Cognate Vocabulary in Language Acquisition and Use: Attitudes, Awareness, Activation 9781783094394

Table of contents :
Contents
Acknowledgements
Introduction
Part 1: Bilingual and Multilingual Language Use
1. Language Users and Language Use
2. Attitudes Towards Bilingual and Multilingual Language Use: The Western and Central-European Perspective
Part 2: Defining Lexical Crosslinguistic Similarity
3. Where Does Crosslinguistic Similarity Come From?
4. Crosslinguistic Lexical Similarity
Part 3: Lexical Crosslinguistic Similarity in Use
5. Crosslinguistic Similarity and Crosslinguistic Influences
6. Cognate Vocabulary in Language Processing
7. Cognate Vocabulary in Second Language Acquisition
8. Language-related Factors in Lexical CLI
9. Learner-related Factors Affecting Lexical CLI
10. Reliance on Lexical CLI as a Strategic Behaviour
Part 4: Investigating Lexical Crosslinguistic Similarity in Language Learning
11. Introduction: Researching the Awareness of Cognates in Polish Learners of English
12. Investigating the Awareness of Cognate Vocabulary: Polish Adult Learners of English
13. Using Cognates as a Vocabulary Learning Strategy: Polish Adult Learners of English
14. Awareness of Cognates as a Motivational Strategy: The Age Factor
15. Towards Plurilingual Language Teaching with
References
Index

Citation preview

Cognate Vocabulary in Language Acquisition and Use

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

SECOND LANGUAGE ACQUISITION: 93

Cognate Vocabulary in Language Acquisition and Use Attitudes, Awareness, Activation

Agnieszka Otwinowska

MULTILINGUAL MATTERS Bristol • Buffalo • Toronto

To the memory of Professor Håkan Ringbom

Library of Congress Cataloging in Publication Data Otwinowska, Agnieszka, author. Cognate Vocabulary in Language Acquisition and Use: Attitudes, Awareness, Activation/ Agnieszka Otwinowska. Second Language Acquisition: 93 Includes bibliographical references and index. 1. Vocabulary—Study and teaching. 2. Cognate words. 3. Word recognition. 4. Languages in contact. 5. Language awareness. 6. Second language acquisition—Study and teaching. I. Title. PE1449.O79 2015 418.0071–dc23 2015019557 British Library Cataloguing in Publication Data A catalogue entry for this book is available from the British Library. ISBN-13: 978-1-78309-438-7 (hbk) ISBN-13: 978-1-78309-437-0 (pbk) Multilingual Matters UK: St Nicholas House, 31–34 High Street, Bristol BS1 2AW, UK. USA: UTP, 2250 Military Road, Tonawanda, NY 14150, USA. Canada: UTP, 5201 Dufferin Street, North York, Ontario M3H 5T8, Canada. Website: www.multilingual-matters.com Twitter: Multi_Ling_Mat Facebook: https://www.facebook.com/multilingualmatters Blog: www.channelviewpublications.wordpress.com Copyright © 2016 Agnieszka Otwinowska. All rights reserved. No part of this work may be reproduced in any form or by any means without permission in writing from the publisher. The policy of Multilingual Matters/Channel View Publications is to use papers that are natural, renewable and recyclable products, made from wood grown in sustainable forests. In the manufacturing process of our books, and to further support our policy, preference is given to printers that have FSC and PEFC Chain of Custody certification. The FSC and/or PEFC logos will appear on those books where full certification has been granted to the printer concerned. Typeset by Techset Composition India (P) Ltd, Bangalore and Chennai, India. Printed and bound in Great Britain by Short Run Press Ltd.

Contents

Acknowledgements Introduction

ix xi

Part 1: Bilingual and Multilingual Language Use 1

Language Users and Language Use Introduction. Why do Linguists Often Study Monolinguals? Defining Monolingualism Defining Bilingualism Defining Multilingualism Bilingualism and Multilingualism in Social Settings Defining Plurilingualism

2 Attitudes Towards Bilingual and Multilingual Language Use: The Western and Central-European Perspective Introduction Prejudice Against Bilingualism and Multilingualism The Current Position of English in the World Multilingualism with English: Identity Issues Between Obligation and Need: Attitudes Towards Multilingualism in Poland

3 3 4 6 10 13 15 18 18 19 21 23 24

Part 2: Defining Lexical Crosslinguistic Similarity 3 Where Does Crosslinguistic Similarity Come From? Introduction Contact-induced Changes Borrowing as a Linguistic Process Lexical Borrowing Languages in Contact: Lexical Similarity Between Polish and English

31 31 32 33 34

4 Crosslinguistic Lexical Similarity Introduction

43 43

v

36

vi Cognate Vocabul ar y in L anguage Acquisit ion and Use

Degrees of Crosslinguistic Lexical Similarity Describing the Degrees of Cognate Similarity

43 50

Part 3: Lexical Crosslinguistic Similarity in Use 5 Crosslinguistic Similarity and Crosslinguistic Influences Introduction Transfer and Crosslinguistic Influences Crosslinguistic Similarity and the Levels of Lexical Transfer Conclusions: CLI and Linguistic Relativity

59 59 60 63 70

6 Cognate Vocabulary in Language Processing Introduction The Bilingual Mental Lexicon Research on the Processing of Cognates Cognates in Experimental Studies on Vocabulary Learning

72 72 72 76 80

7 Cognate Vocabulary in Second Language Acquisition Introduction The Assumed Role of Cognates in Language Acquisition Mixed Evidence from SLA Studies Conclusions: Why do we get Conflicting Results in Various Types of Studies?

83 83 83 85

8 Language-related Factors in Lexical CLI Introduction Word-related Criteria Influencing CLI and Processing Contextual Factors Influencing Processing and the Recognition of Cognates Typological and Psychotypological Factors in Lexical CLI

94 94 95

9 Learner-related Factors Affecting Lexical CLI Introduction The Role of L2 Proficiency Advantages of Cumulative Learning Experience Language Awareness, Metalinguistic Awareness and Language Use 10 Reliance on Lexical CLI as a Strategic Behaviour Introduction Strategic Learning in Bilinguals and Multilinguals Metalinguistic Awareness and Reliance on Cognate Vocabulary Attention, Awareness and Noticing in Language Learning Crosslinguistic Lexical Similarities as Affordances

90

98 101 106 106 106 114 117 120 120 120 122 126 128

Content s

vii

Part 4: Investigating Lexical Crosslinguistic Similarity in Language Learning 11 Introduction: Researching the Awareness of Cognates in Polish Learners of English Introduction: Aim and Scope of the Research Methods of Study Structure of the Research Project

137 137 139 139

12 Investigating the Awareness of Cognate Vocabulary: Polish Adult Learners of English Introduction Study 1: Focus on Language Learners. Cognate Awareness Questionnaire Study 2: Focus on Cognate Words. Translation Task

144 165

13 Using Cognates as a Vocabulary Learning Strategy: Polish Adult Learners of English Introduction Study 3: Focus on Beginners of English. A Case Study Study 4: Focus on Advanced Learners of English

178 178 179 194

14 Awareness of Cognates as a Motivational Strategy: The Age Factor Introduction Study 5: Focus on Teenage Learners of English Study 6: Motivating Teenagers. A Large-scale Study

216 216 217 226

15 Towards Plurilingual Language Teaching with English: Conclusions and Pedagogical Implications A Summary of the Research Aims Factors that Affect the Awareness of Cognates Benefits from Raising Awareness of Cognates The Age Factor in Raising Awareness of Cognates Implications for Teaching and Syllabus Design: The Way to Plurilingual Education A Final Comment: Multilingual Constellations with English References Index

143 143

235 235 236 238 240 242 244 263 284

Acknowledgements

Investigating the topic of crosslinguistic similarity and writing this book has been a long process. Along the way, I have been encouraged and helped by a number of people to whom I owe a lot. First and foremost, I would like to express my thanks to my friends and colleagues from the International Association of Multilingualism who some time ago convinced me that my research (and my university career) makes sense. My special thanks are due to Larissa Aronin, Gessica De Angelis, Danuta Gabrys´-Barker, Ulrike Jessner and Jean-Marc Dewaele. Profound thanks are also due to David Singleton, who supported the book when it was still an early manuscript. Warm thanks are due to Aneta Pavlenko, who saw the early version of the manuscript and assured me that my perspective in writing about cognates was worthwhile. I would also like to thank my reviewer, who provided detailed, tactful and constructive advice on the manuscript regarding its content and the clarity of presentation within the chapters. I do believe that the manuscript has improved considerably as a result. Many thanks to Laura Longworth, the editor, for facilitating the reviewing process and for her patience in answering my questions. I have been really fortunate to have several Polish friends who were eager to read the whole manuscript, or its parts, and who gave me suggestions on how to improve my work. My very special thanks are due to Aniela Korzeniowska for her kindness and detailed language comments, and to Ewa Haman and Maciej Haman for their feedback and ongoing statistical support whenever I needed to consult them. Last but not least, I am extremely grateful to Jakub Szewczyk, who demanded precise answers to all his irritating questions concerning cognateness, forced me to learn statistics and assisted me in changing one of the research chapters at the very last moment. Without my friends’ invaluable advice this book would not have taken the present shape. I would also like to thank all the teachers and nearly 1000 students from various schools and universities who agreed to take part in the research. Lastly, most special thanks are due to my family for their help, patience and encouragement throughout the years.

ix

Introduction

Although published in the SLA series, this book brings together linguistic, psycholinguistic and educational perspectives on the phenomenon of cognate vocabulary across languages. It deals predominantly with Polish-English cognates and their use by bilingual and multilingual Polish learners/users of English. However, due to the widespread use of the English language, as well as the universality of the processes discussed here, I hope that the topic and the scope of the book will appeal to international readers of various language backgrounds. Let me begin by saying that I am Polish and I live in Poland, a central European country that is surprisingly monolingual in comparison with many other countries. This monolingualism often shapes the way languages are viewed and taught in Poland. Why did I then decide to write a book on bilingualism, multilingualism and the use of crosslinguistic similarity? My interest in cognate vocabulary and multilingual language learners rose thanks to three milestone observations made within a decade of language teaching and learning. They well illustrate why crosslinguistic similarity may be a fascinating topic of study and a highly useful tool in language learning. Observation 1. At the end of the 1990s, as an enthusiastic English language teacher with a fresh MA diploma in applied linguistics, I worked in a private English language teaching (ELT) school for adult learners in Warsaw, Poland. What was rare for such schools, the groups we taught were small and the learners were well matched in terms of their English language proficiency. Since classes aimed at communicativeness, we conducted a lot of speaking activities and closely monitored our students’ discussions. I noticed that one of my intermediate-level learners clearly differed from the rest of the group. Whenever she spoke, she sounded more mature, proficient and sophisticated than the rest of the group. I loved the way she could elaborate on more abstract topics when the rest were lost for words. I tried to find out why she seemed so ‘special’ although her written tests did not differ much from the rest, and she made intermediate-level mistakes. As it turned out, this lady was a teacher of French. In her speech she often resorted to using French-English cognates, which boosted her fluency. xi

xii Cognate Vocabul ar y in L anguage Acquisit ion and Use

Observation 2. Sometime later, I went to live in Germany with my (pre) intermediate German, my husband (who does not speak the language) and two toddlers. I tried to speak English, but my neighbours, of course, preferred using German when talking. What helped me survive were my metalinguistic knowledge and the awareness of crosslinguistic similarity between Polish, English and German. Obviously, falling into the trap of false friends was never a nice experience, but the benefit of trying to use some familiar words and constructions was stronger than the embarrassment of saying something that was not entirely correct. After a year in Germany, my friends were amazed how well I was able to communicate, although half of my articles and endings followed the Polish morphological system and my prepositional phrases strangely resembled those in English. Observation 3. The final experience that aroused my interest in crosslinguistic comparisons and multilingual learners was teaching practical English classes to advanced learners at the University of Warsaw. One of the groups clearly stood out from the rest: they asked questions about language constructions that were different from my other groups, they compared English with other languages and they seemed to know more words. These students were multilingual and they were advanced in two or three foreign languages. These three findings led me to studying the topic of crosslinguistic similarity and its role in multilingual language acquisition (MLA). Amazed by the ‘discovery’ of cognates, I decided to investigate how cognateness ‘works’ across languages and how it can be helpful in individual language learning. I mainly focused on Polish and English because Polish is my native language, while teaching English is my main area of research. Thus, the entire research presented in this book deals with the role of crosslinguistic lexical similarity and multilingualism with English as a part of a language constellation. My work presented here consists of four parts. The first part opens with a discussion of the terminology used in studies on bilingualism, multilingualism and language acquisition. It shows how ambiguous some terms may be when taking into account a variety of language acquisition routes and settings. This part continues by presenting an array of attitudes towards bi- and multilingualism and the spread of English in the context of Western Europe versus the context of central Europe and Poland. The second part of the book focuses on the processes of language contact and the origins of crosslinguistic similarity. It discusses definitions of cognates, internationalisms and false friends, as well as various classifications and scales of crosslinguistic similarity. Finally, it zooms in on Polish-English cognate vocabulary. It shows cognates to result from macro-scale historical processes of borrowing and change. These, in turn, depend on attitudes and social prestige assigned to languages. The third part of the volume dwells on the role of crosslinguistic similarity in the micro-scale processes of language acquisition and use. It presents

Introduc t ion

xiii

the notions of crosslinguistic influences in the area of lexis, as well as the various levels of transfer. Then it passes on to the role of cognates in language processing, presenting psycholinguistic studies on the mental lexicon and the cognate facilitation effects. These are contrasted with the mixed results of studies on the use of cognates in the second language classroom. Finally, this part of the book provides a detailed discussion of the item-related factors in the transferability of lexis, and learner-related factors that have an impact on the results of the studies discussed. The last part of the volume presents a large-scale, long-term research project focusing on the awareness and learning of cognates by Polish monolingual, bilingual and multilingual learners of English. The project employs both quantitative and qualitative methodologies and features several studies that form a coherent whole. The research presented shows, on the one hand, how cognate vocabulary can be examined from a wide array of methodological perspectives, and on the other hand, how it may be dealt with in practice, when teaching learners of various proficiency levels, ages and backgrounds. Eventually, let me come back to my experiences with cognates. Admittedly, the maturation of the ideas presented here was a slow process: it also took me about a decade to make my three milestone observations, to carry out my research and to finalise this writing project. I truly hope, however, that within another decade we will be able to propose practical solutions to using crosslinguistic similarity in language teaching and the potential of cognates across various languages will find its due place in school syllabuses. With this goal in mind, I intended this book to appeal to people involved in studying SLA and MLA, as well as those interested in educational research, although I believe that the book should also be of interest to those involved in the field of crosslinguistic influences. To my knowledge, no other publication provides so many different theoretical and empirical perspectives on cognate vocabulary, bringing together research on language contact, sociology of language, psycholinguistics and language acquisition.

Part 1 Bilingual and Multilingual Language Use

1

Language Users and Language Use

Introduction. Why do Linguists Often Study Monolinguals? Knowing and using several languages is a norm in many societies and countries around the world, a fact which escaped the attention of linguists for a long time. Since the late 1950s, the main object of many linguistic studies has been a set of abstract rules of one language, in accordance with the assumption that all normal adult human beings have linguistic competence in their first language (e.g. Chomsky, 1959, 1965) and language is ‘represented as a speaker’s mental grammar, a set of abstract rules for generating sentences’ (Larsen-Freeman & DeCarrico, 2010: 20). Although the concept of ‘communicative competence’ (Hymes, 1971) drew linguists’ attention to language use in particular contexts, which opened ground for the development of sociolinguistics and pragmatics, for many decades linguistics focused on studying abstract monolingual native speaker’s competence. Chomsky’s concept of language competence and performance, the observable manifestation of the underlying competence, was based on a much earlier dichotomy of langue and parole proposed by a Swiss linguist, Ferdinand de Saussure (1916). According to him, linguistic analysis was not supposed to focus on the use of language (parole, or speech), but rather on the underlying system of language (langue), namely on how the elements of language relate to each other synchronically. It was de Saussure’s understanding of language that strongly influenced structural linguistics, philosophy and literary criticism in the first half of the 20th century. Interestingly, structuralists did not adopt the views of de Saussure’s great contemporary, Baudouin de Courtenay, a Polish linguist (e.g. known for the widely-used notion of the phoneme). Baudouin de Courtenay not only distinguished between language as an abstract group of elements, and speech and its implementation by individuals, but also pioneered research on the use of multiple languages by 3

4

Par t 1: Bilingual and Mult ilingual L anguage Use

speech communities. Sadly enough for studies of multilingualism, structural linguistics and then Chomsky’s tradition followed the work of de Saussure, rather than that of Baudouin de Courtenay, excluding people who use more than language as their abstract competence was not ‘pure’ enough as the object of study (Chomsky, 1986). On the other hand, to return to the sentence opening this chapter, knowing and using several languages is a norm for many people around the world. Thus this book predominantly deals with language acquisition and use by those who use more than one language. In this chapter we will first say why the concept of the ‘ideal’ monolingual native speaker of a language is inadequate as an object for language study. Next, we will define someone who knows, learns or uses another language or languages i.e. becomes bilingual or multilingual. Finally, we will look at multilingual speech communities and define the notion of plurilingualism.

Defining Monolingualism It is worth understanding why the idealised native speaker is now considered an inadequate model for linguistic studies. A native speaker is traditionally defined as a person who has ‘learnt language in a natural setting from childhood as a first or sole language’ (Ellis, 2008: 297). This definition suffices from the linguistic point of view, however, today it is assumed that the ideal situation of knowing and using one language, i.e. being a monolingual native speaker of that language, may be true only of minority of language users. Using two, or even several languages, is natural in many regions of the world (cf. Aronin & Singleton, 2012a; Cook, 1991; Graddol, 1997; Herdina & Jessner, 2002). Thus, it does not suffice to concentrate on understanding and to explain the grammatical knowledge and the language processing of an isolated ideal native speaker. Considering multiple language abilities and changing social identities of language users, one may even question the very notion of the native-speaker and native-like proficiency (Siegel, 2003), which has become vague and fuzzy: Today’s ideal speaker lives in a heterogeneous society (stratified along increasingly globalized lines) and has to negotiate interactions with different people representing all sorts of power and solidary positions on a regular basis. What is this ideal speaker a native speaker made of, but a polyphony of codes/languages working cumulatively (and sometimes complementarily) rather than a single, first-learned code? (Mesthrie, 2010: 74) Therefore, it is proposed that ‘native speakerness’ is not a simple notion. For instance, Leung et al. (1997) claim that terms such as ‘native speaker’ and

L anguage Users and L anguage Use

5

‘non-native speaker’ should be replaced with the criteria of language expertise, language inheritance and language affiliation. Earlier, Skutnabb-Kangas (1984) proposed defining language users according to the origin of their language, their language identification, their competence and the function the language performs in their lives, as summarised in Table 1.1. Considering the criteria of defining the mother tongue as proposed above, we can see that identifying a monolingual native-speaker by his/her mother tongue poses several problems. As suggested by Kachru and Nelson (1996), if approached from the sociolinguistic perspective, a ‘genetic native speaker’ has to be distinguished from a ‘functional native speaker’ – someone who functions as a native speaker in a variety of a language. Kachru (1999, after Ellis, 2008) further argues that, for instance, an educated speaker of a non-native variety of English can achieve the same degree of functionality in the language as someone who acquired it in the natural setting. Cook (2007: 240, original emphasis) states that ‘within the past decade the term native speaker has been deconstructed, partly by recognising that people are multi-dimensional; the role of a native speaker is comparatively a minor part of one’s identity.’ Neither is it easy to define ‘monolingual’ and ‘monolingualism.’ According to Kemp (2009), first it has to be distinguished whether the terms refer to the language use of the individual, or individuals in communities and societies. The term ‘monolingual’ will be used with reference to individuals, the alternative terms being ‘monoglot’ and ‘unilingual.’ Kemp (2009: 13) defines ‘monolinguality’ as the ‘psycholinguistic state of an individual knowing one language’, and ‘monolingualism’ as the use of one language by societies and individuals in those societies (based on Harmers & Blanc’s 1989, 2000 opposition between ‘bilinguality’ and ‘bilingualism’, to be discussed in more detail in the next section). Monolinguals are ‘individuals who use one language and may be proficient at using a number of different varieties of the language together with different registers in the variety or varieties they know, and of switching between varieties and between registers in the appropriate context’ (Kemp, 2009: 13). Interestingly, even such a broad definition may be questionable when one considers Wandruszka’s (1979) claim that people are innately Table 1.1 Criteria of defining the mother tongue Criterion

Definition

ORIGIN IDENTIFICATION a. internal. b. external COMPETENCE FUNCTION

the language one learned first a. the language one identifies with b. the language one is identified as a native speaker of by others the language one knows best the language one uses most

Source: After Skutnabb-Kangas (1995: 44)

6

Par t 1: Bilingual and Mult ilingual L anguage Use

multilingual, just because they are capable of shifting between a number of language variants, such as standard language, dialect, colloquial language, specialist jargon and knowledge of earlier linguistic forms of their own language. Thus, the homogeneity of the monolingual speech community is not clear, considering the existence of language varieties, dialectal differences and differences in the register use within monolingual communities (Labov, 1969; Wandruszka, 1979). Since in many European countries people still identify themselves with one language, monolingualism is often seen by members of western cultures as the unmarked case, and a point of reference for comparison with bilingualism and multilingualism. Conversely, about half the world’s population is bilingual (Grosjean, 1982), and there are about 30 times more languages than countries (Romaine, 1989). Doughty and Long (2003: 4) state that ‘[i]n many parts of the world, monolingualism, not bilingualism or multilingualism, is the marked case.’ The changing perspective on the role of monolinguals has also been acknowledged by applied linguistics. In a recent textbook we can read that ‘because bilinguals outnumber monolinguals in the world’s population, bilinguals more than monolinguals provide a genuinely universal account of the cognitive mechanisms that underline language performance’ (De Bot & Kroll, 2010: 124).

Defining Bilingualism Defining bilingualism poses several problems, and the definition of bilingualism evolved throughout the whole of the 20th century (Ewert, 2009). Most importantly, the definitions vary in relation to criteria of classifying someone as bilingual or not. Older definitions dealt either with the person’s L2 proficiency, or with the functional aspects of language use. In more recent definitions the defining criteria also stress the qualitative differences between the monolingual and the bilingual mind. Finally, bilingualism can be defined in terms of the sequence and completeness of L2 acquisition, as discussed in the following section. It is worth noting that in the present book the terms acquisition and learning are used synonymously, as opposed to Krashen’s (1982) position. However, learning will always refer to classroom settings. Also, symbols L1 will be used to denote the first language, L2 to denote the second language, and L3-Ln to denote the third and other languages of the learner.

Bilingualism and L2 proficiency A popular belief is that a bilingual is someone who speaks two languages perfectly well, and that this situation applies only to people brought up with two languages and cultures. This view probably stems from Bloomfield’s classic definition (1933: 56): ‘In cases where perfect foreign language learning

L anguage Users and L anguage Use

7

is not accompanied by loss of the native language, it results in bilingualism, native-like control of two languages.’ Applying Bloomfield’s definition to describing people who use two languages poses several problems. First of all, it is difficult to define perfection of language use, and what is meant by knowing a language perfectly. Secondly, it raises the question of defining ‘native’ and ‘native-like’ control of languages, leaving it unclear where one ends and the other begins. This definition excludes not only language learners, but also all those who use two languages even on a daily basis, if their use of L2 does not meet idealised native-speaker standard (Ewert, 2009). Interestingly, Cook (1997, original emphasis)1 points out that Bloomfield’s very assumption is wrong, because ‘(…) the one thing that the L2 learner cannot be by definition is a native speaker.’ Imperfect as it may be, using Bloomfield’s definition for years resulted in ‘punishing’ language learners for their failure in achieving nativelike control of their L2, even though they could use their L2 quite efficiently and effectively. In an attempt to define the minimum proficiency needed, Macnamara (1967: 59-60) called bilinguals ‘persons who possess at least one of the language skills even to a minimal degree in a second language.’ This definition obviously points to the other end of the proficiency spectrum, similarly to a more recent statement by Edwards (1994: 55) that ‘[e]veryone is bilingual. There is no one in the world (no adult, anyway) who does not know at least a few words in languages other than the maternal variety.’ Although both these definitions allow us to treat even those L2 learners who are just beginning to learn the second language as bilinguals, it seems that both Macnamara and Edwards are as extreme as Bloomfield. It is difficult to define a bilingual only in terms of his/her maximal or minimal language proficiency.

Bilingualism: L2 use and the bilingual mind There are many more functional definitions of bilingualism. The earliest were coined by structuralists, as the one presented by Weinreich (1953) in his landmark book Languages in Contact. According to this definition: the practise of alternatively using two languages will be called here BILINGUALISM, and the persons involved BILINGUAL. Unless otherwise specified all remarks about bilingualism apply as well to multilingualism, the practice of using alternately three or more languages. (Weinreich, 1953: 5, original emphasis) Weinreich did not make a clear distinction between bilingualism and multilingualism, treating the latter as a variety of bilingualism. Another classic functional definition of bilingualism was proposed by Haugen (1953: 7, original emphasis), who assumed that bilingualism is ‘the point where a speaker can first produce complete meaningful utterances in the other language.’

8

Par t 1: Bilingual and Mult ilingual L anguage Use

On the other hand, Hockett (1958: 16) suggested that a bilingual person may have no productive control over the second language, but still may be able to understand utterances in the language. He called such instances ‘semibilingualism’, referred to as ‘passive’ or ‘receptive’ types of bilingualism in other definitions (Romaine, 1989: 10-11). According to Mackey (2000: 22), ‘bilingualism is not a phenomenon of language; it is a characteristic of its use. It is not the feature of the code but of the message. It does not belong to the domain of langue but parole.’ Contrary to contemporary beliefs concerning language processing in bilinguals, Mackey (1962, 2000: 23) saw bilingualism as a ‘behavioural pattern of mutually modifying linguistic practices varying in degree, function, alternation and interference.’ This remains in sharp contrast to contemporary views first expressed by Grosjean (1989: 6, original emphasis): ‘the bilingual is an integrated whole which cannot easily be decomposed into two separate parts. The bilingual is NOT the sum of two complete or incomplete monolinguals.’ Grosjean’s classic statement emphasises the qualitative differences between the language competence of a monolingual and a bilingual. More recent definitions see a bilingual through the prism of language use as ‘someone who operates during their everyday life in more than one language and does so with some degree of self-confidence’ (Miller, 1983: x). Further, bilingualism is defined as ‘the ability to use two or more languages sufficiently to carry out limited casual conversations,’ which ‘does rule out people who can use a second language only in [very limited and] specialised ways’ (Myers-Scotton, 2005: 44-45). The factors of proficiency and use may be combined. Table 1.2 gives an overview of some types of bilingualism as distinguished by Baetens Beardsmore (1986) and Wei (2000) that may be useful for the discussion in the following chapters. An interesting question is whether language learners can be called bilinguals. Although Hakuta (1986) postulated including all L2 learners in the category, researchers are not univocal in this respect. A crucial point in the discussion of whether L2 learners can be called bilingual comes from Cook (e.g. 1991, 1992, 1997, 1999, 2002), who points to the ambiguity of the very term ‘bilingualism.’ Instead, Cook proposes his own term ‘multicompetence’ to denote the state of knowing/using more than one language by an individual. Originally, multicompetence was defined as the ‘compound state of mind with two grammars’ (Cook, 1991: 112), but was later redefined: Multi-competence is the knowledge of two or more languages in the same mind. It extends the concept of interlanguage by recognising the continual presence of the LI in the learner’s mind alongside the second language, assuming that there is little point in studying the L2 as an isolated interlanguage system since its raison d’etre is that it is added to a first language. (Cook, 2007: 241)

L anguage Users and L anguage Use

9

Table 1.2 Some types of bilinguals (given in alphabetical order) Type of Bilingual

Definition

Additive Bilingual

An individual whose two languages combine in a complementary and enriching fashion. An individual whose ability to function in a second language is developing due to increased use. An individual whose mastery of two languages is roughly equivalent. An individual who can operate in two languages with or without full fluency for the task in hand. An individual who not only understands but also speaks and possibly writes in two or more languages. An individual who understands a second language, in either its spoken or written form, or both, but does not necessarily speak or write it. An individual whose second language is added at some stage after the first has begun to develop.

Ascendant Bilingual

Balanced Bilingual (Symmetrical Bilingual, Ambilingual, Equilingual) Functional Bilingual Productive Bilingual

Receptive Bilingual (Semi-bilingual, Passive Bilingual, Asymmetrical Bilingual) Successive bilingual (consecutive bilingual)

Source: After Baetens Beardsmore (1986: 5-18) and Wei (2000: 6-7).

This holistic view of competence remains in accordance with Grosjean (1989) and with the Dynamic Model of Multilingualism (Herdina & Jessner, 2002), to be discussed further in this chapter. It also allows for discussing various qualitative differences between monolinguals and bilinguals/multilinguals. According to Herdina and Jessner (2002: 59) ‘A bilingual is not the sum of two complete or incomplete monolinguals; s/he rather has a specific linguistic configuration characterised by the constant interaction and coexistence of the two languages involved.’ Essentially, with our current knowledge of how languages are acquired, it is necessary to redefine the goals of language teaching, which so far have implicitly assumed that students have to approximate native speakers. Cook (1997) discusses non-native speakers’ ‘failure’ to achieve native speaker competence, stating that a second language learner and a monolingual native speaker are completely incomparable. The concept of a native speaker as a language-learning goal is inadequate because ‘by definition you cannot be a native speaker of anything other than your first language’ (Cook, 2007: 240). Thus, alongside the concept of multicompetence, Cook introduces the distinction between L2 learners who ‘are acquiring a system for later use’ and L2 users who ‘are exploiting whatever linguistic resources they have for a real-life purpose’ (Cook, 2007: 242) and who ‘know and use a second language at any level’ (Cook, 2007: 240). The majority of people in the world are L2 users and the roles of L2 learner and L2 users may overlap.

10

Par t 1: Bilingual and Mult ilingual L anguage Use

Cook (1999) suggests that skilled non-native speaker users of English are better models for teaching than monolingual native speakers, while Siegel (2003: 193) points out that non-native speakers may have greater proficiency than native speakers in numerous situations. An interesting comment, which wittingly summarises the discussion of whether bilinguals should be native-like in their production, was made by Ringbom (2007: 102): ‘[i]f a learner is too native-like in his production, inevitable pragmatic errors involve a risk that he will be regarded as a stupid native rather than as an intelligent foreigner.’

