordance with Giddens’s claims that the third generation of migrants is the one which most likely embark on a search for their roots, native language and culture (cf. Giddens, 2007: 122). Interviews conducted in Polish bookshops in Lille and Paris have confirmed such interest displayed by the second and third generations of Polish migrants to be increasing. There are more and more young people aged 20+ coming and explaining in native French they are Polish and want to find out about the region they come from. . . (Representative of the Polish Bookshop in Paris, personal communication, May 2007) We observe an increased demand now for Polish language textbooks and Polish guidebooks. (Representative Polish Bookshop in Lille, personal communication, May 2007) Even though those who visit Polish bookshops, by definition, will be a selfselecting sample of persons with a specific interest in Polish, the ob served increase in number of such inquiries may indeed confirm Giddens’s hypothesis. Similarly, there are more and more French citizens of Polish origin who are interested in tracing their family pedigrees as an employee of a vast Polish resource centre in Henin Beaumont notices. A special service has even been set up there to provide those interested with information concerning their origins (Representative of La Maison de la Polonia, personal communication, May 2007).

Links with the home country, returns Owing to the political situation in Communist Poland prior to the year 1989, visits to Poland were impossible for the migrants and even telephone

16

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

communication was not easy to establish. This is probably the main differentiating feature between the Solidarity and present-day migration. The Poles who migrated to France in the 1980s stayed there for five or six consecutive years on average before they first went on a visit to their native country. During the first years of our stay in France, the only form of contact with the family was writing letters. (Female, late 40s, Dunkirk) Visits to Poland were also hindered by the high cost of travel. Only after low-cost flights were established between France and Poland did many Polish migrants decide to visit Poland and often to spend their entire holidays there. As the Polish post-communist reality has changed greatly since 1989 and Poland has become an attractive holiday destination, travel agencies offer more trips to Poland. They are gaining popularity with the French citizens of Polish origin. As a result of this ease of travel, many subjects would often send their children to Poland for the entire holidays and are glad they enjoy going to Poland and improve their Polish during such stays. By contrast, as reported by some of our subjects, those migrants who came to France in 1989 and after usually had no intention of staying more than five years, a time long enough to save up and then invest in property in Poland. Most of them, however, were later on trapped in the currency squeeze between rapidly appreciating Polish ‘zloty’ and the continuing deterioration of the falling franc. As already proved in this preliminary stage of our research, the group of post-1989 migrants, even though they ultimately decided to make their lives in France, would maintain closer links with the home country and the Polish circles in France than Poles who arrived prior to 1989. Another phenomenon worth investigating is the issue of ‘returns’. As many Polish migrants who came to France in the 1980s are now reaching retirement age, some decide to return, invest in property and spend their retirement in Poland; the cost of living, still much lower than in France being an additional inducement. In the course of our research project interviews with employees of moving companies are envisaged with a view to obtaining more reliable data as to how widespread this phenomenon is.

The Polish Community in France in the 1980s: Summary As the preliminary findings presented above demonstrate, Polish people who migrated to France in the 1980s have made every effort, or were forced

Emigrat ion Versus Mobilit y

17

by circumstances, to assimilate into French society, which is clearly visible in the following: • • • • •

attempts to disperse in French society rather than form Polish circles; investment in the mastery of French; assimilation of cultural norms and routines in French society; limited contact with the home country; limited presence of Polish infrastructure or Polish language media in French mainstream culture.

Similarly, the French strategy for integration deployed during the period in question was focused on facilitating the assimilation of the new arrivals. Towards this aim, the following initiatives were undertaken by the French government: • • • • • • •

facilitated procedures for obtaining citizenship; attribution of HLM (council housing) accommodation; financial assistance in the initial period; providing access to re-qualification professional programmes; providing free French language courses to unemployed members of the Polish Community; providing language support classes for migrant children (known as classes d’accueil for recent immigrant children who speak little French) (cf. Schnapper et al., 2003: 31); encouraging French people to engage in charity drives and providing the Polish newcomers with help and thus encouraging a more rapid assimilation.

By contrast, little value was placed on Polish native language and culture maintenance. As a consequence, Polish language and culture was not given recognition in the public domain and met with no major interest on the part of the French. The experience of the Polish Community during the period of Solidarity, will serve as the backdrop to the discussion (following) of the post-accession migration of Polish nationals to Ireland.

The Polish Community in Ireland Remarkable differences can be observed between migration of Polish nationals in the 1980s and the massive influx of Polish people to Ireland after

18

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Poland’s accession to the EU. Before going on to present the situation of Polish migrants in Ireland, it has to be observed that many of differences characterising those two respective migrations result from the fact that the political situation and times have changed. The major changes in that domain include much greater access nowadays to home culture through satellite television and greater ease of return through low-cost air and bus travel. One may speculate that this affects the likelihood of retaining home culture or the propensity to return. The intrinsic differences between the circumstances in which the two groups migrated has resulted in different models for integration. These areas of divergence will be discussed below with a focus on the following factors: demographics, (declared) length of residence, degree of contact with the home country, presence of Polish culture in the public domain, attitudes of the host society members towards new arrivals.

Demographics Reliable data as to the exact number of Poles who have arrived and settled in Ireland from 2004 to 2008 (the time of writing) are for the most part unavailable and, likewise, there is a shortage of quantitative data concerning the make-up of the Polish Diaspora. According to the 2006 Census, there were 63,276 Polish nationals living in Ireland (1.5% of the total population). These data are, however, in disagreement with statistics concerning the quantity of PPS numbers allocated to Polish nationals to date. As The Irish Times reports in 2007, ‘Since the EU expanded in 2004 to include Poland and nine other countries, almost 200,000 Poles have moved here’ (Monaghan, The Irish Times, 16 Feb 2007) which would mean that Poles now constitute 5% of the total population. Even though the figure 200,000 is most frequently cited, there is a basic confusion here of flows and stays. 200,000 persons may well have applied for PPSNs but this does not mean that they call came, stayed and are still here (cf. Kropiwiec & King O’Riain, 2006). GrabowskaLusin´ska estimates that in September 2006 alone only 307 Polish nationals were employed, out of 8515 who received PPS numbers (Grabowska-Lusin´ska, 2006: 7). It is often the case that holders of PPS numbers have already left the country and no record is kept of people’s returns. The absence of such data does not solely apply to Ireland but is a worldwide phenomenon. ‘Given the fluidity of international migratory patterns’, as remarked by Castles and Miller, ‘it is not surprising that we don’t have precise figures on the number of people who can be defined as migrants.’ (Castles & Miller, 1993 cited in Kivisto, 2002: 2)5. In a similar vein, Grabowska-Lusin´ska points to the transitional nature of today’s migrations and to the fact that a large number of

Emigrat ion Versus Mobilit y

19

Polish migrants are simultaneously living in a number of countries, hold more than one citizenship and use various administrative entitlements in many countries. Similarly, the former Polish Ambassador to Ireland, Tadeusz Szumowski repeatedly emphases that the term ‘migration’ is a misnomer when used with reference to the Polish nationals and ‘mobility’ should be used instead (e.g. personal communication, October 2007). A number of Polish statistics institutions have also been monitoring the outflow of Polish citizens. On 1 January, 2007 the Polish Central Statistic Office (GUS) estimated that 1.95 million Poles were living outside of Poland in other European countries for a period longer than two months. According to the same office, the number of Polish people who left their country between the EU enlargement in 2004 and the end of the year 2006 numbered to 1.1 million. Ireland is the third most popular destination for post-accession Polish migrants, chosen by 10% of the whole number of those who have left Poland since May 2004 (Grabowska-Lusin´ska, 2008: 7, 8, data drawing upon ‘BAEL’, Economic Activity Survey, Migrants’ Base). According to The Irish Times, Poles are most likely to live around towns, cities and transport networks. There is a significant concentration of Poles in the Dublin area (around 80% of Polish nationals in Ireland) (O’Brien, The Irish Times 3 July 2007) as well as clusters within commuting distance of cities such as Cork, Limerick and Galway. It is also possible to see concentrations of Poles along major road networks such as the N7 and N4. Smaller urban areas such as Killarney, Monaghan and Roscommon also display clusters of Polish nationals (ibid). The SAPS (Small Area Population Statistics) data from the 2006 Census suggest a very widespread dispersal of Poles in almost every part of Ireland (Central Statistics Office, Theme 2–2: Usually resident population by nationality, 2006).

Interview-based study The discussion of the situation of Polish people in Ireland below will draw upon insights from conversational interviews conducted between November 2005 and October 2007 among 30 Polish people (aged 20–45) who came and settled in Dublin in 2004 and 2005. The sample consisted of: employees of Polish organisational structures (Polish media in Ireland, Polish cultural, social organisations), shop assistants, employees of call-centres, students and staff members at Dublin universities and vocational schools, Polish partners (spouses) of Irish citizens enrolled in Polish language programmes in the Department of Russian and Slavonic Studies, Trinity College Dublin. All the informants were met by the author in 2005, through networks developed in the course of her work as a Polish Government Lector in the

20

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Department of Russian and Slavonic Studies at Trinity College and also people met in the streets, shops, and on planes. All of the subjects whose testimony is reported here remain in touch with the author to the present day. Spoken to on various occasions in relaxed circumstances, they provided the author with insights into their particular situations in Ireland, their envisaged length of residence, problems encountered, attitudes towards the Irish, as well as the positive and negative consequences of their decision to emigrate. All the interviews were conducted in Polish.

Length of residence; degree of contact with the home country; returns Results of a survey conducted by a Warsaw-based agency Kinoulty Research, which interviewed 109 Poles living in Ireland, indicate that 50% of the Poles questioned would like to live in Ireland as long as possible, while 18% want to stay here permanently. Barely one-third treat their stay in Ireland as temporary (Monaghan, The Irish Times, 16 February 2007). Such trends are further substantiated in an interview-based survey, executed five months later among 1389 Polish people (in Britain and Ireland) by a polling agency for the Gazeta Wyborcza newspaper, according to which, ‘nearly half of the estimated 200,000 Poles living in Ireland today (i.e. in 2008) say they do not intend to return home [while] some 49 per cent of those questioned said they wanted to stay in Ireland for at least the next five to ten years’ (Scally, The Irish Times, 5 July 2007). These findings are in a sharp contrast to more anecdotal evidence obtained from Poles interviewed in more informal contexts. ‘Most of my friends here plan to stay five or six years in Ireland but the longer they stay, the more difficult it is to move back to Poland.’ (A Polish man interviewed for The Irish Times, McDonagh, 14 August 2007). Since Poland joined the EU many people go abroad to work. Most plan to stay for a few years and then go home. (Polish priest interviewed for The Drogheda Independent 18 July 2007; cf. Kropiwiec and King O’Riain, 2006: 34) The intention to return to Poland was also expressed by the Poles interviewed in a longitudinal case-study conducted over the last two years among 30 Poles (cf. 6.2). Unlike Poles who migrated to France in the times of Communism, the Poles who have been interviewed in Dublin declared that they saw their stay in Ireland as having a definite endpoint. The majority of speakers (20 out of 30) said they were going to stay in Ireland for ‘up

Emigrat ion Versus Mobilit y

21

to 5 years’ and were then planning to return to Poland. Five respondents who admitted that they would like to stay in Ireland permanently were partners (spouses) of Irish citizens Five remaining informants (students and university staff members) admit that they are ‘keeping their options open and are waiting to see what time brings’. ‘I don’t know how long I’m going to stay in Ireland,’ was their most common answer and, even though leaving Ireland presents itself as one of the options, they do not necessarily equate it with a return to Poland. Those young, educated people are usually interested in gaining new experience and enhancing their qualifications, thus increasing their competitiveness on international labour markets. Interest in the Polish cultural events and in the political situation in Poland (as shown in the high turnout at the parliamentary elections in October 2007 with 20,000 registered for voting in Ireland) is also indicative of Polish migrants’ concern with current developments in their home country and most likely their plans to return. Over 75% of the author’s respondents said they went to Poland at least twice a year to visit their families and friends. Scheduled flights on 32 routes between Ireland and Poland offered by at least seven airlines have been developed since 2004 to facilitate such visits. As a consequence, Polish people can remain in close contact with their country and families. Therefore acculturation is not a pressing need, as it was the case in the subjects interviewed in France. Over 60% of all the respondents admit that they often host their Polish friends and family members. 80% of the informants frequently talk to their relatives and friends in Poland and benefit from low-cost phone calls and Internet calls (cf. ‘transnational practices’, Kropiwiec & King O’Riain, 2006: 44). Those Poles who migrated to France in the 1980s probably had little expectation that an early return to a democratic and economically vibrant country was likely and may have made the best of a situation where their future most likely lay, in France. This attitude would certainly have been encouraged by a country which expected migrants to fit in. By contrast, if there is a characteristic feature of migration within the EU, it is fluidity, with much coming and going and high rates of return by EU migrants to their own countries. Earlier phases of EU expansion show a clear pattern in which economic convergence occurs as living standards of newer Member States rise towards the norm and migrants return as those improved economic circumstances offer greater possibilities in the home country. This is demonstrably the case for such countries as Ireland itself, Spain, Portugal and Greece. One can expect that a similar pattern will emerge as economic and social conditions improve in the New Accession States and that a certain proportion of migrant workers from those countries will return home in the

22

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

future, although the precise patterns which may emerge are a matter of conjecture at present. In future stages of research an attempt will be made to investigate the interrelation between the perceived need to become acculturated and the intention to remain.

Attitudes of the host community towards new arrivals The position of Polish post-accession migrants in Ireland is entirely different from the position of Polish nationals who migrated to France in the 1980s. Even though a positive attitude towards Polish newcomers was observed in the French society in the Solidarity period, it was not accompanied by an increased interest in the Polish language or culture. The primary aim, as seen both by the newly arrived Poles and the French, was to facilitate assimilation. Additionally, in light of the fact that the population of France was over 60 million and the Polish Community was one of many minority ethnic groups at the time, the Polish presence was unlikely to make itself particularly visible. Estimates indicate that there are (at the time of writing) more speakers of Polish in Ireland than there are native Irish speakers. As a consequence, the Polish Community found itself at the centre of public attention. The strong presence of Polish migrants and the representation of Polish culture have been met with increasing interest on the part of the host society. A unique phenomenon is witnessed in Ireland at the time following the influx of Polish nationals in 2004, namely, a great effort on the part of the host community to master the language of a migrant community. Increased numbers of enrolments of Irish people in Polish language courses are proof of the increasing interest in Polish language and culture among the general public. Polish cultural events held on a regular basis not only facilitate contact between Poles and their native culture, but also contribute to promoting it amongst the Irish. Polish events which have taken place in Dublin over the last six months (e.g. IFI Polish Film Festival, Polish Art exhibition in the National Gallery in Dublin) were met with strong interest on the part of the host community. In the three years since Poland’s accession to the EU, one can say that the Polish Community has established its strong presence in the host society, and greatly enriched the cultural scene in Ireland. Results of the analyses of Irish media coverage conducted since January 2007 are indicative of this interest. The following areas concerning the Polish Diaspora are most pertinent: • •

estimates of the number of Polish Community members (based on PPS counts); declared length of stay;

Emigrat ion Versus Mobilit y

• • • •

23

employment sectors; Polish cultural events; Polish infrastructure; Polish contribution to the newly developing multicultural Irish society.

The Polish language was rapidly appreciating in value on the linguistic market in Ireland. A wide and elaborate linguistic infrastructure developed over the years directly following the accession around the Polish Community in Ireland, with more than 15 Polish-language publications or websites operating here and a network of schools, shops, bars and religious groups established. The medium of Polish occupies (in 2007) a sizeable share of radio and TV broadcast time and advertisement. State exams can be sat in Polish, both at the Junior Certificate and Leaving Certificate levels; Polish is one of 15 mother-tongue languages examined at Leaving Certificate level (cf. Debaene, 2008; Mac Cormaic, 2007). There are a couple of explanations for the vast popularity of Polish language and culture in Ireland. First of all, the Polish Community constitutes around 8% of the local workforce. It therefore becomes a potentially profitable target for service providers (e.g. in publishing and advertising). As opposed to the situation of Polish people in France in the 1980s, when the tendency was to avoid contacts with other Polish Community members, sometimes out of fear of Communist agents, sometimes in an attempt to accelerate integration into French society and mastery of the French language, Polish migrants in Ireland are willing to maintain their Polish lifestyle. In the future stages of the research it will be investigated whether greater reliance on Polish language media does not slow down Polish people entering networks or limit second language acquisition.

The Irish Model for Integration Contrary to the well-documented integration policy in France, Irish approaches to integration have yet to be translated into effective official policy. The Minister of State for Integration, Conor Lenihan said at the Trinity Immigration Fair (24 September 2007) that Ireland aimed for a ‘a mid-way point between the excesses of the multicultural model and the excesses of the assimilationist model’. The primary goals in 2008, as seen by Lenihan, are (a) the provision of English language services to adult migrants; (b) funding lines with regard to faith-based organisations involved with migrants and the large membership organisations that drive civic participation; (c) encouraging

24

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

integration at local levels; (d) appointing a standing ministerial council for integration drawn exclusively from the migrant community; (e) appointing a new task force on integration to recommend policy solutions for the medium to long term; (e) a standing commission or forum on integration (Lenihan, 2007). The term which is most frequently used in the ongoing Irish debate on integration is ‘interculturalism’, which seeks to avoid the problems associated with earlier approaches to integration (criticism of some conceptualisations of multiculturalism have focused on the fact that it constructs minority communities as homogenous entities with no internal divisions). As characterised by the National Consultative Committee on Racism and Interculturalism (NCCRI): Interculturalism suggests the acceptance not only of the principles of equality of rights, values and abilities but also the development of policies to promote interaction, collaboration and exchange with people of different cultures, ethnicity or religion living in the same territory. Furthermore interculturalism is an approach that sees difference as something positive that can enrich a society and recognises racism as an issue that needs to be tackled in order to create a more inclusive society. Submission to the all-party oireachtas committee on the Constitution. The Family, 2005: 6 As it is often emphasised in the course of the Irish debate, integration is generally regarded as a two-way process between minority ethnic groups and the majority populations (cf. NCCRI submission to the InterDepartmental Working Group on the Integration of Refugees in Ireland, 1999). ‘Integration means the ability to participate to the extent that a person needs and wishes in all of the major components of society, without having to relinquish his or her own cultural identity.’ (Integration: A two way process’, p. 9) The example of the post-accession Polish Community can be said to have made a significant contribution to the ongoing intercultural dialogue. The value placed on the Polish contribution to the mainstream culture is reflected in: • • • • • •

strong presence of Polish language in public spaces; strong representation of Polish; well-developed Polish infrastructure; Polish media operating in Ireland; particular attention paid to the Polish Community in the Irish media; rising interest in the Polish language and culture on the part of Irish people.

Emigrat ion Versus Mobilit y

25

Conclusions and Future Directions Based upon the preliminary results the following factors which bear upon the process of integration can be identified: • • • • • • •

reasons to migrate; (envisaged) length of residence in the host country; demographics, i.e. the number of immigrant community members and their concentration in specific areas; status enjoyed by the migrant community in the host society; institutional support, i.e. use of a community language in mass media, as the medium of education, or for religious purposes; presence of native culture in the mainstream cultural scene; degree of contact with the home country.

In the course of further studies envisaged both in France and Ireland, an attempt will be made to identify the positive and negative outcomes of the two migration patterns (migration for settlement and mobility) on Polish migrants particularly with respect to their sense of well- being in the host community, identity, pace of second language acquisition, professional careers and family life.

Notes (1) This paper draws on data collected in France under the auspices of the IRCHSSfunded project ‘Second Language Acquisition and Native Language Maintenance in the Polish Diaspora in Ireland and France,’ Investigators: D. Singleton, V. Regan and E. Debaene. (2) The author hereby wish to acknowledge the help of Sarah Smyth (Department of Russian and Slavonic Studies, Trinity College Dublin) who suggested this methodological approach be used in this study. (3) The author hereby wishes to thank Henryk Giedroyć and Wojciech Sikora (The Polish Literary Institute, Maisons-Laffitte) for their most valuable insights into the situation of Polish migrants in France in the 1980s. (4) All translations have been done by the author who accepts full responsibility for their accuracy and appropriacy. 5Estimates suggest that 175 million people (3%of the world’s population) live outside their country of birth (Golgin and Reinert, 2006: 157)

References Central Statistics Office Ireland, SAPS – Small Area Population Statistics. Usually resident population by nationality, 2006, accessed 18 February 2008. http://beyond2020. cso.ie/Census/TableViewer/tableView.aspx?ReportId = 12647 Costa-Lascoux, J. (1989) De l’immigré au citoyen. Paris: La documentation Française.

26

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Debaene, E. (2008) Community language policy, planning and practice. The case of Polish in the Republic of Ireland. Scottish Languages Review 17, 1–3. Inter-Departmental Working Group on the Integration of Refugees in Ireland. (1999) Integration: A two way process. Dublin: Department of Justice, Equality and Law Reform. Drogheda Independent (News) 18 July 2007, Fr Peter finds lots in common with homeland, accessed 18 February 2008. http://www.drogheda-independent.ie/news ˙ycie s´wiata: Dzieje s´wiata. Poznan´: Polskie Media Dybkowska, A. (1996) ‘Rok 1968’ Z Amer. ˙ aryn, J. and Z ˙ aryn, M. (2002) Polskie dzieje. Warszawa: Wydawnictwo Dybkowska, A., Z Naukowe PWN. Dziwniel, M. (2007) Znad Niemna przez Odre˛ nad Sekwne˛. Ruda: Mamiko. García, O. Peltz, R. Schiffman, H. and Fishman, G.S. (2006) Language Loyalty, Continuity and Change: Joshua A. Fishman’s Contributions to International Sociolinguistics. Clevedon: Multilingual Matters. Geremek, B. and Frybes, M. (2004) Kaleidoscope Franco-Polonais. Paris: Instytut Adam Mickiewicz. Giddens, A. (2007) Europe in The Global Age. Cambridge: Polity Press. Goldin, I. and Reinert, K. (2006) Globalization for Development: Trade, Finance, Aid, Migration and Policy. World Bank and Palgrave Macmillan, Washington, DC and New York. Grabowska-Lusin´ska, I. (2008) Najnowsza Emigracja z Polski – Raport OBM. Biuletyn Migracyjny 16, accessed 18 February 2008. http://www.migration-news.uw.edu.pl/ BiuletynMigracyjny16.pdf Grabowska-Lusin´ska, I. (2006) ‘Irlandia –migracyjny tygrys.’ Biuletyn Migracyjny 9, accessed 18 February 2008. http://www.migration-news.uw.edu.pl/Biuletyn Migracyjny9.pdf Habielski, R. (1995) Emigracja. Warszawa: WSiP. Hładkiewicz, W. (2002) Emigracja z Polski po 1989 roku. Zbiorowos´ci polonijne lat dziewie˛ćdziesia˛tych w Europie Zachodniej: Sposoby opisu i wyjas´niania zjawiska; Kraków: Bohdan Grell i córka. Jendrowiak, W. (2002) Histoire de l’immigration Polonaise dans la région du Nord et du Pas-de-Calais des origines a nos jours (185 pp. Maîtrise d’histoire contemporaine), Université d’Artois. Judycki, A. and Judycka, Z. (1996) Les Polonais en France. Dictionnaire Biographique, Volume I. Paris: Editions Concorde. Kastroyano, R. (2002) Negotiating Identities. States and Immigrants in France and Germany. Princeton : Princeton University Press. Kivisto, P. (2002) Multiculturalism in a Global Society. USA: Blackwell Publishing. Kropiwiec, K. and King O’Riain, R.C. (2006) Polish Migrant Workers in Ireland. Polscy Migranci Pracujacy w Irlandii. Community Profiles Series. Dublin: National Consultative Committee on Racism and Interculturalism (NCCRI), accessed 18 February 2008. http://www.nccri.ie/pdf/06_Polish_Report.pdf Kubasiewicz-Houée, E. (2007) De l’Est à l’Ouest. Itinéraire d’une femme engagée. Translated from Polish by Ewa Houée, Yves Houée and Teresa Le Meur. Saint-Brieuc: Association Warmie/Mazurie. Lenihan, C. (2007) Addressing the needs of immigrants. Metro Eireann. 13 December, accessed 18 February 2008.http://www.metroeireann.com/index.php?option = com_ content&task = view&id = 936&Itemid = 51 McDonagh, M. (2007) ‘Poles apart from home.’ The Irish Times, 14 August, p. 38.

Emigrat ion Versus Mobilit y

27

Michel, S. (2007) Immigration: the plight of Poles in France. Cafebabel.com. European Current Affairs Magazine, accessed 18 February 2008.http://www.cafebabel.com/en/ article.asp?T = T&Id = 13007 Mac Cormaic, R. (2007) Lenihan supports ‘mid-way’ strategy on migration. The Irish Times, 25 September, p. 6. Monaghan, G. (2007) Poles here and here to stay. The Irish Times 16 February, p. 44. Nowa Encyklopedia Powszechna. (1995) Volume 2. Warszawa: PWN. O’Brien, C. (2007) Patterns of Polish and UK settlement here differ. The Irish Times, 3 July. p. 8. Overview of integration policy in France 2002–2005. Le rapport du Haut Conseil a l’integration, accessed 18 February 2008. http://www.eukn.org/binaries/eukn/ france/policy/2006/9/policy-overview-integration.pdf Parekh, B. (2006) Rethinking Multiculturalism, Cultural Diversity and Political Theory.London: Macmillan Press Ltd. Ponty, J. (2004) Immigration au XXe siecle, phenomene nouveau, l’immigration devient un enjeu dans les relations internationales. In B. Geremek and M. Frybes (eds) Kaleidoscope Franco-Polonais (123–129). Paris: Instytut Adam Mickiewicz. Scally, D. (2007) Half of Poles in Ireland say they intend to stay. The Irish Times, 5 July, p. 10. Schnapper, D., Kreif, P. and Peignard, E. (2003) French immigration and integration policy. A complex combination. In F. Heckmann and D. Schnapper (eds) The Integration of Immigrants in European Societies: National Differences and Trends of Convergence (15–44). Stuttgart: Lucius and Lucius. Submission to the all-party oireachtas committee on the Constitution. The Family. National Consultative Committee on Racism and Interculturalism. January 2005, accessed 18 February 2008. http://www.familydiversity.ie/resources/nccri%20submission%20jan%2005.pdf Submission to the Working Group on the Integration of Refugees. February 1999, accessed 1 February 2008. http://homepage.eircom.net/ ~ racismctee/submit.html ˙ alin´ski, H. (2005) Polacy we Francji. Historia i współczesnos´ć. ‘Konspekt’, March 2005, Z accessed 18 February 2008. http://www.wsp.krakow.pl/konspekt/21/polfranc.html

2 The Bookseller and the Basketball Player: Tales from the French Polonia Vera Regan

Introduction This chapter presents the story of some individuals from French Polonia, focusing in particular on two speakers from the group studied, through the lens of close quantitative and qualitative analysis of their speech. The mixed method approach used to throw a spotlight individuals’ lives, histories, aspirations, experience of migration, stances in relation to dual and multiple identities helps in the telling of their individual stories and has implications for migrants globally. It has been customary to see quantitative analysis as not always picking up the grain necessary to access the fine detail of individuals’ situations. It is hoped to demonstrate that quantitative analysis when sensitive to currently evolving categories in the current world situation, and when combined with qualitative investigation of the data gathered, will propose an effective method of gaining access to peoples’ life histories in a way that using only qualitative analysis, or case studies, or indeed quantitative analysis on its own, do not. For an overview see Regan (2013). To this end, the paper addresses issues of the methodological appropriateness of quantitative versus qualitative data, particularly in relation to second language acquisition (SLA) and migration. The issue of individual variation has been a theme in SLA literature almost since its inception, evoked in general overviews of the field (for example Ellis, 1994; Myles & Mitchell, 1998), as well as in individual studies (Dewaele & Furnham, 2000; Regan, 1995). 28

The Book seller and the Basket ball Pl ayer: Tales f rom the French Polonia

29

There is a general acceptance that individual variation is a reality in SLA, but quantitative studies of second language (L2) speakers, which deal in aggregations of large numbers of informants and data, are sometimes said to ‘swallow up’ the detail of individual behaviour in their analyses. These issues have been addressed, for example, by Bayley (2002) and Regan (2004), both of whom show by quantitative analysis that individual variation patterns frequently follow group patterns. This paper aims to further explore this relationship between quantitative (specifically variationist) analysis and individual variation, in relation to L2 Polish speakers of French. In relation to this theoretical issue, the paper demonstrates that, whereas the quantitative results provide a general picture of the overall trends in the Polish participants’ speech variation, further qualitative analysis reveals an interesting picture of individual stances against the backdrop of these overall trends. So although it remains true that individual speaker patterns tend to replicate group patterns, it appears that qualitative investigations can provide crucial detail, which amplifies the emerging picture of peoples’ language use. The participants in the study, as migrants, differ crucially from the speakers described by much previous sociolinguistic research, who are usually less mobile and often live in a relatively ‘static’ physical situation. A consequence of migrants’ mobile lives in a globalised world is language practices and usages different from practices and usages of more traditionally situated speakers. Research approaches today need to take into account the trajectories of peoples’ lives, and the subsequent variation in their use of language resources, for instance Pennycook (2012). According to Blommaert (2010), ‘Mobility is the great challenge: it is the dislocation of language and language events from the fixed position in time and space attributed to them by a more traditional linguistics and sociolinguistics. . . It is the insertion of language in a spectrum of human action which is not defined purely in relation to temporal and spatial location, but in terms of temporal and spatial trajectories.’ This paper investigates how Polish migrants living in France use the linguistic resources at their disposal as they negotiate different time and space zones. For this an investigation which touches on indexicality in relation to their acquisition and use of variation patterns in their L2 (French) was put in place. In this context, a quantitative analysis of the variation patterns in their speech was first carried out. Following this detailed investigation of speech patterns, is an exploration of their function for the communicative needs of these mobile speakers currently living in France, having come from Poland and, in some cases, intending to return to Poland. In this sense, these speakers are very different from the more settled ‘speech community’ so central to traditional variationist studies of L1 speech.

30

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Quantitative results of the study of the Polish speakers show a detailed picture of their speech, but as we noted earlier, further unpacking the individual variation was found to deepen and broaden the picture. Analysis of the use of one particular variable, ne deletion, in French, by the Polish speakers show general patterns but also indicate individual agency in the face of social structural norms. Intriguing individual differences indicated by the quantitative results suggest the need for further probing. What follows now is a description of the social and historical context from which our participants emerge: a short history and description of Polish emigration to France in the past few hundred years and the recent story of those particular Polish people who participated in our study.

Poles in France Emigration has been a staple feature of the Polish societal landscape for at least two hundred years (for a historical overview of Polish migration to France see Debaene, this volume; Nestor, this volume; Regan, forthcoming), Poles have emigrated to France in great numbers for centuries. It is estimated that French Polonia amounts to 800,000 or even 1 million; only in the United States is the number of Poles greater than the numbers in France. Various waves of migration took place since the 19th century. This study focuses on Poles living currently in France, and deals with one of the main migration phases in France. The first period, not addressed in this study, is the post World War Two migration, when great numbers of Poles moved from Poland to work in the mining regions in Alsace and northern France. ‘Push and pull’ mechanisms were primarily economic in this period. The majority of newcomers found employment in mining and agriculture. This study addresses post-1980 migration, which consists of two phases: before 1989 and afterwards. Migration during the period 1980–1989 is frequently referred to as ‘Solidarity migration’. The highest outflow of people took place during and in the aftermath of martial law (1981–1983), when the Communist Party in power in Poland enabled one-way cross-border movement. In terms of factors which influenced the decision to emigrate, Solidarity migration is a complex phenomenon. Despite the common notion that this outflow of people was made up of anti-communist activists, forced to escape political repression (Zalinski, 2005), economic factors also played a non-negligible role in this migration wave as well (Habielski, 1995). Polish people who came to France before 1989 usually intended to stay permanently. Emigration in 1989 and afterwards took place after the round-table talks and the collapse of the Iron Curtain. Free movement of people across the borders facilitated migration, which was primarily economically

The Book seller and the Basket ball Pl ayer: Tales f rom the French Polonia

31

motivated. Those Poles who came to France after 1989 were, in the vast majority of cases, economically motivated and intended to stay for a few years, save up enough to invest in Poland and then to return. No completely reliable data appear to exist as to how many Polish people entered France in the 1980s. Anecdotally, the number of people who came and settled down in France between 1980 and 1985 is 100,000. There are two explanations for the absence of such data, as Elz˙bieta Sayegh (Consul at the Polish Embassy in Paris) suggests (personal communication) – (a) the period of the 1980s is still too recent for historical or even sociological studies; (b) the general tendency for the Poles who arrived in France prior to the collapse of communism in 1989 was to immediately integrate into French society. They were rapidly dispersed into the host society, rather than forming recognisable Polish ‘communities’. Also, Polish people were effectively granted political asylum, work permits and, in a matter of three years, French citizenship. No systematic attempts were made by the French to count Polish newcomers, as confirmed in the interviews with the employees of the Polish Consulate and the Polish Literary Institute (Elz˙bieta Sayegh, personal communication, Polish Embassy in Paris, May 2007; Wojciech Sikora, Polish Literary Institute, Maison-Laffitte, personal communication, May 2007). Around 90% of Polish people who entered France before 1989 applied for political refugee status and the procedure was operated by OFPRA – l’Office Français de Protection des Refugiés et Apatrides. Ponty provides earlier data from OFPRA (e.g. for the 1950s when 20,000 Polish immigrants benefited from the right of asylum), but does not, however, provide figures relevant to the period in which the informants of our study emigrated to France (Ponty, 2004).1

The Study A central issue of the wider study was an investigation of language as an indicator of integration. Sociological literature frequently cites language as one of the most important indicators of integration by migrants. Despite this claim, detailed analyses of language are rare in fact. For this reason, the study focuses on the close analysis of language acquisition and use as an instrument for gauging degrees of integration, stances of the speakers in relation to integration and, ultimately, a fuller picture of the lives of Polish migrants to France. Case studies of particular groups of migrants are valuable in the creation of a more complete picture of what it means to be a migrant, which is perhaps the more usual situation of people globally in the 20th century.

32

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

As noted earlier, the focus of this chapter is on language practices and language use by migrant speakers. A related, more specific question, is which causal variable or variables are most important in both the development and use of speech patterns and whether these play a part in or are indexical of the integration process of the individual into the host society. L2 acquisition literature has always found that length of residence and proficiency play an important role in language development and integration. Based on previous findings, this study hypothesised that length of residence in France, as well as proficiency in French, would correlate with use of the informal variant, ne deletion, in contemporary spoken French.2

Informants The informants for the study are a group of speakers of Polish nationality at two research sites: Mitre (anonymised name) and Paris. Mitre was chosen as it is situated in the northern mining regions where many Poles had settled in the 19th century. Migrants settling in the North benefited from the ‘chain migration’ phenomenon in that many had relatives already settled in the areas around Lens or Dunkerque. Paris has been a consistent location for Polish migration for many years. Polish migrants found that Paris presented many opportunities in terms of work and accommodation as well as Polish organisations and support centres which facilitate the initial contact with France. The informants in this study emigrated to France between 1960 and 1995. The length of residence for the speakers varies from 40 to 15 years at the time of interview. Their ages ranged from 40 to 70 years and they worked in a range of different occupations.

Methodology The speakers were interviewed in their own home, place of work or in public spaces such as cafés and restaurants, by and large the sort of places which would have figured normally in the daily lives of the speakers. The interviews were free conversations, very ‘lightly’ directed by the interviewers. The conversations frequently elicited narratives of personal experience, either present or past. Several speakers spoke of their life in Poland; some of the older people told stories of the Second World War. Many told of their experience of leaving Poland and migrating to France, as well as stories of parents or relatives who had come to France in previous generations, to the mining areas in the north of France. Each interview lasted at least two hours, and each was conducted by two interviewers: one Irish and one Polish. The presence of a co-ethnic interviewer had the advantage of creating a relaxed relationship, based on shared

The Book seller and the Basket ball Pl ayer: Tales f rom the French Polonia

33

experience and cultural knowledge.3 In addition to the ‘sociolinguistic interviews’, prior meetings with the speakers had been held by the Polish L1 interviewer; this elicited both ethnographic data as well as further reflections in Polish on the issues of identity, integration and experiences in France. Both sets of interviews were analysed subsequently and the data from both were integrated in the final analysis.

Data Analysis The interviews were transcribed, coded and analysed using GoldVarb 2001 (Robinson et al., 2001), a PC version of Goldvarb 2.0 for Macintosh. The original multivariate program (called Varbrul, Rand & Sankoff, 1990) was developed by Sankoff and Labov and others to model variation in language.4 Multivariate analysis of speech from a variationist perspective seeks to model variation in speech by calculating the relative effect of multiple affecting factors on the production of particular variants. For a more complete account of variationist analysis see Poplack and Tagliamonte (2001), Guy (1993), Tagliamonte (2006, 2012), Walker (2010) and Meyerhoff (2006). The constructs and methods of variationist analysis have been used in the analysis of L2 speech in the past few decades (Adamson, 1988, 2009; Bayley & Preston, 1996; Bayley & Regan, 2004; Young, 1991).

Variationist analysis and L2 speech The Varbrul computer program is a set of computer programs designed to analyse naturally occurring data and the social and linguistic distribution of variable forms. It is essentially based on a statistical regression model. Many statistical procedures are not appropriate for these data because naturally occurring data tend to have a large number of empty cells (because many combinations are linguistically impossible or very unlikely, and so this gives rise to lots of empty cells). A procedure such as Anova, for instance, is good for balanced data that emerge from controlled experimentation. It is inadequate to handle the kind of naturally occurring data that are collected in studies of interlanguage variation. So in relation to variable linguistic data, what we need is a way of modelling simultaneously the multiple and cross-cutting contextual and linguistic constraints on variation. Furthermore, we need to determine the relative weight of the factors (linguistic and social) which affect it, whether these social factors are ethnicity, gender, age or effect of context in acquisition. Versions of the Varbrul program have been described by Cedergren

34

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

and Sankoff (1974), Rand and Sankoff (1990) Sankoff, (1988), Guy (1988, 1993), Tagliamonte (2006), Paolillo, (2002) and Walker (2010). For SLA researchers, a comprehensive ‘how to’ is found in Young and Bayley (1996). The program quantitatively models variable yet systematic data. Native speech shows patterned variability (Labov talks of ‘orderly heterogeneity’); variable rule analysis models this variability. The approach variationists in SLA have taken is that ‘it is unlikely any single contextual factor alone can explain the variability in the data.’ (Young & Bayley, 1996). Multivariate analysis best deals with the multifactorial aspect of L2 data. Initially it used what was referred to as ‘variable rules’, a term which was part of early terminology in relation to categorical and optional rules. In fact the program probabilistically models choices made by the speaker and calculates the strength of the factors which constrain these choices. ‘Variable rules’ tell us not only what is possible, but what is likely and unlikely. They provide precise quantitative information. And they are probabilistic in focus rather than deterministic. Given that many factors contribute to the occurrence of a linguistic entity, we need to try to see the pooled contribution of each of these elements. The ultimate aim is to add more information to the model, to account for everything in the data and to explain or adequately represent spoken language. In sum, variation analysis examines the alternating forms of a linguistic variable, for example, je ne sais pas vs. je sais pas, and notes which features, linguistic and social, co-occur with these forms. These correlations are modelled quantitatively, so that the ‘best fit’ is found for the model proposed. Also, the model provides the maximum amount of information possible on the choice of variant. The program tells us the overall likelihood of the phenomenon occurring; not just whether something occurs or not, but how often, and what is the weight of the factors which affect it. We need to know other information, for instance, whether certain groups of speakers or indeed individual speakers are more likely to use it more than others, or if it is more likely to appear in certain linguistic contexts than others. Varbrul works by determining in a principled way the probabilistic weight each constraint (or factor) contributes to the operation of a rule, especially in the sort of data sociolinguists treat. The program calculates the weight for each factor and assigns a value ranging from 0.00 to 1.00. That range indicates the degree to which a factor promotes the operation of the rule (the higher the value, the greater the influence). Theoretically, weights below 0.50 inhibit the operation of the rule and those above it promote it. As with all figures and numbers in statistics there can be a certain arbitrariness in ‘cut-off’ points. It is more informative to see the entire range of the scale as an indication of increasing likelihood that the rule applies, and an indication of the relative strengths of the factors.

The Book seller and the Basket ball Pl ayer: Tales f rom the French Polonia

35

The variable: ne deletion Deletion and retention of ne is a stable variable in spoken French. It is not particularly undergoing change (although rates have increased in the recent past in native speech) and so is a feature the L2 speaker would become aware of as an intrinsic part of spoken French which they hear from native speakers. However, unlike a fluctuating, incoming variable undergoing change, it may be more easily acquired along with the constraint patterns which accompany it (that is, the patterns of relative strength of the factors which affect its use). In spoken French, ne is deleted variably to a great or lesser extent by all members of French-speaking communities throughout the world. It is a powerful indicator of formality, power and solidarity, style, register and so on and has a network of relationships with sociolinguistic factors such as age, sex and social class. It often co-occurs with other such stable sociolinguistically sensitive variables in French as the variable use of tu as opposed to vous, or on as opposed to vous and nous. Based on previous research on L1 as well as L2 speakers (Ashby, 1981, 2001; Dewaele & Regan, 2002; Regan, 1996; Sankoff & Vincent, 1977; Sankoff & Vincent, 1980) factor groups (or factors) hypothesised to constrain the variability in the data were lexicalisation, following phonological segment, preceding phonological segment, subject of sentence, subject of verb, verb type and presence/absence of clitic. Social factors were length of residence, gender and age (see Table 2.1). We excluded occurrences such as ‘ils ne parlaient que polonais’, and all infinitival propositions (‘de ne pas sortir’). We also excluded consecutive repetitions of the negative particle, geminates such as «on entend pas» «on en veut pas, personne n’en veut».

