Bilingualism and Minority Languages in Europe : Current Trends and Developments [1 ed.] 9781443891660, 9781443819435

Citation preview

Bilingualism and Minority Languages in Europe

Bilingualism and Minority Languages in Europe: Current Trends and Developments Edited by

Fraser Lauchlan and Maria del Carmen Parafita Couto

Bilingualism and Minority Languages in Europe: Current Trends and Developments Edited by Fraser Lauchlan and Maria del Carmen Parafita Couto This book first published 2017 Cambridge Scholars Publishing Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2017 by Fraser Lauchlan, Maria del Carmen Parafita Couto and contributors All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-1943-3 ISBN (13): 978-1-4438-1943-5

CONTENTS

Acknowledgements .................................................................................. viii List of Tables .............................................................................................. ix List of Figures............................................................................................. xi Chapter One ................................................................................................. 1 Bilingualism and Minority Languages: The Current Context in Europe Fraser Lauchlan and M Carmen Parafita Couto Section A: Attitudes, Identity and Perceptions towards Minority Languages Chapter Two .............................................................................................. 12 Contesting the Conventionalising of Castilian: The Role of Galician Parents as Counter-Elites Anik Nandi and Ashvin I. Devasundaram Chapter Three ............................................................................................ 34 Practitioner Engagement with Research in Minority Language Contexts: Evidence from Research regarding Gaelic Medium Education Sarah MacQuarrie and Fiona Lyon Chapter Four .............................................................................................. 48 Extralinguistic Factors Influencing the Pronunciation of English by Welsh-English Bilinguals Charles Wilson and Margaret Deuchar Chapter Five .............................................................................................. 70 Modeling the Life and Death of Competing Languages within Populations of Speakers from a Physical and Mathematical Perspective Luís F. Seoane and Jorge Mira

vi

Contents

Chapter Six ................................................................................................ 94 Attitudes amongst Children and their Parents to Speaking and Learning a Minority Language in School: Perspectives from Scotland and Sardinia Fraser Lauchlan, Marinella Parisi and Roberta Fadda Section B: The Benefits of Being Bilingual in Minority Language Areas Chapter Seven.......................................................................................... 120 The Linguistic, Cognitive and Emotional Advantages of Minority Language Bilingualism Nia E. Young, Mirain Rhys, Ivan A. Kennedy and Enlli M. Thomas Chapter Eight ........................................................................................... 141 Language Balance and Cognitive Advantages in Frisian-Dutch Bilingual Children Evelyn Bosma, Elma Blom and Arjen Versloot Section C: Grammatical Aspects amongst Bilinguals and Multilinguals in Minority Language Areas Chapter Nine............................................................................................ 160 The Nominal Domain in Dutch-Papiamento-Spanish Multilinguals in the Netherlands M Carmen Parafita Couto, Rocío Pérez-Tattam and Pedro Guijarro-Fuentes Chapter Ten ............................................................................................. 183 Subject Advantage in the Production of Relative Clauses by Basque Speakers in Contact with Spanish Maria José Ezeizabarrena and Amaia Munarriz Chapter Eleven ........................................................................................ 212 Early Minority Language Acquirers of Spanish Exhibit Focus-related Interface Asymmetries: Word Order Alternation and Optionality in Spanish-Catalan, Spanish-Galician and Spanish-English Bilinguals Timothy Gupton

Bilingualism and Minority Languages in Europe

vii

Chapter Twelve ....................................................................................... 240 Testing Alternative Theoretical Accounts of Code-Switching using Event-related Brain Potentials: A Pilot Study on Welsh-English M Carmen Parafita Couto, Bastien Boutonnet, Noriko Hoshino, Peredur Davies, Margaret Deuchar and Guillaume Thierry Chapter Thirteen ...................................................................................... 255 Bilingual Grammars and the Formation of a New Language in a Minority Language Setting Pedro Guijarro-Fuentes and Acrisio Pires Contributors ............................................................................................. 274

ACKNOWLEDGEMENTS

The editors of this volume would like to thank the following people who gave their time freely in order to peer review the submitted manuscripts. Indeed, some of whom were generous of their time by looking at the same manuscripts after revisions were made. Their help was invaluable in helping us ensure that there would be a consistently high quality of the work and research presented in this book. They are the following: Claire Wilson, Antje Muntendam, Lucía Loureiro Porto, El Mauder, Gerardo Fernandez-Salgueiro, Kate Bellamy, Luis Lopez, Gorka Elordieta Alcibar, Jonathan Morris, Amaia Munarriz, Maria José Ezeizabarrena, Cassie Smith-Christmas, Dick Smakman, Leticia Pablos Robles, Hanna Lantto, Rosa Guzzardo We would also like to thank the various authors who have contributed to this volume. We appreciate the patience shown over the two years that this book has been in preparation. A full list of the authors is detailed in the List of Contributors.

LIST OF TABLES

Table 4-1 Table 4-2 Table 4-3

Table 4-4 Table 6-1 Table 6-2 Table 6-3 Table 6-4

Table 6-5

Table 6-6 Table 7-1 Table 7-2 Table 7-3 Table 8-1 Table 8-2 Table 8-3 Table 9-1 Table 9-2

Phonological variables studied Summary of expected influence of factors Rbrul results listing each factor group that was significant in the choice of Welsh English phonological variant in order of significance Summary of results showing factors favouring traditional Welsh English variants Sub-division of children involved in the research Sub-division of parents involved in the research Breakdown of responses to Q1: Should Gaelic (Sardinian) be taught in school? Breakdown of responses to Q2: Do you think it is more important to speak English (Italian), Gaelic (Sardinian), or are they equally important? Breakdown of responses to Q3: Do you think that learning Gaelic (Sardinian) at school is just as important as learning modern internationally recognised languages, for example, French, German, Spanish? Breakdown of responses to Q4: Do you consider yourself to be Scottish (Sardinian) or British (Italian)? Study 1 – number of participants per age group and school language background Study 2 – number of participants per age group and home language background Study 3 – number of participants per home language background Mean and standard deviation characteristics of the Dutch-dominant and the balanced bilingual children Correlations between the cognitive measures Descriptive statistics for cognitive task measures by language group Spanish Grammaticality Judgment Task - Sample items Papiamento and Dutch Grammaticality Judgment Task Sample items

x

Table 9-3 Table 10-1 Table 10-2 Table 10-3 Table 11-1

Table 12-1 Table 13-1 Table 13-2

List of Tables

Spanish, Papiamento and Dutch Context Based Collocation Task - List of adjectives Details of six studies designed for SR and OR elicitation in Basque and/or Spanish Rates of target-like SRs and ORs produced across elicitation studies in Basque Rates of target-like SRs and ORs produced across elicitation studies in Spanish Summary of possible (transitive predicate) word orders by language and information packaging/pragmatic context Example stimuli, characteristics and associated predictions Number of main language Spanish speakers in England and Wales, London and South West England Mean rate of accuracy, standard deviation, and range of overall accuracy – Acceptability Judgement task

LIST OF FIGURES

Figure 4-1 Figure 4-2 Figure 5-1 Figure 5-2 Figure 5-3 Figure 5-4 Figure 6-1 Figure 6-2

Figure 6-3

Figure 6-4

Figure 6-5

Figure 7-1 Figure 7.2 Figure 7-3 Figure 7-4 Figure 9-1 Figure 9-2 Figure 9-3

Percentage of population (aged 3 and over) able to speak Welsh by local authority, 2011 Location of participants’ home towns. Contains Ordnance Survey data A model for racial segregation Abrams-Strogatz model for language competition Geography affects the competition A bilingual AS-model Should Gaelic (Sardinian) be taught in school? Should Gaelic (Sardinian) be taught in schools (in total and by group: monolingual/bilingual and Scottish/Sardinian)? Do you think it is more important to speak English (Italian), or Gaelic (Sardinian), or are they both equally important? Do you think that learning Sardinian (Gaelic) at school is just as important as learning modern internationally recognised languages, for example, French, German, Spanish, etc.? Learning Gaelic (Sardinian) at school is just as important as learning modern internationally recognised languages such as French, German, Spanish, etc. (Monolingual parents only). Study 1 - English vocabulary standard scores Study 1 - Standardised NARA accuracy scores Study 2 - English vocabulary standard scores Study 2 - English reading standard scores Grammaticality Judgment Task - Mean scores for nounadjective agreement across the three languages Grammaticality Judgment Task - Mean scores for determiner-noun agreement across the three languages Grammaticality Judgment Task - Mean scores for adjective placement across the three languages

xii

Figure 9-4 Figure 10-1 Figure 10-2 Figure 11-1

Figure 11-2

Figure 11-3

Figure 11-4

Figure 11-5

Figure 11-6

Figure 11-7

Figure 11-8

Figure 11-9

Figure 12-1

List of Figures

Context Based Collocation Task - Proportions for adjective placement across the three languages (Approximate) mean rates of target-like SR and OR production in Basque, by participants’ profile (Approximate) mean rates of target-like production of SRs and ORs in Spanish, by participants’ profile. Appropriateness Judgment Task mean acceptability ratings of SVO and VSO word order by bilingual group for broad focus contexts with transitive predicates Appropriateness Judgment Task mean acceptability ratings of SV and VS word order by bilingual group for subject narrow-focus contexts with transitive predicates Mean Word Order Preference Task response proportions with 95% confidence interval indicating VS word order is preferable in subject narrow-focus contexts Mean Word Order Preference Task response proportions with 95% confidence interval indicating SV and VS word order are equally preferable in subject narrowfocus contexts Word Order Preference Task response proportions with 95% confidence interval indicating VO word order is preferable in object narrow-focus contexts Word Order Preference Task response proportions with 95% confidence interval indicating VS word order is preferable in subject contrastive focus contexts Word Order Preference Task response proportions with 95% confidence interval indicating SV and VS word order are equally preferable in subject contrastive focus contexts Word Order Preference Task response proportions with 95% confidence interval indicating VO word order is preferable in object contrastive focus contexts Word Order Preference Task response proportions with 95% confidence interval indicating OV and VO word order are equally preferable in object contrastive focus contexts Event-related brain potentials elicited

CHAPTER ONE BILINGUALISM AND MINORITY LANGUAGES: THE CURRENT CONTEXT IN EUROPE FRASER LAUCHLAN AND M CARMEN PARAFITA COUTO

Europe: the current context It is interesting times regarding the value and status of minority languages in Europe. On the one hand, in recent years, we have seen a revitalisation of many minority languages across Europe, whereby some governments are trying hard to combat their diminishing popularity by introducing legislation to promote and protect these languages. On the other hand, there is evidence that less and less people are being raised to speak these minority languages, and the very existence of many languages is in increasing danger (Romaine, 2007). Scotland could be considered a case in point. While legislation was introduced in 2005 to promote the Gaelic Language (the Gaelic Language (Scotland) Act, which gave effect to the principle that the Gaelic and English languages should be accorded equal respect), at the same time, the numbers of Gaelic speakers in Scotland have continued to decline, albeit at a slower rate than before (Office for National Statistics, 2016). This is not only a problem in Europe. There are predictions that of the approximately 5000 languages that are currently spoken in the world, no more than 500 will survive by the end of the present century (Crystal, 2002). It is unsurprising that most of those that are in danger of disappearing are minority languages in the sense intended in the present volume. By minority languages for the present volume, we mean languages that are spoken by a minority of people in a region or country in Europe, usually in competition with the more dominant, established, and internationally-recognised language of the State. Thus, in the following

2

Chapter One

chapters there are a number of research studies presented that have focused on the following contexts: the speaking of Galician in Spain, Irish in Ireland, Scottish Gaelic in Scotland, Sardinian in Italy, Welsh in Wales, Frisian in the Netherlands, Catalan in Spain, Basque in Spain, Papiamento in the Netherlands, and finally the speaking of Spanish in the UK. Thus, by minority languages we mean both indigenous languages to the community and also immigrant languages (e.g. Papiamento in the Netherlands and Spanish in the UK). While we cannot pretend that this is an exhaustive list of the minority languages spoken in Europe, it does nevertheless represent a significant contribution in the discussion of the place of minority languages in Europe at the current time. While it could be argued that the advantages of being raised bilingual are being increasingly recognised across the world (see for example, Bhatia & Ritchie, 2005; Bialystok, Criak & Luk, 2012; Garcia, 2008; Escudero, Mulak, Fu & Singh, 2016; Lauchlan, Parisi & Fadda, 2013)1, at the same time, there is not the same level of 'uptake' regarding some minority languages in Europe as one might expect. In Chapter Two of this volume, which opens section A "Attitudes, identity and perceptions towards minority languages", Anik Nandi and Ashvin Devasundaram discuss their research investigating the current status of Galician in Galicia, in northern Spain. They interviewed and undertook focus groups with parents who were positively predisposed to the Galician language. It is interesting to note the difficulties that one faces when there is an unequal power relationship between a majority language of the State (in this case, Castilian) and a minority language (in this case, Galician). For example, the parents in the study highlighted the State's failure to provide adequate resources to support the assimilation of Galician amongst children and the next generation. The difficulties that children face when speaking Galician at school are outlined, for example, in a group of friends it might be the case that only one of the group doesn't speak Galician, but in this case the language of the group would revert to Castilian rather than Galician. The role of parents in promoting minority languages in the home domain is discussed in Nandi and Devasundram's chapter. It is the parents in the Galician community who are attempting to revitalise the language through bottom-up policies, such as the formation of social groups, language immersion schools, extra-curricular activities, and even the use of more modern technology such as social media outlets (e.g. WhatsApp). 1

In writing the above statement the authors exert caution and acknowledge the recent evidence (de Bruin, Treccani & Della Sala, 2015) that has highlighted the possibility of publication bias in studies that have demonstrated an advantage of bilinguals over monolinguals in executive-control tasks.

Bilingualism and Minority Languages: The Current Context in Europe

3

The importance of attending formal education in a minority language, in terms of ensuring its survival over the generations, cannot be overemphasised. In Scotland, the introduction of legislation to support the formation of Gaelic-medium education in schools in 1985 has played a major role in the maintenance of the language. The number of schools offering Gaelic-medium education has steadily increased and there are now over 150 pre-school, primary and secondary establishments across Scotland that offer Gaelic-medium education (O'Hanlon, Paterson & MacLeod, 2012). In Chapter Three, Sarah MacQuarrie and Fiona Lyon discuss their research that has investigated the level of engagement from teachers who work in Gaelic-medium schools, and how such a high level of support and engagement has positively impacted on the quality of the various research projects undertaken. The lack of resources available in the minority language is again raised as an issue in this chapter, in particular regarding the identification of literacy difficulties. However, progress is being made, for example, the recent development of a phonological awareness test in Gaelic by Lyon is discussed, but the need for further progress is highlighted. The role of staff training is discussed as being crucial in supporting this process. In Chapter Four, Charles Wilson and Margaret Deuchar outline their investigation into the potential extralinguistic factors that may influence Welsh speakers' pronunciation: in terms of whether speakers decide to use the Standard British pronunciation against the Welsh English variant. Based on previous research they decided to explore the following nine factors: national identity, attitudes to Welsh, attitudes to English, style, gender, home language, region of upbringing, social class, and age. They gathered information on these factors through means of interviews with 41 participants. They found the following extralinguistic factors to be significant: style, home language, gender, and region of upbringing. Specifically, informal, Welsh-speaking home, male, and northern (Welsh) upbringing favoured the production of the traditional Welsh English variant. Exploring predictions about language survival from a physical and mathematical perspective may be considered unusual, however Luís Seoane and Jorge Mira make a convincing case for its value in Chapter Five. They review the most recent research that has attempted to offer mathematical models that can capture the dynamics of dying languages. The work done in this area has expanded exponentially in the last 10 years, and there have been modifications of these models to include a bilingual component. Seoane and Mira discuss the particular case of GalicianSpanish as a minority language-dominant language context, in order to

4

Chapter One

demonstrate how these mathematical models can offer interesting insights and practical implications. Recent advances in this line of research have offered up the following important factors that can be considered when assessing the fate of minority languages. For example, it is through weak interactions between competing cultures that these different minority cultures can survive, although weak interactions can often mean isolation. The geographic distribution of speakers matters: a dense stronghold of speakers of a minority language can reverse an adverse situation where the language is in danger of extinction. Languages can survive if human mobility is reduced by physical barriers. These and other factors are described in Seoane and Mira's chapter. Identity, and its relationship to attitudes towards minority languages, is explored in Chapter Six by Fraser Lauchlan, Marinella Parisi and Roberta Fadda. In a cross-cultural study, they investigated the attitudes and perceptions of 236 bilingual and monolingual parents and children in Scotland and Italy towards the respective minority languages of Scottish Gaelic and Sardinian. They found that generally they were positive attitudes towards both minority languages, both amongst the bilingual group (those who speak Gaelic/English or Sardinian/Italian) and the monolingual group (those who speak English only or Italian only). Perhaps unsurprisingly, attitudes were even more positive amongst the bilingual group. An interesting result was the differences found between the two contexts: attitudes were significantly more positive amongst the Scottish sample compared to the Sardinian sample, with even more pronounced differences between the parents who do not speak the minority languages. In Scotland, monolingual parents displayed positive attitudes towards the teaching and learning of Scottish Gaelic, whereas Sardinian parents were not as positive about Sardinian. The research did not find any connection between identity and attitudes, although more Scottish participants did identify themselves with being Scottish compared to those in Sardinia who identified themselves as Sardinian. The implications of the results are discussed by Lauchlan, Parisi and Fadda, in which they highlight that, despite the generally positive attitudes discovered by their study, it does not necessarily mean that there is wider practice and dissemination of the minority languages under examination. There are other factors that will influence the possible growth and development of these languages, some of which have already been cited above in the context of Chapter Two, namely the unequal power relationship that exists between the minority language and the language of the State, and the lack of suitable resources and opportunities to promote the language.