Defining Multilingualism Multilingualism versus bilingualism It is still a common belief that being multilingual involves adding yet another language to the linguistic repertoire. In structuralist studies, multilingualism was viewed as a version of bilingualism (Weinreich, 1953; Haugen, 1953) and appears as such also in later definitions. For instance Baetens Beardsmore (1986: 3) argued that the ‘term bilingualism does not necessarily restrict itself to situations where only two languages are involved but is often used as a shorthand to embrace cases of multi- or plurilingualism.’ This approach is still present today in psycholinguistic literature. For instance, De Bot and Kroll state that psycholinguists have recognised the importance of extending the study of language processing to individuals who are acquiring or actively using more than one language. (…) [T]he term ‘bilinguals’ is used to refer to such individuals, even though their additional languages are not as strong as their L1. (De Bot & Kroll, 2010: 124, emphasis added) However, a multilingual cannot be judged in accordance with monolingual or bilingual standards. One of the first to draw attention to the fact that multilingualism is a more complex phenomenon than bilingualism was the Czech scholar Vildomec (1963). Following, came the studies of multilingual learning strategies by Naiman et al. (1978) and Ringbom’s (1987) classic book on trilingualism. According to De Angelis and Selinker (2001: 44), ‘it should be said that a multilingual is neither the sum of three or more monolinguals, nor a bilingual with an additional language.’ Today, it is rather bilingualism which is often seen as a specific case of multilingualism, and not vice versa (Herdina & Jessner, 2002). Although these phenomena remain interconnected, multilingualism stretches beyond bilingualism because introducing another language into a bilingual system changes the whole system with respect to the new configuration of mutual

L anguage Users and L anguage Use

11

interactions between all the elements. Therefore, the differences between bilingualism and multilingualism are not only quantitative but, most importantly, qualitative in nature, which pertains to knowledge, processes and ways of linguistic functioning (e.g. Herdina & Jessner, 2002; Hoffmann, 2001). Being multilingual is the very state of the co-existence of the languages in the user’s mind and their mutual interactions. Cenoz and Genesee (1998: 2) describe multilingualism dynamically as ‘the process of acquiring several non-native languages and the final result of this process.’ It is possible to define multilingualism from a purely functional perspective. For instance, Gabrys´-Barker (2005: 17) calls it, pragmatically, ‘the ability to use or function in more than two languages.’ However, if we apply Skutnabb-Kangas’ (1984: 81) criteria, it turns out that a multilingual can be defined in multiple dimensions (Jessner, 2006). Multilinguals will differ in terms of levels of proficiency in their languages, the origin of their languages, the functions of the languages for the individual, and their own identification with the languages, as well as the way others identify them. Combining some of the criteria discussed above, De Angelis and Selinker (2001: 44) say that ‘a multilingual is a speaker of three or more languages with unique linguistic configurations, often depending on individual history.’ Some authors (e.g. Hoffmann & Ytsma, 2004) oppose the terms of multilinguality, trilinguality and bilinguality associated with the process of acquisition, with the terms of multilingualism, trilingualism, and bilingualism, understood as the final product of acquisition. Others (e.g. Dewaele, 2008, 2010) are careful not to describe all individuals who use more than two languages as multilingual, but differentiates them on the basis of the number of languages used into trilinguals, quadrilinguals, pentalinguals, etc., acknowledging that multilinguals do not display equal proficiency in all their languages. Following Haugen, the terms polyglot and polyglotism can also be found in literature to denote individual multilingualism (e.g. Jessner, 2008b). Perhaps it is worth finishing this discussion with a quote from Kemp (2009: 24), who concludes that ‘it would be useful if researchers were to give a detailed definition of multilingualism as part of their study.’ Explicit definitions would allow others to understand the principles behind the study, and how each study relates to the existing literature. De Angelis (2007) suggests including in the definition of multilingualism as many of the following aspects as possible: the number of languages known to the speaker, the age of acquisition (AoA) for each non-native language, the sequence of acquiring all languages, proficiency level in all non-native languages including how this level was measured, whether the use of the languages is active or passive, what productive and receptive skills are used for each language and how these were measured, the time of exposure to native and non-native language environments, the manner of acquisition (formal/instructed acquisition versus natural acquisition), the amount of formal instruction in each nonnative language (years and hours per week), the classroom language of

12

Par t 1: Bilingual and Mult ilingual L anguage Use

instruction for each non-native language (if learned in a formal setting) and the context in which each language is or was used, for example at home, at school, with peers etc. Definitely, we have to be aware that individuals are unique and differ in their mastery of language skills, cognitive and metalinguistic abilities, as well as their backgrounds.

Individual multilingualism and language acquisition Comparing bilingualism and multilingualism from the perspective of language acquisition mechanisms, Cenoz and Genesee remark that multilingual acquisition and multilingualism (…) implicate all the factors and processes associated with second language acquisition and bilingualism as well as unique and potentially very complex factors and effects associated with the interactions that are possible among the multiple languages being learnt and the process of learning them. (Cenoz & Genesee, 1998: 16) Whereas in second language acquisition (SLA) L2 can be acquired after L1 or simultaneously with L1, it is not difficult to notice that in MLA languages can be acquired simultaneously or consecutively in various configurations. According to Cenoz (2000: 40), with three languages involved, there at least four different routes of acquisition: (1) (2) (3) (4)

L1, L2 and L3 can be acquired simultaneously. L1, L2 and L3 can be acquired consecutively. L2 and L3 can be acquired simultaneously after the acquisition of L1. L1 and L2 can be acquired simultaneously before the acquisition of L3.

How complicated and varied the process of MLA might be as compared with SLA, is illustrated by Table 1.3. It is worth noting that MLA, similarly to SLA, may also take place in various settings ranging from fully naturalistic to fully formal. The symbol L3 is widely used in literature to refer to the third language of an individual, in the sense of the third language the person has contact with. Another term and acronym widely accepted in the English-speaking research community is third language acquisition (TLA). Although TLA covers a variety of developmental patterns, its result is called trilingualism (Jessner, 2006). Hammarberg (2010, 2014) points to the fact that the use of the terms first, second and third language (L1, L2, L3) have become somewhat inconsistent for cases that go beyond the settings of one L1 – one L2 – one L3. The reason he mentions for this is primarily the ad hoc development of this terminology when the terms first language (L1) and second language (L2) were coined. According to Hammarberg (2014), the extension of the L1-L2 terms went in

L anguage Users and L anguage Use

13

Table 1.3 Possible variants of MLA Second language acquisition L2 (SLA) L1 ⇒ L2 (compound bilingualism) L1 + L2 (coordinate bilingualism)

Multilingual language acquisition MLA L3 (TLA) L1 ⇒ L2 ⇒ L3 L1 ⇒ L2/ L3 L1 /L2 ⇒ L3 L1/ L2/ L3

L4 L1 ⇒ L2 ⇒ L3 ⇒ L4 L1 ⇒ L2/ L3 ⇒ L4 L1 ⇒ L2 ⇒ L3/ L4 L1 ⇒ L2/ L3/ L4 L1/ L2 ⇒ L3 ⇒ L4 L1/L2 ⇒ L3/ L4 L1/ L2/ L3 ⇒ L4 L1/ L2/ L3/ L4

Source: After Cenoz, 2000: 40 and Gabrys´-Barker, 2005: 24.

two directions: one was to redefine L1 and L2 to denote ‘native language’ and ‘non-native language’, while the other was to add the category of a third language (L3) to focus on L3 acquisition. However, the terminology still could not capture the status of the multilingual’s languages in various complex settings of acquisition. That is why Hammarberg (2010, 2014) argues in favour of the cognitively based, rather than chronological models of MLA, claiming that there are qualitative differences between the native and the non-native languages acquired in complex situations. He proposes to take into account the following factors when describing MLA: NL(s)/Prior NNL(s)/Current NNL, or in other words: L1(s)/Prior L2(s)/L3 (where NL = native language, NNL = non-native language). He also defines L3 in the following way: In dealing with the linguistic situation of a multilingual, the term third language (L3) refers to a non-native language which is currently being used or acquired in a situation where the person already has knowledge of one or more L2s in addition to one or more L1s. (Hammarberg, 2010: 97, original emphasis) While currently the terms L1, L2 and L3 are used for describing the gradual expansion and size of a linguistic repertoire, as well as for characterising the status of a speaker’s languages in situations of performance, Hammarberg postulates using different sets of terms for the two purposes. He also proposes explicitness about the function in which the terms L1, L2 and L3 are being used in a given paper or document.

Bilingualism and Multilingualism in Social Settings A basic distinction must also be made between bilingualism and multilingualism at the individual and societal levels. An older term used for

14

Par t 1: Bilingual and Mult ilingual L anguage Use

societal bilingualism was ‘diglossia’, originally referring to communities where two or more varieties of the same language were used by some speakers and later extended to situations where two languages or two language varieties are used by the same population (Kemp, 2009). The early definitions by Bloomfield (1933), Weinreich (1953) and Haugen (1953) did not make it clear whether the term ‘bilingualism’ referred to individual, or societal bilingualism. One of the first to draw a clear line between the individual and societal uses of the term was Baetens Beardsmore (1986). His distinction was repeated in Hamers and Blanc (1989, 2000: 6), who opposed ‘bilingualism’, defined as the property of social groups, and ‘bilinguality’, which they called the psychological state of an individual able to communicate in two languages. Similarly, some authors differentiate between societal ‘multilingualism’ and ‘multilinguality’, i.e. individual multilingualism (Jessner, 2006; Ó Laoire & Aronin, 2004). It is worth noting that societal bilingualism and multilingualism often do not depend on the people’s own choosing, but rather may be forced upon them by other circumstances, including politics, religion, culture, education, economy and even natural disasters which cause major migrations (Crystal, 1997a). Following Siegel (2003: 179-180), one can distinguish five major types of sociolinguistic settings for L2 acquisition resulting in societal bilingualism. The dominant L2 setting (often called the majority language context), means that L2 is the dominant language of the majority of the population and is used in all domains of everyday life. The bilinguals learning/using the L2 are predominantly immigrants, visitors or indigenous peoples (e.g. Turks in Germany, or Aborigines in Australia). The minority L2 setting is when the speakers of the dominant L2 teach the minority L2, which usually happens in the naturalistic context (e.g. English speakers learning Welsh). In the external L2 setting, the speakers of a dominant language learn a foreign language, or a lingua franca (e.g. Poles learn English in Poland). The coexisting L2 settings are environments where L1 and L2 users are of similar status and the languages are used in similar domains (e.g. Italian and German in South Tyrol). Finally, the institutional L2 setting (sometimes called the official language context) is when L2 is widely used in some domains and institutions, but for most of the population it is the additional language (e.g. English in Nigeria, or Russian in the USSR). Societal multilingualism is sometimes divided into vertical and horizontal (Mansour, 1993). According to Mansour’s model, speakers can be viewed in terms of their organisation in space i.e. the type of pattern that multilingualism takes in the societies. Horizontal multilingualism is exemplified in communities, where multilingualism is present at a higher (i.e. macro) level of society, but this does not imply that all citizens are multilingual (Cenoz & Genesee, 1998). They live in their own geographic spaces, like separate groups not integrated into the larger multilingual society. This is the case with countries such as Switzerland or Belgium where there are many official

L anguage Users and L anguage Use

15

languages, but due to the territorial principle which is based on population separation, there are regions where people are nevertheless monolingual (Franceschini, 2009). Vertical multilingualism is displayed by societies where ‘two or more ethnolinguistic groups share the same territory and participate in joint socio-economic activities’ (Mansour, 1993: 19). This can be observed in many of the major European cities which are considered multilingual and multicultural, in some regions which are occupied by a bilingual or multilingual population or among people living in the borderland. In these speech communities, other languages (e.g. English) are taught as L3, which leads to multilingualism. Aronin and Singleton (2008, 2012a) claim that contemporary societal multilingualism is characterised by the spread of multiple language use through the entire social range and different professional groups. They also argue that since contemporary multilingualism is strictly connected with globalisation, contemporary societies are so inseparably linked with multilingualism that contemporary multilingualism is a prerequisite for society’s functioning and progress on a world scale.

Defining Plurilingualism Living in a multilingual society implies mobility and means having a bulk of linguistic and cultural experience which adds up to overall communicative competence. Such a way of looking at multilingualism is strongly connected with the notion of plurilingualism, widely recognised in European documents concerning language policy (cf. Komorowska, 2004, 2007a), which can be included under the umbrella term of multilingualism (Jessner, 2008b). The very word ‘plurilingualism’ is new to the English language. While in English the same term can be used for societal and individual multilingualism, French and German use different words for referring to an individual’s ability to use several languages (plurilinguisme/Mehrsprachigkeit, respectively) and to the multilingual nature of a given society (multilinguisme/Vielsprachigkeit, respectively). According to Mackiewicz (2002), the Council of Europe has translated the French terms literally into English, using ‘plurilingualism’ in the individual and ‘multilingualism’ in the societal sense. The European Union, on the other hand, uses the term ‘multilingualism’ when referring to an individual, and ‘linguistic diversity’ when referring to societies. Interestingly, the European policy has forced some local policy makers to coin new words to denote plurilingualism. For example in Polish, the word wieloje˛zycznos´ć (i.e. multilingualism) used to function with reference to both individual and societal multilingualism. However, the new term róz˙noje˛zycznos´ć (i.e. use of different languages) has been introduced to cope with the term of plurilingualism reoccurring in European documents. It is

16

Par t 1: Bilingual and Mult ilingual L anguage Use

also possible to find the new coinage of kilkuje˛zycznos´ć (i.e. use of several languages), which is supposed to be more neutral (Wilczyn´ska, 2007). The Common European Framework of Reference (CEFR), which is one of the best known documents formulated by the Council of Europe, provides an exhaustive definition of plurilingualism as: the ability to use languages for the purposes of communication and to take part in intercultural interaction, where a person, viewed as a social agent, has proficiency of varying degrees, in several languages, and experience of several cultures. This is not seen as the superposition or juxtaposition of distinct competences, but rather as the existence of a complex or even composite competence on which the user may draw. (Council of Europe, 2001: 168) This definition captures the core characteristics of multilingual language knowledge and use, stressing the importance of communicativeness rather than the mastery of language. In fact, the definition implies that a plurilingual individual naturally presents different levels of knowledge of each of their languages and explains the pluri-/multilingual system in terms of mutual influence among a person’s languages. A metaphor of a plurilingual person proposed by Christ (2001: 3 emphasis original) states that ‘[a] person is plurilingual if, with respect to a number of languages, he/she has learned to cross the threshold into these different language houses.’ By this he implies that there exists some ‘minimal’ competence which one has to achieve in order to be counted among plurilinguals. However, plurilingualism seems to mean more than multilinguality, when one takes into consideration the socio-cultural factors interwoven into the definition. It is often specifically underlined that plurilingualism cannot be considered separately from pluriculturalism, which promotes interlingual tolerance, respect and cooperation. Still, some researchers prefer to use the terms of plurilingualism and multilingualism used interchangeably (e.g. Hufeisen & Neuner, 2004). In this book, both multilingualism and plurilingualism will be used with reference to individuals rather than to societies. Due to the situation described in Chapter 2, which concerns the nature of language use in Polish society, societal bilingualism and multilingualism are not the main focus of this work, which deals predominantly with individual bilingualism and multilingualism acquired in external and institutional settings. In the light of the discussion above, it is understood that an individual’s multi- or plurilingualism does not imply knowing several languages perfectly, but trying to use this knowledge in communicating with other people in various situations. This returns to the term of functional bilingualism/ multilingualism, referring to how well a person may function with his/her languages, which varies from minimal ability to accomplish a restricted set of social activities to being able to perform all the activities in the languages.

L anguage Users and L anguage Use

17

Such functional bilingualism/multilingualism is a feature of any language user/learner in Cook’s (2007) understanding. In practice, plurilingualism manifests itself in the ability to function in a multicultural community, and thus implies an increased linguistic and cultural awareness, as well as metalinguistic sensitivity to similarities and differences between languages.

Note (1) http://homepage.ntlworld.com/vivian.c/Writings/Papers/MonolingualBias.htm

2

Attitudes Towards Bilingual and Multilingual Language Use: The Western and Central-European Perspective

Introduction The last three decades have witnessed a growing interest in bilingualism and multilingualism as a societal and individual phenomenon. The development and spread of multilingualism is shaped by historical processes, current events and language policies. It also depends on the beliefs and attitudes of the members of particular societies. In the past centuries, multilinguality with Latin, Greek or French reflected the high social status of the speaker and his/her belonging to the social elite. Today, especially in the European context, people’s bi- and multilinguality is a reflection of their needs and lifestyles and multilingualism is recognised as a characteristic feature of the European identity. Worldwide, such recognition has been aided by the processes of electronic communication, globalisation, and the resulting increased mobility of individuals. This has led to reconceptualising approaches to language acquisition and use. However, recognising the multilingual nature of societies and the plurilingualism of individual speakers as commonplace does not happen overnight. Linguistic diversity stand in strong opposition to the one country-one language policy, which used to be promoted in Western civilisation for over two centuries and resulted in strongly negative attitudes towards bilingualism and multilingualism. Thus, in the West, including the European Union, there seems to be a clash between the understanding of bilingualism and multilingualism by researchers and by ordinary people. This chapter briefly discusses how such attitudes were evoked and how they are changing. Next, it brings to the foreground the issue of language prestige and the current 18

At t itudes Towards Bilingual and Mult ilingual L anguage Use

19

position of English in the world and in Europe. Finally it presents how Poland, once a multilingual society, became monolingual and why attitudes towards multilingualism are different in Poland than in Western Europe.

Prejudice Against Bilingualism and Multilingualism The monolingual bias in the Western research tradition Until the 1980s, bilingual and multilingual speakers were regarded as an exception in linguistic studies. Such an approach was also rooted in the – so called – fractional view of multilinguals understood as ‘several monolinguals in one person’, i.e. speakers with separate competencies for each of their languages (Jessner, 2006). The fractional view probably resulted from Bloomfield’s (1933) definition of bilingualism (see Chapter 1), but for a long time had an impact on researching language acquisition which focused on speaker’s individual and separate languages. Under such interpretation, interactions between the languages would be seen as accidental and anomalous (Grosejan, 2008). Any crosslinguistic influence would be regarded as negative, and codeswitching was seen as a result of carelessness and ‘contamination’ of the language used (cf. Jarvis & Pavlenko, 2007). The fractional view still manifests itself in a number of beliefs, the consequences of which are limiting for multilinguals themselves. One of them is that the less fluent speakers of L2-Ln are not considered real bilinguals/multilinguals (Grosjean, 2008). In language teaching, the idea of a perfect command of all languages sets goals which are highly unrealistic for many language learners (e.g. Cook, 2002, 2007). The fractional view of bilingualism and multilingualism led to encouraging language learners to strive for the linguistic competence of idealised monolingual native speakers interpreted as full language mastery. The monolingual bias also involved using native-speaker monolingual norms of the target language to explain the acquisition of language systems and to evaluate learners’ production, despite evidence for learners forming interlanguage grammars of their own. This attitude is known as the ‘comparative fallacy’ in interlanguage studies carried out in the 1970s and 1980s (Bley-Vroman, 1983). Another issue resulting from using monolingual norms is the ‘long-standing Western tradition of prejudice against bi- and multilingualism, ascribing a negative and harmful effect on the cognitive development of bi- or multilingual children’ (Jessner, 2008: 15). For a long time the users of two or more languages were considered linguistically or even intellectually inferior to monolinguals (cf. Baker, 1988; Cummins, 2000; Jessner, 2006, 2008). Such prejudice against bilinguals and multilinguals may have resulted from studies of bilingual children from the 1920s and 1930s, which demonstrated harmful effects of bilingualism on intelligence. Conducted mostly in Britain on Welsh/

20

Par t 1: Bilingual and Mult ilingual L anguage Use

English bilinguals and in the United States on immigrants, the studies continued until 1950s, showing ‘there can be no doubt that the child who is exposed and reared deliberately in a bilingual environment is handicapped in his language growth’ (Thompson, 1952: 367). The ‘language handicap’ was interpreted in terms of linguistic confusion that affected children’s intellectual development and academic performance (Hakuta & Diaz, 1985). These findings are now considered unreliable because they suffered from a range of methodological problems. Since bilingual language and cognitive skills were measured according to monolingual standards, bilinguals scored much lower than monolinguals because they were less proficient in the language. Tests also contained culturally-bound items and were politically biased since they were carried out with the intention to support the Anglicisation policies of the governments (Baker, 1988; Edwards, 2004). The positive effects of bilingualism began to be acknowledged after the landmark research of Peal and Lambert (1962) conducted in Canada with schoolchildren. In contrast to previous research, this methodologically rigorous study showed that bilinguals outperformed their monolingual peers on both verbal and nonverbal intelligence tests. The results served to fight prejudice concerning bilingualism and bilingual education, which was feared to produce retarded, poorly educated individuals unable to function in either language (Hakuta & Diaz, 1985). Today, it becomes clear that the respective situations of a monolingual language user and a multilingual language user cannot be objectively compared, which is supported by an increasing number of studies. However, it is worth remembering that at the beginning of 21st century there are still no reliable norms, standards and tests measuring bilingual child development in the first years of life (Armon-Lotem, 2012).

Attitudes to bilingualism and multilingualism in society The origins of the problem why bilingualism and multilingualism may be reluctantly received by Western cultures also relates to the one countryone language policy popularised from the rise of the European and American nation-states in the 18th and 19th centuries (Hornberger, 2002; Mesthrie, 2010). The introduction of the one country-one language policy is connected with periods of intense nationalism and the dominance of European powers. Nation-states, which functioned on the basis of political power and conquering new territories, ideologically claimed the people within as a nation. The state policies strived for the homogenisation of the society, including its linguistic unification (Hornberger, 2002; Mesthrie, 2010; Singleton & Aronin, 2007b). The nation-state was based on the assumption that a country is occupied by a nation which should be unified by one common language, so emphasis was placed on promoting or imposing one national language spoken by the majority of the population. Official monolingualism was supposed to create

At t itudes Towards Bilingual and Mult ilingual L anguage Use

21

a national identity based on the common language and culture, as opposed to other languages and cultures. From this perspective, bilinguals and multilinguals may have been perceived as a threat of questionable identity and thus, questionable loyalty (Hornberger, 2002; Mesthrie, 2010). As a result, we are currently used to thinking about countries in terms of the official languages spoken within their territories. This has also led to regarding certain languages as more prestigious than others, especially those that have become the official languages of schooling. Today, in Western societies the position and understanding of multilingualism is changing, which is directly connected with appreciating advantages of language knowledge and with language policies (Aronin & Hufeisen, 2009; Hufeisen & Neuner, 2004; Jessner, 2008). In Europe it is stressed that the knowledge of more than one language (or at least one European lingua franca), is indispensable in education, culture and social life, and it often becomes the condition for finding employment (Grin, 2002; Klein, 2007). AllEuropean political recommendations stress the value of linguistic richness and diversity, while national curricula recommend developing cultural identities in young people and developing their intercultural competence. As shown in various documents (Multilingualism: an asset for Europe and a shared commitment, 2008; Recommendation R (98) 6 Linguistic diversification; Recommendation 1539, 2001) and guidelines (Beacco & Byram, 2002, 2007), the ultimate goal for European citizens is to achieve plurilingual and pluricultural competence, defined as ‘the language communication and cultural interaction skills of a social player, who, at various levels masters several languages and has experience of several cultures’ (Council of Europe, 2001: 168). The term ‘European identity’ (Hufeisen & Neuner, 2004) is used to express the idea of belonging to the greater community as a representative of a smaller, native one. On the other hand, questions about the meaning of national identity and citizenship are still valid. Faced with the unprecedented wave of migrations, western societies tend to be increasingly concerned with the issue of maintaining social cohesion and preserving national, cultural and linguistic heritages. This is reflected in the inner policies regarding citizenship, which is often granted to immigrants on the basis of their language proficiency in the official state language and the knowledge of the moral values of the nation state. The knowledge of the national language is still viewed as a part of national identity and associated with social inclusion, patriotism, and loyalty (Van Avermaet & Pulinx, 2012).

The Current Position of English in the World It is claimed that the rapid development of multilingualism in the last decades of the 20th century was not accidental, but resulted from a historically unique combination of social, political and economic factors and is

22

Par t 1: Bilingual and Mult ilingual L anguage Use

directly related to globalisation (Aronin, 2005; Aronin & Singleton, 2008, 2012a; Graddol, 1997, 2006). Globalisation also poses problems connected with the spread of English, which involves the issues of power, identity and education (Graddol, 1997, 2006; Kachru, 1991). From a historical point of view, the current position of English can be attributed to two factors: the expansion of British colonial power, which had a major impact on the global status of English as it is today, and the emergence of the USA as the leading economic state of the 20th century (Crystal, 1997b). To illustrate the scale of this phenomenon, a decade ago the total number of speakers of English (as first, second and foreign language) was estimated to come up to 1000 million (Crystal, 1997b). However, according to Graddol (2006: 99), within the next 10-15 years it is likely to reach around 2000 million. The figures correspond to the expanding circle of Kachru’s (1985; Kachru & Nelson, 1996) model of concentric circles, where each circle represents a different way English is acquired and used. In the model, the Inner Circle reflects the traditional basis of English where it functions as the primary language of all domains of social and private activity, i.e. Great Britain, the USA, Canada, New Zealand and Australia. The Outer or Extended Circle illustrates the territories where English plays a crucial administrative role (e.g. in education or governance), largely a result of 19th century British colonial undertakings. The largest of the circles, the Expanding Circle, embraces those nations that recognise the global, political, technological and especially economic importance of English as an international language. The magnitude of the spread appears particularly impressive when we consider that the current demand for English comes largely from people’s need to communicate without obeying any prescribed standards. Thus, apart from Kachru’s model, other conceptualisations of the global use of English have been proposed, including English as an International Language and English as a lingua Franca (ELF) (Seidlhofer, 2002). International English is a variety of English used for cross-cultural communication by native speakers in all their dialects, as well as by nativised and bilingual speakers of English (McKay, 2002: 132). The term lingua franca is an umbrella concept. It denotes using a language for a number of varied purposes employed by heterogeneous groups of speakers for whom the language does not always have the status of a second or foreign language (Knapp & Meierkord, 2002: 9-10). English performs the role of a lingua franca: its global variety does not have native speakers, is not subjected to external standardisation and is influenced by a wide range of its users’ native languages (Seidlhofer, 2001, 2005). The descriptions and analyses of ELF in the field of pragmatics and culture (e.g. Firth, 1996; James, 2000, 2005), phonology (e.g. Jenkins, 2006, 2007) and lexicogrammar (e.g. Seidlhofer, 2001, 2005; Prodromou, 2006) describe its linguistic features, such as simplifications in lexis, grammar and phonology which contribute to the common use of ELF.

At t itudes Towards Bilingual and Mult ilingual L anguage Use

23

The popularity of ELF may result from the fact that it does not conform to the strict monolingual norms of English taught as a foreign language (Jenkins, 2006). ELF is not learnt to communicate with native speakers, but used as a tool to communicate mainly with other non-native speakers. Therefore, such phenomena as code-switching, or transfer, are seen as exploring bilingual and multilingual resources, rather than deviating from monolingual norms. According to Jessner (2006: 9), ‘the increasing number of speakers of ELF presents a crucial factor in the growth of global multilingualism.’

Multilingualism with English: Identity Issues The current position of English as an international language evokes various attitudes. The idea that a quarter of the world’s population can communicate in the same language seems appealing and the spread of English is interpreted as reflecting the desire of global speech communities to exchange their experience and ideas, which contributes to bilingualism and multiculturalism (Brutt-Griffler, 2002; Widdowson, 1997). Moreover, it is stressed that using English as an international language does not impose identity change, or a desire to imitate native speakers (House, 2003). However, positive attitudes towards the spread of the ‘world Englishes paradigm’ (Phillipson, 2007) are balanced by emphasising negative aspects of the phenomenon. The ‘global English paradigm’ (Phillipson, 2007) may endanger other languages and cultures through the undue glorification of the English language and policies adopted in relation to promoting English. There are tendencies to acknowledge the dominant role of English as a prestigious language and present the economically and politically dominant English-speaking groups as the only right, correct, standard and norm (Atkinson, 1999; Byram & Feng, 2004; Holliday, 1999). Such negative repercussions of the spread of English are called ‘linguistic imperialism’ (Phillipson, 1992) because the policy of the ELT industry is seen as a direct continuation of the imperialistic practices of the Inner Circle nations, which distribute their ideologies and cultures globally, and thus raise their commercial power (Phillipson, 1992; Pennycook, 2007). Since parents demand earlier and more intensive teaching of English encouraged by the prospect of social and economic benefits for their children, is that English is continuously replacing other languages as the second language taught in school (Cummins & Davidson, 2007). A result of the spread of English and its use as an international language motivation to learn English is also changing (Dörnyei, 2006). The traditionally understood integrative motive (Gardner, 1985) has to be reinterpreted because the integrative aspect of motivation does not make sense without a clear context and identification with the native L2 community. Global English identity will thus be related to instrumental aspects of English

24

Par t 1: Bilingual and Mult ilingual L anguage Use

language use, i.e. the Ought-to L2 self, as opposed to Ideal L2 Self, related to hopes, aspirations and needs. More people learn English because of the attributes they ought to possess in their professional lives, developing their Ought-to English selves. Therefore, English connected with the ‘global English identity’ becomes a survival skill of purely utilitarian, or functional nature (Dörnyei, 2006). It is also postulated that the notion of linguistic identity should not be represented by single languages, but sets of interacting languages called Dominant Language Constellations (DLC) of languages known to the speaker (Aronin, 2006; Aronin & Singleton, 2008, 2012a; Singleton & Aronin, 2007b). Such DLCs become the basic ‘unit of consideration’ or ‘the unit of measure’: for any individual the Dominant Language Constellation is the group of his/ her most important languages that, functioning as an entire unit, enable him/her to act in a multilingual environment in such a way as to meet all his/her needs. (Aronin & Singleton, 2012a: 59) A DLC would include languages of a different status, and either consist of a global language, an official language and a locally important language, or of an international language, the state language and the immigrant language. English often remains of major importance, but increasingly functions as only one ingredient of individual’s repertoire and is inseparable from other languages within the DLC, as one of the elements of a ‘package’ at the individual’s disposal (Singleton & Aronin, 2007b).

Between Obligation and Need: Attitudes Towards Multilingualism in Poland A multilingual nation from a historical perspective Although Poland is a country in central Europe, and has always belonged to the western tradition in terms of religion and education, Polish attitudes towards societal and individual multilingualism have only partly been shaped by the socio-historical processes prevailing in the West since the 18th century. The attitudes towards multilingualism have always been based on openness and tolerance towards other ethnic groups (Komorowska, 2014), evident at the turn of the 15th century, when Jewish refugees and Germans settled in Poland where they enjoyed many privileges. From the 14th century, multilingualism was assumed a status symbol. Apart from Polish, the knowledge of Latin, promoted by both church and state, was obligatory among the élite. While the nobility eagerly added French, German and Italian to their linguistic repertoire, for the lower social strata intercomprehension was a survival skill in a multi-ethnic nation. At the end of the 18th century, Poland

At t itudes Towards Bilingual and Mult ilingual L anguage Use

25

was a multilingual and a multinational country characterised by high religious and linguistic tolerance (Komorowska, 2014). The situation changed dramatically with the fall of the Commonwealth. With the partitions at the end of the 18th century, when the territory of Poland was divided between Prussia, Austria and Russia, Poland disappeared from the map of Europe for over 100 years. Throughout the 19th century, Poles were forced to use the official state languages: German and Russian. Thus, considerable effort was made by Polish elites to preserve Polish national identity and the Polish language, which was treated by the occupants as a minority language and was banned from use in schools and public administration from the second half of the 19th century. Speaking Polish became a manifestation of patriotism and respect for traditional values (Davis, 2005). On the other hand, it was still common among the Polish nobility, intelligentsia and bourgeoisie to be multilingual. Both men and women communicated in their native Polish, the state language (Russian or German), as well as French or English, which started gaining importance as the language of trade. These languages were taught at school and at home, together with Latin and classic Greek, which were still highly valued (Schramm, 2008). An outstanding example of such élite multilingualism was the great English writer of Polish origin, Joseph Conrad (Józef Konrad Korzeniowski), who was fluent in several languages. The main attitudes towards languages and language learning in 19th century Poland under the partitions can be summarised in several main points (Schramm, 2008). First of all, Poles manifested strong tendencies in defending the use of the Polish language in opposition to the acculturation and assimilation policies of the Russian, Prussian and Austrian nation-states. Second, knowledge of Latin in the upper classes was a must, as it was associated with Catholicism and the traditional values of the Polish nobility. Third, the assimilation policy of the occupants paradoxically opened the way for Poles, males and females, to study at German and Austrian universities and learn other modern languages from course books in German and in Russian. For instance, it is claimed that the double Nobel Prize winner, Maria Skłodowska-Curie, would never have been able to indulge in science if it had not been for her thorough education in Poland. At that time such education for women was not possible in France. Poland regained its independence in 1918 after 123 years of partitions and active opposition to the occupants. Not only did the Polish language and culture survive the assimilation policies, but actually they flourished, giving Poles a modern sense of nationality (Davis, 2005). Although, between the First and the Second World Wars, there was a rise of Polish nationalism and negative attitudes towards linguistic and cultural ‘otherness’ (e.g. Belarusian, Ukrainian, Lithuanian), in 20th century interwar Poland, individual multilingualism was still highly valued and modern western languages, as well as Latin and Greek, were commonly taught in secondary schools (Schramm, 2008). Official teaching programmes put emphasis on modern languages,

26

Par t 1: Bilingual and Mult ilingual L anguage Use

French, English and German, i.e. languages considered important for cultural, educational and commercial reasons (Iwan, 1972).