Results Table 2.2 presents the results of the Varbrul analysis. The Polish speakers of French L2 have an 83% deletion rate, which is similar to French L1 rates (Armstrong, 2002; Coveney, 1998). By and large, in the case of the French Poles, the constraint ordering is similar to L1 constraint order. Constraint order is the hierarchical arrangement of the relative strength of factors which affect the variation and can thus produce a fine-grained picture of the structure of variation in speech. While in the case of the Polish speakers in France, the constraint order was relatively similar to native speaker patterns, there were nevertheless some differences. For instance, the factor ‘presence of clitics’ was significant for the Polish speakers, but non-significant for native speakers (Ashby, 1981), and in the case of lexicalisation, the Polish speaker

36

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Table 2.1 Factor groups with examples from the corpus Gender • Male • Female Age • 50+ • −50 Following phonological segment • Vowel • Consonant Preceding phonological segment • Vowel • Consonant Structure of verb • Main • Copula • Modal/Auxiliary Clause type • Main • Subordinate Sujet • Pronoun • Noun • Zero Object clitic • Zero • Present Lexicalisation • Non-lexicalised phrase • Lexicalised phrase Length of residence Post WW2 (1945) • . . .. . .. . ... (1981–1983) • Recent (1989, fall of communism)

Maintenant il est pas bien avec Atena Je savais pas quoi Tu fais pas la tête pour ça Elle a pas dit pauvre Je souviens plus le prix À 5 heures c’est pas possible Je peux pas dire quelle note Je voulais pas repasser et parler Parce que au début c’était pas possible J’ai pas encore fini l’école pour aujourd’hui L’armée de résistance ne voulait pas le reconnaître Ne fermez surtout pas cette institution ! Les agences polonaises sont pas tout à fait honnêtes Le Polonais il se tient pas ensemble Ils téléphonent pas chez nous pour la chercher C’est pas toujours ça ; il faut pas faire ça ; il y a pas ; je sais pas.

The Book seller and the Basket ball Pl ayer: Tales f rom the French Polonia

Table 2.2 Varbrul results: ne deletion in speech of Polish L2 speakers of French Factors Gender • Male • Female Age • 50+ • −50 Following phonological segment • Vowel • Consonant Preceding phonological segment • Vowel • Consonant Structure of verb • Main • Copula • Modal/Auxiliary Clause type • Main • Subordinate Sujet • Pronoun • Noun • Zero Object clitic • Zero • Present Lexicalisation • Non-lexicalised phrase • Lexicalised phrase Length of residence Post WW2 (1945) • . . .. . .. . ... (1981–1983) • Recent (1989, fall of communism)

Occurrences of deletion N : 974

%

p

465 509

84 80

ns Ns

541 433

80 85

ns ns

457 517

81 83

0.379 0.611

308 666

84 78

ns ns

563 235 176

79 88 60

0.374 0.757 0.575

824 150

84 72

0.533 0.350

915 22 37

85 30 37

0.548 0.063 0.448

816 158

80 92

0.445 0.786

620 348

80 87

0.558 0.259

159 815

90 82

0.682 0.418

37

38

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Participants • MARIUSZ • HASSAN • ANNA • Daniel • Elena • Gaby • Henri • Joanna • Kieran • Louis

0.343 0.376 0.793 0.276 0.249 0.226 0.808 0.478 0.891 0.874

Input 0.898 Log likelihood = − 407,865 Significance = 0.073

order is the reverse of L1 speaker order. Similarly, Meyerhof et al. (2009) found, in relation to Polish speakers living in Edinburgh and in London, that whereas many of the factor groups had similar constraint ordering to that of L1 English speakers, some were different. However, although the group results for constraint hierarchies were, by and large, similar to native speaker patterns, on the other hand, the rates are very variable according to individual speakers. Inter individual variation is to be expected in L2 speakers, as noted at the outset, and is frequently found in studies of second language acquisition.

Effect of factors on ne deletion in the speech of Polish L2 speakers of French In this section first the effect of the various factors is discussed in relation to the group results and then two individual speakers are foregrounded.

Length of residence (migration wave) A specifically ‘Polish’ factor group, ‘Wave of migration’ is significant. Those speakers who arrived in France earlier deleted more (0.682) than those who arrived later (0.418). Length of residence is therefore an important constraining factor in the use of ne deletion in spoken French.

Age Despite the fact that the results for the analysis of age as a factor were non-significant, the percentages showed that the tendencies were those of native speakers. Those speakers who were aged less than 50 years omitted ne a bit more than those who were older than 50 (85% vs. 80%). These figures

The Book seller and the Basket ball Pl ayer: Tales f rom the French Polonia

39

approach those of native speakers in relation to age. Ashby (2001), in relation to L1 speech, found that 86% younger speakers deleted ne as opposed to 76% of older speakers.

Gender Likewise, although the factor group ‘gender’ was not significant in the Varbrul analysis, indications show that men use the vernacular variant slightly more than women, as is the case with native speakers, and indeed with other L2 speakers (84% for men as opposed to 80% for women).

Following segment Results for ‘following sound segment’ were in the expected direction: the Polish speakers follow the constraint pattern of native speakers in relation to this factor. They tend to omit ne when the following segment begins with a consonant (0.611). As Ashby (1976) points out, the fact that a following vowel disfavours deletion is in accord with universal CV structure.

Verb Syntactic structure of the verb showed that the Polish speakers omitted ne significantly more when using a copula (0.757) than modals or auxiliaries (0.575).

Clause type This factor was found to be significant and is similar to L1 and L2 usage in French in previous studies: Ashby (1976), for L1 speakers, found main clause caused deletion at 0.70 and subordinate clause at 0.40, and Regan (1996), for L2 speakers who spent a year in France, found main clause deleted at 0.64 and subordinate clause at 0.32. The Polish speakers in the present study show the same constraint order: main 0.53 subordinate 0.35.

Subject type The range in the figures (from highest to lowest) in relation to subject type is significant. If the subject is a pronoun, 0.548, but if a noun phrase, 0.063. The speakers frequently used imperatives which favoured retention of ne 0.448. Table 2.3 is a comparison of the Polish speakers ne deletion in relation to noun phrase with other L2 speakers (Regan, 1996) and L1 speakers (Ashby, 1976).

Object clitic The presence of an object clitic favoured omission of ne (0.786). In relation to L1 speakers, Ashby (1976) did not find this factor to be significant but Regan (1996) in relation to L2 speakers did find it was significant. L2 speakers may well be behaving differently in relation to clitics, perhaps for processing reasons.

40

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Table 2.3 Comparison of Polish speaker variable use of subject type with L1 and other L2 speakers

Noun phrase Pronoun No subject (imperatives)

L1 speakers Ashby (1981)

L2 speakers (Irish) Regan (1996)

L2 speakers (Polish)

0.28 0.64 N/A

0.02 0.53 N/A

0.063 0.548 0.448

Table 2.4 Ne deletion in L2 speech and lexicalisation

Non-lexicalised phrases Lexicalised phrases

Irish speakers Regan (1996)

Polish speakers

0.33 0.78

0.558 0.259

Lexicalisation Interestingly, the results for deletion of ne in lexicalised phrases (Table 2.4) run counter to what has been previously found in relation not only to L1 speakers (Ashby, 1981) but also for L2 speakers (Regan, 1996). As for previous studies, the factor group itself was significant in the Polish L2 data. However, unlike the speakers in the previous studies, the Polish speakers retained ne in the lexicalised phrases. It is possible that unlike the Irish speakers the Polish speakers use the formal version of L1 ‘chunks’. It is also true to say that pragmatic factors can play a role in use of chunks and at times ne is retained for emphasis. It is not impossible that these adult speakers who are frequently talking about their futures and their childrens futures might have used more emphasis than Irish student speakers.

Individual variation As mentioned earlier, in common with all groups of L2 speakers there is considerable individual variation in the rates of variant usage. The Polish speakers vary considerably in their rates of ne deletion, a point which will be discussed in greater detail later.

Insights from individual conversations Both the pre-interviews in Polish and the main interviews in French revealed much about stances of the speakers regarding France, the French,

The Book seller and the Basket ball Pl ayer: Tales f rom the French Polonia

41

their own place in France, their identity, their attitudes to the French language and their own language acquisition process. The Poles who came to France during the Solidarity movement intended to settle permanently in France and build a life for themselves there. Their wish to integrate was very strong and they felt that learning French was crucial in this enterprise. Most of the speakers in our study had little or no French when they arrived in France. They commented that it took them a year or two to acquire enough French to cope with daily life and another couple of years to feel comfortable interacting with French people. The few who had taken some French in school in Poland found that the school French they had learnt was very different from the French they found themselves using in France and so they virtually had to start over in the learning process. They actively enhanced their learning process by watching television, reading newspapers and magazines, interacting with French people as much as possible, avoiding Polish speakers and taking language courses. Many found the process difficult and frustrating at times: « surtout qu’on pouvait pas communiquer (. . .) c’était horrible c’est vrai pendant deux trois premiers mois où (. . .) on a tas de choses à dire aux gens et puis eux ils ont tas de choses à nous dire aussi et bon finalement y a toujours un blocage parce que bon on était avec des dictionnaires pendant deux mois on se déplaçait toujours partout avec des bouquins comme ça dans un sac parce qu’on pouvait pas faire autrement » [female speaker] « au début sans travail je savais pas parler (. . .) les premiers mois on est comme les nouveaux nés ». [Elena] An initially striking feature of the informants’ reflections was their own feeling that they are perceived by the French as a ‘model minority’ type group. Many told us of the good reputation Polish people have in France and their feeling that this perception facilitated their integration into French society. They felt that the French perceived them as a hardworking community who have contributed to the economy of the region without adding to crime levels or compromising the public and social good. « ils voient que ces immigrés polonais (. . .) ils sont catholiques pratiquants plus que nous les Français / et ils sont justes ils sont bien élevés les hommes sont bien habillés ils se bagarrent pas ils se cultivent et ils se forment et en plus ils reviennent avec des diplômes ». (Alfred)

42

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Discussion The quantitative results revealed an interesting general picture of the speech patterns of these Polish L2 speakers of French. We now know that, by and large, the Polish speakers are following the general patterns of French L1 speakers, both at the levels of rates and of constraint ordering. The factors which affect the variation of ne deletion in the speech of native speakers are the same ones which constrain the variation in the speech of the Polish informants, with a few exceptions. So, we have a detailed picture of the variation patterns among the Polish speakers of French. This tells us that the Polish speakers are accommodating to French norms (whether consciously or unconsciously). These general results are invaluable information in our investigations. Amongst many other findings, they confirm our hypotheses that length of residence is important in the adoption of vernacular speech norms. They also indicate that proficiency affects use of vernacular variants. The fact that the speakers as a group are accommodating in so wholesale a fashion both in rates of deletion and in patterns of constraint hierarchies, seems wholly in keeping with the speakers’ attitude towards France, the French, and their view of French perception of themselves as a model minority. The figures resulting from the Varbrul analysis provide a wealth of empirical detail about speech patterns of the speakers. Despite the overall patterns that quantitative analysis has clearly demonstrated, a close look at individual deletion rates reveals a high degree of variability. When we look at deletion rates for the whole group of speakers, we find interesting facts in relation to those individuals at either end of the continuum. Whereas for the group as a whole we find that the longest stay in France correlates with the highest deletion rates and that those most proficient in French also delete most, we also find that the figures for two individuals run counter to this direction. At the two ends of the continuum, we see that the bookseller (Mariusz), a very high proficiency speaker and who has the longest length of residence, has, in fact, the lowest deletion rate, and on the other hand, the basketball player (Anna), who has been in France for less time, is deleting at a very high rate. Anna’s high deletion rate is all the more interesting as women tend to have lower rates of vernacular variants in the case of stable variables, such as ne deletion. The intriguing case of these two speakers prompted a further investigation into the trajectories of their lives; qualitative evidence provides further information about these lives and the choices of the individuals. Brief portraits of the two speakers show contrasting experiences and profiles. In France, they have very different occupations and their interests, aspirations and aims are very different. The bookseller presents himself as a

The Book seller and the Basket ball Pl ayer: Tales f rom the French Polonia

43

highly educated person; he was a directeur des études in an engineering school in France, and in Poland, he had been awarded a doctorate in engineering in automatisme. He was born in 1948 and arrived in France in 1983. He first left Poland to teach his speciality (informatique) in Algeria. Having an advanced educational level, after retirement he decided to open a Polish bookshop in Mitre in 2003 with his daughter-in-law who complains of a « désert culturel » in the region. He has a near native proficiency level of French (only noticeable were some omission of articles and non-liaisons in a few phrases). He generally uses a relatively standard register of French. He left Poland, seeking a better economic situation. Now in France, he runs the Polish book shop in Mitre. He presents himself as the embodiment of Polishness: he is an expert in Polish literature, Polish history, Polish culture, and his shop is a hub of Polish social life. During the two hours we spent talking to him, people stopped in to say hello. He was a great storyteller and told many stories about the history of Polish migrants in France. He laments the lack of knowledge and interest in Polish heritage on the part of Polish people. He talks of those who ‘pretend’ to be Polish and yet speak an ‘inferior’ Polish: ‘ce..n’est qu’« un argot de rase campagne » he says. Or later, when speaking to a Polish academic compatriot: « madame si avec un tel polonais vous allez donc là-bas au Congrès où il y a que des professeurs de votre niveau je vous conseille quand même de le présenter en français ». Here we see him alluding to a high standard of Polish which the average, or even the educated, French Pole does not reach. Better to speak French at the conference than present in substandard Polish. He reports his speech in formal French and uses subordinate clauses, a rarity in spoken French (‘si, avec un tel polonais..’). He sees his function and aim as that of setting standards: leading and encouraging people to discover their Polish heritage and identity. He talks about the history of Poland, and informs us of various facts about history and language during the conversation. He prides himself on his education and knowledge of Polish: « mais vous savez quand je lis la poésie française / et la poésie polonaise je trouve le vocabulaire de la poésie française bien plus modeste que la poésie polonaise notre langue est difficile mais a une richesse de vocabulaire telle (. . .) mais grâce à ça justement elle peut transmettre de la finesse que j’ai trouvé dans aucune autre langue moi moi je suis fier d’être Polonais c’està-dire d’abord de parler une langue compliquée pour apprendre ensuite les langues qui sont toutes plus simples que la mienne ». In fact several of those speakers with a higher level of education in Polish obtained in Poland before arrival in France perceive Polish as ‘above’ French

44

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

in terms of difficulty. He talks of his ambition to build a house back in Poland and retire there. In contrast, Anna, the basketball player, is a more recent economic migrant. She has two children, and is very focused on her future life and that of her family. Whereas the bookseller has invested heavily in intellectual and cultural domains to negotiate his trajectory through migration, she has invested in sport. It is her physical prowess which was her passage to her new country and life. She and her husband (both basketball professionals) came to France as soon as it was permissible. Under the communist government it was impossible to leave Poland as an athlete before the age of 27. They came with a 2-year contract intending to return. At the time of coming they had a 6-year-old daughter. Their son, a second child, was born in France. Despite the elder child’s initial reluctance to speak French, the children became the reason for the couple’s remaining in France. They found the French welcoming, by and large, and invested heavily in making their lives work in their new country. Anna talks about the fact that it is perfectly possible for her to stay in touch with the new technologies such as social media as well as the older ones like television and radio. However despite her access to these technologies, she maintains that she does not really keep in touch with happenings in Poland. In addition, although they make visits to Poland where they have family, she still comments that the links between the relatives are relatively tenuous. Her focus is squarely on France and her family’s future in France. Where Mariusz talks about building a house in Poland, and discusses construction, Anna talks about builders she is employing to work on her house in France. The two speakers, Anna and Mariusz, are positioned in opposite directions in space and time. Mariusz orients himself towards Poland, and wants to take people back to Poland and its cultural and literary past. Anna orients herself towards contemporary France and the future. She speaks a casual contemporary French and has a high rate of ne deletion. Whereas Mariusz told us he wanted to return to Poland, Anna seems totally caught up in making a future for herself and her family in France. Between them they represent a Janus-like positioning on the part of migrants in general; on the one hand looking back towards the lost country and on the other looking towards the future in the new receiving country. These two individuals represent, at a local level, structural patterns which characterise migration as a whole.

Conclusion Our detailed Varbrul analysis shows, as expected, that Length of Residence is a crucial predictor of ne deletion. However, when we looked

The Book seller and the Basket ball Pl ayer: Tales f rom the French Polonia

45

closely at individuals we found that there were some whose deletion rates did not correspond exactly to this generalisation, with two individuals standing out from the overall general pattern. This contrasting pair of speakers, at the opposite ends of the quantitative continuum, is an interesting example of how qualitative data and ethnographic detail provide complementary ‘layers’ of individual lives. Although the quantitative data demonstrate general patterns of L2 acquisition, and identify exceptions to these general patterns, it is the qualitative data that provide explanation for the anomalies revealed through analysis of the quantitative data. The two research processes each contribute something to the ultimate picture we build up of language use and how it relates to peoples’ life journeys. Quantitative data provide an accurate picture of a group of speakers’ language patterns. Qualitative analysis provides the personal data that makes each individual a unique case. Combing the quantitative and qualitative data shows how individuals can stand out against the background of the broader social structural categories, depending on the stances they take. These two contrasting speakers demonstrate that, whereas overall categories are a useful predictor of patterns of L2 acquisition, individuals are more than simply exemplars of such overall categories. Their personal biographies and current relationships with others in their society lead to significant variations from overall patterns. Our qualitative findings do not undermine earlier quantitative findings that LOR may be the best predictor of use of vernacular variants, but demonstrate that these generalisations are true only at a general level. Other factors may be important, as humans are complex and we cannot explain what they do by reference to broad social and linguistic generalisations. Each person has their own story which helps account for outcomes in their language usage. In the case of the basketball player and the bookseller, we have a generalisation which is true, but we discover on closer analysis that we must allow for individual variation explained by their individual stories: their stances in particular contexts, the construction of various identities, their aims and ambitions for the future. The reality of life is ‘kaleidoscopic, complex and complicated, often a patchwork of overlapping activities’ (Blommaert & Dong, 2010). Qualitative analysis helps us to capture more closely the complexity of this reality. This individual variation may be more significant for migrant speakers, who are more mobile than L2 learners in previous studies. The language repertoires of migrants have recently been described as ‘emergent’ (Blommaert, 2010). The use of ne deletion is part of the linguistic repertoire available to the L2 speakers, to be used, perhaps not always in the same way as by L1 speakers, but for different functions. The bookseller is indexing a

46

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

cultured self – a cultured Polish self, paradoxically, and the basketball player is using a highly native-like feature (ne deletion) to index a ‘young very native French’ identity. She is using a native form but for a slightly different function – to index identity. She is using language as a resource, in fact. The basketball player is using a specific ‘bit’ from her repertoire to fulfil a particular need; to present herself as a contemporary French person with a future in France. In fact both speakers are using variation patterns of ne deletion in an emblematic fashion rather than simply a communicative one. So it may be true in the generality that L2 acquisition and, specifically, the adoption of vernacular norms is an indication of integration but, in the specific, individual cases may run counter to the generalisation, as in these cases of language practices as indicators of identity rather than integration.

Acknowledgements The author thanks Niamh Nestor and Isabelle Lemée who read drafts of this chapter. Any errors are completely the author’s own.

Notes (1) This paper is an account of a subset study of a project entitled, ‘Second language acquisition and native language maintenance in the Polish Diaspora in Ireland and France’. Principal investigator, David Singleton, Co-investigators Vera Regan and Ewelina Dabaene, funded by the IRCHSS (Irish Research Council for the Humanities and Social Sciences. 2006–2009). (2) Both interviewers gauged proficiency by agreeing on criteria of high or medium proficiency, using use of articles and use of verb tenses, both problematic to Polish L1 speakers of French L2, as general indicators. (3) The interviews were conducted by Regan and Debaene during three field trips to Mitre and Paris (May–Oct. 2007). (4) The computational analysis was carried out by Lemée and Regan.

References Adamson, H.D. (1988) Variation Theory and Second Language Acquisition. Washington D.C.: Georgetown University Press. Adamson, H.D. (2009) Interlanguage Variation in Theoretical and Pedagogical Perspective. New York; London: Routledge. Armstrong, N. (2002) Variable deletion of French ne: A cross-stylistic perspective. Language Sciences 24 153–173. Ashby, W. (1976) The loss of the negative morpheme ne in Parisian French. Lingua 39, 119–137. Ashby, W. (1981) The loss of the negative particle in French. Language 57, 674–687. Ashby, W. (2001) Un nouveau regard sur la chute du ne en francais parle tourangeau: s’agit-il d’un changement en cours? Journal of French Language Studies 11, 1–22.

The Book seller and the Basket ball Pl ayer: Tales f rom the French Polonia

47

Bayley, R. (2002) In variation in the group and the individual: Evidence from SLA and language shift. Paper presented at the 14th Sociolinguistics Symposium, Ghent. Bayley, R. and Preston, D. (1996) Variation in Second Language Acquisition. Amsterdam: Benjamins. Bayley, R. and Regan, V. (2004) The acquisition of sociolinguistic competence: Special issue. Journal of Sociolinguistics 8(3). Blommaert, J. (2010) The Sociolinguistics of Globalization. Cambridge: Cambridge UP. Blommaert, J. and Dong, J. (2010) Ethnographic Fieldwork: A Beginners Guide. Bristoel: Multilingual Matters. Cedergren, H. and Sankoff, D. (1974) Variable rules: performance as a statistical reflection of competence. Language 50(2), 333–355. Coveney, A. (1998) Awareness of linguistic constraints on variable ne omission. Journal of French Language Studies 8, 159–188. Dewaele, J-M. and Furnham, A. (2000) Personality and speech production: A pilot study of second language learners. Personallity and Individual Differences 28, 355–365. Dewaele, J-M. and Regan, V. (2002) Maitriser la norme sociolinguistique en interlangue francaise: le cas de l’omission variable de ‘ne’. Journal of French Language Studies 12, 123–148. Ellis, R. (1994) The Study of Second Language Acquisition. Oxford University Press. Guy, G. (1988) Advanced Varbrul analysis. In K. Ferrara (ed.) Linguistic Change and Contact (pp. 124–136). Austin: University of Texas. Guy, G. (1993) The quantitative analysis of linguistic variation. In D. Preston (ed.) American Dialect Research. Amsterdam: Benjamins. Habielski, R. (1995) Emigracja. Warsaw: WSiP. Meyerhof, M., Schleef, E. and Clark, L. (2009) Sociolinguistics and Immigration: Variation among Polish Adolescents Living in the UK, NWAV 38. University of Ottawa. Meyerhoff, M. (2006) Introducing Sociolinguistics. Routledge. Myles, F. and Mitchell, R. (1998) Second Language Learning Theories. London New York Sydney: Arnold. Paolillo, J.C. (2002) Analyzing Linguistic Variation. Stanford: CSLI Publications. Center for the Study of Language and Information. Pennycook, A. (2012) Language and Mobility. Bristol: Multilingual Matters. Ponty, J. (2004) Immigration. Au XXe siecle, phenomene nouveau, l’immigration devient un enjeu dans les relations internationales. In Kaleidoscope Franco-Polonais, pp. 123–129. Paris: Instytut Adam Mickiewicz. Poplack, S. and Tagliamonte, S. (2001) African American English in the Diaspora. Oxford: Blackwell. Rand, D. and Sankoff, D. (1990) Goldvarb Version 2: A Variable Rule Application for the Macintosh. Centre de recherche mathématiques, Université de Montréal. Regan, V. (1995) The acquisition of sociolinguistic native speech norms: Effects of a year abroad on second language learners of French. In B. Freed (ed.) Second language Acquisition in a Study Abroad Context, pp. 245–267. Amsterdam; Philadelphia: Benjamins Publishing Co. Regan, V. (1996) Variation in French interlanguage: A longitudinal study of sociolinguistic competence. In R. Bayley and D. Preston (eds) Second Language Acquisition and Linguistic Variation, pp. 177–203. Amsterdam and Philadelphia: Benjamins Publishing Co. Regan, V. (2004) The relationship between the group and the individual and the acquisition of native speaker variation patterns: A preliminary study. IRAL International Review of Applied Linguistics in Language teaching, 42, 335–348.

48

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Regan, V. (2013) Variation and second language acquisition. In J. Herschenson and M. Young-Scholten (eds) The Cambridge Handbook of Second Language Acquisition (pp. 272–291). Cambridge: Cambridge University Press. Regan, V. (forthcoming) In Linguistic Vitality and the Polish community in France. Royal Irish Academy. Robinson, J., Lawrance, H. and Tagliamonte, S. (2001) Goldvarb 2001, online at http:// www.york.ac.uk/depts/lang/webstuff/goldvarb/ Sankoff, D. (1988) Variable rules. In U. Ammon, N. Dittmar and K. Mattheier, J. (eds) Sociolinguistics: An International Handbook of the Science of Language and Society, pp. 984–997. Berlin: Walter de Gruyter. Sankoff, G. and Vincent, D. (1977) L’emploi productif du Ne dans le français parlé à Montréal. Le francais moderne XLV, 243–254. Sankoff, G. and Vincent, D. (1980) The productive use of ne in spoken Montréal French. In G. Sankoff (ed.) The Social Life of Language, pp. 295–310. Philadelphia: University of Pennsylvania Press. Tagliamonte, S. (2006) Analysing Sociolinguistic Variation. Cambridge: Cambridge University Press. Tagliamonte, S. (2012) Variationist Sociolinguistics: Change, Observarion, Interpretation. Wiley-Blackwell. Walker, J.A. (2010) Variation in Linguistic Systems. New York; London: Routledge. Young, R. (1991) Variation in Interlanguage Morphology. New York: Peter Lang. Young, R. and Bayley, R. (1996) VARBRUL analysis for second language acquisition research. In R. Bayley and D. Preston (eds) Second Language Acquisition and Linguistic Variation, pp. 253–306. Amsterdam, Philadelphia: John Benjamins. Zalinski, H. (2005) Polacy we Francji. Historia i wspolczesnosc. Konspekt.

3 The Positional Distribution of Discourse like – A Case Study of Young Poles in Ireland1 Niamh Nestor

Introduction Ireland has experienced momentous change over the last decade and a half, with a sharp reversal in its traditionally outward-migration patterns to a very rapid increase in inward-migration, particularly noticeable after the accession of ten new EU Member States, including Poland, on 1 May 2004. Mac Éinrí and White have characterised Ireland’s experiences of migration as ‘unique, at least in European terms’ (2008: 153), and the impact of this population change has been particularly striking in the pupil profiles of Irish schools. At present, approximately 12% of primary school pupils (Naughton et al., 2010) and 9% of post-primary school students were born outside of Ireland (Department of Education and Skills and the Office of the Minister for Integration, 2010). Since 2008, Ireland’s economic woes have been welldocumented, both nationally and internationally, and the onset of recession has seen a return to double-digit unemployment figures and net outward migration. This has led many to assume, perhaps simplistically, that the migrants who arrived in their thousands since the mid-1990s will now decide to return ‘home’ – in the popular consciousness, their country of birth. The recent publication of data from Ireland’s 2011 Census implies the opposite. In the case of Polish nationals, there has been an almost 100% increase, from 63,276 Polish nationals in 2006 to 122,585 in 2011. There are also increases reported for other non-Irish nationalities (e.g. Lithuanian, Latvian, Nigerian, 49

50

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Romanian, Indian, etc.), indicating that, contrary to public opinion, many migrants have chosen to remain in Ireland, hence making migration a ‘permanent feature of Irish society’ (Ní Chonaill, 2010). Other studies have shown that the presence of children in families has a significant impact on a family’s decision to remain or return, and that migrant parents do not wish to disrupt their children’s schooling, often for a second or subsequent time, despite the challenging economic situation they may find themselves in (cf. Ryan & Sales, 2011; Suárez-Orozco & Suárez-Orozco, 2001). There is a growing body of scholarship on migration into Ireland, with a particular focus on integration and social cohesion. Within this work, there is a dearth of material on young people (Mac Éinrí & White, 2008; Ní Laoire et al., 2009; Smyth et al., 2009). Although some research is available,2 e.g. on migrant children’s wellbeing (e.g. Fanning et al., 2011), migrant children’s experiences at school (e.g. Devine, 2009; Kitching, 2011), immigration as understood by children from the host community (e.g. Devine & Kelly, 2006; Devine et al., 2008), the linguistic and social worlds of migrant children (e.g. McDaid, 2011; Nestor & Regan, 2011) and the provision of English language support at school (e.g. Nowlan, 2008), there continues to be the need for improved policy-making to be better informed by more evidence-based academic research, with particular reference to enhanced social cohesion and inclusion policies. This is particularly necessary given the fact that non-Irish-born children represent a significant percentage of the school population. Equally, the question of integration into the host community continues to be pertinent. One indicator of integration is language. The acquisition of the language of the majority is necessary to participate at all levels of society – educationally, politically and in the labour market. While research into the acquisition on standard varieties, often acquired in a classroom setting by second language learners, is widely available, there is less research on the acquisition of the ‘non-standard’ aspects of the first language (L1) by second language (L2) learners. One such ‘non-standard’ aspect which is receiving increasing attention is the discourse marker3 like. Sankoff et al. (1997: 191) claim that the fluency with which a speaker uses discourse markers is an indication of how integrated he/she is into the local speech community. This chapter investigates the use of discourse like by young Polish L2 speakers and aims to contribute to the growing literature on young migrants living in Ireland, particularly from the perspective of the acquisition of sociostylistic variation. The L2 variety of English under discussion in this chapter is Irish English.4 The research participants are young Poles aged 9–19 who have migrated to Ireland since Poland’s accession to the European Union in 2004 and who are in full-time education in Ireland. The chapter is a case study of six of

The Posit ional Distr ibut ion of Discourse like

51

these participants and their use of discourse like, particularly its positional distribution. The data form part of a larger study in which I analyse a corpus of second language speech data for the production of a number of different variables, including discourse like. The research is as yet at a very early stage which prevents the presentation of a fine-grained analysis. However, in future analyses, I hope to contribute to the growing literature on the acquisition of variation in naturalistic contexts, particularly the acquisition of a number of different variables, including discourse like, by L2 speakers in Ireland. Both academic and non-academic sources confirm that discourse like is one of the most ‘salient features of present-day vernacular English’ (D’Arcy, 2005: ii), and it is held up as an example of ‘a burgeoning change in progress’ (Levey, 2006: 413). What is also noteworthy is that the spread of discourse like throughout the global varieties of English ‘affords valuable opportunities to chart the transnational spread of a discourse-pragmatic innovation’ (Levey, 2006: 413). Much of the academic research on discourse like has been carried out on L1 speech and I review some of this work below. This chapter, however, concerns itself with L2 speech and the use of discourse like by young Poles learning English as an L2 in Ireland. A preliminary analysis of the speech data has shown that discourse like is a salient feature of the speech of many of the young Poles in the broader study but what interests us in this paper is the way in which discourse like patterns in the speech of a small group of these L2 learners. I focus on whether or not they are using discourse like, and, if so, how they are using it, and whether this use corresponds to L1 patterns of usage. Hickey (2007: 376) notes that like is present in the speech of all age groups in Ireland. Therefore, we expect that the young Poles in this study would be sufficiently exposed to the use of like in the L1 Irish English they hear. Whether they are acquiring the patterns of variability of the wider L1 English speech community is what is ultimately under investigation. There is not, however, a substantial body of research on L1 use of like in Irish English (although see Amador Moreno, 2012; Kallen, 2006; Nestor et al., 2012; Schweinberger, 2012; Siemund et al., 2009). By considering what information there is available on L1 Irish English, I attempt to compare L2 usage with L1 input. The chapter is divided into a number of sections. First, I review some of the literature on discourse like, with a particular focus on the use of discourse like in Irish English. I then briefly outline a recent history of Polish migration to Ireland and introduce the participants in this study. Next, I describe the methodologies and coding schema. Following that, I present the results of the ethnographic questionnaire which formed part of the broader study and I analyse a small sample of speech data for the occurrence of discourse like. Finally, I draw some preliminary conclusions.

52

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Discourse like Discourse like in Irish English, as in other varieties, has been receiving increasing attention both within and outside of academia. Many studies note the disregard with which discourse like is met: it has been described as ‘nonstandard’ and ‘intrusive (. . .) [making] sentences seem disjointed to many listeners’ (Underhill, 1988: 234), as occurring ‘grammatically anywhere’ (Siegel, 2002: 64), as ‘a randomly occurring item devoid of semantic or pragmatic significance and functioning as a mere filler where speakers have difficulty encoding their message’ (Miller & Weinert, 1995: 366), and as ‘a meaningless interjection or expletive’ (Dailey-O’Cain, 2000: 60) which is deemed to ‘indicate either vagueness in thinking or general inarticulacy’ (Levey, 2003: 24). In Ireland, as elsewhere, the belief that American English has negatively influenced Irish English and other varieties of English, as well as other more specific attitudes to like, are both propagated by and reflected in media reports. An article on language use in an Irish daily broadsheet described like, along with other lexical items such as all, whatever, and my bad, as ‘the instant noodles of everyday language: bland, readily available and requiring the minimum of effort’ (The Irish Times, 2009). The article claims that we use these words because we are lazy, ‘when we can’t be bothered to engage our brains’. In this case, not all that unfamiliar, the media seems to be drawing on the broader narrative of the perceived increasing Americanisation of Irish culture and culture more generally. Meyerhoff and Niedzielski (2003: 535) claim that this type of response is ‘familiar’. The US is perceived as the powerful and dominant partner in globalised exchanges of all kinds and, in line with this, American English is ‘discursively constructed as a “category killer”, that is, a competitor that enters a market, “with so much buying power that they almost instantly kill the smaller competitors”’ (Klein, 2000: 134, cited in Meyerhoff & Niedzielski, 2003: 535). Like is often believed to have been recently ‘invented’ by young people in its current form. It is associated with adolescents, particularly adolescent girls, and it is commonly proffered as an abundantly available example of the inarticulacy of youth (however, see D’Arcy (2007) for a comprehensive dismantling of the many myths around the development and usage of like). Previous research has challenged the many myths around like and has demonstrated that, in fact, it has a long history in the English language. D’Arcy (2008) provides references to like from the early part of the nineteenth century and points out that first-generation native New Zealanders who were born between 1851 and 1919 used like (D’Arcy, 2007, 2008). Currently, in England, Ireland and Scotland, seventy- and eighty-year-olds

The Posit ional Distr ibut ion of Discourse like

53

use like (D’Arcy, 2005, 2007, 2008). D’Arcy (2005: ii), in her variationist treatment of discourse like, concludes that, far from being unconstrained by syntax, discourse like is, in fact, highly constrained by syntax and occurs in specific positions among speakers of all ages. Like, in its ‘vernacular’ uses (D’Arcy, 2007: 392), i.e. in its discourse, quotative and approximative adverb functions, has been researched in a number of varieties of English, including American English (e.g. Barbieri, 2007, 2009; Buchstaller, 2001; Cukor-Avila, 2002; Ferrara & Bell, 1995; Fuller, 2003; Romaine & Lange, 1991; Underhill, 1988), Canadian English (e.g. D’Arcy, 2004, 2005, 2007, 2008; Tagliamonte, 2005; Tagliamonte & D’Arcy, 2004; Tagliamonte & Hudson 1999), Scottish English (e.g. Macauley, 2001; Miller & Weinert, 1995), London English (e.g. Andersen, 1997, 1998; Levey, 2003, 2006), Australian English (e.g. Sharifan & Malcolm, 2003; Winter, 2002) and Irish English (e.g. Amador Moreno, 2012; Kallen, 2006; Nestor et al., 2012; Schweinberger, 2012; Siemund et al., 2009). Dailey-O’Cain (2000: 61) points to several studies which conclude that like is undergoing grammaticalisation, a process whereby ‘particular items become more grammatical through time’ (Hopper & Traugott, 2003: 2). In this respect, Romaine and Lange’s 1991 paper considers the case of quotative like in American English in detail. In its more ‘traditional “grammatical”’ forms, like can function as a lexical verb, noun, preposition, conjunction and suffix (D’Arcy, 2006: 339). As noted earlier, like also has a number of ‘vernacular’ uses: the discourse marker, quotative be like, and the approximative adverb (D’Arcy, 2007: 392). The following examples are taken from the corpus under discussion in this chapter: (1) the quotative complementiser be like which is used to introduce constructed dialogue (Tannen, 1986), e.g. (i) Do you know, she was like, “No, do it now”, like. Never, no one ever said that, do you know and, like, it was not, it was not even busy nothin’, like you know (Patryk, 631a). (ii) ...and she, like, em, drawing on the bo- book and the things like that so if somebody is like, em, want to, like, give back her, to do the same thing, and to show her how, like, how is have the, like, book like that, she is always taking the book and she is like, em, “Stop drawing on my books” and things like that (Agnieszka, 179 h). (2) like when used as an approximative adverb. D’Arcy (2008: 126) states there is a process of lexical replacement underway in Toronto English. The apparent time results in her study demonstrate that like when used in approximative contexts is favoured by speakers under 30 and that, in

54

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

fact, about is being ousted by like in numerical contexts among this group, e.g. (i) Yeah, I, I lived here. I lived there, em, for like three years and the-. Before I kind of lived, like, in smaller city outside it. Like, it’s kind of, I dunno, that’s like four thousand people who live there so that’s pretty small (Benedykt, 14d). (ii) Yeah, I live in [house number] fifty-nine, so I’ve like three minutes to school (Bronisław, 870). (3) discourse like. Following D’Arcy (2008), we treat both the discourse marker and the discourse particle in this chapter under the combined heading of discourse like. D’Arcy (2006: 339) states that a discourse marker is clause-initial and functions to relate the utterance (over which it has forward scope) to discourse which has gone before, e.g. (i) Like we didn’t yet talk about this whole, like, to take a bus and, like, this whole, em, stuff, but we did, em, we did the thing about this, how to, like, save energy and, em, like, em, to turn off the, turn the every, like, device when you don’t use it and (sic) (Benedykt, 285a). (ii) Yeah, like, the, the old one give us one homework for, for, for week (Mateusz, 55c). A discourse particle, on the other hand, occurs clause-medially and functions pragmatically (e.g. by indicating ‘a speaker’s epistemic stance to the form of the utterance’) (D’Arcy, 2006: 340), e.g. (i)

Maybe, maybe they boring and have, like, nothing to talk so they thinkin’, “Oh, oh, let’s start saying something like that” (Klara, 1436a). (ii) I think that every person has, like, a talent do you know and some people would be good at drivin’ anyways and some people won’t, no matter how many lessons they have. Just, they just won’t be good drivers (Patryk, 218). In this chapter, I am concerned with the positional distribution of discourse like in the corpus of L2 speech I have collected. As well as discourse like as it occurs in clause-initial and clause-medial position, I also include clause-final position, e.g. (i)

Like, in Poland, you’d be beepin’ the horn all the time and, do you know, usin’ bad language, but, you know, and here people won’t say anythin’, like you know (Patryk, 212b).

The Posit ional Distr ibut ion of Discourse like

55

(ii) Yeah and if, em, I will be done dancing, just, I will be never, like, go to Warsaw or Gdan ´ sk. That’s- I was two three times in Warsaw, one time in Gdan ´ sk. That’s like- We will be never, like, with my family go, like (Agnieszka, 477d). (iii) I just hate when everyone speaks Irish, like. Except the teacher cos, the Irish, it’s kinda weird to me because, em, I don ‘t know any of the words and, you know, when Irish people speak Irish I don ‘t understand them (Bronisław,1034). Schweinberger (2012: 187) notes that Irish English is ‘particularly interesting’ because of the ‘unusual’ positioning of discourse like, namely the frequency with which it is used in clause final position. D’Arcy (2005) considers clausefinal like, which she describes as ‘the “traditional” (and obsolescing) British pattern, where LIKE takes backward scope’ (2005: 5), and notes (D’Arcy, 2005: 4) that the Oxford English Dictionary documents usage of like in the late eighteenth and early nineteenth centuries, and that, in these records, like ‘generally occurs in clause-final position’ (D’Arcy, 2005: 4). Siemund et al. (2009) analysed the positional distribution of discourse like in different varieties of English and use the ICE (International Corpus of English) components for the purposes of comparability. They conclude that there are two distinct distributional patterns for discourse like: Irish English behaves similarly to Indian English (called the ‘Group A’ varieties) insofar as discourse like mainly occurs in clause-marginal positions (clause-initial or clause-final) (Siemund et al., 2009: 29–30); on the other hand, the ‘Group B’ varieties, Philippine English and East African English,5 prefer discourse like in clause-medial position (Siemund et al., 2009: 30). In later research, Schweinberger (2010) adds British English, American English, Singaporean English and Jamaican English to the Group B varieties. Amador Moreno (2012), in an analysis of discourse like in Paul Howard’s novel The Curious Incident of the Dog in the Night-dress, notes that, contrary to the dominant patterns of positional distribution in Irish English where discourse like appears more frequently in clause-marginal positions and particularly in clause-final position, the majority of the occurrences of discourse like in the novel occur in clause-medial position. In fact, there are only two occurrences of clause-final like and one occurrence of clause-initial like. Amador Moreno notes that Paul Howard, in his rendering of the novel’s characters, would have been particularly aware of the social and geographical differences played out through the variation in Dublin English. She goes on to say that ‘the fact that sentence-medial position like is favoured by the characters in the novel confirms Hickey’s [2005] argument that “fashionable” Dublin speakers actively try to distance themselves from local Dublin speakers’ (2012: 29). In this case study, five of the participants live in rural

56

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

areas outside of Dublin; the remaining participant (Benedykt) is resident in Dublin. In light of the above, it will be interesting to compare the positional distribution of discourse like across the speakers. Schweinberger (2012: 182) notes that the ICE components under analysis in his study are not accompanied by audio recordings and, thus, ‘the classification of various instances of LIKE (…) relies heavily on morpho-syntactic features and meta-linguistic information provided in the transcriptions. Such information encompasses, for example, the presence of pauses or utterance boundaries.’ Hasund (2003) used intonation in order to analyse the discourse markers like in English and liksom in Norwegian (Hasund, 2003: 193): [. . .] when like or liksom occurs on the border between two clausal or nonclausal units, it is sometimes difficult to determine whether the markers occur in initial position in relation to the following unit or in final position in relation to the preceding unit. In such cases I have used intonation to determine the syntactic status of the markers: all the instances that have been classified as final are pronounced with a terminal intonation, in most cases rising, in a few cases falling (both like or liksom). In this study, the audio-files are available and, therefore, I was able to use intonation to dissipate any ambiguity with respect to the positional distribution of each occurrence of discourse like. This was particularly pertinent in determining the position of discourse like at clause-marginal boundaries.

Polish Migration to Ireland There have been previous migration waves to Ireland from Poland, but not by any means as great as that during the Celtic Tiger years. Grabowska (2005: 32) summarises these waves as follows: (i)

Post World War II migration. The Irish government offered approximately 1000 third-level scholarships to Polish people who had been forced to leave Poland. (ii) ‘Solidarity migration’ in the early 1980s. In the aftermath of the imposition of martial law in Poland (1981–1983), the ruling Communist Party only allowed one-way cross-border movement (Regan & Nestor, 2010: 146). (iii) ‘Migration of hearts’ in the mid 1980s. This wave was made up mostly of young Polish women who emigrated to Ireland to marry Irish men. These women often became Irish citizens through marriage.