Bilingualism and Minority Languages: The Current Context in Europe

5

Section B, "The benefits of being bilingual in minority language areas", opens with Chapter Seven where Nia Young, Mirain Rhys, Ivan Kennedy and Enlli Thomas outline three recent research studies they have undertaken in Wales and Ireland that have explored the cognitive, linguistic and emotional benefits of being bilingual in these minority language areas. They explored the relationship between the language used in the home and in school amongst children attending Welsh and Irish medium education. They demonstrate clearly that bilinguals in Ireland and Wales (whose L1 is English) perform just as well as their monolingual English counterparts on several measures of English reading and vocabulary. However, there were differences found in those bilingual children who were L1 Welsh, whereby they performed at a lower level than their LI English and monolingual counterparts. They argue that much work needs to be done to support bilinguals' development of English vocabulary and reading for those children who speak Welsh at home and attend a Welsh-medium school. Young and colleagues also report on the cognitive benefits explored in their research. Results were mixed: generally there was not a consistent pattern found, although two bilingual groups did outperform the monolingual groups on Executive Function tasks, thus demonstrating an advantage in favour of the older (11-12 years old) bilingual children. These bilingual children included the L1 English bilinguals attending English-medium education in Ireland (and learning Irish as L2), and the balanced bilinguals in Wales (children who attended Welsh-medium education and had one Welsh-speaking parent and one English-speaking parent at home). The authors outline the implications of this result in Chapter Six. Finally, Young and her colleagues explored the possible emotional benefits of bilingualism by investigating the relationship between their self-esteem and their measured and self-ratings of language ability. Again, their results did not reveal a consistent pattern, and age seemed to play a role. Younger bilingual children were shown to have lower self-esteem than monolingual children, however, in the older age group, the opposite was true. The bilingual children reported higher ratings of self-esteem than the monolinguals. An interesting result was that the bilinguals, in general, demonstrated that they used more solution-focused coping strategies than monolinguals when faced with certain difficulties. The authors suggest that, even if there is the possibility that there might be issues of selfesteem that affect bilinguals at a younger age, the difference appears to be levelled-out by the time they are around 11-12 years old. These issues do need further investigation, as the authors point out, since there are several

6

Chapter One

factors that need to be considered in this complex relationship between bilingualism, general self-esteem and self-perceptions of ability in literacy (in both of their spoken languages). It is an area that is under-researched and Young and her colleagues should be applauded for attempting to unravel some of the issues involved. In Chapter Eight, Evelyn Bosma, Elma Blom and Arjen Versloot report on their research that explores the possibility of cognitive benefits in bilingual children who speak Frisian and Dutch. The importance of language balance is highlighted: the children involved in the study were selected according to their language abilities in both languages, thus creating two groups: a Dutch-dominant group (who were more proficient in Dutch than in Frisian) and a balanced bilingual group, i.e. those children who had a ‘relative similarity’ in proficiency across the two languages. The two groups were matched on their ability in Dutch, as assessed by various standardised Dutch-language tests. The authors used four Executive Function tasks to explore the possibility of differences in the two groups. Two tasks measured skills in attention (selective attention and executive attention), and two measured working memory (verbal working memory and visuospatial working memory). Similar to Young and colleagues' research, the results were mixed. While there was some evidence of an advantage in favour of bilinguals on the selective attention and the verbal working memory tasks, the effects were small to medium. There were no differences found in the other two tasks. The authors discuss the implications of their results, and highlight the importance of establishing language balance in order to potentially profit from any possible advantages in being bilingual. In Section C, "Grammatical aspects amongst bilinguals and multilinguals in minority language areas", a number of various issues are explored. The five different research studies reported in this volume have explored the interaction of the grammatical systems of different minority-majority languages across Europe. Firstly, in Chapter Nine, Maria del Carmen Parafita Couto, Rocío Pérez-Tattam and Pedro Guijarro-Fuentes report on their investigation of nominal constructions in Dutch-Papiamento-Spanish trilinguals who live in the Netherlands. The three languages differ with regard to gender and noun-adjective order in the nominal domain. The authors used two tasks in order to explore any evidence of cross-linguistic priming in the nominal domain, and to what extent the three languages are affected. One was a receptive task (a grammaticality judgment task) and the other was a productive task (a context-based collocation task). Results showed that the trilinguals seemed to experience more difficulties identifying gender violations on the adjective than on the determiner, both

Bilingualism and Minority Languages: The Current Context in Europe

7

in Spanish and Dutch. However, in terms of identifying gender violations on the determiner, the trilinguals were more accurate in Spanish than in Dutch. Findings also indicated that the trilinguals apply the relevant adjective placement rules to Dutch and Papiamento, but interestingly, they over-generalize post-nominal adjective position in Spanish, regardless of the pragmatic conditions. These results, as well as the possibilities for future research in these areas, are outlined in detail in Chapter Nine. It has been established that subject relatives (SRs) are generally “easier” to process and to produce than object relatives (ORs) in many (mostly VO) languages. However, in an OV language like Basque, previous research has concluded that ORs are easier for adults and children than SRs for processing and comprehension, though with the exception of production, where the SR advantage still holds. This conflicting result is explored in Chapter Ten by Maria José Ezeizabarrena and Amaia Munarriz. They present results from six separate studies that were designed to investigate SR and OR elicitation in Basque and/or Spanish in four groups of participants: (a) Monolinguals (tested in Spanish); (b) Simultaneous 2L1 bilinguals (raised from birth with Basque and Spanish simultaneously); (c) Successive L1 Spanish-L2 Basque bilinguals (raised with Spanish-speaking parents but exposed to Basque from age 4 onwards, or even later, but mainly through the educational system); and (d) Successive L1 Basque-L2 Spanish bilinguals (raised with Basque-speaking parents, who acquired the Spanish language through their exposure to it in the street and neighbourhood, or alternatively, through the educational system in Spanish). The data presented indicate that the advantage, if any, is an SR-advantage, although restricted to children’s production. Thus, this apparent SR-advantage in Basque is argued as being a developmental issue, rather than as a consequence of the inter-linguistic effect of the majority language (in this case, Spanish) on the minority language (Basque). In Chapter Eleven, Timothy Gupton investigates word order alternation and optionality amongst Spanish-Catalan, Spanish-Galician, and Spanish-English bilinguals, focusing on the majority language (Spanish) of these minority language speakers. The research focused on the bilinguals' performance on two different tasks: an Appropriateness Judgment Task (AJT) and a Word Order Preference Task (WPT) that were designed to examine competencies and preferences related to word order and focus type with transitive predicates. Results indicated that those who acquired Spanish as a second language appear to have acquired word order variation with transitive predicates in a native-like manner similar to native Spanish speakers, but only in comparison to the Spanish L1-English

8

Chapter One

L2 group. Compared to the other native Spanish speaker groups, results are rather more varied. Gupton discusses these results and argues that particular language pairings affect Spanish competencies in different ways. For example, the preferences of Spanish-Catalan bilingual speakers for subject narrow-focus replies differ when compared to the English L1Spanish L2 group and the Spanish-Galician bilinguals. Gupton states the case that Spanish-Catalan speakers are not producing such differences as a result of cross-linguistic interference. The research offers some fascinating insights into the language learning of these diverse bilingual groups who speak the minority languages of Catalan and Galician. The innovative adoption of a neuro-scientific approach, through the measurement of event-related brain potentials (ERPs), to investigate codeswitching is described in Chapter Twelve by Maria del Carmen Parafita Couto, Bastien Boutonnet, Noriko Hoshino, Peredur Davies, Margaret Deuchar and Guillaume Thierry. They tested the acceptability of codeswitched nominal constructions in Welsh and English based on contrasting predictions of two linguistic theoretical models of code-switching, namely the matrix language framework (MLF) and the minimalist program (MP) model. The MLF predicts acceptability for a Welsh adjective inserted before a noun in a sentence with an English morphosyntactic frame, and also a violation for English adjectives inserted in the same position in a sentence with a Welsh morphosyntactic frame. The minimalist model instead predicts the exact opposite. The results showed that it was the MLF that appeared to provide a better account of the linguistic mechanism involved, however, there was some ambiguity revealed in the control contrasts, where no differences were found in the two control conditions. Further research is required. However, the authors argue that their research represents evidence that the analysis of adjective placement in WelshEnglish bilingual speech may align with online measures of the comprehension of such constructions. The research in this area is still in its infancy but represents an innovative and exciting area of bilingual investigation for the future. In our final chapter, Chapter Thirteen, Pedro Guijarro-Fuentes and Acrisio Pires describe their research that explores language change and language acquisition of a minority language from the perspective of generative linguistics. In particular, they highlight the interaction between the linguistic features underlying Differential Object Marking (DOM) in Spanish. Their participants were bilingual English-Spanish speakers from different backgrounds across the UK, and they were interested in how this distinct bilingual group develop their internal grammatical system distinctively from monolingual speakers. The results are interpreted as

Bilingualism and Minority Languages: The Current Context in Europe

9

representing a type of new language development that involves language change, rather than due to incomplete language acquisition or attrition. In other words, the group of bilingual speakers that were under study in Guijarro-Fuentes and Pires' research demonstrated the acquisition of new grammars that are distinct from the grammars of previous generations. The authors conclude that this is the result of an ongoing natural process of language change. In the pages that follow, the full details of the research summarised above are provided in the corresponding twelve chapters. This volume represents a large body of research that is current and cutting-edge in the field of bilingualism and minority languages in Europe. The motivation underlying the preparation of this volume was that previously many of these minority languages have been in danger of becoming obsolete, mainly because of negative attitudes towards the speaking of these languages, and in particular the potential negative impact that learning may have had on the person's development in the dominant language of the State. In recent times, there has almost been a complete reversal of this position whereby the benefits of being raised bilingual are beginning to be understood (Bhatia & Ritchie, 2005; Bialystok, Criak & Luk, 2012; Garcia, 2008; Escudero, Mulak, Fu & Singh, 2016; Lauchlan, Parisi & Fadda, 2013). This fact, alongside the resurgence of national and regional identity across many areas of Europe has resulted in there being more interest, and more positivity about minority languages in recent times. However, is this enough to ensure that many of these minority languages will survive for another 100 years, given the small numbers that speak them? Only time will tell. As Thomas and Roberts (2011: 106) state: "The ultimate use of a minority language is mediated by the characteristics of the individual, the nature of the linguistic interchange and opportunities at school, and the availability of the language in the wider society"

These issues may be different across the different countries and regions of Europe, and it is difficult to speculate which languages will survive and which will fall by the wayside. It has been argued that it will be no great loss if Europe were to lose many of its minority languages, since the continent only accounts for about 3% of the world's languages (Romaine, 2007). Such a view may be considered harsh, though it represents the cold reality of the context in Europe. While it may be true that some minority languages in Europe may be close to extinction, it seems, for the moment at least, that research into various aspects of these languages is very much alive and well. Such research, as outlined in this volume, explores diverse issues such as the cognitive and linguistic benefits of speaking these

10

Chapter One

languages, attitudes and perceptions towards them, and exploring the finer grammatical aspects of the languages. It remains to be seen whether such research will play a key role in the continued survival of these minority languages in Europe in the years to come.

References Bhatia, T.K. & Ritchie, W.C. (Eds.) (2005). The Handbook of Bilingualism. Oxford: Wiley-Blackwell. Bialystok, E., Craik, F.I.M & Luk, G. (2012). Bilingualism: Consequences for Mind and Brain. Trends in Cognitive Sciences,16(4), 240-250. Crystal, D. (2002). Language death. Cambridge, UK: Cambridge University Press. De Bruin, A., Treccani, B. & Della Sala, S. (2015). Cognitive Advantage in Bilingualism: An Example of Publication Bias? Psychological Science 26: 99-107. Escudero, P., Mulak, K.E., Fu, C.S.L. & Singh, L. (2016). More Limitations to Monolingualism: Bilinguals Outperform Monolinguals in Implicit Word Learning. Frontiers in Psychology. Published online 2016 Aug 15. doi: 10.3389/fpsyg.2016.01218 Garcia, O. (2008). Bilingual Education in the 21st Century: A Global Perspective. Oxford: Wiley-Blackwell. Lauchlan, F., Parisi, M. & Fadda, R. (2013). Bilingualism in Scotland and Sardinia: exploring the cognitive advantages of speaking a ‘minority’ language. International Journal of Bilingualism, 17(1), 43-56. O'Hanlon, F., Paterson, L., & McLeod, W. (2012). Language Models in Gaelic Medium Pre-School, Primary and Secondary Education. Inverness: Soillse. Office for National Statistics (2016). National Records of Scotland: 2011 Census aggregate data. UK Data Service (Edition: June 2016). Romaine, S. (2008). Preserving Endangered Languages. Language and Linguistics Compass 1/1–2: 115–132. The Gaelic Language (Scotland) Act (2005). Edinburgh, UK: Scottish Parliament Thomas, E.M. & Roberts, D.B. (2011). Exploring bilinguals’ social use of language inside and out of the minority language classroom, Language and Education, 25(2), 89-108.

SECTION A: ATTITUDES, IDENTITY AND PERCEPTIONS TOWARDS MINORITY LANGUAGES

CHAPTER TWO CONTESTING THE CONVENTIONALISING OF CASTILIAN: THE ROLE OF GALICIAN PARENTS AS COUNTER-ELITES1 ANIK NANDI AND ASHVIN I. DEVASUNDARAM

Abstract Recent LPP research reveals how policy-makers endorse the interests of dominant social groups, marginalise minority languages and perpetuate systems of socio-lingual inequality. This paper examines the Castiliandominated Galician linguistic landscape, perceiving the rise of grassroots level actors or agents. These include teachers, parents, family members, language activists and other speakers of minority Galician who play a significant role in interpreting and implementing language policy on the ground. This study locates these individuals as ‘counter-elites’ (Higley 2010; Beard and Phakphian 2012), generally comprised of the educated Galician demographic, who if disillusioned with policy decisions of ruling state elites may develop alternative discourses of resistance to hegemonic ideologies. This analysis centres on Galician parents as counter-elite intermediaries, who implement individual language policy in diverse arenas and collective mobilisations including co-operative funded Galician medium schools. Drawing from in-depth fieldwork interviews, this paper demonstrates that in Galicia’s shrinking Galician speaker pool, counter-elite parents can play an important role in the language revitalisation process. The endeavour is to ascertain whether Galician 1

This chapter is based on an unpublished paper entitled ‘Contesting the Conventionalising of Castilian: Galician Newspeaker Parents as Counter-Elites’, which was presented at the Fifth Cambridge Conference on Endangered Languages, July 31, 2015, the University of Cambridge, United Kingdom.

Contesting the Conventionalising of Castilian

13

counter-elite parents can restore intergenerational transmission and if their microcosmic interrogation of the dominant Castilian discourse could lead to bottom-up language policies.

1. Introduction Until recently, Language Policy and Planning (henceforth LPP) research has largely centred on state-run language policy formulation and planning programmes. The role of the actors for whom LPP is purportedly designed is often understudied or overlooked. The interpretation and implementation of LPP by the subjects of its discourse including parents, students, teachers and other members of civil society has received diminished attention from researchers (Casales-Johnson, 2013; Ricento, 2015). Gaps between policy rhetoric and policy implementation by ruling elites often leave many language policies ineffectual. Therefore, the aforementioned actors, if disillusioned with top-down language policies originating from the state, may resist from the bottom-up and create their own language agenda (McCarty, 2011; Tollefson, 2013). This chapter conceives and locates these resisting individuals and groups as ‘counter-elites’, a concept that will be elucidated further. In this regard, this analysis focuses on the enduring dominance of Castilian as the actively conventionalised language in Galicia. It delves into current efforts by Galician counter-elites to challenge the majoritarian narrative of Castilian. The formulation of an autonomous language agenda in the face of disillusionment with supervening state policy is frequently enacted at the micro-level, particularly within the family. The home-use of a language can facilitate its intergenerational transmission whilst simultaneously spilling over and interacting with the outer social and public spheres (Schwartz and Verschik, 2013). This chapter concentrates on the interaction between macro-level language policy and parental agency in relation to Galician parents from the urban domain. Through their own linguistic behaviour, these parents play a prominent role in the revitalisation and maintenance of Galician outside the home and school space, particularly in the context of framing family language policies (henceforth FLP). FLP, as Fogle (2013: 83) defines, “refers to explicit and overt decisions parents make about language use and language learning as well as implicit processes that legitimize certain language and literacy practices over others in the home”. Ultimately, we will demonstrate how these Galician parents are part of a multitude of urban counter-elites taking the discourse of minority Galician

14

Chapter Two

beyond the home space and engaging in contesting the conventionalising of Castilian. Language policy whether macro or micro, as Spolsky (2009: 1) notes, “is all about choices” and one of the major objectives of a language policy is “to account for the choices made by individual speakers on the basis of rule-governed patterns” (ibid.). Therefore, the family as a micro social unit can be considered as a ‘community of practice’ (Lanza, 2007) with its own norms for language use. Since parents are the in situ language managers in the home domain, language management at the family level refers to the choices and attempts that parents make to maintain a language or languages (Schwartz, 2010; Curdt-Christiansen 2009, 2014). We will examine how parents perform their role as counter-elites contesting Castilian’s linguistic dominance by dissolving boundaries between interior and exterior spatial and discursive contexts such as home and outside, family and community, local and national. In this study, we will focus on parents who have gone through the Galician education system since 1980, and experienced the aftermath of the post-Franco political regime’s (1939 – 1975) language policies. These parents are the embodiment of language revitalisation strategies in Galicia since the 1980s, following Spain’s transition to democracy and the inclusion of Galician in domains of use from which it was previously absent, such as education, public administration and mass media. Traditional speakers of Galician have largely been represented as an aging rural based population with little or no formal training in the language (O’Rourke and Ramallo 2013, 2015). However, it is important to adopt a more flexible, inclusive paradigm that takes into account a middle-class and urban-based demographic that speaks both traditional and standardised varieties of Galician. In this regard, the target research samples of this study are Galician parents from urban/semi-urban backgrounds between the age group of 3550 years, from various occupations. The upper age-range of the sample ensures the inclusion of parents who have experienced the education system’s transition from the Franco regime to Galician autonomy. Data is drawn from fieldwork in three different urban/semi-urban areas of Galicia including Santiago de Compostela, Bertamiráns and Vigo. This chapter is derived from a larger body of doctoral research, which draws from eighteen in-depth semi-structured interviews and two focus group discussions with individual parents and couples. This chapter uses data from two individual interviews and the two abovementioned focus groups conducted in Santiago de Compostela and Vigo. Thematic analysis has been used for data interpretation. This study commences with a brief

Contesting the Conventionalising of Castilian

15

overview of the Galician sociolinguistic milieu, top-down language policy, its immediate impact on the urban realm and the role of counter-elites, culminating with a thematic analysis of the collected data.

2. The Galician sociolinguistic milieu Galician (also Gallego in Castilian or Galego in Galician) is a language variant of the western Ibero-Romance branch, spoken in Galicia, an autonomous region in the north-western part of the Iberian Peninsula. Galicia’s present day population is around 2,800,000; the total number of speakers of Galician is approximately three million including the population who speak it as a second language (Instituto Nacional de Estatística/Galician Institute of Statistics, henceforth IGE 2014). According to the last demo-linguistic survey entitled Enquisa de Condiciones de Vida das Familias (Questionnaire on the Conditions of the Livelihood of the Families) carried out in 2013 by IGE, around 98% of the total Galician population claim to understand Galician and around 90% claim that they can communicate in Galician at varying degrees. However, these statistics can be specious when analysed through the prism of recent trends. Of the population, 41% reported to have Galician as their L1, which signals a reduction of 22% in the last ten years. By contrast, 31% of the total population claimed to have Castilian as their first language, which indicates an increase of 12% since 2003. Another 25% of the population stated that they were brought up speaking both languages. This bilingual demographic has witnessed an increase of 5% since 2003. In relation to daily language use, in 2013, 31% of the population reported using only Galician; a reduction of around 12% in the last ten years. Comparatively, 26% of the total population use only Castilian, the use of which has increased 6% since 2003. Another 42% use both languages on daily basis, signifying an increase of 5% during the last decade (37% in 2003). As the above data demonstrates, there is a constant increase in Castilian speaking monolinguals and the people who use both Castilian and Galician to varying degrees in contemporary Galician society. Concurrently, macrolevel data also registers a continuous language loss amongst Galician speakers, whether monolingual or bilingual, underscoring a seemingly inexorable language shift towards the dominant language, Castilian. The vulnerability of Galician is still largely attributable to an aging and rural-based population (Observatorio da Cultura Galega, 2011a, 2011b). It is worth mentioning here that 52% of the present day Galician monolinguals are more than sixty-five years old (IGE 2014). Until the first half of the 20th century, more than 90% of the Galician population lived in

16

Chapter Two

rural areas, where Galician was the sole language of communication (ReiDoval, 2007). Gradual emigration to urban areas since the mid-twentieth century has destabilised the rural demographic base of Galician. As Castilian was already the predominant language in the cities, the process of language shift towards Castilian in the urban milieu seemed a fait accompli (Ramallo, 2012). Demo-sociolinguistic surveys carried out by IGE between 2003, 2008 and 2013 support the aforementioned claim of language shift from Galician to Castilian (Loredo-Gutiérrez, 2015). These surveys record a constant loss amongst active users of Galician (from 61% in 2003 to 51% in 2013) and a gradual increase in the number of Castilian speakers (from 38% in 2003 to 48% in 2013). This language shift is more prominent in the age group between five to fourteen years. There is a significant increase in children communicating exclusively or mostly in Castilian: 21% in the last ten years (from 53% in 2003 to 75% in 2013). On the other hand, children who speak only or mostly Galician decreased 15% (from 40% in 2003 to 25% in 2013). In this Castilian-dominated topography, the impact of recent language policies on the vitality of Galician, or lack thereof, bears scope for further investigation.

3. Top-down LPP in Galicia In the Galician context, ‘prestige’ is afforded to the dominant language – Castilian (Monteagudo, 2012a; O’Rourke and Ramallo, 2015). Franco’s dictatorship made the use of Castilian obligatory as the only language for administration, education and media. This marked an era of repression and discrimination for the Galician language and the region’s culture (O’Rourke, 2011). During this period, the use of Galician was mostly restricted to the home domain and to informal conversations. After Franco’s death in 1975, democracy returned to Spain. The Spanish Constitution (1978) was written recognising Galicia as one of the autonomous communities of Spain, with Galician designated as the region’s ‘co-official’ language. Later, in 1983, top-down language policy models were put in place in line with the Law of Linguistic Normalisation of Galician (Lei de Normalización Lingüística). Whilst critically analysing the state-driven LPP models designed for Galicia, Lorenzo-Suárez (2005) argues that these LPP models are built on erroneous conceptions about the linguistic vitality of Galician. These misconceptions contribute to an inaccurate analysis of the true numerical and territorial strength of Galicians. Additionally, ever since LPP was put into practice in Galicia, policy stakeholders of the

Contesting the Conventionalising of Castilian

17

ruling centre-right wing Government of Partido Popular de Galicia (Popular Party of Galicia, henceforth PPdeG2), took very little interest in implementing their policy initiatives at the grassroots level. This is mainly because they were more interested in preserving the status quo and not upsetting the Castilian-speaking urban middle class in Galician society (Álvarez-Cáccamo, 2011; Monteagudo, 2012b). Almost all top-down language policy documents in the Galician political milieu purport to achieve what is sometimes referred to as a “balanced or harmonious bilingualism” (Regueira, 2006). In this idealised state of bilingual equilibrium, both Galician and Castilian would co-exist as official languages of the community without influencing or interfering with each other. However, after more than three decades of implementation of top-down LPP in Galicia, macro-level sociolinguistic accounts continue to register a significant language shift towards Castilian, especially amongst the younger generation (Observatorio da Cultura Galega, 2011a, 2011b; IGE 2014). In 2010, the incumbent Galician government introduced changes to the existing language education policy through a new decree entitled O Decreto de Plurilingüismo (The Decree of Plurilingualism, henceforth DDP). According to the government, this decree is primarily based on a survey carried out with Galician parents, based on what the parents want in pre-school as a medium of instruction for their children. Although this new policy claims to ensure the continuation of Galician in primary and secondary schools along with Castilian, it allows the medium of instruction to be that of the children’s home language. There is a contrary, and indeed, a quixotic element to this policy. Since Castilian has been, and is, the most widely spoken language in urban/semiurban areas, a majority of Galician children tend to be brought up speaking Castilian by Castilian-speaking parents. Therefore, with the application of the DDP, Castilian automatically becomes the medium of instruction in the urban pre-primary education curriculum. Ultimately, this present policy towards language in education further constricts the conduits of access to Galician among pre-school students in urban/semi-urban arenas. It is also important to note that ever since this top-down LPP was put into 2 Partido Popular de Galicia (PPdeG): Partido Popular (PP) is a conservative centre-right wing Spanish political party founded in 1989. Partido Popular de Galicia (PPdeG), on the other hand, is an affiliated branch of PP which has been in power in Galicia uninterruptedly between 1990-2005 and then again since March 2009-present. Present Galician Government which is ruled by PPdeG strongly maintains a centralist perspective which is evident from their pro-Castilian discourses.