From World War II until the present: Forced monolingualism After World War II, which forced large-scale migrations, deportations and territorial changes, Poland became a politically monolingual country, with the official Polish spoken by a vast majority of the society and vernacular varieties used by the lower classes. During World War II the Nazi policies of the German occupants destroyed the Jewish communities, while the speakers of Belarusian, Ukrainian and Lithuanian were largely deported to the territories of the USSR. Since the country was incorporated into the Communist bloc, Poles were made to learn Russian at schools. Despite the efforts, Poles never actually used this language: attitudes towards Russian were strongly negative due to the still lingering memory of the Russian occupation before World War I and during World War II. Until the late 1960s, teaching modern western languages was reduced to a minimum (Grucza, 2001). All this resulted in the prevailing monolingualism in Polish society. However, the model of the Polish intelligentsia knowing at least one western language still prevailed through the 44 years of Communism. The ability to speak a foreign western language was highly appreciated by the society, which contrasted with the Communist government’s policy of making contact with the West as difficult as possible for its citizens (Komorowska, 2010; Wilczyn´ska, 2007). After the fall of the USSR, Russian quickly lost its privileged status of the lingua franca of the Communist bloc. In countries formed from the former Soviet republics, attitudes towards speaking and teaching Russian became visibly negative (Pavlenko, 2009b). In Poland, after the fall of Communism in 1989, foreign languages regained their prestige, as language education was again considered very important, but the teaching of Russian was practically abandoned as it lost the status of the language officially taught in Polish schools (Moz˙ejko, 2009). In the 1990s Poles turned to learning and using English associated with western capitalist values.

Attitudes towards multilingualism with English in Central Europe and in Poland During the Communist period, English symbolised freedom and modernity to Poles, and since 1989 it has quickly become the most prestigious foreign language across the society. In the 1990s, although still well-known only to the best educated social strata, English started to be extensively used in the media and advertising. Such intense language contact (see Chapter 3) resulted in numerous English affixes, words and phrases that penetrated into Polish, influencing the lexis, semantics and grammar (Man´czak-Wohlfeld,

At t itudes Towards Bilingual and Mult ilingual L anguage Use

27

1994, 1995; Otwinowska-Kasztelanic, 2000a; Zabawa, 2007). Linguists often expressed their negative attitudes towards Polish being inundated with English borrowings overused in daily life (Markowski, 1992). Since joining the European Union in 2004, English has still affected Polish. It is also the most highly-valued and widely-taught foreign language (EURYDICE, 2012a). However, the tendency to value English very high was not unique to Poland, but was visible also in other Central European countries. A study which is of great relevance to portraying the growing role of English in Central Europe after the fall of Communism is a survey of learner attitudes and motivation to learn foreign languages run in Hungary (Dörnyei et al., 2006). The study covers a period of transition from a closed, Communist society to becoming a member of the European Union. Due to the exceptionally large sample size (over 13,000 teenage Hungarian language learners) and measures repeated in 1993, 1999 and 2004 the findings provide unique information about how and why language globalisation took place in a postCommunist society. From the Polish perspective, the Hungarian study is of vital importance because of parallels between the socio-historical processes described by Dörnyei et al. (2006) and those taking place in Poland within the same period. Both Poland and Hungary witnessed large-scale socio-political changes after the collapse of the Iron Curtain in 1989 that had a substantial and pervasive language impact on attitudes and motivation to learn. The three dates chosen for the Hungarian survey are also milestones of change for both countries. The early 1990s witnessed the collapse of the Communism and radical transformation into an open, market-oriented democracy. During the transition three processes had a tremendous impact on the changing nature of the society and language knowledge. First, Russian, the compulsory before 1990, was replaced by a selection of western languages. Second, foreign television channels and English-language films became widespread. Thirdly, both countries opened up their borders, which resulted in a dramatic increase in foreign tourism and new economic relationships. The end of the 1990s marks the end of the transition process and joining NATO. Finally, May 2004 is the moment when both Hungary and Poland joined the European Union. The study by Dörnyei et al. (2006) gives a unique account of how the significant socio-political changes affected attitudes and motivation to learn five different target languages: English, German, French, Italian and Russian. The Hungarian data clearly reveal the abandonment of Russian and the strengthening of the position of English. At the same time they providing evidence that learning languages other than English may be impeded by the ‘Englishisation’ process. English gradually pushed to the side other foreign languages traditionally present on the Hungarian language learning scene, including the regional lingua franca, German. English started to be perceived as the natural option, or a part of the obligatory, ‘ought-to-have package’ of the teenager. The 2004 data reveal that 15 years after the fall of communism

28

Par t 1: Bilingual and Mult ilingual L anguage Use

languages like Italian and Spanish gradually became of interest for the most motivated learners. Such tendencies can also be noticed in Poland, where multilingualism and proficiency in languages other than English is gradually being perceived as an advantage: a constellation of the local Polish language, global English and French, Spanish, Italian, German, or Portuguese turns out to be an important asset for a professional career in adulthood (Gabrys´-Barker & Otwinowska, 2012). The new generations of multilingual students notices that the knowledge of other languages can be a serious advantage on the job market. Today, language teaching in Polish schools is in line with the recommendations of the Council of Europe. However, despite the positive attitude towards individual multilingualism in Poland, the ability to use several languages on a daily basis still relatively rare. Most Poles would not call themselves bilingual or multilingual and Polish society is still perceived as linguistically and culturally homogeneous in comparison with the western societies (Wilczyn´ska, 2007). According to a report by The Council for the Polish Language, ‘98% of Polish citizens claim that Polish is their native language, while members of the language minority communities are mostly bilingual, i.e. they also speak Polish’ (Pisarek, 2007: 1). The percentage of schoolchildren who speak at home a language other than Polish is only 2%, which is one of the lowest figures in Europe (EURYDICE, 2012a). Interestingly, linguistically homogenous Poland has the highest proportion of citizens (75%) who believe that everyone in the EU should be able to speak two languages in addition to their mother tongue (Eurobarometer Report 243). Considering the history of Polish society it is obvious that such positive attitudes towards individual multilinguality are culturally well grounded. That is why Poland, although perceived as a western democracy, is quite different from the western countries in terms of language use and attitudes towards multilingualism.

Part 2 Defining Lexical Crosslinguistic Similarity

3

Where Does Crosslinguistic Similarity Come From?

Introduction The languages we use are not stable monoliths, but undergo constant changes due to complex intra-linguistic (intra-systemic), and extra-linguistic (contact) factors, which are often of a socio-political and economic nature. Both types of factors are at constant interplay depending on the language and the circumstances of its use. As for the intra-systemic factors, languages have an internal structure which itself is dynamic, so they change to increase the economy of speech and to conform to the regular rules and patterns (Crystal, 1997). Languages also change as a result of contact with other languages, which is sometimes called foreign interference (Thomason & Kaufman, 1988). For instance, migrations cause divergence in pronunciation, vocabulary use and syntax under the influence of other languages or vernaculars. On the other hand, the features of one language are likely to converge under the influence of the features of other languages if the contact of the speech communities is frequent and intensive enough. Such interactions take place in all languages and the process of taking elements from one language to other languages has existed over all historical periods. In the European context, a special role in the process of language change was played by the linguae francae of their times: Greek, Latin, German and French, which have strongly influenced the vocabulary of modern languages in certain historical periods. Today this role has been taken over by English and the interaction between languages is increased by the mobility of people within and between countries (see Chapter 2). Chapter 3 will mainly focus on contact-induced changes and on the processes of linguistic borrowing. Resulting from language contact, borrowing leads to different levels of crosslinguistic similarity, important for the discussion in the following chapters. For the sake of simplicity, in Chapter 3 we will assume a bilingual perspective of language contact, which is often adopted in 31

32

Par t 2: Def ining Le x ical Crosslinguist ic Simil ar it y

the literature on contact linguistics. However, it is worth remembering that the processes described here may pertain to more than two languages. In the second part of the chapter we will point to the fact that thanks to language contact, numerous lexical similarities between languages can be found not only within language families, but also across typological boundaries. We will draw special attention to lexical influences from multiple languages present in English and Polish and briefly account for the language contact between these two languages. English is important here as a global language that is widely taught and used. Crucially, the similarities it bears with other languages may be useful for teaching and learning languages beyond English as L2 (see Chapter 15). Polish will be in focus as the L1 of the learners/users examined in the empirical part of the book (see Chapters 11–14).

Contact-induced Changes Contact-induced language changes operate when people who use different languages come into contact. However, the necessary factor for changes to occur is interaction when language users try to bypass ‘the communicative barriers facing them by seeking compromise between their forms of speech’ (Winford, 2003: 2). Thus, in most cases, changes into languages are introduced by bilingual or multilingual speakers of the languages involved. There are two general categories of contact-induced changes: those due to borrowing and those due to transfer (Winford, 2010). Contact-induced changes can be viewed in the macro scale, at the societal level (borrowing), and in the micro scale, at the level of individual language users (transfer and code-switching). The possible results of language contact differ depending on the internal (linguistic) and external (social and psychological) factors playing a role in the contact situation. The linguistic predictors of language change include the degree of typological distance and similarity between the languages in contact (Winford, 2003), as well as markedness influencing the ease of learning and the degree of integration of the elements within a linguistic system (Thomason, 2010). The relevant social factors involved in language change include the length and intensity of the contact, the respective size of the groups, the power or prestige relationships, the patterns of interaction between them and functions served by intergroup communication. The type of contact may directly influence the type of language change: the longer the contact period is and the greater the level of bilingualism, the more likely it is that some lexical and syntactic features will be adopted (Thomason, 2010). The type of contact is, in turn, affected by the attitudes of speakers, and by language policy. The features of a language regarded to be more prestigious will be eagerly adopted and used by the speakers of a language considered less prestigious. Deliberate change connected with language planning also depends on the

Where Does Crosslinguist ic Simil ar it y Come From?

33

attitudes: the more positive the attitudes within the community, the more likely the change will take place.

Borrowing as a Linguistic Process Borrowing can be understood as ‘the incorporation of foreign features into the group’s native language by speakers of that language: the native language is maintained but is changed by the addition of the incorporated features’ (Thomason & Kaufman, 1988: 37). In common understanding borrowing is limited to vocabulary. According to a recent definition: items/structures are copied from language X to language Y, but without speakers of Y shifting to X. In this simple form borrowing is characteristic of ‘cultural’ contact, e.g. Latin and English in the history of the latter, or English and other European languages today. Such borrowings are almost exclusively confined to words and phrases. (Hickey, 2010: 18) However, restricting borrowing to words is insufficient, since the process can involve much more than simply adopting foreign vocabulary items. As was widely discussed by the classics of language contact literature in the 1950s and 1960s (e.g. Haugen, 1950; Hockett, 1958; Weinreich, 1953), there is a much broader range of what can be borrowed between languages. In Haugen’s (1950: 212) definition of the process, it is language ‘patterns’ rather than vocabulary items: ‘borrowing is then the attempted reproduction in one language of patterns previously found in another.’ Clearly, this definition goes beyond the lexicon because the patterns to be reproduced do not have to be of a lexical kind. The new element, or pattern first penetrates from one language to another, then its use spreads till it is assimilated by the second language community and stops being perceived as foreign. Languages in contact are variously referred to in literature on the subject, as the ‘source’ or the ‘lending’ language and the ‘recipient’ or ‘borrowing’ language (e.g. Haugen, 1950; Thomason & Kaufman, 1988). Some other researchers prefer calling them the ‘donor’ language and the ‘recipient’ language (e.g. Winford, 2003), or the ‘source’ language and the ‘receiving’ language (Thomason, 2010). Of interest is also the very meaning of the term ‘borrowing,’ the inappropriateness of which was first pointed out by Haugen (1950: 211) ‘The metaphor implied is certainly absurd, since the borrowing takes place without the lender’s consent or even awareness, and the borrower is under no obligation to repay the loan.’ The inadequacy of term is also often noted in recent sources: [g]ranted, the term ‘borrowing’ is imprecise (nothing is ‘borrowed’ from A to B) but the term is established in the field and its use ensures

34

Par t 2: Def ining Le x ical Crosslinguist ic Simil ar it y

continuity with existing literature. The term ‘copying’ is in fact more accurate: speakers of language A copy features of language B into their own language. (Hickey, 2010: 20) Depending on the type of contact situation, the process of borrowing can take various degrees of intensity resulting in various types of borrowing (Thomason & Kaufman, 1988). If language contact is only casual and there is minimal cultural pressure, one can expect only lexical borrowing to take place. If societal bilingualism/multilingualism is wide-spread and if there is strong long-term cultural pressure from the source-language, a borrowing of structural features may happen as well, including phonological and syntactic elements and even features of the inflectional morphology. The more intense the contact, the more structural items can be borrowed, as presented in Table 3.1. Social values, purist attitudes of the speakers towards their native (recipient) language, and their loyalty to this language are responsible for certain resistance trends to borrowing foreign language words and patterns. The more the recipient language is regarded to be a prestigious one, the less intensive the process of borrowing is.

Lexical Borrowing Every language borrows lexical material from other languages, either by absorbing the items which have no native equivalents, or by literal translation. Words are the first foreign elements to enter the recipient language (Thomason & Kaufman, 1988), the process of borrowing lexical items is the simplest (Haugen, 1968), and changes in the area of words and their pronunciation are the most noticeable and frequent (Crystal, 1997; Weinreich, 1953). As far as types of lexical borrowing are concerned, Weinreich (1953) basically Table 3.1 Intensity scale of borrowing Kind of contact

Type of borrowing

Casual contact Slightly more intense contact

Lexical borrowing only Slight structural borrowing, conjunctions, adverbial particles Slightly more structural borrowing, derivational affixes Moderate structural borrowing (major structural features that cause relatively little typological change) Heavy structural borrowing (major structural features that cause significant typological disruption)

More intense contact Strong cultural pressure Very strong cultural pressure

Source: After Winford (2003: 30)

Where Does Crosslinguist ic Simil ar it y Come From?

35

divided lexical borrowings into simple words and compound words, whereas Haugen (1950) distinguished types of borrowing, based on whether source language morphemes are imported into the borrowing, and whether they are substituted by the elements of the borrowing language. The division into types is illustrated in Table 3.2 with examples from various languages. A loanword, term was coined in the 1870s from the German word Lehnwort (McArthur et al., 1992: 623), is a simple word or an unanalysed phrase transferred from the donor language to the recipient language. Loanwords can also be divided into ‘popular’ and ‘learned’ ones (Pyles & Algeo, 1982). ‘Popular’ are those transmitted orally and used in everyday communication, while ‘learned’ owe their adoption to scholarly influences. Such ‘learned’ loanwords, otherwise called quotes, are used only by some speakers of the recipient language, such as journalists, or scientists. Quotes do not constitute part of the lexical system of the recipient language, so they are not inflected and retain their original spelling (Winford, 2003). As for compound words and phrases, there are loanblends and loan shifts. Loanblends are kinds of borrowings which combine a foreign loan with a native form, so that only one element is of foreign origin. The category of loanshifts comprises loan translations and semantic loans: extensions, Table 3.2 Types of lexical borrowings

morphemic importation without substitution morphemic importation and substitution

morphemic substitution without importation

Simple words

Compound words/phrases

loanword, e.g.:

loanword, e.g.:

robot (borrowed from Czech to a number of languages) loanword, e.g.:

leitmotiv in English borrowed from German loanblend (combines a foreign loan with a native form), e.g.: Polish pracoholik from English workaholic [praca – work] + [holic]

Polish anioł, or English angel from Latin angelum (adapted phonologically and morphologically) loanshift: semantic loan (meaning extension under the influence of English), e.g.: Polish mysz (under the influence of English meaning extended from animal to a computer device)

loanshift: loan translation/calque (foreign elements replaced by semantically equivalent native ones), e.g.: English superman from German Übermensch [super – über] + [mensch – man] Polish nastolatek from English teenager [nasto – teen] + [latek – ager] French gratte-ciel from English skyscraper [gratte – scrape] + [ciel – sky]

36

Par t 2: Def ining Le x ical Crosslinguist ic Simil ar it y

narrowings, or shifts (Haugen, 1968) occurring when some words in the recipient language begin to be used in conformity with a foreign semantic pattern. Loan translations or calques are, in most cases, compound or complex loans, where foreign language elements are replaced by semantically equivalent native ones. All these categories are also used today, not only in studies on bilingual language contact, but also with reference to lexical crosslinguistic influences in researching bilinguals and multilinguals (e.g. Ringbom, 2006, 2007, see Chapter 5). Since the process of borrowing is dynamic, lexical borrowings undergo several stages of adaptation before they become fully assimilated. Unassimilated loans and quotes are taken over by the society and are adapted at various levels: graphical, phonological, morphological and semantic (Rusiecki, 1980). Partly assimilated loans are inflected and often take derivational affixes, but retain their original spelling. Fully assimilated lexical borrowings may become essentially indistinguishable from the native vocabulary (Winford, 2003).

Languages in Contact: Lexical Similarity Between Polish and English Multiple influences: Similarities across typological boundaries Thanks to the historical process of borrowing, i.e. adopting elements of other languages to the language of a given speech community, it is possible to find numerous similarities between contemporary languages. Obviously, such similarities can be traced between languages which are typologically close, for instance within the Romance, Germanic and Slavonic groups in the European context. However, similarities can also be found between languages which are typologically quite distant, such as Finnish and English (e.g. Ringbom, 2007), or English and Polish (Otwinowska-Kasztelanic, 2000a). From the discussion of borrowing, similarities between such languages are most evident in the area of vocabulary. When we look at the origin of English words, we can see that they were borrowed from a number of foreign languages, as in the example below (the choice of words and comments by Trask, 1994: 9–10): (a) kayak: from Eskimo. (b) cafeteria: from Spanish. (c) soprano: from Italian. Italian has long been known as the language of music, and English has borrowed dozens of musical terms from Italian. (d) tulip: from Turkish. (e) coach: from Hungarian. (f) kangaroo: from an unidentified Australian language.

Where Does Crosslinguist ic Simil ar it y Come From?

37

(g) ballet: from French. French was long known as the language of high culture, and many English words in the domain of art and literature come from French. (h) tea: from Chinese. The Chinese were the first to drink tea, and two of their words for it, te and cha, have been borrowed throughout the world. (i) palaver: from Portuguese. (j) democracy: from Greek. (k) khaki: from Urdu. (l) tsunami: from Japanese. (m) mayonnaise: from French. The great prestige of French cuisine has brought many French culinary terms into English. (n) yacht: from Dutch. The Dutch have long been noted seamen, and many English nautical terms derive from Dutch. (o) sex: from Latin. The word is derived from the Latin verb meaning ‘to divide’: the human race is ‘divided’ into two sexes. The words ‘section’ and ‘sector’ come from the same Latin root. (p) waltz: from German. (q) ukulele: from Hawaiian. (r) sauna: from Finnish. (s) denim: from French. (t) ski: from Norwegian. (u) juggernaut: from Hindi. (v) algebra: from Arabic. The medieval Arabs were distinguished mathematicians. Interestingly enough, the list can also serve us as an illustration of the fact that language distance does not guarantee complete lexical divergence between languages. For instance, most of the words enumerated above can also be found in contemporary Polish, a Slavonic language which typologically is quite distant from English. Apart from points (e), (i) and (u) above, in Polish we can find all the remaining words (Polish spelling): (a) kajak; (b) kafeteria; (c) sopran;, (d) tulipan; (f) kangur; (g) balet; (h) herbata (‘herb’ + Chinese ‘te’); (j) demokracja; (k) khaki; (l) tsunami; (m) majonez; (n) jacht; (o) seks; (p) walc; (q) ukulele; (r) sauna; (s) denim; (t) ski (recently borrowed, found especially in loanblends like ski-serwis); and (v) algebra. Although the words have been adapted graphically to fit the spelling of each language, one can easily see that the form of the English and the Polish words is very similar, and so is their semantics. Such a similarity is not accidental. It can be explained by the fact that new lexical elements are often borrowed by multiple languages independently. This often happens with the names of designates and notions unknown in the recipient language, internationalisms (see Chapter 4) and ‘exotics,’ i.e. expressions connected with the specific culture (Fisiak, 1970). What is more, borrowings may be introduced to the recipient language either

38

Par t 2: Def ining Le x ical Crosslinguist ic Simil ar it y

directly from the donor language, or by the medium of other languages. The direct importation is a result of simple contact between two languages. If a borrowing penetrates to the recipient language through other languages, it is a result of a complex contact (Fisiak, 1964). In the case of English borrowings in Polish this can be illustrated by the following examples: (a) budz˙et comes straightforwardly from the English word budget. (b) rumsztyk from the English rump steak came to Polish through German. (c) bonanza from the Spanish bonanza came to Polish through English, which serves as an intermediate link in this case. Fisiak (1964: 287) Additionally, many languages contain new words created from Latin and Greek stems and suffixes (e.g. Pl. telewizja – En. television, Pl. wideofon – En. videophone). These are the so called Neo Latin/Greek internationalisms coined especially from the 19th century onwards and adapted in many languages (Denison et al., 2006).

Foreign influence on the vocabulary of English Although English is a Germanic language, 75% of modern English words have been borrowed from other languages in the course of history (Winford, 2003: 29). The origin of 10,000 of the most frequent words in modern English are not Germanic but rather Romance: while 32% of the words come from Old English and 4% from other Germanic languages (due to the Scandinavian invasion between the 8th and 10th centuries), as much as 45% comes from French and 17% from Latin (Van Gelderen, 2006: 4). The remaining 2% is a result of borrowings from languages other than the ones mentioned above. Table 3.3 presents some of the sources of contemporary English words. The influence of Latin on English involved both the ‘popular’ and ‘learned’ loanwords (Pyles & Algeo, 1982). The Roman occupation made a lasting impression on British society and language. Consequently in Modern English, there are words of Latin origin from the times of the Great Roman Empire (e.g. wine – vinum, street – via strata), mostly of the ‘popular’ kind (Van Gelderen, 2006: 9). Christianizing in the Middle Ages resulted in the most extensive borrowing of the ‘learned’ kind, especially in the domains of religion, church institutions and law, education, books, learning (e.g. cleric, school, paper, library, scribe, complete, imaginary), as well as clothing, household goods, medical terms, animals and foods. However, the results of the influence were strongest in the 15th and the 16th centuries (e.g. area, data, gradual, medium, series, strict, ultimate, urban). The period of New Learning, when Greek and Latin were established as the main languages of education, science and culture resulted in a large number of semantic loans and loan translations (Kastovsky, 2006: 222; Pyles & Algeo, 1982: 297).

Where Does Crosslinguist ic Simil ar it y Come From?

39

Table 3.3 Some sources of words in contemporary English Some sources of the English words Latin French Scandinavian Spanish Dutch (including Afrikaans) Arabic Turkish Hindi Hungarian Cherokee

24,940 9470 1530 1280 860 615 125 120 26 1–3

Source: After Denison and Hogg (2006: 2)

The onset of the French influence on English is clearly marked historically by the Norman Conquest. The Normans, who spoke French, ruled both the country and the church, while the English-speaking population preserved the language and tradition of the country, which provided natural settings for bilingualism (Fisiak, 1993; Pyles & Algeo, 1982). The French influence affected mostly vocabulary and left traces in English phonology, spelling and morphosyntax. Borrowing from French manifested itself ‘from ecclesiastic terms to words relating to law, army and the navy, fashion, meals, social life, art, learning, and medicine’ (e.g. sacrifice, government, authority, army, captain, letter, literature, question, mirror, beef, mutton, fruit) (Flippula, 2010: 437). Through French, English was also indirectly influenced by Latin. As for other languages, Modern English came into contact with Dutch mediated by the political, commercial and cultural contacts. Thus, the main spheres of borrowing were maritime terminology (dock, gin, commodore, etc.), drawing and painting (sketch, landscape, etc.). Italian influenced English in the Renaissance period with the main spheres of borrowing being business (bank, risk, bankrupt, etc.), music, and architecture. Other influences on English came from Spanish, German, Russian, Czech, Indian, Japanese, and Native American languages, due to colonisation in the 18th and 19th centuries, as well as the spread of science and technology from the 19th century till the present day (Van Gelderen, 2006: 10).

Foreign influence on the vocabulary of Polish Even a casual look at Table 3.4 suffices to see that English and Polish, although typologically distant, were influenced by similar languages, both in simple and complex contact situations.

40

Par t 2: Def ining Le x ical Crosslinguist ic Simil ar it y

Table 3.4 Languages that exercised influence on Polish Languages influencing Polish

Period

Germanic dialects Latin Czech German Tatar and Turkish Hungarian Russian dialects Italian French Russian English

9th–10th c. 10th–20th c. 10th–16th c. 12th–20th c. 13th–17th c. 14th–16th c. 14th–19th c. 16th–17th c. 16th–20th c. 19th–20th c. 19th–20th c.

Source: After Dubisz (2005: 213).

Since Latin was one of the most important source languages for lexical borrowing in Europe, it also extensively affected the vocabulary of Polish, resulting in c.a. 10,000 borrowings. The influence lasted for 11 centuries because throughout the period Latin functioned as the language of the intellectual elite. The first Latin words connected with religion (ołtarz – altar, poganin – pagan, anioł – angel) entered Polish as a result of complex contacts with Czech and German. In the Middle Ages, numerous words connected with medicine (lawenda – lavender, cynamon – cinnamon) law, administration and education (data, statut) were borrowed (Dubisz, 2005; Klemensiewicz, 1961). Since the 15th century, Latin started to be widely used by the upper classes both in speech and in writing and there was even a fashion among the nobility to overuse Latin in speech (Dubisz, 2005). The borrowings from that period include those connected with institutions (abdykacja – abdication), art and literature (aforyzm – aphorism, absolut – absolute), artefacts (kolumna – column, tron – throne, aparat – apparatus) names of professions (architekt – architect, patron – patron, profesor – professor), and abstracts (apetyt – appetite, decyzja – decision, deklaracja – declaration, natura – nature, propozycja – proposition, reforma – reform) (Klemensiewicz, 1965: 146-148). In the 18th century, using Latin was diminished by the spread of PolishFrench bilingualism among the nobility (Dubisz, 2005). There are over 2500 German borrowings in Polish. In the late Middle Ages the period of intensive German settlement resulted in borrowings connected with buildings, commerce and administration, as well as technical terminology. The second wave of German borrowings came during the Partitions in the 19th century and brought about numerous calques and loan translations. The Renaissance period saw the increasing influence of Italian,

Where Does Crosslinguist ic Simil ar it y Come From?

41

with borrowings connected with architecture, art and names of plants (e.g. pałac – palace, kalafior – cauliflower). It is estimated that there are over 1200 Italian words in contemporary Polish. Contact with French, used as a European lingua franca in the 18th and 19th centuries, resulted in over 3500 borrowings in many domains of social life, art and literature (e.g. bukiet – bouquet, uwertura – overture, afera – affair). The Polish lexicon was also influenced by Russian and Ukrainian, and some words were borrowed from Hungarian, Spanish, Arabic, Turkish, Finnish, Dutch, Icelandic, Norwegian, Swedish and Japanese.

English influence on modern Polish English influence on Polish dates back to the turn of the 19th century and is characterised by the growing number of English loanwords. The oldest borrowings include those connected with social life (lord), trade (budz˙et – budget, dolar – dollar) and sports (bokser – boxer, dz˙okej – jockey). Dictionaries of foreign words from mid-19th century noted about 180 borrowings from English, (Man´czak-Wohlfeld, 1995: 31), and over 530 in the 1930s (Koneczna, 1936) mostly connected with film, mass culture, and sport (Walczak, 1994). A rapid expansion of borrowings from English was noted after the Second World War (e.g. bestseller, dubbing, fan, folder, happy end, hot-dog, quiz, rock, thriller), although the administration made efforts to forbid or to ridicule the use of English words which penetrated from behind the iron curtain (Buttler, 1981; Grabowska, 1972; Rybicka, 1976). Since the 1970s, the number of Anglicisms in Polish has been constantly growing already reaching about 700 in the early 1980s, together with a growing number of internationalisms coined in English, French or German (Se˛kowska, 1993; Walczak, 1994; Waszakowa, 1994, 1995; Witaszek-Samborska, 1992, 1993). Most borrowings from English underwent morphosyntactic and phonetic adaptation (Fisiak, 1961, 1962, 1963, 1964, 1968, 1970). It is currently estimated that the number of English borrowings in Polish exceeds 3500 words (Man´czak-Wohlfeld, 1994, 1995) and still expands, especially in the domains of technical and scientific terminology (e.g. serwer, skaner, trend), business (e.g. biznes, boom, leasing, menadz˙er) and daily life (e.g. piercing, hamburger) (Markowski, 1992). After the transformation of 1989 a fashion for English resulted in the tendency to overuse English (Waszakowa, 1994, 1995). There were even changes in the spelling of some Polish names, graphically ‘styled’ to look more English (k would be replaced by c, ks by x, f by ph and w by v). This tendency was connected with the high status of English (Sawicka, 1995). The intense contact resulted not only in lexical borrowing from English, but also in numerous semantic shifts and frequent instances of structural borrowing (Walczak, 1994; Waszakowa, 1994, 1995, 2005; Otwinowska-Kasztelanic, 2000a, 2000b; Rusiecki, 2002; Waszakowa, 2005; Zabawa, 2007, 2008, 2009, 2012).

42

Par t 2: Def ining Le x ical Crosslinguist ic Simil ar it y

Having analysed the processes which underlined the formation of contemporary vocabularies of English and Polish, there is no doubt that even across typological boundaries languages overlap to a certain extent. The processes responsible for that are complex language contact phenomena, involving cultural influences of various kinds, the evolution of both languages, as well as the intensity of the contemporary influence of high-prestige English. The result of language contact, crosslinguistic similarity is highly interesting from the point of view of language teaching and learning and will be analysed in the following chapters.

4

Crosslinguistic Lexical Similarity

Introduction Convergence and divergence processes within languages triggered by multiple language contact and use lead to the existence of numerous structural and lexical similarities between European languages. The existence of such common patterns and words is often referred to as crosslinguistic similarity, which can be generally defined as the degree of congruence between the languages involved (Jarvis & Pavlenko, 2007). Traditionally, the notion of crosslinguistic similarity is inherently connected with language typology, which results in a classification of structural types among languages (Comrie, 1981). The typological distance, connected with objective language similarity, is determined by linguistic genealogy, i.e. relatedness of languages within the same family. At the syntactic level, typological distance is determined by the degree of formal structural parallels in the grammatical patterns, while at the lexical level by the number of shared word forms inherited from a common ancestor, including both free and bound morphemes (Hall et al., 2009: 157). It is assumed that typologically close languages, and also those unrelated typologically (e.g. English and Polish), share a considerable number of words which are very similar or even identical. This chapter deals with such crosslinguistically similar lexis, in particular with cognates, internationalisms, and false friends. It also discusses how formal lexical crosslinguistic similarity can be measured. Finally, it presents a database of English-Polish cognates that is used in further research in Part 4 of the book.