The Posit ional Distr ibut ion of Discourse like

57

(iv) Post-1997. Migration during the Celtic Tiger boom years, which Grabowska (2005: 32) describes as having a ‘dual character’. Some migrants were outsourced by their multinational or Irish company bosses but most came to Ireland through the process of chain migration. This is a process whereby personal contacts and developed networks of migrants in the destination country lead to an induction process: ‘One migrant inducts another. Whole networks and neighbourhoods leave to work abroad, bringing back stories, money, know-how and contacts’ (Hochschild, 2006: 214). Factors influencing an individual’s decision to migrate are complex (Grabowska, 2005: 31) but the post-1997 migration flow is characterised as primarily economically motivated in nature (Grabowska, 2005: 32). This was facilitated by the fact that Ireland (along with the UK and Sweden) did not require Polish citizens to hold a work permit in order to gain employment after Poland’s accession to the EU in 2004. Previous to the publication of the results of Census 2011, there had been much debate about the actual number of Polish nationals living in Ireland. Census 2006 returned a figure of 63,276, but this was widely disputed by community sources and in the media (cf. Smyth, 2010), both of whom suggested that the number of Poles in Ireland hovered somewhere closer to 200,000. Census 2011 shows an increase of almost 100% in the number of Polish nationals living in Ireland to 122,585, making Poles the largest nonIrish group in Ireland. The total number of non-Irish nationals living in Ireland has risen to almost 545,000, accounting for c. 12% of the total population of the country. This is a substantial increase, amounting to more than a doubling of the estimated 5% of migrants who lived in Ireland before the mid-1990s (OECD, 2009: 15). Within this figure of 545,000 the great majority came from European Union countries (almost 387,000), with 122,585 from Poland, meaning that Poles represent, on paper at least, 22.5% of the total non-Irish population. Of these 122,585 Polish people, 10,011 are aged 0–4 years old, 10,355 are aged 5–12 years old (primary school age-group) and 4925 are aged 13–18 years old (post-primary school age-group) (personal communication, the Central Statistics Office).

The Young Polish People in this Study The young people in this study have all migrated to Ireland from Poland since 2004. The decision to migrate was always taken by one or both parents, sometimes with but sometimes without the agreement of the young people.

58

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

The young people generally live with both parents, and, in some cases, other members of the extended family live in the same house or close by. In other cases, the family remains fragmented due to migration (one parent in Poland and one in Ireland), or has fragmented since arrival. Since leaving Poland, these young people have experienced the various ups and downs of the migration process. They have felt excitement at the prospect of living in a new country, making new friends, attending a new school, and having new experiences. They have also felt homesickness and loneliness, the anxiety brought on by the separation from family and friends, the insecurities associated with ‘fitting in’, and the often tricky dynamics of friendship-forming while maintaining a consistent commitment to old friends in Poland.6 Because these young people are sometimes more proficient in the language of the receiving country, they often become the mouthpieces of their parents in various situations (cf. Suárez-Orozco & Suárez-Orozco, 2001). Since the beginning of Ireland’s economic downturn, many of these young people have shared the anxieties of financial and job insecurity borne by their parents. Not least, these young people have had to carry the often overwhelming burden of coping with the demands of a new curriculum through a second language in a very different school system.

Ethnographic Questionnaire7 The ethnographic questionnaire was designed to elicit biographical, linguistic, socio-psychological and linguistic-educational data. There were two versions of the questionnaire (one for younger children and one for teenagers), and the participants (N = 103; 61 males and 42 females) could complete the questionnaire in Polish or English. The mean age of the participants was 13 years and two months. The mean length of residence in Ireland was one year and five months. The majority of the participants (86%) returned to Poland between one and three times per year, and throughout the interviews, this return to Poland was an important area of discussion. The participants generally displayed confidence in their own language abilities (Table 3.1, i). When considering home language usage (Table 3.1, ii), it is no surprise that Polish is dominant but the results for Englishlanguage usage with family members are interesting. One reason for this could be that some of the young people’s close relatives have non-Polishspeaking partners and therefore English has become one of the languages of communication at home. Equally as unsurprising is the result for English-language usage at school (Table 3.1, iii) but Polish is also used quite

The Posit ional Distr ibut ion of Discourse like

59

frequently, indicating involvement in Polish-speaking networks. This result is further supported by the balance of daily language use (Table 3.1, iv). The participants regard their multilingual language abilities very positively (Table 3.1, v). When asked about national identification (Table 3.2), the majority of the participants chose ‘Polish’, but 20% chose to identify as ‘Irish-Polish’. Two of the participants identified as ‘Other’. One chose a regional identification, while the other identified as something ‘between’ Polish and Irish (evoking Bhabha’s (1994) third space – cf. Nestor and Regan (2011) for a discussion of this case).

Table 3.1 Language First contact with English (mean age) Receiving formal language instruction?

Self-reported language proficiency(i) How well do you speak English? (like a native/ very well/well/quite well/a little/not at all)

Language use in the family(ii) How often do you use Polish with your family? (always/often/sometimes/rarely/never) How often do you use English with your family? (always/often/sometimes/rarely/never) Language use at school(iii) How often do you use English at school? (always/often/sometimes/rarely/never) How often do you use Polish at school? (always/often/sometimes/rarely/never) Language use and bilingualism(iv) Balance of Polish and English in your daily life? How happy are you about the fact that you can speak both Polish and English?(v) (very happy/happy/neutral/unhappy/very unhappy)

8yrs 9mths 65% attending English classes (for the most part English as an Additional Language (EAL) at school) 28% quite well 37% well 17.5% very well 6% like a native 96% always or often 26.2% often or sometimes

84.5% always or often 54% often or sometimes

40% a balanced use of both languages 54% use more Polish

95% very happy or happy with their bilingualism

60

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Table 3.2 Identity How do you identify? (Polish, Irish-Polish, Other)

78% Polish 20% Irish-Polish 2% Other

Scope and Research Methodologies There was a total of 103 young people aged 9–19 involved in the wider study – 61 males and 42 females. Each participant had moved to Ireland since EU accession in 2004. There were two research sites, one urban (Dublin) and one rural, and the research was conducted in the young people’s schools for the most part. On the rare occasion that this was not possible, the researcher gathered data in the young person’s home. Seventy-five of the young people were living in a rural area and 28 lived in an urban area. The difference in sample sizes was attributed to the nature of access in both research sites. The researcher was given permission to conduct research in more schools in the rural area. Of the total number of participants, 77 (45 males and 32 females) agreed to take part in the interview process. The interviews lasted between 35 minutes and 2 hours. In general, where possible, the researcher spoke to the young person for an hour or longer. Sometimes the age of the young person indicated that one hour might be too long and tiring; on other occasions, the natural dynamics of school life meant that the interview lasted for less than one hour. Speakers were recorded in semi-directed interviews (cf. Tagliamonte, 2006) based broadly on Labovian modules; however, for the most part, the interviews were largely guided by the young people’s own interests. In the broader research project, the interview data will be analysed quantitatively and qualitatively for the production of particular linguistic variables. However, in the present chapter, the scope of the analysis will be somewhat more limited owing to the fact that it is in the preliminary stages. The author speaks Polish fluently after a significant amount of time spent in Poland. She translated any Polish used during the interviews or in the ethnographic questionnaire. Her Polish language proficiency also gave her the status of ‘insider’ during the field study period.

Coding There were a number of steps in the coding process. First, I extracted all occurrences of like from the corpus. These were then coded for

The Posit ional Distr ibut ion of Discourse like

61

function: ‘traditional’ categories (lexical verb, noun, preposition, conjunction and suffix; cf. D’Arcy, 2006: 339), discourse like, quotative be like, like when used as an approximative adverb, it’s like (frozen form), like when used as part of a general extender, false starts, incomplete clauses, and crosslinguistic like. Instances of like were coded as a ‘false start’ when the participant started to formulate a clause using like but did not complete this clause. He/she then reformulated in a second complete clause and did not use like. Instances of like were coded as ‘incomplete’ when there was not enough syntactic information available to unambiguously code the occurrence of like. The speaker did not reformulate in a second clause. Instances of like were coded as ‘crosslinguistic’ when it was clear that there was an influence from Polish on the speaker’s formulation in English.8 Occurrences of like which were coded as ‘incomplete’ or ‘crosslinguistic’ were rare in the corpus. In order to achieve a more accurate normalisation for the purposes of the comparability of each speaker’s linguistic performance, the frequency of the occurrence of discourse like was calculated per 1000 words for each speaker (see results section, Table 3.4). As the main focus of this chapter is on the positional distribution of discourse like in the corpus, discourse like was coded for position within the clause. This entailed coding every occurrence of discourse like for whether it occurred in clause-initial, clause-medial or clause-final position. Following Siemund et al. (2009: 23), a clause ‘refers to either a syntactic unit consisting of minimally a subject–verb construction which may be accompanied by syntactically related constituents that modify the subject– verb construction or to elliptic utterances which can be rephrased as subject–verb constructions’. The authors go on to define a clause semantically as referring to ‘basic propositions that cannot meaningfully be segmented further without interfering with the truth conditions of the proposition in question’ (Siemund et al., 2009.). This analysis primarily focuses on the positional distribution of discourse like within the clause. Later research will also focus on the functional role of like in the clause; however, this is beyond the scope of this chapter. I determined discourse like to be in clause-initial position if it occurred ‘before the first obligatory constituent of the clause, i.e. clause-initial LIKE may not be the first linguistic unit of the clause construction since other optional expressions (e.g. other discourse markers) may precede its occurrence’ (Siemund et al., 2009: 23–24). Likewise, discourse like was coded as clause-final if it occurred ‘after the last propositionally functional constituent of the clause’ (Siemund et al., 2009: 24). As noted earlier, intonation was important in determining the position of discourse like, particularly at clause

62

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

boundaries. What follows are examples from this study of discourse like in clause-initial, clause-medial and clause-final positions. •

Discourse like in clause-initial position Like, we had court here so she was playing, like, maybe rounders (Agnieszka, 192c). Like, in metalwork, we, em, now, we d-, we do, like, a car, like (Benedykt, 203a).

•

Discourse like in clause-medial position You know, she’s not, like, givin’ out loads of work (Patryk, 104). We got new science teacher and she’s, like, quite different than the other (Mateusz, 55b).

•

Discourse like in clause-final position Well, sometimes we go in there. We sitting or we walking. It’s nothing to do, so, yeah, we walking around, like (laughs). (Klara, 892) I’m goin’ to Disneyland Paris with my brother, like. He wan-. He never seen one and he wants to go there (Bronisław, 1391)

Finally, for the purposes of comparison, clause-initial and clause-final occurrences of discourse like were grouped together as clause-marginal. Thus, I was able to observe any differences between the use of discourse like in clause-medial and clause-marginal positions. Siemund et al. (2009) found that discourse like in Irish English occurred in clause-marginal position more frequently (58.17%) than in clause-medial position (31.45%).9 I wished to determine whether the young Poles in this study are following similar trends.

Results and Discussion As mentioned earlier, this chapter presents preliminary findings on the use of discourse like, particularly the positional distribution within the clause, by six young Polish L2 learners of Irish English. The questions motivating the analysis were whether or not these young Poles are using discourse like, and, if so, how they are using it, and whether this use corresponds to the positional distribution patterns attested for L1 usage. Table 3.3 provides brief ethnographic details for the participants.

Analysis of discourse like in the speech data Each occurrence of like (in both its discourse and non-discourse functions) was extracted from the corpus. There was a total of 1079 occurrences

The Posit ional Distr ibut ion of Discourse like

63

Table 3.3 Ethnographic details Name10

Sex

Age

Age on arrival

Length of residence

Location

Patryk

Male

18yrs

14yrs 9mths

3yrs 3mths

Rural

Benedykt

Male

14yrs 8mths

11yrs 7mths

3yrs 1mth

Urban (Dublin)

Agnieszka

Female

12yrs 6mths

11yrs

1yr 6mths

Rural

Klara

Female

16yrs 8mths

15yrs 3mths

1yr 5mths

Rural

Bronisław

Male

10yrs 3mths

7yrs 3mths

3yrs

Rural

Mateusz

Male

14yrs 4mths

12yrs 11mths

1yr 5mths

Rural

of like in all functions in this corpus. Of these, 690 were tokens of discourse like, which represents 63.948% of all the occurrences of like (see Figure 3.1).11

1200

Total occurrences of LIKE 1079

LIKE

No. of tokens

1000 800

690

600 400 200 0

LIKE (all functions)

DISCOURSE LIKE

Figure 3.1 Total occurrences of LIKE in the corpus

Interspeaker variation Interspeaker variation is to be expected in the speech of L2 learners and has been well-attested in previous research (cf. Sankoff et al., 1997). Figure 3.2 further illustrates the oftentimes great difference in the usage of discourse like by the various participants in this study. The frequency of the occurrence of discourse like in all functions has been calculated per 1000 words for each speaker. This was deemed a more accurate normalisation for the purposes of the comparability of each speaker’s linguistic performance (see Table 3.4). Interspeaker variation is also clear when the occurrences of discourse like are viewed as a percentage of the use of like in all functions (Table 3.5).

64

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

No. of tokens

Total LIKE and Discourse LIKE 400 350 300 250 200 150 100 50 0

Total LIKE Discourse LIKE

Patryk

Benedykt

Agnieszka Klara Participants

Bronislaw

Mateusz

Figure 3.2 Use of total LIKE and discourse LIKE by participants. Interspeaker variation Table 3.4 Frequency of discourse LIKE per 1000 words

Patryk Benedykt Agnieszka Klara Bronisław Mateusz

Word count Total discourse LIKE [N = 690]

Frequency per 1000 words

9221 3779 5824 2830 4524 2606

29.498 41.281 26.786 29.329 3.979 1.919

272 156 156 83 18 5

Table 3.5 Discourse LIKE as a % of all LIKE

Patryk Benedykt Agnieszka Klara Bronisław Mateusz

Word count

Total LIKE [N = 1079]

Total discourse LIKE [N = 690]

Discourse LIKE as a % of total LIKE

9221 3779 5824 2830 4524 2606

371 258 248 144 41 17

272 156 156 83 18 5

73.315 60.465 62.903 57.639 43.902 29.412

The above data seem to demonstrate that, even where there are very few occurrences of like, discourse like has a relatively significant presence in the L2 speech of these participants. Taking one case from the lower end of the scale, e.g. Mateusz, it seems clear that, although he uses like (all functions) only 17 times during his interview, almost 30% of these are tokens of discourse like.

The Posit ional Distr ibut ion of Discourse like

65

On the upper end of the scale, when Patryk uses like, a striking 73% of these are tokens of discourse like. Further research is necessary to elucidate the differences in the usage of discourse like among these participants and, particularly, to investigate the impact of various factors such as sex, age, length of residence, age on arrival, proficiency, and social networks.

Other functions of like As noted earlier, there was a total of 1079 tokens of like in all functions extracted from the corpus for analysis. There were 690 tokens of discourse like. The remaining 389 tokens of like had different functions, as is illustrated in Table 3.6. Table 3.6 All functions of LIKE Total like Discourse like Other like ‘Traditional’ categories Quotative be like Approximative adverb It’s like (frozen form) General extender Excluded from analysis

1079 690 389 151 9 45 13 38 133

Positional distribution of discourse like I now turn to an analysis of the positional distribution of discourse like in these data. Until these data can be compared to L1 data for the impact of various other factors such as sex and age, a more refined comparison cannot be drawn. For now, using data already available (e.g. Siemund et al., 2009), I will present a preliminary exploration of a correspondence between the speech of the L2 learners in this study and the patterns of positional distribution of discourse like in the L1 speech community. Figure 3.3 shows that discourse like is used is each of the positions indicated by Siemund et al. (2009) – clause initial, clause medial and clause final. If we collapse clause-initial and clause-final into one category – clausemarginal – we arrive at a clearer picture of the clause-medial versus clausemarginal breakdown for this corpus of speech data (see Figure 3.4). Figure 3.4 shows the difference between the number of tokens of clause-medial and clause-marginal discourse like. In their data, Siemund et al.

66

Linguist ic and Cultural Acquisit ion in a Migrant Communit y Positional distribution of discourse LIKE [N = 690] 350

No. of tokens

POSITION

293

300

235

250 200

162

150 100 50 0 Clause initial

Clause medial Position in the clause

Clause final

Figure 3.3 Positional distribution of discourse LIKE Positional distribution of discourse LIKE [N = 690] 450

No. of tokens

400 350

397

MARGINAL AND MEDIAL

293

300 250 200 150 100 50 0

Clause marginal Clause medial Marginal and Medial

Figure 3.4 Marginal and medial positions

(2009: 27) found that speakers in the ICE Ireland corpus used discourse like in a clause-marginal position in 58.17% of cases and in clause-medial position in 31.45% of cases,12 a clear favouring for clause-marginal positions in Irish English. In this corpus, it seems that the L2 speakers are broadly following this pattern: discourse like is used in clause-marginal positions in 57.54% of cases and in clause-medial position in 42.46% of cases. However, these are raw figures. Statistical analysis is needed in order to determine the significance of these findings. Also, five of the speakers are from rural areas outside of Dublin. Therefore, it is necessary to look at the differences between speakers in the positional distribution of discourse like. Table 3.7 shows the number of tokens of discourse like used by each speaker.

The Posit ional Distr ibut ion of Discourse like

67

Table 3.7 Positional distribution of discourse LIKE

Patryk Benedykt Agnieszka Klara Bronisław Mateusz

Word count Total discourse LIKE [N = 690]

Clause-initial Clause-medial Clause-final

9221 3779 5824 2830 4524 2606

53 31 39 31 5 3

272 156 156 83 18 5

27 107 112 45 1 1

192 18 5 7 12 1

There are observable differences in the positional distribution of discourse like between the speakers. Patryk, Bronisław and Mateusz13 all favour clause-marginal positions, while Benedykt, Agnieszka and Klara favour clause-medial position. As noted earlier, further research is necessary to elucidate the differences in the usage of discourse like among the participants and, particularly, to investigate the impact of various factors such as sex, age, length of residence, age on arrival, proficiency, and social networks. For now, it is interesting to note the broader trends in the data. One of the most striking of these is the contrast between Benedykt and Patryk. In this case, it seems that sex has no impact nor does length of residence as both participants had been living in Ireland for approximately three years at the time of the interview. What may be pertinent is location. Patryk lives in rural area; Benedykt lives in Dublin. While Benedykt lives in what would not be considered to be ‘south’ Dublin or, in Hickey’s (2005) terms, an area which would typically be associated with a more ‘fashionable’ variety of Dublin English, he nonetheless favours clause-medial discourse like and, thus, is following the patterns noted in Amador Moreno (2012). However, Agnieszka and Klara, who both live in a rural area, also favour clause-medial position. Whether sex has a more significant impact than location will be further investigated in future analyses of the data.

Conclusion This chapter focused on the positional distribution of discourse like in the L2 speech of young Poles living in Ireland. A preliminary analysis of the speech data has shown that discourse like is a salient feature of the speech of many of the young Poles in the broader study. Broad patterns in the data would suggest that the L2 speakers are adopting trends of the

68

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

wider L1 speech community. A more detailed analysis elucidates notable differences between speakers and indicate that sex and location may have a potentially significant impact. As other research in second language acquisition has shown, there is a large degree of intra- and interspeaker variation in L2 speech. This is no different in this study. Sankoff et al. (1997) commented on the ‘extreme variation’ (p. 202) in the use of discourse markers in French by the participants in their study. This ranged from participants who used almost no discourse markers to those who used discourse markers effortlessly. We can observe the same in this small sample here. If we extend here Sankoff et al.’s (1997) claim that the fluency with which a speaker uses discourse markers is an indication of how integrated he/she is into the local speech community, then we can hypothesise that social networks are likely to play a key role in the acquisition of sociostylistic variation and, furthermore, that the personal motivation to mark one’s identity in English as being part of these networks could also a factor. The analyses presented above are still in the preliminary stages and whether other factors are significant will become clear once a more complete analysis has been carried out.

Notes (1)

(2)

(3) (4) (5) (6)

(7)

This research has been made possible through the support of the Irish Research Council for the Humanities and Social Sciences. I would like to thank the Central Statistics Office for their generous assistance with population statistics. I would also like to thank the reviewers. Needless to say, any errors are my own. This list is by no means exhaustive. See Mac Éinrí and White (2008) for a review of research available on migrant children in the Irish context. A more recent source of research on migrant children in Ireland can be found in Darmody et al. (2011). In this analysis, I follow Schiffrin’s description of discourse markers as ‘sequentially dependent elements that bracket units of talk’ (1987: 31). Ireland is officially a bilingual country. Irish and English are recognised as its official languages. This research focuses on English. However, for research on the use of discouse like in Irish (as an L2), see Ní Chasaide and Regan (2008). East African English includes both Kenyan English and Tanzanian English. For a discussion on emotions and human mobility, see the 2008 Special Issue of the Journal of Intercultural Studies, 29(3) entitled ‘Transnational families: Emotions and belonging’. Also, Suárez-Orozco and Suárez-Orozco (2001) devote a chapter to the psychosocial impact of immigration. These results have been reported elsewhere – cf. Nestor and Regan (2011). The ethnographic questionnaire is designed to elicit important background information about the participant, particularly the kind of information which variationist sociolinguists believe may have an impact on the way the person speaks, e.g. sex, age, length of residence in the host country, previous contact with English, use of Polish and so on.

The Posit ional Distr ibut ion of Discourse like

69

(8) What follows is an example of the use of like in the corpus which was coded as crosslinguistic like. It is clear from the Polish translation that the speaker has directly translated ‘jako’ here into English as ‘like’: (ENG) I think I still treat it like a slang. (6-m-103-70a) (PL) Mys´le˛, z˙e cia˛gle traktuje˛ to jako slang. (9) Siemund et al. (2009: 42) explain that the percentages do not add up to 100% as they omitted all ‘unclassifiable cases of LIKE’. In this analysis, I focus on only the classifiable cases of LIKE. Therefore, the percentages add up to 100%. (10) To protect the privacy of the participants, pseudonyms have been used throughout. (11) Siemund et al. (2009: 25) found that discourse like represented 46.53% of all cases of like (all functions) in their data. (12) See footnote 9. (13) Bronisław and Mateusz use very few tokens of discourse like and, thus, I have decided to focus on the other four participants for this section of the analysis.

References Amador Moreno, C.P. (2012) A corpus-based approach to contemporary Irish writing: Ross O’Carroll-Kelly’s use of like as a discourse marker. International Journal of English Studies 12 (2), 19–38. Andersen, G. (1997) ‘They gave us these yeah, and they like wanna see like how we talk and all that’: The use of ‘like’ and other discourse markers in London teenage speech. In U-B. Kostinas, A-B. Stenström and A-M. Karlsson (eds) Ungdomsspråk i Norden. MINS 43 (pp. 83–95). Stockholm: Stockholms Universitet, Institutionene für Nordiska Språk. Andersen, G. (1998) The pragmatic marker ‘like’ from a relevance-theoretic perspective. In A. Jucker and Y. Ziv (eds) Discourse markers: Descriptions and theory (pp. 147–170). Amsterdam and Philadelphia: John Benjamins. Barbieri, F. (2007) Older men and younger women: A corpus-based study of quotative use in American English. English World-Wide 28, 23–45. Barbieri, F. (2009) Quotative be like in American English. Ephemeral or here to stay? English World-Wide 30 (1), 68–90. Bhabha, H. (1994) The Location of Culture. London: Routledge. Buchstaller, I. (2001) An alternative view of ‘like’: Its grammaticalization in conversational American English and beyond. Edinburgh Working Papers in Applied Linguistics 11, 21–41. Central Statistics Office, Ireland (CSO). Online document: http://www.cso.ie/statistics /nationalityagegroup.htm. Cukor-Avila, P. (2002) ‘She say, she go, she be like’: Verbs of quotation over time in African American Vernacular English. American Speech 77, 3–31. D’Arcy, A. (2004) Contextualizing St. John’s youth English within the Canadian quotative system. Journal of English Linguistics 32, 323–345. D’Arcy, A. (2005) Like: Syntax and Development. PhD thesis, University of Toronto. D’Arcy, A. (2006) Lexical replacement and the likes. American Speech 81 (4), 339–357. D’Arcy, A. (2007) Like and language ideology: Disentangling fact from fiction. American Speech 82 (4), 386–419. D’Arcy, A. (2008) Canadian English as a window to the rise of like in discourse. Anglistik: International Journal of English Studies 19 (2), 125–140.

70

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Dailey-O’Cain, J. (2000) The sociolinguistic distribution of and attitudes towards focuser like and quotative like. Journal of Sociolinguistics 4 (1), 60–80. Darmody, M., Tyrrell, N. and Song, S. (eds) (2011) The Changing Faces of Ireland: Exploring the Lives of Immigrant and Ethnic Minority Children. Rotterdam: Sense publishers. Department of Education and Skills and the Office of the Minister for Integration. (2010) Intercultural Education Strategy, 2010–2015. Online document: http://www.education.ie/servlet/blobservlet/mig_intercultural_education_strategy.pdf. Department of Social Protection, Ireland. Online document: http://www.welfare.ie/EN /Topics/PPSN/Pages/ppsn_all_years.aspx. Devine, D. (2009) Mobilising capitals? Migrant children’s negotiation of their everyday lives in school. British Journal of Sociology of Education 30 (5), 521–535. Devine, D. and Kelly, M. (2006) ‘I just don’t want to get picked on by anybody’: Dynamics of inclusion and exclusion in a newly multi-ethnic Irish primary school. Children & Society 20 (2), 128–39. Devine, D., Kenny, M. and McNeela, E. (2008) Naming the ‘other’ – children’s construction and experience of racisms in Irish primary schools. Race, Ethnicity and Education 11 (4), 369–385. Fanning, B., Haase, T. and O’Boyle, N. (2011) Immigrant child well-being and cultural capital. In M. Darmody, N. Tyrrell and S. Song (eds) The Changing Faces of Ireland: Exploring the Lives of Immigrant and Ethnic Minority Children (pp. 167–182). Rotterdam: Sense Publishers. Ferrara, K. and Bell, B. (1995) Sociolinguistic variation and discourse function of constructed dialogue introducers: The case of be + like. American Speech 70, 265–290. Fuller, J.M. (2003) Use of the discourse marker like in interviews. Journal of Sociolinguistics 7 (3), 365–377. Grabowska, I. (2005) Changes in the international mobility of labour: Job migration of Polish nationals to Ireland. Irish Journal of Sociology 14, 27–44. Hasund, I.K. (2003) The Discourse Markers ‘Like’ in English and ‘Liksom’ in Norwegian Teenage Language: A Corpus-based, Cross-linguistic Study. PhD thesis, University of Bergen and Agder University College. Hickey, R. (2005) Dublin English. Evolution and Change. Amsterdam: John Benjamins. Hickey, R. (2007) Irish-English: History and Present-Day Forms. Cambridge: Cambridge University Press. Hochschild, A.R. (2006) Love and gold. In D.M. Newman and J.A. O’Brien (eds) Sociology. Exploring the Architecture of Everyday Life. Readings (6th edn, pp. 212–220). Thousand Oaks, CA: Pine Forge Press. Hopper, P. and Traugott, E. (2003) Grammaticalization (2nd edn). Cambridge: Cambridge University Press. Kitching, K. (2011) The mobility of racism in education: Contested discourses and new migrant subjectivities in Irish schooling. In M. Darmody, N. Tyrrell and S. Song (eds) The Changing Faces of Ireland: Exploring the Lives of Immigrant and Ethnic Minority Children (pp. 89–104). Rotterdam: Sense Publishers. Klein, N. (2000) No Logo, No Space, No Choice, No Jobs. London: Flamingo. Levey, S. (2003) He’s like ‘Do it now!’ and I’m like ‘No!’. English Today 73, 24–32. Levey, S. (2006) The sociolinguistic distribution of discourse marker like in preadolescent speech. Multilingua 25, 413–441. Macauley, R. (2001) You’re like ‘why not?’ The quotative expressions of Glasgow adolescents. Journal of Sociolinguistics 5 (1), 3–21.

The Posit ional Distr ibut ion of Discourse like

71

MacÉinrí, P. and White, A. (2008) Immigration into the Republic of Ireland: A bibliography of recent research. Irish Geography 41 (2), 151–179. McDaid, R. (2011) GŁOS, VOCE, VOICE: Minority language children reflect on the recognition of their first languages in Irish primary schools. In M. Darmody, N. Tyrrell and S. Song (eds) The Changing Faces of Ireland: Exploring the Lives of Immigrant and Ethnic Minority Children (pp. 17–34). Rotterdam: Sense Publishers. Meyerhoff, M. and Niedzielski, N. (2003) The globalisation of vernacular variation. Journal of Sociolinguistics 7 (4), 534–555. Miller, J. and Weinert, R. (1995) The function of LIKE in dialogue. Journal of Pragmatics 23, 365–393. Naughton, B., O’Mahony, A. and Miley, S. (2011) How to combat primary school segregation in a multicultural world: Ireland’s experience. In J. Bakker, E. Denessen, D. Peters and G. Walraven (eds) International Perspectives on Countering School Segregation (pp. 33–49). Antwerp, Apeldoorn: Garant. Nestor, N. and Regan, V. (2011) The new kid on the block. A case study of young Poles, language and identity. In M. Darmody, N. Tyrrell and S. Song (eds) The Changing Faces of Ireland: Exploring the Lives of Immigrant and Ethnic Minority Children (pp. 35–52). Rotterdam: Sense Publishers. Nestor, N., Ní Chasaide, C. and Regan, V. (2012) Discourse ‘like’ and social identity – a case study of Poles in Ireland. In B. Migge and M. Ní Chiosáin (eds) New Perspectives on Irish English (pp. 327–353). Amsterdam: John Benjamins. Ní Chasaide, C. and Regan, V. (2008) Sociostylistic variation in the L3 French and L2 Irish speech of adolescents in an Irish language immersion school. Paper presented at Sociolinguistics Symposium 17, Micro and Macro connections, Amsterdam, April 3–5. Ní Chonaill, B. (2010) Race pressure points no longer can be ignored. The Irish Times newspaper article, 4 April 2010, accessed 5 April 2010. Online at http://www.irishtimes.com/newspaper/opinion/2010/0409/1224267967713.html. Ní Laoire, C., Bushin, N., Carpena-Méndez, F. and White, A. (2009) Tell me about yourself. Migrant children’s experiences of moving to and living in Ireland. Online document: http://migration.ucc.ie/children/finalreport.html. Nowlan, E. (2008) Underneath the band-aid: Supporting bilingual students in Irish schools. Irish Educational Studies 27 (3), 253–266. OECD (2009) OECD Reviews of Migrant Education: Ireland. Online document: http:// www.oecd.org/dataoecd/1/50/44344245.pdf Regan, V. and Nestor, N. (2010) French Poles, language and identity: An intergenerational snapshot. In V. Regan and C. Ní Chasaide (eds) Language Practices and Identity Construction by Multilingual Speakers of French L2. The Acquisition of Sociostylistic Variation (pp. 145–58). Modern French Identities, Vol. 80. New York: Peter Lang. Romaine, S. and Lange, D. (1991) The use of ‘like’ as a marker of reported speech and thought: A case of grammaticalization in progress. American Speech 66, 227–279. Ryan, L. and Sales, R. (2011) Family migration: The role of children and education in family decision-making strategies of Polish migrants in London. International Migration, ‘early view’ (electronic) article, first published online 19 January 2011. DOI: 10.1111/j.1468-2435.2010.00652.x. Sankoff, G., Thibault, P, Nagy, N., Blondeau, H., Fonollosa, M-O. and Gagnon, L. (1997) Variation in the use of discourse markers in a language contact situation. Language Variation and Change 9, 191–218. Schiffrin, D. (1987) Discourse Markers. Cambridge: Cambridge University Press.

72

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Schweinberger, M. (2010) LIKE revisited. A quantitative analysis of the distribution, position, and function of the discourse marker LIKE in Hiberno-English and other varieties of English. Paper presented at New Perspectives on Irish English, University College Dublin, Dublin, Ireland. 11–14 March. Schweinberger, M. (2012) The discourse marker LIKE in Irish English. In B. Migge and M. Ní Chiosáin (eds) New Perspectives on Irish English (pp. 179–202). Amsterdam: John Benjamins. Sharifian, F. and Malcolm, I.G. (2003) The pragmatic marker like in English teen talk. Pragmatics & Cognition 11 (2), 327–344. Siegel, M. (2002) ‘Like’: The discourse particle and semantics. Journal of Semantics 19, 35–71. Siemund, P., Maier, G. and Schweinberger, M. (2009) Towards a more fine-grained analysis of the areal distributions of non-standard features of English. In E. Penttilä and H. Paulasto (eds) Language Contacts Meet English Dialects: Studies in Honour of Markku Filppula (pp. 19–45). Newcastle: Cambridge Scholars Publishing. Smyth, E., Darmody, M., McGinnity, F. and Byrne, D. (2009) Adapting to diversity: Irish schools and newcomer students. Dublin: The Economic and Social Research Institute. Online document: http://www.esri.ie/publications/latest_publications/view /index. xml?id = 2783. Smyth, J. (2010) Poles based in Ireland repatriate €841 million. The Irish Times newspaper article, 26 October 2010, accessed 12 December 2010. Online at http://www. irishtimes.com/newspaper/ireland/2010/1026/1224282006224.html. Suárez-Orozco, C. and Suárez-Orozco, M.M. (2001) Children of Immigration. Cambridge, MA: Harvard University Press. Tagliamonte, S. (2005) So who? Like how? Just what? Discourse markers in the conversations of young Canadians. Journal of Pragmatics 37, 1896–1915. Tagliamonte, S. (2006) Analyzing Sociolinguistic Variation. Cambridge: Cambridge University Press. Tagliamonte, S. and D’Arcy, A. (2004) He’s like, she’s like: The quotative system in Canadian youth. Journal of Sociolinguistics 8, 493–514. Tagliamonte, S. and Hudson, R. (1999) Be like et al. beyond America: The quotative system in British and Canadian youth. Journal of Sociolinguistics 3 (2), 147–172. Tannen, D. (1986) Introducing constructed dialogue in Greek and American conversational and literary dialogue. In F. Coulmas (ed.) Direct and Indirect Speech (pp. 311– 332). Berlin: Mouton de Gruyter. Underhill, R. (1988) Like is, like, focus. American Speech 63, 234–246. Winter, J. (2002) Discourse quotatives in Australian English: Adolescents performing voices. Australian Journal of Linguistics 22 (1), 5–21. [no author] ‘Do you, like, totally want chips with your rice going forward?’ The Irish Times newspaper article, 10 October 2009, accessed 14 October 2010. http://www.irishtimes.com/newspaper/features/2009/1014/1224256602033.html.

4 The Polish Diaspora in Austria: Mutual Attitudes, Processes of Integration and Identity Formation Among Immigrant Poles B. Bidzin´ska

Introduction In any serious discussion on identity of a particular national or ethnic group, and its relations with other ethnic groups, it seems vital to consider a number of questions regarding the whole historical and socio-political contexts of societies. These questions, among other, are: is the reader familiar with the recent history of both the host and receiving societies? Does he or she have an understanding of the specific nature (or its lack) of their past historical relations? These aspects are necessary to fully grasp the complexities of individual immigrant experiences and their impact on identity formation.

Part One: Theoretical Considerations The Aim of the Study For the purposes of the project, the individuals who emigrated from Poland in the 1980s were the target of the study. However, in order to make the study 73

74

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

more comprehensive, it has been decided that some of our participants’ children who were born in Austria would also be included. Details of the methodology used, and a description of our participants, can be found in the empirical part of the analysis in later sections. However, the reader is offered a theoretical introduction to the study before he or she proceeds to the empirical part of the project. Whereas the subject of Polish immigrants to Ireland has recently been a very popular direction in socio-linguistic research, Polish immigration to Austria, although its history is much longer and more complex, does not seem to have been the focus of any larger academic project to date. The following study seeks to introduce the reader to this rather neglected aspect of Polish immigration, and demonstrate how different immigrant experiences can be depending on time and place. The main research questions we posed throughout the study related to the nature of individual immigrant experiences of the participants. • • •

How welcoming and open to Poles in general was the receiving (Austrian) society considered by our participants? What was the nature of inter-Polish relations in Austria? Were Poles forming a hermetic community or rather integrating with the larger society? What were our participants’ attitudes to Austria, the receiving country?

The Complexities of Identity Before we start with an empirical analysis of our data, some theoretical introduction to the concept of identity is necessary. One has to bear in mind that various factors play a role in the process of identity formation, both on the individual and societal level. These, among others, seem to be: • • • • • •

the historical background of mutual relations between the respective societies; the general economic situation of the region under investigation; the current political situation of the region; the social status of the group(s) in question; cultural and religious values of the society; personal choices of the individual regarding his/her place in the society, for example, their job;

The Polish Diaspora in Austr ia

• •

75

the individual’s attitudes to their own native group and cultural heritage; the individual’s attitude to other national or ethnic groups.

We will refer to these aspects throughout our study, especially in the empirical section where the reader will find a tabular presentation of the obtained data from the interviews conducted with our participants. It seems that identity may be analysed on two levels – sociological, where the environment dictates and conditions the status of a particular group, and psychological reflected in an individual’s choices regarding their own understanding of their role in the society. The individual’s social identity (Jenkins, 1996) relates to their own concept of who they, and the people around them, are or represent. In other words, it reflects the individual’s way of thinking about themselves in the society – their role in the community on the basis of which, perhaps, the individual positions themselves in relation to other members of the society. He or she categorizes themselves and others. This categorization originates in the perception of similarities and differences between people (Weeks, 1991). In some cases such as, for example, certain orthodox religious communities, these differences may be clear cut and members of respective groups may be easily identifiable as distinct from members of the mainstream society. Moreover, while for some people such religious, cultural, or ethnic, differences are not important at all, for others they may form a good reason for unfavourable categorization which may lead, eventually, to prejudice. Another interesting theoretical concept on which our discussion will draw is that of cultural identity. Adler (2002) distinguishes two levels of cultural identity: collective and individual. Whereas the first refers to the groups’ behaviour dictated by particular culture-specific values, the second relates to the interplay of personality and collective factors which are present in an individual. These two levels seem to complement each other and shape the individual’s understanding of themselves as an independent – that is, different from others – member of a particular group, being national, ethnic or religious. Identities are often understood as flexible, unsettled and fluid (Goffman, 1959). Over time, people can change their understanding of themselves and their attitudes to others. For example, immigrants have to adapt to a new environment, which is not an easy and straightforward process. This, in turn, is often related to psychological and social aspects of immigration. Whether an individual integrates with or rejects the host community seems to reflect their attitude to both the heritage and host cultures. It shows how

76

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

the individual understands their place in the larger society, their roles and, perhaps, obligations. In our analysis, we considered not only the situation of our participants as they perceived it, but also the socio-political conditions of the host country in which they lived at the time of the interviews – Austria.

Polish Identity In order to understand the concept of Polish identity one has to consider not only the history of Poland, but also various directions of Polish political thought, reflected not only in essays, but also in literature. One important theme often present in Polish literature is that of national martyrdom. The idea of a constant need to fight for independence, security and protection of national identity has been very popular throughout decades. The theme of martyrdom was especially popular during the time of the three partitions of Poland between Austria, Prussia and Russia in 1772, 1793 and 1795. The reader should be informed that in the years between 1795 and 1918 Poland did not formally exist as a recognized nation state and did not feature on the map of Europe. Poles shared, at that time, the fate of Jews – a nation without the homeland. Another important aspect of Polish identity may be described as a strong link to the Slavic heritage and its symbols, the Polish language and Catholicism. Of course, it should be remembered that not all Poles would subscribe to these ideas of Polishness. Nevertheless, they are present in many important works of philosophers and writers, and even contemporary artists. ‘Kraj nad Wisła˛’ (The land upon Vistula) exists as a popular metaphorical way of referring to Poland, for example in songs by popular music bands. The Polish language has always been an identity marker for generations of Poles, especially during turbulent times of political instability, partitions, and wars. The reader is referred, among other, to the famous children’s protest in Wrzes´nia in Prussia (1901–1902) against the ban of the Polish language in schools, and secret classes held in Polish for school children during the Second World War. The reader should bear in mind that attendance at such forbidden educational activities for Poles carried a huge risk of being caught or turned in and, in consequence, punished. Even today, in 21st century Europe, language issues are, at times, very sensitive. The reader is referred to famous law cases where immigrant parents were forced to fight for their right to speak Polish in front of their children (Stowarzyszenie Polskie Przeciw Dyskryminacji Dzieci T.Z.).

The Polish Diaspora in Austr ia

77

Finally, the Catholic Church has always played an important role in Poland, and in the formation of identity of many generations of Poles. Brian Porter, in his essay on Polish Catholic identity cited Cardinal Wyszyn ´ski (p. 289): Nowhere else is the union of Church and nation as strong as in Poland. Although some may not agree with this statement or disapprove of the idea of a strong influence of Church on the state, the fact is that the majority of Poles declare themselves as Catholic. (Porter, 2001: 289)

The Case of Austria – Selected Immigrant Policies Austria is not officially considered a traditional country of immigration. Nevertheless, it has hosted many national and ethnic groups within its borders. Slovens, Hungarians, Czechs, Slovaks, Roma and Croats represent the recognized minorities (Böse et al., 2002). In the 1960s Austria introduced guest worker programmes for immigrants who were welcome in the labour sector, whereas the service sector was generally reserved for Austrians. Of course, these guest worker programmes were conditioned by the state of the economy in the region. There was a need for temporary low-skilled workers to fill jobs which Austrians could not do, or did not particularly wish to perform. Böse et al. (2002) argue that by the introduction of those guest worker programmes, Austria opened its borders to immigrants and gradually started to become a country of immigration (Böse et al., 2002: 3). Those migrant workers who decided to stay longer in Austria were joined by their families. In order to protect Austrian employees from the growing competition on the labour market, in 1975 the Foreign Worker’s Occupation Act was introduced. It stipulated that Austrians should be preferred over foreign workers when a prospective employer was filling a job vacancy. Following the 1956 Hungarian Uprising and the 1968 Prague Spring, in 1981 Austria introduced a visa requirement for Poles escaping the communist regime and the Martial Law (Jandl & Kraler, 2003). As a result of the massive influx of immigrants into Austria, in 1990 a quota for foreign workers was introduced by the Austrian Government. Further legislations introduced in order to better control immigration into Austria were the 1992 Asylum Act, the 1992 Aliens Act, and the 1993 Residence Act. In 1997 both became a single law and were replaced by the 1997 Aliens Act. In 1998 the Naturalisation Act was introduced. In addition to a tenyear waiting period for naturalization, immigrants who wished to become Austrian citizens were obliged to prove their integration with the

78

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

mainstream society. It must be noted here that special language courses were offered to immigrants to better prepare them for future naturalization. The Freedom Party led by Jörg Heider played a significant role in forming anti-immigrant atmosphere and panic in the media resulting in, among other, an Anti-Foreigner Referendum in 1992 (Böse et al., 2002). The reader must bear in mind that the above-mentioned legislation seeks to illustrate the political situation and social status of Polish immigrants who entered Austria in the 1980s, as well as their children who were born in Austria. The accession of Austria into the European Union forced significant changes since Austria had to comply with the rest of the EU legislation dealing with sensitive immigration issues. In the following sections the reader will be introduced to the methodology used for data collection as well as the results of the empirical part of the project.