18

Chapter Two

practice, language shift in the urban contexts has proportionally gained momentum (Loredo-Gutiérrez, 2015). Children in the age group of five to fourteen years are directly affected by this language policy model. The macro data provided by the IGE (2014) reveals that the number of adolescents who never speak Galician have increased by 17% in the last five years. As soon as this data was made public, DDP came under the critical scanner. Government stakeholders, such as the President of the Xunta de Galicia (Government of Galicia), Alberto Núñez Feijóo stated during a press release that the present top-down LPP is pro-Galician and by no means discriminatory towards the language. Whilst defending this language policy model, he further argued that it should be the family, and not the education system, which is responsible for intergenerational transmission of Galician. In his view, speaking Galician or Castilian is a question of individual choice. The Galician government’s former Education Minister Jesús Vázquez Abad and Dario Villanueva, Director of Real Academia Española (Royal Academy of Spanish Language) echo the President’s claim that “individual liberty of language selection in a bilingual society” is a justification, if not an exoneration of the incessant language shift to Castilian (Hermida, 2014; Álvarez, 2014; La Opinión, 2014). There appears to be a marked disjuncture between Núñez Feijóo’s endeavour to separate the ideological dimensions of political discourse and the ‘individual’ parental home space. It could be argued that the latter is always-already linked to ideology. In essence, the state is responsible for creating and implementing the LPP and injecting it into the public domain through various Ideological State Apparatuses (Althusser, 1971) such as schools, religious institutions, mass media, and indeed, through the institution of the family itself. These apparatuses are often used to perpetuate top-down ideologies as a ‘false consciousness’ amongst civil society (Eagleton, 1991). So, when there is a perennial transference of majoritarian influences, state policy and media messages from the public sphere into the home space, the school and home spaces become intertwined. In this regard, the statements by the President of the Xunta de Galicia appear all the more contradictory and specious, because he empowers Galician families with false agency, when in actuality they are always-already ideologically controlled. Therefore, his segregation of the interior home domain as a space of individual language choice, distinct from the exterior or broader dimensions of society (Foucault, 1994) is contestable. Ultimately, statedriven macro-level language policies are designed to address and regulate

Contesting the Conventionalising of Castilian

19

social structures at all possible levels. Home language choices and practice are therefore invariably influenced by the individual family’s perception of these dominant macro political policies and their reverberations in wider social structures. The above situation, where government stakeholders make family members responsible for language loss leads to considering whether it is the individual agency of Galician parents that determines the intergenerational transmission of Galician, or whether there are larger socio-political, economic, cultural and structural variables at play, all implicated in determining the fate of Galician.

4. Counter-elites: the voice of bottom-up resistance Top-down language policies, as discussed above, can use governing strategies to exert control over language beliefs, myths, attitudes, practices and ideologies based on the linguistic culture of a community or individual. In this regard, meso (i.e. school and/or family) and micro (i.e. individual) can be considered as the grassroots level, where the interpretation, negotiation and implementation of top-down language policies take place on a day-to-day basis. The exercise of power, therefore, will not only be found at the macro level of state policy, but also at meso and micro levels, amongst family members or even between partners. For instance, it is mostly parents, as LPP arbiters inside the family domain who decide what language(s) should be spoken at home. Significantly, the grassroots level comprises of various actors or agents, including teachers, principals, parents, civil servants, family members and language activists inter alia. These participants play a significant role in negotiating and implementing language policy on the ground. Amongst all these LPP agents, the free-floating and often fragmented members of a pro-Galician collective could be located as ‘counter-elites’. ‘Elites’ have been defined in scholarly literature as powerful individuals [or groups] who can exert “undue influence over communitylevel processes and outcomes” (Beard and Phakphian, 2012: 145; c.f. Higley, 2010: 163). They reach power “through publicly recognised merit, inheritance, or even [by] force” (Fumanti, 2004: 2). Elites often perpetuate their authority over society through material possessions, family connections, employment status, political and religious affiliation, academic credentials, personal history and personality (Bourdieu, 1984, 1996). In this regard, counter-elites, as Murugova et al. (2015: 274) argue, have relatively “the same characteristics as the ruling elites, except of one - the access to instruments of power (…)” (our italics).

20

Chapter Two

In the context of this study, counter-elites are comprised of the educated Galician demographic, who if disillusioned with policy decisions of ruling state elites, may develop alternative discourses of resistance to hegemonic ideologies. This may then lead to enactments of de facto language policies at the grassroots level. Many of the Galician parents interviewed in relation to this study map onto this definition. These parents act as counter-elites by deploying multifarious mechanisms to contest the ruling elites’ implicit and paradoxical patronage of Castilian in Galicia. These measures include language management in the home space, interaction with social groups through social media and technological interfaces, formation of co-operative mobilisations, and informal social interactions with other parents and their children, outside the school space. We posit this Galician collective as counter-elites significantly due to their own paradoxical construction. Even though the counter-elites function literally and pragmatically as a mode of resistance to the hegemonic ideologies of ruling state elites, they are largely elites themselves (albeit bereft of access to the instruments of power). They can often attempt to occupy a possible power vacuum, by being ready to usurp the policy-making position occupied by ruling elites, if the elites can be dislodged from their privileged position (c.f. Pareto, 1935; Mosca, 1939; Higley, 2010). The actions and motivations of counter-elites align with the Foucauldian claim that the field of power is always in motion, circulation and dispersion, constantly changing its agents (Foucault, 1991). Higley and Burton (2006) observe that the influence of elites in modern society is neither incidental nor intermittent. Instead, it is continuous, systematic and in most cases substantial, as ordinary citizens tend to follow elite authority in an unconscious way (Lewis and Hossain, 2008). In other words, elites exert their power less often by compulsion, and more through social, cultural and symbolic domination (Bourdieu and Passeron, 1977). Unlike Foucault (1978), who interprets power as ‘ubiquitous’ and largely beyond agency or structure, Bourdieu (1984) finds power as culturally and symbolically created, and often legitimised through an interaction between agency and structure. This is precisely the case with Castilian, its official validation by an ambiguous LPP, the elite systems that implicitly promote it, and the absorption of this majority discourse by significant swathes of the polity. However, it is also important to consider the economic stimuli that often form a sine qua non for the governing elite to patronise Castilian over Galician.

Contesting the Conventionalising of Castilian

21

For Bourdieu, the term ‘profit’, should not only be restricted to economic theories but must also broach the exchange of particular forms of capital such as social (achieved through social connections and group membership), cultural (academic credentials) and symbolic (source of prestige) that could subsequently act as a motor for material or economic gain (Bourdieu, 1984, 1986). This is particularly true of Castilian as the dominant symbolic and cultural language that facilitates commerce in Spain, signposting how language as a commodity or service can be considered as a form of labour. In other words, if people are getting paid for their language-related work, the discourse of language invariably takes on a materialistic dimension (Shankar and Cavanaugh, 2012). Therefore, when users of a national language such as Castilian, enjoy economic benefits for using it as the common parlance of national commerce, as well as to gain access to social mobility in the extant status quo, their linguistic practices can accrue symbolic capital. This perpetuation of a majoritarian linguistic practice that overtly or covertly favours a mono-language model as a metonym for capital gain, “can further maximise the conversion [of language] to economic capital” (Urciuoli and LaDousa, 2013: 176). Again, this is one of the justifications for the top-down LPP model to be skewed towards Castilian, which is seen as a more commercially viable language in a neoliberal market-driven economy. Bourdieu’s research is invested in the social reproduction dynamic of how dominant elites continue to retain their position by exploiting these aforementioned forms of capital: social, symbolic, cultural and economic. Social reproduction occurs through socialised norms, commonly accepted or normalised codes and customs that guide human behaviour and thinking - what Bourdieu terms as ‘habitus’. Habitus has been described as a cognitive or mental system of structures rooted within an individual which is often created and reproduced unconsciously, “without any deliberate pursuit of coherence (…) without any conscious concentration” (Bourdieu, 1984: 170). Family, culture, environment and educational milieu of an individual play a significant role in the shaping of one’s habitus. In other words, it can be considered as the internal interpretations of the external structures of society, the latter of which are ideologically influenced, if not controlled, by the elites in power. Family plays a significant role in the reproduction and transference of exogenous social structures, ideologies and discourses. This is largely because parents provide their children with material, human, social and cultural capital “whose transmissions create inequalities in children’s educational and occupational attainment” (Tzanakis, 2011: 76). This is especially pertinent to the choice, application and use of language in the

22

Chapter Two

home domain. In Galicia, the education system, which is always already (again the use of this term: please consider using one or other) controlled by ruling Castilian-speaking elites, influences and reinforces the family’s dominant language reproduction process. This is done by privileging and promoting the achievement of Castilian-oriented cultural and linguistic capital among pupils. Bourdieu’s proposition of symbolic capital, filtered through the Galician context, is reflected in the pre-eminence of Castilian as a medium of instruction. This in turn invokes Foucault’s notion of the power/knowledge nexus. The interaction between social and symbolic capital manifests in the intangible ‘prestige’ afforded to the dominant language - Castilian, which is widely perceived to possess a greater degree of symbolic capital. If the knowledge of Castilian is imbued with symbolic capital, this knowledge becomes susceptible to the discourses and dictates of power. This intricate knowledge/power relationship plays out both in the domains of school and home, where teachers and parents under the custodianship of the state could become arbiters of the use of either Galician or Castilian or both. As argued earlier, the statistical evidence for language shift towards Castilian among children and young adolescents indicates that the pendulum of elite power is swinging towards Castilian in the exterior space of the socio-political sphere. Therefore, grassroots counter-elites, such as parents, from the interiority of their private sphere, can use their individual apprehension and interpretation of this shift to raise public awareness, in order to contest the domination of ruling elites. Their bottom-up contestation through alternative discourses of power further underscores the aforementioned claim of a school/state knowledge/power nexus. In the absence of access to the ‘instruments of power’ mentioned earlier, we shall now inspect the alternative pragmatic modes through which counter-elite Galician parents blur the exterior/interior dichotomy. We shall look at how they bridge inner and outer spatial and social spheres, so as to create new conversations around the revitalisation of Galician.

5. Contesting the conventionalisation of Castilian from the ground Reported language practices seem to confirm that each of the parents exercises his/her individual language practices at home. In the context of this study, a large proportion of counter-elite couples confirmed that they had no reservations in choosing Galician as their home language, whilst

Contesting the Conventionalising of Castilian

23

other pro-Galician individual parents are in constant negotiation about their language choice within the family. For instance, Ana, a mother from Bertamiráns who did not have Galician as her L1 and learnt the language in school, states in an interview that she intends to incorporate Galician into her day-to-day life. However, her family speaks mainly Castilian at home: Ana: Yes, it is true that I am on my own [in her attempt to use Galician]. I have tried to learn and tried to apply Galician in my day-to-day life, not just knowing it academically. But well, my main language is Castilian. I try to speak Galician every moment though. (emphasis added)

Even though her knowledge of Galician is mostly imbibed from books, Ana clearly states that she attempts to exceed her academic awareness of Galician and expand its horizons by incorporating it practically, in her daily life. She further claims to speak with her daughter in Galician for at least an hour every day. Ana adds that if they start speaking in Galician during dinner, her partner, Manolo, and their eldest son, normally participate in the conversation, switching from Castilian to Galician: Ana: Influenced by my own efforts, my daughter, who speaks mostly Castilian, said: Why don’t we find a moment when all of us will speak in Galician? We liked the proposal, and we selected dinner time for this exercise, since that’s the time when we all are together. (emphasis added)

Notably, Ana’s conscious personal decision to expand the ambit of Galician in her everyday life is diversified and instantiated through its adoption in the home domain, as a Language Planning mechanism amongst the family. Fernando and Marisa, a couple from Bertamiráns, mostly converse in Castilian. Both of them learnt Galician at school and currently have two school-going children, with whom Fernando often speaks Galician. Sometimes, when his children come back home from school, Fernando makes it a point to speak with them only in Galician. He states that he wants his children to understand that Galician is not an unfamiliar language and can be used in everyday life, akin to the Castilian they are familiar with in school and the outside world. Fernando: I often speak to them [the children] in Galician so that they can see that this language is no different. It is something normal and serves a purpose in everyday life and is not something that is only studied in school.

24

Chapter Two

Fernando considers that the children may not understand the difference between Galician and Castilian, and that they respond in the language they want to speak in at that moment. He also adds that he does not want to force them to speak any specific language. In essence, Fernando claims to adopt an approach of tolerance in his home linguistic practice. Apart from speaking in Galician, Fernando devotes a considerable amount of time to maintaining his own written Galician and attempts to implement the language in his everyday routine. He has a kitchen garden and writes about it in his diary in Galician. Fernando states: When I write any official document, I do it in Galician. When, I have something private, I also try to document it in Galician...the list for shopping, I do it in Galician. Now I am also writing a sort of diary, you know... ‘I went there today’...and all that. I am doing it in Galician…

The above account could be interpreted as an interesting linguistic practice for language revitalisation and maintenance at the individual level. Reported language practices in the above case studies also demonstrate how the exterior Castilian-dominated sphere ideologically penetrates the interior home space, creating fluctuations, vacillations and inconsistencies in language practice on the ground. Fernando and Ana’s articulations of fidelity to Galician, and their hope for its furtherance in their home space seem to be contradicted by continuous interruptions and intrusions. In this regard, their claimed commitment to Galician is offset by the couple’s and their children’s intermittent ‘lapses’ into Castilian. These reversions could be interpreted as dominant linguistic reverberations - the traces and residues of the family’s Castilian-suffused social, occupational and scholastic interactions and experiences in the outer domain. This incommensurability between the family’s affirmation of Galician and their occasional capitulation to Castilian reiterates the potency of Castilian’s practical, symbolic and ideological dominance. It nonetheless underscores the agency exerted by counter-elite parents, such as Fernando and Ana, to consciously interrogate the Castilian mono-lingual monopoly, particularly within the contours of their home, and increasingly, as we shall see, beyond its precincts. Several pro-Galician parents are involved in a bottom-up discourse that melds their individual efforts with broader collective mobilisations. For instance, parents from the focus group conducted in Santiago de Compostela, underscore the insufficiency and inadequacy of learning materials in Galician available to children. These parents concomitantly highlight their continuing efforts to locate and access such resources:

Contesting the Conventionalising of Castilian

25

Virgilio: What happens is that there is a numerical disadvantage for Galician. Elena: Of course. Bea: Indeed Lara: The thing is that available resources are overwhelmingly in Castilian… Elena: Television weighs in a lot too. On television most of the programmes that you see are in Castilian… Mercedes: The thing is that we don’t have these resources… Virgilio: We have to make a lot of effort as parents to promote Galician. Mercedes: Of course. Virgilio: That we often do... as is the case, I hear, with other parents. Salvador: Yes. Adam: We go to the library almost every week to search for films in Galician Lara: Yes!

The above scenario underscores the active engagement by parents’ to broaden the remit of their language management (Spolsky, 2009). It exemplifies their conscious efforts to continue intergenerational transmission of Galician by bridging the disjuncture and discrepancy in Galician use between the exterior and interior spheres. At the same time, their critique of the paucity of resources and learning materials in Galician, censures the government’s non-attainment of its own language planning goals. It spotlights the state’s failure to provide adequate audio-visual implements and other learning support mechanisms to augment the assimilation of Galician amongst children and the next generation. These counter-elite parents are also self-aware of the cultural and ideological dominance of Castilian in the urban domain. Since Castilian is predominantly the language of children’s socialisation, these parents often place emphasis on the status planning of Galician in the family domain. Aspects of status planning are evident in the following comment by Dario, a discussant from Vigo, who underscores the parental role in developing prestige for Galician at home so that his children do not feel ashamed of speaking Galician: Dario: I believe that the role of families, or at least what I feel I have to carry out with my children, is to instil pride in them about who they are… not only linguistically, but... they should be dignified. They should feel courageous and brave and feel that they have to be independent. I have hope that if they are educated in this way, they can defend Galician, which we speak at home.

26

Chapter Two

To a large extent, Dario’s approach involves accentuating the symbolic capital of Galician at home, by infusing a sense of pride and selfidentification in the use of the language. This could be analysed as being tantamount to a call for Galician ‘national’ pride; noteworthy whilst considering that national’ pride would perhaps conventionally be associated with or at least prescribed to be linked with Castilian. Ultimately, the teleological basis of Dario’s active proliferation of Galician at home is for his children to enunciate and enact their Galician identity without ‘fear’ or ‘shame’, in the outer Castilian-dominated social sphere. Echoing Dario’s experience and perspective, Elena, a mother from Santiago de Compostela, whose son studies in a public school, explains to the group that her son should not be alienated for speaking Galician: Elena: The other day, for instance, he was telling us “sometimes I am uncomfortable and I prefer to change to Castilian and I don’t know what to do”. Then we told him, “look we also went through this. But, it [Galician] is part of your personality, which you don’t need to renounce. You are a Galician speaker and there is no reason why you should abandon it when you are in the playground. Your friends will respect you as will others who don’t like you for this”. I would like to say, yes, we do try to rationalise this (situation) in a way… not simply lecture them and go ahead. If we don’t, we know that he will get confrontational about this.

This statement from Elena further underlines the parental role in status planning of a minority language in the home domain. It also demonstrates how counter-elite parents prepare their children to face the hegemony of Castilian in an exogenous field, equipping their progeny with the psychological strength and linguistic skills to counter the dominant discourse. Bea, another mother from Santiago de Compostela, states that as parents, they are keenly aware of the socio-lingual dominance of Castilian. Therefore, to prevent language shift during children’s early years, they formed a pro-Galician parents group entitled Tribo (‘Tribe’ in Galician), comprising a collective that wants its children to socialise in Galician: Bea: If you have a group of friends that normally communicates in Castilian, most usually the Galician-speaking children shift from Galician to Castilian. But, we have formed a group … a group of Susi’s friends who socialise in Galician. And the medium of instruction in school [cooperative funded immersion schools, discussed later] is also Galician and there are many Galician speaking kids in school…

Contesting the Conventionalising of Castilian

27

The above mobilisation can be considered as a heterodox, stylised strategy of bottom-up resistance by pro-Galician parents contesting the conventionalisation of Castilian. It also points to parental language governmentality (Foucault, 1978) through parents selecting peers for their children, and can be considered as an aspect of language management. This scenario also evokes the discourse of bio-power (Foucault, 1994), whereby parents as physical progenitors often assume the role of custodians over their children’s language practice, perceiving this ‘ownership’ as a putative parental right. Apart from the above micro-level bespoke social groups being formed by like-minded Galician parents, larger cohesions are emerging that straddle the school, social and home spheres. During field research, it was observed that Raiola, a co-operative funded school, has been offering Galician as a medium of instruction for more than forty years (O proxecto pedagóxico/The pedagogic project, Raiola, 2015). On the other hand, Escola Semente (Semente School), started more recently, in 2011, as an exigent response to the state’s extant language policy - ‘The Decree of Plurilingualism’. Escola Semente, akin to Raiola, is a language immersion school initially funded by a group of pro-Galician counter-elites who sought to educate their children in Galician. At present, the Parents Association of Semente has more than 90 members, who actively support the endeavour to defy official language policy from the grassroots (O Projeto/The Project, Semente en Compostela, 2015). One of this study’s focus group respondents from the Semente Parents Association also belongs to the aforementioned pro-Galician parents’ collective, Tribo. Tribo, started as a WhatsApp messenger group in July-August 2013, and now includes more than 40 families who meet several times a week to enable their children to socialise and converse in Galician (Bal and Rodríguez, 2014). The main intention of Tribo, as described by its members in focus group discussions, is to prevent language shift during their children’s early years. Parents interested in joining the collective generally contact the group members though WhatsApp, a technological interface that allows Tribo members to collectively exchange messages. The members of this group also communicate informally amongst themselves, meeting in different places in Santiago to organise or participate in various extracurricular or cultural activities that involve their children’s interaction in Galician (Bal and Rodríguez, 2014). During visits to Escola Semente it was observed that the school has two classrooms redolent with children’s learning materials in Galician including posters, lyrics of traditional Galician songs, storybooks, audio and video resources. A representative from the school explained that the

28

Chapter Two

parents can borrow these materials for a short period and use them to develop an effective home literacy environment in Galician. This illustrates how Galician is being promoted at Escola Semente, with the school addressing the grassroots-level vacuum in the form of the state’s failure to supply adequate and accessible learning materials in Galician. Research also revealed that after school hours, the school space is transformed into a language learning zone, often hosting different Galician-oriented extracurricular activities with the children. Since children from other pre-primary and primary schools can also participate in these group activities by paying the course fee, several group members of Tribo bring their children to participate in these events. In effect, such initiatives further dissolve distinctions between counter-elite parents and opens up alternative avenues for their children to compare notes, interact and bolster their Galician verbal and written skills. In a Castilian-dominated urban terrain, the activities organised by Tribo can also be interpreted as providing an organisational basis for the bottom-up contestation of the conventionalisation of Castilian orchestrated by individual pro-Galician parents. Among the families interviewed in Santiago, four of them are active members of this group.