Degrees of Crosslinguistic Lexical Similarity What is understood by cognate words In historical and contact linguistics, cognates will be understood as words that have a common etymological origin. The very word ‘cognate’ 43

44

Par t 2: Def ining Le x ical Crosslinguist ic Simil ar it y

derives from the Latin cognatus – blood relative. From this perspective, crosslinguistic lexical similarity involves ‘words of similar structure and of similar or in many instances identical meanings in the various languages of the Indo-European group (…) recognised as cognate, that is, of common origin’ (Pyles & Algeo, 1982: 79, original emphasis). In linguistic studies, cognates are often defined as ‘words in different languages which have descended from a common parent word’ (Schmitt, 1997: 209), or words that have similar meaning, spelling and form, and have been inherited from the same ancestor language (Whitley, 2002). Cognates defined in accordance with etymological criteria include those words which have descended from earlier derivatives of the Indo-European language family. Numerous instances of such cognates, similar in form and meaning will be present in typologically close languages (e.g. verbs können in German, kunne in Danish and can in English). But in the case of languages typologically distant, the common etymological Proto-Indo-European origin is often hard to detect on the basis of the word form (e.g. młyn in Polish, mill in English, Mühle in German, and moulin in French). Although etymologically such words are cognates, most language users will not find them strikingly similar. When we compare the definitions of cognates used by historical linguists on the one hand, and linguists dealing with language processing and acquisition on the other, there seems to be a lot of discrepancy in how cognates are defined. Contrary to historical linguists preoccupied with genealogy, in psycholinguistic studies the term ‘cognate’ may be defined only in accordance with formal criteria. For instance, it may be reserved for orthographically identical words that share form and meaning in two languages (e.g. bed in Dutch and in English) (Lemhöfer & Dijkstra, 2004; Lemhöfer et al., 2008). Such definition is useful for typologically close languages abundant in similar words, but in the case of Polish and English, cognates understood in such a way are restricted to a small subset of words (e.g. radio, zebra) and some recent unassimilated borrowings. In psycholinguistic studies, cognates may also be defined as phonologically and/or orthographically similar translation equivalents (e.g. De Groot & Comijs, 1995; Van Hell & Djikstra, 2002), or as translation equivalents with high orthographic overlap with no regards to phonology (Schepens et al., 2012). All these definitions of cognates are far from the etymological criteria of cognateness, so it seems that the category of cognates can be defined very broadly to include any ‘words with similar form and meaning in two languages’ (de Bot, 2004: 19).

Internationalisms Translation equivalents (cognates) which exist in several unrelated languages (e.g. English optimism, Polish optymizm, German Optimismus, French optimisme) are sometimes called internationalisms, especially in the

Crosslinguist ic Le x ical Simil ar it y

45

Central- and East-European linguistic tradition (Kolwa, 2001; Maćkiewicz & Siatkowski, 1992). The definition of such words was coined in the mid1950s, thanks Soviet and German linguists interested in ‘international words’ in language teaching: ‘International words can be regarded as similar in their orthographic and phonological form, are fully or partly semantically congruent, and express meanings/notions internationally understood’ (Internacional’nymi slovami javliajuttsja shodnye do stepeni uznanvaemosti v ortografičieskom ili fonologičieskom otnošenii slova s polnost’ju ili častično obščej semantikoi, vyražajuščie ponjatia meždunarodnogo značenija. Akulenko, 1958, after Kolwa, 2001: 5, English translation mine). Defined in such functional terms, internationalisms should be present in at least three unrelated languages, should be one another’s translation equivalents and their written or spoken form should allow for their recognition even by a person who does not know the language. Internationalisms (Polish: internacjonalizmy, Russian: internacionalizmy, German: Internationalismen) are often based on Greek and Latin stems and come from a variety of sources, including the historic interrelatedness of Indo-European languages, mutual borrowing, borrowing from non-European languages, linguistic conventions in institutions such as the church, scientific and technological standardization, and international news exchange. Some were coined from Latin and Greek morphemes in the 19th century (e.g. English humanism, socialism, Polish humanizm, socjalizm) and spread over various European languages. In a wider sense, the category of internationalisms can also include internationally used morphemes like anti-/anty(Maćkiewicz, 1984).

Defining cognates in language acquisition research In language acquisition research, as suggested by Odlin (1989: 78), ‘the term cognate is broader than the use of the term by historical linguists.’ In the light of the discussion in the sections above, it is perfectly justified to disregard genealogy in defining cognates, but to take into account the formal and semantic similarity of given words. Clearly, there are cognates of genetic and non-genetic origin, and language learners and users rarely know the difference between true cognates, loanwords and accidentally similar words, so essentially, all three categories may have the same effects in language learning (Jarvis, 2009). Therefore, Ringbom (2007: 73) proposes a very broad definition: ‘cognates in two languages can be defined as historically related, formally similar words, whose meanings may be identical, similar, partly different, or occasionally, even wholly different.’ He states that cognates may be found in ‘related languages, and to a minor extent also in unrelated languages because of possible loanwords’ (Ringbom, 2007: 73). While the definition captures various similarity relations between word forms, the type of historical relation between them remains unimportant.

46

Par t 2: Def ining Le x ical Crosslinguist ic Simil ar it y

For the purpose of the present study cognates will be defined as words similar in their form and meaning, which have descended from a common parent word, have been borrowed from Lx language to Ly language, or are internationalisms borrowed independently by languages Lx, Ly and Lz. They do not have to be identical since they have been adapted to fit the rules of spelling, phonology and morphosyntax of Lx, Ly and Lz. Defined in such a way, cognates exist in European languages, both close and distant typologically (e.g. English – pessimistic, computer, zebra; Polish – pesymistyczny, komputer, zebra; German – pessimistisch, Computer, Zebra; Spanish – pesimista, komputer, cebra, etc.). Such a broad, definition of cognates will be assumed for all studies presented in the following chapters.

Defining false friends, or deceptive cognates It is obvious that not all the words which have a similar form in two languages share exactly the same meaning. In any two languages there will be pairs of words which sound or look similar but whose meaning is different. An example of such a word is the English adverb actually, whose meaning is entirely different from the meaning of the Polish adverb aktualnie (currently). Such deceptive words, that is words in two languages which have the same, or a similar orthographic form but not have common semantics are commonly called ‘false friends.’ Their similarity may be purely accidental (e.g. pain in English and pain French) or, in the case of genetically related words, it may result from the dynamics of language change and divergence processes. Such pairs of words in two languages, whose stems and affixes show incidental grapho-phonemic similarity, are variously named in the literature on the subject, for instance, faux amis du traducteur (in French: false friends of the translator) (Odlin, 1989), irreführende Fremdwörter (in German: misleading foreign words) or fałszywi przyjaciele (in Polish: false friends). In psycholinguistics, depending on the modality studied, such pairs of words can also be called ‘interlanguage homophones’ de Bot (2004), or ‘interlingual homographs’ (e.g. Lemhöfer & Dijkstra, 2004; Lemhöfer et al., 2008). Haugen (1956: 47) called such pairs of words ‘synonymous diamorphs’, while Lado (1957: 83) used the term of ‘deceptive cognates’ and defined them as ‘words that are similar in form but mean different things.’ Wełna (1977) finds Lado’s (1957) definition of ‘deceptive cognates’ too restrictive, arguing that only words which do not show any semantic correspondence can be called ‘deceptive cognates.’ He calls for a broader definition: ‘a deceptive word is a word in the lexicon of some language which exhibits easily identifiable graph-phonemic similarity to a word (words) in another language. The similarity is accompanied by either partial correlation in the meaning or by the absence of any direct semantic correspondence’ (Wełna, 1977: 79). Along similar lines, Rusiecki (2002) discusses the existence of ‘absolute false friends’ and ‘partial true friends.’ While the meaning of partial true friends may

Crosslinguist ic Le x ical Simil ar it y

47

overlap, absolute false friends are those words in one language whose meaning is entirely different than it would seem from their similarity to words in another language.

The interaction of meaning and form The discussion above points to an important fact: false friends and cognates, both genetic and nongenetic, are similar to a varying degree. As de Bot (2004: 19) admits, ‘similarity between the words of the two languages may vary somewhat and there is no clear definition of what “similar” is, both in orthography and in phonology.’ Ways of dealing with the degrees of formal and semantic similarity have been put forward in different studies. For instance, Rusiecki (2002: 73) suggested a functional category of ‘interlingual analogues’ defined as ‘a class of words in any language (Lx) that sound and/or look familiar to speakers of another language (Ly).’ The class would comprise both false friends and cognates of various degrees of similarity. Berthele (2011) proposed that genetic cognates in typologically related languages can be grouped into a radial category with prototypical, identical (or almost completely identical) examples at its centre and increasingly different examples further on. The model takes account of both allographic and allophonic variation between different pairs of words, but the category has fuzzy boundaries since it remains unclear where interlingual cognateness ends. Figure 4.1 presents the model with examples from Danish and English verbs. Cognates of genetic and nongenetic origin can also be placed on a continuum, or a scale of crosslinguistic formal and semantic similarity (Jarvis, 2009; Ringbom, 2006, 2007). In terms of semantic relations, the scale includes cognates (words with similar form and meaning), partial cognates (words sharing formal similarity but only partial semantic overlap) to false friends or deceptive cognates (words sharing formal similarity but no

Danish English

give give

mena mean

kunna can

skulle should

Figure 4.1 Degrees of cognateness as a radial category (adapted from Berthele, 2011: 201)

48

Par t 2: Def ining Le x ical Crosslinguist ic Simil ar it y

semantic overlap). In terms of formal relations there is a continuum of crosslinguistic similarity, crosslinguistic difference/contrast relations, and crosslinguistic zero relations between word forms (Ringbom, 2006, 2007). Thus, the meaning of the formally similar words in two languages may be equivalent, the Lx meaning may be narrower and included in a wider range of Ly meanings, the Ly meaning may be narrower and included in a wider range of Lx meanings, the meanings may overlap to a certain extent, or remain in complete contrast. The scale is schematically presented in Figure 4.2. It is also possible to present the degrees of similarity between PolishEnglish cognates and false cognates in terms of the formal and semantic criteria devised by Jarvis (2009). Ranking words according to the degree of form and meaning overlap, we can established six groups of words, as presented in Figure 4.3. According to the classification, true Polish-English cognates fall into the categories of ‘the same or similar form and the same meaning.’ There are c.a. 3000 such true cognates in various semantic fields (Otwinowska-Kasztelanic, 2004). Absolute false friends (deceptive cognates) belong to the categories of ‘the same or similar form and dissimilar meaning.’ An example of a pair of absolute false friends is eventually, whose meaning is entirely different from the meaning of the Polish word ewentualnie (perhaps). Finally, partial false friends, or (partial cognates) belong to the category of ‘similar meaning and the same or similar form.’ Partial false friends are words which have several meanings, one of which may overlap across the two languages, while the other meanings are different. A good example of such partial true/false friends is the English word declare and the Polish deklarować. One of its meanings is the same as in Polish (‘declare help’ ‘deklarować pomoc’), but two other meanings are different (‘declare war’ – ‘wypowiadać wojne˛’, or Form

Meaning

Same Similar Dissimilar

Same Sw. finger En. finger Sw. offer (victim; offer) Eng. offer Sw. strand (beach) Eng. strand

Similar Sw. fader En. father Sw. ben (leg, bone) En. bone Sw. gris (pig) En. grease

Dissimilar Sw. hjul En. wheel Sw. täcka (cover) En. thatch Sw. växt (plant) En. Waist

c og n at e s partial cognates/ partial false friends words of related meaning but different in form fal se frien ds

Figure 4.2 The continuum of similarity/difference relations in form and meaning between English and Swedish words (adapted from Jarvis, 2009: 107)

Crosslinguist ic Le x ical Simil ar it y

49

Form Same

Meaning

Same Similar

Dissimilar

Pl. radio En. radio Pl. student (at university) Eng. student (at school or university) Pl. hymn (national anthem) Eng. hymn (religious song)

Similar Pl. alfabet En. alphabet Pl. adres (of a place) En. address (noun or verb) Pl. nowela (short story) En. novel (noun - book or adjective - new)

cognates partial cognates/ partial fal se friends false friends

Figure 4.3 Polish-English words presented as the continuum of similarity/difference relations in form and meaning

‘nothing to declare’- ‘nic do oclenia’) (Otwinowska-Kasztelanic, 2004). Two current dictionaries of Polish-English false friends define over 1000 items including absolute and partial false friends (Rudolf, 2003; Szpila, 2005).

Conceptual equivalence Pavlenko (2009a) draws attention to yet another aspect of the semantic continuum, namely to the non-equivalence of the concepts related to the meanings of words in the different languages. Lexical concepts, understood as multimodal mental representations that include visual, auditory, perceptual and kinaesthetic information stored in the memory of language users may differ from language to language in terms of structure, boundaries and prototypicality. This means that even close translation equivalents may not always constitute conceptual equivalents, even in the case of formally similar items. The conceptual relationships within the continuum of words in two languages fall into three broad categories. Conceptual equivalence, or near equivalence is understood as a relationship where ‘linguistic categories mediated by languages A and B share both category structures and boundaries’ (Pavlenko, 2009a: 133). An example of such equivalence are cognates tas in Dutch and tasse in French: their meaning and conceptual representation are the same, roughly similar to the English cup. Partial (non) equivalence refers to those pairs where concepts partially overlap, but are not the same in two languages. Two or more categories in language A can be nested (subsumed) within a larger category of language B, or a single category in language A may be roughly divided between the conceptual categories in language B.

50

Par t 2: Def ining Le x ical Crosslinguist ic Simil ar it y

For example the Russian category chashaka, or the Polish category of filiz˙anka (restricted to a china cup with a handle) are nested within the broader English category of cup, which additionally includes plastic and paper cups. Finally, in the case of conceptual non-equivalence a linguistic category of language A does not have a meaning/conceptual counterpart in language B (Pavlenko, 2009a). Along similar lines, Wełna (1977: 73) proposes that English-Polish pairs of words which are related on a grapho-phonemic level may remain in four different types of semantic relations: (1) Equivalence e.g. En. alphabet : Pl. alphabet. (2) Inclusion a) with the En. unit having more meanings e.g. En. fiction : Pl. fikcja, and b) with the Pl. unit having more meanings e.g. En. protocol : Pl. protokół. (3) Overlapping e.g. En. platform : Pl. platforma. (4) Contrast, e.g. En. lecture : Pl lektura. Effectively, equivalence refers to cognates, inclusion and overlapping to partial cognates (Ringbom, 2007), partial true friends (Rusiecki, 2002) and to partial (non) equivalence in Pavlenko’s (2009) terms. Contrast relations refer to false friends, or deceptive cognates (Lado, 1957).

Describing the Degrees of Cognate Similarity Measuring the objective formal similarity of cognates For the purpose of linguistic studies and psycholinguistic experiments it is often necessary to establish objective measures of cognate similarity. For instance, similarity of form and meaning between pairs of words is often estimated on the basis of ratings by competent bilingual judges (e.g. Tokowicz et al., 2002; Dijkstra et al., 2010). The formal similarity of word pairs to be used in experiments may also be captured in terms of orthographic word neighbourhoods: how many words differ from a given word in one, or two letters (e.g. English bud – but, Polish dom – dym). What may also help in establishing the degrees of similarity is word length, morphological family size and bigram frequency, which is the orthographic typicality of a word in the context of its own language (Lemhöfer et al., 2008). Recently, Levenshtein distance (LD), a string metric used in information theory, has gained interest as a tool for measuring orthographic similarity between pairs of cognates in various languages (e.g. Beijering et al., 2008; Schepens et al., 2012; Vanhove, 2014). Levenshtein distance counts the minimal number of substitutions, insertions, and deletions to edit one string into another of any length. In the case of words from different languages, LD

Crosslinguist ic Le x ical Simil ar it y

51

helps in establishing degrees of orthographic (formal) similarity between pairs of cognates. While LD = 0 means full formal overlap for identical words, the greater the LD, the less similar the words are. To give an EnglishPolish example, the pair guitar – gitara has LD = 2, while the pair physical – fizyczny has LD = 8. The LD measure can be applied to account for the distances between the orthographic and phonological forms of words. Sometimes the raw LD is used (e.g. Berthele, 2011) in the case of written stimuli (letters) and aural stimuli (phonemes). However, it seems that for written words the raw LD does not simulate letter correspondences exactly, as it does not take into account the length of the words involved. Hence, normalised LD can be calculated (Schepens et al., 2012: 159), which simulates orthographic similarity between 0 and 1, where the score is 1 minus the relation between the distance (minimal number of insertions, deletions and substitutions) and the lengths (maximal length of the source expression and the length of the destination expression). Schepens et al. (2012) present a convincing proof that the LD metric is a reliable measure of cognate similarity. They applied their normalised LD metric to a translation database (Euroglot professional 5.0), aiming to approximate the orthographic similarity of translation pairs from different languages and to study the distribution of cognates in six European languages. As the first step, translation equivalents from a huge translation database were retrieved to identify cognates between all 3–8 letter words in English, German, Dutch, Spanish, French and Italian. In order to verify whether the conceptual structure of the cognate database was consistent with that of language users, the automatically extracted translation equivalents were compared with ratings by competent bilingual judges obtained in studies by Tokowicz et al. (2002) and Dijkstra et al. (2010). Since normalised LD highly correlated with those ratings (correlations of .88 and .96), orthographic similarity norms obtained with normalised LD are claimed to be reliable measures of orthographic similarity for given word pairs (Schepens et al., 2012: 160). Next, the normalised LD measure was applied to translation equivalents in six languages in order to automatically identify orthographic similarity distributions of cognates across language pairs. The simulation was carried out only for words with a word length between 3 and 8 characters and normalised LD smaller or equal to 0.5. The simulation rightly predicted that languages belonging to the same language family (e.g. Germanic or Romance languages) share more cognates than languages belonging to different families. The exceptions were combinations of English with French, Spanish or Italian because English bears high lexical similarity to Romance languages. The simulation revealed that the combination English–French shared 2727 identical cognates, which is comparable to languages from the same typological family, where the average is 2374. From the simulations with

52

Par t 2: Def ining Le x ical Crosslinguist ic Simil ar it y

normalised LD it turned out that most languages share a relatively high number of cognates with only one other language (e.g. Dutch with German, Italian with Spanish, English with French) and vice versa, so the similarity is truly symmetrical (as theorised by Ringbom, 2007). So, cognate distributions based on the orthographic similarity continuum simulated with normalised LD reveal that the number of form-similar and identical cognates is correlated with language typology. In addition, the simulation reveals that a particular Romance, or Germanic language generally shows a high overlap within the whole Romance or Germanic language family, but also with English: ‘in contrast to the general case, only English shares many cognates with all other languages’ (Schepens et al., 2012: 162). Unfortunately, the LD measure also has some limitations, as it may not always simulate human perceptions of word-to-word correspondences. Since the measure operates on orthographic input only, it does not take into account some grapheme-to-phoneme regularities and is sometimes unable to approximate human similarity judgments. For instance, Schepens et al. (2012) admit finding in their study some word pairs as cotton – katoen, which were overlooked by the computerised model, but were rated as similar by bilingual human judges. The probable mechanism involved in recognising grapheme-to-grapheme correspondences can be explained by certain phonological regularities between the languages involved. If a grapheme (or a grapheme cluster) from one language corresponds to a phoneme realised as a grapheme in another language, then such a correspondence of graphemes in the two languages can be regarded as systematic. Bilingual/multilingual language users probably perceive those most regular and frequent grapheme-to-grapheme correspondences and those which simultaneously correspond to a common phoneme in both languages. Therefore, it seems that incorporating some grapheme-tophoneme mappings may add to the psychological reality of the computerised measure of orthographic similarity used.

Establishing the measure of formal similarity for Polish-English cognates As a reliable model of formal crosslinguistic similarity the LD function was also used for describing the orthographic relations between PolishEnglish cognate pairs. However, taking into account the limitations of the normalised LD (Schepens et al., 2012), a similarity index used here was established in a slightly different way than the one described above: it also accounts for some grapheme-to-phoneme regularities between the languages. First, the Normalised Levenshtein Distance (NLD) metric (Otwinowska & Szewczyk, 2013), calculates the relation between the LD (minimal number of insertions, deletions and substitutions) and the length of the longest word of in the pairs of Polish-English cognates. Thus, NLD simulates orthographic

Crosslinguist ic Le x ical Simil ar it y

53

similarity between 0 and 1, where score 0 means full formal overlap, while score 1 means the highest degree of difference. Normalised LD ( NLD) =

LD maximum length of the longer word

Additionally, the model accounts for some grapheme-to-phoneme regularities. We assume that from the language learners’ perspective, some grapheme-to-grapheme mappings between English and Polish may be more transparent than other. An example of such a frequently occurring correspondence is the English grapheme cluster and its Polish counterpart , both highly frequent and systematically realised as the /f/ phoneme. On the other hand, not all correspondences are systematic and frequent. The English grapheme corresponds to several phonemes in English and also to several graphemes in Polish, such as , and . That is why it cannot be said to form a highly systematic correspondence with the Polish . To account for such regularities, the NLD index for English-Polish cognates includes some most probable grapheme-to-phoneme mappings present in English and in Polish. The NLD index incorporates changes for the most frequent grapheme-to-phoneme mappings, but only on condition that the change decreases the measure. As for English, the changes introduced to NLD are based on an index of the strength of particular grapheme-to-phoneme associations (Berndt et al., 1987), as presented in Table 4.1. As for Polish, we decided to recalculate only those graphemes and grapheme clusters which were typical of Polish orthography and resulted in dual orthographic realisations of the same phoneme. These included the / orthographic distinction, realised as /z˛/, and the / distinction realised as /u/. We decided to resign from such changes for the / distinction, as it would have increased the NLD measure in the case of pairs such as e.g. chemia – chemistry. Additionally, the changes included single phoneme realisations mapped onto graphemic clusters: as /tz˛/, as //, and as //.

Similarity categories within the database of Polish-English cognates A corpus of 2747 Polish-English cognates was extracted from a dictionary of cognate pairs (Otwinowska-Kasztelanic, 2004). Further on, the pairs of words from the database were grouped in several ways in order to account for the categories of semantic and formal similarity. The whole database of 2747 words contained 1873 nouns, 283 verbs and 591 adjectives listed together with information about their semantic overlap. Adverbs, as derived from adjectives, were not taken into account in the study.

54

Par t 2: Def ining Le x ical Crosslinguist ic Simil ar it y

Table 4.1 Grapheme-to-phoneme mappings introduced to NLD English grapheme, or grapheme cluster

Corresponding phoneme

Probabilities for grapheme-to-phoneme correspondences

Example from the database of cognates











/k/ /kt/ /f/ /l/ /m/ /n/ /f/ /p/ /k/ //

1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 0.952 1.000 1.000 0.885

dock verdict effective dollar grammar tennis graphic opponent qualification shocking discuss globetrotter version dyslexia

/s/ /t/ /v/ /ks/

Source: Probabilities from Berndt et al. (1987: 8–9)

It was established that in the database there were 1992 words which displayed full semantic overlap, including 1463 nouns, 132 verbs and 402 adjectives, as presented in Figure 4.4. Next, the whole sample of 2747 words was sorted according to their NLD. In the database there were 294 words with NLD score 0, i.e. words which displayed full formal similarity (e.g. alarm, alibi, media, medal, minimum, problem, protest, referendum, sport, turban, ultimatum). The sample included 290 nouns (50% of them concrete and 75% of them with full semantic overlap), but only four adjectives (blond, incognito, khaki, super) and no verbs. The

Figure 4.4 Percentages of nouns, verbs and adjectives a) in the whole database of English-Polish cognates (N = 2747); b) within cognates with full semantic overlap (N = 1997)

Crosslinguist ic Le x ical Simil ar it y

55

full list of words with NLD = 0 can be found in Appendix 1 (p. 246). These words belong to semantic fields of everyday life: clothes and cosmetics, entertainment, sport, business and food. They also include exotics, (e.g. curry, gyros, igloo, cf. Chapter 3). Some of them are relatively recent loanwords that have not been adapted graphically to the Polish orthographic system for reasons of language prestige and market value of foreign goods (e.g. dealer, spray). Since formal similarity of cognate pairs may aid to their recognition, it can be assumed that the lower the NLD, the easier the English words should be for a Polish learner (this hypothesis will be explored in Chapter 12). That is why, in the next step the proportion of words possibly least and most difficult from the learners’ point of view were extracted from the database. It was assumed that words with NLD lower than 0.2 would be potentially easiest to learn as they are most similar to their Polish counterparts, and that words of NLD higher than 0.7 would be the most difficult. The distribution of words in the cognate database according to their NLD and word category is presented in Figure 4.5. In the databease there are 669 English words whose NLD to the Polish equivalent is smaller than 0.2. These cognates, potentially easiest to recognise by a Pole, included 646 nouns, which constituted 34% of all the nouns from the database, and 23 adjectives, which constituted 4% of adjectives. There were no verbs in this group. On the other hand, among the 142 English words of NLD higher than 0.7, i.e. words potentially most difficult to recognise, there were 53 nouns, which constitutes 3% of all the nouns in the database, 61 verbs, which is as much as 22% of all the verbs in the base and 28 adjectives, which is 5% of the adjectives in the database. The distribution of the NLD feature points to an interesting regularity. We can assume that English language learners may first notice cognate nouns, due to their higher

Figure 4.5 The distribution of nouns, verbs and adjectives in the database sorted according to the NLD categories

56

Par t 2: Def ining Le x ical Crosslinguist ic Simil ar it y

formal similarity to their Polish orthographic neighbours. On the other hand, adjectives and verbs may be more difficult to notice and learn because they tend to be less similar to their Polish counterparts. Concluding, we can say that the database of English-Polish cognates is an important source of cognate words, sorted and classified to be used in future studies. The words from the database are going to be used in experiments presented in Chapters 12, 13 and 14.

Part 3 Lexical Crosslinguistic Similarity in Use

5

Crosslinguistic Similarity and Crosslinguistic Influences

Introduction Apart from contact situations that can result in borrowing and language change (Chapter 3), languages are also said to be ‘in contact’ if they are used alternately by the same persons (Crystal, 2006), which involves both reception and production. Such micro-scale language contact in bilinguals and multilinguals is sometimes labelled language mixing (e.g. Romaine, 1989; Thomason, 2001). Language mixing is an umbrella term for a range of phenomena, including (among others) code-switching, code-mixing, code alternation, and various transfer processes. Code-switching, which is inter-sentential (occurs across sentence boundaries), involves inserting an utterance in another language within a speech event. Code-mixing, which is intrasentential, involves switches within a single sentence. Code alternation is the use of two (or more) languages by the same speaker and involves the use of the different languages known to a bilingual or multilingual in different situations, etc. (Thomason, 2001). Of our special interest here are the processes of transfer, or mutual influences between the languages that an individual can use. In contact linguistics literature, a distinction is often made between borrowing transfer and substratum/substrate transfer (Thomason & Kaufmann, 1988; Odlin, 1989). Borrowing transfer is defined as ‘the influences found in the use of a person’s native language that are due to the subsequent acquisition of another language’ (Odlin, 1989: 165), which may even lead to language attrition. Substrate/substratum transfer is understood as ‘the influences found in a second language that are due to the native language’ (Odlin, 1989: 169). In fact, if more languages are involved these processes may be more complex and their nature goes far beyond ‘interference’ from one language when producing another. Therefore, in language acquisition studies transfer is often labelled as ‘crosslinguistic influences.’ 59

60

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

This chapter briefly discusses the early view of transfer as interference and introduces the notion of crosslinguistic influences in relation to bilingualism and multilingualism. Then it focuses on the phenomenon of lexical transfer. It stresses the role of objective, perceived and assumed crosslinguistic similarity, and consequently, the role of cognates and false friends in bilingual/multilingual language acquisition, as well as in reception and production of languages. The chapter finishes with a discussion of lexical transfer at the lexemic, lemmatic and conceptual level. Throughout the chapter the terms of acquisition and learning may be used synonymously, but learning will always refer to classroom situations.

Transfer and Crosslinguistic Influences Language transfer as interference At first glance it may be difficult to distinguish between transfer and codeswitching as both involve language mixing. However, whereas switching normally involves overt contrasts between two languages, transfer phenomena are often more difficult to detect (Odlin, 2009). In common understanding, transfer and code-switching phenomena are regarded as a kind of ‘contamination’ of the language used (see Chapters 1 and 2). This negative attitude towards transfer often results from the lack of a deeper understanding of bilingualism and multilingualism (Jarvis & Pavlenko, 2007), but was cemented, paradoxically, thanks to the first systematic research on bilingualism. Weinreich (1953) defined language transfer as linguistic interference, where the elements which interfere with the main language elements are ‘against the rules’ of that language. Therefore, many early studies of transfer concentrated on ‘deviations from the norms of either language that occur in the speech of bilinguals as a result of their familiarity with more than one language’ (Weinreich, 1964: 1). In this definition Weinreich implied that negative aspects of transfer were more prominent than the positive ones. Weinreich was, in fact, the father-founder of the way transfer was understood in the next decades and studied within the Contrastive Analysis Hypothesis (CAH) framework (Odlin, 2003). The CAH framework (Lado, 1957) emphasised the role of differences between languages leading to learning difficulties, and hence the role of errors seen as bad habits transferred from the L1. Such transfer where negative interaction of areas in L1 and L2 would result in errors was named ‘negative transfer’ and ‘positive transfer’ would be the interaction of similar areas resulting in facilitating effects on L2 acquisition. The CAH explored structural relationships between languages to solve the problem of how to prevent interference from the native language when learning another. Its pedagogical conclusions were that language teachers should focus on the structural

Crosslinguist ic Simil ar it y and Crosslinguist ic Inf luences

61

differences between languages understood as synonymous with difficulties, and drill the new L2 habits to eradicate interference from the learner’s L1. As stressed today, there are numerous reasons why the linguistic notions of similarity and difference cannot be equated in a simple way with the psychological notions of ease and difficulty of learning. As the CAH predictions and conclusions were rather bold, a counter-reaction followed during the 1970s. Thanks to early experimental evidence on the morpho-syntactic development of linguistically heterogeneous L2 learners of English, the hypothesis that L1 influences L2 acquisition was largely rejected (Dulay & Burt, 1974; Krashen, 1982), which resulted in abandoning transfer studies (Ellis, 2008). They were reappraised at the end of the 1970s and the beginning of the 1980s, together with the growing interest in the positive aspects of transfer and in the area of lexis.