Methodology A number of research methods were employed in the project in order to obtain data of high quality. The aim was not only to verify the data gathered, but also to examine the research topic from as many perspectives as possible, in order to obtain a comprehensive overview of the sociological processes taking place within the studied community. In addition to semistructured one-to-one interviews with the recruited subjects, a questionnaire was circulated among the participants. The interviews were recorded for future transcription. All participants were given an information letter and, if they agreed to participate, were asked to return a signed consent form. The interviews with the participants were held between 2010 and 2011.

The Participants The target of our study were individuals who emigrated from Poland in the 1980s and sought asylum in Austria. At the time when the interviews were conducted they had spent between twenty and thirty years in Austria. For the second group of our participants we chose young people in their twenties who were born and raised in Austria, held Austrian citizenship and grew up speaking Polish at home, and German outside the home. Due to the sensitive nature of the project, we were not able to find many individuals who were willing to take part in our research. We managed to recruit 20 individuals who immigrated to Austria in the early 1980s, and five who were born of Polish parents and raised in Austria.

The Polish Diaspora in Austr ia

79

Part Two: Empirical Study The Research Questions We repeat here the main research questions which guided our study of Polish immigration to Austria. Those were: • • •

How welcoming and open to Poles in general was the receiving (Austrian) society considered by our participants? What was the nature of inter-Polish relations in Austria? Were Poles forming a hermetic community or rather integrating with the larger society? What were our participants’ attitudes to Austria, the receiving country?

We were interested in a number of issues related to the Polish immigration to Austria such as, for example, reasons for leaving Poland, the general perceived status of Polish immigrants, inter-Polish attitudes and relations, frequency of contacts with other Poles, and attitudes to the Polish language. The data which we obtained from the interviews and questionnaires provided an interesting insight into the quality of our participants’ immigration experiences over years.

Results of the Study The dramatic developments which took place in Eastern Europe during the period of the Cold War triggered often desperate responses and decisions. These, in turn, often shaped the participants’ lives and, in consequence, often forced them to redefine their own identity – immigrant identity. Psychological and social adaptation to a new environment, even if it is culturally proximate, requires a lot of psychological effort on an individual’s part. Therefore, a presentation of results of any sociological study should be well contextualized to be informative. Having briefly introduced the reader to the conditions of Polish immigration in Austria, we now proceed to the empirical part of our study. In order to better organize our data, we divided our presentation into sections.

Number of Participants The reader is reminded that, owing to the sensitive nature of our study, we were not able to find many individuals who were willing to take part in

80

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

our project. The number of recruited individuals who immigrated into Austria in the early 1980s, and young individuals who were born and raised in Austria and grew up speaking both Polish and German was 20 and five respectively.

Tabular Presentation of the Obtained Data The reader will find a tabular presentation of the obtained data in Tables 4.1 and 4.2. The data were also presented in the unpublished PhD thesis (Bidzin´ska, B., 2012) submitted at Trinity College Dublin.

Discussion and Conclusions The study seeks to contribute to a popular topic of sociological research of Polish immigration yet another, and somewhat neglected, Austrian, perspective. We divided our participants into two separate groups to better Table 4.1 Attitudes to Poland and Polish Issues Concepts

Reasons for leaving the homeland Associations with the homeland Involvement in Polish life & culture

Number of Polish friends in general Citizenship Problems discouraging the participants from active involvement with the Polish issues

Participants The 1980s immigrants to Austria

Poles born in Austria

Economy (66%) Political situation in Poland (60%) The past (72%) Family & friends (30%) Contacts with Polish institutions (85%) Frequent (twice a year or more) travels to Poland (85%) Described as ‘many’

Not applicable to this group of participants

Austrian citizenship (65%) Polish politics (80%) Inappropriate behaviour of fellow Polish immigrants (52%)

Austrian citizenship (100%) Polish politics (41%) Inappropriate behaviour of fellow Polish immigrants (60%)

Politics (40%) Contacts with Polish institutions (40%) Frequent (twice a year or more) travels to Poland (60%) Described as ‘many’

The Polish Diaspora in Austr ia

81

Table 4.2 The immigrants as members of the Austrian Society Concepts

Polish immigrants in the Enlarged European Union

Participants The 1980s immigrants to Austria

Poles born in Austria

Austrians are open to the integration of Poles (15%) Perceived change in Poland’s status after its accession to the EU (90%) Feeling a part of the host society (90%) No (90%)

Austrians are open to the integration of Poles (20%) No perceived change in Poland’s status after its accession to the EU (100%) Not feeling part of the host society (80%) No (100%)

Participants’ attachment to the host society Planned return to the homeland in the future Yes (100%) Polish language as an important part of the participants’ identity

Yes (100%)

contextualize our study and enable the reader to analyse the obtained data in an organized and structured way. As can be seen from the tables, responses to some questions differed significantly for the two groups of participants. Economy and unstable political situation in Poland in the eighties were most often mentioned by the 1980s immigrants to Austria as the main reasons for emigrating from Poland and seeking new home outside its borders. The question related to emigration from Poland was not applicable to the other group of our participants who were born and raised in Austria. Similarly, the great majority of our 1980s immigrants to Austria mentioned their past as their main association with Poland. A more private sphere of life, that is, family and friends, were the second most often cited concept, however, by only one third of our respondents (30%). The group of individuals who were born in Austria mentioned politics as their main association with the homeland of their parents. As for the involvement in Polish life and culture, the majority of the 1980s immigrants to Austria kept in touch with the Polish institutions in Austria, while less than half of our respondents born in Austria admitted active involvement in them. However, most of our participants in both groups travelled to Poland at least twice a year.

82

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Problems which sometimes discouraged our subjects from involvement in Polish life and culture in Austria were the quality of Polish politics and inappropriate behavior of fellow Polish immigrants. The vast majority of our participants in the group of the 1980s immigrants held Austrian citizenship. All five individuals in the group of Poles born in Austria held only Austrian citizenship. Most of our participants in the group of the 1980s immigrants to Austria perceived changes for the better in Poland’s image after its accession to the European Union in 2004. Interestingly, no such change was registered by the members of the other group who declared that Poland’s status had not change at all. In addition, our subjects did not consider Austrians as open to the integration of Poles into their society. Nevertheless as many as 90% of the 1980s immigrants to Austria declared attachment to their new homeland. To the contrary, the majority of the interviewed individuals in the other group (80%) did not feel a part of the mainstream society. Most of our participants in both groups did not plan to return to Poland and settle there. However, for all our participants – 100% in both groups – the Polish language featured as a very important part of their identity. One conclusion which might, perhaps, be drawn from the abovepresented data is that some participants felt somewhat torn between the two cultures – native and host – and developed their own identity which incorporated elements of both Polish and Austrian culture. Here, the idea of a shifting, flexible identity is demonstrated, especially for the 1980s immigrants to Austria. Polish issues, such as maintenance of the Polish language in the family and active involvement in Polish institutions in Austria, were still important to them. However, they developed genuine affection to Austria – they considered themselves part of the mainstream (Austrian) society. What is interesting is that our participants in the other group had quite opposite feelings. The reasons for this were revealed later, when the participants in this group described their often painful childhood experiences with other kids. Offensive nicknames and racist comments from other children (including immigrant children) were the participants’ ‘daily bread’ at school. The majority of our participants in both groups did not plan to go back to Poland and (re)start their new lives there. They had accustomed and, over time, adapted to the Austrian environment, or did not wish to start everything from the beginning in Poland. One can only imagine that moving back to Poland would have been, at least for some, a risky decision. Poland, although a member state of the European Union, differs from Austria in many ways. For example, the welfare system, or salary levels in some sectors of the job market are still worlds apart.

The Polish Diaspora in Austr ia

83

Future Research Directions Researching a particular minority community across different countries is very informative for at least two reasons. On the one hand a researcher can observe how a particular ethnic, or religious, group adapts to, sometimes culturally proximate and sometimes very different, host environment. On the other hand, one can learn something not only about the immigrants themselves, but also about the host society in general. For example, legal regulations which apply to citizens of the European Union (or any other federation of countries, etc.) may differ significantly from those which apply to other immigrants, in this case, the so called non-EU citizens. Therefore, it may be metaphorically said that the host country, understood as a political and economic entity, sets the stage, and both the mainstream society and immigrants are actors playing their roles. However, certain topics may be considered sensitive by members of some ethnic or cultural/religious groups. Care should be taken in order to structure any sociological research in a way which is not offensive to prospective participants. In research like ours, owing to delicacy of some matters discussed with the participants, finding many individuals willing to share their experiences with the researcher may not be an easy task. In such cases, a research method known as ‘snowball sampling’ (Dörnyei, 2003) where the participants inform their friends and acquaintances about the nature of the study and help recruit other participants who meet the necessary criteria. It may also help to inform about the project relevant institutions like, for example, cultural institutes, embassies, schools, societies etc., and talk to people who run or visit them. More research exploring the topics discussed above should be conducted in the future, including more participants belonging to the groups discussed above. It would be interesting to carry out a study with recent (post-2004) immigrants who are already permanent residents in Austria and see whether, and how, their experiences change over time. A study of Austrians who are resident in Poland would also make an interesting project.

References Adler, P. (2002) Beyond Cultural Identity: Reflections on Multiculturalism. Originally published in Culture Learning. East-West Center Press, Richard Brislin, Editor, 1977, pp. 24–41, then republished Intercultural Communication edited by Larry Samovar and Richard Porter, Wadsworth Publishing Company, 1976, pp. 362–378. Böse, M., Haberfellner, R. and Koldas, A. (2002) Mapping Minorities and their Media: the National Context – Austria. Centre for Social Innovation. http://www.lse.ac.uk/collections/EMTEL/Minorities/reports.html Brian Porter (2001) Slavic and East European Journal SEEJ, 45 (2), 289–299.

84

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Dörnyei, Z. (2003) Questionnaires in Second Language Research. Construction, Administration and Processing. Lawrence Erlbaum Associates. Goffman, E. (1959) The Presentation of Self in Everyday Life. Garden City, NY: Anchor Books. Jandl, M. and Kraler, A. (2003) Austria: A Country of Immigration? International Centre for Migration Policy Development. http://migrationinformation.net/Feature/ Jenkins, R. (1996) Social Identity. London: Routledge. Jones, E.E. and Nisbett, R.E. (1971) The Actor and the Observer. Divergent Perceptions of the Causes of Behavior. Morristown, NJ. General Learning Press. Weeks, J. (1991) Against Nature: Essays on History, Sexuality and Identity. London: Rivers Oram Press.

5 Divergence, Convergence and Passing for a Native Speaker: Variations in the Use of English by Polish Migrants in Ireland Ewelina Debaene and John Harris

...You recognised my accent, And asked where I was from. Just like that. So ask, ‘just like that,’ About anything else. – My shoe size × Do you like wild strawberries? [...] But don’t ask, ‘WHERE ARE YOU FROM?’ Because that question creates an abyss between us And it makes me feel As tiny As a dwarf Next to you. And I thought I was grown-up. Do you really Want to be A giant Next to me?

(Krzysztof Wodiczko (1999) Critical Vehicles: Writings, Projects, Interviews. Cambridge, MA.: MIT Press.) 85

86

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Introduction This paper presents an exploratory study of the relationship between attitudes and identity on the one hand and situationally related variations in language use by non-native speakers on the other. Our subjects are Polish nationals who migrated to Ireland after the accession of Poland to the European Union in May 2004 and had lived there for at least three years. We also examine the related issue of the extent to which Polish non-native speakers of English in Ireland aim for, or believe they approximate, native speaker performance in English in different settings (or aspire ultimately to pass for native speakers of English). We are primarily interested in the speakers’ own views on these issues and their assessment of the importance of such issues in relation to identity, the maintenance of solidarity with Polish friends and groups, and their integration into Irish society generally. Specific attention is given to speech accommodation strategies – divergence, convergence and passing for a native speaker – their use and functions. The paper draws on background data collected within the framework of the IRCHSSfunded project ‘Second Language Acquisition and Native Language Maintenance in the Polish Diaspora in Ireland and France’ (Singleton et al., 2008a). The participants’ views, practices and interpretations of divergence, convergence and passing phenomena in their own and other Polish immigrants’ use of English were gathered by means of focus group meetings, a questionnaire and one-to-one interviews.

Speech Accommodation Strategies: Taxonomy and Functions Research in the field of language attitudes and speech accommodation provide plenty of evidence that language is a powerful social force which does considerably more than just convey the intended referential information (Cargile et al., 1994, cited in Giles & Billings, 2004: 188). Considering bilingual communication within the broad frame of social interactions between individuals and groups, language may be seen as a paintbrush by virtue of which personal identities are represented. Research designed to investigate interaction patterns between in-group and out-group members is situated on the borderline between, on the one hand, sociology and social identity theory and, on the other, psycholinguistics and social accommodation theory. Social identity theory proposes that people adopt and combine multiple identities, some of which are more personal and idiosyncratic, and others of which are

Divergence, Convergence and Passing for a Nat ive Speaker

87

group-related (Meyerhoff, 2006: 70). On this view, the extent to which people are more individualist and/or collectivist in this sense is a strong determinant of their linguistic interactions, their non-verbal communication and their choice of strategies to establish, contest or maintain relationships through talk. Accommodation theory, in turn, is defined as ‘[a] constant movement toward and away from others by changing one’s communicative behaviour’ (Sachdev & Giles, 2004). Accommodation is seen as the process by which speakers consciously or unconsciously attune or adapt their linguistic behaviour in the light of both their interlocutors’ behaviour and their attitudes towards their interlocutors (Meyerhoff, 2006: 72). Accommodation theory refers, furthermore, to a balance of social psychological processes focusing on social integration and differentiation. Interactants are viewed as having certain expectations regarding optimal levels of bilingual accommodation, these anticipations being based on stereotypes about outgroup members, ways of conducting appropriate intergroup business, and the prevailing social and situational norms (Sachdev & Giles, 2004: 355). Among the many different accommodative strategies that speakers are observed to use to achieve their goals, convergence has so far attracted the most research attention. This is a strategy whereby individuals adapt their communicative behaviours in terms of a wide range of linguistic features (e.g. speech rate, accents), paralinguistic features (e.g. pauses, utterance, length), and nonverbal features (e.g. smiling, gazing) in such a way that their behaviour becomes more similar to their interlocutor’s. The phenomenon of convergence, however, may easily escape researchers’ observation, as there may be a discrepancy between, on the one hand, speakers’ own impressions and feelings about their communicative performance (subjective convergence), and, on the other, their objective linguistic behaviours (see Thakerar et al., 1982: 210). One can furthermore differentiate between upward convergence (in terms of societal valence), when people converge with those in superordinate positions and downward convergence, when those in higher societal positions alter their speech in order to facilitate understanding (cf. Maass & Arcuri, 1996: 218; Sachdev & Giles, 2004: 356). Convergence is most likely to occur when out-group speech reflects high status or high social desirability. The most radical form of linguistic convergence is passing, a frequently overlooked form of language ability described as ‘a temporary, context-, audience- and medium-specific performance’ (cf. Piller, 2002). While all types of convergence characterize movement towards the speech of one’s interlocutor and accentuate commonality between the interlocutors, the opposite type of linguistic behaviour, divergence, is the movement away from the speech of the interlocutor and accentuating difference

88

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

between the speaker and their addressee(s). Giles and Ogay describe this latter type of speech accommodation as a strategy whereby individuals emphasize distinctiveness from their interlocutors, usually on the basis of group membership (Giles & Ogay, 2007). On the continuum between the dichotomous extremes described above (i.e. convergence and divergence), there is a range of linguistic strategies, including inter alia partial convergence. Partial convergence is neither full convergence nor divergence and involves ‘the alternate use of two or more languages in the same utterance or conversation’. Otherwise referred to as code switching, it involves the alternation between varieties, or codes, across sentences or clause boundaries’ (Meyerhoff, 2006: 116). Sachdev and Giles argue that the reasons for code switching may include not finding the right word at a particular moment, allowing others to have practice in one’s language, the desire to impart a cosmopolitan identity or a community identity through humour. They emphasize further exploration of this form of partial accommodation is needed. Despite the fact that code switching has continued to evoke unfavourable evaluation from linguists and language users themselves, these negative perceptions are now changing with more and more scholars perceiving it as a balanced approach between convergence and divergence (Sachdev & Giles, 2004: 355). The phenomenon of speech accommodation is frequently discussed in the context of intergroup relations, mutual perceptions and cultural stereotypes. Intergroup situations tend to activate sociolinguistic stereotypes, which in turn induce divergent or convergent shifts in language use. Depending on the type of intergroup relation and the strength of ethnolinguistic identity, people will tend to shift their speech patterns either towards or away from that of the person being addressed. Thus, both convergence and divergence may be driven by stereotypic beliefs about how in-group and out-group members typically talk (Giles et al., 1987). Closely related to these stereotypical intergroup perceptions, a number of potential gains and losses associated with convergence may be identified. First of all, the strategy of convergence is frequently motivated by a desire to gain approval from the recipient (and hence, perhaps, access to the social networks to which the recipient is believed to belong), the premise being that of similarity attraction (Byrne, 1969). This is the notion that the more similar we are to our interlocutor, the more he or she will like or respect us and the more social rewards we can expect (Sachdev & Giles, 2004: 356). Converging on a common linguistic style also improves the effectiveness of communication. This has been associated with increased predictability of the other and hence a lowering of uncertainty and interpersonal anxiety, leading to better mutual understanding (cf. Gudykunst, 1995; Sachdev &

Divergence, Convergence and Passing for a Nat ive Speaker

89

Giles, 2004: 357). Finally, in terms of Tajfel and Turner’s (1979) social identity theory, convergent language shifts have been considered a tactic for achieving a positive self-image. Seeking social approval is seen as the main element of the individual mobility strategy that may be adopted by certain members of subordinate groups who perceive their in-group’s social position as legitimate and immutable. Despite its numerous potential benefits, convergent strategies have the potential to significantly undermine one’s individual ways of expression and ultimately lead to the loss of personal or social identity (Sachdev & Giles, 2004: 357). Turning our attention now to reasons for divergent speech strategies, it has been substantiated in many studies that non-standard styles of speech are commonly perceived as indicative that the speakers in question would be unsuitable as partners in close personal relationships, and also likely to hold many dissimilar beliefs (e.g. Stewart et al., 1985). They are furthermore perceived as being less in control communicatively (Bradac & Wisegarver, 1984). Viewing divergence solely from the point of view of its negative outcomes might lead, however, to an incomplete understanding of the phenomenon, which in certain contexts may in fact grant a person substantial gains. Positive effects of divergent linguistic behaviours are of a psychological nature. They stem from the fact that divergence may be regarded as a tactic of intergroup distinctiveness for bilinguals wishing to feel good about their ethnicity, and thereby about themselves (cf. Sachdev & Giles, 2004: 358). Divergence can also be adopted in order to shape receivers’ attributions and feelings (Simard, Taylor & Giles, 1976, cited in Sachdev & Giles, 2004: 367) and to remind interlocutors that one does not belong to the same ethnolinguistic group (cf. reasons for code switching). This brief theoretical background relating to communicative strategies serves to contextualize the examination of migration theory to which we now turn. We will be concerned particularly with the linguistic behaviours of immigrants entering on interactions with members of their host societies.

Speech Accommodation in Migrants Recent decades have seen a great proliferation of research into the linguistic behaviours of migrants in interactions with host societies. What is more, the achievement of good communication and increased interactions between migrants and home communities are seen as key factors in reducing conflict, reversing negative stereotypes and developing positive mutual relationships. Within migration theory, linguistic and cultural encounters

90

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

between immigrants and host community members are frequently discussed in the context of the contact hypothesis (cf. Cook, 1962; Pettigrew, 1998; Williams, 1947). Proponents of the unconditional contact hypothesis emphasize the role of ‘acquaintance potential’ (Cook, 1962) or ‘friendship potential’ (Pettigrew, 1998). The argument is that members of the host community are more fearful and hostile toward migrants if they interact with them less, and that contact across ethnic divides in and of itself mitigates mutual animosities and prejudice, irrespective of group status, goals and political, social, economic and cultural context. Friendship with even one out-group member would logically create expectations that friendship with other out-group members is acceptable, if not desirable. Potential for friendship should therefore reduce hostility and negative stereotypes. Because contact is more likely if the size of the migration group is larger, this may explain why hostility towards immigrants may actually diminish when migration levels increase or stay high. Other studies (Forbes, 1997) found, however, that more frequent and widespread contact between culturally dissimilar groups is likely to increase intergroup hostility in many social contexts. Language is a particular topic in discussion of mobility and intergroup relations in the contemporary world. The strategies of speech accommodation in migrant groups perceive linguistic behaviours as intrinsically related to the processes of integration and acculturation as well as strongly impacting on identity formation. Studies of the linguistic assimilation of immigrants in dominant cultures with findings indicating that language shifts in interactions between migrants and members of the host societies are typically unilateral with the subordinate groups converging with the dominant groups in their use of language far more than vice-versa (Giles, 1977). Moreover, convergence by immigrants has typically been considered an indication that they have acculturated to the host culture’s values (Kim, 2001), a process that may be described as unilateral, unidirectional and asymmetric (cf. Giles, 1977; Sachdev & Giles, 2004: 357). Attempts at social integration, linguistic acculturation or identification by means of convergence have been generally accorded positive evaluation by their receivers (Giles et al., 1973) and perceived to validate the recipients’ own ways of expressing themselves, to increase attractiveness and compliance, communicative efficiency, and cooperativeness. More recently, however, alternative approaches to the issue have been proposed. As Shohamy (2006: 36) observes, contrary to the assimilative patterns adopted by immigrants in the past, current immigrants demand recognition and the right to disagree with the hegemonic ideology as they strive for the legitimacy of maintaining their differences. They may refuse therefore to acquire the host national language (or local dialects or accents) and

Divergence, Convergence and Passing for a Nat ive Speaker

91

insist on maintaining their home languages. Increasingly, the maintenance and expression of ethnic identity is becoming optional, a matter of voluntary commitment on the part of individuals rather than being obliged to subscribe to an identity which is imposed on them (Parsons, 1975: 64–65). The shifts and fluctuations in language ideologies and in the range of identities available to individuals have become particularly visible in the light of recent sociopolitical and socioeconomic trends: globalization, consumerism, and the exploitation of media technologies. Giles and Byrne (1982, cited in Pavlenko & Blackledge, 2004: 4) developed a theory of ethnolinguistic identity which considers language to be a salient marker of ethnic identity and group membership. The researchers identified three factors that contribute to a group’s ethnolinguistic vitality and determine which members of the minority group were most likely to acquire the language of the dominant group: (i) in-group identification; (ii) identification with other groups; and (iii) ethnolinguistic vitality. Discussing transnationalism, Extra and Yağmur (2004: 12) remark that, with national boundaries being eroded, local background and origin have become more important. Shain (1999) refers to those who change their place of residence but at the same time continue to maintain close connections with their kin society (cf. Meinhof & Triandafyllidou, 2006: 205). Complex loyalties and affiliations are a part of the current nation-state and so are the languages. Reflecting on these paradigm changes, Pennycook (2004, cited in Shohamy, 2006: 10) remarks that ‘the moment has arrived to argue that the language concept […] has served its time.’ As a result of unrestricted international travel and globalization, languages tend to mix and form various hybrids. In the same vein, Gubbins and Holt (2002, cited in Shohamy, 2006: 38) claim that language identity in Europe is diverse, complex and changing as boundaries of language, citizenship and borders are breaking up, resulting in emerging languages which are not distinct and homogenous (cf. Singleton et al., 2008b: 196–197). As we hope to demonstrate later, many aspects of these theoretical perspectives relate quite directly to the history and experiences of Poles migrating to Ireland in recent years. They also appear to resonate with the views of the Poles in our study about certain aspects of language and accommodation in an Irish context. As a new immigration country, Ireland has seen many new forms and arenas of interaction between the host community and the numerous minority communities established here. The present study provides a unique chance, therefore, to analyse some sociolinguistic aspects of these interactions as they are still taking form, in particular any speech accommodation which may be part of them. To understand the relevance of the social identity and social accommodations issues in the Polish–Irish

92

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

context, however, we first need to present some historical and demographic background information concerning what is the largest minority group in Ireland since EU enlargement in 2004.

Migration of Poles to Ireland: Generation Nothing Strongly represented in the current Polish community in Ireland are member of the so-called ‘generation nothing’. This is the transition cohort who grew up in communist Poland and whose entry into adulthood coincided with the country’s transformation from communism to democracy. Most of the participants in our study are also part of this generation. The term ‘generation nothing’ (or ‘generation zero’) was coined by Wandachowicz (2002: 11) who in his seminal article in the leading Polish national newspaper Gazeta Wyborcza used it to describe those who comprised the demographic baby boom in the second half of the 70s and the early 80s. He paints a detailed portrait of this large group of Polish people and provides insights into the common elements in their life-stories, their collective identity, expectations, shared memories and experiences. It will be useful to summarize his most important claims at this point as they serve to illuminate the decision to migrate made by many of those people and help to situate this decision in the wider context of their historical and socio-cultural background. Owing to the particular demographic, economic and political conditions obtaining at the time, ‘generation nothing’ are Poles now aged from their mid 20s to their mid 30s who have been obliged from the early stages of their lives to compete with a large cohort of peers. As a result, they have experienced a constant sense of lack as there were generally not enough places for them at schools, universities and in labour markets. What is more, they lived their childhood and early adulthood in the shadow of Western Europe, which, through the chinks in the iron curtain, revealed itself to them as a land of prosperity and a symbol of unattainable pleasures. After the fall of communism in 1989 there was a general expectation that this generation, because it was so numerous, would change Poland. However, as Wandachowicz comments, they did not seem to have any idea for society. Instead, they were preoccupied with finding something for themselves as individuals, not as a group. Wandachowicz’s view is that this generation missed its great chance for freedom, opting rather too quickly to accept the rules of consumerist capitalism while failing to question its values and leaving themselves little space for individual choice. The vast majority decided to pursue business careers, particularly favouring advertising, which was

Divergence, Convergence and Passing for a Nat ive Speaker

93

commonly perceived as a symbol of professional and financial success. Many commentators remark on the great impatience characteristic of this group, and their urgent drive to achieve better standards of living, and access the full range of freedom and material goods previously denied to them. In a sense this generation can be seen as a cross-linking element between two political systems, and also equally affected by both systems, a kind of ‘comm-pitalist’ hybrid. Developing Wandachowicz’s thesis, one can say that members of ‘generation nothing’ were brought up and educated in a communist country but were destined to live and work in an entirely different reality and according to new rules. Education originally designed to meet the needs of pretransformation Polish society did not prepare them for the life they were/are to lead in the ‘post-transformation Poland’. There was a general crisis of values during the transformation itself and, closely related to that, a lack of role models. In consequence, ‘generation nothing’ had very limited contact with representatives of the older generation and thus there was limited intergenerational transmission of experience, perspectives and values. Older people were not willing, or not able, to pass on their experience, as it did not seem to be applicable to the new reality. It is significant that many Polish migrants, including subjects in the present study, testify that it was in Ireland that they were first given a chance to learn from older people and to establish good relationships with them. It may be inferred from this socio-cultural context that many representatives of ‘generation nothing’ feel they have never fully realized their potential. They remain very attached to the idea of learning, as if trying to compensate for the ‘gap in education/preparation for life’. As Wandachowicz remarks further, ‘this generation has no identity’, or rather no collective identity. They are very individualistic in their ways, committed to a constant search for themselves and for things worthy of pursuit, constantly (re-) formulating and questioning their values. Defining and maintaining an identity seems like an incessant meaning-making activity and, subjectively at least, a process primarily concerned with differentiating oneself from others rather than trying to make oneself like others. For ‘generation nothing’ Poles, EU-enlargement and the consequent possibility of migrating and working legally in other EU markets provided an entirely new domain in which their identities and future could be explored and defined. For many ‘generation nothing’ Poles, the decision to migrate represents just another, even if a bigger, step in their constant self-development, rather than a primarily economically driven strategy. The dominance of ‘generation nothing’ in the Polish community in Ireland, as well as something of their vocational and educational profile, is

94

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

revealed in a number of surveys. Results of market research commissioned by a Dublin-based Polish magazine Polski Express in 2007 indicate that 60% of Poles living in Ireland are between 20 and 35 years of age (Polski Express, 2007). The situation in Britain is broadly similar. A British Home Office report published in 2006, found that eight out of ten Polish immigrants had not reached the age of 32. Most were single and only 3% moved to Britain with their families (Mac Cormaic, 2007). Another, interview-based, survey conducted in 2007 by a Warsaw-based agency (Kinoulty Research), among 109 Poles, described Polish people in Ireland as ‘eminently employable’. Some 57% of the Poles interviewed were under 25, 34% had higher education and the majority of subjects were over-qualified for their Irish jobs. According to a report prepared for the National Economic and Social Council of Ireland (2006: 21), most A8 migrants are employed in construction, and farming and farm-related industries such as meat processing. As observed by Ruhs, the majority of Polish post-accession migrants working in Ireland, although employed across a wide range of occupations, are carrying out relatively lowskilled and low-waged jobs in a variety of sectors including hotels, restaurants, construction, agriculture and care (Ruhs, 2007).

The present study It might be expected that a generation which had already experienced the impact of social and political transformation at an individual level would be well equipped to adapt to subsequent change, for example, adapting to and integrating in a new environment. One may speculate that young Poles, having grown up in a competitive environment distinguished by scarce resources and perhaps consequently wishing to escape homogenization, might be prompted to migrate not just by economic considerations but also by the prospect of exploring other lifestyles and cultures. If we are correct in this, it would follow that having arrived in Ireland, individual Poles might have some reluctance to network only within Polish migrants’ circles in the host country and indeed might feel some degree of ambivalence about being part of such a large immigrant Polish community. These considerations prompt a variety of questions about Polish migrants’ views and behaviour in the linguistic/communicative domain, including their acquisition of (Irish) English speech and language norms, and their attitudes to issues of speech accommodation and identity. The present study consisted of an exploratory investigation of their own views on the relationship between individual and ethnic identity on the one hand and situationally related variations in their use of English in Ireland on the other. Specific attention was given to the phenomena of convergence and

Divergence, Convergence and Passing for a Nat ive Speaker

95

accommodation. We also examine the related issue of the extent to which young Polish non-native speakers of English in Ireland aim for, or believe they approximate, native speaker performance in English in different settings (or ultimately aspire to pass for native speakers of English). We are primarily interested in the speakers’ own views on these issues and on their importance in relation to identity, the maintenance of solidarity with Polish friends and groups, and their integration into Irish society generally. Participants’ views, practices and interpretations of convergence and accommodation phenomena in their own and other Polish immigrants use of English have been gathered by means of focus group discussion, followed – where necessary – by individual interviews. The study sets out to establish young Poles’ views on the following questions: • •

• • •

Has their accent in speaking English since coming to Ireland changed? How determined are their attempts to acquire native (Irish) English speaker norms and levels of linguistic performance in relation to accent, syntactic structure and content, non-linguistic interaction, register, idiom and communication generally? Do they aspire to master any other varieties of English? How are interactions in English different in terms of convergence and accommodation relative to whether the participants consist of native/ native, native/non-native or non-native/non-native groups or pairs? How does Polish speakers’ accentedness in English, where it is present, limit their ability to adapt their performance in English to native norms? How are convergence and accommodation related to issues of identity for young Poles in Ireland?

Methodology Because our study constitutes an initial investigation of the experiences and adaptation of recently established migrant communities, there are advantages to inductive, ethnographic approaches to collecting data. We believed that focus group discussions would be the best way to obtain insights into the participants’ collective identity and their shared views, while at the same time providing space for individual self-reflection and exchange of comments and ideas. As Orbe (1998) notes, focus group discussions present a good opportunity to collect descriptions of lived experiences. They generate data by which the participants can consider their own views in the context of the views of others. In the present case, they provide a time-efficient method of

96

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

drawing on several persons’ experiences at once, while also creating a context that encourages synergistic insights unattainable during individual interviews (Cooks & Orbe, 1993). Apart from group dynamics contributing to the description of lived experiences (Patton, 1983, cited in Orbe, 1998), focus groups organized into homogenous clusters also appear uniquely appropriate for research on co-cultural communication, as in-group peers typically play a central role in social constructions of reality and identity formation. They present a more natural environment than that of an individual interview because participants are influencing and are influenced by others – just as in real life (Kreuger, 1988). The value of a co-cultural group perspective is apparent in the opportunity provided to marginalized group members (migrants in our case) to see dominant societal structures through the eyes of a ‘stranger’ (Collins, 1986). To survive and succeed in society, those persons marginalized by dominant structures must be attentive to the perspective of the dominant group and their own. As a result, a ‘double vision’ is established that advocates an awareness of and sensitivity to both the dominant worldview and the individual’s own perspective. An ‘outsider-within’ position is inherent in the perspective of co-cultural group members who function in a dominant society (Orbe, 1998: 29).

Participants and procedure The subjects are 25 Polish nationals who came to Ireland after the accession of Poland to the European Union in May 2004 and have lived in their adopted country for at least three years. In our conduct of the focus groups we adhered to Kreuger’s (1988) recommendations regarding the appropriate number and make-up of participants, duration of meetings, choice of questions, the roles of the researcher, ethical issues and – at the data processing stage – the development of transcripts of audio recordings and analysis of results. In order to obtain quality results, we held three focus groups (in the period between February and May 2009) and made sure the groups in question were relatively homogenous as regards the participants’ age and length of stay in Ireland. Multiple groups with similar participants were needed to help us detect patterns and trends across groups. The size of groups was conditioned by two factors: they had to be small enough for everyone to have an opportunity to share insights and yet large enough to provide diversity of perceptions and a pool of ideas. As familiarity among the participants might inhibit disclosure, we chose to recruit subjects who were unfamiliar with one another (cf. Kreuger, 1988: 20). The topics of discussions were predetermined, sequenced and followed a so-called questioning route or interview guide,

Divergence, Convergence and Passing for a Nat ive Speaker

97

arranged in a natural, logical sequence. Also, elements of triangulation were introduced, with mini-questionnaires being distributed to the participants at the beginning of each focus group meeting and follow-up individual interviews being conducted where necessary. The short questionnaires were meant to provide space for reflection, to signal the issues to be discussed in the course of the focus group meeting and also to help to avoid the so-called ‘me too phenomenon’ when participants do not provide original answers but repeat the testimony of the speaker preceding them. Before proceeding to the analysis, it should be noted that a parallel focus group investigation was conducted with the longer-standing population of Polish immigrants in France. The participants in the French study were representatives of the so called Solidarity generation. Preliminary results of this French study will be referred to briefly in passing in reporting the Irish results below whenever such comparison may help to delineate common or divergent experiences and views relating to L2 use and speech accommodation in these two Polish migration contexts.

Analysis of Results and Discussion As regards the topics of our focus group discussions, language and speech accommodation was initially located within the broader context of issues such as cultural stereotypes, mutual perceptions, identity, attitudes towards the host society, the Polish diaspora and the general well-being or otherwise of Poles in Ireland. It was notable that participants were more willing, and appeared to be in a better position to readily comment on these more general issues than on those related specifically to language and accents (including speech accommodation). Their comments on the latter topics seemed, relatively speaking, less organized and less premeditated, as if perhaps being consciously reflected on and verbalized for the first time. It was clear also that certain comments expressed by one subject would often lead to others reflecting, affirming or sharing their views in response. Thus, the focus group discussions provided a springboard and framework for reflection and a sharing or exchange of views in a collaborative, supportive spirit which was truly enjoyed by our participants.

Self-assessed English proficiency Almost all participants came to Ireland with a prior knowledge of English, usually self-assessed as being at a lower to upper intermediate level. Some had previously migrated to other English speaking countries. Many

98

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

emphasized that the English they had been studying at school was entirely different from the language they were exposed to on arrival to Ireland. In addition, because Ireland is the country when they first embarked on their professional career, much of the work-related vocabulary in their repertoire, for example, is English and they find it easier to discuss work-related issues in English. Even those who felt they were very proficient in English, however, initially found it very difficult to understand Irish speakers of English. Many participants are by now well aware of the existence of a great variety of Irish accents and can differentiate between some of them. They are aware that their way of speaking has changed since they first came to Ireland. One point of reference for such perceptions consists of conversations with native speakers of English from outside Ireland, mainly American and English. Polish migrants often receive comments and questions from their interlocutors such as: ‘You speak English with an Irish accent.’ ‘Have you lived/learnt English in Ireland?’ Even though such comments give our Polish participants a lot of satisfaction as proof of the linguistic proficiency they have attained and their progress in approximating native speaker norms, the vast majority have not consciously aimed at acquiring the Irish accent. They see language in this context as a means rather than an aim in itself, and they believe their approach to language use and language learning is essentially pragmatic, with communicative proficiency being perceived as the most desirable objective. Only a few participants (3 out of 25) reported that, in addition to improving their communicative competence, they would like to acquire an Irish accent when speaking English. This, however, was motivated by their efforts to achieve language proficiency, part and parcel of the effort to model their speech on their Irish (native speaker) interlocutors rather than arising from any desire to lose their Polish accent in English or to escape any potential negative stereotypes associated with being Polish. Our participants did refer to Polish friends who speak with a strong Irish accent, though these were individuals who had come to Ireland with no command of English, so that the only variety of English known to them was Irish English.

Convergence There were also a few anecdotal accounts, however, of other Poles (acquaintances or friends of our participants) who did exert serious, conscious efforts to pass for native speakers. The fact that this is a minority experience among Irish Poles contrasts somewhat with the experiences of those of the Solidarity generation who migrated to France, as recounted in

Divergence, Convergence and Passing for a Nat ive Speaker

99

the parallel French focus group study mentioned earlier. The latter Poles report they had often been taken for Belgian French language users by members of the host French community and such comments pleased them as an indication of their near-native proficiency and their successful eradication of their Slavic accent in French (cf. section ‘Passing for a native speaker’ below). Even if the English acquired by Poles in Ireland has some Irish features, very few of our participants have ever succeeded in passing for a native speaker and most say they are easily identified by Irish people as non-native English speakers. Their experience is that Irish people can easily classify their accent as Eastern European, though they cannot usually differentiate Russian, Czech, Latvian or Ukrainian accents in English. Our participants do not feel particularly happy about being treated as members of such a large undifferentiated group. They would prefer it if Irish people were more aware of the substantial differences between Poland and other countries. They often say to their Irish interlocutors that Poland is a Central European as opposed to an Eastern European country.

Identity, stereotype and perceived country of origin Another question which was advanced for discussion in the course of focus groups concerned subjects’ feelings when they were asked by Irish people about their country of origin. This topic was met with a certain surprise and hesitation. It appeared that either it was something that previously had not been broached or that they had never tried to verbalize their feelings concerning questions from Irish-born people about their ethnic origins. Most agreed that they are usually met with positive reactions from Irish people when they say they are Polish: When I say I am Polish, Irish people often come up with some Polish words they know and sometimes say that they know a Polish person living in their neighbourhood or working with them and how they like that person. . . (A comment expressed by one of participants in focus group discussion, Dublin February 2009). As regards the existence, or formation, of positive and/or negative stereotypes concerning Polish nationals in Ireland, our participants’ comments strongly suggested that the positive comments outnumbered the negative. Stereotypical descriptions of Pole in Ireland were much more likely to be, for example, hard-working, friendly, religious, having strong family values than to be grumpy, too emotional, too direct or heavy drinkers.

100

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Polish group identity versus individuality Even though the majority of our participants relish the good reputation enjoyed by Poles in Ireland and appreciate its advantages, they are nevertheless, as individuals, ambivalent about being part of such a large Polish community. They find it difficult to come to terms with the fact that there are so many Polish people in Ireland and the fact that, as a consequence, the initial keen interest in Poles and Polish culture once displayed by the host community has since flagged somewhat. When someone asks me where I’m from, and when I say I am Polish, silence falls, there are no more questions. . . As if they have known enough about Polish people by now and were not interested in finding out more. I remember when I first came to Ireland a lot of people would ask me about Poland, about our economy, political scene, about our cuisine. . . (Comment by a focus group participant) We ceased to be exotic for Irish people. I think they are much more interested in other nationalities now, Italians, Spanish. . . (Comment by a focus group participant) It seems that our participants would like to escape the rather too clear-cut division between Poles and their Irish hosts – the too easy classification and labelling – which the possession of a Polish accent in English facilitates. They would like to be able to contribute more as individuals to Ireland’s cultural diversity. They admit that, as a consequence, they would like on occasion to be able to hide their Polish accent to the extent that it would not automatically reveal their membership of a group with which they do not necessarily identify in all respects. One could conclude that the outgroup homogeneity bias on the part of the host Irish community may be a strong motivation in some cases for our Polish participants to wish to lose their Polish accent. This resistance among ‘generation nothing’ Poles to being seen as part of any larger collective in Ireland, of course, is the very same feeling which prompted many of them to come to Ireland in the first place as part of an exploration of their own individual identity. They desire their stay in Ireland to be a new experience, an opportunity to learn about other cultures and societies, to ‘come in and mingle’ at this international party, rather than to become associated with, or to form, a parallel Polish society in Ireland. Another related issue is of a more universal, psychological nature, although it is part of the context within which we must view our participants’ views on convergence and the acquisition of an Irish accent when

Divergence, Convergence and Passing for a Nat ive Speaker

101

speaking English. This relates to our participants’ complaint that almost every conversation they have in Ireland starts with ‘where are you from?’ Notwithstanding the fact, as we have seen, that our participants are not particularly concerned to hide their Polish origins, and do not see revealing their status as Poles as a problem, an opening question of this kind has the effect of directing the conversation on a different trajectory from the one which would normally be followed in interactions between ‘unmarked’ native speakers from the host community. The Poles feel that this has the consequence of making initial interactions and relationships with host society members shallow.