6. Conclusion Breaking the binary of interior/exterior domains in the above-described scenarios involving pro-Galician parents demonstrates the innate plasticity in the deliberation between Castilian and Galician. In essence, the sites for intersection and overlapping of the two linguistic discourses are subject to daily negotiations, variations and mutations. The habitus of the various participants in these arbitrations constitute a diverse and shape-shifting topography. Various modifiable vectors also play a role, with multiple endogenous and exogenous factors influencing the connection between the interior and exterior realms that impact on the counter-elite parents and their families’ linguistic lived experience. These reagents include the media, peerinfluence, tacit social mores and language practices that are linked to governmentality and bio-power. The latter two can stem not only from the state but also from parental control over children. These interlinked multilayered factors map out in concentric circles of power, from the microlevel of parental jurisdiction to the custodians of political power at the executive policy-making echelons of the state (including regional to federal levels and the layers in-between). This configuration could be linguistically located as a Castilian meta-hegemony (Devasundaram,

Contesting the Conventionalising of Castilian

29

2016), which is essentially a system of hegemony within larger hegemonies, or a hierarchy of interconnected hegemonies. Castilian with its Franco-era historiography dominates the indigenous Galician linguistic and cultural context, also exerting influence over the Galician political machinery. However, both the Castilian master narrative and its Galician government proponents are being contested by counterelites, who are themselves largely urban educated elites. Galician counterelite parents, as we have seen, are not averse to orchestrating their own hegemony of language and identity within the precincts of family and home. They often transpose this exclusionary Galcian-based socialisation to the outside sphere. The above inter-relationship between various layers of linguistic dominance in Galicia could be analysed through the prism of power and a hierarchical ‘meta-hegemony’, and bears scope for further scholarly exploration. Due to Galician’s growing presence in the education system, a generation of counter-elite parents such as Paloma, Bea, Elena and Fernando has emerged, who do not have Castilian as their home language, but are influenced by an ideological bond with Galician. These counterelite parents have started widening the symbolic space for Galician. In a community, such as Galicia, where the existence of orthodox native speakers perceptibly shrinks due to language shift, counter-elite parents, by creating alternative bottom-up language policies, can occupy an important role in the language revitalisation process from the ground. However, it still remains to be ascertained whether counter-elite parents can effectively restore the process of intergenerational transmission of Galician in the home domain and within the community, and in turn, create new generations of native speakers. The profiles discussed in this chapter further reveal that symbolic capital, evident in the privileging of Castilian as the language of communication is to some degree offset by pro-Galician counter-elite parents at home and through their cooperative contestation (involving Raiola and Escola Semente) from the micro-level. These endeavours could be interpreted as counter-hegemonic strategies to destabilise the normalisation and legitimisation of the dominant discourse. However, it must be reiterated that the circuits and corridors of ideology, linguistic culture and power dynamics informing the conventionalising of Castilian in the Galician urban terrain are polymorphous and cannot be reduced to a bipolar model of interior and exterior. In essence, the instantiation of the multi-modal arbitration between the discourses of power, ideology and language is played out on a daily basis through various actors and stakeholders who are implicated in a constantly fluctuating field of agonistic negotiation. An analysis of the parental

30

Chapter Two

linguistic practices in this chapter further indicates that these parents, based on their individual language beliefs, ideologies and management decisions become stakeholders or implementers of language policy within the family domain. Their ‘under-the-radar’ participation in LPP may appear extremely intermittent and ad hoc, but their individual attitudes towards language use, when galvanised into collective mobilisations, can cause an impact in their immediate society’s language behaviour.

References Althusser, L. (1971) ‘Ideology and Ideological State Apparatuses’ in Lenin and Philosophy and other Essays, London: New Left Books, pp 170186. Álvarez, E. (2014) ‘Vázquez sitúa el problema del gallego en que no se emplea al salir de la escuela’. 13 December 2014. Available from: [12/01/16]. Álvarez-Cáccamo, C. (2011) ‘Contra o capitalismo linguístico: perante a crise da língua na Galiza’, Agália, Revista de Estudos na Cultura, 104 (2), 11-28. Bal, R. and Rodríguez, I. (2014) ‘Pais e profesionais organízanse para impulsar o ensino en galego’. 06 March 2014. Available from: http://www.galiciaconfidencial.com/noticia/17979-pais-profesionaisorganizanse-impulsar-ensino-galego [19/06/16]. Beard, V. and Phakphian, S. (2012) ‘Community-based planning in Thailand. Social capital, collective action and elite capture’, in Daniere, A. and Luong, H. (eds.) The dynamics of social capital and civic engagement in Asia, London: Routledge, pp145-162. Bourdieu, P. (1984) Distinction: A Social Critique of the Judgement of Taste, London: Routledge. —. (1996) The State Nobility: Elite Schools in the Field of Power, Oxford: Polity Press. Bourdieu, P. and Passeron, J. C. (1977) Reproduction in Education, Society and Culture, Beverly Hills: Sage. Cassels-Johnson, D. (2013) Language Policy, London: Palgrave Macmillan. Curdt-Christiansen, X. L. (2009) ‘Visible and invisible language planning: Ideological factors in the family language policy of Chinese immigrant families in Quebec’, Language Policy, 8 (4), 351–375. —. (2014) ‘Family language policy: sociopolitical reality versus linguistic continuity’, Language Policy, 12 (1), 1-6.

Contesting the Conventionalising of Castilian

31

Devasundaram, A. (2016) India’s New Independent Cinema: Rise of the Hybrid, New York: Routledge. Eagleton, T. (1991) Ideology: An Introduction, London: Verso. Fogle, L. W. (2013) ‘Parental ethnotheories and family language policy in transnational adoptive families’, Language Policy, 12 (1), 83102. Foucault, M. (1978) The History of Sexuality, an Introduction, Volume I, New York: Randome House, Inc. —. (1991) ‘Governmentality’, in Burchell, G., Gordon, C. and Miller, P. (eds.), The Foucault effect: studies in governmentality, Chicago: The University of Chicago Press, pp 87–104. —. (1994) The Order of Things: An Archeology of the Human Sciences, New York: Vintage. Fumanti, M. (2004) ‘The Making of the Fieldworkers: Debating Agency in Elites Research’, Anthropology Matters Journal, 6 (2), 1–9. Hermida, X. (2014) ‘Feijóo atribuye a las familias la caída del uso del gallego entre los niños’, El País, 14 December 2014. Available from: [12/01/16]. Higley, J. (2010) ‘Elite Theory in Political Sociology’ in Leicht, K. T. and Jenkins, J. C. (eds), Handbook of Politics: State and Society in Global Perspective, Handbooks of Sociology and Social research, New York: Springer, pp 161-176. Higley, J. and Burton, M. (2006) Elite Foundations of Liberal Democracy, Lanham: Rowman and Littlefield. Instituto Galego de Estatística. (2014) Enquisa de Condicións de Vida das Familias. Coñecemento e Uso do Galego. Ano 2013, Santiago de Compostela: Instituto Galego de Estatística (IGE). Retrieved from [18/06/16] La Opinión. (2014) ‘Villanueva atribuye la pérdida de ´galegofalantes´ a la falta de transmisión en las familias’. 16 December 2014. Available from: [19/06/16]. Lanza, E. (2007) ‘Multilingualism and the family’, in Auer, P. and Wei, L. (eds.), Handbook of Multilingualism and Multilingual Communication, Berlin: Mouton de Gruyter, pp 45-66. Lewis, D. and Hossain A. (2008) ‘A Tale of Three Villages: Power, Difference and Locality in Rural Bangladesh’, Journal of South Asian Development, 3 (1), 33–51.

32

Chapter Two

Loredo-Gutiérrez, X. (2015) ‘Aproximación á situación sociolingüística en Galicia. A evolución da transmisión lingüística interxeracional do galego’, in A letra miúda- Revista de sociolingüística da CGENDL, No. 6 (2015). Available from:

[10/01/16] Lorenzo-Suárez, A. M. (2005) ‘Planificación lingüística de baixa intensidade: o caso galego’, Cadernos de Lingua, 27, 37-59. McCarty, T. L. (Ed.) (2011) Ethnography and language policy, New York and London: Routledge. Monteagudo, H. (2012a) Facer país co idioma. Sentido da normalización lingüística, A Coruña: Real Academia Galega. —. (2012b) ‘Política lingüística en Galicia: Apuntes para un nuevo balance’ in Kremnitz, G., Cichon, P., and Czernilofsky-Basalka, B. (eds.) Quo vadis, Romania? Wien: Institut für Romanistik, Universität Wein. Mosca, G. (1939) The ruling class, New York: McGraw-Hill. Murugova, V. V., Skorobogatov, A. V., Bagaeva, I. V., Volkova, N. V., Pavlova, V. A., Zhuravlev, V. P. and Apraksina, N. D. (2015) Positional Approach to the Political Elite Criteria Selection in Democratic States, Journal of Sustainable Development, 8 (3), 271276, available from doi:10.5539/jsd.v8n3p271 [accessed 21/03/16]. O Projecto, Semente Compostela: Escola de ensino Galego, 2015. Available from: [22 November 2015] O’Rourke, B. (2011) Galician and Irish in the European Context: Attitudes towards Weak and Strong Minority Languages, Basingstoke: Palgrave Macmillan. O’Rourke, B. and Ramallo, F. (2013) ‘Competing ideologies of linguistic authority amongst newspeakers contemporary Galicia’ in Language in Society, 42, 287–305. O’Rourke, B. and Ramallo, F. (2015) ‘Neofalantes as an active minority: understanding language practices and motivations for change amongst new speakers of Galician’, International Journal of the Sociology of Language, 231, 147–165. Observatorio da Cultura Galega. (2011a) O idioma galego na sociedade: A evolución sociolingüística 1992-2008, Santiago de Compostela: Concello da Cultura Galega —. (2011b) A(s) lingua(s) a debate. Inquerito sobre opinións, actitudes e expectativas da sociedade galega, Santiago de Compostela: Consello da Cultura Galega.

Contesting the Conventionalising of Castilian

33

Pareto, V. (1935) The mind and society. A treatise on general sociology, New York: Dover. Proxecto Pedagóxico: o noso lugar, a nosa lingua, CPR Raiola, 2015. Available from: [22 November 2015] Ramallo, F. (2012) ‘El gallego en la familia: entre la producción y la reproducción’, Caplletra, 53: 167-191. Regueira, X. L. (2006) ‘Política y lengua en Galicia: la ‘normalización’ de la lengua gallega’, in Castillo Lluch M. and Kabatek J. (eds.) Las lenguas de España. Política lingüística, sociología del lenguaje e ideología desde la Transición hasta la actualidad, Madrid / Frankfurt am Main: Iberoamericana / Vervuert, pp. 61-93. Rei-Doval, G. (2007) A lingua galega na cidade no século XX: unha aproximación sociolingüística, Vigo: Xerais. Ricento, T. (2015) ‘Foreword’, in Hult, F. M. and Cassels-Johnson, D. (eds.) Research Methods in Language Policy and Planning: A Practical Guide, West Sussex, UK: Wiley Blackwell. Schwartz, M. (2010) ‘Family language policy: Core issues of an emerging field’, Applied Linguistics Review 1 (1): 171-192. Schwartz, M. and Verschik, A. (eds.) (2013) Successful Family Language Policy: Parents, Children and Educators in Interaction, New York: Springer. Shankar, S. and Cavanaugh J. (2012) Language and materiality in global capitalism, Annual Review of Anthropology, 41: 355–369. Spolsky, B. (2009) Language management, Cambridge: Cambridge University Press. Tollefson, J. W. (Ed.) (2013) Language policies in education: critical Issues, Second edition, Oxon: Routledge. Tzanakis, M. (2011) ‘Bourdieu’s Social Reproduction Thesis and the Role of Cultural Capital in Educational Attainment: A Critical Review of Key Empirical Studies’, Educate 11 (1), 76-90. Urciuoli, B. and LaDousa, C. (2013) ‘Language Management/Labor’, Annual Review of Anthropology, 42, 175–190, available doi: 10.1146/ annurev-anthro-092412-155524 [accessed 31/03/16].

CHAPTER THREE PRACTITIONER ENGAGEMENT WITH RESEARCH IN MINORITY LANGUAGE CONTEXTS: EVIDENCE FROM RESEARCH REGARDING GAELIC MEDIUM EDUCATION SARAH MACQUARRIE AND FIONA LYON

Abstract Following the formal inception in 1985 Gaelic medium education is now more widely available across Scotland. While it is clear that provision being offered in 60 schools is a considerable achievement, it does mean that research projects have had a finite number of teachers and establishments to work with. This chapter presents an overview of relevant research that includes Gaelic medium education teachers who volunteered to be involved. Given the well-known difficulties of conducting research in applied settings (such as education) this suggests Gaelic medium teachers are particularly supportive of research and this ought to be recognised as a specific characteristic of Gaelic medium education when it is viewed amongst other minority languages.

Introduction The last 30 years has seen a transformation in the provision of Gaelic medium education (GME) in Scotland. GME uses an immersion model so that competency in Gaelic and English is achieved by the end of primary school. In the 30 years since GME was made available, a steady increase is evident in the number of children enrolled; between 2010-2014 the number rose by 35% (Scottish Government, 2014). It is an exciting and challenging period when an increase in the number of Scottish Gaelic speakers is fundamental to the development of the Gaelic language in

Practitioner Engagement with Research in Minority Language Contexts

35

Scotland in the attempt to arrest decline in the number of Gaelic speakers in Scotland -currently 57,600 reported in the 2011 census (National records of Scotland, 2015). The combination of educating children in the language and ensuring a robust support structure is vital so that Gaelic, despite being labelled as a minority language1, is regarded as relevant to the majority (Garcia, 2010) and considered important enough to survive and strengthen its position in Scotland.

The context of immersion education in the UK When compared with other minority languages in the United Kingdom, such as Irish and Welsh, it would appear that Gaelic has not enjoyed a high status. It was only in 2005, with the introduction of the Gaelic Language Act, that formal recognition was given to the Gaelic language. In contrast, in the Republic of Ireland, Irish was recognised as the main official language in 1937 (Baker, 1997) and bilingual education was decreed by law in 1922. Welsh medium primary education was formalised in 1988, offered as a core subject in the National Curriculum and in 1990 became a compulsory subject for all pupils in Wales from ages 5 to 14. The commitment and support regarding the development of the language is evident through the establishment of Welsh medium schools in areas where Welsh is less strong that led to in excess of 50,000 pupils being enrolled in such provision within 10 years (Welsh Government, 2015). Since the formal inception in 1985, Gaelic medium education (GME) has become more widely available across Scotland. Although provision available in 60 schools is a considerable achievement, in no way is such progress commensurate with Irish and Welsh in the UK (McLeod, 2008). Variation is evident in the census figures regarding those with Gaelic language capacity (National records of Scotland, 2015) and the patterns of Gaelic use by children inside and away from classrooms (O'Hanlon, Paterson, & McLeod, 2012). Given this variation not every child will have Gaelic at home and this has been identified as a challenge for practitioners (Stephen, McPake, McLeod, Pollock, & Carroll, 2010). The recent pupil enrolment figures make it clear that Gaelic medium education accounts for 2.9% of national provision in Scotland (Bòrd na Gàidhlig, 2015). Thus, while there has been growth in Gaelic medium education since the mid1980s, this does mean that research projects have had, and will continue to have, a finite number of establishments and teachers to work with. This 1

UK ratified the European charter for regional and minority languages in 2001, including Scottish Gaelic in the list of languages.

36

Chapter Three

chapter will offer examples of studies that were successful because of the enthusiasm and engagement of teachers who volunteered their time to be involved.

Gaelic Medium Education and teacher engagement Our focus is on relevant research that considered GME provision offered within compulsory education. In Scotland, the Standard for Initial Teacher Education (General Teaching Council Scotland, 2006, p. 3) defines a core professional interest and requirement for teachers in Scotland as “engaging with current educational issues and contributing to the processes of curriculum research and development, staff development and school development”. Despite this, Menter, Elliot, Hulme, Lewin, and Lowden (2011) point out that there is no legal obligation for teachers to engage with researchers, whether to complete a questionnaire or undertake an interview. Within education, research(ers) rely on teachers; teachers however do not depend on research, rather they are encouraged to engage with it. Teacher engagement played an important part in the development of GME. Indeed, the first attempt at Gaelic bilingual education began in 1975 in the Western Isles where the Bilingual Education Project tried to build on the home language of Gaelic-speaking pupils but used both Gaelic and English as the medium of instruction. Murray and Morrison (1984), the project directors, agreed to conduct a three year pilot programme of bilingual education, developing a Gaelic-English curriculum in a few primary schools in Gaelic-speaking areas. Twenty schools were involved initially amounting to some 1,100 pupils, or 20% of the total primary roll of the Western Isles. Initial impressions of the project team were that there was “a bewildering mixture of uncertainty, contradictory attitudes and practice among teachers that affects profoundly the use of Gaelic in schools” (Murray & Morrison, 1984, p. 10). There were doubts as to how fluent pupils were becoming following this model of bilingual education, and thus, educationalists, language activists and parents realised that a different approach was necessary (Mitchell, 1992; Robertson, 1999). This was an example of how the term ‘bilingual education’ did not in fact fully deliver fluency in Gaelic. Some teachers doubted their own ability to teach in the bilingual medium and were unsure of the project. A similar scheme was attempted in 1978 in five schools in Skye but although its aim was for children to access a school curriculum in Gaelic, as well as in English, it did not appear to produce fluency in Gaelic among the pupils (Robertson, 1999). The bilingual model, in a largely

Practitioner Engagement with Research in Minority Language Contexts

37

Anglophone environment, did not appear to work. They took their inspiration from Welsh medium education that originated in the 1940s. It was perhaps apt that Thompson (1976, p. iii) claimed “this is a good time for fresh analysis of Gaelic’s role in the life of Scotland”. The 1970s was a challenging time; attempts at bilingual education were not always as successful as was intended and evaluation proved challenging (Mitchell, 1992) yet this did not diminish practitioner engagement as a resulting activity was the production and publication of several books and resources by teachers who were willing to help prepare materials. The introduction of Comhairle nan Sgoiltean Araich (CNSA) in 1982, the Gaelic Pre-School Playgroup Association, was significant in the advancement of GME. Gaelic-speaking parents ran the playgroups that combined with parental pressure, led to demands for Gaelic medium primary education. The first Gaelic medium state-funded nursery class was established in Inverness in 1988 and the rapid growth of GME was such that, by 2010, there were 60 primary schools, involving 2256 pupils (Bòrd na Gàidhlig, 2010), 58 nursery providers and some 115 pre-school groups, showing that CNSA was a critical influence in the formation of GME provision. Gaelic medium education was first made available in 1985 in two Scottish primary schools: Glasgow and Inverness. It was perhaps no surprise that such provision was initiated on the mainland and not in the Western Isles (a region assumed to have strong links with the Gaelic language and is commonly referred to as the Gaelic “heartland”). Interestingly, the first Irish medium school of recent times was established in Belfast, separate to the main regions linked with the Irish language, by a small number of Irish-speaking families who wanted to raise their children through Irish medium education. In Scotland, numbers of pupils were small initially (for example 24 pupils in 1985) but increased steadily as more people became aware of the existence of GME.