The notion of crosslinguistic influence By the 1980s, attention had shifted to redefining the notion of transfer and exploring the conditions under which transfer takes place (Gass & Selinker, 1983; Selinker, 1983; Thomason & Kaufman, 1988). Studies also aimed at explaining interaction of various systems of languages leading to different paths of acquisition, avoidance and overproduction of certain elements when learning another language (Gass, 1988). Odlin (1989: 29, emphasis added) presented his well-known definition stating that ‘transfer is the influence resulting from the similarities and the differences between the target language and any other language that has been previously (and perhaps imperfectly) acquired.’ Odlin’s definition clearly acknowledges that the learner/user of a language may be in fact, multilingual. This has paved the way to studying transfer in the acquisition of more than one language. The changes in the approach to transfer in the 1980s were also reflected in the terminology. Sharwood-Smith and Kellerman (1986) suggested using the term of ‘crosslinguistic influence’ (CLI) rather than ‘transfer’ as more adequate to capture complex phenomena. They argued that the term is ‘theoryneutral, allowing one to subsume under one heading such phenomena as “transfer”, “interference”, “avoidance”, “borrowing” and L2-related aspects of language loss and thus permitting discussion of the similarities and differences between these phenomena’ (Kellerman & Sharwood Smith, 1986: 1). This new nomenclature reveals a shift in focus from emphasising how the knowledge of a previous language (originally only considered in terms of the learner’s L1) impeded the learner’s progress in the new language, to a more neutral stance. Whereas transfer was previously perceived as a one-way phenomenon with L1 as the source and L2 as the target, CLI indicates that all languages available to the learner mutually affect one another in various directions. Hence the new notions of reverse transfer, bidirectional transfer and lateral

62

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

transfer began to be used to describe the dynamic processes of CLI. The interest in reverse transfer, or bidirectional transfer (Pavlenko & Jarvis, 2002), was probably influenced by psycholinguistic research into the integration of the bilingual lexicons (e.g. De Groot et al., 1994; De Groot, 1995; De Groot & Comijs, 1995; Kroll & Stewart, 1994; Van Hell & De Groot, 1998). The main bulk of research into reverse transfer of L2 on L1 started with Cook’s (1991) notion of multicompetence (see Chapter 1). Cook’s (2003) volume on the effects of L2 on L1 was the first on reverse transfer and the contributions presented evidence on reverse transfer in areas as diverse as collocational knowledge (Laufer, 2003), lexical diversity (Dewaele & Pavlenko, 2003; Jarvis, 2003), semantic and conceptual changes (Pavlenko, 2003; Kecskes & Papp, 2003) pragmatics (Cenoz, 2003) and syntax (Cook et al., 2003). The impact of L2 English on language users’ L1 Polish was studied by Ewert (2009). The discussion so far has been generally constrained to the L2 context, but the very term ‘crosslinguistic influence’ implies that all the languages known to the language learner/user can influence one another in the case of multilinguals (Odlin, 2009). Such transfer relations which involve L2 influence on L3 in language learning, sometimes called ‘lateral transfer,’ were first documented by Ringbom (1987) in his study of L1 Finnish - L2 Swedish and L1 Swedish - L2 Finnish learners of English as L3, and confirmed by Williams and Hammarberg (1998) and numerous other studies that followed (see Chapter 8 for an exhaustive discussion): e.g. Bardel and Falk (2007) and Falk and Bardel (2011) on syntactic structures; Cenoz et al. (2003) and Gabrys´Barker (2005) on lexical influences; Gabrys´-Barker (2012) on syntax and lexis; Wrembel et al. (2010) on phonology, just to mention several publications. De Angelis and Selinker (2001: 43) called the process of L2 influence on L3 ‘interlanguage transfer’ and defined it as ‘the influence of a non-native language on another non-native language – i.e. the documented transfer from one interlanguage to another.’ Whether the traditional conceptualisation of transfer is extensive enough to cover for phenomena specific to multilingualism remains an open question. Herdina and Jessner (2002) argue that the whole term of crosslinguistic influence needs redefining as ‘crosslinguistic interaction’ in order to account for the qualitative differences of multilingual CLI in comparison with CLI in bilinguals. De Angelis (2007) proposes a new classification of crosslinguistic influence into ‘one-to-one’ (L1-L2), and ‘many-to-one’ (combined CLI) because transfer in multilingual processing may come from more than one language. It occurs either when two or more languages interact with one another and influence the target language together, or alternatively, when one language influences another, which in turn influences the language being acquired. Considering all the developments of the notion of CLI, in current literature on the subject CLI is defined very simply as ‘the influence of a person’s knowledge of one language on that persons’ knowledge or use of another

Crosslinguist ic Simil ar it y and Crosslinguist ic Inf luences

63

language’ (Jarvis & Pavlenko, 2007: 1). The definition does not specify the number of languages involved, their level of advancement and the order of acquisition. However, the old term of transfer is still widely used. It is often stressed that ‘transfer’ and ‘crosslinguistic influence’ can in fact be used interchangeably, as the most convenient cover terms to refer to the phenomenon (e.g. De Angelis, 2007; Jarvis & Pavlenko, 2007; Odlin, 2003, 2009; Ringbom, 2007).

Crosslinguistic Similarity and the Levels of Lexical Transfer Objective and perceived/assumed crosslinguistic similarity CLI may be enhanced by crosslinguistic similarity. The role of congruity between languages was first captured in the ‘Transfer to Somewhere Principle’ (Andersen, 1983), which emphasises that typological similarity between languages with regard to a particular syntactic feature is a necessary condition for transfer to take place. Although the principle paved the way to considering transfer as a process based on the learner’s linguistic awareness, typology cannot account for all instances of transfer (Murphy, 2003). Thus, an important step further was the consideration of the learner’s perception of congruity between languages, i.e. the notion of psychotypology (Kellerman, 1977, 1983, 1986, 1995). Psychotypology refers to how the learner perceives the relationship between particular structures and words of the L1 and the L2, and whether the learner recognises the forms as congruent (see Chapter 7 for discussion). If languages are perceived as similar with regard to a particular structure, transfer will be more likely to occur; if they are perceived as dissimilar, this may lead to avoidance (Kellerman, 1983, 1986). Sometimes, transfer will not take place: ‘learners may not be able to capitalise on crosslanguage correspondences because some type of “thinking for speaking” may be beyond individual awareness’ (Kellerman, 1995: 142). This is an important statement pointing to the fact that learners may not even be aware of crosslinguistic similarities, or may fail to make mental associations, or ‘interlingual identifications’ (Odlin, 1989: 113) between language structures or words. As suggested, ‘the actual similarity or dissimilarity of forms and meanings is only one factor at work in transfer; the judgement of each individual learner matters as much’ (Odlin, 2002: 260, original emphasis). Therefore, CLI largely depends on the judgements of language learners/users, which are always subjective, as opposed to the objective levels of congruence between languages (Odlin, 2003). According to Ringbom (2006, 2007), it is difficult, if not impossible, to measure objective similarity between languages, although such attempts have been made (e.g. Ard & Homburg, 1983). Since an objective estimation

64

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

of language distance can also be misleading about the likelihood of transfer (Odlin, 1989), Ringbom (2006: 38) introduces the notions of ‘perceived’ and ‘assumed similarity,’ as opposed to objective similarity of language items and forms. The objective and perceived similarity can, in fact, be fully corresponding when the learner perceives the objective similarities, yet such accordance is reserved to language close typologically. If measured, objective similarity would necessarily be symmetrical, i.e. the same for a pair of two languages (see Chapter 4 for the discussion on measuring objective crosslinguistic similarity of cognates by Schepens et al., 2012). Objective similarity could imply that speakers of language X find understanding language Y as easy, as in the same way speakers of language Y find understanding language X. However, perceived similarity in language comprehension is not symmetrical, as ‘speakers of language X may find it easier to understand language Y than speakers of language Y to understand language X’ (Ringbom, 2007: 7). Ringbom and Jarvis (2009) describe three reasons for this asymmetry between the objective and the perceived: the learners may fail to notice the objective similarity, they may also misinterpret the nature of the similarity that they notice, or they may notice similarity where there in fact is none. Perceiving formal similarity to an existing L1/Ln word occurs in the process of comprehension and before the learner is conscious of the word’s meaning. This is where the ‘assumed similarity’ comes into play. If the learner finds the form of a lexical item familiar (in the case of cognates or false friends), he also tends to assume that its meaning is similar to another one existing in his or her lexicon, although such assumptions may be wrong (cf. false friends and partial cognates discussed in Chapter 4). In production, learners encode their ideas into language structures they already know by assuming that there exist some similarities between the languages in question. This assumption of similarity may prove so fundamental that wherever there is L1/Ln influence in target language production, it is always due to the assumed similarity between the L1/Ln and the target language (Ringbom & Jarvis, 2009). Concluding, although the objective similarity accounts for the learner’s rate of acquisition and the amount of time needed to achieve certain proficiency levels (Ard & Homburg, 1992; Odlin, 1989; Ringbom, 1987; Ringbom & Jarvis, 2009), the assumed, or perceived similarity, may exert influence on language learning and use, and this influence may be more direct than that of objective crosslinguistic similarity (Ringbom & Jarvis, 2009; Odlin, 2006). Where similarities are merely assumed in typologically distant languages, this may lead to a considerable risk of errors.

Overall transfer, item transfer and system transfer Learners’ perceptions and assumptions interact with the continuum of crosslinguistic similarity-difference relations, labelled as ‘similarity,

Crosslinguist ic Simil ar it y and Crosslinguist ic Inf luences

65

difference/contrast, and zero relations’ (Ringbom, 2006, 2007). A ‘similarity relation’ occurs when an item (or pattern) in the target language is ‘perceived as formally and/or functionally similar to a form or pattern in L1 or some other pattern known to the learner’ (Ringbom, 2007: 5). A natural tendency, especially in the early stages of learning another language, is to try and establish a one-to-one relationship with L1 or any other language known. Across typologically related languages obvious candidates for such relationships will be cognates, as defined in Chapter 4. At the early stages they will facilitate both comprehension and learning, even if some of the relationships may later turn out to be false, or partially false (Ringbom, 2006). Full-scale similarity of form and function is rare except for languages very close typologically such as Swedish and Norwegian, or Polish and Slovak, which in practice are mutually comprehensible. Numerous groups of European languages (e.g. Slavonic and Germanic, or Germanic and Romance) stand in ‘contrast relations.’ Here, the learner perceives the target language pattern as ‘in important ways differing from an L1 form or pattern, though there is also an underlying similarity between them’ (Ringbom, 2007: 6). A Polish learner learning English or German will encounter both similarities and differences leading to positive and negative transfer. For instance, although syntactic patterns may be different, lexical similarities will be present even across typological boundaries in the form of borrowings which found their way to those languages for various historical reasons (see Chapter 3). Thus, in contrast relations, the learner encounters problems in learning the language, but is generally aware of the existence of the system and will eventually be able to figure out how this language works. If languages are typologically very distant, they remain in ‘zero relations’ where, apart from some language universals, ‘the learner finds nothing at all that is relevant to L1 as the learning progresses’ (Ringbom, 2007: 6). At the early stages the target language patterns and items seem to have very little relation to the L1 or any other language known to the learner: the concepts present in the target language may be unknown, or there is a totally different writing system, like in the case of a European trying to learn Chinese. Consequently, the learner has to spend a considerable amount of time to figure out how the language works and there is not much positive transfer involved. Any possible similarities have to be pointed out to the learner in an explicit way. The continuum of similarity/difference relations as perceived by the learner result in the different levels of transfer that may occur, labelled as overall transfer, item transfer and system transfer. Ringbom (2007: 57) defines ‘overall transfer’ as an umbrella term including learners’ reliance on both the formal similarity of individual items and the extent of functional equivalence between the languages. Overall transfer depends on the learners’ judgment of similarity between their languages, ‘beginning from a common alphabet, phonemes in common and similar phonotactics, over the

66

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

division into grammatical categories (case, gender, word classes) to the number of cognates and other lexical similarities’ (Ringbom, 2006: 39). The more closely related the languages are, the more overall transfer can be found. If languages stand in contrast or zero relations, overall transfer does not take place. ‘Item transfer’ is based on the underlying perceived similarity of form combined with the assumed similarity of meaning. It often occurs at the early stages of learning, as learners with a limited knowledge of the target language rely on form rather than on meaning and establish oversimplified equivalence hypotheses (Ringbom, 2007: 55). Whenever learners perceive a target language item to be similar in form to an Ln item, they assume those two are also similar in meaning or function. Such interlingual identifications may have a positive influence on comprehension, but may also lead to negative transfer. This is the case of false friends and partially similar cognates whose use in the target language is restricted syntactically and morphologically in comparison with the language known. Partial identification of two lexical systems may lead to negative transfer because the assumed similarity gives the learner a false sense of security, which may cause problems at the phonological, morphological and semantic level (Odlin, 1989). Thus, negative transfer could be better defined as the ‘absence of relevant concrete (positive) transfer, leading to subsequent wrong assumptions about cross-linguistic similarities’ (Ringbom, 2007: 30-31). This is an important shift in the approach to crosslinguistic influences in language learning when one thinks back to assumptions of the CAH and the search for differences between languages that would hinder the learning process. The mechanism of item transfer is such that formal similarity in the shape of similar spelling, pronunciation, the morphological make-up of words or multiword structures triggers the activation of lexical forms from one of the languages in target language production (Ringbom & Jarvis, 2009). Importantly, when more languages are involved, learners activate lexical items from a language that they perceive as typologically closest to the target language. This transfer of form can be observed in language switches, coinage of hybrids, and the use of false friends (for numerous examples see Ecke, 2014). In complete language switches, the form of the code-switched word may be either entered into the target language production unmodified, or adapted partially to the rules of the target language (e.g. an L2 word with L3 morpheme or pronunciation). Item transfer may also involve transfer of meaning realised as sematic extensions and calques in the form of hybrids and blends (see Chapter 3) which do not exist in the target language (Ringbom, 2006). Yet, as the knowledge of the target language increases, those oversimplified assumptions are re-examined and accordingly revised. System transfer, or ‘procedural transfer’ (Ringbom, 2001, 2006, 2007) involves the assumption of functional, or semantic similarity when no formal similarity is involved. The learner assumes sameness of meaning, but

Crosslinguist ic Simil ar it y and Crosslinguist ic Inf luences

67

not necessarily of form, which often results in negative transfer due to noncongriuty of language systems. Procedural transfer nearly always occurs from the L1 rather than another non-native language, as ‘apparently semantic properties must be well internalised, preferably automatised, in order to be transferred’ (Ringbom, 2007: 56). However, this may also depend on the L2 proficiency and the stage of L3 learning, where ‘foreign language effect’ (as discussed in Chapter 9) may come into play.

Lexemic and lemmatic transfer The mental lexicon is understood as the database containing all the words in the mind of the language user. Since the mental lexicons of bilingual and multilingual language users constitute an amalgam of all the languages known (Pavlenko & Jarvis, 2001, 2007), they are different from those of monolingual speakers of a given language and are prone to crosslinguistic influences at the lexical level. The phenomenon of lexical transfer can be defined as ‘the influence that a person’s knowledge of one language has on that person’s recognition, interpretation, processing, storage and production of words in another language’ (Jarvis, 2009: 99). Research on the architecture of the mental lexicon suggests that word knowledge is stored at three general levels of representation: lexemes, lemmas and concepts (see for example the Distributed Features Model of the mental lexicon discussed in Chapter 6). A word’s formal, i.e. orthographic and morpho-phonological specifications are stored at the lexemic level, while semantic and syntactic information is stored at the lemmatic level. The third level is the level of concepts (to be discussed in the next section). The three dimensions of word knowledge will also influence transfer processes in the minds of bilinguals and multilinguals. Consequently, crosslinguistic influences may happen at the three level of representation: lexemes, lemmas and concepts. The lexemic-lemmatic-conceptual types of transfer may also be labelled as the occurrence of form, frame, and meaning-based CLI (e.g. Ecke, 2014) When transfer involves the graphemic and the phonological structure of a particular word form, including its morphology, it can be called lexemic transfer (Jarvis, 2009). Thus, whenever lexical transfer reflects lexeme-level links, it will be enhanced by formal (lexemic) crosslinguistic similarities (e.g. similar orthography of cognates and false friends). This is due to the fact that mental links between words within a language, which are formed within, or across levels of representation (i.e. from lexeme to lexeme, or from lexeme to lemma) can also be established across two or more languages. The links between lexemic representations are of varying strengths and depend on factors such as crosslinguistic similarity and frequency of use. That is why, thanks to formal morpho-phonological similarities at the level of lexemes, the strength of the links in the case of cognates and false friends may be increased (Jarvis, 2009). The strengths of the links also affect the degree to

68

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

which lexemes will be activated during the processing and use of a given language by a multilingual (de Bot, 2004). A discussion of bilingual and multilingual lexical processing can be found in Chapter 6. The scope of lemmatic transfer involves the semantic and syntactic properties of words (Jarvis, 2009). The semantic properties include mental associations between lemmas and concepts (accounting for polysemy and the semantic ranges of words) and mental associations between lemmas and other lemmas (accounting for synonymy, antonymy and other types of word to word associations). Syntactic properties concern word category (e.g. noun, verb, etc.), its subcategorisation frame and other features (e.g. grammatical gender). When collocations, the knowledge of compound words and fixed expressions are transferred, they also come under the umbrella of lemmatic transfer. Lemmatic transfer will take place if, for instance, the learner assumes the syntactic frame of a verb in the target language on the basis of the syntactic frame of its cognate verb in another language (e.g. Hall et al., 2009; Ecke, 2014). According to Jarvis (2009), lexical transfer takes place via one of the two mental processes in the bilingual/multilingual mental lexicon: the formation of learned crosslinguistic associations and processing interference. The first of these mental processes involves establishing conscious mental links between stored representations of elements (e.g. lexemes or lemmas) from two, or more, different languages. Again, the formation of learned crosslinguistic associations should be easiest in the case of cognates, where formal similarity leds to common semantics. Consequently, cognates should be easiest to learn for those learners who are able to recognise them. This issue is discussed in Chapters 6–10 and tested in Chapter 12. On the other hand, processing interference is the main cause of observable manifestations of lexemic transfer (Jarvis, 2009). It results from the activation of words (lemmas or lexemes) in one language when the speaker is trying to use another language, often independently of previously formed mental associations between specific pairs of words from two (or more) different languages. Therefore, in language production, lexemes can emerge as lexical intrusions into the selected language in the form of false friends, or other formally similar lexical words. Such intrusions can also take the form of crosslinguistic lexical blends, created online due to processing interference from competing activated lexemes, as well as in unintentional language switches probably caused by a high level of activation of the intruding lexeme (see e.g. Dewaele, 1998; Ecke, 2014; Poulisse, 1999; Ringbom, 2001; Williams & Hammarberg, 1998 and discussion in Chapter 9). This probably reflects the pre-existing mental associations between words that are superficially similar. Besides learned crosslinguistic associations, lexical transfer can also occur in the form of intentional and strategic uses of the other language known. These include intentional code-switches to another language or borrowing

Crosslinguist ic Simil ar it y and Crosslinguist ic Inf luences

69

lexemes from another language to fill in the gap in the lexical knowledge, which is the subject of studies presented in Chapters 13 and 14.

Conceptual and semantic transfer ‘Concepts reflect the level of thought and experiential knowledge, and they consist of various types of mental images, image schemas, mental scripts and forms of knowledge that are organised into structured categories of thought and categories of meaning’ (Jarvis, 2009: 100-101). Concepts store the visual, auditory, perceptual, and sensory-motor kind of information, as well as other types of schemata, which are all organised into categories and integrated into a person’s ‘web of knowledge’ about the world. Most, though not all, of the mental concepts we have are mapped onto words, but can be linked to multiple lemmas (Jarvis & Pavlenko, 2007). According to Pavlenko (1999), languages may differ not only on semantic, but also on the conceptual level, i.e. in the structure of particular conceptual categories and links between words and concepts. Since word-related knowledge in the mind is stored at the level of lexemes, lemmas and concepts, transfer can take place not only at the lexemic and lemmatic level, but also at the conceptual level. Pavlenko and Jarvis (2001: 288) define conceptual transfer as ‘all instances where conceptual representations are involved in linguistic manifestations of cross-linguistic influence.’ They argue that linguistic transfer is largely driven by the conceptual need to find adequate linguistic means of expression in the less developed language. In other words, conceptual transfer involves the reliance on L1-mediated concepts because the bilingual’s or multilingual’s mind may lack some conceptual representations in the less developed languages (Jarvis & Pavlenko, 2007; Pavlenko, 1999). In situations of oral communication when the time available for response is limited, language users/learners commonly rely on the implicit links between the L2/Ln forms and their L1 conceptual representations, which may lead to crosslinguistic influence in the form of conceptual transfer. In the classroom situation, if enough time is provided, learners may use their explicit knowledge and come up with the target forms congruent with the L2/Ln conceptual representations in their minds. Conceptual transfer is to this end connected with Slobin’s (1996) idea of ‘thinking for speaking’, which suggests that speakers organise and structure their thoughts in language-specific ways before verbalisation, even if at a more general level their patterns of thinking are not language-specific. Although the notions of conceptual transfer and thinking for speaking overlap, the latter is narrower: ‘thinking-for-speaking is based on the assumption that language specificity occurs only during speech-planning processes, whereas the scope of conceptual transfer is not restricted by such an assumption’ Jarvis (2011: 3). Thus, conceptual transfer deals with more than only selecting and structuring

70

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

information in the mind, as it is also concerned with the possible differences in how information is stored in the mind: ‘reliance on L1-mediated concepts leads to instances of CLI theorised here as conceptual transfer. In turn, reliance on the links established between L1 words (e.g. synonymy), or between L1 words and L1-mediated concepts is theorised here as semantic transfer’ (Jarvis & Pavlenko, 2007: 120, original emphasis). To exemplify what may lead to conceptual and semantic transfer let us give an example of a Finnish learner who said in English *He bit himself in the language, instead of ‘in the tongue.’ The interference from Finnish is semantic in nature and can be labelled as semantic transfer because Finnish (similarly to Polish) has only one polysemous word for both tongue and language, i.e. a verbal label linked to discrete concepts. Thus the interference occurs not at the level of conceptual representations but at the point of semantic mapping. On the other hand, if an English-speaking learner of Russian asks for a paper cup referring to it as chashka, the transfer is both semantic (involves an inappropriate link) and conceptual. It reveals inadequate knowledge of what the concept of chashka (cup) involves: a paper cup in Russian would be referred to as a small glass, stakanchik, which is a concept different from the one of a cup and non-existent in English. On the other hand, Odlin (2005, 2008) calls for caution in calling transfer conceptual, even if it is defined very broadly as ‘cross-linguistic influence involving relativistic effects’ (Odlin, 2008: 310). Therefore, he makes a distinction between ‘conceptual transfer’ and ‘meaning transfer,’ which he suggests as a neutral term between semantic and pragmatic influence (Odlin, 2005: 5). It will involve cases of influence from the semantics or pragmatics of the native language (or L2 in L3 acquisition) on the target language production, which may not constitute a case of conceptual transfer. For instance, Poles who otherwise speak excellent English, often tend to say: *We were at the theatre with my brother yesterday, instead of My brother and I went to the theatre yesterday (Odlin, 2008: 311). Still, it is hard to believe that Poles fail to grasp the conceptual difference between singular and plural referents. Rather than that, Poles base their utterances on the predominant Polish pattern where a first-person plural inflection (-my) can serve as a marker of first person plural reference: Wczoraj bylis´my z bratem w teatrze (Yesterday were-1st PL with brother-INS at theatre-LOC). Therefore, we can conclude that ‘all conceptual transfer involves meaning transfer but not all meaning transfer involves conceptual transfer’ (Odlin, 2008: 311).

Conclusions: CLI and Linguistic Relativity Transfer is an ultimately subjective phenomenon, both as a conscious and a subconscious process. The scope of lexical transfer largely depends on learner judgements of perceived and assumed similarity. Thus, lexemic,

Crosslinguist ic Simil ar it y and Crosslinguist ic Inf luences

71

lemmatic and conceptual transfer often stem from psychotypology, rather than sheer typology of languages used and acquired by a given language user/learner. The notion of conceptual transfer is an important corollary to emphasising the role of learner judgments and perceptions in CLI, referred to as ‘the differences in the way languages predispose their speakers to conceptualise experience’ (Kellerman, 1995: 137). This calls back to the notion of language relativity, i.e. the view that language reflects the thought-world of its speakers, which dates back to Wilhelm von Humboldt (1767–1835). It also bases on the hypothesis of Whorf et al. (2012) of linguistic relativity, which states that users of different grammars develop somewhat different views of the world: Linguistic relativity is often defined as the hypothesized influence of language on thought. Such influence might affect either comprehension or production, and such influence could, of course, affect comprehension or production in a second language (or a third, a fourth, etc.); moreover, the influence might be where the L1 is influenced by the L2. (Odlin, 2005: 5, original emphasis) A knowledge of additional languages has the power to transform someone’s worldview, which facilitates certain ways of thinking (Jarvis & Pavlenko, 2007). Therefore, we must accept the fact that although certain transfer processes are generalisable, various levels of transfer strongly depend on individual learner factors. If we accept the fact that transfer, both linguistic and conceptual, stems from individual learner factors, it also seems obvious that individual learner judgements and perceptions will determine the way in which crosslinguistic similarity affects language processing, language acquisition and language use.

6

Cognate Vocabulary in Language Processing

Introduction Among words which share common semantics across languages, cognates have a special status. As discussed in the previous chapters, the number of shared cognate forms determines the level of crosslinguistic lexical similarity between the languages involved. Thus cognates, i.e. words whose form and meaning considerably overlap across two or more languages, will constitute one of the most salient similarities between the lexicons of those languages, as perceived by the learner (see Chapter 5). Other words that are formally similar to their L1 equivalents are false friends (interlingual homographs) where crosslinguistic orthographic similarity between forms interferes with the divergent meanings (see Chapter 4). Thanks to formal similarity, both cognates and false friends are somewhat ‘special’ for bilinguals, multilinguals and language learners. That is why these words are within the range of interest of experimental psycholinguistics and language acquisition studies. The present chapter often adopts a bilingual perspective, mostly due to the character of the studies discussed. First, it deals with the status of cognates and false friends in psycholinguistic experiments carried out mostly on bilingual subjects, where cognates helped to demonstrate that both languages of a bilingual are constantly active when processing words in one or the other language. In laboratory experiments on word processing, bilinguals process cognate words faster and more effectively than non-cognates. Such studies are presented together with the possible explanation of the processes and conclusions concerning the structure of the bilingual and multilingual mental lexicon.

The Bilingual Mental Lexicon The methods of study The mental lexicon is understood here as a database of all the words in the mind of a language user, and lexical access is the process of entering the 72

Cognate Vocabul ar y in L anguage Processing

73

mental lexicon to retrieve information about words (Dijkstra, 2005: 180). Models of the monolingual mental lexicon have been created since the late 1960s, while models of the bilingual mental lexicon followed much later when scientists started to be interested in the bilingual processing of language. The main question to be answered was whether words in the two languages known to the bilingual are stored together in one system, or in separate, isolated ‘compartments.’ The problem of words storage in the bilingual mental lexicon, i.e. the integration/separation debate, goes back to Weinreich’s (1953) early distinction between ‘subordinative’, ‘compound’ and ‘co-ordinate’ relationship between the L1 and the L2 mental lexicons of a bilingual. In ‘subordinative’ bilingualism, L2 word forms are connected to L1 meanings via primary connections to L1 forms. In ‘compound’ bilingualism the L1 and L2 forms are connected at the meaning level. In ‘co-ordinate’ bilingualism there are separate systems of form-meaning links for each language, which means that the lexicons are separate. According to the model, the different types of bilinguals and their mental lexicons are associated with different kinds of learning experience. Following Weinreich’s model, traditionally, there were two alternative theories concerning bilingual lexical representation. The ‘common memory’ theory postulated a single integrated memory system for both languages, which reflects Weinreich’s ‘compound’ bilingualism. The ‘multiple-memory’ theory claimed that words from each language are represented separately, which reflects Weinreich’s notion of ‘coordinate’ bilingualism. For a long time it was not possible to state which of the model was correct. At first, conflicting research evidence was produced in favour of both: the common, conceptual, language-independent storage of words, as well as the separate, language-dependent storage of words. In laboratory conditions lexical access is investigated with the use of various types of tests and tasks. Most of these are online experiments, where the participant is sitting in front of a computer screen and is instructed to press buttons/keys in reaction to stimuli, white the time of reaction is measured. The experimental tasks include word, or picture naming, word translation, and semantic categorisation, i.e. deciding if the word belongs to a particular semantic category. Another type of task is lexical decision, where subjects have to decide whether a particular string of letters is a word in the target language or a non-word (pseudoword) in a given language (Traxler, 2012). It is precisely experiments on bilingual processing of cognates with the use of reaction time (RT) measurements which helped to modify the monolingual models of the mental lexicon and create a bilingual model. Retrieving information from the mental lexicon about the characteristics of a word takes up to a few hundred milliseconds. It was assumed that measuring RT of how bilinguals read and perceive cognates and non-cognates (matched for various features) would help to determine if cognates and non-cognates are

74

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

processed in a different way. The RT differences between cognates and noncognates would mean that the fact that cognates exist in two languages known by a bilingual have an impact on their processing (Traxler, 2012). If the processing of cognates was ‘boosted,’ in comparison to non-cognates, it would mean that their representation was accessed simultaneously from two languages. This would also prove the nonselective, integrated ‘compound’ storage of words. This finding would help in the debate on the architecture of the bilingual mental lexicon, as discussed below.

The Revised Hierarchical Model of the bilingual mental lexicon On the basis of picture naming and word translation experiments Potter et al. (1984) proposed the Hierarchical Model of bilingual representation, in which words in L1 and L2 are stored at two levels: the surface level corresponding to their formal features and at the common, deep level of concepts which stores their sematic properties (compare with the discussion in Chapter 5 of lexical storage at the level of lexemes, lemmas and concepts). It was developed into the Revised Hierarchical Model (RHM) proposed by Kroll (1993), which postulates that during lexical processing, a bilingual accesses the two levels of the mental lexicon: the lexical (formal) store through lexical access and the conceptual (semantic) store through conceptual access. Consequently, we can differentiate the conceptual and the lexical processing routes. If processing occurs solely at the lexical (formal) level, it follows the lexical processing route. If processing occurs between the lexical level and the conceptual level and involves conceptual access, then it follows the conceptual processing route (Kroll, 1993). Kroll and collaborators (Kroll, 1993; Kroll & Stewart, 1994) tried to work out the connections between the levels, as presented in Figure 6.1. On the basis of word translation experiments, the RHM postulates that the conceptual store has connections to the L1 lexical store (L1 conceptual connection) and the L2 lexical store (L2 conceptual connection) which hold lexical representations at the lexical level. At the lexical level, the L1 and L2 lexical stores are also interconnected. The strength of connections in

Figure 6.1 The asymmetrical model (adapted from Kroll, 1993: 69)

Cognate Vocabul ar y in L anguage Processing

75

Figure 6.1 are illustrated by the dashed lines, representing weaker connections, while the continuous lines are used for stronger connections. The L1 conceptual connection is stronger, and the L2 conceptual connection is weaker. As regards the lexical connections, the strength differs depending on the direction of processing. The difference in the strength of connections depends on the magnitude of activation. For example, every time a representation at the lexical level connects with its conceptual representation, the connection between them becomes stronger. This also applies also to the lexical connections: each time a bilingual translates from L1 to L2 or from L2 to L1, the lexical connection in the particular direction of translation becomes stronger. Moreover, the conceptual connection in L2 (the target language) is less developed than its equivalent in L1. The situation changes with the development of the second language proficiency. When the L2 conceptual connections are more frequently used, they become stronger. Bilinguals whose L1 and L2 conceptual connections are equally strong are referred to as balanced bilinguals (see Chapter 1), as opposed to unbalanced bilinguals whose L2 is less developed.

The Mixed-Representational Model Simultaneously with the RHM, De Groot (1993) put forward a MixedRepresentational Model. She suggested that various pairs of words across languages differ in the extent to which they share semantic information (conceptual features) in L1 and L2 and also the particular formal characteristics (word type features): frequency, familiarity, definition accuracy, cognateness and imageability/concreteness (De Groot, 1992, 1993; De Groot et al., 1994; De Groot & Comijs, 1995). Cognateness involves phonological and/or orthographical similarity and translation equivalence. Imageability describes the ease with which a mental image of a particular word can be recalled: there are concrete words of high imageability and abstract words of low imageability. Words characterised by high definition accuracy are easy to define, while it is difficult to think of a definition for words characterised by low definition accuracy. Word frequency and familiarity are highly similar, but the difference lies in the objectivity of the variable. Familiarity refers to the subjective opinion of a language user as to how well-known a certain word appears to be. Word frequency, on the other hand, is a more objective, statistical variable and describes how often a particular word is used in a language. If we assume that various classes of words differ in terms of their form and meaning mapping at the two levels, their processing will also differ. In a series of experiments on Dutch-English bilinguals, De Groot and colleagues proved that all word type characteristics described in the MixedRepresentational Model have an impact on the way words are processed. For instance, high frequency words will be easier to retrieve, and concrete words of high imageability will be processed faster than abstract words. Most

76

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

interestingly however, the experiments pointed to robust effects when bilinguals process cognates, relative to non-cognates and false friends (De Groot, 1992, 1993; De Groot et al., 1994; De Groot & Comijs, 1995).