Divergence versus passing for a native speaker As already mentioned, only a few subjects (2 out of 25) reported that they have been taken for native speakers of English by their Irish interlocutors – one is an almost balanced bilingual English/Polish language user brought up in the English speaking part of Canada. They do not, however, perceive passing as a generally desirable goal and do not know many Polish people who would pursue such a goal/strategy. In fact, they make anecdotal reference to Polish people they know who deliberately speak with a strong Polish accent with a view to expressing their ethnic identity and distinctiveness. This might be seen as an example of divergence, an important tactic of bilinguals to maintain intergroup distinctiveness and to affirm their own ethnicity and to enhance their self-esteem (cf. Sachdev & Giles, 2004: 358). Some participants nevertheless anecdotally comment that some Polish friends do indeed on occasion try very hard to speak with an Irish accent so as not to be taken for Poles or ‘to shed their Polishness’: I know a Polish man who wants to act and to sound Irish. He is doing all he can to lose his Polish accent and to speak like the Irish do. Even when interacting with other Poles, he speaks English. He is very unhappy when someone takes him for a Pole. I think it is due to the fact that he was fed up with his life in Poland and wants to start on a new slate here, as an entirely new person. (Comment by a focus group participant) In the light of this and similar comments, it may be inferred that conscious attempts to acquire Irish speaking norms (or lack of interest in such acquisition) are strongly related to (a) our subjects’ level of satisfaction with their life in Ireland and (b) the quality of their life in Poland prior to departure. Borrowing the phrasing of a famous Polish playwright, Slawomir Mroz˙ek, one can differentiate between the phenomena of (i) migrating (escaping) from

102

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

and (ii) migrating towards something (Mroz˙ek, 1975: 53; author’s translation). Comparing the results of the present Irish study of Poles with the preliminary findings of interviews with Solidarity-era migrants in the parallel French study lends weight to Mrozek’s formulation. Solidarity-era Polish migrants to France were forced to emigrate to escape the communist-imposed regime, and went through an assimilative pattern of integration in the host French society. This was reflected in their approach to the French language. Unlike the majority of present-day, post-accession Polish migrants to Ireland, our participants in France admit they have been strongly committed to acquiring native speaker norms in French, and have exerted considerable effort to pass for a native speaker. Their ultimate aim was to become indivisible from and fully part of their new homeland. As they had consciously cut off their ties with Poland – returning there not being an option – they did not emphasize or derive their sense of identity from their ethnicity. Neither did they attempt to emphasize their individual/Polish group distinctiveness, or to diverge from the majority French host community, through accented French language use or through the use of the Polish language in the public domain. Such efforts to acquire native speaker norms are not nearly as important to the Polish post-accession migrants in Ireland. In their decision to migrate, these young people are usually motivated by curiosity about and willingness to experience new cultures and to gain new kinds of expertise. In Mroz˙ek’s terms they are migrating towards. The new experience, is expected to add to and to enrich their individual personas and identity, rather than to entirely alter their identity. Their patterns of linguistic and communicative behaviour reflect these expectations, goals and experiences.

Subtle changes in use of Polish Many participants report that during their short visits back home to Poland, subtle changes which have occurred in their non-verbal communication and choice of words and conversational topics in Polish are noted both by the participants themselves and by their family members and friends. In particular, it is clear that they use gestures more than they did before they migrated. They are also using Polish equivalents of such words and phrases as ‘sorr,excuse m,thank yo,how are yo, more often than is customary in Poland or is required by the rules of politeness. There are also a number of words they cannot find equivalents for in Polish and therefore they sometimes resort to code-switching in conversations with other Polish people or use polonized versions of English words. Such cross-linguistic influences are worthy of further investigation.

Divergence, Convergence and Passing for a Nat ive Speaker

103

Conclusions Polish migrants to Ireland view their approach to learning and using English as primarily pragmatic, with a strong emphasis on communicative efficiency. Various types of speech accommodation have been identified in this study, ranging from conscious divergence to deliberate efforts to pass for a native speaker. Factors which seem to be most influential on speech accommodation in our participants include: the original reasons for migrating, level of satisfaction with life in Ireland, level of satisfaction with life in Poland prior to departure, level of English on arrival, attitude to acquiring an Irish accent in speaking English, familiarity with varieties of English other than Irish English, sense of ethnic identity, the interlocutors involved and contexts of language use. Being the object of an outgroup homogeneity bias seems to be the only factor which might prompt a desire in individuals to lose their Polish accent in speaking English, or to have any desire to pass for a native speaker of (Irish) English. In contrast to the experience of Polish Solidarity-era migrants (in the parallel French study) there is a relaxed acceptance among our participants in Ireland of their status as migrants (as outgroup members), a status which is not accompanied by any sense of inferiority or lower social position. Nevertheless, a certain ambivalence is felt by our participants about one aspect of the migrant experience – being part of such a large Polish community in Ireland. This ambivalence may be explained by a very specific sociocultural background and history in Poland - membership of the so-called ‘generation nothing’. As regards future research directions, it would be valuable to examine, in the course of a follow-up study, the extent to which the subjective opinions expressed by our participants are reflected in their actual language performance. Videotaped interviews would be one way to determine this. We also see the need to expand and to diversify the sample with a view to focusing and drawing comparisons between the experiences of particular groups of Polish migrants in Ireland (e.g. Polish migrants employed in selected professional sectors; Polish nationals who are married to host community members). This would help to further explore the relationship between language use, cultural stereotypes and ethnic identity.

References Bradac, J.J. and Wisegarver, R. (1984) Ascribed status, lexical diversity, and accent: Determinants of perceived status, solidarity, and control of speech style. Journal of Language and Social Psychology 3, 239–255.

104

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Byrne, D. (1969) Attitudes and attraction. Advances in Experimental Social Psychology 4, 35–89. Collins, P.H. (1986) Learning from the outsider within: The sociological significance of Black feminist thought.Social Problems 33,514–532. Cook, S.W. (1962) The systematic analysis of socially significant events: a strategy for social research.Journal of Social Issues18, 66–84. Cooks, L.M. and Orbe, M. (1993) Beyond the satire: Selective exposure and selective perception in ‘In Living Color’. Howard Journal of Communications 4 (3), 217–233. Extra, G. and Yagmur, K. (eds) (2004) Urban Multilingualism in Europe: Immigrant Minority Languages at Home and School. Clevedon: Multilingual Matters. Forbes, D. (1997) Ethnic Conflict: Commerce, Culture, and the Contact Hypothesis. New Haven, CT: Yale University. Giles, H. (1977) Social psychology and applied linguistics: Towards an integrative approach. ITL: Review of Applied Linguistics 33, 27–42. Giles, H. and Billings, A.C. (2004) Assessing language attitudes: Speaker evaluation studies. In A. Davies and C. Elder (eds) The Handbook of Applied Linguistics, pp. 187–209. Oxford: Blackwell. Giles, H., Coupland, J. and Coupland, N. (1991) Contexts of Accommodation: Developments in Applied Sociolinguistics. Cambridge: Cambridge University Press. Giles, H., Mulac, A, Bradac, J.J. and Johnson, P. (1987) Speech accomodation theory: The first decade and beyond. In M.L. Mclaughlin (ed.) Communication Yearbook 10, pp. 13–48. Beverly Hills, CA: Sage. Giles, H. and Ogay, T. (2007) Communication accommodation theory. In B.B. Whaley and W. Samter (eds) Explaining Communication: Contemporary Theories and Exemplars, pp. 326–344. Mahwah, NJ: Lawrence Erlbaum. Giles, H., Taylor, D.M. and Bourhis, R.Y. (1973) Toward a theory of interpersonal accommodation through speech: Some Canadian data. Language in Society 2, 177–192. Gudykunst, W.B. (1995) Anxiety/uncertainty management (AUM) theory: Current status. In R.L. Wiseman (ed.) Intercultural Communication Theory, pp. 8–58. Thousand Oaks, CA: Sage. Kim, Y.Y. (2001) Becoming Intercultural: An Integrative Theory of Communication and CrossCultural Adaptation. Thousand Oaks, CA: Sage. Kreuger R.A. (1988) Focus Groups: A Practical Guide for Applied Research. London: Sage. Maass, A. and Arcuri, L. (1996) Language and stereotyping. In N. Macrae, M. Hewstone and C. Stangor (eds) The Foundations of Stereotypes and Stereotyping, pp. 193–226. New York: Guilford. Mac Cormaic, R. (2007) Following dreams far from home. Irish Times Online. 25 April. http://www.irishtimes.com (accessed 19 January 2009) Meinhof, U.H. and Triandafyllidou, A. (2006) (eds) Transcultural Europe. Cultural Policy in a Changing Europe, pp. 3–23. Basingstoke: Palgrave Meyerhoff, M. (2006) Introducing Sociolinguistics. London: Routledge. Mrozek, S. (1975) Emigranci. Krakow: Noir sur Blanc. National Economic and Social Council of Ireland (2006) Managing Migration in Ireland: A Social and Economic Analysis. A report by the International Organization for Migration for the National Economic and Social Council of Ireland, no. 116. Orbe, M. (1998) Conceptualizing Co-cultural Theory: An Explication of Culture, Power and Communication. Thousand Oaks, CA: Sage. Parsons, T. (1975) Some theoretical considerations of the nature and trends of change. of ethnicity. In N. Glazer and D.P. Moynihan (eds). Ethnicity: Theory and Experience, pp. 53–83. Cambridge, MA: Harvard University Press.

Divergence, Convergence and Passing for a Nat ive Speaker

105

Pavlenko, A. and Blackledge, A. (2004) Negotiation of Identities in Multilingual Contexts. Clevedon: Multilingual Matters. Pettigrew, T.F. (1998) Intergroup contact theory. Annual Review of Psychology 49, 65–85. Piller, I. (2002) Passing for a native speaker: Identity and success in second language learning. Journal of Sociolinguistics 6 (2) 179–206. Ruhs, M. (2007) Migrant workers face tough choices. Irish Times Online. 5 April. http:// www.irishtimes.com(accessed 19 January 2009). Sachdev, I. and Giles, H. (2004) Bilingual accommodation. In T. Bhatia and W. Ritchie (eds) The Handbook of Bilingualism, pp. 353–378. Oxford: Blackwell. Shain, Y. (1999) Marketing the American Creed Abroad: Diasporas in the US and their Homelands. Cambridge: Cambridge University Press. Shohamy, E. (2006) Language Policy: Hidden Agendas and New Approaches. London: Routledge. Singleton, D., Regan, V. and Debaene, E. (2008a) Second Language Acquisition and Native Language Maintenance in the Polish Diaspora in Ireland and France. Final report on a project funded by the Irish Research Council for the Humanities and Social Sciences. Singleton, D., Smyth, S. and Debaene, E. (2008b) ‘Polish diaspora in Ireland and France’ and ‘Our languages: Who in Ireland speaks and understands Russian?’ The rationale, structure and aims of two Dublin-based research projects. Paper presented at the Conference: ‘Contemporary migrations: Dilemmas of Europe and of Poland.’ Centre of Migration Research. November, University of Warsaw. Stewart, M.A., Ryan, E.B. and Giles, H. (1985) Accent and social class effects on status and solidarity evaluations. Personality and Social Psychology Bulletin 11, 98–105. Tajfel, H. and Turner, J.C. (1979) An integrative theory of intergroup conflict. In W.G. Austin and S. Worchel (eds) The Social Psychology of Intergroup Relations, pp. 33–47. Monterey, CA: Brooks/Cole. Thakerar, J.N., Giles, H. and Cheshire, J. (1982) Psychological and linguistic parameters of speech accommodation theory. In C. Fraser and K.R. Scherer (eds) Advances in the Social Psychology of Language, pp. 205–255. Cambridge: Cambridge University Press. Wandachowicz, K. (2002) Generacja Nic. Gazeta Wyborcza, 2002/09/05 p. 11. Williams, R.M., Jr. (1947) The Reduction of Intergroup Tensions. New York: Social Science Research Council.

6 The Direction of Causality in the Relationship Between Phonological Short-term Memory and L2 Lexical Knowledge: The Case of Adult Polish Learners of English in Ireland Agnieszka Skrzypek

Introduction In the last two decades the relationship between the rate of vocabulary learning and the ability to repeat nonwords (phonotactically licit made-up words that do not exist in the language on which they are based) has been the focus of intense research. Significant positive correlations between scores representing vocabulary knowledge and scores representing the ability to repeat nonwords or lists of nonwords have been reported (for an overview see e.g. Baddeley, 2007; Gathercole, 2006a, 2006b). As correlations are not in themselves indicative of the direction of causality, this relationship can be interpreted as either supporting the claim that phonological short-term 106

Phonological Short-term Memory and L2 Lexical Knowledge

107

memory (PSTM, conceptualized here as immediate memory for verbal wordlike material tapped by nonword tasks) causally affects vocabulary learning or that the reverse holds true. The current paper examines this issue in the context of adult L2 learning using the cross-lagged correlational paradigm.

The Conceptualization of Working Memory (WM) and Phonological Short-Term Memory (PSTM) A number of commonalities can be identified in current working memory (WM) models in relation to the conceptualization of WM and its processes. Current WM models conceptualize WM as ‘mechanisms and processes that are involved in the control, regulation, and active maintenance of taskrelevant information in the service of complex cognition’ (Miyake & Shah, 1999: 450). WM is viewed as a limited capacity construct, the efficiency of which can vary substantially among individuals. WM is typically assumed to comprise multiple components that account for some domain-specific effects, and the current study focuses on a component of WM that has been identified as particularly relevant in the context of L2 lexical learning. The component in question – phonological short-term memory (phonological STM, also referred to as the phonological loop and verbal STM) – is believed to be responsible for the manipulation and retention of verbal material. Within the most widely researched WM model, the so-called multi-component WM model (Baddeley & Hitch, 1974; Baddeley, 2000, 2007), phonological STM is viewed as a subsidiary system along with two other subsidiary systems (the visuospatial sketchpad and the more recently incorporated episodic buffer (Baddeley, 2000) and one supervisory attentional system (the central executive). Even though the terms short-term memory (STM) and working memory (WM) appear to be used interchangeably at times, there is a distinction between these two constructs according to current theoretical approaches to WM (see e.g. Baddeley et al., 2009). The term verbal or phonological STM is normally used to refer to performance on specific types of task that involve the retention of a small amount of verbal information, which is normally tested immediately after the stimulus is presented. The term WM, on the other hand, is used to denote mental operations that involve temporary storage of information and also some manipulation of that information. In this paper the term phonological STM (PSTM) is used to refer to the ability to store small amounts of verbal information over a brief interval.

108

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

PSTM Capacity and Vocabulary Knowledge: Correlational Data Correlational studies reveal that the relationship between PSTM and L1 vocabulary knowledge is particularly strong at lower levels of L1 proficiency. The diminishing role of PSTM in L1 vocabulary knowledge as familiarity with L1 increases has been documented in a number of studies (Gathercole et al., 1992; Papagno & Vallar, 1995; Speciale et al., 2004: Experiment 1). Strong correlations between PSTM and vocabulary knowledge at lower levels of L1 proficiency have led researchers to suggest that PSTM is responsible for long-term learning of novel word forms (Baddeley et al., 1998; Gathercole et al., 1997). At the beginning stage of L1 learning, each word form that a child encounters is a completely novel word form. Gradually the child acquires more and more robust L1 phonological knowledge, on the basis of which increasingly robust phonological generalizations are made (Edwards et al., 2004: 434). The more proficient the child is, the more robust his/her phonological generalizations become. To facilitate learning a new word, L1 learners can then access phonological representations of close neighbours of that word. For example, while learning the adjective unpleasant L1 users can access the adjective pleasant and the prefix un from LTM. It would, therefore, seem possible that, when a phonological representation of a close neighbour of a new word is available, a reduced amount of burden is placed on the limited-capacity PSTM component, rendering correlations between PSTM scores and L1 vocabulary scores weaker and in many cases non-significant. The picture presented in studies involving adult L2 learning is less conclusive than that emerging from developmental studies of children learning their L1. In a longitudinal study by Speciale et al. (2004: Experiment 2) no significant link between PSTM and L2 vocabulary learning was detected in adult subjects at the very beginning of L2 learning. A longitudinal study by O’Brien et al. (2006), on the other hand, suggests that a similar pattern to that observed in child L1 acquisition may be present in adult L2 learners, namely that the link between PSTM and vocabulary is stronger at lower levels of proficiency. In a group of adult L2 learners at an unspecified level of L2 proficiency (possibly more advanced beginners or lower-intermediate students) the link between PSTM and productive use of vocabulary in spoken output was stronger in those subjects who were classified as lower ability students. There are also a number of laboratory-based studies involving adults that indicate that at the onset of L2 learning the link between PSTM and L2 vocabulary learning is very strong (Craik & Watkins, 1973; Gupta, 2003; Papagno & Vallar, 1992; Speciale et al., 2004, Experiment 1).

Phonological Short-term Memory and L2 Lexical Knowledge

109

Debate on Casual Determinants in the Relationship Between PSTM and Vocabulary Knowledge There are two alternative accounts that propose opposite directions of causality in the relationship between PSTM tasks and vocabulary knowledge, namely the so-called PSTM account (Baddeley et al., 1998) and the linguistic account (Snowling et al., 1991). The PSTM account posits that individual differences in PSTM capacity causally determine performance on nonword tasks and vocabulary tasks, whereas the linguistic account proposes that it is vocabulary knowledge, and not PSTM, that is a causal determinant of performance on nonword tasks. Of the two accounts only the PSTM account subscribes to this notion that PSTM is causally linked with the growth of vocabulary knowledge. Both the PSTM account and the linguistic account acknowledge that the level of lexical/phonological knowledge affects performance on nonword tasks. One of the documentations of the beneficial effect of LTM on STM recall includes the lexicality effect, that is enhancement in serial recall accuracy for words as compared to nonwords (Hulme et al., 1991). Recall advantage explained in terms of LTM contribution is also found for high as compared to low-frequency words (the word frequency effect, e.g. Gregg et al., 1989), for lists presented in L1 as opposed to L2 (the language familiarity effect, e.g. Thorn et al., 2002), for nonwords containing highly frequent phoneme combinations (the phonotactic frequency effect) and for nonwords that are subjectively rated by listeners as resembling real words (the wordlikeness effect, e.g. Bailey & Hahn, 2001). The above-mentioned effects that reflect the influence of LTM on STM recall of verbal information can be explained in terms of reconstructive mechanisms involved in recalling partially degraded traces of verbal information. When phonological traces of words become partially degraded, it is argued that they may be reconstructed on the basis of long-term knowledge about the constituent words provided that such knowledge exists in the mind of the language user. The term redintegration was adopted in Schweickert’s (1993) multinomial processing tree model of serial recall to refer to this process of restoring partially decayed traced of words on the basis of available LTM representations. This term was then subsequently used by Hulme and colleagues (1991) to explain why recall of words from STM is better than recall of nonwords. Hume and colleagues’ argument was that in the case of word recall there are lexical representations available in LTM that may be used to repair partially decayed traces of words, whereas in the case of nonwords no such representations are available. Gathercole

110

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

et al. (1999) extended the idea of redintegration to nonwords by illustrating that recall of nonwords of high wordlikeness (that is words of high relative similarity to existing words in a given language) is better than recall of nonwords of low wordlikeness, and thus providing evidence of the operation of LTM influence of nonword tasks.

Attempts to Resolve the Debate on the Direction of Causality One frequently cited study that addressed the issue of the direction of causation in the link between PSTM and child L1 vocabulary learning is a longitudinal study by Gathercole et al. (1992). Eighty 4-year olds from a middle class background were subjected to PSTM and receptive vocabulary testing. The tests were re-administered when the children reached the ages of 5, 6 and 8. To explore the directionality of the association Gathercole and colleagues applied a cross-lagged partial correlational analysis, which involved calculating and comparing the strength of the correlations between early PSTM and later receptive vocabulary, and between early vocabulary and later PSTM scores. According to the assumptions of this type of analysis, the correlation should be stronger in the causal than in the noncausal direction. The pattern obtained for the link between PSTM and vocabulary at ages 4 and 5 suggests that PSTM capacity drives vocabulary learning rather than vice versa. However, this pattern shifted for age range 6–8 suggesting that expanding vocabulary knowledge might drive PSTM performance at higher levels of familiarity with the language. French (2003, 2004) also used a cross-lagged partial correlational analysis to investigate the directionality of the link between PSTM (Arabic nonword repetition) and L2 learning in 54 French children (mean age = 11.2 years) at a beginning level of L2 proficiency (presumably A2) enrolled in a five-month intensive English course in Quebec’s Saguenay Region. Partial cross-lagged correlations revealed that PSTM played a causal role in the children’s L2 development over time (including the development of L2 vocabulary), which is in line with the Gathercole et al. study (1992) in relation to the relationship between PSTM and L1 vocabulary learning at ages 4 and 5. The Gupta and Tisdale (2009) neural network model was developed to computationally addresses the issue of the interaction between phonological STM and existing lexical knowledge in a longitudinal setup. The model offers an account of nonword repetition performance and novel word form learning (referred to by the authors as phonological vocabulary learning), as

Phonological Short-term Memory and L2 Lexical Knowledge

111

well as the interaction between the above variables. Each word form is presented to the model one syllable at a time. After the word is presented, and when there is no longer any phonological information in its input layer, the model attempts to produce the word in its output layer. Over the course of many learning instances, the model learns new phonological sequences and also acquires generalized phonological knowledge. Gupta and Tisdale (2009) carried out a number of simulations within their neural network architecture in order to explore the opposition between the PSTM account and the linguistic account. One of their simulations (simulation 1) revealed that phonological vocabulary grows as a function of exposure in the simulated linguistic environment. Another simulation (simulation 5) showed that variation in PSTM functionality affects vocabulary learning and that with no PSTM functionality no vocabulary growth takes place. Gupta and Tisdale interpret this as support for the PSTM account, which posits that PSTM is a critical and causal determinant of vocabulary learning. However, even when PSTM functionality is not explicitly manipulated, nonword repetition ability increases as a function of increasing vocabulary knowledge (simulation 4). This could be interpreted as support for the PSTM account which acknowledges increased impact of lexical/phonological knowledge on performance on nonword tasks as vocabulary size increases. This could also partially be interpreted as support for the linguistic account which posits that lexical/ phonological knowledge plays an important causal role in performance on nonword tasks. The Gupta and Tisdale model indicates that this may be indeed the case, but only at higher levels of proficiency, when vocabulary knowledge has expanded to a sufficient level. Notably, the Gupta and Tisdale simulations do not support the view that at the beginning of language learning the involvement of PSTM in vocabulary learning is relatively minor, as the linguistic account stipulates.

Research Questions This paper addresses two research questions that focus on the relationship between PSTM and L2 vocabulary knowledge in adults. The first question examines the issue of the strength of this relationship at two levels of L2 proficiency defined in accordance with CEFR (Council of Europe, 2001), elementary (A2) and pre-intermediate (B1). The second question homes in on the direction of causality in the relationship between PSTM and L2 vocabulary knowledge. These two questions and respective predictions are outlined below.

112

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Research question 1: Does the relationship between PSTM and L2 vocabulary knowledge decrease with an increased level of L2 proficiency in adults, as has been shown to happen in the case of child L1 learning? It would seem plausible that a pattern of weakening of the relationship between PSTM and vocabulary knowledge should apply to adult L2 learning, as has been shown to happen in child L1 learning. If PSTM is a language learning device as Baddeley and colleagues argue (1998), its mechanisms of learning novel word forms should be applicable not only to the L1 context but also to the L2 context. However, it is also possible that the strength of the link between PSTM and subsequent L2 vocabulary knowledge may be affected by the fact that adults have another language system (their L1) at their disposal. Before embarking on the task of L2 learning, adult learners already have extensive knowledge of L1 phonological regularities and a rich network of L1 semantic information. When there is a high degree of resemblance between L1 and L2 phonological regularities, adults may facilitate the process of learning novel L2 words by consulting phonological regularities in their L1. In such a case, therefore, it would seem plausible that the impact of individual differences in PSTM capacity on vocabulary knowledge may be weaker than it is at initial stages of child L1 learning. When the degree of resemblance between L1 and L2 phonological regularities is lower (as is the case in Polish (L1) and English (L2) – a the languages under scrutiny in the current study), L1 long-term knowledge may in some instances still be consulted, but the extent of such consultations would be expected to be lower. As a result, the burden placed on PSTM while learning novel L2 words may produce significant correlations between variables in question.

Research question 2: What is the causal direction of the relationship between PSTM and the knowledge of high-frequency L2 vocabulary in the case of adult learners of L2 at two levels of proficiency – elementary (A2) and pre-intermediate (B1)? Is it the PSTM scores at Time 1 (the onset of the study) that contribute to the L2 vocabulary scores at Time 2 (six months after the onset), or is it the L2 vocabulary scores at Time 1 that contribute to the PSTM scores at Time 2? As indicated earlier, most studies linking PSTM measures with L1/L2 development are correlational in nature. As correlation does not imply causation, the direction of causation in this link is still open to debate. Even

Phonological Short-term Memory and L2 Lexical Knowledge

113

though PSTM is generally regarded as a strong predictor of word learning (e.g. Baddeley, 2003b; Papagno et al., 1991; Service, 1992), in principle it is plausible either that good PSTM functioning supports vocabulary development or that the reverse holds true. Provided that some significant correlations are detected in this study, there are at least three possible patterns that could emerge from this study. First, the cross-lagged correlations obtained in the current study could support the PSTM account, namely that individual differences in PSTM capacity are causally linked with subsequent vocabulary knowledge. Second, the obtained cross-lagged correlations could support the linguistic account, according to which the involvement of PSTM in vocabulary learning may be relatively minor, if any at all. Finally, it is also plausible the obtained cross-lagged correlations may reveal that the direction of causality is different at different proficiency levels. At the A2 level of proficiency, for example, in line with Baddeley and colleagues’ thinking individual differences in PSTM capacity may affect the efficiency of novel L2 word learning, thus causing substantial differences in subsequent knowledge of L2 vocabulary. At the B1 level of proficiency, when knowledge of L2 phonological regularities is broader than at the A2 level, PSTM may turn out to be no longer causally linked with subsequent L2 vocabulary knowledge.

Participants The sample comprises 60 adult Polish learners of English resident in Ireland (age range 25–35), 30 of which were at the A2 level of proficiency and 30 at the B1 level. Proficiency levels defined in relation to the Common European Framework of Reference (CEFR; Council of Europe, 2001) and were measured by the OUP Pen and Paper Placement Test (2001). The longitudinal data obtained in the course of this study came from 24 males and 36 females (see Table 6.1 for more information about the participants). The individuals who expressed willingness to participate in our project were offered a six-month English language course at Trinity College Dublin (TCD) free of charge (see Skrzypek, 2010 for a detailed description of the TCD course). None of our subjects attended any other English language course for the duration of this study. As a certain amount of dropout from the TCD course was anticipated owing to motivational factors or other unforeseen circumstances, we recruited over 100 subjects. Approximately 40% of the learners who signed up for the course had to be excluded from our analyses – a few individuals reported having hearing problems or dyslexia, some subjects dropped out before completing the course, whereas others failed to attend one of the testing sessions.

114

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Table 6.1 Background information about participants Group

A2 (n = 30)

B1 (n = 30)

Gender

13 males 17 females M = 29.3 SD = 4.091 M = 17.16 SD = 7.61 Primary Secondary Vocational Tertiary Other (e.g. private tuition) Secondary Vocational Tertiary Russian German French Italian Spanish One Two

11 males 19 females M = 30.4 SD = 3.654 M = 24 SD = 11.76 Primary Secondary Vocational Tertiary Other (e.g. private tuition) Secondary Vocational Tertiary Russian German French Italian Dutch One Two

Age (years) Residence in Ireland (months) Context of first exposure to English

Education

Other foreign languages

Number of other foreign languages per student

8 8 0 5 9

26.67% 26.67% 0% 16.67% 30%

14 3 13 15 14 4 1 1 25 5

46.67% 10.0% 43.33% 50% 46.67% 13.33% 3.33% 3.33% 83.33% 16.67%

8 7 0 8 7

26.67% 23.33% 0% 26.67% 23.33%

3 1 26 13 20 4 2 1 23 7

10% 3.33% 86.67% 43.33% 66.67% 13.33% 6.67% 3.33% 76.67% 23.33%

Operational Definitions and Research Instruments The current study identifies seven variables in each of the two proficiency groups. PSTM capacity was tapped by two types of nonword tasks (serial nonword recall and by serial nonword recognition) before the commencement of the TCD English language course (Time 1) and at the end of this six-monthlong course (Time 2), thus yielding four sets of scores at each proficiency level (see Table 6.2). Knowledge of high-frequency L2 vocabulary was measured by means receptive and productive vocabulary size tests from relevant frequency bands (1 K–2 K for the A2 level; 2 K–3 K for the B1 level) at Time 1 and Time 2, and yielded two sets of scores at each proficiency level. The last variable representing the amount of exposure to L2 outside the classroom between

Phonological Short-term Memory and L2 Lexical Knowledge

115

Table 6.2 Constructs and corresponding measures Underlying construct

Corresponding test scores A2 (elementary)

Serial nonword recall scores at Time 11 Serial nonword recall scores at Time 22 Serial nonword recognition PSTM capacity scores at Time 1 (measured without articulation) Serial nonword recognition scores at Time 2 Exposure to L2 outside Exposure scores3 to L2 the classroom between Time 1 and Time 2 1K–2K Vocabulary Levels Test Knowledge of highscores at Time 1 frequency L2 vocabulary (not 1K–2K Vocabulary Levels Test directly related to scores at Time 2 class content)

PSTM capacity (measured with articulation)

1 2 3

B1 (pre-intermediate) Serial nonword recall scores at Time 1 Serial nonword recall scores at Time 2 Serial nonword recognition scores at Time 1 Serial nonword recognition scores at Time 2 Exposure scores to L2 between Time 1 and Time 2 2K–3K Vocabulary Levels Test scores at Time 1 2K–3K Vocabulary Levels Test scores at Time 1

Time 1 – before the commencement of the TCD English language course Time 2 – after the end of the six-month-long TCD course Exposure scores – the average number of hours per day recorded in students’ journals between Time 1 and Time

Time 1 and Time 2, was included in our research design as a potentially confounding variable. A detailed description of the piloting of all research instruments and can be found in Skrzypek (2010: 144–172) along with reliability coefficients obtained during the main study (Skrzypek, 2010: 213).

PSTM tasks: Serial nonword recall and recognition For the purpose of this study two operational definitions of PSTM capacity were formulated – one involving articulation of lists of nonwords and the other involving passive recognition sets of nonword lists. The two operational definitions of PSTM adopted here are as follows: (1) serial nonword recall performance – the ability to retain and repeat L1-based nonword lists of varying lengths immediately after the presentation of each list with the correct nonword order maintained at Time 1 and Time 2;

116

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

(2) serial nonword recognition performance – the ability to retain sets of two L1-based nonword list of varying lengths and to judge, immediately after the presentation of a given set, whether the nonwords within each set are presented in the same order at Time 1 and Time 2. It should be pointed out that from a methodological perspective there are merits to using two PSTM measures instead of relying on one measure exclusively (Gathercole & Pickering, 1999; Service, 2008, personal communication). One of the merits stems from the fact that, contrary to the earliest writings on the subject, PSTM tasks do not provide a pure measure of the construct. Apart from memory functions PSTM tasks are also known to tap other processes, such as e.g. speech-motor output processing (Gathercole, 2006a: 528–531). One of the implications is that, for example some individuals could obtain substantially lower serial nonword recall scores than their ‘true’ PSTM score as a result of some minor articulation problems. To ensure that the PSTM capacity of subjects with some (even minimal) output problems is not underestimated by using a recall measure only, the use of serial nonword recognition alongside Serial Nonword Recall has been strongly recommended (Gathercole et al., 1999: 66). In subjects who do not have any output problems, the two measures should be highly correlated (provided that nonwords of low wordlikeness are used, as explained below). The serial nonword recall task used in this study consisted of three sets of nonword lists, each made up of 15 lists of the same length. Set one, two and three contained 2-item, 3-item and 4-item lists respectively (see Skrzypek, 2010: 288–289). Before the commencement of the task the subjects were involved in a 5-minute trial session during which they were familiarized with the task procedure. No corrective feedback was provided during testing. The measure of performance on the serial nonword recall test was the number of correctly repeated lists across all list lengths tested. A repeated list had to contain no mistakes to be accepted as correct. The testing was discontinued if a subject failed to repeat eight out of 15 lists of a given length. The serial nonword recognition task comprised 30 pairs of nonword lists with 10 pairs of nonword sequences at each of three list lengths, that is at 4-item, 5-item and 6-item lengths (see Skrzypek, 2010: 290–291). The position of each nonword was controlled to ensure the nonwords occurred in a variety of positions within the pool of lists. At each list length, five of the ten pairs of nonword sequences were identical and the remaining five shared exactly the same nonwords but two of the nonwords in question were transposed in the second sequence. The initial and final pseudowords never changed their position. The participants were instructed to listen to each

Phonological Short-term Memory and L2 Lexical Knowledge

117

set of lists and tick either ‘the same’ or ‘different’ (or ‘not sure’) on a designated webpage. The subjects were told they would listen to each set only once. One point was scored for a correctly recognized set of lists as either the same or different. The nonwords created for use in this project were pronounceable phonological sequences according to Polish phonotactics. Each item is a three-syllable CV–CCV–CV combination. The constituent syllables were selected from a Polish language dictionary (Polish Scientific Publishers PWN – CD-ROM edition) and a frequency dictionary of Polish (Kurcz et al., 1990). The position in which these syllables appear in the constructed nonwords reflects their position in real three-syllable Polish words (i.e. initial, medial and final). Wordlikeness ratings were obtained for all nonwords prior to the commencement of the study in order to exclude those items that were too reminiscent of real Polish words. Wordlikeness ratings reflect the degree to which novel syllable sequences resemble existing words. It is believed that the lower the perceived degree of wordlikeness, the more limited the involvement of long-term memory and the purer the measure of PSTM (see e.g. Gathercole & Pickering, 1999). The nonwords with the lowest wordlikeness ratings were selected for inclusion in the recall and recognition tasks (see Skrzypek, 2010: 362). The reason why nonwords were based on L1 (Polish) phonotactics, and not L2 (English) phonotactics, derived from the fact that this was an attempt to employ PSTM measures that would produce consistent results when administered over a period of time. In adult L2 learning both L1- and L2based nonwords have been employed (e.g. O’Brien et al., 2006; Speciale et al., 2004). In the serial nonword recognition format, L1-based nonwords have been shown to produce stable results over a period of time (O’Brien et al., 2006), and it is not unlikely that the same pattern may apply in the recall format. Additionally, were L2-based nonwords to be used, the fact that some participants spoke English with a very strong Polish accent could have affected the accuracy of scoring L2-based nonword tasks.

Vocabulary Levels Test In many frameworks (e.g. Nation, 1990; Qian, 2002) there is a consensus that vocabulary knowledge comprises at least two dimensions, that is vocabulary breadth/size and vocabulary depth. Vocabulary breadth refers simply to the number of words that the learners are familiar with (even if the understanding of the meaning is only superficial), while depth of vocabulary refers to the learners’ knowledge of various aspects of words under investigation, such as e.g. word associations.

118

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

This study focuses on gains in high-frequency L2 vocabulary breadth, that is selected L2 words from frequency lists containing the first (1 K), second (2 K) and third 1,000 (3 K) most frequent L2 words. The operational definition of L2 vocabulary knowledge used in the current study is as follows: knowledge of the receptive and controlled productive general vocabulary breadth defined in relation to the English word frequency lists used in the creation of Vocabulary Levels Test (Nation, 2001; Schmitt et al., 2001) – at Time 1 and Time 2. The 1 K–2 K Vocabulary Levels Test and the 2 K–3 K Vocabulary Levels Test were administered to the A2 and B1 groups respectively. This 1 K–2 K test battery included the receptive 1000-word test (Nation, 2001), the receptive 2000-word test (Nation, 2001 modified by Schmitt et al., 2001; Schmitt, 2000), the productive 1000-word test (Skrzypek, 2010) and the 2000-word productive test (Laufer & Nation, 1999). The 2 K–3 K test battery included the receptive 2000-word and 3000-word tests (Nation, 2001 modified by Schmitt et al., 2001; Schmitt, 2000), the productive 2000-word and 3000word tests (Laufer & Nation, 1999). A detailed justification as to why these frequency bands were chosen is presented in Skrzypek (2010). Each of the above tests was scored on a dichotomous scale (0,1) and in the productive component grammatical and minor spelling mistakes were disregarded. Because of the fact that the number of items on the receptive and productive components differed, the obtained scores were turned into percentages, added and averaged.

Exposure to L2 Outside the Classroom The participants were resident in Ireland (Dublin) throughout the duration of the TCD language course and were, therefore, exposed to some amount of L2 outside the classroom. As Poles were the most dominant migrant group in Ireland in the two years preceding 2008 (see Census (2006, 2009), we could not assume that the patterns of exposure to L2 would be similar for all of participants. A considerable number of the Polish migrants interviewed in 2006 and 2007 under an umbrella of the larger Polish Diaspora Project reported that they could go about their daily routines interacting mainly in L1 (Skrzypek et al., in press). They pointed to the fact that a number of services were available in Polish (Polish shops, Polish schools, Catholic masses in Polish, Polish legal advice, Polish film festivals, etc.). Some of them reported socialising mainly with other members of the Polish community in Ireland owing to a common linguistic and cultural background. Among the Poles we had interviewed prior to 2008, there were also

Phonological Short-term Memory and L2 Lexical Knowledge

119

a number of individuals who reported interacting through L2 mainly and using hardly any L1 on a daily basis. In the current study exposure (to L2 outside the classroom) was defined as: • •

interacting in L2 with L2 speakers – face-to-face, via the phone or on the internet, watching TV in L2, reading books, newspapers, etc. in L2, listening to radio programmes in L2.

Exposure was measured in terms of hours per day between Time 1 and Time 2. The subjects were asked to keep a diary for six weeks (one selected week of each month during the TCD language course), in which they were requested to note the number of hours of exposure to L2 outside the classroom per day.

Research Methods As the first research question does not make any assumptions about the direction of causality, correlations and some partial correlations were computed and interpreted. In relation to the second research question, the same statistical method was employed as that used in the Gathercole et al. (1992) study, i.e. cross-lagged correlational analysis of longitudinal data. By way of background, the rudiments of a cross-lagged panel correlation paradigm require two constructs, X and Y, measured at two different points in time. As shown in Figure 6.1, the two constructs generate four variables

Figure 6.1 Cross-lagged panel correlation paradigm

120

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

(Variable X at Time 1, Variable X at Time 2, Variable Y at Time 1 and Variable Y at Time 2) and six correlations among the four variables. Two of the six correlations, Px1y2 and Px2y1, are referred to as cross-lagged. Px1y2 represents a correlation between Variable X at Time 1 and Variable Y at Time 2, while Px2y1 represents a correlation between Variable X at Time 2 and Variable Y at Time 1. According to the assumptions of a cross-lagged correlational analysis, the correlation should be stronger in the causal than in the noncausal direction (e.g. Campbell, 1988; Campbell & Stanley, 1963; Crano et al., 1972; Crano & Mellon, 1978). To test for the significance of the differences between the cross-lagged correlations in this study a formula attributed to Pearson and Filon (cited in Kenny, 1975: 896) was used. As the formula is not easily accessible, it is reproduced in Equation 1. z=

N (r14 − r23 ) 2 2

(1 − r14 ) + (1 − r232 )2 − k

k = (r12 − r24r14 )(r34 − r24r23 ) + (r13 − r12r23 )(r24 − r12r14 ) + (r12 − r13r23 )(r34 − r13 r14 ) + (r13 − r14r34 ) + (r24 − r34r23 ) N–sample size 1,2,3,4–variables r–correlation coefficient r14 –correlation between variable 1 and variable 4

Results This paper presents results of statistical tests for parametric data. A failure to properly check data assumptions can result in the use of inappropriate statistical procedures and, consequently, flaw the logic behind hypothesis testing. Assumptions of parametric data were, therefore, checked thoroughly before conducting the relevant tests. These checks included, inter alia, standardizing the values of skewness and kurtosis, and running the Shapiro–Wilk test (see Tables A.2 and A.3 in the Appendix), analysing boxplots and histograms, generating normality plots, and where relevant also checking the assumption of homogeneity of variance (Levene’s test) and the assumption of linearity (scatterplots). All sets of data reported below are normally distributed with no outliers present (see Skrzypek, 2010, for more information).

Phonological Short-term Memory and L2 Lexical Knowledge

121

Group Characteristics in Terms of PSTM and L2 Vocabulary The PSTM characteristics of the participants at Time 1 were as follows. In serial noword recall the A2 group and the B1 group obtained mean scores of 15.6 (SD = 4.9, min/max = 6/25) and 17.5 (SD = 6.3, min/max = 6/30) respectively. In serial noword recognition the A2 group and the B1 group obtained mean scores of 16.4 (SD = 3.9, min/max = 9/24) and 18.1 (SD = 3.9, min/ max = 10/25) respectively. The means of the A2 and B1 serial nonword recall scores are not statistically different (t(58) = − 1.30, p > 0.05), and neither are the means of A2 and B1 serial nonword recognition scores (t(58) = − 0.63, p > 0.05). Serial nonword recall and recognition scores are highly correlated in both groups (r = 0.683 and r = 0.645 in the A2 and B1 group respectively; ps < 0.001). The group means of PSTM scores at Time 2 were very similar to those obtained at Time 1 (see Table A.2 in the Appendix). The differences between group means at Time 1 and Time 2 were non-significant for each pair of measures. This implies that group means did not change significantly between Time 1 and Time 2 (t(29) = 0.290 for the A2 serial nonword recall scores at Time 1 and Time 2; t(29) = 0.976 for the A2 serial nonword recognition scores at Time 1 and Time 2; t(29) = 0.094 for the B1 serial nonword recall scores at Time 1 and Time 2; t(29) = − 0.088 for the B1 serial nonword recognition scores at Time 1 and Time 2; ps > 0.05). The 1 K–2 K Vocabulary Levels Test scores in the A2 group at Time 1 ranged from 30.4% to 70.3% (see Table A.3 in the Appendix). The minimum score obtained at Time 2 was over 1% higher than that obtained at Time 1, namely 31.7%, while the maximum score was over 10% higher than that at Time 1 reaching 80.9%. The difference between Time 1 mean (49.2%) and Time 2 mean (56.1%) is highly significant (t(29) = 9.243, p < 0.001).