Gaelic medium education teaching staff As GME is still relatively new, resources are scarce; teachers do not have as large a number or variety of resources as English medium teachers. This necessitates more teacher time sourcing and developing teaching materials that limits the time available to participate in research projects. This is a theme that regularly features in teachers’ reports in GME related research (Lyon & MacQuarrie, 2014; MacQuarrie & Lyon, 2015). A further limitation for researchers in GME is the small workforce; there are 193 teachers in 60 GME primary school establishments (Bòrd na Gàidhlig,

38

Chapter Three

2015). Of these, many are relatively new to the profession; recruitment of teachers is an on-going problem but steps are being taken to address this by Bòrd na Gàidhlig (an executive non-departmental public body, responsible to the Scottish Parliament). For example, Gaelic speaking teachers are encouraged to transfer to Gaelic medium education; immersion courses are available on a full and part-time basis; a 12 month on-line course is available for existing Gaelic speaking teachers. GME attracts international interest from research projects involving minority language studies and given the small population, GME teachers can be inundated with requests to participate in research studies. Strict regulations and guidance about the conduct of research is an important element of ethical practice. Universities or agencies involved in research must ensure that rigorous ethical scrutiny is applied to the design of research projects. Teachers should have assurance in the confidentiality of their responses as practitioners, but they will also be aware of their role and responsibility for the protection of pupils and parents linked to the research taking place. In Scotland, local government is organised through 32 local authorities2 that act as gate-keeper and must be approached in all cases of research involving schools. It is at this point that there is an opportunity to limit the number of requests for GME teacher involvement in research.

Findings of the studies Several studies have been successful because of the enthusiasm and engagement of teachers who volunteered their time to be involved. GME has at its heart a minority language that has influenced teachers’ eagerness to participate in a variety of projects despite the extra work involved. O'Hanlon (2015) investigated factors influencing parental choice of Gaelic and Welsh medium education. She distributed a bilingual (Gaelic-English) questionnaire to all 62 Gaelic medium primary education providers in Scotland and achieved a 55% response rate – with 34 schools responding from 11 of the 14 local authority areas which provide Gaelic medium education. A Welsh–English version of the questionnaire was distributed to 62 of the 466 Welsh medium primary education providers in existence in 2007 (Welsh Assembly Government, 2007, p. 64). Schools were sampled from eight local authorities representing a range of community 2

These local authorities are distributed across Scotland serving rural and urban populations. The locations of schools offering GME provision in these areas tend to be singular and set apart from other GME establishments.

Practitioner Engagement with Research in Minority Language Contexts

39

language contexts. The Welsh survey received a 26% response rate – with responses from 16 schools from 6 local authority areas. The more than double response rate from GME schools is an indicator of Gaelic primary school teachers’ enthusiasm, when researchers make requests of their time and knowledge. GME teachers are willing to engage with researchers where they believe that the study and their contribution can make a difference to the curriculum and to pupils’ attainment and the remainder of this chapter looks at specific examples of such research.

Resource availability and assessment There is an abundance of curricular materials available in English, some of which can be successfully translated or adapted for use in GME. However, literacy materials are particularly vulnerable and likely to suffer from translation (unlike materials aimed at numeracy). Pollock (2007) examined the challenges in teaching literacy in 6 GME classrooms, involving 35 initial questionnaires (43% response rate), 18 follow-up questionnaires (86% response rate) and 15 interviews with teachers. Her research came at a time when in-service teacher training specifically aimed at GME was scarce. Pollock reports that there was little practical guidance; teachers were working in isolation and there was little consistency in the support and provision of resources and reading materials across local authorities. Indeed, many schools were still using ‘paste-overs’ (text within English language books and resources that had been translated into Gaelic and placed on top of the original text giving a sequential reading scheme). This finding was echoed in a larger scale Scottish Government review of Gaelic medium education (Stephen et al., 2010), that noted the additional effort, time and support required on the part of teachers to support the availability of appropriate resources. To some extent, as well as the lack of resources, the isolated geographical situation of GME schools can provide some uncertainty and a lack of confidence as was found in Lyon (2003) when it came to gauging attainment. In this study, 45 schools providing both Gaelic and English medium education were approached. 20 schools responded (offering a response regarding Gaelic, English or both forms of provision) leading to: 18 replies from Gaelic teachers, and 17 replies from English teachers. English medium teachers reported being more confident when identifying pupils with reading difficulties and were more likely to recognise literacy related weaknesses earlier than their Gaelic medium counterparts. English medium schools relied on area support specialists and Support for Learning staff who administered commercially produced

40

Chapter Three

assessments, whereas GME staff identified pupils who were experiencing problems through normal class work. Gaelic medium teachers do not often have the services of a Support for Learning teacher to carry out such assessments nor are specific assessment resources available in Gaelic. Given the situation noted by Lyon (2003), it would appear that little progress has been achieved in the intervening years, as evident in an audit by MacLullich (2014)3. This double disparity continues in that GME has few Support for learning staff with the knowledge and language capabilities to undertake specialised assessment, even if such resources were available. This is a significant issue for GME that needs to be urgently addressed. A difficulty noted by teachers in the early immersion stage in Gaelic medium education (Lyon, 2003) was the assessment of phonological skills and early reading ability within a class of children with different cultural and linguistic backgrounds. This led to further research by Lyon (2012) into the role and assessment of phonological awareness in GME that could not have taken place without the support and co-operation of GME class teachers. In the 2012 study, pupils completed an assessment of phonological awareness in Gaelic created solely for this purpose. Following the pupil testing that teachers completed individually with pupils, teachers also participated in questionnaires and interviews. At that time, 11 of the then 14 local authorities offering GME participated (86% response rate) and 48 schools took part (80% response rate). 368 pupils were involved in the data collection, in effect 60% of the population in Primaries 2 and 3. Some 91% of teachers returned questionnaires and it had been hoped that 10% of the respondents would volunteer to be interviewed but, in fact, 40% of the respondents approached the researcher to be interviewed indicating that they valued the impact that the research had on their teaching and pupils’ attainment. Many teachers demonstrated limited knowledge of phonological awareness and its terminology in relation to early reading development and phonics, and regarded the study as beneficial in terms of their own continuing professional learning. Here, practitioner engagement was noted not just because of positive attitude or self-interest in developing the field in which teachers work, but such involvement can have a wider impact, contributing to the development of knowledge within the population taking part in studies. Perhaps such opportunities have particular value for GME teachers given the scarcity of little specific in-service training 3

The audit covers a wide range of themes pertaining to ASN, GME and teacher education. The full audit can be accessed at: http://www.gaidhlig.org.uk/bord/en/our-work/gaelic-education/additional-supportneeds/

Practitioner Engagement with Research in Minority Language Contexts

41

delivered for GME (Pollock, 2007). The phonological awareness test that was developed as a result of the research was subsequently published by GL-Assessment, and is widely used in GME classrooms today. Teacher engagement continues with the resource as it is available and used in GME settings across Scotland. Earlier research conducted by Müller (2007) examined language use and writing ability with the cooperation of teaching staff in Skye, indicating that pupils’ writing skills in Gaelic were influenced by their knowledge of the English language. Here teaching staff benefitted from this research as the outcome indicated that schools require to improve the writing ability of pupils in Gaelic in order to ensure the survival of the language. A workshop and seminar held at a conference in 2013 indicated the interest in holding discussions with GME teachers (Lyon & MacQuarrie, 2013) and other research conducted on a national scale demonstrated teachers’ eagerness to share their experiences in order to support the development of materials being readied for use in their classrooms (Lyon & MacQuarrie, 2014). A group discussion was held at the annual conference for GME teachers, An t-Alltan, in October 2014 (MacQuarrie & Lyon, 2015). At this event teachers come together from a variety of authorities offering the researchers insight into different practice across Scotland. Participants commented on the strengths of adaptable resources that have particular value in composite classes where levels of Gaelic language ability may vary. The influence of the home context and family to the development of literacy was also emphasised. Participants made comparisons with the breadth of resources when pupils focused on the English language and noted the scarcity of books (particularly fiction), in Gaelic. This was considered a particular challenge as it narrows the links possible between work completed in school and reading that takes place outside of school. Participants’ description of avenues they would employ to encourage engagement with families revealed that participants accessed different resources to manage the school-home link. Suggestions for resources teachers felt would add value included resources being made available in forms such as e-books, or graphic novels, or in the form of software such as apps. At present, there has been little change in the availability of such resources since the review of GME by Stephen et al. (2010). Participants noted the widespread availability of such technology at home, indicating that e-books could be a suitable avenue to explore and would be of value to link school and home experiences.

42

Chapter Three

Career long professional learning Training is also an area that fosters practitioner engagement. In 2002 in Nairn, the first Gaelic Teachers’ Conference, A’ Chuisle, was arranged following teacher demand and a further two events (2004 and 2005) were held, each supported by a number of stakeholders that included the Scottish Government and Her Majesty's Inspectorate of Education. These events gave an opportunity for GME teachers to come together and share their ideas and experiences, and are reflected in a report that aimed to address aspects of the provision of learning support in Gaelic medium education and support achievement in GME (Her Majesty’s Inspectorate of Education, 2005, p. 15). The first A’ Chuisle conference provided the initial impetus for the establishment of an on-line website for Gaelic teachers – Gàidhlig airloidhne – that led to increased confidence for teachers working in often isolated locations. In the intervening years, when Pollock (2007) conducted her doctoral research, there was little formal professional development activity that targeted the specific needs of Gaelic medium teachers. Only in 2008 were practitioners offered the opportunity to engage at An t-Alltan, the Gaelic teachers’ conference, that has been made available as an annual event by Stòrlann, a limited company established by the Scottish Executive in 1999 to co-ordinate the production and distribution of curriculum resources for Gaelic education. At this annual event, teachers from primary and secondary schools throughout Scotland are introduced to new resources, developments in learning and teaching and policy updates. The fact that each school automatically receives a sponsored place for a Gaelic medium teacher to attend is evidence of the on-going commitment to support practitioners. There is recognition and appreciation of training opportunities, including in-service training at the annual conference for Gaelic medium education teachers (MacQuarrie and Lyon, 2015, MacLullich, 2014). A theme visible in recent years at An t-Alltan is that practitioners lead workshops sharing resources they have developed. However, not all teachers feel comfortable in doing so; some practitioners, while eager to develop resources, noted barriers. The availability of time was a concern but more pressing was their reluctance to share resources they had developed through concerns of inaccuracies in the material (MacQuarrie & Lyon, 2015). One avenue to counter this barrier was warmly received; teachers would welcome the opportunity to seek peer review prior to sharing a resource. This would help distribute resources developed by teachers in response to their classroom needs but allow a form of quality

Practitioner Engagement with Research in Minority Language Contexts

43

assurance, offering support with aspects such as grammar and spelling that have been reported as barriers. Since the completion of the study, such a mechanism was implemented in May 2015 on Storlann’s website (National Gaelic Resource Agency). Teachers are often supportive of research if they perceive it has value, leading to a benefit for them or the pupils in their care. This is the case particularly where teachers have pupils with additional support needs (ASN). MacLeod and MacLeod (2001) evaluated learning support in four Gaelic medium schools in the Highland region (the then largest authority offering GME). They looked at the areas giving rise to ASN, the background of these needs and offered suggestions as to how these needs could be addressed in GME. Macleod and Macleod identified the urgency for assessment and diagnostic tools developed specifically for Gaelic as well as the need for resources to support differentiated teaching. Practitioners were eager to support the evaluation as they hoped that it might address the dearth of materials in GME. Although the demand for such parity is still evident, it took a further 13 years for a national level audit of ASN to be conducted (Lyon & MacQuarrie, 2014). Despite the gap, practitioners were keen to participate and were involved in large numbers in the study commissioned by Bòrd na Gàidhlig (MacLullich, 2014). The aim of the study was to ascertain the nature and extent of ASN in GME. The conclusions and recommendations in the report were derived from the results of an online survey completed by 110 staff in 33 schools. Additional questionnaires were sent to education officers, educational psychology teams and speech and language therapy services and a total of 45 interviews were conducted. An event to launch the report was held in Edinburgh4 bringing together both teaching and health professionals, to raise awareness of the assessment of language and communication in GME. Work is in progress so that contributions from the event are brought together to form an online repository or handbook that evidences the continuing impact this study has had and will have. Again teachers perceived that their participation could improve outcomes for their pupils.

4

Lyon. F. and MacQuarrie, S (2014) Developing phonological and reading tests for children within Gaelic medium education. Additional Support needs in Gaelic Medium education Conference. University of Edinburgh, Scotland. http://www.storlann.co.uk/asn-slt/

44

Chapter Three

Future directions Looking ahead there are a number of opportunities where practitioner engagement would be particularly advantageous. There is a range of research evidence indicating that supporting early readers by offering a variety of resources, so that children can read around a topic that interests and engages them, has substantial benefits in the short- and longer-term (Topping, 2015). This may have particular value within immersion models of education (such as GME) where a breadth of resources allows for the variability in reading expected in children who are becoming bilingual (Kovelman, Salah-Ud-Din, Berens, & Petitto, 2015). Given the developments in policy in Scotland, where the 1+2 language learning approach is being established (Education Scotland, 2015)5, the value of a broader range of materials available in Gaelic will be valuable for GME but will also benefit Gaelic learners across Scotland, including those learning Gaelic within the 1+2 approach. The use of technology in particular seems opportune and if sensitively managed (Parish Morris, Mahajan, Hirsh Pasek, Golinkoff, & Collins, 2013) may positively support literacy and reading skills. Technology would also be helpful for interactions that take place between children and adults as well as between children with their peers in class-time or at home with siblings/friends (Wood, Pillinger, & Jackson, 2010), addressing areas that practitioners have indicated need to be targeted. As a minority language there is variation in the age distribution of Gaelic speakers with the number of older speakers declining but a notable rise in younger speakers (most likely as a result of GME) so it is important to have evidence from such research to inform future planning and methodologies in schools. Such evidence would not be available without the direct participation of teachers, who recognise they are an increasing, albeit small, population and this knowledge encourages rather than diminishes engagement and cooperation with research activity. Given the well-known difficulties of conducting research in applied settings (such as education) evidence brought together in this chapter suggests Gaelic medium teachers are particularly supportive of research. The willingness of teachers to engage with research has had a central role in project outcomes successfully being achieved and a mutual beneficial relationship is evident between teachers and researchers. The eagerness amongst GME teachers to engage with research and support the 5

This policy is aimed at ensuring that every child has the opportunity to learn a modern language from P1 onwards. Additionally, each child should have the right to learn a second modern language from P5 onwards

Practitioner Engagement with Research in Minority Language Contexts

45

development and implementation of content, assessments and materials derived in Gaelic for use in GME is noteworthy and ought to be recognised as a specific characteristic of Gaelic medium education when it is viewed amongst other minority languages.

References Baker, C. (1997). Bilingual education in Ireland, Scotland and Wales. In J. Cummins & D. Corson (Eds.), Encyclopaedia of language and education, Vol 5: Bilingual Education. Netherlands: Kluwer Academic Publishers. Bòrd na Gàidhlig. (2015). Education Data Retrieved 30 September, 2015, from http://www.gaidhlig.org.uk/bord/en/our-work/research/educationdata/ Education Scotland. (2015, February 2015). Language learning in Scotland. A 1+2 approach. Further guidance on L3 within the 1+2 policy., from http://www.educationscotland.gov.uk/Images/L3SupplementaryGuidan ce160215_tcm4-852848.pdf Garcia, O. (2010). Minority Language Education. In B. M (Ed.), Concise encyclopedia of applied linguistics: Elsevier. General Teaching Council Scotland. (2006). Standard for Initial Teacher Education. Edinburgh: GTC Scotland. Her Majesty’s Inspectorate of Education. (2005). Improving achievement in Gaelic. Livingston: Her Majesty’s Inspectorate of Education. Kovelman, I., Salah-Ud-Din, M., Berens, M. S., & Petitto, L.-A. (2015). “One glove does not fit all” in bilingual reading acquisition: Using the age of first bilingual language exposure to understand optimal contexts for reading success. Cogent Education, 2(1), 1006504. doi: 10.1080/2331186X.2015.1006504 Lyon, F. (2003). Reading delay in Gaelic immersion education. Unpublished MSc Thesis, University of Strathclyde, Glasgow. —. (2012). Early Intervention in Gaelic medium Education: Creating and Evaluating a Tool for Assessing Phonological Awareness. Unpublished PhD thesis, University of Strathclyde, Glasgow. Lyon, F., & MacQuarrie, S. (2013). Support for children with additional support needs. Paper presented at the An t-Alltan Conference, Aviemore, Scotland. http://storlann.co.uk/an-t-alltan/2013-gaidhlig. html

46

Chapter Three

Lyon, F., & MacQuarrie, S. (2014). Assessment of reading skills in Gaelic medium education: Exploring teachers’ perceptions and present practice. Educational & Child Psychology, 31(2). MacLeod, D. M., & MacLeod, D. J. (2001). Learning support in Gaelicmedium primary schools in Highland. Inverness: Highland Council. MacLullich, A. (2014). Audit on additional support needs in Gaelic medium education and staff training needs analysis. Inverness: Bòrd na Gàidhlig. MacQuarrie, S., & Lyon, F. (2015). Seeking teachers’ views regarding literacy resources in Gaelic medium education. Final report: Soillse research network. McLeod, W. (2008). An opportunity avoided? The European Charter and UK language policy. In P. Gwynedd & D. Robert (Eds.), The European Charter for Regional or Minority Languages: Legal Challenges and Opportunities. (pp. 201-218). Strasbourg: Council of Europe Publishing. Menter, I., Elliot, D., Hulme, M., Lewin, J., & Lowden, K. (2011). A guide to practitioner research in education. London: Sage. Mitchell, R. (1992). The "independent" evaluation of bilingual primary education: a narrative account. In C. J. Alderson & A. Beretta (Eds.), Evaluating Second Language Education. Cambridge Applied Linguistics. Cambridge: Cambridge University Press. Müller, M. (2007). Language use, language attitudes and Gaelic writing ability among secondary pupils in the Isle of Skye. In W. McLeod (Ed.), Revitalising Gaelic in Scotland. Edinburgh: Dunedin Academic Press. Murray, J., & Morrison, C. (1984). Bilingual primary education in the Western Isles. Stornoway: Acair. National records of Scotland. (2015). Scotland’s Census 2011: Gaelic report (part 1). Retrieved September 30, 2015, from http://www.scotlandscensus.gov.uk/news/gaelic-analytical-report-part-1 O'Hanlon, F. (2015). Choice of Scottish Gaelic-medium and Welshmedium education at the primary and secondary school stages: Parent and pupil perspectives. International Journal of Bilingual Education and Bilingualism, 18(2), 242-259. doi: 10.1080/13670050.2014.923374 O'Hanlon, F., Paterson, L., & McLeod, W. (2012). Language Models in Gaelic Medium Pre-School, Primary and Secondary Education. Inverness: Soillse. Parish-Morris, J., Mahajan, N., HirshǦPasek, K., Golinkoff, R. M., & Collins, M. F. (2013). Once upon a time: parent–child dialogue and

Practitioner Engagement with Research in Minority Language Contexts

47

storybook reading in the electronic era. Mind, Brain, and Education, 7(3), 200-211. Pollock, I. E. (2007). The acquisition of literacy in Gaelic-medium primary classrooms in Scotland. University of Edinburgh, Edinburgh. Robertson, B. (1999). Gaelic education. In T. G. K. Bryce & W. M. Humes (Eds.), Scottish education. Edinburgh: Edinburgh University Press. Scottish Government. (2014). Summary Statistics for Schools in Scotland A National Statistics Publication for Scotland, Statistical Bulletin. Education Series. No.5 2014 Edition. Stephen, C., McPake, J., McLeod, W., Pollock, I., & Carroll, T. (2010). Review of Gaelic medium early education and childcare. www.scotland.gov.uk/Publications/2010/06/22090128/14 Thompson, D. S. (1976). Gaelic in Scotland. Glasgow: Gairm. Topping, K. J. (2015). Fiction and Non-Fiction Reading and Comprehension in Preferred Books. Reading Psychology, 36(4), 1-38. Welsh Assembly Government. (2007). Welsh in Schools 2007. Cardiff: Welsh Assembly Government. Welsh Government. (2015, 23/07/2015). National Education Statistics. Retrieved 29/09/2015, 2015, from https://statswales.wales.gov.uk/v/8RN Wood, C., Pillinger, C., & Jackson, E. (2010). Understanding the nature and impact of young readers’ literacy interactions with talking books and during adult reading support. Computers & Education, 54(1), 190198.