Research on the Processing of Cognates The cognate facilitation effect in various RT experiments As predicted, it turned out that cognates are easier and faster to translate than non-cognates, which is known as the ‘cognate advantage’ or the ‘cognate facilitation effect.’ The effect has been observed in word translation both in the forward (L1 to L2) and backward (L2 to L1) direction (De Groot, 1992, 1993; De Groot et al., 1994; De Groot & Comijs, 1995). This led researchers to the assumption that there might be asymmetries in the lexical access, i.e. in the processing routes for cognates and non-cognates and that cognates might be specially represented in the bilingual mental lexicon. It was argued that cognates in bilingual processing may be accessed not only via their representation in the conceptual memory (the conceptual processing route in RHM), but also more directly via the lexical level of representation (the lexical processing route in RHM), i.e. through their form (De Groot, 1992; De Groot et al., 1994; De Groot, 1995; De Groot & Comijs, 1995; Kroll & Stewart, 1994). For example, in explaining the faster translation of cognates relative to non-cognates, De Groot (1992) suggested that lexical-level representations of cognates in the two languages are more strongly interconnected than in the case of non-cognates. In consequence, cognates are more likely to be translated via a direct route connecting the two lexical representations, without relying on a slower route via conceptual representation of the translated word. If we assume that cognate processing in both directions proceeds via the lexical route of access to the mental lexicon that is strongly connected with the formal features of words, this explains why cognates are processed faster than non-cognates. For instance, if translating involves a conscious search through the lexicon, formal crosslinguistic similarities might strongly facilitate the process. However, this was only one of the possible explanations (as discussed below). The cognate facilitation effect was shown not only in translation, but also in other types of RT experiments. For example, Dijkstra et al. (1998) demonstrated that in lexical decision tasks performed by Dutch-English bilinguals in L2 on a set of identical cognates, interlingual homographs (i.e. false friends), and matched controls, the reaction times to the cognate stimuli were significantly shorter than those on the matched controls. In the study by Lemhöfer and Dijkstra (2004), Dutch-English bilinguals performed a lexical decision task involving isolated word recognition. In the experiment cognates were recognised faster than the matched English and Dutch

Cognate Vocabul ar y in L anguage Processing

77

controls, whereas interlingual homographs (false friends) did not show any effects, relative to Dutch controls. All in all, the cognate facilitation effect has been corroborated by a whole range of recognition and production tasks performed by bilinguals, including such diverse methodologies as progressive demasking (Dijkstra et al., 1999), picture naming (Costa et al., 2000) and word association (Van Hell & De Groot, 1998). The effect has been obtained also with multilinguals (trilinguals), situated in an L1 setting and performing lexical decision or word association tasks in their L1 (Van Hell & Dijkstra, 2002).

The Distributed Lexical/Conceptual Feature model The interpretation of earlier experiments led to merging De Groot’s (1993) Mixed-Representational Model and Kroll’s (1993) RHM into a single Kroll and De Groot’s (1997) Distributed Lexical/Conceptual Feature Model. It locates representational elements at the two levels: conceptual and lexical, like in Kroll’s (1993) RHM. However, each lexical and conceptual element is represented as a node, and each word as a collection of nodes, as in De Groot’s (1993) Mixed-Representational model (Figure 6.2). The level of lexemes includes only the aspects of form, i.e. orthography, while the level of concepts stores aspects of word meaning. Language-specific lemmas mediate between the two levels and reflect the mapping between the lexical and conceptual levels.The model, which can be easily extended to include more than two languages (L1, L2, L3, etc.), assumes that various classes of words differ in terms of their form and meaning mapping, i.e. they differ in the extent to which they share representational elements at the two levels. In the model, any translation equivalents in L1 and L2 will share nodes at the conceptual level. However, cognates will also share nodes at the formal, lexical level, which gives them an advantage in how they are recognised and processed. Together with the results of earlier experiments, this assumption, led to further conclusions concerning the cognate advantage in word processing.

Figure 6.2 Kroll and De Groot’s (1997) lexical/conceptual feature model (adapted)

78

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

Explaining cognate advantage in bilingual/multilingual word processing The studies of cognates have pointed to a very high degree of connectivity and dynamic interplay between the L1 mental lexicon and mental lexicons in additional languages. The cognate facilitation effect is treated as a very strong proof of the integration if the lexicons in a bilingual mind and their ‘compound’ character, to use Weinreich’s (1953) terminology. It is justified to say that bilinguals access words in their mental lexicons in a language-nonselective way. However, it was still to be proved whether the effect for cognates depended on their formal similarity across languages, or the way their semantic representations in the lexicon were accessed. The two possible explanations for the facilitating effects in the processing of cognates were put forward by Dijkstra and his colleagues. First, Dijkstra et al. (1999) proposed that all words which share common orthographic representation across two languages (i.e. cognates and interlingual homographs/ false friends) could be recognised faster, relative to L2 control words. Dijkstra et al. (1999) suggested that only thanks to their formal similarity bilinguals may have a greater experience with these orthographic forms. This type of explanation was not confirmed by data (see for example Lemhöfer & Dijkstra, 2004; Van Hell & De Groot, 2008). Experiments have shown that the advantage held only for cognates, but not for interlingual homographs/ false friends. This led to the formulation of an alternative explanation of the advantage of cognates. This alternative explanation does not assume that cognates are represented in any way different than non-cognates. The second explanation of cognate advantage in processing by Dijkstra and colleagues pertains to the semantic representation of cognates in the mental lexicon. It bases on the observation that bilinguals recognise words in a language-nonselective way, i.e. during the initial stages of word recognition word candidates from several languages are often co-activated (see Dijkstra & Van Heuven, 2002, for an overview). When a cognate is presented, orthographic (formal) representations of the word in both languages are activated, which then leads to the activation of the corresponding semantic representation. In the case of cognates the semantic representation is common for the two languages. Since two orthographic representations of the cognate words activate the same semantic representation, the activation strength is greater, relative to when a non-cognate word is processed. Such facilitated activation of conceptual representation speeds up the translation process (Van Hell & De Groot, 2008), and also feeds back to orthographic representation speeding up the lexical decision process (Lemhöfer & Dijkstra, 2004). False friends/interlingual homographs also lead to the non-selective activation of two orthographic representations of the word in two languages. However, these representations do not lead to the activation of convergent semantic representations, giving false friends no advantage over

Cognate Vocabul ar y in L anguage Processing

79

orthographically dissimilar words. Thus, cognates and false friends are processed differently, despite their formal orthographic similarity to their counterparts in another language.

The BIA+ Model The explanation of the cognate effect described above fits in with such recent models of isolated-word processing in bilingual memory such as the Distributed Features Models (see Figure 6.2), and interactive accounts of bilingual word recognition, i.e. the Bilingual Interactive Activation (BIA) Model, its successor BIA+ (Dijkstra & Van Heuven, 2002), and extension to the Multilingual Interactive Activation Model (Dijkstra, 2003) when three or more languages are involved in the processing. In sum, the BIA+ model rests on several ideas. It assumes that the bilingual lexicon is integrated across languages and is accessed in a language nonselective way, while word recognition is affected by cross-linguistic orthographic, phonological and semantic overlap. The processing is initiated by visual input from the given text and proceeds in a bottom-up manner from letter features, to letters, and then to words. For languages that share similar orthographies activation is parallel and results in competition at the lexical and sublexical levels. During recognition, several lexical candidates are activated depending on their similarity to the input word, regardless of the language. The activation depends on other individual factors such as use frequency, subjective frequency, the participant’s L2 proficiency, etc. Only the layer of language nodes is responsible for the selection of the language because the language nodes inhibit words in the non-target language in a top-down manner (cf. Dijkstra & Van Heuven, 2002). Because the BIA+ model also introduces lexical and sublexical phonology to account for the observed patterns of orthographic and phonological interaction, recognition is affected not only by crosslinguistic orthographic and semantic similarity effects, but also by crosslinguistic phonological overlap. If access to the bilingual mental lexicon is non-selective, then the presentation of an isolated cognate word in one language co-activates orthographic, phonological, and semantic representations of parallel languages of a bilingual, or all languages known to a multilingual. This is most likely when the words that are presented are highly similar in orthography, semantics, and phonology. Since cognate translation equivalents share orthographic, semantic and phonological information, while non-cognate translations share only semantic information, the activation of the three codes in cognate translations is responsible for the advantage of cognates, compared to non-cognates. Importantly, since activations depend on individual factors, this implies that the phonological and semantic L2/Ln codes may be delayed in their activation in relation to the L1 codes. This is related to the type of task and the language level of students. Also, when a word recognition task is inserted

80

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

into a sentence context, the process is sensitive to syntactic and semantic context information; for example, context information might inhibit or reduce the activation of lexical candidates or induce a more flexible activation of lexical candidates in the two languages. So, the BIA+ model points to the fact that processing may be influenced by various factors related to individual language users/learners and the role of linguistic and non-linguistic context in which word recognition and processing takes place.

Cognates in Experimental Studies on Vocabulary Learning The keyword method, word association, and picture association Since an abundance of evidence points to the connectivity of the mental lexicons in L1 and L2/Ln, it seems that cognates should also be easy to learn. Indeed, vocabulary learning studies that manipulated word-type features showed considerable effects of the cognate status on learning words (De Groot & Keijzer, 2000; Ellis & Beaton, 1993; Kroll et al., 1998; Lotto & De Groot, 1998). These findings also significantly correlated with shorter RTs in the case of cognates, relative to non-cognates. Some experiments with the word association, and picture association vocabulary learning methods corroborated the assumption that cognates are easier to learn. However the keyword method, otherwise effective, did not show good results (see an overview by De Groot & Van Hell, 2005). Although the advantage of the keyword method over rote rehearsal has been demonstrated for vocabulary in various languages in laboratory and classroom conditions, the method turned out unsuitable for learning cognates. The keyword method is a mnemonic technique in which learning takes place in two stages. In the first stage, learners learn to associate the new word (e.g. cot) to a keyword in their native language, i.e. a word that looks, or sounds like the new word to be learned (e.g. kot - Polish for ‘cat’). In the second step, the learner is encouraged to create a mental image in which both the keyword and the L1 translation of the novel word interact (here: a cat in the cot). The method enhances learning as it activates both the verbal system and the image system in memory. However, for cognates it is not necessary to find a keyword that looks or sounds like the L2 word and the mental images triggered by words with close orthographic neighbours in L1 and L2 are convergent. Therefore, it ‘seems an unnecessarily laborious and ineffective method for learning cognates, particularly considering the large advantage that cognates have over noncognates in the more straightforward learning methods’ (De Groot & Van Hell, 2005: 12). On the other hand, vocabulary learning within the word association and picture association paradigm revealed a huge advantage of cognates over

Cognate Vocabul ar y in L anguage Processing

81

noncognates. The methods consist in presenting the learner with pairs of stimuli – two words or a word and picture (De Groot & Van Hell, 2005). The stimuli differ in length and word-type features (e.g. frequency, imageability), and may be presented auditorily or visually. The testing phase involves a cued recall task where one of the elements in a pair, i.e. the cue is presented, while the participant has to recall the second element of the pair. Alternatively, testing may involve recognition. Participants are presented with stimulus pairs that they have, or have not learned earlier and they must indicate whether the presented stimulus pair is ‘old’, i.e. was presented during the learning phase or ‘new’, i.e. not presented during the learning phase. Two studies (Kroll et al., 1998; Lotto & De Groot, 1998) contrasted the word association and picture association methods manipulating various word-type criteria. These studies report robust cognate effects: cognates are easier to learn and retrieve. The two elements within the cognate pairs used in studies by Kroll et al. (1998) and Lotto and De Groot (1998) were typically similar in orthography and in phonology, so it was concluded that the learner’s recognition of either type of relationship (phonological or orthographic), or both, might facilitate learning. More importantly, cognate status materialised not only in the word association condition, but also in the picture association condition, and was equally large in these two conditions. Although Lotto and De Groot (1998) suggest that the picture representing a cognate word activates the corresponding L1 word form, this effect may also give evidence for the integration of the sematic representation of cognates.

Explaining cognate advantage in experimental learning paradigms The facilitating effects of cognateness on vocabulary learning can be explained by the theories of how they are stored and retrieved from the mental lexicon. Lotto and De Groot (1998) and De Groot and Keijzer (2000) suggest three possible reasons for the enhanced vocabulary learning performance for cognates. The first and simplest explanation given for the cognate advantage is that in the case of learning a target language cognate, which is the orthographic neighbour of the L1 word, there is less to be learned. Secondly, the advantage of learning cognates may stem from the formal overlap in the case of cognates and the lack of such an overlap in the case of noncognates. Basically, when a cognate is presented as the testing stimulus, it will constitute a strong cue for the retrieval of its translation equivalent in the target language. Finally, assuming that bilingual/multilingual memory, similarly to monolingual memory, is organised by morphology, a cognate relation between two words can be considered special case of a morphological relation that may exist between words within the same language. Such a relation is reflected in the joint storage of morphologically related words, or

82

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

lexemes in memory (e.g. go, goes, going and for cognates e.g. house, haus). According to this view, the learning of a new non-cognate word relation involves creating a new representation, while learning a cognate word may only involve adapting a representation already stored there before the learning took place. Thus, the learning process is less demanding for the learner, which results in the effects of enhanced learning. This explanation ties up with the notion of lexemic transfer (Jarvis, 2009), discussed in Chapter 5. De Groot and Van Hell (2005) point out that the three explanations are not mutually exclusive and may all contribute to the enhanced learning of cognates relative to non-cognates.

7

Cognate Vocabulary in Second Language Acquisition

Introduction Considering experimental evidence on bilingual/multilingual word processing and foreign vocabulary learning, crosslinguistic similarities between the known and the target language should be decisive in defining how easily the language will be learnt. Since learning consists in relating new knowledge to what we already know, or can easily associate with something known, cognate vocabulary seems ideal to learn in a fast and effortless way (Ringbom, 2007). For the learner, the mere presence of cognate forms should be the most salient feature of the newly learnt language. Moreover, there would be no difference, whether the crosslinguistically similar forms are true cognates deriving from linguistic genealogy or non-genetic cognates deriving from language contact (Jarvis, 2009; see Chapter 4). This chapter focuses on the role of cognates in second language learning. It points to the differences between the assumed role of cognates in SLA and how, surprisingly, some language learners tend not to be aware of cognates in the target language. Finally, conclusions are drawn from the studies presented, which point to certain methodological flaws of SLA research on cognates.

The Assumed Role of Cognates in Language Acquisition In SLA the presence of crosslinguistic lexical similarity is considered to be a real asset. It is claimed that, apart from the obvious disadvantages of false friends, ‘learners will find one language far easier to learn than another if the one language shows many lexical similarities with their native language and the other does not’ (Odlin, 1989: 79). The very idea that similarities and differences in word forms and word meanings play a role in how 83

84

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

quickly a particular foreign language may be learned by speakers of another language has been recognised for at least a century. Already in 1899 Sweet observed that [m]astering the vocabulary of most European languages means simply learning to recognize a number of old friends under slight disguises, and making a certain effort to learn a residue of irrecognizable words, which, however, offer less difficulty than they otherwise would through being imbedded in a context of familiar words. The higher vocabulary of science, art, and abstract thought hardly requires to be learnt at all; for it so consists either of Latin and Greek terms common to most European languages or of translations of them. (Sweet, 1899, quoted in Odlin, 1989: 77–78) In 1887, Zamenhof constructed Esperanto, an international auxiliary language whose vocabulary is largely based on cognates. Before World War II, Jespersen (1928) developed the idea of a language based on words common to all people. Within the framework of contrastive analysis, Lado (1957) introduced a scale of difficulty in learning foreign language vocabulary, and distinguished seven categories. The words easiest to learn would be those similar in their form and meaning to the native equivalents; next came those similar in form, but with slight meaning differences. As Lado (1957: 2) himself claimed ‘those elements that are similar to [the learner’s] native language will be simple for him, and those elements that are different will be difficult.’ The interest in studying the potential of cognate vocabulary and internationalisms (see Chapter 4) in language teaching rose in the 1980s and 1990s (Kolwa, 2001). Studying a core of internationalisms, or common loan words, in many European languages (Braun, 1986, 1989; Braun et al., 1990) resulted in creating a dictionary of nearly three thousand internationalisms in ten European languages: German, English, Dutch, French, Italian, Spanish, Portuguese, Russian and Polish. In the last two decades, the interest in internationalisms was fostered by the united-Europe policy. It gave rise to attempts to create programmes aimed at intercomprehension, i.e. training receptive language competence on the bases of similarity relations between typologically close languages (e.g. The EuroCom project, Hufeisen & Marx, 2007). As demonstrated by research on receptive bilingualism/multilingualism in closely related Germanic languages, crosslinguistic similarities provided by cognate pairs aid understanding an unknown language (e.g. Gooskens, 2007; Gooskens & Van Bezooijen, 2006; Gooskens et al., 2011; Vanhove, 2014). Having discussed the historical processes of borrowing (Chapter 3), we might assume that in the case of other European languages from the major families, learning vocabulary should be extremely easy. A considerable bulk of vocabulary present in most modern European languages derives from

Cognate Vocabul ar y in Second L anguage Acquisit ion

85

Latin and Greek. Such words are also common in formal styles and registers used by educated European speakers in their L1s and are characteristic of academic discourse (Nagy & Townsend, 2012). Meara (1993) points to the existence of c.a. 3000 English-Spanish cognates. He estimates the number of English-French homographic cognates (identical in their form) as 6500, and non-homographic as 17,000. Nation and Meara (2002: 49) claim that ‘almost all the basic Anglo-Saxon words have parallel forms based on Latin and Greek, which are used in particular, in specialist discourse.’ There exist cognate words across typological boundaries, common to Germanic, Romance and Slavonic languages. As claimed by language educators, adults normally have about 5000 of these easily recognisable words in their vocabulary. [. . .] [T]hese words provide that part of a newspaper article on, say, international politics that can be immediately understood: this vocabulary usually forms the larger part of such articles. (www.eurocomcerner.de) Such data, as well as the results of experimental studies discussed in the previous chapter, suggest that for language learners cognate words should constitute considerable help in bridging the gap between the vocabularies of the known and the target language. However, contrary to those claims and the experimental findings, evidence from the language classroom is far from univocal. As it turns out, some learners in classroom conditions do not take advantage even of obvious cognates: they do not automatically accept formally similar words as equivalent to the L1 orthographic neighbours, especially if the L2 language they are learning is distant typologically from L1.

Mixed Evidence from SLA Studies Positive evidence from the language classroom In comparison with research on cognate processing, there are surprisingly few pedagogically oriented studies on the use of cognates, and the evidence they produce is mixed. Some studies report that learners indeed recognise cognates with ease and benefit from such words in vocabulary acquisition. More specifically, there are SLA studies showing that the L1 background of the learner is important when mastering L2 vocabulary. Learners from language backgrounds that bear more formal lexical similarity to the target language are more successful in learning target words. For example, a well-known study by Ard and Homburg (1983) compared the acquisition of vocabulary by learners of English with L1 Spanish and L1 Arabic. Students with the Spanish L1 background considerably outperformed

86

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

learners with L1 Arabic background. As the authors suggested, the learners benefitted not only from recognising cognates, but also from having more time to concentrate on unfamiliar vocabulary. When tested, the students with the Spanish L1 background did best whenever the test items had spelling similar to that of the Spanish cognate forms (Ard & Homburg, 1983). In another study Holmes and Ramos (1993) found that even with beginniners, cognates could aid reading comprehension. The project ‘Brazilian English for Specific Purposes’ studied recognising cognates in reading comprehension tasks performed by undergraduate Brazilian students, mostly beginners in learning English. Data was gathered through think-aloud protocols. Students read silently through an English text, orally summarised their understanding in Portuguese and they simultaneously commented on the task. Later the wording of the subjects’ summaries was matched with the original text to identify cognates and other vocabulary items which were recognised or misrecognised in the original text and were transferred to the summary. Holmes and Ramos (1993) had accurately predicted that learners would approach cognates with suspicion due to folk-linguistic ideas about foreign language learning. However, they concluded from their study that cognate recognition was a ‘natural strategy’ in learning English, and that their learners spontaneously sought out cognate vocabulary in texts. Another piece of positive classroom evidence comes from English children learning Spanish in bilingual immersion classes. Research shows that L2 cognates help expand children’s L1 vocabulary. Cunningham and Graham (2000) studied the effects of Spanish L2 immersion on children’s English L1 vocabulary. Thirty 5th- and 6th-grade immersion students and 30 monolingual children from a control group did the same standardised productive and receptive vocabulary tests. Additionally, they did a Spanish–English cognate test, which aimed at recognising low-frequency English words that had high-frequency equivalents in Spanish. The immersion group did similarly to monolinguals on the productive vocabulary tests, but significantly better than monolinguals on the standardised receptive vocabulary test because of the presence of cognates. Immersion students also outperformed the control group on the test of cognates. Cunningham and Graham (2000) conclude that Spanish immersion children revealed benefits in their L1 English receptive vocabulary knowledge due to positive transfer from Spanish as a foreign language.

Difficulties in recognising cognates Not all researchers, however, are so enthusiastic because some language learners do not seem to benefit from the presence of cognates. According to Swan (1997), treating cognates as a sure source of positive transfer is an obvious oversimplification. He points out that the very existence of cognates does not always lead to an enhanced mastery of L2 vocabulary, similarly, not all dissimilarities between languages cause learning problems. Moreover,

Cognate Vocabul ar y in Second L anguage Acquisit ion

87

Odlin (1989: 79) claims that ‘more and more research on contrastive lexical semantics shows that recognition of cognates is often a problem. Learners may not always note the formal similarities that mark a cognate relation, and they may not always believe that there is a real cognate relationship.’ According to Schmitt (1997: 209), although cognates are ‘an excellent resource for both guessing the meaning and remembering new words (. . .) learners do not automatically accept cognates as equivalent.’ Overall, for at least three decades studies have repeatedly shown that cognates remain unnoticed, or tend to be avoided, by learners (Banta, 1981; Dressler et al., 2011; Kellerman, 1983; Lightbown & Libben, 1984; Nagy et al., 1993; Odlin, 1989; Otwinowska-Kasztelanic, 2009, 2011a; Schmitt, 1997; Singleton, 2006; Swan, 1997). A lot of the evidence which points to the fact that cognates do not constitute the expected asset in learning vocabulary comes from studies involving English. Potentially, learning English vocabulary should not be a problem for learners from Germanic, Romance and Slavic backgrounds. As shown in Chapter 3, although historically English is a Germanic language and typologically it is close to German and Dutch, it underwent considerable transformation under the influence of Norman French. Consequently, it also shares hundreds of cognates with Romance languages, as well as numerous cognates with Slavic languages, which borrowed extensively from Latin, Greek, French, German and English itself (cf. Nation & Meara, 2002; Schepens et al., 2012). At the same time, though, research evidence shows that learners often find it difficult to benefit from cognates shared by English and other European languages. For instance, in a study on the acquisition of German vocabulary by English students, Banta (1981) found some learners’ inability to recognise cognates quite ‘astonishing’ and opted for teaching cognates as a strategy for expanding learners’ vocabulary. Schmitt (1997) examined Japanese learners of English (n = 600) with a survey of language learning strategies. He noted that although Japanese has borrowed numerous words from English, checking for an L1 cognate was considered to be the least helpful strategy by a vast majority of learners. In a classic study, Kellerman (1983) pointed to the importance of learners’ perceptions in accepting cognateness. In the study, Dutch L1 learners of English as L2 were confronted with idiomatic expressions containing Dutch-English cognate verbs. They were to judge whether a list of 17 expressions containing the Dutch verb ‘breken’ could be translated into English with the verb ‘to break.’ It turned out, that those expressions whose meaning was closer to the primary meaning of the verb displayed higher acceptance rate than those whose meaning was more distant from the primary sense of the verb. For instance, the acceptance rate for an expression ‘he broke his leg’ was 100%, while for ‘some workers have broken the strike’ it was only 11%. The study implies that the less representative of the prototypical meaning a usage of a given form is, the harder it is for learners to

88

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

accept cognateness and translation equivalence of expressions. Learner judgements will be thus be influenced by two interacting factors, the learnerrelated psychotypology and the item-related prototypicality. Let us now focus on some classroom-oriented studies in more detail. Lightbown and Libben (1984) have found evidence that cognates do not always help learners in L2 vocabulary production. Their research dealt with teenage French L1 learners studying English as L2. They compared the French learners with native speakers of English of the same age. Both groups had to watch a film featuring some obvious cognates (e.g. sheriff, cowboy, guitar) and had to write a composition narrating the events. Later, a portion of these ESL students also wrote another composition about the same film in their native language, French. The authors of the study expected cognates to reappear in the French and English essays, but the results were disappointing. The French-speaking participants did not use the same words in the same contexts as did English native speakers. Apart from free compositions, both groups performed a cloze test and a word acceptability judgment task to force learners to activate cognate vocabulary. However, it again turned out that students were reluctant to accept words with similar spellings in the two languages to be equivalent. Lightbown and Libben (1984) conclude that L2 learners tend to avoid cognates encountered for the first time, especially if earlier warned about the existence of false friends. They do not trust that words may have the same meaning even though their spelling is similar in L1 and L2. Learners will not accept, or attempt to use cognates, unless they have encountered them in a specific context in the target language. Several studies focused on cognate recognition. For instance, Nagy et al. (1993), studied Spanish bilingual and biliterate teenage students of English (n = 74). They were first given yes/no and multiple choice vocabulary tests in Spanish and English to check their word knowledge, including cognates. Later, they were asked to search for cognates in texts that contained the words tested. However, when reading, participants were able to find and circle only half the cognates they had shown they knew in vocabulary tests. In another study by Soufra (in Singleton, 2006), English adult beginners of Modern Greek were given a number of Greek-English written translation tasks and a multiple-choice collocation recognition task to investigate whether and how they benefit from crosslinguistic similarity. While the study overall produced plenty of evidence for language transfer, some students proved completely unable to consciously recognise even obvious cognates in written texts (e.g. γαλαξιασ – galaxy, αυθεντικοσ – authentic). Singleton (2006) explains that the students may have perceived the two languages to be relatively distant and thus did not pay attention to similarities. Admittedly, in the case of Soufra’s study, the differences in the script may have played a role in the recognition process, but the problem applies also to languages sharing the same alphabet.

Cognate Vocabul ar y in Second L anguage Acquisit ion

89

Need for cognate strategy training The studies discussed above show that learners differ in intuitively acknowledging that crosslinguistic lexical similarities are helpful. As noticed in research on learning strategies, ‘although cognates may be deliberately used in language texts, if they are not explained, many students never see the relationship of these cognates to words in their own language. Once the student’s attention is drawn to the relationship, the same student may learn several hundred words in a very short time’ (Rubin, 1987: 16). It is claimed that learners need cognate strategy training that will help them make intelligent guesses. Such cognate training would also support vocabulary learning: ‘ears and eyes trained to recognise (. . .) cognates and common loans will help brains to build new passive vocabulary more rapidly in the target language’ (Banta, 1981: 136). It is also suggested that teachers who are native speakers of the target language are to be blamed for disregarding the issue of cognate training. If they do not share the learners’ knowledge of their L1, it is almost impossible for them to help learners take advantage of cognate vocabulary (Holmes & Ramos, 1993). The claims that learners need training in using cognates are corroborated by classroom evidence. For instance, Tréville (1996) worked with university level beginners and false beginner English learners of French as the L2 (n = 105) during a course focused on listening and reading comprehension. The experimental group participated in a modified schedule of the course focused on recognising English-French cognates, while the control group did not receive such instruction. At the end of the course, all students were given the Test of Aptitude in Recognising Written Cognates. The exam involved seven scales and consisted of nine tasks concerning cognate recognition and use. Tréville (1996) found that the experimental group obtained better results in the recognition, application of interlexical correspondence rules, identification of grammatical categories, generalisation of interlexical rules and in the selection of appropriate words in given contexts. In conclusion, the study points out that that existence of cognates between languages does not guarantee that L2 learners will use cognates properly: students need training. A recent study that pinpointed the need to train recognising cognates was presented by Dressler et al. (2011). It involved a small group of SpanishEnglish bilingual teenagers (n = 8) half with and half without prior vocabulary learning strategy training. The study was part of a vocabulary intervention programme designed to enhance the vocabulary development of teenage students. The programme included a component of cognateawareness raising. In addition to teaching the meanings of 168 English target words to children, the intervention explicitly modelled strategies for inferring the meanings of unknown words: using context clues, word-part analysis and cognate-awareness. The cognate awareness raising was the strategy

90

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

of focus for three of the 15 weeks of the course and cognates constituted a subset of the selected academic target words. The testing phase involved a think-aloud procedure of data collection, while learners were presented with short passages containing words whose meaning had to be resolved. The texts included target cognates (e.g. amicable, obscurity) which the students did not know in English. All Spanish–English cognate pairs were semantically related, but had varying degrees of orthographic and phonological overlap. Dressler et al. (2011) noted that strategy training gave a distinct advantage in inferring meanings to those students who had been taught to search for cognates. At the same time, it appeared from the data that mere familiarity with the Spanish word in a cognate pair was not a sufficient condition for the cognate strategy to be used. Following the claim that learners need to be trained in the awareness of crosslinguistic similarity, whole educational programmes such as EuroCom were developed, even for typologically related languages which abound in cognates (www.eurocomresearch.net). The project aimed at showing that the relations between languages in the same group can be utilised in language teaching practice and at creating teaching methods and teaching tools for intercomprehension within Romance, Slavic and Germanic language groups. The methods rely on teaching receptive competence via interlingual transfer. As the proponents of the method suggest, a network of transfer correspondences can make written language within each group transparent to users or related languages. The network of correspondences is called the Seven Sieves (www.eurocomcerner.de) and focuses on lexical and morphosyntactic similarities between texts in European languages. The First Sieve extracts international vocabulary from the text. The Second Sieve then extracts the words belonging to the pan-Romance, pan-Germanic or panSlavonic vocabulary, the Third Sieve deals with sound correspondences and the Fourth Sieve deals with spelling. The other Sieves are concerned with syntactic structures and morphosyntactic elements to provides the basic formulae for recognising the different ways different grammatical elements have developed in a particular group of languages. Thanks to EuroCom learners are supposed to become aware a bulk of familiar knowledge they already had subconsciously, and not just for one language, but for several languages within the group (cf. Hufeisen & Marx, 2007; Meißner et al., 2004; Zybatow & Zybatow, 2002).