Correlations and Selected Partial Correlations The first research question addresses the issue whether the relationship between PSTM and subsequent L2 vocabulary knowledge in adults decreases with an increased level of L2 proficiency. Our data show that the correlations between serial nonword recall (Time 1) and vocabulary measures in the B1 group are indeed weaker than those obtained in the A2 group (see Tables 6.3 and 6.4). In the B1 group serial nonword recall (Time 1) is linked with the 2 K–3 K Vocabulary Levels Test at the 0.05 level (Time 2) (r = 0.413, p = 0.023). In the A2 group serial nonword

122

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Table 6.3 Simple intercorrelations among variables in the A2 group Variables 1. Serial nonword recall (Time 1) 2. Serial nonword recognition (Time 1) 3. Exposure to L2 (between Time 1 & 2) 4. 1K–2K Vocabulary Levels Test gainsA 5. 1 K-2 K Vocabulary Levels Test (Time 1) 6. 1 K-2 K Vocabulary Levels Test (Time 2) Note: A

1 r sig. r sig. r sig. r sig. r sig. r sig.

2

3

4

5

6

– 0.683*** 0.000 0.281 0.133 0.459* 0.011 0.420* 0.021 0.548** 0.002

– 0.294 0.155 0.384* 0.036 0.359 0.052 0.466* 0.010

– 0.461* 0.010 0.023 0.904 0.164 0.385

– −0.056 – 0.769 0.254 0.952*** 0.176 0.000

–

*p < .05; **p < .01; ***p < .001

1K – 2K Vocabulary Levels Test gains – scores at Time 1 subtracted from scores at Time 2

Table 6.4 Simple intercorrelations among variables in the B1 group Variables 1. Serial nonword recall (Time 1) 2. Serial nonword recognition (Time 1) 3. Exposure to L2 (betweenTime 1 & 2) 4. 2K–3K Vocabulary Levels Test gainsA 5. 2 K-3 K Vocabulary Levels Test (Time 1) 6. 2 K-3 K Vocabulary Levels Test (Time 2) Note: A

1 r sig. r sig. r Sig. r sig. r sig. r sig.

2

3

4

5

6

– 0.645*** 0.000 0.555* 0.001 0.367* 0.046 0.227 0.228 0.413* 0.023

– 0.185 0.328 0.218 0.246 0.189 0.317 0.304 0.102

– 0.473* 0.008 0.214 0.257 0.447* 0.013

– −0.067 0.726 0.036 0.850

– 0.908*** 0.000

–

*p < .05; **p < .01; ***p < .001.

2K – 3K Vocabulary Levels Test gains – scores at Time 1 subtracted from scores at Time 2

recall (Time 1) and the 1 K–2 K Vocabulary Levels Test (Time 2) are correlated at the 0.01 level (r = 0.548, p = 0.002). In the B1 group the correlation between serial nonword recall (Time 1) and the 2 K–3 K Vocabulary Levels gain scores is significant at the 0.05 level (r = 0.367, p = 0.046). In the A2

Phonological Short-term Memory and L2 Lexical Knowledge

123

group the correlation between the corresponding variables is also significant at the 0.05 level but noticeably stronger (r = 0.459, p = 0.011). The correlations between our second PSTM measure, serial nonword recognition (Time 1), and vocabulary measures present a similar picture to that sketched above. In the A2 group serial nonword recognition (Time 1) correlates with the 1 K–2 K Vocabulary Levels Test scores (Time 2) (r = 0.466, p < 0.05) and with the 1 K–2 K Vocabulary Levels Test gain scores (r = 0.384, p < 0.05). In the B1 group serial nonword recognition (Time 1) correlates neither with the 2 K–3 K Vocabulary Levels Test gain scores (r = 0.218, p = 0.246) nor with the 2 K–3 K Vocabulary Levels Test at Time 2 (r = 0.304, p = 0.102). In the light of the above findings one could argue that, as L2 proficiency increases, the relationship between PSTM and subsequent L2 vocabulary knowledge becomes weaker. The gains on the 1 K–2 K Vocabulary Levels Test between Time 1 and Time 2, apart from being linked to the subjects’ PSTM scores, were also linked to their exposure to highly frequent words outside the classroom. The exposure scores are correlated with the 1 K–2 K Vocabulary Levels Test gain scores at the 0.05 level (r = 0.461). As in the A2 group, the exposure variable in the B1 group is significantly linked to the 2 K–3 K Vocabulary Levels Test gain scores (r = 0.473, p < 0.01) and the 2 K–3 K Vocabulary Levels Test scores at Time 2 (r = 0.447, p < 0.05). The above correlations indicate that the B1 learners with a higher level of exposure to L2 outside the classroom (between Time 1 and Time 2) obtained significantly higher scores on the 2 K–3 K Vocabulary Levels Test at Time 2, and that their gains on the test were also significantly higher than the gains of the remaining B1 students. This finding is in keeping with the Gupta and Tisdale (2009) computational model of STM (simulation 1). It should also be noted that although the correlations between serial nonword recall (Time 1) and L2 vocabulary measures (Time 2 and gain scores) in the B1 group are significant at the 0.05 level, they become non-significant when exposure is partialled out (ps > 0.05). This is caused by the fact that there is a high correlation between serial nonword recall (Time 1) and exposure in the B1 group. As both serial nonword recall (Time 1) and exposure correlate with L2 vocabulary measures (Time 2 and gain scores), it is statistically impossible to separate the influence of serial nonword recall (Time 1) and exposure on L2 vocabulary scores.

Cross-lagged Correlations The second research question addresses the issue of the direction of causality in the relationship between PSTM and knowledge of L2 vocabulary.

124

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

In order to discriminate between the PSTM account and the linguistic account, cross-lagged correlations were calculated, and the differences between these correlations were tested to establish whether they were statistically significant (see Equation 1). The cross-lagged correlational analyses involving serial nonword recall and L2 vocabulary measures provide a very clear indication of the direction of causality at the A2 level, but fail to do so at the B1 level (see Figure 6.2). In the A2 group the cross-lagged correlation between serial nonword recall (Time 1) and the 1 K–2 K Vocabulary Levels Test (Time 2) is significant at the 0.01 level (r = 0.548), and appears stronger than the converse cross-lagged correlation between the 1 K–2 K Vocabulary Levels Test (Time 1) and serial nonword recall (Time 2) (r = 0.389, p < 0.05). The difference between these cross-lagged correlation coefficients is statistically significant (z = 2.497, p < 0.05). In the B1 group the cross-lagged correlation between serial nonword recall (Time 1) and the 2 K–3 K Vocabulary Levels Test (Time 2) is not significantly different from the cross-lagged correlation between the 2 K–3 K Vocabulary Levels Test (Time 1) and serial nonword recall (Time 2) (r = 0.413 and r = 0.364 respectively, both ps < 0.05; z = − 0.198, p > 0.05). The pattern of results obtained at the A2 level can be interpreted as support for the PSTM account, according to which PSTM is a causal determinant of subsequent vocabulary knowledge.

Figure 6.2 Cross-lagged correlations between PSTM scores and L2 vocabulary scores in A2 and B1 groups

Phonological Short-term Memory and L2 Lexical Knowledge

125

The cross-lagged correlational analyses involving serial nonword recognition and L2 vocabulary measures do not allow us to make inferences about the direction of causality in this relationship. In the A2 group the correlation between serial nonword recognition (Time 1) and the 1 K–2 K Vocabulary Levels Test (Time 2) appears to be stronger than the correlation between the 1 K–2 K Vocabulary Levels Test (Time 1) and serial nonword recognition (Time 2) (r = 0.466, p < 0.01 and r = 0.333, p > 0.05 respectively). However, the difference between these cross-lagged correlations is not significant (z = 1.254, p > 0.05). In the B1 group the cross-lagged correlation between serial nonword recognition (Time 1) and the 2 K–3 K Vocabulary Levels Test (Time 2) and the cross-lagged correlation between serial nonword recognition (Time 2) and the 2 K–3 K Vocabulary Levels Test (Time 1) is nonsignificant (ps > 0.05), and therefore the difference between these two correlations was not statistically tested.

Conclusions PSTM and L2 vocabulary knowledge at two proficiency levels The current longitudinal study points to the existence of a strong link between PSTM and subsequent L2 vocabulary knowledge in adults at a relatively low level of L2 proficiency (A2). This strong link was established for both of the measures used to tap individual differences in PSTM capacity, namely L1-based serial nonword recall (involving articulation) and L1-based serial nonword recognition (not involving overt articulation). Our data reveal that in adult L2 learners at the A2 level of proficiency a lower ability to retain small amounts of verbal information in STM has a detrimental impact on the efficiency of the learning of phonological forms of new L2 words. Adult L2 learners with a lower PSTM capacity are, therefore, likely to need more exposure to L2 and more repetition to be learn new phonological forms. The present study also indicates in B1 learners the link between PSTM and subsequent L2 vocabulary knowledge is noticeably weaker as compared to A1 learners. In Gathercole’s (2006a) words, at higher levels of L2 proficiency learners may no longer need to be reliant on PSTM, as they can use their mental lexicons to ‘capitaliz(e) on knowledge of structures (which may be semantic, conceptual, or phonological in form) that have already been constructed’ (p. 515). The above results are in contrast with results obtained in two other longitudinal studies that explored the link between PSTM and adult L2 language learning. Speciale et al. (2004: Experiment 2), did not detect a significant link between PSTM at Time 1 (nonword repetition based neither on L1 nor L2) and

126

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

receptive L2 vocabulary knowledge at Time 2 in complete beginners. O’Brien et al. (2006) also failed to detect a significant link between PSTM at Time 1 (L1-based serial nonword recognition) and free productive knowledge of L2 vocabulary in speech at Time 2 (L2 proficiency level not clearly specified). However, when O’Brien et al. divided their subjects into lower and higher ability learners, a significant correlation was detected in lower ability learners. One reason why our study yielded different results from those conducted by Speciale et al. (2004: Experiment 2) and O’Brien et al. (2006) could be the nature of PSTM and L2 vocabulary measures used. Regarding L2 vocabulary, O’Brien et al. measured free productive knowledge of L2 vocabulary in spoken output, while Speciale et al. measured the students’ ability to arrange some high-frequency words into 10 pairs that could be reasonably contrasted (Speciale et al. do not provide examples). As only certain words can be contrasted in terms of their meaning, it is not unlikely that the Speciale et al. receptive L2 vocabulary task was too restricted in its content to reflect the amount of progress their students made over a seven-week period of L2 instruction. In the O’Brien et al. study, the type of PSTM measure used (L1-based serial nonword recognition) might not have adequately reflected ‘true’ PSTM capacity in the population of adult learners. O’Brien and colleagues did not raise any concerns about the use of this particular PSTM measure in adults. Our concerns in this connection – based on our observations from the current study – are briefly outlined later. In the current study, all research instruments were extensively piloted and amended before the main study where required. O’Brien et al. and Speciale et al. make no mention of piloting, and nor do they make any mention of checking the assumptions of parametric data. Regarding the O’Brien et al. study, it should also be mentioned that there were some confounding variables that the authors did not consider at the time, including the possible impact of learning environment on the learning outcomes (some of their subjects learnt L2 in an L1-speaking environment, while others in an L2-speaking environment). Considering the above shortcomings, it is rather difficult to assess the extent to which correlations obtained in these two studies can be generalized to the underlying population.

Direction of causality in the relationship between PSTM and L2 vocabulary: The PSTM account and the linguistic account The pattern of results obtained at the A2 proficiency level lends support to the PSTM account (Baddeley et al., 1998), according to which PSTM capacity casually affects subsequent vocabulary knowledge. Apart from indicating that there is a strong link between the variables under scrutiny, the crosslagged correlations we generated provide evidence that PSTM is a causal

Phonological Short-term Memory and L2 Lexical Knowledge

127

determinant of subsequent L2 vocabulary knowledge. The cross-lagged correlations obtained in the B1group also appear to be in keeping with the PSTM account, according to which the link between PSTM and vocabulary learning is likely to weaken as language proficiency increases. As a result, the finding by Gathercole et al. (1992) in relation to PSTM and L1 learning in children aged 4 and 5 can be extended to early stages of L2 vocabulary learning in adult L2 learners. It should be pointed out, however, that only our serial nonword recall data allow us to draw conclusions about the direction of causality at the A2 level. The cross-lagged correlation between serial nonword recognition and relevant L2 vocabulary measure was stronger than the converse cross-lagged correlation, but the difference between these two correlations turned out to be non-significant. Our data also support a recent computational model by Gupta and Tisdale (2009), which demonstrates that variation in PSTM capacity has an impact on the speed of vocabulary learning. According to the Gupta and Tisdale model ‘PSTM functionality is a critical and causal determinant of phonological vocabulary learning – without PSTM functionality, there is no vocabulary learning/growth’ (Gupta & Tisdale, 2009: 496). The Gupta and Tisdale model also demonstrates the impact of the level of exposure on vocabulary learning, which is also supported by our data in the context of adult L2 vocabulary learning. If our PSTM measures contained L2-based nonwords, significant correlations between the L2 vocabulary scores (Time 1) and serial nonword recall (Time 2) could be interpreted as an indication of the influence of the lexical/ phonological knowledge on subsequent PSTM performance. In the Gupta and Tisdale (2009) computational model, nonword repetition ability increases as a function of increasing vocabulary knowledge and increasing knowledge of phonological regularities. In the current study serial nonword recall is not L2-based, and therefore it does not share the phonological content with L2 vocabulary. As a result, the impact of L2 lexical/phonological knowledge (Time 1) on performance on L1-based serial nonword recall (Time 2) should be expected to be minimal, if any at all. The occurrence of this significant correlation in our dataset could be linked to the fact that the recall scores remain stable over time, and also to the high correlations between the recall scores at Time 1 and Time 2 and between the L2 vocabulary scores at Time 1 and Time 2.

Concerns about reliability and construct validity of a recognition-based measure of PSTM in adult learners One of our concerns relates to the level of reliability of the serial nonword recognition measure. The reliability coefficients of the recognition measure

128 Linguist ic and Cultural Acquisit ion in a Migrant Communit y

were lower than we would ideally wish for, despite having adhered to guidelines that have been followed by other PSTM researchers using nonword tasks. The reliability coefficients for serial nonword recognition ranged from 0.704 to 0.831 (Spearman-Brown split-half reliability coefficients), which is relatively low. The reliability coefficients for serial nonword recall, however, ranged from 0.897 to 0.918, indicating a much higher level of reliability. Another cautionary note relates to how the recall and the recognition task appear to be conceptualized within the multi-component WM model. In this model PSTM is theorized to comprise two sub-components, the phonological store that is responsible for creating new phonological traces and the articulatory rehearsal process that supports the trace creation. According to Baddeley (2003a), serial recall of phonological information relies on the phonological store and the articulatory rehearsal process, while serial recognition of phonological information relies on the phonological store only. Potentially, therefore, it would seem that comparing the results obtained by means of these two task formats might provide information about how these two subcomponents operate in relation to L2 learning. This has been attempted in relation to child L1 acquisition (Gathercole et al., 1999). Normal adults unlike young children, however, are known to employ rehearsal strategies that rely on subvocal rehearsal for lists that exceed the capacity of the phonological store. Many of the subjects who participated in the current study reported engaging in subvocal rehearsal while completing the recognition task. It would, therefore, seem hasty to conclude that in adults serial nonword recognition taps the phonological store but not the articulatory rehearsal process. This brings up the question of whether serial nonword recognition taps the same construct in adults as it is believed to tap in children. The serial nonword recognition task has mainly been used in research on PSTM and child L1 learning, and has not yet been extensively explores in studies of adult L2 learning. The task was employed by Gathercole and colleagues in a series of innovative studies (e.g. Gathercole et al., 2001) to tap PSTM capacity in children who might have (even minor) articulatory difficulties. To the best of our knowledge, serial nonword recognition has so far been used in only one other longitudinal study of adult L2 learning (O’Brien et al., 2006), and the study in question did not aim to compare this procedure with a more widely used serial nonword recall procedure. In a recent study of artificial L2 learning in adults (Martin, 2009) a relatively weak correlation between the serial nonword recognition task and the nonword repetition task was detected (r = 0.34, p < 0.05), which the author interprets as a possible problem with reference to the construct validity of the serial nonword recognition task (Martin, 2009: 47–48). Despite the fact that in the current study the correlation between serial nonword recall and serial

Phonological Short-term Memory and L2 Lexical Knowledge

129

nonword recognition was significant at the 0.001 level, we feel it is essential that the issue of construct validity of the serial nonword recognition task be explored in future studies when this is used in an adult population.

References Baddeley, A.D. (2000) The episodic buffer: A new component of working memory? Trends in Cognitive Sciences 4 (11), 417–423. Baddeley, A.D. (2003a) Working memory and language: An overview. Journal of Communication Disorders 36 (3), 85–107. Baddeley, A.D. (2003b) Working memory: Looking back and looking forward. Nature Reviews Neuroscience 4, 829–839. Baddeley, A.D. (2007) Working Memory, Thought, and Action. Oxford: Oxford University Press. Baddeley, A.D., Eysenck, M.W. and Anderson, M.C. (2009) Memory. Hove: Psychology Press. Baddeley, A.D., Gathercole, S.E. and Papagno, C. (1998) The phonological loop as a language learning device. Psychological Review 105 (1), 158–173. Baddeley, A.D. and Hitch, G.J. (1974) Working memory. In G. Bower (ed.), The Psychology of Learning and Motivation, Vol. 8, pp. 47–90. New York: Academic Press. Bailey, T.M. and Hahn, U. (2001) Determinants of wordlikeness: Phonotactics or lexical eighborhoods? Journal of Memory and Language 44, 568–591. Campbell, D.T. (1988) Methodology and Epistemology for Social Science: Selected Papers. London: University of Chicago Press. Campbell, D.T. and Stanley, J.C. (1963) Experimental and Quasi-experimental Designs for Research. Boston: Houghton Mifflin Company. Council of Europe (2001) Common European Framework of Reference for Languages: Learning, Teaching, Assessment. Cambridge: Cambridge University Press. Craik, F.I.M. and Watkins, M.J. (1973) The role of rehearsal in short-term memory. Journal of Verbal Learning and Verbal Behavior 12 (6), 599–607. Crano, W.D., Kenny, D.A. and Campbell, D.T. (1972) Does intelligence cause achievement: A cross-lagged panel analysis. Journal of Educational Psychology 63 (3), 258–275. Crano, W.D. and Mellon, P.M. (1978) Causal influence of teachers’ expectations on children’s academic performance: A cross-lagged panel analysis. Journal of Educational Psychology 70 (1), 39–49. Edwards, J., Beckman, M.E. and Munson, B. (2004) The interaction between vocabulary size and phonotactic probability effects on children’s production accuracy and fluency in nonword repetition. Journal of Speech, Language and Hearing Research 47 (2), 421–436. French, L.M. (2003) Phonological Working Memory and L2 Acquisition: A Developmental Study of Quebec Francophone Children Learning English. PhD thesis, Université Laval. French, L.M. (2004) Phonological working memory and L2 acquisition: A developmental study of Quebec Francophone children learning English. Dissertation Abstracts International, A: The Humanities and Social Sciences 65 (2), 487. Gathercole, S.E. (2006a) Nonword repetition and word learning: The nature of the relationship. Applied Psycholinguistics 27 (4), 513–543. Gathercole, S.E. (2006b) Complexities and constraints in nonword repetition and word learning. Applied Psycholinguistics 27 (4), 599–613.

130

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Gathercole, S.E., Frankish, C.R., Pickering, S.J. and Peaker, S.M. (1999) Phonotactic influences on short-term memory. Journal of Experimental Psychology: Learning, Memory, and Cognition 25 (1), 84–95. Gathercole, S.E., Hitch, G.J., Service, E. and Martin, A.J. (1997) Phonological short-term memory and new word learning in children. Developmental Psychology 33 (6), 966–979. Gathercole, S.E. and Pickering, S.J. (1999) Estimating the capacity of the phonological short-term store. International Journal of Psychology 34 (5 and 6), 378–382. Gathercole, S.E., Pickering, S.J., Hall, M. and Peaker, S.M. (2001) Dissociable lexical and phonological influences on serial recognition and serial recall. Quarterly Journal of Experimental Psychology 54A (1), 1–30. Gathercole, S.E., Service, E., Hitch, G.J., Adams, A.-M. and Martin, A.J. (1999) Phonological short-term memory and vocabulary development: Further evidence on the nature of the relationship. Applied Cognitive Psychology 13 (1), 65–77. Gathercole, S.E., Willis, C.S., Emslie, H. and Baddeley, A.D. (1992) Phonological memory and vocabulary development during the early school years: A longitudinal study. Developmental Psychology 28 (5), 887–898. Gregg, V.H., Freedman, C.M. and Smith, D.K. (1989) Word frequency, articulatory suppression and memory span. British Journal of Psychology 80 (3), 363–374. Gupta, P. (2003) Examining the relationship between word learning, nonword repetition, and immediate serial recall in adults. The Quarterly Journal of Experimental Psychology 56A (7), 1213–1236. Gupta, P. and Tisdale, J. (2009) Does phonological short-term memory causally determine vocabulary learning? Toward a computational resolution of the debate. Journal of Memory and Language 61, 481–502. Hulme, C., Maughan, S., and Brown, G.D.A. (1991) Memory for familiar and unfamiliar words: Evidence for a long-term memory contribution to short-term memory span. Journal of Memory and Language 30 (6), 685–701. Kenny, D.A. (1975) Cross-lagged panel correlation: A test for spuriousness. Psychological Bulletin 82 (6), 887–903. Kurcz, I., Lewicki, A., Sambor, J., Szafran, K. and Woronczak, J. (1990) Słownik frekwencyjny polszczyzny współczesnej. Kraków: Instytut Je˛zyka Polskiego PAN. Laufer, B. and Nation, P. (1999) A vocabulary-size test of controlled productive ability. Language Testing 16 (1), 33–51. Martin, K.I. (2009) Phonological and Working Memory and L2 Grammar Learning. BA thesis, University of Michigan. Miyake, A. and Shah, P. (1999) Models of Working Memory: Mechanisms of Active Maintenance and Executive Control. New York: Cambridge University Press. Nation, I.S.P. (1990) Teaching and Learning Vocabulary. New York: Newbury House. Nation, I.S.P. (2001) Learning Vocabulary in Another Language. Cambridge: Cambridge University Press. O’Brien, I.S., Segalowitz, N., Collentine, J. and Freed, B. (2006) Phonological memory and lexical, narrative, and grammatical skills in second language oral production by adult learners. Applied Psycholinguistics 27, 377–402. OUP. (2001) Paper and Pen Test: User Manual. Oxford: Oxford University Press. Papagno, C., Valentine, T. and Baddeley, A.D. (1991) Phonological short-term memory and foreign-language vocabulary learning. Journal of Memory and Language 30 (3), 331–347. Papagno, C. and Vallar, G. (1992) Phonological short-term memory and the learning of novel words: The effect of phonological similarity and item length. The Quarterly Journal of Experimental Psychology 44A (1), 47–67.

Phonological Short-term Memory and L2 Lexical Knowledge

131

Papagno, C. and Vallar, G. (1995) Verbal short-term memory and vocabulary learning in polyglots. The Quarterly Journal of Experimental Psychology, Section A: Human Experimental Psychology 48A (1), 98–107. Qian, D.D. (2002) Investigating the relationship between vocabulary knowledge and academic reading performance: An assessment perspective. Language Learning 52 (3), 513–536. Schmitt, N. (2000) Vocabulary in Language Teaching. Cambridge: Cambridge University Press. Schmitt, N., Schmitt, D. and Clapham, C. (2001) Developing and exploring the behaviour of two new versions of the Vocabulary Levels Test. Language Testing 18 (1), 55–88. Schweickert, R. (1993) A multinomial processing tree model for degradation and redintegration in immediate recall. Memory and Cognition 21 (2), 168–175. Service, E. (1992) Phonology, working memory and foreign-language learning. The Quarterly Journal of Experimental Psychology 45A (1), 21–50. Skrzypek, A. (2010) The Relationship between Phonological Short-term Memory and the Development of L2 Vocabulary and Collocational Knowledge: A Study of adult Polish Learners of English at the A2 and B1 Levels of English Proficiency. PhD thesis, Trinity College Dublin. Skrzypek, A., Kopečková, R., Bidzin´ska, B. and Singleton, D. (in press) Language and culture: Attitudes towards and perceptions of English L2 acquisition among adult Polish migrants in Ireland. In B. Geraghty and J. E. Conacher (eds) Cultures in Contact: Contemporary Issues in Language Use and Language Learning. London: Continuum. Snowling, M., Chiat, S. and Hulme, C. (1991) Words, nonwords and phonological processes: Some comments on Gathercole, Willis, Emslie and Baddeley. Applied Psycholinguistics 12 (3), 369–373. Speciale, G., Ellis, N.C. and Bywater, T. (2004) Phonological sequence learning and shortterm store capacity determine second language vocabulary acquisition. Applied Psycholinguistics 25 (2), 293–321. Thorn, A.S.C., Gathercole, S.E. and Frankish, C.R. (2002) Language familiarity effects in short-term memory: The role of output delay and long-term knowledge. The Quarterly Journal of Experimental Psychology 55A, 1363–1383.

Appendix Table A.1 Nonwords used in PSTM tasks PSTM tasks

Nonwords

Serial nonword recall

biznale, bywlera, ceptyko, dokrapu, fezbino, foksela, gaflosy, gibruta, jestabyjukloty, kosmuca, lagroki, letrumi, mudrygo, neglika, raplido, roblewy, seprody, woskane, wycmosa bafroce, bostagi, ceplira, dagryko. fimrosy. gadlipu. gubrawi, jedmuda, jubleny. kacmoba. Iizmato. meksona, nazdoty, pegwika, ruspale, rysnudo, sopryla, toskaby, wekluga, zudrami

Serial nonword recognition

132

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Table A.2 PSTM scores in the A2 and B1 groups Time 1 Statistics

A2 Serial Recall

B1 Serial Recall

A2 Serial Recognition

B1 Serial Recognition

k N mean min/max SD Shapiro-Wilk test Skewness Kurtosis

45 30 15.6 6/25 4.944 D(30)= 0.979, ns* −0.087, ns −0.644, ns

45 30 17.5 6/30 6.257 D(30) = 0.977, ns −0.038, ns −0.584, ns

30 30 16.4 9/24 3.839 D(30) = 0.979, ns 0.036, ns −0.694, ns

30 30 18.1 10/25 3.912 D(30)= 0.974, ns −0.235, ns −0.644, ns

Statistics

A2 Serial Recall

B1 Serial Recall

A2 Serial Recognition

B1 Serial Recognition

k N mean min/max

45 30 15.5 6/24

45 30 17.5 7/30

30 30 16 8/25

30 30 18 9/28

Statistics

A2 Serial Recall

B1 Serial Recall

A2 Serial Recognition

B1 Serial Recognition

SD Shapiro-Wilk test Skewness Kurtosis

4.883 D(30)= 0.971, ns −0.175, ns −0.883, ns

6.246 D(30) = 0.959, ns −0.141, ns −0.781. ns

4.135 D(30) = 0.981, ns 0.232, ns −0.216, ns

4.513 D(30)= 0.985, ns 0.142, ns −0.075. ns

Time 2

Time 2

* ns – non-significant (p > .05)

Phonological Short-term Memory and L2 Lexical Knowledge

133

Table A.3 Vocabulary and exposure scores in the A2 and B1 groups A2 group Statistics

1 K–2 K vocabulary (Time 1)

1K–2K vocabulary (Time 2)

Exposure scores (between Time 1 & 2)

N Mean Min/max SD Shapiro-Wilk test Skewness Kurtosis Levene’s test

30 49.2 % 30.4/70.3% 12.881% D(30)= 0.930, ns* 0.158.ns −1.337, ns F(1,28) = 0.809, ns

30 56.1 % 31.7/80.9% 13.296% D(30) = 0.971, ns 0.025, ns −0.966. ns F(1,28)= 0.161, ns

30 4.4 (h/d)* 1/8 (h/d) 2.235 (h/d) D(30)= 0.932, ns 0.059, ns −1.169 ns F(1,28) = 0.018, ns

2 K–3 K vocabulary (Time 1) N 30 Mean 52.1 % Min/max 20.3/76.7% SD 16.432% Shapiro-Wilt test D(30) = 0.949, ns Skewness −0.660. ns Kurtosis −0.111. ns Levene’s test F(l,28) = 3.645, ns

2 K–3 K vocabulary (Time 2) 30 64.5 % 31.3/91.9% 15.164% D(30)= 0.95 9, ns −0.484. ns −0.344. ns F(1,28) = 1.956,ns

Exposure scores (between Time 1 & 2) 30 5.3 (h/d) 2/9 (h/d) 2.204 (h/d) D(30)= 0.940, ns 0.098. ns −1.095,ns F(1,28) = 0.108, ns

B1 group Statistics

* ns – non-significant (p > .05) h/d – hours per day

Table A.4 Reliability coefficients of the research instruments Research instruments

Serial nonword recall (A2) Serial nonword recall (B1) Serial nonword recognition (A2) Serial nonword recognition (B1) 1K–2K Vocabulary Levels Test(A2) 2K–3K Vocabulary Levels Test(B1)

Spearman-Brown split-half reliability coefficient Time 1

Time 2

0.918 0.897 0.831 0.704 0.876 0.951

0.905 0.923 0.812 0.792 0.845 0.912

7 Segmental Acquisition in Polish Child and Adult Learners in Ireland Romana Kopecˇková

Introduction Following Poland’s accession to the European Union in May 2004, large numbers of Polish migrants arrived in Ireland in search of better career prospects and greater opportunities to improve their skills in speaking the language that is arguably the lingua franca of the globalizing world. In numerous cases these migrants brought with them their school-aged children, who also experienced a major shift in their life circumstances. One of the greatest challenges faced by Polish migrants on arrival in Ireland may have been the host community language. Irish English is characterized by a degree of distinctiveness in vocabulary, construction, and pronunciation; particularly in terms of certain sounds that are peculiar to it, this variety of English can pose a great challenge for the unaccustomed ear. It was thus of interest in the present study to examine how Polish children and adults have come to grips with specific English vowel sounds after about three years of stay in Ireland. This was considered an issue worth examining for at least two reasons. First of all, the Polish learners of English in this study were acquiring the target language in a context in which opportunities for use of their native language (henceforth L1) were plentiful, as was exposure to the target language spoken by other migrant learners. Although previous research has been rather inconclusive about the role of second language (henceforth L2) experience in L2 phonological acquisition (see e.g. Piske et al., 2001, for discussion), infrequent exposure to native-speaker 134

Segment al Acquisit ion in Polish Child and Adult Lear ners in Irel and

135

input has been claimed to significantly affect the accuracy with which L2 speech is produced (Flege, 2008; Moyer, 2008), as has continuous and frequent use of the native language (Guion et al., 2000; Piske et al., 2002). Second, to include child and adult learners in such an investigation could provide greater insights into issues surrounding the age-related differences in L2 speech learning. It is still uncertain, for example, to what extent differences in L2 experience and/or in L1 sound system development at the time when L2 learning begins contribute to age effects in L2 phonological acquisition (see Flege & MacKay, 2011, for discussion). The present study, therefore, seeks to examine whether the specific migrant experience of Polish children and adults in Ireland – characteristic of a fairly balanced exposure to and use of both their languages in their everyday life in the L2-speaking country – can represent a learning constellation advancing our understanding of agerelated effects in L2 phonological attainment.

Age and Experience Effects in L2 Segmental Production The effect of age on learning an L2 in immersion settings has been demonstrated in many studies to date (see Singleton & Ryan, 2004 for a review). In the realm of L2 speech production, the early age benefit has been noted for the production of L2 consonants (e.g. MacKay et al., 2001), the production of L2 vowels (e.g. Flege et al., 1999; Piske et al., 2002), as well as for the perception of the overall degree of foreign accent (e.g. Aoyama et al., 2003; Flege et al., 1999). This evidence has mostly been based on studies which used a retrospective research design, where adult L2 learners – differing in their age of arrival in the L2-speaking country – were compared in terms of their eventual L2 production abilities after many years of stay in the L2-speaking country. For example, Flege et al. (1999) examined intelligibility scores obtained for ten English vowel sounds produced by Italian speakers living in Canada for most of their lives, and found that those learners who arrived in the country at the age of seven performed with native-like ability, whereas those who arrived in Canada as young adults differed in their performance from native speakers significantly. Similar findings were also reported in studies which directly compared children to adults as they were learning their L2 (Baker et al., 2002; Oh et al., 2011; Tsukada et al., 2005). In a study by Baker et al. (2002), Korean children outperformed Korean adults in their productions of several L2 vowels as early as after one year of residence in the US, and attained native-like accuracy in producing all the tested vowels after about nine years of life in the country. The adult learners, in contrast, were

136 Linguist ic and Cultural Acquisit ion in a Migrant Communit y

found to accurately produce only those L2 vowel sounds which involved close L2–L1 counterparts. In another study, Tsukada et al. (2005) compared the acquisition of English vowel sounds by native Korean children and adults, and showed that the young learners were more native-like in the production of all the tested vowels than were the adult learners after both three and five years of residence in North America. Although early starters thus appear to be more successful L2 speech learners in global terms, it is not necessarily the case that they will produce the full range of L2 sounds exactly like native speakers. In fact, studies which closely scrutinized L2 segmental acquisition in learners of diverse age groups have pointed out that there are significant differences in what might be called the ‘learnability’ of individual L2 segments (Baker et al., 2002, 2008; Munro & Derwing, 2008; Piske et al., 2002) . For example, Piske et al. (2002) examined the production of 11 English vowels by early Italian–English bilinguals, who were further divided into two subgroups according to the amount of L1 use (low vs. high) in their everyday life in Canada. In comparison to native speakers, the ‘early low’ bilinguals were found to be indistinguishable in their performance; however, the ‘early high’ groups failed to produce five of the 11 tested vowels in a native-like manner, all of which had no phonetic counterparts in the Italian language. The authors speculated that the L1 use effects on the segmental acquisition of the ‘early high’ group may have been related to the learners’ presumably less frequent L2 use, inversely linked to their more frequent L1 use, as well as to an increased influence of their L1 sound system on forming new phonetic representations for L2 sounds. This interpretation is broadly in line with the predictions of the Speech Learning Model (SLM) (Flege, 1995, 2002, 2007), emphasizing the interactive nature of L2 phonological acquisition. In SLM, children are predicted to be better L2 speech learners thanks to the state of development of their L1 sound system at the time of L2 learning. Their developing L1 sound system is expected to have a weaker influence on the formation of the L2 sound system than it is the case with an established L1 sound system of adults. Children are thus predicted to be less likely to perceptually associate L1 and L2 sounds, and as such better able to discern cross-language phonetic differences between the sounds. Consequently, they are more likely than adults to perceive and produce L2 speech accurately (Baker et al., 2002, 2008; Kopečková, 2011). Applied to the findings reported by Piske et al. (2002), one could argue that the ‘early low’ Italian–English bilinguals’ evolving L1 sound system exerted minimal influence on the formation of their L2 sound system, whereas the ‘early high’ Italian–English bilinguals’ L1 sound system was frequently activated, and as such influenced their acquisition of certain specific L2 vowels to a greater extent. Where the findings reported by Piske

Segment al Acquisit ion in Polish Child and Adult Lear ners in Irel and

137

et al. (2002) disagree with the SLM, however, is in respect to the question of learnability of specific L2 sounds. The SLM hypothesizes that those L2 sounds that are perceived as distinct sounds will be easy to acquire because phonetic differences between such sounds and their L1 counterparts can be easily detected; they may be difficult to produce initially, however (Baker et al., 2008; Oh et al., 2011). On the other hand, those L2 sounds that are perceived as similar to L1 sounds are predicted to be difficult to acquire in the model, as L2 learners might not detect the subtle phonetic differences between the L2 and L1 sounds concerned, and will therefore produce such L2 sounds with a more obvious foreign accent. Thus, this view proposes that the outcome of L2 speech learning depends on the accuracy of perceived phonetic similarity between L1 and L2 sounds, which will differ for child and adult L2 learners as a result of their L1 sound system development at the time of L2 acquisition. The present study, therefore, seeks to investigate whether the Polish learners’ ability to perceive differences between specific L1 and L2 sounds can be related to their L2 vowel acquisition. Another perspective on L2 segmental acquisition in learners of diverse age groups is that children are commonly exposed to L2 input that is qualitatively different from that enjoyed by adults, thus aiding their successful L2 production more. According to Jia and Aaronson (2003), younger learners often use the target language with their L2-speaking peers and interact with the host culture to a much greater extent than older learners. They are thus more likely to become dominant in their L2 and to attain native-like abilities in the language. It is also possible that owing to the more intensive exposure to the target language, children might also be more motivated to use their L2 frequently and accurately. Indeed, it is inevitable that experiential and affective factors (interwoven with age of L2 learning) work together in L2 phonological acquisition (Moyer, 2008). The purpose of the present chapter, therefore, is to evaluate these views of child–adult differences in L2 speech learning. It does so by examining the effect of cross-language phonetic similarity perception and L2/L1 use on the production of specific English vowels by Polish migrant children and adults after about three years of stay in Ireland.

Method Participants A group of 40 native Polish (NPI) participants was recruited from the Polish Diaspora Project during the summer of 2009, and divided into two

138

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

subgroups according to their age on arrival in Ireland: 20 children (between the ages of 7 and 12 years) and 20 adults (aged 21 years and older). The participants had all resided in Dublin for about three years at the time of the study (range: 1 to 5 years). Before coming to Ireland, they had all received some formal instruction in the English language in Poland; when tested formally using a standardized pen-and-paper placement test (Anglia Placement Test, 2009), the NPI participants were evaluated, on average, at an upperintermediate level in terms of their ability in English vocabulary, structure and reading comprehension. A control group of 19 native Polish (NPP) speakers (10 children and 9 adults) living in Poland and with no English immersion experience was also included in the project. The members of this group were mostly attending formal English language classes with non-native (Polish) teachers at the time of the study, and their proficiency in the language was self-reported as ranging from ‘beginner’ to ‘upper-intermediate’ levels. These participants were recruited in order to provide something of a linguistic resemblance to a group of Polish migrants as they might have been on the first day of their arrival in the host country. More importantly, the NPP child and adult participants were also included in this study to test the SLM prediction that children are more accurate cross-language perceivers than adults thanks to the state of development of their L1 sound system at the time of L2 speech learning. If this is the case, child L2 learners without prior L2 immersion experience will also demonstrate greater ability in accurately relating L2 sounds to L1 sounds, in comparison to adult L2 learners without L2 immersion experience. Finally, a control group of native English-speaking participants (NS) comprised 10 children and 10 adults chosen on the basis of their place of birth (Dublin) and of not speaking or learning Polish as an additional language. The NS groups were included to ensure that the Polish learners’ performance in English reflected their L2 abilities, rather than developmental differences. A summary of the background information for the six participant groups is provided in Table 7.1.

Stimulus material and procedure The stimuli chosen for the present study included the , , , , , , , and 1 vowel sounds, representing different degrees of learnability for native Polish speakers. For example, the vowel sounds ,  and  are phonetically very similar in both languages, although the back vowel  can also be realized frontally in the variety of English spoken in Ireland (Hickey, 2005). In contrast, the vowel  is a specific sound of Irish

Segment al Acquisit ion in Polish Child and Adult Lear ners in Irel and

139

Table 7.1 Participants in the study Group

Age

AOA

AOE

LOR

English proficiency

Polish adults in IRL (NPI) Polish children in IRL (NPI) Polish adults in PL (NPP) Polish children in PL (NPP) Irish adults (NS) Irish children (NS)

24–53 (29.7)

21–49 (26.7)

5–29 (14.0)

1–5 (3.2)

1–6 (4.15)

12–15 (13.2)

7–12 (9.9)

6–12 (8.4)

1–5 (3.4)

1–5 (3.5)

27–48 (33.1)

9–26 (14.8)

10–12 (11.1)

4–10 (6.2)

25–42 (29.5) 9–14 (11.4)

Note: Means for age, age of arrival (AOA), age of first exposure to English (AOE), length of residence (LOR), and English proficiency (where 1 = elementary level, 2 = beginner, 3 = intermediate, 4 = upper-intermediate, 5 = advanced, and 6 = native-like) are provided in brackets.

English, which is typically produced as a mid, centralized, back, somewhat rounded vowel, and as such stretching between Polish ,  and . In some Dublin accents, this back vowel approximates to the production of  (Hughes et al., 2005). The English front vowel  is close in articulatory vowel space to Polish , although Polish speakers are known for producing the vowel sound closer to English . The vowel  does not occur in Polish and is realized between the Polish vowels  and . Finally, the English vowel  is somewhat similar to Polish , whereas the monophthongized Irish English  is quite different from the Polish  in that it is pronounced higher and more frontally in the oral cavity. It is notable that vowel duration does not distinguish vowels in the Polish language. The tested vowels were placed in a bVt word context, which is frequent and productive to a comparable extent in both Polish and English. In total, eight stimulus words (beat, bit, bet, bat, bought, boat, boot and but) were produced by a female native speaker of English, who was born and lived in Dublin all her life. Another three native speakers of English, all of whom were linguists from Trinity College Dublin, identified the bVt words and rated the goodness of the vowel production in the words on a 7-point Lickert

140

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

scale. Only those instances of the English vowel productions that were uniformly identified and judged to be excellent exemplars were selected and used in the stimulus material presented to the study participants. The data from the adult participants in the study were collected in a quiet language laboratory room at Trinity College Dublin, whereas the young participants were visited in their school or home environment, where the tasks were administered in a quiet room. The participants’ productions were recorded as digitized sound files (22.05 kHz, 16-bit resolution), using a professional digital recorder, ZOOM Handy Recorder H4.

Data Collection Production task The Polish learners’ production of the eight vowel sounds were tested in a delayed repetition task. In this production task, the participants were asked to listen to a stimulus sentence ‘I say for you’ in English, and then to repeat the sentence as if in a reaction form: ‘And now I say for you’. The sentences were elicited twice from the participants, in order to ensure that productions with errors caused by the testing situation could be taken out of the analysis. If no such removal of erroneous productions was needed, the second production – assumed to be the more naturally and comfortably produced – was chosen for use in a subsequent intelligibility task. To facilitate the presentation of the participants’ productions for judgement, the eight CVC tokens from each participant were digitally extracted, using a waveform editing programme.