CHAPTER FOUR EXTRALINGUISTIC FACTORS INFLUENCING THE PRONUNCIATION OF ENGLISH BY WELSH-ENGLISH BILINGUALS1 CHARLES WILSON AND MARGARET DEUCHAR

Abstract It is well known that minority languages are influenced by the majority languages with which they come into contact, but influence can also operate in the opposite direction. This is the case with the pronunciation of Welsh English, which to varying degrees is influenced by the phonological system of Welsh (cf. Wells, 1982). One might expect Welsh speakers to show particular influence from Welsh phonology in their English, but this is not always the case. Our research question was which extralinguistic factors have a significant influence on a Welsh-English bilingual’s choice of Standard English vs. Welsh-influenced pronunciation when speaking English. Data were collected from 41 participants in both north and south Wales by means of a sociolinguistic interview and a questionnaire to obtain extralinguistic information. A multivariate analysis of the linguistic and extralinguistic variables was performed using Rbrul (Johnson 2009) in order to identify the significance of the extralinguistic factors. The following factors were significant: age, gender, parents’ language, attitudes to English, region and style.

1

We gratefully acknowledge the AHRC, ESRC and Bangor University for their financial support of this study. We are also grateful for feedback received from participants at the Language Diversity in Wales conference, National Library of Wales, 18-19 July 2014, II Southern Englishes Workshop, University of Cambridge, 23-24 March 2015, and XV International Congress of Celtic Studies, University of Glasgow, 13-17 July 2015.

Pronunciation of English by Welsh-English Bilinguals

49

Introduction According to Penhallurick (2007), English was first spoken in Wales as early as the twelfth century. However, it did not become widely established in Wales until the nineteenth century, with the advent of compulsory education in English and the industrial revolution. Even then, it was not until about the 1960s that all adults in Wales were able to speak English. By the 1960s, monolingualism in Welsh had largely disappeared, while monolingualism in English had taken over and become characteristic of three quarters of the population (Jones and Williams 2009, 662). Nevertheless, 19% the population remains bilingual in Welsh and English today (Office for National Statistics, 2012). Their geographical distribution varies, and is shown in Figure 4-1. The English spoken in Wales incorporates a considerable amount of variation which shows influence from several competing influences: standard British English pronunciation, the phonology of English varieties spoken on Wales’s borders, and the phonology of the Welsh language. Thomas (1984) draws attention to influences from e.g. the west Midlands, the southwest and northwest England. Wells (1982) considers a Welsh accent in English to have two striking characteristics: Welsh intonation and Welsh vowel quality: in particular, Welsh English tends to have more monophthongs and fewer diphthongs than other accents of English. Wells (1982, 377): “There can be little doubt that the main influence on the pronunciation of English in Wales is the substratum represented by the phonological system of Welsh”. Paulasto (2013) provides a synthesis of what is known about the phonology of traditional Welsh dialects insofar as this differs from standard British English, drawing on a range of sources (e.g. Parry 1999, Penhallurick 2004, Walters 2003). In this chapter, we shall focus on some variations in the pronunciation of vowels as produced by bilingual speakers of Welsh and English, focusing on two vowels which are variably produced as diphthongs in Wales (as in FACE, GOAT) and two vowels which are variable in quality (as in BATH, STRUT). We shall be interested in determining what makes Welsh speakers choose (albeit not consciously) phonological variants which are either closer to traditional Welsh English phonology or to standard British English. In other words, our aim is to identify the significant extralinguistic factors in the choice by Welsh speakers of what we shall call a traditional Welsh English vs. standard British English phonological variant when speaking English.

50

Chapter Four

Fig. 4-1 Percentage of population (aged 3 and over) able to speak Welsh by local authority, 2011 © Crown copyright and database right 2012 (Office for National Statistics, 2012).

Phonological features studied Four vowels were selected as dependent linguistic variables, as shown in Table 4-1.

Pronunciation of English by Welsh-English Bilinguals

Phonological variable (Wells’s (1982) labels)

51

BATH

Standard British English Variant [‫ܤ‬:]

Traditional Welsh English Variant [a]

FACE

[eܼ]

[e‫]ޝ‬

GOAT

[ԥ‫]ݜ‬

[o‫]ޝ‬

STRUT

[‫]ݞ‬

[ԥ]

Table 4-1 Phonological variables studied

The BATH and STRUT vowels have in common that they are back vowels in standard British English (cf. Wells 1982, 281), whereas the Welsh variants are fronted, with the most common Welsh English variant of BATH being [a] according to Penhallurick (2007, 156). Both Wells (1982, 387) and Penhallurick (2004, 104) point out that a longer vowel [a:] is sometimes used in Welsh English, but for our purposes it is the frontness of the vowel which interests us as being the main characteristic of a Welsh English variant. The STRUT vowel is more central in traditional Welsh English than in standard British English, merging with schwa ([ԥ]) in stressed syllables. This can be accounted for by the fact that [‫ ]ݞ‬does not occur in Welsh (cf. Penhallurick 2007, 155), and that [ԥ] can appear in stressed syllables in Welsh, unlike in standard British English. In the case of FACE and GOAT, the distinctive characteristic of Welsh English is that monophthongs tend to be used instead of the diphthongs of standard British English. Penhallurick (2007, 157) points out that Welsh does not have the diphthongs [eܼ] and [ԥ‫ ]ݜ‬and suggests that the use of the monophthong variants in north Wales has also been influenced by their occurrence in varieties spoken in neighbouring northwest England. In a similar vein, he argues that the more frequent use of diphthongs in south Wales may have been influenced by the use of diphthongs in neighbouring southwest England (2004, 105). More details of the regional variation within Wales in the pronunciation of FACE and GOAT are provided by Penhallurick (2004).

Previous work on variation in English pronunciation in Wales Previous work on variation in English pronunciation in Wales has been conducted mainly within social psychology, and benefited considerably

52

Chapter Four

from the pioneering work of Wallace Lambert on the social psychology of bilingualism (cf. Lambert 1967). It has made extensive use of Lambert’s “matched guise” technique according to which participants are asked to evaluate two or more different accents which (unbeknown to them) are actually produced by the same speaker. The advantage of this approach is that it elicits covert rather than overt attitudes to language which should be uninfluenced by the respondent’s notion of what they should think or what they think is politically correct to say. The application of his approach to English in Wales led to several influential publications in the 1970s, e.g. Giles (1970, 1971), Bourhis, Giles & Tajfel (1973), and Bourhis & Giles (1976). In a review article on the social meaning of Welsh English Giles (1990, 275) reports on a complex picture “wherein WE [Welsh English] varieties are downgraded on competence traits generally but upgraded on those of social attractiveness”. For example, not only did Giles (1970) find that Received Pronunciation (RP)2 was more prestigious than Welsh English, but Giles (1971) found that the personalities attributed to RP speakers included superior ratings in terms of their assumed ambition, intelligence, self-confidence, determination, and industriousness. On the other hand, Welsh English guises were rated as superior in integrity and social attractiveness. Bourhis, Giles & Tajfel (1973) investigated the role of Welsh and Welsh English in reflecting Welsh identity. They found that Welsh speech was perceived by both bilinguals and English monolinguals more favourably than Welsh-accented English, which in turn was perceived more favourably than RP English speech “in terms of being less conservative, more patriotic, nationalistic, romantic, emotional, desirable as an immediate superior and as a model to be emulated” (Bourhis et al 1973, 456). On other traits3, such as friendliness and trustworthiness, there was no significant difference between the evaluation of Welsh and Welshaccented English, but both were evaluated more favourably than RP. Bourhis et al (1973, 458) interpret these results as suggesting that in addition to the Welsh language, “the Welsh accent can also serve as a marker of ethnic identity”. Giles (1990, 261) reports further on research 2 Giles (1970; 1971) and Bourhis et al (1973) use the term ‘RP’, but to avoid making any assumptions about the specific variety of English or Welsh that influences Welsh-English bilinguals to favour the competing variants in our study, we will use the terms ‘standard British English’ and ‘traditional Welsh English’. 3 The list as given in Table 5 (Bourhis et al 1973: 457) was “Understanding, Arrogant, Sociable, Sports-loving, Trustworthy, Self-confident, Snobbish, Friendly, Similarity to speaker, Desirable as colleague”.

Pronunciation of English by Welsh-English Bilinguals

53

which shows that in intergroup contexts, where an RP and a Welsh English speaker are in conversation (as indeed in our study) “maintaining or emphasizing one’s perceived nationality by means of accent….is a valued tactic asserting cultural identity”. This interpretation finds support in the results of a study by Bourhis & Giles (1976) which tested the relation between Cardiff theatre-goers’ compliance with a request to fill in a questionnaire and the language/variety (Welsh, Welsh English, or English with RP pronunciation) in which an announcement of the request was made. For bilinguals4 an announcement in Welsh led to the greatest degree of cooperation, but Welsh English led to a greater degree of cooperation than RP. Although the earliest and largest amount of work on social aspects of Welsh English appears to have come from social psychology, Coupland’s (1980; 1988) work has been an important contribution in the framework most relevant to this study, variationist sociolinguistics. Whereas the work in social psychology referred to “Welsh English” without examining much in detail what this variety actually was, Coupland, building on the descriptive work of Wells (1982) and Thomas (1984), investigated variation in Welsh English, particularly focusing on data from Cardiff, the capital. Coupland (1988) describes the phonological variables (ng), (h), (ai), (au), (ԥ‫( )ݜ‬r) and intervocalic (t). A study of 51 clients of a Cardiff travel agency showed that the realizations of (h), (ng) and intervocalic (t) varied with occupational class, educational background and gender of speaker. The same phonological variables were found to be related to the contextual style of the travel agent, where “casual”, “informal work-related”, “client” and “telephone styles” could be distinguished, progressing from least to most formal. Little further work on variation in Welsh English had been conducted until recently, when Morris (2013) reported on his study of variation in the Welsh and English speech of Welsh-English bilinguals aged 16-18 in two locations in north Wales. He focused on the variables (l) and (r) in both languages. Gender was found to be a significant extralinguistic variable in that tokens of (l) in English (the language of our focus here) were lighter (with lesser degrees of velarization) in female speech. Variation in (r) was found to be related to home language (Welsh vs. English) as well as to style (word list vs. interview) and area (northwest vs. northeast Wales). 4

The monolingual English speakers (who are less relevant to our study) heard announcements instead in RP, Welsh English with a broad accent and Welsh English with a mild accent. They were most cooperative with the mild Welsh English announcement and least cooperative with the broad Welsh English announcement.

54

Chapter Four

One extralinguistic variable which does not appear to have been extensively investigated in sociolinguistic studies of Welsh English pronunciation to date is that of age. This is an important factor for the study of language change, which may be inferred from age differences in language usage by means of the “apparent time construct” (Bailey 2002, 312). In this approach, as Tagliamonte (2012, 43) explains, “Age differences are assumed to be temporal analogues, reflecting historical changes in the progress of the change”. Indeed, Paulasto (2013, 257) shows her use of the apparent time construct in her study of Welsh English syntax, reporting from her comparison of speakers of different generations that there is evidence for dialect levelling over time. The use of apparent time in pronunciation is illustrated by Trudgill’s (1974) classic study of sociolinguistic variation in Norwich English, he finds that the younger the speaker, the more likely they are to produce a centralized variant of (e) before /l/ in words such as help. From this finding he infers, assuming apparent time, that change in the pronunciation of (e) is in progress (Trudgill 1974, 104–5). Our use of age as an extralinguistic variable should help to determine whether the pronunciation of English in Wales is changing along the lines reported by Paulasto for syntax.

Research question Our research question was the following: which extralinguistic factors have the greatest influence on the choice of standard British English vs. traditional Welsh English variants of the selected phonological variables? (cf. Table 4-1). That is to say, what makes speakers pronounce a standard British English or traditional Welsh English phonological variant?

Extralinguistic factors examined We chose to examine nine extralinguistic factors that we expected to be linked to phonological information on the basis of previous literature. The nine factors were as follows: national identity, attitudes to Welsh, attitudes to English, style, gender, home language, region of upbringing, social class, and age. Information on these factors was obtained by means of a questionnaire. 1) National identity: participants were asked to choose from the categories Welsh, English, Scottish, Irish, British and Other. Given the work reported above by Bourhis et al (1973) and Giles (1990)

Pronunciation of English by Welsh-English Bilinguals

2)

4)

5)

6)

7)

8)

9)

55

we expected self-identification as Welsh to favour the choice of traditional Welsh English variants. and 3) The same question elicited information on participants’ attitudes to the Welsh and English languages. Because of studies such as that by Bourhis & Giles (1976) outlined above, we expected positive attitudes to Welsh and/or negative attitudes to English to influence choice of the Welsh English variant. The importance of style in the variationist work by Coupland (1980, 1988) led us to expect a more formal style to enhance the likelihood of speakers choosing the standard British English variant (cf. also Trudgill 1974). Morris’s (2013) study led us to expect gender to be significant in the choice of variant, and for this to favour standard British English given other studies like Trudgill (1974) which demonstrated a greater tendency for women to approximate standard speech. We thus expected men to be more likely to choose traditional Welsh English variants than women. Morris (2013) also demonstrated the influential effect of home language, and we expected participants with Welsh as their home language to be more likely to choose traditional Welsh English variants in their English than participants with English or both Welsh and English as their home languages. Traditional dialectology, as well as sociolinguistic studies, have established regional background to be a major factor influencing speech. Because Welsh is more prevalent (see Figure 4-1) as a community language in the north (especially the northwest, where many of our participants were brought up: see Figure 4-2) than the south of Wales, we expected a northern upbringing to favour the choice of traditional Welsh English variants. Thomas (1984, 178) identifies a “northern” model of Welsh English which is more influenced by Welsh than the “southern” model. Furthermore, Morris (2013) has provided evidence that the pronunciation of Welsh in northwest Wales appears to influence the pronunciation of English in that region. As occupational class was significant in Coupland’s (1988) study (and mentioned by Penhallurick (2004, 104) in relation to the BATH vowel, we decided to investigate whether a higher occupational class would be related to the choice of standard British English variants. Regarding age, the speech of younger people may be interpreted as indicating the direction of change in pronunciation as outlined

Chapter Four

56

above. Given the extent to which bilingualism and thus exposure to English has increased during the twentieth century (see Jones 1993; Deuchar 2005) and what Coupland & Thomas (1990, 5) call “progressive Anglicisation” (cf. also Williams 1990) we expected older people to favour the more traditional Welsh English variants more than younger people.

Expected results In Table 4-2 we summarize the expected influence of each extralinguistic factor. Extralinguistic factor group

Factor expected to favour the traditional Welsh English variant

1) National identity

Self-reported Welsh identity

2) Attitudes to Welsh 3) Attitudes to English

Positive attitudes to Welsh, and/or Negative attitudes to English

4) Style

Informal

5) Gender

Male

6) Home language

Welsh5

7) Region of upbringing

Northern

8) Occupational class

Lower

9) Age

Older

Table 4-2 Summary of expected influence of factors

5

There were three variants of home language: Welsh, English and both languages. Other than age, which was categorized as a continuous factor, ‘home language’ was our only extralinguistic factor group that was not binary.

Pronunciation of English by Welsh-English Bilinguals

57

Methodology Data collection The data were collected by means of sociolinguistic interviews (see Labov 1972) with 41 bilingual participants, 20 from north Wales and 21 from south Wales. Figure 4-2 shows the locations where the speakers had been brought up, many of them still living in those places. All participants were over 18, 21 being male and 20 female. There was considerable variation in age, with ages ranging from 18 to 73. A table showing the participants together with their ages, gender and regional background is included in Appendix 1. All participants reported confidence in their ability to conduct extended conversations in Welsh and English6. All had learned Welsh either in early childhood or at primary school. The interviewer presented as a standard English speaker who did not speak Welsh. If he had been known to be Welsh-speaking, this would probably have elicited Welsh speech from the participants, whereas the focus of the study was their speech in English. The sociolinguistic interview lasted approximately 30 minutes, with about 25 minutes of casual conversation, and 5 minutes for the reading of a word list. The interview was recorded on the interviewer’s iPhone device, which has high quality recording technology. The casual conversation was guided by the interviewer asking questions about each speaker’s childhood and holidays, the aim being for participants to pay as little attention as possible to their speech. However, the word list (see Appendix 2) was designed to make participants pay more attention to their speech, and included words with the relevant phonological variables for our study. The questionnaire mentioned above was administered after the interview. This elicited general information about the speaker, including the extralinguistic factors which we expected to have an impact on the speakers’ choice of standard British English vs. Welsh English phonological variant.

6

Two participants reported confidence in conducting only basic conversations in English.

58

Chapter Four

Key Upbringing locations coded as ‘north Wales’ Upbringing locations coded as ‘south Wales’ Fig. 4-2 Location of participants’ home towns. Contains Ordnance Survey data © Crown Copyright and database copyright 2012 (created from 1: 1 000 000 scale digital data)

Pronunciation of English by Welsh-English Bilinguals

59

Data analysis After excluding the first five minutes of the recordings of informal speech, the first 507 tokens taken from words containing the phonological variables being studied, and the first 108 words from the word list containing the phonological variables being studied were transcribed. The phonological variables were labelled ‘traditional Welsh English’ or ‘standard British English’ by the first author on the basis of auditory perception supplemented by Praat analysis (Boersma and Weenink, 2014) and discussion with the second author. Identification of the first and second formants in Praat helped us to identify vowel height and frontness vs. backness, while movement in the formants helped to distinguish diphthongs from monophthongs. The coded variants (i.e. traditional Welsh English or standard British English phonological variants) and the extralinguistic factors (i.e. gender, region of upbringing) were coded on Excel spreadsheets, and a multivariate analysis was performed on the coded data using Rbrul (Johnson 2009). Speaker and linguistic variable categories were analysed as random effects. The output from Rbrul provides information about significant predictive factors through quantitative analysis of the relation between the dependent phonological variables and the independent extralinguistic factors.

Results Table 4-3 lists the significant factors from the Rbrul multivariate analysis. A p number below 0.01 shows that a factor is significant at this level, and the numbers in the “log odds” and “factor weight” columns allow us to see which factor favoured the production of the traditional Welsh English variant. In each factor group, the factor listed first is the one which favoured the choice of the traditional Welsh English variant. This factor can also be identified by the fact that it has a factor weight above 0.5 and log odds (a measure equivalent to factor weights) of above 1. The “tokens” 7

Some speakers did not produce as many as 50 tokens of the variables studied, in which case all tokens produced were used. In the case of BATH, none of the speakers produced 50 tokens. 8 Some speakers sped up their reading towards the end of the list as they became more comfortable with the exercise, and some speakers seemed to be uncomfortable reading English and therefore misread words, so it was decided that, in order to control for this, the first 10 tokens of correctly-identified words would capture sufficient data on careful speech (i.e. the style factor).

Chapter Four

60

column shows us how many traditional Welsh English phonological variants were numerically chosen altogether, with the “percentage of total” column showing the proportion of traditional Welsh English variants for each factor. Deviance Intercept df Grand Mean r2 FACTOR GROUP (in order of significance)

7454.224 -0.392 8 0.508 0.556 LOG ODDS

TOKENS

TRADITIONAL WELSH ENGLISH VARIANT % OF TOTAL

FACTOR WEIGHT

0.278 -0.278

6298 1637

52% 48%

0.52 0.48

HOME LANGUAGE Welsh 1.292 Mixed -0.326 English -0.966

4866 2517 552

61% 35% 28%

0.784 0.419 0.276

GENDER Male Female

4037 3898

57% 44%

0.652 0.348

REGION OF UPBRINGING North 0.582 4041 South -0.582 3894

62% 39%

0.641 0.359

STYLE Informal Formal

p

2L1 bilinguals’ >

206

Chapter Ten

Basque dominant L1B-L2Sp bilinguals), or lower in the Spanish of bilinguals with less exposure to that language, indicates that the advantage pattern itself is not affected by the amount of exposure to the language. This does not rule out, however, the influence of the amount of exposure on the production of target-like SRs and ORs, the mastering of which requires a high lexical, morphosyntactic and pragmatic competence (Gutierrez-Mangado and Ezeizabarrena, 2012, Munarriz et al., 2016). The variability in profiles reported throughout this study is only a sample of the diversity of bilingual profiles existing in the Basquespeaking community, which has experienced an increase in numbers during the last three decades (Eusko Jaurlaritza-Gobierno Vasco, 2013, Eustat, 2012). This context provides an interesting field of investigation for the factors that may affect the use of a minority language (Basque) in long-standing contact with the majority language (Spanish). In order to approach the study of the potential effects of the type of bilingual profile and the methodology on the elicited production of RCs in early Basque, we focused on the 4- to 7-year-olds’ stratum of this Spanish-Basque bilingual community, which has opted for an educational system which aims to ensure that every child achieves some competence in the use of the two official languages spoken in the community.