Conclusions: Why do we get Conflicting Results in Various Types of Studies? Although experimental psycholinguistic research points to robust cognate facilitation effects when cognates are processed by bilinguals and multilinguals, classroom evidence gives mixed results: some learners benefit

Cognate Vocabul ar y in Second L anguage Acquisit ion

91

from cognates, and some seem to be ‘immune’ to their existence. There may be complex reasons for the discrepancy between the facilitating effects of cognates in experimental settings and difficulties in noticing them in the classroom. The first group of factors pertain to the specificity of experimental settings as opposed to classroom settings, including aspects of methodology of language teaching. The second group of factors involve methodological flaws of some SLA classroom research. Finally, the third group of factors involve those that add to ‘transferability’ and noticing of particular cognate items on the basis of individual judgements of language learners. Let us first focus on the classroom settings and experimental settings. In the classroom settings, learners’ inability to take advantage of cognates may result from language teaching methodology which disregards crosslinguistic similarity. For instance, as a far echo of the influential CAH hypothesis (Lado, 1957), some researchers emphasise that uninstructed experimentation with cognates by L2 learners might be harmful due to the existence of false friends (Holmes & Ramos, 1993; Sheen, 1977). Therefore, many teachers tend to draw students’ attention to false friends, rather than cognates, believing that the presence of false friends will lead students to learning difficulties (cf. Lightbown & Libben, 1984; Meara, 1993). In consequence, L2 learners, if warned about the existence of false friends, tend to avoid cognates: they do not trust unknown words with similar spelling in L1 and L2, not believing that they may have the same meaning (Lightbown & Libben, 1984). Secondly, apart from programmes aiming at intercomprehension (e.g. Hufeisen & Marx, 2007; Meißner et al., 2004; Zybatow & Zybatow, 2002), teaching materials rarely focus on L1-L2 crosslinguistic similarities. This is due to the fact that course-book series are often constructed in a way that allows for their use worldwide. Consequently, they disregard individual differences between learners, including their various L1 backgrounds (Dörnyei & Skehan, 2003; Meara, 1993). If learners are to be treated as homogeneous and books are to be sold in various countries, there is no possibility for coursebook authors to focus on crosslinguistic similarities. Therefore, cognates remain unnoticed in textbooks, apart from the most obvious recent borrowings from English. Yet another reason for the conflicting evidence may be the type of methodology used in the SLA research and language processing research. Psycholinguistic experiments are meticulously planned and implemented, but for the same reason, they may have low ecological validity. All experimental stimuli (cognates and non-cognates) are strictly controlled in terms of the orthographic overlap, word frequency and/or word class, which never happens in the class setting. Also the equipment used to run experiments is far from whatever can be found in the language classroom. On the other hand, in psycholinguistic experiments less attention is paid to individual learner factors, their language backgrounds and the constellations of languages known. Also, online processing experiments are usually carried out

92

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

on people who are quite proficient in the languages examined, which has an impact on the results (see Chapter 9 for a discussion). As for classroom research, it may be hard to draw univocal conclusions from the SLA studies precisely because they used varied methodologies and research instruments, and they examined learners from various backgrounds and age groups. Although traditionally, many SLA studies emphasised learner-related factors (i.e. age, language background, proficiency), they usually disregarded such important learner-related factors as multilinguality and constellations of languages known (see Chapter 9 and 10 for discussion on learner-related factors). Moreover, SLA research usually fell short of reliably measuring the objective similarity of cognates used in research. Most SLA studies neglected the characteristics of the word items used and therefore were not able to show if formal crosslinguistic similarity between L1 and L2 words truly helps L2 word acquisition in the case of cognates, or really hinders acquisition in the case of false friends. Overcoming some of these shortcomings is attempted in studies presented in Chapters 12–14. Quite importantly, however, some studies presented conclusions that were not always justified by the results. As an example, let us come back to the study by Holmes and Ramos (1993) that used think-aloud protocols to show positive results of cognate recognition in texts. Although cognate recognition was considered to be learners’ ‘natural strategy,’ there was a lot of variation from one student to another. To explain this variation, the authors stated that ‘cognate identification seemed to be personal, with some subjects inclined to be more liberal than others admitting a word cognate’ (Holmes & Ramos, 1993: 89). They also pointed out that previous knowledge might have been important because students had clearly relied on their background knowledge of the topic to predict the content of the text. It seems that in discussing the results many individual learner factors were disregarded. Furthermore, the design had serious methodological flaws. Apart from the think aloud method, the researchers asked students to prepare a L1 group summary of a given text. The finding was that almost all cognates were recognised by each of the groups. However, the very task may have confounded the results obtained. Without closely monitoring the group work, it is not clear, if cognates were universally recognised by all students within the groups. Considering the variation between students, as reported earlier, possibly cognates were recognised by one student in each group, and then he/ she passed the knowledge on to the other group members. All in all, it is not clear whether success claimed in the study by Holmes and Ramos (1993) was really due to the learning of cognates. It seems that some authors’ conclusions may have been slightly haphazard and resulted from the lack of methodological rigour in the study. On the other hand, in the studies which presented negative evidence for cognate recognition, it is equally unclear whether learners really did not benefit from the presence of cognates. If benefitting from cognates is not

Cognate Vocabul ar y in Second L anguage Acquisit ion

93

measured by specific tests but by tasks depending on awareness (noticing in texts, explaining meaning), it is hard to conclude that learners really did not know the words. It may well have been that cognates were easier to learn than other words, but the learners were not able to consciously explicate their knowledge when asked to enumerate cognates, judge their similarity, or comment on similarities with the use of the think-aloud method. Let us pass on to the last and very important group of factors, which involve ‘transferability’ of particular cognate items. As already discussed in Chapter 5, the nature and type of crosslinguistic influences is strictly connected with individual judgements of language learners. Thus, students’ willingness to accept that elements of one language can be useful in learning another language depends on transferability, which is ‘the probability with which a structure will be transferred relative to other structures in the L1’ (Kellermann, 1983: 117). Since ‘not everything that looks transferable is transferable’ (Kellermann, 1983: 113), there must be constraints that inhibit transfer between languages even in the case of obvious congruency of words and structures. A constraint on transfer can be defined as ‘anything that prevents the learner from either noticing a similarity in the first place or from deciding that the similarity is a real and helpful one’ (Odlin, 2003: 454). Jarvis and Pavlenko (2007: 175) enumerate five categories of factors that may affect crosslinguistic influence (CLI) and transferability, and hence also the learner’s success or failure in noticing crosslinguistic correspondences. They pertain to some language-related and some learner-related factors: (1) (2) (3) (4) (5)

Linguistic and psycholinguistic factors. Cognitive, attentional and developmental factors. Factors related to cumulative language experience and knowledge. Factors related to the learning environment. Factors related to language use.

The first category relates to characteristics of the languages, their genealogy, structure and the characteristics of language items (words, morphemes). The next two categories depend on the individual characteristics of the language learner: the person’s ability to use the cognitive capacities and resources at the time of learning (quality of the working memory, age related factors) and the effects of the learner’s prior experience in language learning (cumulated amount of knowledge). The last two categories are related to the setting of language learning and the external context of language use. The following three chapters will focus on the two groups of factors mentioned above. Chapter 8 will focus on language-related factors which may make cognates easy or difficult to learn and will examine what properties of cognates may hinder their recognition. Chapter 9 and 10 will focus on learner-related factors, i.e. for whom and why it may be easier to benefit from cognate words when learning a language.

8

Language-related Factors in Lexical CLI

Introduction While differences between languages do not always lead to learning difficulties (contrary to the strong version of the CAH discussed in Chapter 5), similarities may not always lead to enhanced learning. Although a range of psycholinguistic experimental studies on bilinguals indicate that cognates are processed in a language non-selective way, in the language classroom, cognates remain unnoticed by some learners and often do not aid learners in understanding texts, or producing utterances. From the discussion presented in Chapter 7 it seems that the key to understanding when the presence of cognate words may trigger positive crosslinguistic influences in language learning is the learner’s awareness of their existence. Learners’ awareness cooperates with constraints on the effects of objective crosslinguistic similarity which shape learners perceptions and assumptions. Therefore, perceiving the existence of cognate vocabulary may be difficult for reasons connected with the properties of words and the characteristics of the languages involved. The present chapter will discuss such different language-related factors responsible for taking advantage of crosslinguistic similarity. First, it will present word-related characteristics that affect both processing and learning of words, including cognates. Next, it will present contextual factors that add to cognate recognition. Finally, it will try to explain the language-related circumstances when recognising cognates and using them in language learning may pose difficulties for the learner, and when the task is relatively easy. 94

L anguage-rel ated Fac tors in Le x ical CL I

95

Word-related Criteria Influencing CLI and Processing Markedness and prototypicality Learners generally resist transferring L1 marked forms. Kellerman’s studies (1978, 1979, 1983, 1986) on Dutch learners of English and German conclude that a particular L1 item will be less transferable if it is perceived by the learner to be somehow marked: infrequent, irregular, or less prototypical. Kellerman (1978, 1987) has shown that learners rely on certain types of objective similarities, but they do not rely on others, depending on how language specific, or prototypical they consider the language features in question to be. ‘The less representative of the prototypical meaning a usage of a given form is, the lower its transferability’ (Kellerman, 1987: 65, in De Angelis, 2007: 23). Thus, transferability will largely depend on perceived and assumed similarities, involving learner judgements of markedness. Constraints on transferability stemming from learner judgements pertain not only to grammatically marked forms but also to lexical items, whose forms, meaning and use are perceived by the learner as non-prototypical, like in the case of the expressions with ‘breken-break’ Dutch-English cognates discussed in Chapter 7 (Kellerman, 1983, 1986). In the case of lexical items, transferability and utilising crosslinguistic similarity may depend on the characteristics of particular word items. Learners may not believe that a real cognate relationship exists between formally similar words if the words are somehow marked, infrequent, or less prototypical, which includes semantic and structural opaqueness. Further the transferability criteria of markedness and prototypicality correspond with some word-type features already discussed in Chapter 6 (e.g. De Groot, 1992; De Groot, 1993; De Groot et al., 1994; De Groot & Comijs, 1995). For instance, words in two languages may display varied orthographic overlap (cognateness). They may refer to concrete objects or to abstract entities (imageability/concreteness) and may be used seldom or often (frequency of occurrence). The following sections will focus on the impact of these criteria on processing, learning and transferability.

Orthographic similarity The orthographic similarity of cognates can affect the facilitation effects in experimental settings. For instance, in a lexical decision task by Font (2001) French-Spanish bilinguals were presented with identical cognates and cognates differing in one letter. The amount of cognate facilitation in L1 and L2 processing depended on the letter change in the cognate word. The facilitation effect for identical cognates was significantly larger than for pairs of cognates differing in one letter (Font, 2001, in Dijkstra, 2005). Likewise, in

96

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

a translation study by Dijkstra et al. (2010), those cognates which were more like their translation equivalents were recognised faster than those which bore less similarity. In online translation experiments on German-English bilinguals guessing cognates form an unknown L3 Swedish, the increasing orthographic similarity significantly improved translation accuracy (Vanhove, 2014). Lesser orthographic similarity of cognate pairs can also hinder their recognition in classroom conditions, and hence negatively affect transferability. For example, the study on Spanish learners of English by Nagy et al. (1993), found that greater orthographic congruence in Spanish-English cognate pairs resulted in their higher recognition rates in texts (see Chapter 6). Also the case study by Dressler et al. (2011) on Spanish-English bilinguals suggests that a degree of orthographic overlap between cognates might be required for students to access cognate knowledge. The less similar cognates remained unnoticed by learners. Concluding, the amount of facilitation and recognition of cognates results from objective crosslinguistic similarity of cognate pairs. Dijkstra (2003) attributes such orthographic effects to the fact that formal features of words may influence the word selection process in bilingual and multilingual visual word recognition. A lesser degree of similarity may ‘deactivate’ the other languages known to the learners, which will be discussed in more detail below.

Word frequency Word frequency plays a role in processing and learning vocabulary, so frequency effects are reported in experimental studies on cognate recognition and translation (e.g. De Groot et al., 1994; Lemhöfer et al., 2008; Dijkstra, 2005; Dijkstra et al., 2010), to the effect that high-frequency words are processed faster and more accurately than low-frequency words. It is also proved that facilitation effects for non-identical cognates disappear when such cognates are of low frequency in both languages (Font, 2001, in Dijkstra, 2005). Frequency matters to a lesser extent in experimental studies on vocabulary learning. Marginally better performance for high-frequency words than for low-frequency words has been noted in several studies, but the effects are not robust (De Groot & Van Hell, 2005). In SLA, word frequency (together with word length and cognateness) are enumerated as elements of the so-called ‘word difficulty’ which may affect learnability (Laufer, 1990, 1997). For example, Milton and Daller (2007) examined the impact of such word-difficulty factors within a single model. They tested the impact of word frequency, word length (number of syllables) and cognateness (degree of orthographic similarity) on learning French vocabulary by English learners (n = 106). The test set consisted of 100 words from five frequency bands of a lemmatised French vocabulary list.

L anguage-rel ated Fac tors in Le x ical CL I

97

On average, the words of the most frequent band were shorter and less cognate than in the other bands, whereas the words of the least frequent band were the longest and the most orthographically similar. The study showed that the frequency band of the word was a significant predictor of learning difficulty (Milton & Daller, 2007, in Milton, 2009). Across SLA studies on vocabulary learning, high-frequency words are reported to be acquired faster than low-frequency words (cf. an extensive summary by Singleton, 1999). If a word is of higher frequency, it is also more likely to have been encountered and learned faster, as presented in a model by Meara (1992) and proved empirically for learners with various combinations of L1 and L2 languages (Meara & Milton, 2003; Milton, 2009). On the other hand, studies have also demonstrated that under certain circumstances it does not matter how often a particular word is presented to learners. The effect of frequency of presenting particular word items may be negligible when the learner’s proficiency in L2 is too low, when the L2 form is not salient to the learner, or when it requires explicit learning (cf. VanPatten et al., 2004; Vidal, 2011). The frequency of the forms in L1 and/or L2 may also affect crosslinguistic influences: learners tend to transfer their frequency preferences from L1 to L2 (Andersen, 1983; Jarvis & Pavlenko, 2007).

Word class Semantically, the different word classes encode different types of meaning. For instance, nouns typically denote entities and have meanings that are less dependent on the specific linguistic context than verbs, or adjectives (Wierzbicka, 1988). Consequently, nouns will carry more ‘core’ and less ‘marked’ meanings than verbs, adjectives and adverbs. This entails that classes of words may also be acquired at different rates both in first and SLA and that transferability of particular word items may also depend on their word class. Evidence from L1 contexts shows that children acquire nouns earlier and faster than verbs and adjectives (see Sandhofer & Smith, 2007 for an extensive overview) because they refer to whole entities, and not only to their selected properties (Sandhofer & Smith, 2007; Smith et al., 1997). The noun advantage over other word classes is also present in L2 acquisition. Classroom-based research shows that nouns are learned faster than verbs, adjectives and adverbs, as presented by longitudinal case studies (e.g. Horst & Meara, 1999; Milton, 2009), and studies on incidental vocabulary learning from reading (Kweon & Kim, 2008; Vidal, 2011). This pertains to all learner proficiency levels. For instance, Schmitt and Zimmerman (2002) examined advanced learners of English (n = 106) by testing their productive knowledge of derivatives from the same word family (e.g. V = select, N = selection, Adj. = selective, Adv. = selectively). Their subjects had to fill in a gap in a sentence with the right word. The results of their tests indicated that verb and noun target forms were supplied more often, whereas adjective and

98

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

adverb forms were generally less well known. To explain the noun advantage in L2 acquisition, it is proposed that nouns are easiest to acquire because they are perceptually better grounded than other word classes and may enjoy more crosslinguistic overlap of meanings than other word categories, which also adds to their transferability

Imageability Vocabulary items can be categorised into concrete words of high imageability and abstract words of low imageability. This criterion, connected with how a word is perceptually grounded, affects the ease with which a mental image of a particular word can be recalled (De Groot et al., 1994). Imageability influences the bilingual processing of both cognates and non-cognates of words (Cies´licka & Ekert, 2009; De Groot et al., 1994; De Groot & Comjis, 1995; Van Hell, 1998; Van Hell & De Groot, 1998). For example, Van Hell and De Groot (1998) examined the impact of word class and concreteness on processing with the use of a word association task. They found that concrete words elicited more associations than abstract words, both within and across languages. Interestingly, in their study, word associations for cognates were also more similar across languages than for non-cognates. Concrete words are also translated more accurately than abstract words. Tokowicz and colleagues (2002) studied Dutch-English bilinguals who performed a semantic rating task on translation pairs. They found that concrete translation equivalents are more likely to share meaning than abstract translation equivalents. Relatively few vocabulary learning studies have manipulated concreteness, but all note considerable effects of this variable on learning words (De Groot & Keijzer, 2000; Ellis & Beaton, 1993b; Kroll et al., 1998; Lotto & De Groot, 1998). For instance, the recall scores for concrete words are significantly higher than those for abstract words (De Groot & Keijzer, 2000). Concrete words turn out to be easier to learn and retrieve than abstract words, which results from greater richness, information content, associative embedding and opportunity for their anchoring and retrieval relative to abstract words (De Groot & Van Hell, 2005; Lemhöfer et al., 2008). Therefore, we can assume that concrete words have more chances of being transferable than because they have the more prototypical meaning than abstract words.

Contextual Factors Influencing Processing and the Recognition of Cognates Relative language activation depending on context Not only the features of vocabulary items, but also the context they are embedded in may modulate their processing. Dijkstra (2003) draws attention

L anguage-rel ated Fac tors in Le x ical CL I

99

to the fact that the features of the linguistic and non-linguistic context may inhibit the activation of languages known to the learner. This activation will depend on the person spoken to, or type of text interacted with, the situation, the content of discourse and the function of the interaction. The concept of relative language activation stems from Green’s (1986) model of bilingual language processing and Grosjean’s language mode (1985, 1997, 1998, 2001) models extended to multilinguals (Dewaele, 2001). According to these models, the user’s languages are activated to a varying degree in a given situation, but never completely switched off. The language mode affects the perception of words and sentences, as well as the speed of access to one, two or more lexicons. Usually, the so called ‘base language’ of a bilingual/multilingual language user/learner is active, while the state of the other language(s) may vary from deactivated in the monolingual language mode, to relatively active in the bilingual/multilingual language mode. ‘At any given point in time and based on numerous psychosocial and linguistic factors, the bilingual has to decide, usually quite unconsciously, which language to use and how much of the other language is needed – from not at all to a lot’ (Grosjean, 2001: 2). In the case of multilinguals, the most highly activated language takes over, while the person’s other languages remain less activated along the language mode continuum (Dewaele, 2001). There are also numerous factors that add to the activation, or deactivation of language modes. According to Dijkstra (2003), relative language activation can be modulated by both the top-down effects of context (morphological, syntactic, and language membership information) and the bottom-up linguistic effects of context (phonetic/phonological/orthographic features, etc.). The issue of precisely how and to what extent language activation and the activation of cognates interacts with top-down factors and bottom-up factors still remains an empirical question. However, Dijkstra (2003) points to a number of factors that appear to aid activation in visual processing by bilinguals and multilinguals. These include neighbourhood density and language-specific cues, which are visible even if two languages are closely related and use the same script. For instance, words may contain diacritical markers (e.g. accents in French), language-specific letters (e.g. a˛and e˛in Polish) or letter combinations characteristic of a given language and not for the other (e.g. th in English, or sz in Polish). The context of such clues may ‘deactivate’ the other languages known to the learners. The situation is even more complex if languages do not share the same script.

Sentence context in modulating cognate effects The recognition and processing of non-cognates and cognates also depends on the semantic context in which a particular word is or is not embedded. For instance, studies with monolingual speakers show that the semantic context of a given sentence can modulate the effects of the word

100

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

type features of concreteness and frequency: ‘Generally speaking, words (e.g. “apple”) that are preceded by a semantically constraining context (like “She took a bite of the fresh green . . .”) are recognised faster than words preceded by a neutral sentence context (like “The final word in this sentence is . . .”)’ (Van Hell, 2005: 2298). In reality, not every sentence context is highly predictive with regard to the target word and sentences occurring in natural language situations are often mildly predictive. Although there are few studies on bilingual speakers investigating the processing of words in context, it has been shown experimentally that meaningful contextual information may also modulate the degree of co-activation of cognates in bilingual memory (Schwartz & Kroll, 2006; Van Hell, 2005; Van Hell & De Groot, 2008). For example, Van Hell (2005) studied the role of the semantic constraint of the sentence context and whether sentence context modulates cognate effects. To this end, Dutch-English proficient bilinguals performed a lexical decision task in their L2, or translated words from L1 to L2, or from L2 to L1. The target words were presented in isolation, or were preceded by a predictive, highly constraining sentence context or by a neutral sentence context. It turned out that the cognate effect was eliminated in a lexical decision task in the second language, and was strongly reduced in translation after the participants read a sentence in a highly constraining context condition. The cognate advantage disappeared also in a study by Schwartz and Kroll (2006) where Spanish-English bilinguals read cognate words in the meaningful sentence context that was highly constraining. The sentences in L2 English contained cognates like the word ‘radio.’ When read in isolation, cognates were processed faster than other words, which suggests that even when reading in one language only, the other language was active. When cognates were embedded in the sentence that was only generally constrained, like: ‘My friend wanted to know if the radio we bought came with a warranty,’ the advantage for cognates remained. In constraining sentences like ‘In the car my friend and I listened to songs on the radio and sang along’ the cognate advantage disappeared. It seems that although context may modulate the activation of the bilingual’s languages, it does not switch off the influence of the other language completely. However, the results of these two studies suggest that lexical access can become more language-selective when words are embedded in contextual information (Van Hell, 2005). Consequently, Singleton (2003) draws attention to the fact that the evidence for cognate facilitation effects largely derives from experimental situations where the stimulus words are decontextualised and presented in isolation and that these conditions may be particularly favourable to the activation of cognates across languages. This decreases the ecological validity of such experimental findings because language learners normally function in contextually rich situations, where they are supposed to deal with words embedded in sentences and in spoken and written discourse. Kroll (2008)1

L anguage-rel ated Fac tors in Le x ical CL I

101

suggests that ‘we normally speak and read sentences in the context of spoken discourse and written texts. The rich cues to the intended language should be available in context and might serve to accomplish the task of effectively switching off the other language.’ Singleton (2003) also argues that it has not been sufficiently explored whether a suitably constraining context in normal linguistic interaction could effectively prevent the activation of one language, or another. Definitely, the findings above translate into the classroom conditions in such a way that in contextually rich situations, target language cognates may not be perceptually salient to the learner. When reading, language learners may not notice the formal similarities that mark a cognate relation, if cognates do not occur in highly constraining contexts, or are not presented in isolation.

Typological and Psychotypological Factors in Lexical CLI Language typology in cognate recognition Dijkstra (2003) reminds us of neighbourhood-density effects, i.e. the fact that the number of orthographic neighbours of a given word affects visual recognition. The recognition process may be facilitated or hindered by the differences in the characteristics of lexical items across languages and their lexical distance (see Chapter 4). The smaller the lexical distance between crosslinguistic orthographic neighbours (including cognates and false friends) the more they will be activated in word recognition. The less similar the cognates, the less activation and ‘recognisability’ there will be. Smaller lexical distance between crosslinguistic orthographic neighbours depends on the actual language distance and language typology. Language typology is responsible for the distribution cognates in European languages and the degree of lexical similarity between particular pairs of languages (see Chapter 4). As demonstrated for European languages, for a pair of languages that are typologically close, the orthographic forms of cognates are identical or highly similar. What follows is that the number of orthographically similar and identical cognates are highly correlated with a particular language family (Schepens et al., 2012). If two languages are more distant typologically, the number of close, similar crosslinguistic neighbours will decrease. Consequently, cognate words which are less similar because they come from more typologically distant languages will be activated less than words from closely related languages. If typologically close languages share more similar cognates, we may assume that for a user of a language from that typological group it is easy to notice and learn crosslinguistic lexical similarity. Faced with predominantly similar words, the learner easily guesses their semantic equivalence.

102

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

Conversely, for languages which are typologically further apart, the number of similar cognates decreases, relative to the number of cognates differing considerably in form. Less similar cognates are less activated in visual recognition, hence they are more difficult to notice and learn. Those less similar cognates may remain unnoticed, which is often the case in languages distant typologically. As claimed by Odlin (1989, 2002), if the learner judges two languages to be considerably different, he/she may not even seek crosslinguistic lexical similarities between the languages, i.e. consciously believe in cognate relations.

Language typology in lexical CLI The typological distance between the learner’s L1 and the languages acquired is a very important predictor of both visual recognition of cognates (Dijkstra, 2003) and transferability (Andersen, 1983; Odlin, 1989, 2003). In the case of vocabulary in European languages, positive CLI (noticing and using cognates), as well as negative CLI (overusing false friends) will be strongest within each of the major typological groups: Germanic, Slavonic or Romance. For instance, CLI will be stronger between Czech and Russian than between Czech and English (Duškova, 1984), or Swedish and English, rather than Finnish and English (Odlin, 2003; Ringbom, 1987, 2007; Sjöholm, 1976). Typologically close languages can be even mutually comprehensible for learners and users (Hufeisen & Marx, 2007). Across typological boundaries overusing false friends may be considered a feature of the learner’s idiolect (Otwinowska-Kasztelanic, 2010). The situation is more complex when more than two languages are involved. One of the first pieces of research investigating crosslinguistic influences in learning L3 was a study by Sjöholm (1976) carried out on the SwedishFinnish border, where people are predominantly bilingual and learn English as L3 at school. Swedish and Finnish are typologically distant, but also the populations of Finnish-dominant and Swedish-dominant speakers are otherwise quite comparable in terms of education and culture. Sjöholm (1976) compared the language production of the Finnish-dominant and Swedishdominant bilingual learners of English. He found that the Swedish-dominant learners made more errors traceable to the L1 than the Finnish-dominant group. Kellerman (1983) attributes this difference to the learner’s sensitivity to the relatedness between L1 and L2. He states that the Finnish-dominant group realises ‘that their mother tongue is not a useful basis for making predictions about the forms of English utterances’ (Kellerman, 1983: 114). A lot of evidence on the role of typological proximity comes from further studies on trilingual and multilingual subjects. The learner’s L2, if close typologically, exerts strong influence on understanding the following foreign languages. For instance, Dewaele (1998) studied speakers of Dutch, English and French. The study revealed that learners tended to transfer more from

L anguage-rel ated Fac tors in Le x ical CL I

103

one of the languages that are typologically close to the language learned, not necessarily from their L1. This finding is confirmed by many later studies (e.g. Berthele, 2011; Dewaele, 2010; Hall et al., 2009). In a study on understanding cognates in an unknown language by Berthele (2011), the typological proximity of the learners’ languages was found to be important for inferring the meaning of cognates and noncognates. Berthele asked multilinguals of varied proficiency in L2 and L3 to infer the meanings of words in an unknown language typologically close to the Germanic, or Romance languages they used. The words whose meaning was to be resolved were presented both in sentence context and out of context. Participants performed tasks based on inference and commented in the form of think-aloud protocols. Berthele (2011) found out that those multilinguals who had high proficiency in two languages typologically close to the target performed better than all other participant groups. In a study on cognate verbs, Hall et al. (2009) found that trilingual learners assumed that verbs shared syntactic frames with cognate forms in the languages that are typologically closer, i.e. Spanish and French, rather than Spanish and English, or Spanish and German. In their study, learners of L3 German and of L3 French studied unfamiliar verbs that were cognate with their L1 Spanish equivalents, L2 English equivalents, or neither. In immediate tests, verbs in the typologically closer French yielded a stronger effect for Spanish grammatical frames than German verbs for English frames. In delayed tests, the effect had disappeared for German but was maintained for French. Moreover, also non-cognates were judged more grammatical in the L2 frame in both experiments. Hall et al. (2009) conclude that similarity of form, combined with typological proximity and L2 status can jointly affect learners’ assumptions about the grammatical properties of the newly learned words. This finding leads to the conclusion that not only the actual typological distance, but also the way the learner judges, or perceives the languages can influence transferability.

Psychotypology It is suggested that the actual typological similarity involved in CLI cooperates with the individual learner judgements because typological distance may be viewed differently by language learners (Kellerman, 1977, 1978, 1979, 1983, 1986; Vildomec, 1963). As mentioned in Chapter 5, the learner’s perception of typological distance between languages is referred to as psychotypological distance, which is the individual assessment of the degree of relationship between particular structures and words of L1 and L2. The smaller the psychotypological distance, i.e. the perceived/assumed/subjective language distance, the easier the learner recognises the forms as congruent between L1 and L2, which can facilitate L2 acquisition (Jarvis & Pavlenko, 2007; Ringbom 2006, 2007). The main point of Kellerman’s perspective is

104

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

that transferability does not depend on the congruence between L1 and L2 but refers to a judgment that the learner makes about his L1 before knowing anything about the corresponding L2 structure. What follows is that the learner has a notion of the degree of typological relatedness between languages, which determines transferability and influences the extent to which the learner will transfer elements of lexicon, phonology or grammar from one language to another. It is proposed that in situations where the languages involved are perceived as close, CLI increase (e.g. Kellerman, 1977, 1979, 1983; Ringbom, 1987, 2007; Singleton & Little, 1984). On the other hand, when languages are typologically distant, learner beliefs about the distance between L1 and the target language in a way ‘block’ noticing crosslinguistic similarity. A perceived dissimilarity may also lead to avoidance of a particular L2 structure (Kellerman, 1986, 1995; Odlin, 1989, 2002; Ringbom, 2006, 2007). Moreover, the subjective psychotypological distance may be congruent or incongruent with the actual typological distance. This means that learners may view the languages in question as distant but still be able to spot crosslinguistic similarities, or view the languages as relatively close, but disregard similarities, which ‘blocks’ transferability (De Angelis, 2007). Research indicates that multilingual language learners can use L1 or L2 as the source of influence in L3 production if the languages are perceived as close (Hammarberg, 2001; Williams & Hammarberg, 1998). On the other hand, in trilingual situations where L1 is typologically distant from L2 and L3, but the latter two are related, the learner can make use of the L2 foundation in early stages of L3 acquisition. For instance, Cenoz (2001) investigated the performance of Basque-dominant and Spanish-dominant bilingual learners of English as L3. Children and teenagers were asked to tell a narrative in L3. Their oral productions were analysed in terms of the lexical borrowings used. The results revealed that Spanish was the most common source language for all learners, and Cenoz (2001) concludes that this lends further support to the notion of psychotypology. The Basque-dominant group used Spanish (L2) as the source language to a higher extent than the Spanishdominant group, which leads to the conclusion that L3 learners make use of their L2 in the learning process when the language is related. In the study, they transferred more elements from Spanish because they felt it was more related to English than Basque, their L1. Another study where psychotypological factors turned out to be very important is one by Gibson and Hufeisen (2003). Here multilingual learners of English and German were asked to translate a text from Swedish, a language unknown to them into a known foreign language (German or English as L3, L4 or L5). Although the learners had several languages at their disposal, they tended use lexical knowledge from German and English. The participants of the study were also asked to explicate which of their languages helped in completing the task. They pointed to English and German

L anguage-rel ated Fac tors in Le x ical CL I

105

as the sources they could draw from and explicitly excluded their L1 languages as CLI sources because they belonged to other typological groups. Gibson and Hufeisen (2003) conclude that the learners psychotypologically perceived German and English as languages the most similar to Swedish and thus disregarded their other languages. In conclusion, there is plenty of evidence in favour of interaction between the different languages used by an individual and the role of learner judgments affecting lexical crosslinguistic processes. The perceived distance between languages may determine which language features bilingual and multilingual language users/learners will attempt to transfer from one language to another. It is worth noting a controversy pointed out by Singleton (2003: 169), ‘the very fact of this “psychotypological” dimension runs counter to the notion of straightforward total integration within the mental lexicon, because, precisely, it implies a degree of selectivity in relation to consultation of the languages represented.’ In the following chapter we will find out more about the learner-related factors that have an impact on language activation and CLI processes involving the use of cognates.

Note (1) Psychological Science Agenda, http://www.apa.org/science/about/psa/2008/01/ kroll.aspx

9

Learner-related Factors Affecting Lexical CLI

Introduction Psychotypological judgments are strictly connected with the languagerelated factors in CLI as discussed in the previous chapter. However, they are not fixed, but may be revised when the learner obtains more information about the language learned (Kellerman, 1979) and develop naturally as part of the learner’s language awareness, including metalinguistic awareness. Both types of awareness are conditioned by the person’s cumulative language learning experience, i.e. the growing proficiency in L2, and proficiency in other languages known. Thus, the chapter opens with the main learner-related factors that come to play in shaping CLI, or that can prevent learners from noticing crosslinguistic similarity. They involve phenomena related to language proficiency, automaticity and fluency of processing. Next, we will discuss the issue of the ‘L2 status’ in learning L3, and the factors of frequency and recency of language use. Finally, we will focus on the cognitive differences between monolinguals, bilinguals and multilinguals pointing to the role of cumulative language learning experience in shaping language awareness of the learner.