Intelligibility judgement task In this study, an intelligibility judgement task was used to determine whether the Polish learners were able to produce the target vowel sounds accurately enough for a native speaker to identify it. That is, if a speaker’s production of the vowel in ‘bat’ sounded more like  than any other English vowel, it was deemed intelligible, even if it was not a perfect exemplar of the vowel . A group of seven native-speaker listeners participated in this task; they all came from Ireland, and comprised both of those who had some background in linguistic research and those of a nonlinguistic background. None of them had ever learnt Polish as an additional language. In a self-paced online presentation,2 the listeners were asked to listen to a single word (632 in total: 79 participants × 8 productions) and to identify which vowel sound they had just heard by selecting one of the eight tested vowels from a drop-down menu. The vowels were presented in the IPA

Segment al Acquisit ion in Polish Child and Adult Lear ners in Irel and

141

format, followed by a lexical example to assist those participants who might not have been familiar with phonetic transcriptions of words. The listeners could also select the response ‘other’ for tokens that could not be placed in any of the target English vowel categories. The results of the intelligibility task were calculated in terms of the correct target identifications by the seven listeners.

Perceptual assimilation task In order to determine whether children and adults perceive the phonetic similarity between L1 and L2 vowel sounds differently, the NPI and NPP participants were asked to identify each of the eight stimulus words in terms of one of the following six Polish alternatives presented on a computer screen in their orthographic form (only): bity , byty , buty , bety , baty, and boty . The eight target words were presented in one repetition for the participants. First, they listened to the English word and matched it to one of the six Polish keywords to which they believed it was most similar. Second, they made goodness-of-fit judgements regarding the similarity of the English vowel they had just heard and the Polish vowel they had chosen, using a 7-point Lickert scale, ranging from ‘1’, indicating that the two sounds were not at all alike, to ‘7’, indicating that the sounds were a complete match. Participants were encouraged to use the entire scale and to follow their first impression in completing the task. To lend validity to this task, the participants were asked to read out the Polish keywords from the screen, and to concentrate on how similar they felt the ‘middle sounds’ in the presented words were. They could listen to the English words and pronounce the Polish words several times, if they desired, but they were not allowed to change their answers once they were given, in order to maintain the impressionistic element in the task.

Background questionnaire An extensive background questionnaire was also used in this study. It was developed in two comparable versions in order to reflect the language style of children and adults, as well as the realities of their respective learning environments. Specifically, the Polish participants were asked to report on their learning experience with English and on their use of Polish; for example, they reported on the frequency of use of these two languages in different contexts in their everyday life in Ireland. They also responded to questions tapping into their motivation for and attitudes towards acquiring English, and on aspects of life in Ireland relating to use of the language, such as their perceptions of the difficulty of understanding Irish English, the importance of sounding native-like in English, and levels of happiness about living in

142 Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Table 7.2 Reported L2/L1 contact by Polish children and adults in Ireland Groups

L2 use

L1 use

L2 use with native speakers

L2 use with non-native speakers

NPI adults NPI children

2.77 (0.437) 2.42 (0.467)

2.45 (0.510) 2.66 (0.699)

1.75 (0.786) 1.42 (0.607)

2.10 (0.852) 2.42 (1.170)

Note: Standard deviations for mean values are provided in brackets; 1 = ‘very often’, 2 = ‘often’, 3 = ‘sometimes’, 4 = ‘rarely’, 5 = ‘never’.

Ireland. Table 7.2 provides a summary of the questionnaire data that are pertinent to this study.

Results Intelligibility of L2 vowel productions As some vowel productions might have been ambiguous and as such judged inconsistently, a complete agreement among the listeners was not expected. Of the 632 vowel tokens presented, all seven listeners agreed on 228 (36%), and another 151 tokens (24%) were uniformly identified by six listeners. The agreement rate varied across vowels and speakers, ranging from 100% for  produced by NS adults, to 5% for  spoken by NPI adults, as rated by at least six listeners. This suggests that disagreement among the listeners was caused by ambiguity in the productions, rather than being attributable to listener factors. As explained above, the results of the delayed repetition task were determined by calculating the percentage of correct productions in terms of the intelligibility of each participant’s production of the eight English vowels, as judged by seven native speakers of English. As initial tests of homogeneity of variances on the individual vowel productions revealed that similar variances could be assumed only for ,  and  vowels, non-parametric tests of analysis were applied, whereby individual vowel productions were evaluated. Overall L2 production intelligibility scores, however, did satisfy the assumption of equal variances ( p = 0.085), and therefore were analysed using parametric ANOVA tests. To compare overall L2 production accuracy scores between the participant groups, a one-way ANOVA was performed on the data, yielding a significant effect of group [F(5,73) = 14.526, p = 0.000]. A post-hoc

Segment al Acquisit ion in Polish Child and Adult Lear ners in Irel and

143

Bonferroni test revealed that NPI adults differed significantly from NS adults ( p = 0.000), but not from NPP adults ( p = 0.475) in their English vowel productions. In contrast, NPI children did not differ from agematched NS children ( p = 0.297), but produced the tested L2 vowels significantly more accurately than did NPP children ( p = 0.035). Comparisons between NPI children’s and NPI adults’ performance did not yield significant differences, however ( p = 1.000). These results are graphically displayed in Figure 7.1. To determine which of the tested vowel sounds were particularly challenging to produce for the participant groups, a series of Mann-Whitney U-tests was performed. The results, which are summarized in Table 7.3, indicate that although NPI children and NPI adults did not differ in their production accuracy of the tested segments overall, they did significantly differ in their ability to produce two of the English vowels,  and . These specific L2 vowels were produced by NPI children to native-like levels, whereas NPI adults managed to perform accurately only in the production of those vowels that are similar in respect to the L1 counterparts, i.e. , , and  vowel sounds. Furthermore, NPI children performed significantly more accurately than did NPP children overall, and specifically in the production of the vowels ,  and . In fact, L2 children in this study with three-year-long immersion

Figure 7.1 Mean production intelligibility scores by participant groups

144

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Table 7.3 Production intelligibility of individual L2 vowels by selected participant groups Groups: comparisons

L2 vowel

NPI children > NPI adults NPI children < NS children NPI children > NPP children

NPI adults < NS adults

NPI adults > NPP adults

U-test

p level



97.000

0.005



105.000

0.009



54.500

0.032



55.500

0.046



33.000

0.003



36.500

0.005



32.000

0.002



46.500

0.017



51.000

0.021



29.000

0.001



6.000

0.000



37.500

0.003



45.000

0.032



40.000

0.016

experience did not differ significantly from native speaking children in terms of their overall L2 vowel production; it was their  and  productions only that distinguished them from NS children. NPI adults, for their part, had apparently improved only those productions that were not related to ‘distant’ L2 sounds, i.e. English vowels  and . Although the former L2 vowel  was produced at levels about halfway between the accuracy of the L2 learner without naturalistic L2 experience and that of a native speaker, the vowel  was produced in a native-like manner by the NPI adults. The English vowel  did not pose any challenge for the Polish speakers to pronounce, as this is an L2 segment which is almost identical to their L1 vowel , and thus apparently showed a ceiling effect.

Cross-language phonetic similarity perception In order to determine whether there is a relationship between the Polish learners’ L2 production abilities and their perception of cross-language phonetic similarity, a perceptual assimilation task was administered to the participants. Analyses of this task revealed that the majority of participants heard English  as Polish , English  as Polish , English  as Polish , English  as Polish , English  as Polish , English  as Polish , and

Segment al Acquisit ion in Polish Child and Adult Lear ners in Irel and

145

English  as Polish . It is noteworthy that the English vowel  was also perceived as very similar to Polish  by a third of the NPI children. The only instance in which the participants’ groups differed was in respect to their classification of the Irish English vowel . Whereas the majority of NPI adults perceived this sound to be similar to the Polish vowel , NPI children classified it mostly as Polish . A reverse pattern was found for NPP adults and NPP children. Despite the similarities in overall patterns of cross-language identification between the Polish children and adults, the NPI children tended to rate the similarity between the tested vowels lower than did the NPI adults. In contrast, the NPP children rated the similarity between L1 and L2 sounds higher than did the NPP adults. A one-way ANOVA test yielded a significant group effect for the overall fit index scores [F(3,55) = 3.160, p < 0.05]. When Tukey’s post-hoc test was applied, however, it was found that NPI children were not significantly less likely than NPI adults to perceptually relate each of the tested vowel sounds ( p > 0.10). Nor did NPP children significantly differ from NPP adults in their evaluations of the perceptual match between the eight English vowels and their Polish primary response alternatives ( p > 0.10). Interestingly, the NPP children performed in this task comparably to the NPI children, suggesting that L2 children’s perception of cross-linguistic similarity might show good accuracy levels regardless of L2 learning experience. The NPP adults, in turn, differed significantly from the NPI adults in giving lower similarity ratings to most of the tested vowels, showing a kind of psychological bias towards perception of similarities between Polish and English vowel sounds.

L2 experience data As shown in Table 7.2, NPI child and adult migrants did not significantly differ in terms of their reported contact and use of English and Polish in their everyday life in Ireland. What is noteworthy though is that the NPI children enjoyed a qualitatively different type of contact with their L2 than NPI adults did, mainly thanks to their friendships with both native and non-native speakers of English. The NPI children reported significantly more use of English with their friends than NPI adults did (t = 2.24, df = 38, p = 0.031), and, overall, frequent use of English with friends was associated in the sample (N = 40) with greater use of English with native speakers (r = 0.745, p < 0.01) and nonnative speakers (r = 0.335, p < 0.05); with more use of English at work or at school (r = 0.460, p < 0.001); and with less use of Polish ( r = −0.345, p < 0.05). For example, when reporting on his use of English in Ireland, a 12-yearold Igor indicated that he always used English in his free time because he

146 Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Table 7.4 Reported balance in L2/L1 use by Polish children and adults in Ireland Groups

Balance in L2/L1 use in Ireland

Much more Polish More Polish Half and half More English Much more English Total

NPI adults

NPI children

1 4 9 2 4 20

0 4 12 4 0 20

liked to spend more time with friends who were English-speaking, and because he aimed to be able to speak the language really well for his job ambitions as an IT specialist. Similarly, a thirteen-year-old Paulina indicated that all her friends were English-speaking and therefore she liked using the language very much. In contrast, the adult L2 learners reported either more Polish-speaking friendships, or a balance between a contact with native and non-native friends, the former being mostly colleagues from work. A 28-yearold Izabela, for example, reported that she felt very happy living in Ireland, as she had met a lot of new Polish friends in the country and appreciated her ‘Polish infrastructure’ in Dublin very much; her only contact with native speakers was apparently at work when communicating with her boss. What this brief data thus seems to suggest is that NPI child migrants of this study were involved in a variety of L2 contact domains, and that their investment – both linguistic and affective – into English language learning was of a different nature than that of NPI adult migrants. Although quantitatively comparable and balanced in everyday L1 use (see Table 7.4) for both the groups, the children’s use of L2 was placed in both formal and informal contexts, ranging from attending target language schools to meeting Englishspeaking friends both in their school and leisure time.

Discussion In the current study, segmental acquisition of English vowels in Polish migrants of diverse age groups, with a medium-length of stay in Ireland, was examined. The results showed that the child L2 learners produced the tested vowels as accurately as age-matched native-speaking children, including vowels that are perceptually different from any of their L1 vowels. They

Segment al Acquisit ion in Polish Child and Adult Lear ners in Irel and

147

were also significantly more intelligible in their productions than child L2 learners without any immersion experience. Adult L2 learners, for their part, however, did not produce the tested vowels significantly differently from adult L2 learners without immersion experience, and progressed only in the production of those L2 sounds that are similar to their L1 sounds. The migrant children tended to rate the similarity between specific L2 and L1 sounds lower than the migrant adults did, who, in turn, were more accurate in the evaluations of cross-language phonetic similarity than adults without immersion experience. Both the migrant children and children in Poland distinguished between L1 and L2 sounds with greater accuracy than their adult counterparts, indicating that children may be superior crosslanguage perceivers. Further analyses suggested that migrant children’s experience with learning the L2 differed from those of migrant adults, as the migrant children frequently used the language with their L2-speaking friends. The results of this study agree with previous L2 segmental acquisition research, which directly compared children and adults as they were learning an L2 in typical immersion settings (Baker et al., 2002, 2008; Tsukada et al., 2005). These studies suggested that children’s superior acquisition of L2 segments is related to their greater perceptual abilities in terms of crosslanguage phonetic similarity, which, in turn, is linked to their incomplete state of development of L1 sound system. Indeed, child L2 learners in this study were shown to perceive discrepancies between diverse L1 and L2 vowels with greater accuracy, and tended to produce such L2 vowels more accurately than adult L2 learners; this held especially true for the new L2 vowel sounds  and . These vowel sounds were mapped by the child L2 learners differently, and were produced significantly more accurately than by the adult L2 learners. However, as was also shown in this study, crosslanguage perception does develop in adult learners as a result of naturalistic experience with the L2. After about three years of stay in Ireland, the NPI adults were found to be more accurate in their perception of similarities and differences between L1 and L2 vowels than were the NPP adults. This is, in a broader sense, in agreement with the fundamental claim of the SLM that the learning capacities for L2 acquisition remain available across the life span (Flege, 1995). One possible critical perspective on the inference that children are less likely than adults to perceptually assimilate L2 sounds to L1 categories would be that children may simply be less able than adults to perform crosslanguage perceptual tasks. Perhaps the use of other methods of measuring perceptual relationships between L1 and L2 sounds might have been more revealing and appropriate in this study. For example, Cebrian et al. (2011)

148

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

proposed combining rated discrimination tasks together with perceptual assimilation tasks in reliable assessments of cross-linguistic perception. Baker et al. (2008) suggested that more fine-grained rating scales be used to make it possible to detect significant differences between children’s and adults’ perceptual goodness ratings. Finally, Strange (2007) advocated the use of an ordinal rather than interval scale of quantification in Lickert-scale judgements of similarity in this type of research. Another possibility, not incompatible with the account presented above, is that the NPI children may have received richer L2 input and sought out such input more, as suggested by the results of the questionnaire data, which may have equipped them better for achieving accurate L2 production. As Moyer (2008) argued, early exposure to an L2 is commonly related to a compound of interacting psychological, social and cognitive factors. Using English with their peers and being enrolled in the target language schools, the children of this study may have been more motivated towards the L2 and exposed to more phonological correction and feedback, leading to more advanced L2 production skills, and presumably to a greater sense of attainment, keeping their motivation in relation to frequent L2 use and contact high. Also, as reported by the NPI children, their parents seemed to have extremely high expectations with regard to their future and development of their English skills. A vast majority of the participating children reported an everyday encouragement on the part of their parents to speak English with their peers as much as possible. The Polish parents were reported to stress constantly the importance of English for their children’s future and for good job prospects. It is to be noted, however, that this very high level of motivation towards English language learning was observed in all the study participants, regardless of their age or L2 performance, which may not be surprising, considering their EU migrant realities. A further reason for the Polish children’s superior L2 speech production might have been that their richer L2 input was related to their experience with the L2 lexicon. Baker and Trofimovich (2008) demonstrated that adults exposed to the L2 in a naturalistic environment for one year were affected by lexical familiarity and frequency in their production of L2 vowels significantly more than were child L2 learners, suggesting that thanks to the greater and richer exposure to native-speaker input, children might surpass adults in their progress through L2 word learning. Where the findings of this study disagree with previous research is in the area of L1 use effects on L2 speech learning. The child migrant learners’ L1 use was not found to significantly affect their L2 production intelligibility. The NPI speakers reported frequent and extensive L1 contact in their everyday life in Ireland, including NPI children who enjoyed opportunities to use

Segment al Acquisit ion in Polish Child and Adult Lear ners in Irel and

149

their L1 at school as well as in a Polish language weekend school. Furthermore, frequent flights to Poland, summer holidays in their home country and communication via the Internet with their family members and friends back in Poland were commonplace. This learning constellation would traditionally be viewed as potentially hampering L2 phonetic learning; however, the results of this study suggest that child L2 learners might not be affected by their L1 use during their L2 speech learning to the extent reported in studies with early L2 learners (Flege et al., 1999; Guion et al., 2000; Piske et al., 2002). One reason for this discrepancy might be that the present study examined L2 phonological acquisition of children as they were learning their L2, rather than in retrospect. The children’s inevitably less extensive experience with the use and processing of the L1 sound system, compared with adults and even early L2 learners who continue to use their L1 often, may have led to a weaker interference with their L2 speech learning. A further possibility in this respect might be that at least some children in this study became dominant in their L2, or felt equally comfortable using their L1 and L2, and therefore no L1 effects could be detected (Jia & Aaronson, 2003). Finally, the L1 use data (as well as L2 use data) were self-reported, rather than measured, and might have therefore been subject to error in the study (Flege, 2008).

Conclusion The main purpose of this study was to examine Polish children’s and adults’ productions of a range of English vowel sounds after a medium– long stay in Ireland, and to offer an account for the learners’ different learning outcomes. The results obtained provided evidence for the prediction that children are superior L2 speech learners, although not necessarily fully native-like, who seem to benefit from their ability to discern discrepancies between L2 and L1 sounds even in contexts where frequent exposure to an L1 is maintained. Furthermore, it has been demonstrated that children tend to enjoy and seek out a qualitatively different type of L2 input than adults do, which appears to further aid their successful L2 segmental acquisition. It is important that further research investigates this combined effect on L2 phonological acquisition in learners of diverse age groups by expanding on other phonetic and phonotactic contexts, segments, and language combinations. In this endeavour, developing yet other measures of cross-language phonetic similarity perception and collecting measured, rather than self-reported, L2/L1 use data would be especially worthwhile, if we are to gain a good understanding of child–adult differences in the acquisition of L2 speech.

150 Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Notes (1) To distinguish between the vowel  from mainstream English and its realization in Irish English, the vowel is sometimes transcribed by the use of the phonetic symbols õ or  õ in the context of the Irish English vowel inventory. (2) For presentation of the intelligibility task, QuestionPro Survey Software (Corporate Licence) was used. This online presentation software supports audio presentations in an mp3 format and a wide variety of design tools. In this study, the listeners could save their responses at any time they desired and continue later. On average, it took them 80 minutes to complete the whole task. Two blocks of presentations were designed, which were integrated into two separate surveys, or rather, links; however, these were chained and thus it was possible for the listeners to continue in one go if they so wished. However, the participants were encouraged, also through the design of the task, to take breaks during the task to reduce fatigue. The division of the task into two parts (comprising separate children’s and adults’ audio files) was also motivated by the requirement to administer the task in two different orders.

References Anglia Placement Test (2009) Anglia Examination Syndicate, Chichester University, accessed 15 January 2009. http://www.anglia.org Aoyama, K., Flege, J.E., Guion, S., Akahane-Yamada, R. and Yamada, T. (2003) Foreign accent in English words produced by Japanese children and adults. In M. Solé, D. Recasens and J. Romero (eds) Proceedings of 15th International Congress of Phonetics Sciences, pp. 3201–3204. Barcelona: Casual Productions. Baker, W. and Trofimovich, P. (2008) Lexical and segmental influences on child and adult learners’ production of second language vowels. Concordia Working Papers in Applied Linguistics (1), 30–54. Baker, W., Trofimovich, P., Flege, J.E., Mack, M. and Halter, R. (2008) Child–adult differences in second language phonetic learning: The role of cross-language similarity. Language and Speech 51 (4), 317–342. Baker, W., Trofimovich, P., Mack, M. and Flege, J.E. (2002) The effect of perceived phonetic similarity on non-native sound learning by children and adults. Proceedings of the Annual Boston University Conference on Language Development 26 (1), 36–47. Cebrian J., Mora, J.C. and Aliaga-Garcia, C. (2011) Assessing crosslinguistic similarity by means of rated dissimilarity and perceptual assimilation tasks. In M. Wrembel, M. Kul and K. Dziubalska-Kołaczyk (eds) Achievements and Perspectives in the Acquisition of Second Language Speech: New Sounds 2010, Vol. 1, pp. 41–52. Bern: Peter Lang. Flege, J.E. (1995) Second-language speech learning: Theory, findings, and problems. In W. Strange (ed.) Speech Perception and Linguistic Experience: Issues in Cross-language Research, pp. 229–273. Timonium, MD: York Press. Flege, J.E. (2002) Interactions between the native and second-language phonetic systems. In P. Burmeister, T. Piske and A. Rohde (eds) An Integrated View of Language Development: Papers in Honor of Henning Wode, pp. 217–244. Trier: Wissenschaftlicher Verlag. Flege, J.E. (2007). Language contact in bilingualism: Phonetic system interactions. In J. Cole and J.I. Hualde (eds) Laboratory Phonology 9, pp. 353–382. Berlin: Mouton de Gruiter. Flege, J.E. (2008) Give input a chance! In T. Piske and M. Young-Scholten (eds) Input Matters in SLA, pp. 175–190. Bristol: Multilingual Matters.

Segment al Acquisit ion in Polish Child and Adult Lear ners in Irel and

151

Flege, J.E., MacKay, D. and Meador, D. (1999) Native Italian speakers’ perception and production of English vowels. The Journal of the Acoustical Society of America 106, 2973–2987. Flege, J.E. and MacKay, I.R.A. (2011) What accounts for ‘age’ effects on overall degree of foreign accent? In M. Wrembel, M. Kul and K. Dziubalska-Kołaczyk (eds) Achievements and Perspectives in the Acquisition of Second Language Speech: New Sounds 2010, Vol. 2, pp. 65–82. Bern: Peter Lang. Flege, J.E., Yeni-Komshian, G.H. and Liu, S. (1999) Age constraints on second-language acquisition. Journal of Memory and Language 41 (1), 78–104. Guion, S.G., Flege, J.E. and Loftin, J. (2000) The effect of L1 use on pronunciation in Quichua-Spanish bilinguals. Journal of Phonetics 28 (1), 27–42. Hickey, R. (2005) Dublin English: Evolution and Change. Amsterdam; Philadelphia: J. Benjamins Publishing. Hughes, A., Trudgill, P. and Watt, D.J.L. (2005) English Accents and Dialects: An Introduction to Social and Regional Varieties of English in the British Isles, 4th edn. London: Hodder Arnold. Jia, G. and Aaronson, D. (2003) A longitudinal study of Chinese children and adolescents learning English in the United States. Applied Psycholinguistics 24, 131–161. Kopečková, R. (2011) Learning vowel sounds in a migrant setting: The case of Polish children and adults in Ireland. In M. Wrembel, M. Kul and K. Dziubalska-Kołaczyk (eds.) Achievements and Perspectives in the Acquisition of Second Language Speech: New Sounds 2010, Vol. 2, pp. 137–148. Bern: Peter Lang. MacKay, I.R.A., Flege, J.E., Piske, T. and Schirru, C. (2001) Category restructuring during second-language speech acquisition. The Journal of the Acoustical Society of America 110 (1), 516–528. Moyer, A. (2008) Input as a critical means to an end: Quantity and quality of experience in L2 phonological attainment. In T. Piske and M. Young-Scholten (eds) Input Matters in SLA (pp. 159–174). Bristol: Multilingual Matters. Munro, M.J. and Derwing, T.M. (2008) Segmental acquisition in adult ESL learners: A longitudinal study of vowel production. Language Learning 58 (3), 479–502. Oh,G.E., Guion-Anderson, S., Aoyama, K., Flege J.E., Akahane-Yamada, R. and Yamada, T. (2011) A one year longitudinal study of English and Japanese vowel production by Japanese adults and children in an English-speaking setting. Journal of Phonetics, doi:10.1016/j.wocn.2011.01.002 Piske, T., Flege, J.E., MacKay, D. and Meador, D. (2002) The production of English vowels by fluent early and late Italian-English bilinguals. Phonetica 59, 49–71. Piske, T., MacKay, I.R.A. and Flege, J.E. (2001) Factors affecting degree of foreign accent in an L2: A review. Journal of Phonetics 29 (2), 191–215. QuestionPro Survey, Corporate Licence [software], purchased and accessed 1 December 2009. http://www.questionpro.com Singleton, D.M. and Ryan, L. (2004) Language Acquisition: The Age Factor, 2nd edn. Clevedon: Multilingual Matters. Strange, W. (2007) Cross-language phonetic similarity of vowels: Theoretical and methodological issues. In O-S. Bohn and M. Munro (eds) Language Experience in Second Language Speech Learning: In Honor of James Emil Flege. Amsterdam: J. Benjamins Publishing. Tsukada, K., Birdsong, D., Bialystok, E., Mack, M., Sung, H. and Flege, J.E. (2005) A developmental study of English vowel production and perception by native Korean adults and children. Journal of Phonetics 33 (3), 263–290.

8 Variation in English Lexical Acquisition Among Polish Migrant Children in Ireland Polly Walsh and David Singleton

Introduction The number of Polish migrant children living in Ireland has markedly increased in recent years. This increase has significant implications, of course, for general education, but also for the speech and language therapy profession, which must adapt its practices to encompass individuals for whom English is an additional language. Differences between the vocabulary acquisition patterns of monolingual and bilingual children has been a matter of intense research and debate (cf. Bialystok, 1988; Doyle et al., 1978; Goodz, 1994; Patterson & Zurer-Pearson, 2004; Umbel et al., 1992). No previous comparative research has been carried out, however, on the lexical acquisition in English of Polish children in the process of becoming sequential PolishEnglish bilinguals in Ireland relative to that of their Irish age-peers, the topic of the present chapter. The aim of the present study was to investigate productive/expressive vocabulary development over a period of six months in Polish children acquiring English as an additional language. Our purpose was to explore how these children compared in this domain to their English-speaking Irish peers and the extent to which they differed among themselves, and also to attempt to identify some factors which may have influenced individual differences among the Poles in respect of lexical proficiency in English. 152

English Le x ical Acquisit ion Among Polish Migrant Children

153

The Importance of Vocabulary Acquisition in an Additional Language For children engaging with a new language environment, mastering the vocabulary of the language in question is one of the most important and challenging tasks facing them. Vocabulary proficiency is seen as crucial across the whole spectrum of comprehension and production demands in the additional language (cf. discussion in Singleton, 1999: 130f.), and, consequently, in the case of the school-going migrant child, as being a major determinant of the success or otherwise of a child’s passage into the educational environment of the host country. Van Hell and de Groot (2005: 10) maintain that vocabulary knowledge and access to and retrieval of such knowledge poses an ‘immense and daunting’ challenge to learners of additional languages. Verhallen and Schoonen (1993) report that lower levels of educational achievement reported in Turkish–Dutch bilingual children in the Netherlands are often attributed by educationalists to lower levels of lexical knowledge in Dutch. In any case, it is clear that for Polish children in Ireland, typically being educated through English alone, the development of vocabulary skills in English is absolutely vital to their educational success.

Monolingual Lexical Development The acquisition of vocabulary is not a discrete and finite task. Rather, the child must go through a succession of acquisitional phases before a word can even begin to be said to be truly ‘known’. Clark (2009: 86f.) notes, for example, that children acquire information about adult word-meanings gradually, as they learn more about each term and the various contexts in which the term can appear. Henrikson (1999: 308) stresses the complexity of vocabulary learning, and cites Aitchinson’s (1994) claim that the child is faced with three tasks when acquiring word meaning, namely labelling, packaging and network building. Labelling involves learning that words refer to objects in the environment and determining which sequences of sounds can be used as a name for an entity. Packaging involves the child learning that any particular word may refer to a number of different referents (i.e. discovering the range of meanings for one word). Network building is concerned with the process of fitting words into semantic networks. The child comes implicitly to realize that words must be stored within the lexicon in an organized manner. Words

154

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

from this point onwards begin to be structured hierarchically, which facilitates the child’s categorization and determination of meaning-relations between them. Lexical development is thus a dynamic and evolving process. Verhallen and Schoonen (1998: 456) state that the three tasks discussed above mean that the vocabulary growth of children occurs on two levels: the macro level (relating to the size of the lexicon) and the micro level (the level at which knowledge of individual word-meaning deepens and becomes ‘decontextualized’). The present study looks at L2 children’s vocabulary proficiency across both the micro- and the macro-level, by examining not only the Polish children’s vocabulary scores, as measured by a standardized expressive vocabulary test, but also their use of word-meanings, as gauged by a definition task.

Bilingual Lexical Development Gosy (2007: 40) reminds us that bilinguals cannot be regarded as two monolinguals in one person (cf. Grosjean, 1989). Bilinguals need to be seen in very much their own terms. Baker (2006: 102), in his wide-ranging exploration of bilingualism (both simultaneous and sequential), notes that often children on the way to becoming bilingual hear one language at home and another language outside the home (for example, in an educational setting). Obviously, the vocabularies developing from exposure to the two distinct environments in question may be quite different in terms of semantic coverage. For children in the process of becoming bilingual, lexical proficiency may differ from that of their monolingual peers, at both the micro- and the macro-level. Not only may bilingual children (especially sequential bilinguals) have a smaller or semantically less diverse vocabulary in one or other of their languages than their monolingual peers (cf. e.g. Umbel et al., 1992), but the depth of their word knowledge may be inferior (cf. e.g. August et al., 2005; Verhallen & Schoonen, 1998). It appears that in the development of vocabulary in the bilingual both the rate and type of vocabulary acquisition may be different relative to monolingual lexical development (cf. CruzadoGuerrero & Carta, 2006; Golberg et al., 2008). All of the above may sometimes result in bilingual children being misdiagnosed as having a language disorder (cf. Singleton & Ryan, 2009). In all types of language development certain types of lexis may be less difficult to acquire than others. Nouns are generally considered to be easier to establish in memory than verbs (Meara, 1996: 31), being more likely than verbs to have stable and concrete referents, making it easier for the child to

English Le x ical Acquisit ion Among Polish Migrant Children

155

map phonological form onto meaning (Tomasello, 2003, cited in Golberg et al., 2008: 5). Likewise, there are indications that concrete words are easier to learn than abstract words, as their imageability makes retention and recall easier (DeGroot, 1993; Ellis & Beaton, 1993; cf. Singleton, 1999: 142). In summary, previous research indicates that lexical development in the child acquiring two languages (sequentially in particular) may differ vis-à-vis monolingual lexical development along dimensions of both breadth and depth. We may speculate (i) that Polish children acquiring English in Ireland as an additional language in what for them is a bilingual situation will know fewer English words than their Irish age-peers and (ii) will know less about the meaning of the English words at their disposal. We can also speculate that particular word types may be easier for the Polish children to acquire. For example, in the process of lexical development they may find nouns easier to contend with than verbs, and concrete words easier than abstract words.

Testing and Definitional Issues The first and most important question to ask when testing is always ‘What do we want to test?’ (Nation, 2001: 346). Our study had as its aim to test expressive vocabulary, and it used two standard instruments of expressive vocabulary assessment. The first measure was a standardized expressive lexical test and the second a standardized definition test. According to Cohen and Manion (1980: 172), variables must have two properties. First, they must be measurable, and second, the proxy variable must be a valid indicator of the hypothetical variable which a given study addresses. That is to say, in this case, the vocabulary test results in question had to be a reasonable proxy for expressive vocabulary proficiency. Expressive vocabulary proficiency is not, however, directly measurable unless it has been operationally defined. In order to examine and discuss Polish children’s acquisition of expressive vocabulary, therefore, we needed first to define expressive vocabulary knowledge in operational terms. Nation (2001: 25) defines productive (or expressive) vocabulary use in terms of ‘wanting to express a meaning through speaking . . . and retrieving and producing the appropriate spoken. . .word form’. Nation discusses the role of the ‘word form’ relative to this kind of definition: it is not only necessary that a child should know the base or stem of a given word, he/she must also be able the retrieve the precise relevant form of that word and apply it where appropriate. On the basis of Nation’s definition of productive (or expressive) vocabulary, the standardized expressive lexical test used in the

156

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

present study can be said to provide an adequate representation of productive word knowledge, especially bearing in mind that the aims of the present study are comparative at a broad-brush level, and that the study is not designed to look in a detailed manner at the lexical profiles of individual children. The vocabulary of a language, it should be emphasized, is not merely a ‘list of words’ (Carter, 1998: 29). Clark (2009: 383ff.) notes that learning a word involves much more than just recognizing its form. It involves assigning it meaning, determining what grammatical category it belongs to and tracking the constructions it can appear in. Probing acquirers’ word definitions helps us to analyse their lexical knowledge in more depth and allows us to determine to what degree a word can be said to be ‘known’ (Verhallen & Schoonen, 1993). In word definition tasks, the child must supply meaning associations between words. These have been traditionally divided into two classes: paradigmatic and syntagmatic. Paradigmatic sense relations have been described as referring to hierarchical relations (vertical) and paradigmatic associates have traditionally been taken to belong to the same grammatical class as the stimulus item (for example, given the word ‘cat’, paradigmatic associates include ‘animal’, ‘dog’, ‘pet’ etc.) (Carter, 1998: 197). Syntagmatic associations, on the other hand, have traditionally been seen as deriving from sequential relationships between associate and stimulus (being seen in terms of ‘horizontal’ associations). Syntagmatic associations of the word ‘cat’ might be ‘furry’, ‘plays’, ‘scratches’ and so on. Children have been found to favour syntagmatic associations in their early years, paradigmatic associations taking the upper hand as children grow older, and their word definitions concomitantly becoming more complex and elaborate. The transition from using mainly syntagmatic categories of associations to using more paradigmatic categories (such as superordinates) has been referred to as the ‘syntagmaticparadigmatic shift’ (e.g. Watson, 1985: 182). Acquirers of additional languages in the early stages of this process behave similarly to very young L1 acquirers, being more likely to produce syntagmatic responses than paradigmatic ones (Singleton, 1999: 130ff.) The hierarchies implicated in paradigmatic associations may be either partonomic (when referring to part-whole associations) or taxonomic (when referring to class-inclusion) (Verhallen & Schoonen, 1993). Paradigmatic associations are particularly important as indicators of whether the child can categorize and generalize. Definition tasks require, precisely, a categorization of the stimulus item in question in paradigmatic terms (e.g. a ‘cat’ is an ‘animal’, ‘green’ is a ‘colour’ etc.), as well as use of other defining properties. The definition task used in the present study allowed us to look at the

English Le x ical Acquisit ion Among Polish Migrant Children

157

children’s acquisition of vocabulary and their knowledge of word meanings and their relationships in some depth.

The Migrant Children’s Proficiency in Polish The migrant children’s proficiency in Polish was examined in the present study by means of a Polish translation of the standardized (English language) expressive vocabulary test used. Evaluating lexical competence in both languages is essential when determining the overall lexical proficiency of a bilingual child (Baker, 2006; Cruzado-Guerrero & Carta, 2006: 3). Not only will vocabulary be distributed across both languages, but the sequential bilingual child’s level of lexical proficiency in his/her native language may affect acquisition of vocabulary in the additional language. Cummins’ (1984) ‘developmental interdependence hypothesis’, for example, claims that the degree of development in the L1 will strongly influence attainment in the additional language. Well-developed lexical skills in the child’s L1 in addition to knowledge of a body of vocabulary related to lexical needs in the additional language have been shown to accelerate lexical development and vocabulary size in the latter (Simon, 2007: 65). It is important therefore to consider the child’s native vocabulary profile when evaluating vocabulary proficiency. Determining the bilingual child’s L1 proficiency can also help us to distinguish between a language learning disability and a temporary (second) ‘language barrier’, as high levels of proficiency in the L1, co-existent with low levels in the L2, indicate that the child’s problems are restricted to the acquisition and use of the additional language and therefore unlikely to be disability-related.

Language Background The language background of a bilingual child will have an important impact on his/her acquisition of vocabulary. A bilingual child’s vocabulary (and communicative) competence will be determined by the nature of his/ her exposure to the languages involved (Miller, 1984a). Many different factors, such as age of arrival, length of residence and amount of exposure to the L2 will affect both the type and the amount of linguistic input in either language and the vocabulary proficiency of any given child (Miller, 1984a). The family pattern of language use will also affect the child’s acquisition of both languages. Cruzado-Guerrero and Carta

158 Linguist ic and Cultural Acquisit ion in a Migrant Communit y

(2006), in their study of bilingual English–Spanish children in the United States, found a direct relationship between children’s relative amount of language exposure at home in English or Spanish and their vocabulary outcomes in both of those languages. Situational exposure to and use of a given language, interacting with its status in the wider community, will clearly affect the kind and diversity of lexical input experienced and the degree of practice with the words encountered (Verhallen & Schoonen, 1993: 346). Both parents of all the Polish children in the present study are Polish, and these Polish children are exposed almost entirely to Polish in the home. They diverge, however, in terms of the levels of English input in their everyday lives, which may impact on English lexical proficiency. One notes that Jia (2003) measured the richness of the English (L2) environment of Mandarin L1 children in New York and found that a variety of sources of native speaker input including books, television and English-speaking friends had a positive effect on English development (see also Jia & Aaronson, 1999, 2003).

Sample The sample in the present study consists of nine Polish children with English as a second language and nine Irish children, all native speakers of Irish English. All of the children were aged between six and seven years of age at the time of initial testing (June 2007). There were six girls and three boys in each group. The age-range of the children was six to seven years at initial testing, a maturational stage generally considered to be a period of considerable vocabulary growth. Children are typically acquiring an average of 3000 new words per year at this age (Clark, 2009). The children who participated in the study were reported to be developing in typical manner with respect to language ability, and no child was receiving speech and language therapy at the time of participation in the study.

Polish children At the first occasion of data collection (June 2007), the Polish children had a mean age of seven years and two months, and had received an average of two years’ and ten months’ exposure to English in the context of a school or preschool. All of the children had arrived in Ireland after two years of age. Parental report via questionnaire indicated that all of the children had been exposed almost exclusively to Polish until their entry into school. None of the children knew any language other than Polish and English (plus school-learned Irish).

English Le x ical Acquisit ion Among Polish Migrant Children

159

Irish children The mean age of the Irish children was seven years and two months at the first occasion of data collection. The Irish children were all native speakers of English, had had formal exposure to Irish as a school subject but no foreign language experience at the time of testing.

Procedure and Materials There were two rounds of data collection in the study, separated by six months. The first round of data collection took place in June 2007 and the second in December 2007. This six-month period was necessary not only in order to look at vocabulary development over time, but also to avoid practice effects. A six-month period is generally indicated as an appropriate time interval between assessments. Both the Irish and the Polish children completed two vocabulary assessments. These assessments were (i) a standardized expressive vocabulary test and (ii) a standardized word definition task. The Polish children also completed a Polish translation of the expressive vocabulary test. Owing to logistical constraints, the Polish translation was administered only in the second round of data collection in December 2007. A language-background questionnaire was distributed to the parents of the Polish participants in June 2007. Data from these assessments and questionnaires are reported in what follows. The English-language data were collected by the one of the present authors and the Polish data by two bilingual Polish PhD students of linguistics.

Standardized expressive vocabulary test The standardized expressive vocabulary test used is the Test of WordFinding (TWF-2, German, 2000), which can be administered as an expressive vocabulary measure, as well as a word-finding assessment.

Standardized word definition task The standardized word definition task used is the Aston Index Vocabulary Scale (Newton & Thomson, 1976). Children are asked to define 26 different words (e.g. ‘what is a house’?). Although individual words out of context are examined in the Aston Index Vocabulary Scale, and it in this sense presents ‘decontextualized’ language use (Snow et al., 1991: 90), correct responses require more than simply one-word answers. Analysis of children’s

160

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

answers therefore not only allows us to evaluate the child’s ability to generalize and categorize, but also gives us an insight into their use of vocabulary within a sentence and their ability to relate words to other associated words.

Polish translation (of the TWF-2) As previously mentioned, unless proficiency in the child’s L1 is considered, few conclusions can be drawn about the bilingual child’s general vocabulary skills and learning potential (Abudarham, 1987a: 89). The Polish translation of the TWF-2 was used to determine the children’s level of lexical competence in their L1 and to compare competencies across both languages.

Language background questionnaire The language background questionnaire elicited information about parental level of English, about details concerning the language background and history of the child, and about the use of English in the home. The questionnaire was translated into Polish to ensure adequate understanding of the questions. Mean scores for the two standardized measures used were calculated to determine if there were any quantitative differences between Polish and Irish children’s expressive vocabulary acquisition over the six-month period. In order to investigate whether differences were significant, independent samples t-tests were carried out for the TWF-2 and the Aston Index Vocabulary Scale. The Polish children’s individual scores compared to the Irish mean score for both standardized tests are shown using bar charts. To determine whether there were qualitative differences between Polish and Irish children’s expressive vocabulary acquisition and proficiency, meaning relations investigated in the definition task were analysed, according to Verhallen and Schoonen’s classification model (1998: 469). This model delineates six different categories of meaning relations and is reproduced in Figure 8.1. Verhallen and Schoonen subdivide the paradigmatic category into three subcategories, which mark hierarchical relationships. These categories were discussed earlier. The authors also include a ‘dummy superordinate’ in their classification scheme. Children may substitute a ‘dummy’ (e.g. ‘something’ or ‘thing’) when they fail to find an appropriate superordinate term that matches their definition. The syntagmatic category distinguishes between ‘specific episode’ and ‘association’ meaning relations. Specific episode meaning relations are either spatial or relational, and association refers to diverse meaning relations that

English Le x ical Acquisit ion Among Polish Migrant Children

161

Figure 8.1 Classification scheme of meaning aspects (Verhallen & Schoonen, 1998: 469)

the child might allocate to a stimulus word (such as functional and instrumental associations). A third category involves subjective meaning relations, which are classified as ‘attitudes’ (for example ‘it’s fun’). The number of meaning relations that the children use are referred to as the ‘quantity’ of word knowledge, whereas the types of meaning relations are referred to as the ‘quality’ of word knowledge (Verhallen & Schoonen, 1998: 460). Bar charts were created to show differences in numbers and types of meaning relations across the Polish and the Irish group for two different stimulus vocabulary items. To analyse patterns in expressive vocabulary acquisition across the Polish group, the Polish children’s acquisition of nouns versus verbs and abstract nouns versus concrete nouns was examined. Analysis of acquisition of nouns versus verbs was conducted using the results obtained in the TWF-2. The percentage of correct answers across these vocabulary groups was calculated and illustrated using bar-charts. Paired samples t-tests were carried out to determine whether particular types of vocabulary were significantly easier for the Polish group to learn. However, as the TWF-2 examines only concrete nouns, investigation of concrete versus abstract nouns was conducted using the Aston Index, which contains 11 concrete stimulus vocabulary items and 15 abstract stimulus vocabulary items. A paired samples t-test was conducted on the Polish children’s acquisition of concrete and abstract nouns to determine whether abstract items were more difficult for these Polish children to learn than concrete items. A qualitative analysis of the two highest- and lowest-scoring Polish native speakers was undertaken in order to determine whether any factors in the children’s language backgrounds might underlie their vocabulary performance (this analysis follows the general approach used by Singleton & Muñoz, 2007).

162 Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Results and Discussion The analysis that follows is divided into two sections. Section One presents and discusses the results obtained from the standardized vocabulary measures used. Section Two outlines and discusses the results obtained from a qualitative analysis of the two highest- and the two lowest-scoring Polish children.