5. Conclusion Higher rates of accuracy for ORs as compared to SRs have been attested in adult processing on-line studies, as well as in off-line comprehension studies conducted with adult and child participants in Basque. Such resuls align Basque with the languages showing an OR-advantage, and consequently challenging the universality of the SR-advantage. However, the overview on the production studies conducted with different age and language profile groups of Basque-Spanish bilingual children (n=62, age range 4-6 years) revealed consistently higher rates of target-like SRs than ORs in the RC production in Basque across studies, samples and tasks, which is in line with the SR-advantage attested in many studies conducted in Spanish and other languages. Such apparently contradictory patterns suggest that the advantage, if any, may not be unidirectional or specific for this language, since directionality varies across participants with the same language profile in the same age group, but also across tasks, regardless of whether they were conducted with the same or different individuals. The language internal inconsistency attested in Basque, and observed also in languages like Chinese or Quechua, points towards a language external motivation. We

Production of Relative Clauses by Basque Speakers

207

conclude that participants’ processing skills (linked to age and language profile) play a very important role in the rates of target-like RC production, especially in non-ideal conditions of language use, such as in the case of individuals with developing or impaired grammar(s) or with low exposure to their language(s). However, the variation in the directionality of the advantage seems to be more related to methodological factors (experimental design, procedure), which suggests that the consistency of the SR-/OR-advantage in a language should be relativized and that labelling a language as SR- vs. OR-advantage type requires robust empirical support. Finally, we argue that the SR-advantage pattern found in the production of Basque RCs among 4- to 6-year-olds cannot be convincingly explained as a consequence of the interlinguistic effect of the majority language (Spanish) on the minority language (Basque) of these bilingual children.

Acknowledgements This research has been partially supported by the Basque Government (IT676-13), the University of the Basque Country (UFI11/6) and the Ministry of Economy and Competitiveness / FEDER (FFI2012-37884-C03-02, FFI2015-68589-C2-1-P). This study could not have been done without the collaboration of the following schools: Aranzadi (Bergara), Axular (San Sebastian), Haurtzaro (Oihartzun), Karmelo (Bilbao) , Uzturpe (Ibarra) and Paz de Ziganda (Atarrabia). Thank you very much for your help. The authors are also very grateful to the editors and the two anonymous reviewers for their helpful suggestions, as well as to the native speaker who proofread the text.

References Arosio, Fabrizio, Flavia Adani, and Maria Teresa Guasti. 2009. "Grammatical features in the comprehension of Italian relative clauses by children." In Merging Features: Computation, Interpretation, and Acquisition, edited by José M. Brucart, Anna Gavarró and Jaume Solà, 138-155. Oxford/New York: Oxford University Press. Artiagoitia, Xabier. 1992. "Why Basque doesn't relativize everything." International Journal of Basque Linguistics and Philology 27:11-35. Barreña, Andoni. 2000. "Estructuras subordinadas en niños monolingües y bilingües vasco-españoles." In Spanish Applied Linguistics at the Turn

208

Chapter Ten

of the Millenium, edited by R. Leow and C. Sanz, 157-173. Somerville: Cascadilla. —. 2013. "El efecto del ímput y de la edad al inicio del contacto con la lengua en la adquisición de los complementadores en euskara." VII International AEAL Conference, Bilbao, September, 2013. Belletti, Adriana, and Carla Contemori. 2012. "Subjects in children’s object relatives in Italian." Revue Roumaine de Linguistique / Romanian Review of Linguistics. RRL edited by Larissa Avram, 117142. Betancort, Moisés, Manuel Carreiras, and Patrick Sturt. 2009. "The processing of subject and object relative clauses in Spanish: An eyetracking study." Quarterly Journal of Experimental Psychology 62 (10):1915-1929. Brucart, José María. 1994. "El funcionamiento sintáctico de los relativos en español." In II Encuentro de lingüistas y filólogos de España y México: Salamanca 25-30 de noviembre de 1991, edited by B. Garza, J. A. Pascual and A. Alonso Gonzalez, 443-469. Salamanca: Universiadad de Salamanca. —. 1999. "La estructura del sintagma nominal: las oraciones de relativo." In Gramática descriptiva de la lengua española, edited by Ignacio Bosque and Violeta Demonte, 395-522. Madrid: Espasa-Calpe. Caramazza, Alfonso, and Edgar B. Zurif. 1976. "Dissociation of algorithmic and heuristic processes in language comprehension: Evidence from aphasia." Brain and Language 3 (4):572-582. Carreiras, Manuel, Jon Andoni Duñabeitia, Marta Vergara, Irene de la Cruz-Pavía, and Itziar Laka. 2010. "Subject relative clauses are not universally easier to process: Evidence from Basque." Cognition 115:57-66. de Rijk, Rudolf Pieter Gerardus. 1972. "Studies in Basque syntax: relative clauses." Department of Foreign Literatures and Linguistics, Massachusetts Institute of Technology. Eusko Jaurlaritza-Gobierno Vasco. 2013. Fifth Sociolinguistic Survey, 2011. Vitoria-Gasteiz: Servicio de publicaciones del Gobierno Vasco. Eustat – Basque Statistical Institute. 2012. http://en.eustat.eus/estadisticas/tema_458/opt_0/tipo_1/temas.html#axz z45E7lDnp8. Ezeizabarrena, Maria-José. 2012a. "Children do not substitute object relatives with subject relatives in every romance language: the case of Spanish." Revue roumaine de linguistique LVII (2):161-181. —. 2012b. "The (in)consistent ergative marking in early Basque: L1 vs. child L2." Lingua 122:303-317.

Production of Relative Clauses by Basque Speakers

209

Ezeizabarrena, Maria José, Amaia Munarriz, and Udane Loidi. in press. "Bilinguals’ production of relative clauses in languages with the opposite head-complement directionality." In Multidisciplinary approaches to bilingualism in the Hispanic and Lusophone world, edited by Kate Bellamy, Michael Child, Paz González, Antje Muntendam and M. Carmen Parafita Couto. Amsterdam: John Benjamins. Frazier, Lyn. 1987. "Syntactic Processing: Evidence from Dutch." Natural Language and Linguistic Theory 5:519-559. Friedmann, Naama, Adriana Belleti, Laurice Tuller, Joao Costa, Maria Teresa Guasti, M. Lobo, E. Hamann, A. Gavarró, A. van Hout, G. Hakansson, Maria José Ezeizabarrena, K. Jensen de López, H. K. J. Van der Lely, L. Avram, A. Sevcenco, J. Weston, I. Fattal, M. Yachini, H. Delage, A. Cunill, M. Muntané, M. Reguant, Y. Yatsushiro, U. Sauerland, J. Kainhofer, P. Hummer, H. Haider, A. Pych, A. Miekisz, C. Contemori, K. Grohmann, and Spyridoula Varlokosta. 2010. "The child who prefers not to produce object relatives: The acquisition of relative clause production in 16 languages." COST A33, London, January 2010. Friedmann, Naama, Adriana Belletti, and Luigi Rizzi. 2009. "Relativized relatives: Types of intervention in the acquisition of A-bar dependencies." Lingua 119 (1):67-88. Friedmann, Naama, and Rama Novogrodsky. 2004. "The acquisition of relative clause comprehension in Hebrew: a study of SLI and normal development." Journal of Child Language 31 (3):661-681. doi: 10.1017/S0305000904006269. Friedmann, Naama, Julia Reznick, Dina Dolinski-Nuger, and Katya Soboleva. 2010. "Comprehension and production of movement-derived sentences by Russian speakers with agrammatic aphasia." Journal of Neurolinguistics 23 (1):44-65. Friedmann, Naama, and Ronit Szterman. 2006. "Syntactic movement in orally-trained children with hearing impairment." Journal of Deaf Studies and Deaf Education 11:56-75. Gavarró, Ana, Arnau Cunill, Miriam Muntané, and Marc Reguant. 2012. "The acquisition of Catalan relatives: structure and processing." Revue Roumaine de Linguistique / Romanian Review of Linguistics. RRL:183–201. Grillo, N. 2009. "Generalized Minimality: Feature impoverishment and comprehension deficits in agrammatism." Lingua 119:1426-1443. Gutierrez-Mangado, Maria Juncal. 2011. "Children's comprehension of relative clauses in an ergative language: the case of Basque." Language Acquisition: A Journal of Developmental Linguistics 18 (3):176-201.

210

Chapter Ten

—. 2013. "Ergative case in the production and comprehension of relative clauses in Basque." International Symposium of Psycholinguistics, Tenerife, March 20-23. Gutierrez-Mangado, Maria Juncal, and Maria-José Ezeizabarrena. 2012. "Asymmetry in child comprehension and production of Basque subject and object relative clauses." BUCLD 36 Online Proceedings Supplement:http://www.bu.edu/bucld/proceedings/supplement/vol36/. Hu, Shenai, Ana Gavarró, Mirta Vernice, and Maria Teresa Guasti. 2016. "The acquisition of Chinese relative clauses: contrasting two theoretical approaches." Journal of Child Language 43 (1):1-21. Kidd, Evan. 2011. The Acquisition of Relative Clauses. Amsterdam/ Philadelphia: John Benjamins. Kidd, Evan, Angel Chan, and Joie Chiu. 2015. "Cross-linguistic influence in simultaneous Cantonese-English bilingual children's comprehension of relative clauses." Bilingualism: Language and Cognition 18 (03):438-452. doi: doi:10.1017/S1366728914000649. Laka, Itziar, Kepa Erdocia, Jon A. Duñabeitia, Nicola Molinaro, and Manuel Carreiras. 2011. "Complex syntactic processing in very proficient non-natives elicits N400." Cognitive Neuroscience Society Meeting CNS, San Francisco, 2011. Lukatela, Katarina, Donald Shankweiler, and Stephen Crain. 1995. "Syntactic processing in agrammatic aphasia by speakers of a Slavic language." Brain and Language 49 (1):50-76. Miyamoto, Edson T., and Michiko Nakamura. 2003. "Subject/Object Asymmetries in the Processing of Relative Clauses in Japanese." In Proceedings of WCCFL, edited by Gina Garding and Mimu Tsujimura, 342–355. Somerville,MA: Cascadilla Press. Müller, N., and A. Hulk. 2001. "Crosslinguistic influence in bilingual language acquisition: Italian and French as recipient languages." Bilingualism: Language and Cognition 4 (1):1-21. Munarriz, Amaia. 2008. Afasiko elebidun baten hizkuntzaren azterketa [“Language assessment in a bilingual with aphasia”]. Master’s thesis. Vitoria/Gasteiz, Spain. —. 2015. Hizkuntzen antolaketa burmuin elebidunean: gaztelania-euskara elebidun afasiko baten kasu-azterketa [“Language organization in the bilingual brain: a case study on Spanish-Basque bilingual aphasia”]. UPV/EHU. Munarriz, Amaia, Maria-José Ezeizabarrena, and Maria Juncal GutierrezMangado. 2016. "Differential and selective morpho-syntactic impairment in Spanish-Basque bilingual aphasia." Bilingualism, Language and Cognition 19:810-833.

Production of Relative Clauses by Basque Speakers

211

Novogrodsky, Rama, and Naama Friedmann. 2006. "The production of relative clauses in syntactic SLI: A window to the nature of the impairment." Advances in Speech-Language Pathology 8 (4):364-375. O'Grady, William. 2011. "Relative clauses. Processing and acquisition." In The Acquisition of Relative Clauses, edited by Evan Kidd. Amsterdam/Philadelphia: John Benjamins. O'Grady, William, Miseon Lee, and M. Choo. 2003. "A subject-object asymmetry in the acquisition of relative clauses in Korean as a second language." Studies in Second Language Acquisition 25:433-448. Rojas Nieto, Cecilia. 2009. "Starting small effects in the acquisition of early relative constructions in Spanish." In Syntactic Complexity: Diachrony, acquisition, neuro-cognition, evolution, edited by T. Givón and Masayoshi Shibatani, 277-310. Amsterdam/Philadelphia: John Benjamins. Sevcenco, Anca, and Larissa Avram. 2012. "Romanian-speaking children’s comprehension of relatives." In Revue Roumaine de Linguistique / Romanian Review of Linguistics. RRL, edited by Larissa Avram, 161– 181. Bucureúti. Soto, Roberto. 2012. "La adquisición del euskera: aproximación a una secuencia típica del desarrollo morfosintáctico de 2 ½ a 5 años." Unpublished PhD, University of the Basque Country (UPV/EHU). Soto, Roberto, and Gerardo Aguado. 2015. "Apuntes sobre la adquisición de la morfosintaxis del euskera en niños de 2 y medio a 5 años." Revista de Logopedia, Foniatría y Audiología 5:30-49.

CHAPTER ELEVEN EARLY MINORITY LANGUAGE ACQUIRERS OF SPANISH EXHIBIT FOCUS-RELATED INTERFACE ASYMMETRIES: WORD ORDER ALTERNATION AND OPTIONALITY IN SPANISH-CATALAN, SPANISH-GALICIAN AND SPANISH-ENGLISH BILINGUALS TIMOTHY GUPTON

Abstract In the current chapter, I examine the Spanish syntax-focus interface competencies in a number of multilingual acquirers. The current chapter differs from previous studies of these phenomena (e.g. Domínguez & Arche, 2008, 2014; Hertel, 2000, 2003; Lozano, 2006; Parafita Couto, Mueller Gathercole & Stadthagen-González, 2015) in a number of important ways. Firstly, this study focuses on extremely advanced English L1-Spanish L2 bilinguals (ESB, N=13). Secondly, bearing in mind the Comparative Fallacy (Bley-Vroman, 1983), I compare this bilingual group to three Spanish bilingual groups: Spanish L1-English L2 sequential adult bilinguals (SEB, N=11), early Spanish-Catalan 2L1 bilinguals (SCB, N=9), and early Spanish-Galician 2L1 bilinguals (SGB, N=13). Thirdly, making use of modified appropriateness and preference judgments with transitive predicates, I pay particular attention to optionality, a residual phenomenon found in advanced bilinguals (e.g. Sorace & Filiaci, 2006; Sorace, 2011, 2012) that has been claimed to be indicative of interface instability. The findings are as follows. The ESB group displays convergence when compared with the mirror-image bilingual group, the SEB

Early Minority Language Acquirers of Spanish

213

bilinguals, the speaker type to which these speakers have the most exposure. In comparison with the other Spanish bilingual groups, subtle divergences emerge. In comparing “native-speaker” groups, experimental results reveal no differences when comparing each of the simultaneous bilinguals to the SEB group. However, comparisons of results in speakers of two minority languages—Galician and Catalan—uncover subtle differences related to the interface of syntax and focus.

1. Introduction In the current chapter, I examine Spanish competencies related to the interface of syntax and focus, with a particular emphasis on differences among speakers of two minority languages spoken in Spain: Galician and Catalan. This work compares these early bilingual competencies with those of advanced adult second-language (L2) acquirers of Spanish, thus building upon studies on the English L1-Spanish L2 acquisition of two types of intransitive predicates: unaccusatives and unergatives (e.g. Domínguez & Arche, 2008, 2014; Hertel, 2000, 2003; Lozano, 2006; Parafita Couto, Mueller Gathercole & Stadthagen-González, 2015). These studies confirm that the deterministic task for the adult English L1 individual acquiring intransitive predicate word order in L2 Spanish is a challenging one at lower competency levels. With continued exposure, L2 acquirers are able to acquire word orders not instantiated in the L1, although advanced L2 acquirers exhibit difficulties with certain subtle properties associated with Spanish word order. As surprising as this might seem for the acquisition of the SV/VS (subject-verb/verb-subject) alternation, it is important to bear in mind that this task is complicated further by exposure to multiple predicate types. Transitive and ditransitive predicates in the ambient linguistic data invoke the presence of objects and adjuncts, yielding a greater number of combinatorial possibilities.1 In this chapter, I present data testing competencies related to transitive predicates in Spanish by Spanish-Catalan, Spanish-Galician, and Spanish-English bilinguals. While all predicate types in Spanish allow flexible word order, this variation is not totally free. As I will show in section 2.2, the discourseinformation structure context at hand interfaces with the syntactic component, thus ruling out certain word order possibilities. Questions 1

For example, in a transitive sentence with a subject, verb, and direct object, there are six possible word order combinations: SVO, SOV, VSO, VOS, OVS, and OSV.

214

Chapter Eleven

related to residual optionality have played a central role in previous L2 acquisition studies, of particular interest in light of the Interface Hypothesis (IH, Sorace & Filiaci, 2006; Sorace, 2011, 2012), which proposes that optionality in advanced, adult, steady-state L2 grammars is indicative of instability resulting from the processing and mapping problem created by the interface of the narrow syntactic component with one or more external modules of the grammar, e.g. discourse pragmatics or information structure. This study examines both of these interfaces. Additionally, previous studies have not explored whether variation related to native speakers’ linguistic profiles further complicates the task of acquisition in L2 learners. The current chapter seeks to examine the interaction of this variable with focus and predicate type. In the following section, I review the syntactic assumptions related to the syntax-focus interface in Spanish.

2. Theoretical orientation While it is well known that Spanish word order is flexible, it is perhaps less well known that the acceptability of different orders is dependent on the information structure of the discourse-pragmatic context in which a clause appears.2

2.1. The syntax The current study assumes a Minimalist syntax (Chomsky, 1995 et seq.), by which utterance creation is driven by pair-wise merge of lexical items and feature checking. I assume that subjects of transitive predicates are base-generated in Spec, vP, as in (1) merging with a causative or agentive finite predicate that moves from V to v in Spanish.3 Next, the subject DP in Spec, vP is probed by the T head. If T has a strong [EPP]-feature, the DP will move to Spec, TP to check this feature and nominative Case on T, resulting in SV(O) order. If this feature is weak, the DP may remain in Spec, vP, thus checking [EPP] via long-distance Agree (i.e. without syntactic movement).

2

Caribbean varieties are a well-known exception. See e.g. Ortiz López (2009) for an overview. 3 DP=Determiner Phrase, VP=Verb Phrase, vP= “light” Verb Phrase (agentive/causative portion), TP=Tense Phrase, Spec=Specifier, Xo=head/nucleus.

Early Minority Language Acquirers of Spanish

215

While T is strong in languages with rigid SVO like English, it is optionally strong in many Romance languages with flexible word order. This leaves us with what appears to be optionality with respect to feature strength associated with T. According to Chomsky (2008), syntactic optionality is only apparent. Movement operations out of the first derivational phase (e.g. for subject DPs, movement out of vP/VP to TP) take place for reasons of a unique Interpretation at Spell-out. In the following subsection, I review proposals that seek to explain word order variation via checking of focus features. On the assumption that focus features interact with regular phi- (person, number, gender, etc.) and [EPP]-features, we have a straightforward explanation for word order variation.

2.2. Focus types Many L2 acquisition studies assume that focus-related phenomena involve the checking of focus features in the narrow syntax, an assumption that is not uncontroversial (see e.g. SzendrĘi 2004). In this subsection, I sketch out some of the proposals most standardly assumed to involve focus.4 I also make reference to relevant research on English, Galician, and Catalan, as the participants in the current study are speakers of these languages in addition to Spanish.

4

Definitions of focus are notoriously murky, and a review of the extensive literature is orthogonal to the current study. I direct the interested reader to (e.g. Gupton 2014: 73-82) for an overview. I follow López (2009), assuming that focus calculation and focus marking refer to a constituent that may satisfy a variable opened by a wh-element.

216

Chapter Eleven

2.2.1. Narrow focus Following Zubizarreta (1998), focused constituents are [+F] marked. Crucially, narrow focus must fall on the rightmost clause edge in Spanish (2b) or the result is infelicitous. (boldface font indicates prosodic stress) (2)

Who ate an apple? a. #[Juan[+F]] comió una manzana. (SVO) a’. #[Juan[+F]] comió una manzana. (SVO) b. Comió una manzana [Juan[+F]] (VOS) c. #Comió [Juan[+F]] una manzana (VSO) eat.pst.3sg Juan an apple ‘Juan ate the apple.’ This state of affairs is claimed to result from the interplay between two rules of the grammar in Spanish: the Constituent Nuclear Stress Rule (CNSR) and the Focus Prominence Rule (FPR). The former ensures that prosodic prominence falls on the right clause edge, while the latter ensures that elements marked as focused [+F] are more prominent. These two constraints conspire such that the scope of focus (i.e. the [+F]-marked constituent) and prosodic prominence occur in the same position at the right-most edge. When conflicts arise, [-F] constituents such as una manzana in (3) must undergo p-movement (prosodic movement), to a higher syntactic position in order to avoid the projection of focus.

Within the framework of this proposal, VOS (3b) and OVS are the result of the object scrambling to not violate the C-NSR or the FPR.5 Zubizarreta claims that prosodic stress on Juan (2a’), in a nonrightmost position (similar to English), results in a contrastive reading. In recent years, however, a number of experimental researchers have found important differences with respect to focus-related variable word order among different varieties of Spanish, a number of which accept SVO in 5

I acknowledge that neither of these options for obtaining different word orders is entirely without controversy. For example, following Chomsky (2008), shifted objects only target Spec, vP in order to escape the first derivational phase, and therefore, must move on to a higher structural position in order to obtain some other Interpretation.