Section One: Quantitative and qualitative differences between Irish and Polish children’s acquisition of vocabulary Comparison of mean vocabulary scores – quantitative differences between Polish and Irish samples The expressive vocabulary test (TWF-2, German, 2000) yields three different scores: raw scores, quotients and percentile ranks. The word-finding quotient (WFQ) is a special type of standardized score which has a mean of 100 and a standard deviation of 15. Mean scores for the TWF-2 (see Figure 8.2) show that the Polish children’s scores, although lower than the Irish children’s in both rounds, increased over the six-month period by 12 from a mean of 75 to 87, whereas the Irish children’s mean score increased by 0.3 from a mean of 104.2 to 104.5. A 12-point jump indicates that the Polish children’s mean score increased by almost an entire standard deviation over 120.0

Test of Work-Finding (TWF-2) 104.2 104.5

102.6

Irish Children

Polish Translation

Word Finding Quotient

100.0 87.0 80.0

75.0

60.0 40.0 20.0 0.0

Polish children

Round 1

Round 2

Figure 8.2 TWF-2 – Irish and Polish children’s mean WFQ scores in Round One and Round Two of testing and mean scores of the Polish translation of the TWF-2 (at Round Two of testing)

English Le x ical Acquisit ion Among Polish Migrant Children

163

the six-month period, At Round Two it contained scores within the range of the monolingual peers (82–116). An independent samples t-test showed that there was a statistically significant difference between the Polish children’s acquisition of English vocabulary as compared to the Irish children’s over the six month period (t (16) = 3.166, p < 0.05, one tailed, d = 1.49). The Polish children’s mean WFQ on the Polish translation of the TWF-2 shows that their mean Polish score was similar to the Irish children’s English score in December 2007 (Round Two of testing), at 102.6. In the definition task (the Aston Index Vocabulary Scale – Newton & Thomson, 1976), the Polish children improved by a larger margin than the Irish children over the six-month period, as shown in Figure 8.3. When converted to age equivalents, the Polish children’s scores increased from an ageequivalent of 5/6 to an age-equivalent of seven, whereas the Irish children’s scores increased from an age equivalent of 8;6 to an age equivalent of nine over the period of six months. In six months, then, the Irish children showed an age-equivalent jump equal to their chronological age (6 months), whereas the Polish children’s age equivalent jumped by 1–1½ years. An independent samples t-test indicated that the difference between the Polish and the Irish children’s vocabulary acquisition over the six-month period, as measured by the Aston Index, was not statistically significant (t (16) = 0.855, p = 0.203, one tailed, d = 0.42). (Note: The Irish children performed above chronological age level on the Aston index. However, a control group with a chronological 18.0

The Aston Index

16.7 15.7

16.0 14.0 12.0

Score

12.0 10.0

10.2

8.0 6.0 4.0 2.0 0.0

Polish Children

Irish Children Round 1

Round 2

Figure 8.3 Aston Index – Irish and Polish children’s mean scores at Round One and Round Two of testing

164

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

mean age of 9;0 cited in the Aston Index manual received a mean age score of 10;6. Given this disparity, a mean age score of 8;6 seems appropriate for the Irish control group in round one of testing, as their mean chronological age on round one of testing was 7;2. It is important to note that the Polish group’s age score of seven on round two of testing cannot, therefore, be seen as age-appropriate). As expected, the Polish children learning English as an L2 lagged behind their monolingual peers on measures of both expressive vocabulary proficiency (as measured by the TWF-2) and word definitions (as measured by the Aston Index Vocabulary Scale). However, the lexical gap between L1 and L2 learners decreased significantly over the six-month period for the TWF-2, indicating that the Polish children in the present study were acquiring expressive English vocabulary at a faster rate than their Irish monolingual age-peers. Although the gap between the Irish and the Polish children also decreased in the Aston Index Vocabulary Scale, the decrease was not significant, and we can thus speculate that defining a word is more difficult for the child learning an L2 than simply naming vocabulary items. The average range of scores on the TWF-2 is between 90 and 110 (as stipulated by the test-makers). In Round One, only one of the Polish children was within the average range with one at just below average (see Tables 1 and 2 in Appendix 1). In Round Two, however, all of the children’s scores had increased, and three of the Polish children were within the average range, with three more just below average. When we consider the results obtained in the Aston Index Vocabulary Scale, we can see that although the increase in vocabulary scores across the Polish cohort is not as marked as the increase in the TWF-2, the children did improve overall and more than the Irish group. In Round One, four of the Polish children had age scores below 5/6. In Round Two, however, only one of the children did. The test-makers of the TWF-2 state that a child with a WFQ of below 90 is likely to have vocabulary learning problems which warrant attention (i.e. further assessment or intervention) (German, 2000: 54). Likewise, the authors of the Aston Index Vocabulary Scale (Newton & Thomson, 1976: 19) state that children can be described as ‘poor readers and spellers’ if they receive an age score equal to or below their chronological age (as discussed above, there is a disparity between age scores on the Aston Index and chronological age, the control group in the manual achieving age scores approximately one and a half years higher than their chronological age). We should be aware that a single vocabulary testing session will give us only a snapshot of the child’s L2 vocabulary on a particular day, and will not take into account the faster vocabulary growth of the L2 child. On the basis of the rate of vocabulary growth in the Polish cohort we can speculate that

English Le x ical Acquisit ion Among Polish Migrant Children

165

they were likely (eventually) to catch up with their monolingual age-peers, although assessing their vocabulary on a single occasion may give an unrealistic picture of their L2 vocabulary proficiency. As seven of the Polish children performed within normal limits on the Polish translation of the TWF-2 (with the remaining two performing just below average), their difficulties on the English-language TWF-2 and the Aston Index Vocabulary Scale appear to relate to second language learning, and cannot be ascribed to a general language learning problem.

Qualitative differences between Polish and Irish children’s English vocabulary acquisition In the present study, both the quality and quantity of the Polish children’s word knowledge in their L2 was lower than the Irish children’s in Round One and Round Two. In the case of the stimulus word ‘house’, for example, although all of the Polish and Irish children gave appropriate responses on both rounds, the Polish children gave fewer word meanings and used fewer types of meaning relations in both rounds of data collection (see Figures 8.4 and 8.5), and neither group’s command of meaning relations showed improvement over the six-month period. Differences between Irish and Polish children became starker when the word definitions required became more complex. In the case of the stimulus 20.0

Aston Index 'House' Round One 18.0

Number of Meaning Relations

18.0 16.0

15.0

14.0 12.0 10.0 8.0 6.0 4.0 2.0 0.0

3.0

3.0

3.0

2.0 1.0 1.0 0.0

Dummy

Hier. Tax

1.0 0.0

Hier. Part Irish Children

Specific Episode

0.0

Associations

Attitudes

Polish Children

Figure 8.4 Aston Index Round One – Irish and Polish children’s use of meaning relations for the stimulus word ‘house’

166

Linguist ic and Cultural Acquisit ion in a Migrant Communit y 18.0

Aston Index 'House' Round Two 16.0

Number of Meaning Relations

16.0

15.0

14.0 12.0 10.0 8.0 6.0

6.0 4.0 2.0 0.0

3.0 2.0 1.0 Dummy

2.0

2.0 1.0

Hier. Tax

1.0 Hier. Part. Irish Children

0.0 Specific Episode

0.0 Associations Attitudes

Polish Children

Figure 8.5 Aston Index Round Two – Irish and Polish children’s use of meaning relations for the stimulus word ‘house

word ‘battle’ for example (see Figures 8.6 and 8.7), the Polish children again made use of a sparser range of meaning relations than the Irish children, and quantity and quality of word knowledge did not increase over the sixmonth period. The results show that Polish children not only produced fewer word meaning types than the Irish children, but also appeared to know less about these words. In the definition task (the Aston Index Vocabulary Scale), the Polish children (as expected) used smaller numbers and fewer types of meaning definitions than the Irish children. This trend did not reduce over time. The Polish children, however, improved by a larger margin than the Irish children on the Aston Index Vocabulary Scale over the six-month period. One should note that the nature of scoring in the test (any indication that the child understands the word is taken as a correct answer) may mask important differences in deep vocabulary knowledge between the Irish and the Polish children. According to the scoring conventions for the test a child may score correctly on a given item without necessarily showing any deep meaning knowledge. As shown in Figures 8.4 and 8.5, the Polish children used fewer paradigmatic meaning relations than the L1 children when describing the word ‘house’. They used mainly associatives when describing ‘house’ (e.g. ‘you live in it’, ‘you can go to the toilet there’, ‘you sleep in it’ etc.). The Polish group used

English Le x ical Acquisit ion Among Polish Migrant Children

167

fewer paradigmatic hierarchic attributions (such as superordinates) than the Irish group across the entire sample. For example, they tended to relate a word such as ‘tree’ more to ‘leaves’ and ‘green’ than to ‘plant’. In Round Two, for example, six of the Irish children used the word ‘plant’ to define ‘tree’, whereas only two of the Polish children did so. Differences between Irish and Polish children became more obvious when word definitions had a more abstract complexion. When defining the stimulus word ‘battle’, for example (see Figures 8.6 and 8.7), the Polish children again used smaller numbers and fewer types of meaning relations than the Irish children. Again, the Polish children who answered correctly relied on associative relations to describe the word (for example, ‘where you don’t like someone and you fight with them’, ‘and they don’t like each other’). The Irish children were more likely to include near-synonyms such as ‘war’, and they often added relevant information about perceived restrictions on membership in the vocabulary item ‘battle’ by specifying the tools specific to a battle (e.g. ‘you use swords and shields’, ‘but with swords, not punching’, ‘there’s sometimes stabbers and shields’). These differences between the Irish and the Polish children in the definition task can be seen across the sample. Only one Polish child scored within the mean (15.7) of the Irish children on the Aston Index Vocabulary Scale.

Number of Meaning Relations

12.0

Aston Index 'Battle' Round One 10.0

10.0

8.0

6.0 4.0

4.0 3.0

3.0

2.0 1.0 1.0 0.0

1.0 0.0

Dummy

Hier. Tax

0.0 Hier. Part. Irish Children

0.0 0.0 Specific Episode

0.0 Associations

Attitudes

Polish Children

Figure 8.6 Aston Index Round One – Irish and Polish children’s use of meaning relations for the stimulus word ‘battle’

168

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Number of Meaning Relations

12.0

Aston Index 'Battle' Round Two 10.0

10.0

8.0

6.0 5.0

5.0 4.0

4.0

2.0

0.0

0.0 0.0

0.0

Dummy

Hier. Tax

0.0 Hier. Part. Irish Children

0.0 0.0 Specific Episode

0.0 0.0 Associations

Attitudes

Polish Children

Figure 8.7 Aston Index Round Two – Irish and Polish children’s use of meaning relations for the stimulus word ‘battle

Even those Polish children who performed close to or above the mean on the TWF-2 could not define the more abstract words, and no Polish child, even on the second round of testing, defined words such as ‘skill’ or ‘connect’ correctly. On the other hand, seven of the Irish children defined ‘skill’ correctly and six defined ‘connect’ correctly on the first round of testing. Given the fact that abstract words appear to be more difficult for the L2 child to learn than concrete words (e.g. Milton, 2007: 50), the fact that the Polish children performed worse on abstract definitions is not surprising. The likely effect on the slope of the Polish children’s vocabulary learning profile is obvious. The Polish children seemed to mirror younger L1 children’s word definitions. Snow et al. (1991) state that younger children tend to include information in their definitions that is incidental, idiosyncratic, and highly personal. The Polish children included much more incidental and idiosyncratic information in their definitions than the Irish children; for example: ‘A cat is the opposite of dogs. The cat is not bad and the dog is bad’, ‘a ball makes you happy’; ‘You can give it as a smelly dinner because fish is smelly’. They also included personal information such as ‘I have a goldfish at home’ (when defining ‘fish’). Clark (2009) contends that the L2 child overextends words more than the L1 child, in order to substitute the closest word in their vocabulary when a more appropriate term is lacking. In this sense, overextension can be seen as a strategy that facilitates communication. The present study supports the

English Le x ical Acquisit ion Among Polish Migrant Children

169

finding that L2 children overextend more than their L1 peers, as the Polish group of children used overextensions much more than the Irish group when defining words. Examples include ‘hands’ (instead of ‘paws’), ‘dog’ (instead of ‘cat’), ‘circle’ (instead of ‘sphere’), ‘arms’ (instead of ‘fins’). All of the Irish children said ‘sphere’ or ‘round’ to describe ‘orange’, whereas seven of the Polish said ‘circle’. According to Anglin (1985), as children get older, they tend to define words in terms of internal constituents and origins. If we consider the stimulus word ‘house’, for example, the Irish children produced definitions that include these two properties (for example, ‘there’s bedrooms in it’, ‘it’s made of bricks’). None of the Polish children defined ‘house’ in terms of constituents or origins, but rather in terms of the function of a house (e.g. ‘sleeping, watching television’, ‘you play with your toys’ etc.). Giving functional definitions is also considered to be characteristic of younger children (Watson, 1985). In summary, the L2 children’s meaning allocation to the English stimulus words is both less extensive (i.e. they mentioned fewer meaning aspects) and less well developed (they mentioned fewer paradigmatic meaning aspects) than the L1 children’s and mirrors the acquisition of a younger child learning an L1. The Polish children’s performance on the Aston Index Vocabulary Scale may have implications for our understanding of L2 vocabulary acquisition. It appears that on the route to fluency in the L2, lexical gaps are evidenced not only in terms of the numbers of words that the child can produce, but also in terms of the child’s understanding and use of those words. This finding supports Verhallen and Schoonen’s (1993, 1998) argument that there tend to be restrictions on the number and the range of meaning aspects that the L2 learner child expresses. The fact that a child can produce a word does not indicate that he or she can describe all its properties or understand all of its conceptual connotations.

Patterns in L2 lexical development across the Polish sample: Nouns versus verbs, and concrete nouns versus abstract nouns The results of the present study show that for the Polish sample under consideration, English nouns were learned more readily than verbs, and concrete nouns more readily than abstract nouns In Round One, 54.7% of nouns were answered correctly, in comparison to 49.5% of verbs (see Figure 8.8). In Round Two, 67.8% of nouns were answered correctly, in comparison to 58% of verbs (see Figure 8.8). A paired samples t-test was conducted to evaluate the difference between N1 and N2 (acquisition of nouns between rounds one and two). There was a statistically significant increase in noun acquisition from Round One (M = 54.7, SD = 2.7)

170

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

% Age Correct Answers

80

Nouns Versus Verbs 67.8

70 60 50

58.0 54.7

49.5

40 30 20 10 0

Round One

Round Two Nouns

Verbs

Figure 8.8 TWF-2 – percentages of Polish children’s correct scores in respect of nouns and verbs in Round One and Round Two

to Round Two (M = 67.8, SD = 1.8, t(8) = 11.7, p < 0.001). An additional paired samples t-test was conducted to evaluate the difference between V1 and V2 (acquisition of verbs between Rounds One and Two). There was a statistically significant increase from Round One (M = 49.5, SD = 4.7) to Round Two (M = 58.0, SD = 4.6, t = 5.1, p < 0.001). A final paired samples t-test was conducted to investigate whether nouns were easier to acquire than verbs. There was a statistically significant difference between the two scores; the score for acquisition of nouns (M = 13.0, SD = 3.3) was significantly greater than that for the acquisition of verbs (M = 8.5, SD = 5.0, t = 2.0, p < 0.075). For the Polish children under consideration, nouns appear to have been easier to learn than verbs, over the six-month period. As can be seen in Figure 8.9, the Polish group gave correct responses in respect of 78.7% of concrete nouns in Round One, and in respect of 89.8% of concrete nouns in Round Two. 0.9% of abstract nouns elicited correct responses in Round One and 1.6% in Round Two. A paired samples t-test was conducted to evaluate the difference between CN1 and CN2 (change in knowledge of concrete nouns between Rounds One and Two). There was a statistically significant increase in knowledge of concrete nouns from Round One (M = 78.7, SD = 4.52) to Round Two (M = 89.8, SD = 2.36, t(8) = 11.87, p < 0.001). An additional paired samples t-test was conducted to evaluate the difference between AN1 and AN2 (change in knowledge of abstract nouns between Rounds One and Two). There was a statistically significant increase in knowledge of abstract nouns between Round One (M = 0.9, SD = 0.274) and Round Two (M = 1.6, SD = 0.255, t(8) = 5.17, p < 0.001). A final paired samples t-test was conducted to investigate whether concrete nouns were

English Le x ical Acquisit ion Among Polish Migrant Children

171

Polish Children - Concrete versus Abstract Nouns (Aston Index) 120

100 % Age Correct Answers

89.8 80

78.7

60

40

20

0

1.6

0.9 Round One Concrete Nouns

Round Two Abstract Nouns

Figure 8.9 Aston Index – percentages of Polish children’s correct scores of concrete and abstract nouns Round One and Round Two

easier to acquire than abstract nouns. There was a statistically significant difference between the two relevant scores; the score relating to concrete nouns (M = 11.1, SD = 2.81) was significantly higher than that relating to abstract nouns (M = 0.7, SD = 0.406, t(8) = 11.39, p < 0.001). For the Polish children under consideration, concrete nouns appear to have been easier to learn than abstract nouns. Carter (1998) suggests that learning new words in the L2 may be related to factors such as how meaningful any particular word is perceived to be by the child. If this is so, it stands to reason that more concrete words such as ‘house’ (a place in which a child spends much of his or her time) will be learnt more easily than words such as ‘battle’, an event which a child is unlikely to experience in his or her lifetime. However, abstract words are necessary not only for successful discourse skills, but also for academic achievement (Verhallen & Schoonen, 1993, 1998). The results of the present study suggest that sequential bilingual children might benefit from systematic and intensive instructional input targeted at enriching and enhancing their abstract vocabulary repertoire. As the child learning a second language appears to have more difficulty learning verbs than nouns and more difficulty learning abstract than concrete nouns, more sustained input for verbs and abstract nouns could also be indicated for them in the school setting.

172 Linguist ic and Cultural Acquisit ion in a Migrant Communit y

To conclude Section One of our results and discussion, it is clear that the results of the present study can only be regarded as meaningful with respect to the nine children considered. What the results suggest, however, is that while these Polish children learning English as an L2 lagged behind their Irish age-peers on measures of expressive English vocabulary proficiency, they were acquiring English vocabulary at a faster rate than their Irish peers. Differences across the Irish and the Polish groups were not confined to size and rate of vocabulary development. The Polish children also appeared to know less about the relevant English words than their Irish peers. Certain types of vocabulary seemed easier than others for the Polish children to acquire. Nouns were apparently acquired more easily than verbs, and concrete nouns more easily than abstract nouns.

Section Two A qualitative analysis of four profiles – individual differences in L2 lexical acquisition Certain factors have been found to be particularly associated with individual differences in rates of L2 development (as outlined previously). In the following section, the profiles of the two highest-scoring Polish children (Ola and Magda) are closely examined, followed by the profiles of the two lowestscoring L2 children (Marek and Zuz·a), in order to determine what factors might underlie their vocabulary performance A summary of the relevant information is presented in Tables 1 and 2 in Appendix 1. Table 1 contains information about the participants’ chronological ages at Round One of testing (June 2007), as well as factors relating to their language background: their age of arrival in Ireland (AoA), their number of years schooling in Ireland (LoS), and their number of years’ residence in Ireland (LoR). It also presents information about L1 and L2 use. Vocabulary proficiency as measured by the two standardized tests and the Polish translation of the TWF-2 is shown in Table 2. The names of the children have been changed to protect confidentiality.

Ola The first subject (Ola) was aged seven years and three months in June 2007. English vocabulary proficiency: Ola obtained a WFQ score of 103 on Round One of testing in the expressive vocabulary task (TWF-2). This score rose to 120 in Round Two, moving her from the 88th to the 91st percentile rank. In the definition task (the Aston Index Vocabulary Scale), Ola achieved a score of 13 on both rounds, giving her an age equivalent of 7;6.

English Le x ical Acquisit ion Among Polish Migrant Children

173

AoA: Ola had arrived in Ireland at the age of 4;8. LoR: Ola had been in Ireland for two years and seven months at the time of initial testing. LoS: She had received two and a half years of Irish schooling at the time of initial testing. Use of L1/L2: Ola used only Polish (L1) at home with her family members. However, she took part in activities in which she regularly encountered and used English. These included swimming, dancing, watching television and going to church. Polish vocabulary proficiency: Ola achieved a WFQ score of 111 on the Polish translation of the TWF-2. Parental knowledge of English: Ola’s mother had learnt English for six years at school and had continued to speak English after leaving school in a work environment. Ola’s mother rated her own level of English as 1 on a scale of 1 to 5. Ola’s father had not learnt English at school, but had attended English classes upon arrival in Ireland. He rated his English as at 4 on a scale of 1 to 5.

Magda The second subject (Magda) was aged seven years and two months in June 2007. English vocabulary proficiency: Magda obtained a WFQ score of 89 on round one of testing in the expressive vocabulary task (TWF-2). This score rose to 103 in round two, moving her from the 23rd to the 58th percentile rank. In the definition task (the Aston Index), Magda achieved a score of 15 in round one and 18 in round two, giving her an age equivalent of 8;6 and 10, respectively. AoA: Magda arrived in Ireland at the age of three years and two months. LoS: She had received three years of Irish schooling at the time of initial testing (June 2007). LoR: Magda had been in Ireland for four years at the time of initial testing. Use of L1/L2: Magda used only Polish (L1) 100% of the time at home with her family members. She took part in activities, however, in which she regularly encountered and used English, including swimming, watching television and staying with friends. Polish vocabulary proficiency: Magda achieved a WFQ score of 108 on the Polish translation of the TWF-2. Parental knowledge of English: Magda’s mother had learnt English for six years at school and had continued to speak English after leaving

174

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

school (she attended English classes for two years upon arrival in Ireland). Magda’s mother rated her own level of English as 4 on a scale of 1 to 5. Magda’s father had also learnt English at school for six years. He rated his English as 4 on a scale of 1 to 5.

Marek The third subject (Marek) was aged six years and eleven months in June 2007. English vocabulary proficiency: Marek obtained a WFQ score of 62 on Round One of testing in the expressive vocabulary task (TWF-2). This score rose to 66 in Round Two, moving him from below the 1st percentile rank to the 1st percentile rank. In the definition task (the Aston Index), Marek achieved a score of 5 in Round One and 9 in Round Two, giving him an age equivalent of below age 5/6 in both rounds. AoA: Marek had arrived in Ireland at the age of five years and six months. LoR: Marek had been in Ireland for one and a half years at the time of initial testing. LoS: He had received one year of Irish schooling at the time of initial testing (June 2007). Use of L1/L2: Marek used Polish (L1) 90% of the time at home and always with his parents. He had one brother (aged 11) with whom he sometimes spoke English. He took part in no outside-school activities and watched Polish television. Polish vocabulary proficiency: Marek achieved a WFQ score of 89 on the Polish translation of the TWF-2. Parental knowledge of English: Neither mother nor father had learnt English before arriving in Ireland. Marek’s mother rated her own level of English as at 1 on a scale of 1 to 5. His father also rated his English as 1 on a scale of 1 to 5.

Zuz·a

The fourth subject (Zuz·a) was aged seven years and four months in June 2007. English vocabulary proficiency: Zuz·a obtained a WFQ score of 60 on Round One of testing in the expressive vocabulary task (TWF-2). This score rose to 66 in Round Two, moving her from below the 1st percentile rank to the 1st percentile rank. In the definition task (the Aston Index), Zuz·a achieved a score of 10 in Round One and 11 in Round Two, giving her an age equivalent of 5/6 in Round One, and 6;6 in Round Two.

English Le x ical Acquisit ion Among Polish Migrant Children

175

AoA: Zuz· a had arrived in Ireland at the age of five years and five months. LoS: Zuz·a had received one and a half years of Irish schooling at the time of initial testing (June 2007). LoR: Zuz·a had been in Ireland for one and a half years at the time of initial testing. Use of L1/L2: Zuz·a used Polish (L1) 100% of the time at home. She spent time mostly with Polish friends outside school. Polish vocabulary proficiency: Zuz·a achieved a WFQ score of 122 on the Polish translation of the TWF-2. Parental knowledge of English: Neither mother nor father had learnt English before arriving in Ireland. Zuz·a’s mother rated her own level of English as 3 on a scale of 1 to 5. Her father rated his English as 2 on a scale of 1 to 5.

Ola and Magda compared As shown above, the profiles of Ola and Magda have both similarities and differences. Both children had similar experience of exposure to English. Although mostly Polish was spoken in both homes, both children regularly took part in activities in which they encountered and used English. Parental knowledge of English was also similar, both sets of parents rating themselves as 4 on a scale of 1 to 5, and both mothers having learnt English in school for six years. Polish vocabulary proficiency (as measured by the Polish translation of the TWF-2) was above average for both children. The greatest difference in their profiles related to age of arrival: Ola had arrived in Ireland at age 4;8, whereas Magda had arrived at 3;2. Length of residence was therefore also different, Ola having lived in Ireland for two years and seven months at the first occasion of testing, and Magda for four years. However, the two children had had a similar length of schooling in the host country; Ola had been in school in Ireland for two and a half years and Magda for three years. Neither child had attended school in Poland.

Zuz·a and Marek Compared

Zuz·a’s and Marek’s profiles also show both similarities and differences. Both children had lived in Ireland for 18 months at the first occasion of testing. Age of arrival in the host country was similar for both children: Marek had been 5;6 upon arrival in Ireland and Zuz·a 5;5. Length of schooling in Ireland amounted one year in Marek’s case and one and a half years in Zuz·a’s case. Exposure to the L2 was also similar. Both children were mainly exposed to Polish in the home, and neither child took part in English language-rich activities outside school. Parental knowledge of L2 revealed that neither set

176

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

of parents spoke any English upon arrival in Ireland. Marek’s parents both rated their English at a level of 1 on scale of 1–5 and Zuz·a’s as, respectively, at a level of 3 (mother) and at level of 2 (father). The most obvious difference between Marek and Zuz·a relates to their Polish vocabulary proficiency, Marek achieving a score of 89 on the Polish translation of the TWF-2, and Zuz·a a score of 122. When we compare the two highest scoring Polish children with the two lowest scoring Polish children, the two most striking differences seem to relate to length of residence and age of arrival. Zuz·a and Marek had both arrived to Ireland at a later age (5;5 and 5;6) than Ola and Magda (4;8 and 3;2), meaning that length of residence for Ola and Magda was longer than for Zuz·a and Marek. The literature (e.g. Collier, 1989; Golberg et al., 2008) suggests that a later age of arrival in the host country confers a short-term advantage in the early acquisition of L2 skills. After two-to-three years’ exposure to the L2, however, this ‘later age’ effect is seen to diminish (Collier, 1989: 517). What emerges in the present study is straightforward: the younger the age of arrival, the higher the vocabulary score. The children’s relative length of residence, however, also needs to be taken into account. We can suggest that, given that the two higher-scoring children had both been living in Ireland for longer than two years, in this case the longer length of residence factor trumped the age of arrival factor. The strong relationship between length of residence and L2 acquisition has long been acknowledged (Cummins, 1991: 84). Both Ola and Magda had been living in Ireland for a considerably longer period of time than Marek and Zuz·a. We can conjecture, then, that length of residence played a part in determining the vocabulary proficiency of the four children in question. This finding supports Cummins’s (1984) notion that number of years’ exposure influences L2 vocabulary development. Ola’s and Magda’s longer length of residence in Ireland means, however, that their length of schooling in the L2 was also longer than Zuz·a’s and Marek’s. As length of schooling in the host country has also been associated with higher levels of L2 proficiency (e.g. Iwasaki, 1981, in Bialystok, 1991), we must consider the effect of L2 education on the children’s vocabulary proficiency. Snow, Cancino, DeTemple and Schley (1991: 97) assert that school exposure to an L2 will facilitate performance on school-like tasks (such as vocabulary assessments) in that language. Parental report informs us that the Polish children in the present study were not immersed in an English-speaking environment until they started school. We can speculate, then, that the children’s exposure to English was minimal until entry into the school system in Ireland. Length of residence may therefore not be as

English Le x ical Acquisit ion Among Polish Migrant Children

177

important a predictive factor (for the group of children under consideration) of L2 vocabulary proficiency as length of schooling. (However, as it is not possible to dissociate length of residence and length of schooling, we can only speculate that this is the case). Exposure to the L2 may also play a role in determining a child’s vocabulary proficiency (Cruzado-Guerrero & Carta, 2006; Cummins, 1984; Singleton & Ryan, 2009). The amount of exposure that a child has to the L2 will surely impact on their language proficiency and their vocabulary outcome in that language. Although the language of the home was almost exclusively Polish for all four children, Ola and Magda had a wider experience of, and exposure to, English in the form of outside-school activities. Zuz·a and Marek, on the other hand, did not take part in L2 activities outside school, and socialized mainly with other Polish children. We can hypothesize, then, that Ola’s and Magda’s exposure to English is wider and more varied contexts than Zuz·a’s and Marek’s may have played a role in their English vocabulary proficiency. Cruzado-Guerrero and Carta (2006: 3) state that there is a direct relationship between the relative amount of language experience in the home and vocabulary outcomes. In the present study, none of the children spoke Polish at home with their parents. However, we can speculate that parents’ proficiency in the second language may impact on the L2 proficiency of the child. Ola and Magda had parents with a higher level of English schooling and higher self-rating of English than Zuz·a and Marek. L2 proficiency has often been related to maternal level of education (cf. Golberg et al., 2008). Although we cannot equate parental L2 competence with maternal educational level, we can posit that parental expertise and superior parental education in the second language is helpful to the child learning that language, and may act as a facilitator of vocabulary development. We might expect the children with a high proficiency in Polish vocabulary to have a high proficiency in English vocabulary, given that L1 proficiency is thought to impact positively on the acquisition of the L2 (Appel & Mukysen, 1987; Cummins, 1984; Umbel & Oller, 1995). At first glance, this appears to be the case. Both Ola and Magda showed a high level of Polish expressive vocabulary proficiency as measured by the translation of the TWF-2 (their scores place them in the 77th and 70th percentile, respectively). However, Zuz·a also achieved a very high score on the Polish translation, obtaining a WFQ score of 122. This score places her in the 93rd percentile and is the highest score of the nine Polish children on the TWF-2. Marek, on the other hand, achieved a WFQ score of 89 on the Polish translation of the TWF-2, placing him in the 23rd percentile. When we relate these findings to the group of Polish children as a whole, we see that there does not appear to

178

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

be a consistent pattern of L2 expressive vocabulary proficiency across the Polish cohort in relation to proficiency in their native language. Polish proficiency in this instance does not seem to predict success of L2 vocabulary learning. The two Polish children who performed within the Irish mean on the English TWF-2 scored above the mean on the Polish translation of the TWF-2. Of the other seven children who performed below the Irish mean on the English TWF-2, however, three performed above the mean on the Polish translation. In the present study, an above-average level of L1 vocabulary proficiency emerges, then, as a necessary, but not sufficient condition for an above-average level of L2 vocabulary proficiency. To sum up the ground covered in Section Two of the discussion of results, we cannot unfortunately disentangle confounding language background variables such as length of schooling and length of residence. Ola’s and Magda’s success may relate to both factors, to one of these factors or to neither factor. What is clear in the present study, however, is that the two highest-scoring children Polish children differ from the two lowest-scoring Polish children in terms of the length of time they have lived in Ireland, their length of schooling in the host country, everyday exposure to the L2, and parental expertise in the L2. We can hypothesize, then, that these factors may have impacted on their expressive English vocabulary success.

Conclusions and Implications As a pilot investigation of the expressive vocabulary acquisition of Polish children living in Ireland, the present study was an attempt to explore some of the general assumptions of L2 vocabulary acquisition scholars, including the assertion that L2 vocabulary learners display different patterns (both quantitative and qualitative) of expressive vocabulary development to their same-aged L1 peers, and that particular language background factors will impact on the L2 child’s vocabulary proficiency. The results show that the seven-year-old Polish children in the present study indeed differed from their Irish age-peers on measures of English expressive vocabulary proficiency. Although they lagged behind their L1 peers on standardized expressive vocabulary tests, they did acquire L2 vocabulary at a faster rate than the Irish children over the period of the study. Testing vocabulary just once would have given a false impression of the L2 participants’ lexical development and potential. The findings of the present study also suggest that Polish children know less about the L2 words than their Irish peers. Testing single words on standardized tests may not capture this finding, and teachers and

English Le x ical Acquisit ion Among Polish Migrant Children

179

therapists should be aware that lexical gaps are not restricted to vocabulary size, but also to the quality of knowledge in respect of words thought to be ‘known’. Regarding similarities across the Polish group, certain types of L2 vocabulary seem to have been acquired more easily than others: nouns more easily than verbs, and abstract nouns more easily than concrete nouns. Such findings may have implications for vocabulary teaching. Children may benefit from more intensive input in relation to particular categories of lexis. Despite these general patterns, however, individual variation across the Polish children was noteworthy, and some of the children performed well below even the Polish mean on the L2 vocabulary measures used. As all of the Polish children performed within normal limits on the L1 vocabulary test, we can infer that children’s low scores on the English vocabulary measures relate to L2 language learning only. Therapists and teachers should be aware of the broad range of L2 development, as slower than usual L2 learners may be particularly at risk of being misdiagnosed with a language impairment. Particular language background factors have been shown to affect migrant children’s lexical development in the L2 in this (and other) studies. These include length of residence and length of schooling in the host country, as well as everyday exposure to, and parental expertise in the L2. If we relate these findings to our knowledge of a given child’s language background, we may begin to improve our performance in setting appropriate expectations for that child’s L2 vocabulary development.

References Abudarham. S. (1987) The identification and appraisal of communication problems of children with dual language (DL) Systems – to teach or to therapise? In S. Abudarham (ed.) Bilingualism and the Bilingual: An Interdisciplinary Approach to Pedagogical and Remedial Issues, pp. 75–99. Berkshire: Nfer-Nelson. Anglin, J.M. (1985) The child’s expressible knowledge of word concepts: What preschoolers can say about the meanings of some nouns and verbs. In K.E. Nelson (ed.) Children’s Language, Vol. 5, pp. 77–127. New Jersey: Lawrence Erlbaum Associates. Appel, R. and Muyksen, P. (1987) Language Contact and Bilingualism. London: Edward Arnold. August, S., Carlo, M., Dressler, C. and Snow, C. (2005) The critical role of vocabulary development for English language learners. Learning Disabilities Research and Practice 20 (1), 50–57. Baker, C. (2006) Foundations of Bilingual Education and Bilingualism (4th edn). Clevedon: Multilingual Matters. Bialystok, E. (1988) Levels of bilingualism and levels of linguistic awareness. Developmental Psychology 24 (4), 560–567. Bialystok, E. (1991) Introduction. In E. Bialystok (ed.) Language Processing in Bilingual Children (pp. 1–9). Cambridge: Cambridge University Press.

180

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Carter, R. (1998) Vocabulary: Applied Linguistic Perspectives (2nd edn). London: Routledge. Clark, E.C. (2009) First Language Acquisition (2nd edn). Cambridge: Cambridge University Press. Cohen, L. and Manion, L. (1994) Research Methods in Education (4th edn). Routledge: London. Collier, V. (1989) How long? A synthesis of research on academic achievement in a second language. TESOL Quarterly 23 (3), 509–531. Cruzado-Guerrero, J. and Carta, J.J. (2006) Assessing vocabulary and the bilingual environment in young Latino children. Perspectives of Communication Disorders and Sciences in Culturally and Linguistically Diverse Populations 13 (1), 8–13. Cummins, J. (1984) Bilingualism and Special Education: Issues in Assessment and Pedagogy. Clevedon: Multilingual Matters. Cummins, J. (1991) Interdependence of first and second language proficiency in bilingual children. In E. Bialystok (ed.) Language Processing in Bilingual Children, pp. 70–89. Cambridge: Cambridge University Press. De Groot, A.M.B. (1993) Word-type effects in bilingual processing tasks: Support fora mixed representational system. In R. Schreuder and B. Weltens (eds) The Bilingual Lexicon, pp. 27–51. Amsterdam: John Benjamins. Doyle, A.-B., Champagne, M. and Segalowitz, N. (1978) Some issues in the assessment of the consequences of early bilingualism. In M. Paradis (ed.) Aspects of Bilingualism, pp. 13–21. Columbia, SC: Hornbeam Press, Ellis, N. and Beaton, A. (1993) Factors affecting the learning of foreign language vocabulary: Imagery keyword mediators and phonological short-term memory. Quarterly Journal of Experimental Psychology: Human Experimental Psychology 46A, 533–558. German, D.J. (2000) Test of Word-finding, (2nd edn). Texas: Pro-ed. Golberg, H., Paradis, J. and Crago, M. (2008) Lexical acquisition over time in minority L1 children learning English as an L2. Applied Psycholinguistics 29 (1), 1–25. Goodz, N.S. (1994) Interactions between parents and children in bilingual families. In F. Genesee (ed.) Educating Second Language Children: The Whole Child, the Whole Curriculum, the Whole Community (61–81). Cambridge: Cambridge University Press. Gósy. M. (2007). Speech perception processing in bilinguals and L2 learners. In Z. Lengyel and J. Navracsics (eds) Second Language Lexical Processes: Applied Linguistic and Psycholinguistic Perspectives, pp. 39–59. Clevedon: Multilingual Matters. Grosjean, F. (1989) Neurolinguists, beware! The bilingual is not two monolinguals in one person. Brain and Language 36 (1), 3–15. Henrikson, B. (1999) Three dimensions of vocabulary development. Studies in Second Language Acquisition 21 (2), 303–317. Jia, G. (2003) The acquisition of the English plural morpheme by native MandarinChinese-speaking children. Journal of Speech, Language and Hearing Research 46 (6), 1297–1311. Jia, G. and Aaronson, D. (1999) Age differences in second language acquisition. The dominant language switch and maintenance hypothesis. In Proceedings of the 23rd Annual Boston University Conference on Language Development, pp. 301–312. Sommerville, MA: Cascadilla. Jia, G. and D. Aaronson (2003) A longitudinal study of Chinese children and adolescents learning English in the United States. Applied Psycholinguistics 24 (1), 131–161. Meara, P. (1996) The classical research in L2 vocabulary acquisition. In G.N.Anderman and M.A. Rogers (eds) Words, Words, Words. The Translator and the Language Learner (pp. 27–40). Clevedon: Multilingual Matters.

English Le x ical Acquisit ion Among Polish Migrant Children

181

Miller, N. (1984) Language use in bilingual communities. In N. Miller (ed) Bilingualism and Language Disability (pp. 3–25). London: Chapman and Hall. Milton, J. (2007) Lexical profiles, learning styles and the construct validity of lexical size tests. In H. Daller, J. Milton and J. Treffers-Daller (eds) Modelling and Assessing Vocabulary Knowledge (pp. 47–58). Cambridge: Cambridge University Press. Munoz, C. and Singleton, D. (2007). Foreign accent in two advanced learners: two successful profiles. EUROSLA Yearbook 7, 171–190. Nation, I.S.P. (2001) Learning Vocabulary in Another Language. Cambridge: Cambridge University Press. Newton, M. and Thomson, M.E. (1976) The Aston Index. Wisbech: Learning Development Aids. Patterson, J. and Zurer-Pearson, B. (2004) Bilingual lexical development: Influences, contexts, and processes. In G. Goldstein (ed.) Bilingual Language Development and Disorders in Spanish-English Speakers (pp. 77–104). Baltimore, MD: Paul Brookes. Simon, O. (2007) A comparative study of mother-tongue and foreign language speech perception, lexical access and speech comprehension processes. In Z. Lengyel and J. Navracsics (eds) Second Language Lexical Processes: Applied Linguistic and Psycholinguistic Perspectives (pp. 60–82). Clevedon: Multilingual Matters. Singleton, D. (1999) Exploring the Second Language Mental Lexicon. Cambridge: Cambridge University Press. Singleton, D. and Ryan, L. (2009) Child language development in a multilingual environment. In Z. Lengyel and J. Navracsics (eds) Tanulmányok a mentális lexikonról, pp. 83–96. Budapest: Tinta Kiadó. Snow, C.E., Cancino, H., DeTemple, J. and Schley, S. (1991) Giving formal definitions: A linguistic or metalinguistic skill? In E. Bialystok (ed.) Language Processing in Bilingual Children (pp. 90–112). Cambridge: Cambridge University Press. Tomasello, M. (2003) Constructing a Language: A Usage-Based Theory of Language Acquisition. Cambridge, MA: Harvard University Press. Umbel, V. and Oller, D.K. (1995) Developmental changes in receptive vocabulary in Hispanic bilingual school children. In B. Harley (ed.) Lexical Issues in Language Learning (pp. 59–80). Michigan: Research Club in Language Learning. Umbel, V.M., Zurer-Pearson, B., Fernandez, M.C. and Oller, D.K. (1992) Measuring bilingual children’s receptive vocabularies. Child Development 63 (4), 1012–1020. Van Hell, J.G. and De Groot, A.M.B. (2005) The learning of foreign language vocabulary. In J.F. Kroll and A.M.B. De Groot (eds) The Handbook of Bilingualism (pp. 9–29). Oxford: Oxford University Press. Verhallen, M. and Schoonen, R. (1998) Lexical knowledge in L1 and L2 of third and fifth graders. Applied Linguistics 19 (4), 452–470. Verhallen, M. and Schoonen, R. (1993) Lexical Knowledge of monolingual and bilingual children. Applied Linguistics 14 (4), 344–363. Watson, R. (1985) Towards a theory of definition. Journal of Child Language 12 (1), 181–197.

182

Linguist ic and Cultural Acquisit ion in a Migrant Communit y

Appendix 1 Table 1 Main language background characteristics of the two highest- and the two lowest-scoring Polish children Subject

Agea

Sex

AoA

LoRb

LoS**

% L2 use at home

Ola Madga Marek Zuz·a

7;3 7;2 6;11 7;4

F F M F

4;8 3;2 5;7 5;5

31 48 18 18

30 36 12 18

0 0 10 0

ain

June 2007 in months

blength

Table 2 Vocabulary scores of the two highest- and the two lowest-scoring Polish Children (as measured by the TWF-2 and the Aston Index Vocabulary Scale) on Round One and Round Two of testing Subject

TWF-2 Polish (WFQ)

TWF-2 Round One (WFQ)

TWF-2 Round Two (WFQ)

Aston Index Round One (age score)

Aston Index Round Two (age score)

Ola Magda Marek Zuz·a

111 108 89 122

103 89 62 60

120 103 66 66

7;6 8;6