Early Minority Language Acquirers of Spanish

217

subject narrow-focus contexts (Gabriel, 2010; Hoot, 2012; Muntendam, 2009 among numerous others). Assuming the cross-linguistic existence of a [+F] feature for narrow-focused constituents, it follows that [+F]-marked constituents in Spanish may or may not align with the rightmost syntactic edge. In English, focus-marking never deviates from SVO. Marking of [+F] right-aligns only when an object is narrow-focused, while [+F]marking of subject narrow focus never coincides with the right edge. According to Ordóñez (2000: 70-71), VOS is acceptable in Catalan and VSO is not. According to Vallduvì (2002: 1250), this is because narrowfocus (rheme, in his terms) constituents must appear in a postverbal position at the right clausal edge. Although little on information structure is found in the Galician literature, Gupton (2014: 147) found that SVO and VOS are both possible word orders.6 What emerges then is that in Catalan, [+F]-marking must coincide with the syntactic right edge, while in Spanish and Galician, it may or may not. 2.2.2. Contrastive focus Recall that Zubizarreta (1998) claims that (2a’) is necessarily contrastive in Spanish. Discursively, a contrastive sentence immediately follows an untrue assertion (4). (4)

A: ¿Qué comistéis hoy? B: Julia comió una manzana y nosotros unos perritos calientes. C: JUAN comió una manzana. ¿No te has dado cuenta? A: What did you all eat today? B: Julia ate an apple and we ate hot dogs. C: JOHN ate an apple. Didn’t you notice?

In (4), B asserts that Julia ate an apple and the rest ate hot dogs for lunch. C then corrects the untrue information in B’s statement regarding Julia, thus opening a variable for the incorrect subject and resolving it immediately with the correct information (i.e. Juan).

6

In Gupton’s (2014) results, SVO was rated significantly higher than VSO and VOS on an acceptability judgment task in response to subject narrow-focus contexts with full-sentence replies in Galician. In a preference task, VOS was overwhelmingly (96.88%) preferred to SVO (Gupton, 2014: 160). The objects in this task, however, were object clitics. Therefore, participants preferred VclS (VOS) to SVcl (SVO). See Gupton (2014: Ch. 5) for discussion of Galician enclisis.

218

Chapter Eleven

Following Rizzi (1997), it is typically assumed that contrastive focus readings are encoded in the left peripheral syntactic position FocP (5), with a contrastively-focused XP checking the associated discoursepragmatic feature [+Foc] in a specifier-head configuration.

In this approach, the feature [Foc] is associated with a contrastive reading; not an informational, narrow-focus reading. Although the details of focus feature checking differ, the locus of contrast encoding and checking (i.e. the left syntactic edge) is much the same as in Zubizarreta (1998). Notably, Ortega-Santos (2013) argues that contrastive focus may appear at the right syntactic edge in Spanish (6). (6)

Pero ¿qué dices? Comió una manzana JUAN. ‘What? It was JUAN, not Julia, who ate an apple.’

In his proposal, sentences like (6) involve movement to Spec, FocP followed by verb remnant movement past the landing site of the contrastive subject (i.e. Spec, FocP) to a higher syntactic position. Therefore, similar to narrow focus, left-edge and right-edge contrasts are proposed to involve [Foc] feature checking in Spanish. Putting aside the theoretical details of contrast encoding, we see that L1 and L2 acquirers are potentially exposed to a variety of word order possibilities for contrast expression. This differs from English, which, not counting cleft constructions, only allows in situ contrast. Catalan has been claimed to not allow low contrast (López, 2009: 47), while Galician has been found to favor low, in situ contrast while disfavoring high, ex situ (i.e. left-peripheral) contrastive focus (Gupton, 2012).

Early Minority Language Acquirers of Spanish

219

2.2.3. Broad focus Regardless of whether [+F] feature assignment operates at the sentence level (7), the standard assumption is that broad focus targets the unmarked word order in out-of-the-blue contexts, which is SVO for transitive predicates in Spanish.7 (7)

[F Juan comió una manzana.]

Taking into account unmarked word orders for unaccusative predicates (VS) and unergative predicates (SV), we see that both SV(O) and VS are possible in broad-focus contexts in Spanish. Gupton (2014: 159) found in experimental judgment task data that Galician speakers prefer SVO for broad focus contexts. Wheeler, Yates & Dols (1999: 599) claim that the basic word order for Catalan is SVO. 2.2.4. Acquiring optionality and word order in Spanish

broad focus

Spanish

Galician

Catalan

English

SVO

SVO

SVO

SVO

VOS

SVO

subject narrow focus SVO/VOS SVO/VOS object narrow focus SVO

SVO

SVO

SVO

SVO/VOS

SVO

SVO

SVO

OVS/SVO

SVO

OVS

SVO

subject contrastive focus object contrastive focus Table 11-1 Summary of possible (transitive predicate) word orders by language and information packaging/pragmatic context

To resume, both L1 and L2 acquirers of Spanish are exposed to a variety of word orders targeting focus type (broad vs. narrow), different constituent types (subject vs. object), and contrast type (high vs. low), to name just those examined in the current study. I summarize these variables

7

Broad focus is also referred to in the literature as all-focus, out-of-the-blue, or thetic contexts. The only thing that may be presupposed in such sentences is that something happened.

220

Chapter Eleven

and their possible word orders in Spanish in Table 11-1, alongside possible word orders in Galician, Catalan, and English. As one can see, Galician, Catalan, and English either entirely coincide with or allow subsets of the word orders attested as possible in Spanish for each context. According to Hulk and Müller (2000) cases of structural overlap involving interface phenomena such as these provide potential for cross-linguistic influence and interference in simultaneous early bilinguals. Although word order variation is explicitly taught regarding adjectivenoun combinations, object clitics with non-finite predicates, and affirmative versus negative imperative forms, the sort of syntactic variation examined here presents a particular challenge to adult L2 learners because it does not typically appear in pedagogical texts. Therefore, learners must rely on the ambient linguistic data to converge on native-like norms. In the following section, I examine a number of studies on this interface in adult L2 acquirers of Spanish.

2.3. Studies on the L2 acquisition of SV/VS word order in Spanish A number of studies (e.g. Domínguez & Arche, 2008, 2014; Hertel, 2000, 2003; Lozano, 2006; Parafita Couto et al., 2015) have examined the L2 acquisition of word order variation involving split intransitivity and narrow focus in Spanish by English L1 speakers. Despite their similarities, these studies report differing results, and therefore reach different conclusions regarding learner competencies. While participants in these studies differ with respect to L1, age, and Spanish competence level, the language profiles of control participants in these studies also differ. Hertel (2000, 2003) examined four English L1 learner groups (N=81) ranging from third semester university-level courses to graduate students and instructors at the highest advanced level. Native speaker subjects (N=18) were graduate students from a number of countries. All were living and studying or working in the USA and reported to have advanced English proficiency. Participants completed a linguistic history questionnaire and two linguistic tasks. A contextualized production task provided English prompts to avoid misunderstandings, as in (8) from Hertel (2003: 287, ex. 12): (8)

You and your friend Sergio are at a party. Sergio leaves to use the bathroom. While he is in the bathroom, Sara, the life of every party, arrives. When Sergio returns he notices that everyone seems much more festive.

Early Minority Language Acquirers of Spanish

221

Sergio asks you: What happened? What do you answer? _______________ A subset of the participants who completed this production task also completed an Appropriateness Preference Task (APT), whose results are reported in Hertel (2000). Participants rated paired SV and VS replies to similar contextualized scenarios on a 7-point (3=very natural to -3=very unnatural) Likert scale, and then were asked which of the replies they preferred, with the options A, B, and No preference. For this task, stimuli and sentence pairs were simultaneously read by participants and presented aurally via audio recorded by a native speaker. Following the context in (8), participants were presented with two candidate replies (9):8 (9)

a. b.

Llegó Sara. arrive.PST Sara Sara llegó. Sara arrive.PST ‘Sara arrived.’

Overall, advanced learners showed high preference levels for VS word order, as in (9a), in production and preferences, a result consistent with grammar restructuring following Full-Transfer/Full-Access (FT/FA) approaches (e.g. Schwartz & Sprouse 1996), given that VS order is impossible in English in analogous contexts. However, the advanced group also exhibited important differences suggesting non-target-like acquisition. In production task results, despite otherwise native-like production, the advanced group produced significantly higher proportions of VS in broad-focus replies with unergative predicates. In the APT results, the advanced L2ers indicated a preference for VS at a higher level than the native group for both intransitive predicate types. Lozano (2006) examined 17 British English native speakers and 18 Greek native speakers, all of whom were advanced-level Spanish L2 acquirers. Controls were 14 Spanish native speakers from Spain, Argentina, Mexico, and Venezuela who were pursuing post-graduate studies in the UK. Although their English proficiency is unknown, it is fair to assume a rather high level since all were pursuing advanced study at a British university. Participants completed a contextualized judgment task similar to Hertel (2000), but this employed a 5-point Likert scale ranging from

8

PST=past tense. For more on interlinear gloss conventions, see Leipzig Glossing Rules (https://www.eva.mpg.de/lingua/pdf/Glossing-Rules.pdf).

222

Chapter Eleven

2=completely acceptable to -2=completely unacceptable. The task included many of the same predicates as Hertel’s study. Results indicate that both learner groups distinguished between SV and VS order in broad-focus contexts with both predicate types in a native-like manner, a result he finds to be suggestive of grammar restructuring. For narrow-focus contexts, the results show that learners rated VS replies in a native-like manner, but rated SV replies significantly higher than the control group. Lozano takes this, combined with statistical evidence showing that these learner groups do not distinguish between SV and VS orders, as evidence of optionality consistent with earlier versions of the IH. Although the native controls also show evidence of accepting both word orders, they distinguished between the two candidate word orders. Domínguez & Arche’s (2008, 2014) study of split intransitivity examined three groups of British English L1 students (N=20 in each group) of Spanish at the secondary and tertiary level. Native-speaker control participants consisted of 17-18 year old students (N=20) in their final year of secondary school in Spain. No additional information was provided about the linguistic histories of the native controls. Results of the study, which replicated Hertel’s contextualized preference methodology, showed an increase in acceptance of VS order as Spanish proficiency level increased. Statistical comparisons found significant differences between native controls and the advanced learner group. Although advanced learners accepted VS at the highest rate of all groups, their high response percentages for “both” for all contexts suggested optionality among VS and SV word orders, a result the authors attribute to a representational syntactic deficit, not a purely pragmatic one. Curiously, however, nativespeaker replies also exhibited optionality, with ratings of “both” ranging from 13 to 26 per cent. Although VS was preferred for unaccusative predicates in both contexts, they found that SV was preferred in broad focus contexts with unergative predicates. There was no clear preference for narrow-focus replies. These results strongly suggest that native speakers also have optionality, a result not found in previous studies.

2.4 Summary and research questions Taken together, the L2 studies outlined above find that adult L2 acquirers of Spanish have difficulty acquiring a native-like distribution of SV and VS in varieties of Spanish that instantiate variable word order (although see e.g. Ortiz-López (2009) on the L2 acquisition of Caribbean Spanish). As previous studies on the L2 acquisition of Spanish syntax and related

Early Minority Language Acquirers of Spanish

223

interfaces have neglected transitive predicates, the current study seeks to answer three important questions: 1.) Do advanced L2 acquirers of Spanish acquire word order variation with transitive predicates in a native-like manner? It is currently unknown if the complexity associated with an additional argument results in difficulties for the L2 acquirer. Therefore, the answer to this question will help to uncover any acquisitional asymmetries while providing a more complete picture of the acquisition of Spanish word order. With respect to the acquisition of syntax interface phenomena, previous studies were inconclusive. As a result, they make differing claims regarding this interface as well as the IH. The current study examines two particular focus-related properties. For this reason, the second question that I seek to answer is the following: 2.) Are all focus-related interface properties equally problematic for advanced adult L2 acquirers of Spanish? The answer to this question may shed light on the syntax-focus interface in Spanish, which encompasses a number of distinct linguistic modules, including information structure and discourse pragmatics. The final question of interest to this study involves native speaker variation. The L2 acquisition studies in §2.3 make use of control participants for statistical comparisons, but not all native speakers have the same linguistic background. Many speakers of Spanish acquire a variety of other languages in addition to Spanish in both naturalistic and formal settings. These additional languages may or may not manifest focusrelated word order variation, which, in turn, may result in cross-linguistic interference. Therefore, the third research question I seek to answer is the following: 3.) With respect to the syntax-focus interface in Spanish in bilingual speakers, what differences (if any) exist according to the specific language pairing?

224

Chapter Eleven

3. The current study The data reported on in the current chapter forms part of a larger investigation of the L2 acquisition of word order variation in Spanish, reported in Gupton (in preparation).

3.1. Participants All participants completed Internet-based informed consent and a linguistic history questionnaire. Analysis of the linguistic history data allowed identification of four bilingual speaker groups: English L1Spanish L2, Spanish L1-English L2, Spanish-Galician 2L1, and SpanishCatalan 2L1. The English L1-Spanish L2 bilinguals (N=13) were very advanced learners of Spanish working at a university in the USA as professors, instructors, and graduate students. The latter three groups initially volunteered to participate as native-speaker controls. The Spanish L1English L2 (N=11) bilinguals were also professors, instructors, and graduate students working at the same university in the USA, and were primarily from Spain, Argentina, and Peru. The Spanish-Galician bilinguals (N=13) and the Spanish-Catalan bilinguals (N=9) were living in Spain at time of data collection. All had childhood exposure to the two languages in question. The data-gathering tool in this study consisted of two offline (i.e. untimed), Internet-based linguistic questionnaire tasks administered via pseudo-randomized modules on surveygizmo.com, followed by completion of a shortened version of the University of Wisconsin Spanish placement test. The first linguistic task was a contextualized Appropriateness Judgment Task (AJT), in which participants were presented with a pragmatic context in Spanish followed by two word order response options (10):

Early Minority Language Acquirers of Spanish (10)

225

Estás en la residencia estudiantil cuando llega un repartidor de Pizza Móvil con una pizza. Le dices que espere y vas a preguntarles a los amigos: “¿Quién ha pedido una pizza?” Tu amigo Xabi te contesta: You are in your college dorm when a pizza delivery person arrives with a pizza. You tell him to wait and you go ask your friends: “Who ordered a pizza?” Your friend Xabi answers: a. La ha pedido Esteban. Ya lo busco. (clVS) cl.F has ordered Esteban. now cl.M look-for.1s b. Esteban la ha pedido. Ya lo busco. (SclV) Esteban cl.F has ordered. now cl.M look-for.1s ‘Esteban ordered it. I will go get him.’

Participants rated each response option with the clitic-verb (clV)9 sequence either followed by the subject (clVS) or preceded by the subject (SclV) on a 4-point Likert scale with an additional option of “I don’t know” (11). (11)

4 – totally acceptable 3 – rather acceptable 2 – somewhat acceptable 1 – not acceptable 0 – I don’t know

The above scale, as well as the definitions appeared (in Spanish) for each sentence response option in the task. Each participant rated 20 stimuli and two response options for each context, netting a total of 40 AJT ratings per participant. The independent variables manipulated by the stimuli were focus type (broad focus, subject narrow-focus) and predicate type (transitive: comprar, regalar, prestar, enviar, pedir; and unaccusative: salir, entrar, venir, llegar, volver/venir). There were five tokens for each combination: 1) transitive predicate in broad focus context, 2) transitive predicate in subject narrow-focus context, 3) unaccusative predicate in broad focus context, and 4) unaccusative predicate in subject narrow-focus context. I report only on the first two in this chapter for reasons of space. The 20 stimuli were randomized for order of presentation and pseudorandomized for order of appearance of the response word orders (SV(O) and VS(O)). All response options were accompanied by clickable audio read by a native peninsular Spanish speaker.

9

In Spanish, a clitic pronoun must appear left-adjacent to a finite verb. All response options in this task included finite verbs, thus always resulting in cliticVerb (clV) word order.

226

Chapter Eleven

The second task was a Word Order Preference Task (WPT) with modifications to the preference options available in order to examine potential optionality without relying on extrapolation. In past studies, optionality is extrapolated based on two different statistical indications: 1) AJT results that do not indicate significant differences between word order response options (e.g. Lozano 2006), and/or 2) an indication of “both” on a contextualized WPT (as in Hertel 2000, 2003; Domínguez & Arche 2008, 2014). In my modified WPT, participants were presented with a pragmatic context in Spanish following Hertel’s (2000, 2003) methodology, as in (8) and (10), which was followed by two word order response options, each with accompanying audio stimuli and a preference scale in Spanish (12): (12)

A. Sentence A is clearly preferable to Sentence B. B. Sentence B is clearly preferable to Sentence A. C. Sentences A and B are equally preferable. D. Neither of these sentences is preferable. E. I don’t know.

The 20 different contextual scenarios were randomized for order of presentation. This task manipulated two independent variables: pragmatic focus type (contrastive focus vs. narrow-focus) and constituent target (subject vs. direct object). All response predicates were transitive.10 Again, response options were pseudo-randomized in order to avoid any priming effects related to response order. There were four possible variable combinations and there were five stimuli for each: 1) subject narrowfocus, 2) subject contrastive focus, 3) direct object narrow-focus, and 4) direct object contrastive focus.11,12

10 Since the WPT task included purely transitive predicates, a greater variety was employed. These were: pedir, comprar, dar, enviar, prestar, regalar, tomar, cantar, limpiar, and ganar. 11 Contrastive focus contexts and replies were designed following López’s (2009) description of “true” contrastive focus, thus avoiding the use of negation. 12 Native speaker comments on early versions of this study remarked that constituents previously invoked in questions resulted in unnatural, forced readings. For example, in subject narrow-focus questions (e.g. ¿Quién ha pedido una pizza?) the object DP una pizza appears. Therefore, in the response options, the DP should appear as the clitic pronoun la.

Early Minority Language Acquirers of Spanish

227

3.2. Predictions For transitive predicates in the AJT, SVO word order is predicted to be the highest rated word order for broad focus contexts for all groups since, as we saw in Table 11-1, that is the default order for all four languages spoken by participants. If we assume that the existence of only one option in the L1 or other L1 will influence preferences by favoring that option in the L2/2L1, then for subject narrow-focus contexts, Catalan bilinguals should rate VOS (VclS) order higher, while English bilinguals should rate SVO (SclV) higher. By the same assumption, Galician bilinguals may rate both SVO and VOS highly, perhaps at levels suggesting optionality for subject narrow-focus contexts. For the English L1-Spanish L2 acquirers, successful L2 acquisition will be evidenced by statistically significant native-like appropriateness judgments. Optionality with respect to word order for all groups will be evidenced by statistically similar ratings for the two candidate word order responses on the AJT. With respect to the WPT, participants have to indicate one possible response preference. Preferences for subject narrow-focus should be as discussed for the AJT in the preceding paragraph. For object narrow-focus contexts, all groups should prefer SVO since all languages have been claimed to favor this order. For subject contrastive focus, all groups should prefer SVO order as well, as that is how this pragmatic meaning is expressed in each language. For object contrastive focus, it is expected that Catalan bilinguals will favor OVS order, in line with Catalan language tendencies, while English and Galician bilinguals should favor SVO since neither permits OVS. As with the AJT, successful native-like L2 acquisition will be characterized by statistically similar results. In this task, however, there are two interface-related phenomena examined: pragmatics (contrast calculation, i.e. low vs. high) and information structure (narrowfocus calculation, i.e. left-edge vs. right-edge). Therefore, any asymmetries will be suggestive that not all focus-related interfaces are equally problematic. As we saw in section 2, there is potential for optionality in contexts influenced by information structure type as well as pragmatic context type in Spanish, and not just for L2 acquirers of Spanish. Importantly then, for AJT replies, optionality in response data may not necessarily indicate interface instability. If specific language pairings influence competencies, we expect to find statistical differences between different native speaker groups for the above tasks.

228

Chapter Eleven

4. Results Responses gathered using the methodology outlined in §3.1 were exported from SurveyGizmo and statistically analyzed using Microsoft Excel and JMP Pro 11 in consultation with the Statistics Consulting Center at the University of Georgia. Bilingual group codes for all relevant figures are as follows: English L1-Spanish L2 bilinguals=ESB, Spanish L1-English L2 bilinguals=SEB, Spanish-Catalan bilinguals=SCB, Spanish-Galician bilinguals=SGB.

4.1. AJT results In broad focus contexts with transitive predicates, Related-Samples Wilcoxon Signed Tank Tests revealed that all bilingual groups distinguish between SVO and VSO word order (p