The Role of L2 Proficiency Proficiency and cognate facilitation effects in processing Participants’ language proficiency in L2 is consistently reported to affect their word processing, as demonstrated in studies on the bilingual mental lexicon. However, the majority of research into the architecture of the bilingual mental lexicon and cognate facilitation effects has been carried out on highly proficient adult L2 speakers, which is their major drawback. Van Hell and Tanner (2012) review empirical studies that have examined how differences in L2 proficiency modulate cross-language co-activation and 106

Lear ner-rel ated Fac tors Af fec t ing Le x ical CL I

107

interaction during bilingual lexical processing. They stress that ‘only a few studies have focused systematically on the role of L2 proficiency by using appropriate experimental designs that warrant strong conclusions on how differences in L2 proficiency may modulate cross-language interaction, and the extent to which the co-activation of the non-target language is related to proficiency in that language’ (Van Hell & Tanner, 2012: 151). Here we will specifically focus only on the research which demonstrated the role of L2 proficiency in cognate and false friend (homograph) processing in visual word perception and word production. One of the first studies which explicitly examined the role of differences in L1 versus L2 proficiency on lexical access in production was an experiment by Costa et al. (2000). Highly proficient Catalan-Spanish and SpanishCatalan bilinguals were asked to name pictures whose names were cognates or non-cognates in the two languages. All bilinguals named these pictures in Spanish, which was either their dominant language (L1), or their nondominant language (L2). As a result, cognate pictures were named faster than non-cognate pictures in both L2 and L1, but the cognate facilitation effect was larger for production in the non-dominant language (lower proficiency) than the dominant language (higher proficiency). The study demonstrated that the magnitude of the cognate facilitation effect depends on the proficiency in the languages of the bilingual. Another experiment which showed that differences in L2 proficiency affect the magnitude of L1-L2 cognate effects in word recognition is a recent study by Brenders et al. (2011). It has demonstrated that cognate facilitation is weaker in less proficient L2 speakers, and it is more dependent on such contextual factors as stimulus list composition. The authors presented cognates and non-cognates in L1 and L2 to four groups of native Dutch speaking children. Two groups were beginning classroom learners of L2 English, and two were more advanced learners of English. All four groups of participants recognised cognates faster than non-cognates in L2, but not in L1. This points to the fact that cognate effects in L1 emerge only when proficiency in L2 is sufficiently high. It has also been demonstrated that in the case of multilinguals, when more than two languages are activated, some minimal level of proficiency is necessary for words from each language known to the learner to play a role in processing. For instance, Van Hell and Dijkstra (2002) run a lexical decision study on Dutch-speaking trilinguals in a purely L1 context. They observed that processing in the strongest language (L1 Dutch) was always influenced by the strongest L2 (English) and always resulted in cognate facilitation effects: cognates were processed faster, which resulted from the activation of two cognate forms activating the common mental representation. However, the knowledge of L3 (French) affected processing only if participants’ proficiency in L3 was high enough. For participants with low proficiency in French cognate facilitation effects did not take place. Van Hell and

108

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

Dijkstra (2002) concluded that the access and activation of words in the bilingual or trilingual memory system is nonselective with respect to language, but that a certain level of language proficiency is required before any weaker language effects become noticeable in processing. Another study that showed the effects of language proficiency on cognate facilitation in multilinguals was a series of experiments by Poarch and Van Hell (2012), who examined the co-activation of languages in German L2 learners of English, German-English bilinguals and German-EnglishLanguage_X multilinguals. In the experiments, they compared the naming of cognate and non-cognate pictures by children in their dominant and nondominant languages, then compared the results to those obtained by adult German-English bilinguals. The L2 learners, bilinguals, and trilinguals all exhibited a cognate facilitation effect in their L2 and/or L3. This indicated that lexical processing in the non-dominant language is affected by the knowledge of the dominant language. However, the more balanced bilinguals and trilinguals also showed a cognate facilitation effect in their dominant language, although the magnitude of the effect was smaller than in their nondominant language. In contrast, in the case of the L2 learners their dominant language performance did not show any cognate effects. This indicates that only bilinguals and trilinguals with a relatively high proficiency in the L2 and L3 show activation in all their languages. Poarch and Van Hell (2011) concluded that although the activation of languages is evident in child L2 learners, bilinguals and trilinguals, the level of L2 and L3 needs to be sufficiently developed to allow for any cross-language activation that will result in significant cognate facilitation effects when naming pictures in the dominant language. This clearly points to the role of language proficiency: lower levels of proficiency in L2 and L3 allow only for limited cross-language activation in language production. Summing up, Van Hell and Tanner (2012) point out that both adult and child bilingual and multilingual speakers can demonstrate cognate effects in L1 and L2 (L3) in both directions, provided that their proficiency in the L2 is sufficiently high. All studies using lexical decision, picture naming, or word naming demonstrate robust cross-language cognate effects across orthographic, phonological and semantic codes, especially when participants are performing in L2. The presence of cross-language activation when performing in the L1 depends on participants’ proficiency in the non-dominant language. Effects are measurable only for bilinguals with relatively high L2 (L3) proficiency, or when participants are made aware of the relevance of their L2 when performing in the L1. What is more, cognate facilitation effects are usually smaller in L1 than in L2 (L3) processing. This also demonstrates that the co-activation of target and non-target language codes is related to differences in relative language proficiency in L1 and the non-dominant language. Unfortunately, most theoretical models of the bilingual mental lexicon do not take into account the varying learner proficiency (Singleton, 2003;

Lear ner-rel ated Fac tors Af fec t ing Le x ical CL I

109

Van Hell & Tanner, 2012). Only the RHM (see Chapter 6), proposes differences in processing for less proficient L2 speakers and a transition to higher levels of proficiency (Kroll & Stewart, 1994; Kroll et al., 2010). More precisely, the model postulates that in the early stages of L2 learning, the connections between the L2 word form and its meaning are weak. However, with increased proficiency, the mappings of the L2 word form to its concept become stronger (Van Hell & Tanner, 2012). An experiment by Cies´licka (2000), based on the architecture of the RHM, supports the claim that proficiency in L2 changes the strength of the connections. Moreover, the findings support Cies´licka’s ‘variable interconnection’ hypothesis, which states that ‘associative links linking various nodes will vary in strength according to the type of bilingual person’s experience in his or her L2’ (Cies´licka, 2000: 33). This means that not only L2 proficiency, but also the way of learning words may have an impact on the types of links and mapping. The study used a lexical decision task to investigate Polish non-fluent learners of English and proficient Polish-English bilinguals, who differed with respect to the vocabulary learning strategies that they employed: preferred learning words in semantic sets, or in paired associations. First of all, Cies´licka demonstrated that the strength of the semantic connections in the mental lexicon depends on L2 proficiency. For lower-level language learners the connections within the semantic network in L2 were weaker than for proficient bilinguals. Secondly, she showed that the strength of the semantic connections in the mental lexicon of proficient bilinguals depends on the learning strategies employed. In proficient bilinguals, the semantic connections were stronger for those who learned vocabulary in semantic sets, as opposed to paired associations. The experiment clearly demonstrated the possible role of both proficiency and learning experience in shaping the architecture of the learner’s mental lexicon. All in all, lexical access and retrieval in L2/Ln clearly depends on the learner’s/user’s proficiency in the language which modulates the coactivation of languages. In the lexical systems of the less proficient learners/ users, the activation of the different codes will be delayed due to relatively weak connections between formal and semantic information. This will lead to slower processing of the L2/Ln, as compared to more proficient L2/ Ln learners/users. Thus, at the basic, cognitive level, the lower proficiency of participants has an impact on cognate activation and cognate facilitation effects. However, the question what minimal proficiency level is needed for language activation in processing has not been so far constructively answered (de Bot, 2004).

Proficiency, fluency and the automaticity of processing The increase in someone’s L2/Ln proficiency has consequences for word processing because it involves changes in the cognitive mechanisms that

110

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

control the activation and inhibition of the learner’s/user’s language systems (Van Hell & Tanner, 2012) and are connected with automaticity in processing the target language. According to Segalowitz (2003), automaticity is regarded as an important issue in all theories of cognitive skill acquisition, including language acquisition. The theory of adaptive control of thought (Anderson, 1983; Anderson & Lebiere, 1998) proposes that acquiring a skill involves a transmission from a stage characterised by declarative knowledge to a stage characterised by procedural knowledge (Segalowitz, 2003: 395). So, early stages of skill development rely on declarative knowledge (i.e. the knowledge ‘that’) and involve conscious control. As the learner practises, sequenced components of the skill become chunked and automatic. When ‘chunked,’ the components are processed as wholes, and hence become unavailable to awareness in the processing, which becomes fast and fluent. In Anderson’s (1983) terms, the declarative knowledge becomes proceduralised (i.e. turned into the knowledge ‘how’). Procedural knowledge does not involve conscious attention and becomes automatised. Achieving higher levels of proficiency in L2 acquisition leads to automaticity and fluency, which refers to ‘aspects of productive and receptive language ability characterised by fluidity (smoothness) of performance’ (Segalowitz, 2007: 181). Cognitive fluency is connected with access fluidity and attentional control. Access fluidity is ‘the process of connecting words and expressions to their meanings (often referred to as lexical access), a necessary component of fluid speaking, reading, and listening,’ while attention control is ‘the process by which a language user focuses and refocuses attention in real time as the message being communicated unfolds’ (Segalowitz, 2007: 182). Fluency develops over time, primarily through extensive experience and practice and is associated with a high degree of automaticity in language processing (Hulstijn, 2001). When some degree of automaticity is obtained, a cognitive activity can be performed without much attentional control (Segalowitz, 2003). Thus, automaticity achieved at higher levels of L2 proficiency leads to unconscious, effortless, and smooth processing. Such automaticity reduces the burden on short-term memory and facilitates the chunking of information into higherlevel units Segalowitz (1997). Conversely, at lower levels of L2/L3 development, more processes are attention-based. Such attention-based processing requires consciousness, attentional control and effort, and because of that usually takes more time. Attentional control implies some involvement of attention, and possibly awareness. As for learning foreign language words, the process involves building mental representations in the bilingual mental lexicon. Practising means mapping the L2/L3 segments onto their corresponding mental representations. Retrieving words from memory is first slow, and their production is not fluent. As the learner practises, more processes become automatic, which leads to more fluent and effortless production. In this sense, the processing

Lear ner-rel ated Fac tors Af fec t ing Le x ical CL I

111

of L2/L3 words improves as a result of practice and repetition, i.e. together with gaining automaticity. A language learner must not just establish mental representations for each L2/L3 segment (declarative knowledge) but must also practise enough to make the use of these segments automatic and fluent (procedural knowledge), to attain high levels of L2 proficiency. Thus, experience with a foreign language, connected with fluency and proficiency in the language, typically leads to faster processing. In real life it is visible in faster rates of speaking and reading and in the ability to process rapid speech. In laboratory conditions, experience with a foreign language results in faster lexical decision times, faster word recognition, etc. On the other hand, automaticity in word processing may be different from automaticity when processing texts and discourse. According to Segalowitz (2003), achieving fluency that leads to automaticity implies that some kind of restructuring of processing has taken place. This restructuring may involve fine-grained changes in single word processing, or more holistic, higher-level changes such as restructuring involved in utterance comprehension. However, the restructuring may occur at one level but not on the other. This finding has implications for vocabulary development in the classroom. First, the learner must have fluency of access to his basic vocabulary in order to be able to learn further vocabulary from context (Meara, 1993). This means that a certain threshold of automaticity in accessing basic vocabulary must be crossed before additional learning can take place, because learning vocabulary from context requires a high degree of fluency in reading and listening (Segalowitz, 2003). Higher-level restructuring is needed for the learner to learn words through reading and listening. Possibly, this might also be the case with cognates: before a certain automaticity threshold is passed, cognate vocabulary may not be useful for the learner. Having passed the threshold in L2/Ln proficiency, the learner gains more smooth, automatic access to words, which frees his/her resources needed for taking a conscious, strategic advantage of cognates in further language learning. Interestingly, studies on vocabulary learning show that learners with low proficiency in L2 may not benefit from the cognateness of the L2 words presented if they have not reached the stage when they are ready to learn the word (Milton, 2009; Vidal, 2011). So, it appears that indeed there is a certain threshold in language proficiency which must be passed before a given language can be considered a source for crosslinguistic consultation in the case of cognates. There is yet another crucial conclusion stemming from the discussion of proficiency, automaticity and language activation. When interpreting results of classroom and laboratory experiments and tests on bilinguals, it is worth remembering that psycholinguistic experimental studies measuring RT involve different kinds of knowledge than some classroom-oriented research. While laboratory experiments on L2 processing tap into procedural knowledge at the deep, cognitive level, written language tests reveal the learner’s

112

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

declarative knowledge. Thus, it is quite possible that laboratory experiments will not provide answers concerning the phenomena of CLI which will be entirely compatible with those from classroom-oriented research.

Proficiency and L2 status in CLI In SLA studies, there is ample evidence that learner’s language proficiency and transfer in all language subsystems are related: less proficient L2 learners transfer more from their L1. Such transfer is often form-based (Jarvis, 2009; Jarvis & Pavlenko, 2007; Odlin, 1989; Ringbom, 2007). Research shows that not only proficiency in L2 but also in L3 and other languages may affect crosslinguistic activation and transferability. Language proficiency also seems to be one of the most crucial factors that could interact with psychotypology in conditioning CLI. When acquiring a new language, one of the learner’s dominant, or stronger languages serves as an auxiliary language. As Ringbom (1986: 155, added emphasis) points out, ‘the less the learner knows about the target language, the more he is forced to draw upon any other prior knowledge he possesses.’ The prior language knowledge involves not only the learner’s L1 and L2, but also any other language previously learned. In L3/Ln acquisition, learners tend to transfer elements from L2, or any other non-native languages whose level is high enough (De Angelis, 2005; Singleton, 1987; Williams & Hammarberg, 1998). If the L2 proficiency is very high, or even near native, the L2 influence may be similar to the influence originating from L1. For instance, Williams and Hammarberg (1998) ran a longitudinal case study of a learner whose L1, L2 and L3 were typologically related. They found that L2 influence (form-based) was strongest in the earliest stages of the L3 acquisition but gradually diminished as the L3 acquisition progressed. At the subconscious level, and during the early stages of L3 acquisition, there is a tendency among multilingual learners to activate their L2 instead of L1, which becomes decisive as the source language for transfer of lexical items, especially short function words. Williams and Hammarberg (1998) called this tendency ‘the L2 status.’ According to Hammarberg (2001: 23), ‘there appears to be a general tendency to activate an earlier secondary language in L3 production rather than L1’ as a supplier language. Thus, L3 learners will choose their L2 as the auxiliary language for two reasons: psychotypological and strategic. First, learners unconsciously want to sound more foreign, and their non-native languages all come into the ‘foreign’ category. If their mother tongue does not sound ‘foreign’ enough, it is suppressed in the mind of the learner, and the L3 learner draws upon their L2. The second argument for the L2 status is that the acquisition mechanisms used in L3 learning are, in fact, similar to those of L2 learning. If formal learning takes place, the mechanisms involved include conscious processes and focusing on form (declarative knowledge). For this reason, L2

Lear ner-rel ated Fac tors Af fec t ing Le x ical CL I

113

learning clearly contrasts with L1 acquisition, which is innate and mostly communicative in nature and automatised (procedural knowledge). As a result, learners draw on their experience in L2 learning because learning L3 seems analogous to learning L2. This line of arguments has been further developed by Bardel and Falk (2012). In their studies of L2 status as a factor in the learning of L3 syntax, Bardel and Falk (2007, 2012) and Falk and Bardel (2011) examined learners with L2 German or English learning L3 Swedish or Dutch. They noticed that in the case of typologically similar languages, in the initial stages of learning, L2 will be preferred as a CLI source in L3 syntax. According to Bardel and Falk (2012), the reliance on L2 is an outcome of the higher degree of cognitive similarity between L2 and L3 than between L1 and L3. They explain that formally learned L2 and L3 have many common cognitive features that they do not share with an L1. For instance, L2 and L3 are learned in similar settings and manners. Also, a foreign language learner, contrary to L1 acquirer, is aware of his/her learning and conscious of using various learning strategies. In their further explanations of the L2 status Falk and Bardel (2012) refer to Paradis’ (2004, 2007) declarative vs. procedural memory systems (see further in this chapter). They claim that L2 is elected for transfer because it is declarative, and therefore, more easily accessible. Interestingly, Falk et al. (2013) show that the L2 status may be overcome by L1 even at the beginning of L3 acquisition if the learner has a high level of explicit metalinguistic knowledge of L1 which is typologically related to L3. Then the declarative knowledge of L1 may aid in producing correct L3 structures. The L2 status hypothesis has been explored by qualitative and quantitative studies, e.g. Dewaele (1998), Müller-Lancé (2003), Hall and Ecke (2003) and Gibson and Hufeisen (2003). In contrast, De Angelis and Selinker’s (2001) study reveals that typological proximity is enough by itself to influence the language activation process, while the factor of proficiency does not necessarily have to be present. In a study by Müller-Lancé (2003), based on data form questionnaires and think-aloud protocols, he found that in the production of Romance foreign languages (Spanish and Italian) German native speakers avoided drawing on their L1. On the other hand, native speakers of Romance languages drew on their L1 quite heavily. Hall and Ecke (2003) examined crosslinguistic aspects of the productions in German (L3) speakers of Spanish (L1) with intermediate to high levels of proficiency in English (L2). They discovered evidence of crosslinguistic influence from both Spanish and English, but found that most effects came from English (L2), which is typologically and psychotypologically closer to German than Spanish.

Frequency and recency of use as factors in CLI CLI can also be related to the frequency and recency of language use. Here, we will focus on three studies of the effects of recency on transfer.

114

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

Dewaele (1998) analysed the lexical inventions of Dutch-speaking learners of French. Some of them had learned English as L3 after learning L2 French, and some learned French as L3, after English as L2. In either case, when producing French, the learners relied on the language they had learned prior to French: L1 Dutch in the case of learners of French as L2, or L2 English in the case of learners of French as L3. On the other hand, Hammarberg (2001) argues that learners are more likely to transfer elements from a language they are actively using than from languages they know but are not using actively. Probably, this kind of CLI is caused by higher proficiency in the active language, relative to other languages, or a larger degree of relatedness between the activated and the target language. These factors may also coincide and work together to influence the target language production, as in the already mentioned case study by Williams and Hammarberg (1998). Here, an English-speaking person who had also studied German was followed in her acquisition of Swedish. Surprisingly, the majority of unintentional language switches in her language production came from German. Considering that the subject had been using English more frequently than German while in Sweden, and still favoured German over English in her production of Swedish, the researchers argued that the recency of use can, in fact, override the frequency of use effect. Concluding, it seems that proficiency in L2 and L3/Ln must be high enough for the language to become activated and to become a source of CLI and to aid the learner in using cognate vocabulary. Although the minimal proficiency threshold is still unknown, proficiency clearly cooperates with typology, recency of use and ‘the L2 status’ in shaping language learning and production. In this respect bilingual and multilingual language users will differ in terms of their cumulative learning experience and the number of languages that can interact. These phenomena will be discussed in the following sections.

Advantages of Cumulative Learning Experience Bilinguals versus monolinguals From the psycholinguistic point of view, language learning experience changes the speaker. As extensively discussed in Chapter 1, ‘the bilingual is NOT the sum of two complete or incomplete monolinguals’ (Grosjean, 1989: 6, original emphasis), and so their language knowledge will differ from that of monolinguals. There is an abundance of current psycholinguistic evidence that proficient bilinguals, both children and adults, differ from monolinguals in a number of ways (see e.g. Bialystok et al., 2009 for an extensive overview). This is due to the fact that bilinguals are faced with a challenge of juggling two languages (switching), or suppressing one language while using another

Lear ner-rel ated Fac tors Af fec t ing Le x ical CL I

115

(inhibition). Apart from their experience of monitoring two languages, bilinguals also have to monitor the context, speech of their interlocutors and other environmental circumstances to inhibit the language that they are not using at the moment, and they largely succeed at using one language and not another. On the other hand, even balanced child bilinguals are often less proficient in both of their languages and have more limited vocabularies than monolingual peers. They may also be less fluent than monolinguals (Bialystok et al., 2009), and even proficient L2 users will probably never reach the competence of a monolingual native speaker of that language, but their L2 will also have an impact on their use of their L1 (e.g. Cook, 2002; Ewert, 2009). This all clearly makes bilinguals different from monolinguals with respect to their language knowledge and use. Language acquisition experience shapes not only the linguistic but also the cognitive abilities of people who use more than one language. The list of cognitive benefits of bilingualism includes their increased verbal and nonverbal intelligence, divergent thinking, performance in non-verbal perceptual tasks and in a variety of tasks that involve metalinguistic knowledge (Hamers & Blanc, 1989). Bilingualism leads to increased creativity from the early years. For instance, when studying bilingual children, Bialystok (1991) observed that they manipulate knowledge in a different way to their monolingual peers. While monolingual students are stronger at acquiring informational knowledge, which requires good memorisation skills, bilinguals are better at operational knowledge which pertains to applying the existing knowledge in new situations, thus it requires creativity in its purest sense. A lot of evidence concerning differences between monolingual and bilinguals language users comes from studies of the attentional processes in bilinguals (Bialystok et al., 2009; Costa et al., 2006). Already bilingual children have better control of processing information than their monolingual peers (Bialystok, 1991). Studies on proficient bilinguals who use their languages on a regular basis show that both their languages are active, even when there is no need to use one of the languages (e.g. Dijkstra et al., 1999; Kroll et al., 2006; Van Hell & Dijkstra, 2002). Thus, language production in bilinguals demands constant control to allocate attention to the language that is needed, while suppressing the other one. This results in their superior abilities in tasks that require executive control (see Kroll & Bialystok, 2013 for an extensive overview), switching attention (see Bobb & Wodniecka, 2013 for an extensive overview) and give them advantage over their monolingual peers in tasks that require memory (e.g. Wodniecka et al., 2010). What is important, the experience of monitoring languages results in life-long advantages that also last in adult and elderly bilinguals and even delays the syndromes of dementia (Bialystok et al., 2007; Bialystok et al., 2009). Superior cognitive advantages and analytic skills of bilinguals aided by the fact that they already use more than one language may lead to increased

116

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

easiness in language learning. A selection of SLA studies have demonstrated the advantage of bilinguals learning a foreign language as their L3 over monolinguals learning the same language as their L2. For instance, Ringbom (1987) examined 11,000 English language national exam papers and found out that Finnish-Swedish bilingual students outperformed Finnish monolinguals. Thomas (1988) studied Spanish-English bilinguals and Englishspeaking monolinguals learning French in the classroom context. Bilinguals performed significantly better than on various measures of French language achievement. In research conducted by Cenoz and Valencia (1994) SpanishBasque bilingual children outperformed monolingual children in learning English. In a Swiss context, Brohy (2001) demonstrated that RomanschGerman bilinguals outperformed German monolinguals when learning French. It seems that as learners acquire proficiency in L2, they gain not only accuracy and fluency in using the language, but also cumulate more cognitive experience, which in turn facilitates language acquisition.

Bilinguals versus multilinguals There is an ongoing debate whether bilinguals differ from multilinguals in a substantial way. In the mid-1980s Baetens Beardsmore (1986: 3) stated that ‘there is no evidence to suggest that the fundamental principles affecting language usage are any different whether two, three or more languages are being used by one and the same speaker.’ Indeed, until the late 1990s, multilingualism was frequently treated as a variant of bilingualism. This tendency could be ascribed to the strong focus on researching the acquisition of L2 and the lack of studies on the acquisition of the subsequent languages (Jessner, 2008). In psycholinguistics, multilinguals are still treated as a subset of bilinguals (Cenoz et al., 2003; Van Hell & Tanner, 2012). This is probably due to the fact that there is little evidence from psycholinguistic studies showing that multilinguals differ from bilinguals in a considerable way in language processing. From studies involving cognate vocabulary we know that when three languages rather than two are involved in processing, cognate facilitation effects may still arise between all three languages, provided that proficiency in all three is high enough (Lemhöfer et al., 2004; Van Hell & Djikstra, 2002). Similar cognate facilitation effects also occur in child bilinguals and trilinguals (Poarch & Van Hell, 2012). The effects can also be found for cognates in three typologically unrelated languages such as Polish, English and German. Processing the weakest language (L3) words entails automatic, parallel activation of candidate words in the dominant, stronger languages, both L1 and L2 (Szubko-Sitarek, 2011). Thus, Dijkstra (2003) states that there is no need for a specific multilingual model of language processing as multilinguals do not require any special processing mechanisms during word selection which are different from the proposed bilingual models.

Lear ner-rel ated Fac tors Af fec t ing Le x ical CL I

117

While the extension of the existing bilingual processing models may suffice to account for multilingual word processing in experimental paradigms, intuitively, it seems there must be a difference between the real-world linguistic functioning of bilingual and multilingual language learners and users. The differences are connected with the growing number of the languages involved, growing learning experience, and the assumption of the role played by crosslinguistic influence within a dynamic system of the learner’s knowledge see DMM, (Herdina & Jessner, 2002). However, it may well be that the differences between bilinguals and multilinguals do not pertain to language processing but solely to the qualitative differences in the linguistic functioning of multilinguals, which result from their cumulative learning experience. Each next language in the learner’s repertoire unavoidably increases the risk of complications and the cognitive effort required to maintain the whole system. However, there are also linguistic and affective factors which combine into a set of powerful resources available to L3 learners who already have had bilingual experience. These factors will include the knowledge and awareness of another foreign language, a wealth of learning strategies, growing confidence in the language learning situation, and lesser language anxiety in comparison with less experienced learners (Gabrys´-Barker, 2005). All factors combined, cumulative learning experience may also result in the increase of learner’s/user’s language awareness, and specifically metalinguistic awareness. These issues will be discussed in the following section.

Language Awareness, Metalinguistic Awareness and Language Use Language awareness is more of an umbrella term than a homogeneous notion. What is more, in current literature on the subject ‘awareness’ and ‘consciousness’ can either cover roughly the same range of meanings (e.g. Svalberg, 2007), or they may refer to two different issues. English ‘awareness’ and ‘consciousness’, although both rooted in metacognition, are not regarded as synonyms. Consciousness is seen as a process, and awareness as a state of knowledge resulting from this process (Jessner, 2008a). The following understanding of the notion is proposed: ‘Language Awareness can be defined as explicit knowledge about language, and conscious perception and sensitivity in language learning, language teaching and language use’ (www.lexically. net/ala/la_defined.htm). It focuses attention on what the learner already knows and on making the knowledge explicit. Metalinguistic awareness is a subcategory of language awareness and a part of meta-cognition which is responsible for one’s knowledge about language defined as ‘the ability to focus attention on language as an object in itself or to think abstractly about language and, consequently, to play with or manipulate language’ (Jessner,

118

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

2006: 42). Consequently, metalinguistic awareness is a specific cognitive skill that accounts for an ability to reflect on language, to understand linguistic patterns and rules and to use them. Metalinguistic abilities may be defined in terms of two underlying skill components, which refer to different aspects of language processing, and which are involved in all aspects of language use: the analysis of knowledge and control of processes. The analysis of knowledge can be defined as ‘the ability to construct explicit representations of linguistic knowledge’ (Bialystok, 1987: 155). High levels of analysis of knowledge are associated with conscious and/or explicit knowledge. The control of processes can be defined as ‘the ability to control linguistic processes by intentionally selecting and applying knowledge to arrive at a solution’ (Bialystok, 1987: 155). High levels of processing control are associated with constructs such as intentionality. Further on, Bialystok states that analysis of knowledge is responsible for making explicit those representations that had previously been implicit or intuitive. The knowledge of procedures for solving a variety of language problems and the ability to execute those solutions through appropriate attentional focus is the function of control of linguistic processing. (Bialystok, 1987: 156) Bilingualism involves an ability to use two language systems and develop skills in controlling the processes and analysing knowledge from two language systems. Thus, the logical result of bilingualism is that the use of two languages influences metalinguistic abilities and metalinguistic awareness. Indeed, bilinguals effectively compare and analyse languages, translate words and see their ambiguity. These skills are essential to the problemsolving process concerning language itself (Bialystok, 1988, 1991, 2007). Moreover, bilinguals are better at recognising words that have more than one meaning, as well as performing better in ambiguous communication situations (Bialystok, 2007; Bialystok et al., 2009). Thanks to their greater awareness of language patterns they are also more sensitive to proper language use in interpersonal communication, which is caused by their greater sensitivity in interpersonal communication (Aronin & Singleton, 2012; Jessner, 2006). Language use, however, depends on gradual development of each skill and proceeds in response to different experiences. For instance, although children appear to possess a relatively sophisticated command of linguistic concepts, their explicit knowledge of these concepts is inadequate for solving most metalinguistic tasks (Bialystok, 1987). Metalinguistic awareness depends not only on the fact of being bilingual, but also on the cognitive development and proficiency in both languages. Children’s language systems and cognition may not be developed to the extent that allows metalinguistic awareness to manifest itself at its fullest. Also, their abilities to recognise formal crosslinguistic similarity are smaller than those of adolescents and

Lear ner-rel ated Fac tors Af fec t ing Le x ical CL I

119

adults, as indicated by an extensive study on age-related effects of cognate recognition by Vanhove (2014). Thus, further on, we will exclude children from the discussion and analysis. We will focus on adult and adolescent language learners/users and the role of their increasing metalinguistic awareness in learning subsequent languages thanks to making conscious crosslinguistic comparisons.

10 Reliance on Lexical CLI as a Strategic Behaviour

Introduction According to Paradis (2007: 23), ‘metalinguistic knowledge is learned by paying attention to what is to be learned and by consciously trying to remember it.’ On the basis of discussion in Chapter 9, we can postulate that cumulative language learning experience, i.e. experience with learning and the acquisition of subsequent languages combined with proficiency in them, results in increased metalinguistic awareness because it requires more control processes of the language systems and more conscious analysis of knowledge. A conscious reliance on knowledge from various language systems will, in turn, enhance better language learning strategies. The growing metalinguistic awareness of a bilingual and multilingual language user/learner often shapes his/her learning strategies and a conscious search for crosslinguistic similarity. Therefore, making use of crosslinguistic similarity in language learning may be a conscious strategic form of behaviour developed when the learner becomes aware of similarity, and how it may boost learning. This chapter discusses the issues of strategic search for lexical crosslinguistic similarity. It discusses awareness, attention and noticing of cognates as a learning strategy. Finally, it finishes with a discussion on the theory of affordances and the role of multilingualism in taking strategic advantage of crosslinguistic affordances.

Strategic Learning in Bilinguals and Multilinguals O’Malley and Chamot (1990: 1) define learning strategies as ‘the special thoughts or behaviours that individuals use to help them comprehend, learn, or retain new information,’ while Oxford (1992, 1993: 18) calls them ‘specific actions, behaviours, steps, or techniques that students (often intentionally) use to improve their progress in developing L2 skills. These strategies can 120

Reliance on Le x ical CL I as a Strategic Behav iour

121

facilitate the internalisation, storage, retrieval, or use of the new language. Strategies are tools for the self-directed involvement necessary for developing communicative ability.’ Although there are numerous extensive classifications of strategic learning types, all emphasise the role of cognition, metacognition and affectivity in strategy choice. Research on learning strategies (e.g. O’Malley & Chamot, 1990; Oxford, 1990; Wenden & Rubin, 1987) resulted in the concept of self-regulated learning. Introducing the concept of self-regulated learning for any proactive behaviour on the learner’s part, Dörnyei (2006: 61) states that ‘what makes strategic learners special is not so much what they do as the fact that they choose to put creative effort into improving their own learning and that they have the capacity to do so.’ Strategically self-regulated learners actively participate in learning, achieve goals while controlling various aspect of their learning, choose appropriate strategies for different conditions, purposes and setting, and show awareness of the relationship between strategy use and learning outcomes (see also Oxford, 2011 for an extensive overview). There is plenty of evidence that bilinguals and multilinguals employ more strategic behaviour when learning languages than monolinguals. De Angelis (2007) provides an overview of studies comparing the metalinguistic abilities of monolinguals and bilinguals, where proficient bilinguals have proved to develop heightened awareness of the forms, meaning and rules of the language (e.g. Ben-Zeev, 1977; Bialystok, 1987, 1988, 1991, 2001; Ricciardelli, 1992; Yelland et al., 1993). It also seems that the knowledge of yet another language (L3) results in more metalinguistic awareness: conscious, strategic behaviour is especially enhanced in multilinguals. Jessner (2006: 70) argues that ‘the metalinguistically aware multilingual learner explores and analyses points of commonality between her or his language systems to obtain the target language item. (. . .) multilinguals do this in an enhanced way since they have more resources they can draw on.’ Already at the beginning of the 1990s, Oxford (1990) found that multilinguals demonstrated greater awareness of their preferred learning style and used more learning strategies than bilinguals. In a study by Nayak et al. (1990), multilingual subjects turned out to adjust learning strategies more effectively to the purpose of specific tasks than monolingual subjects. The study was carried out on a group of monolingual and multilingual subjects learning a miniature artificial linguistic system. Participants were also asked to verbalise strategies used while performing the task. The results showed that in learning the rules for syntax, multilingual subjects outperformed monolingual subjects. Moreover, multilinguals were also better at structuring their strategies to the task, and used a wider variety of different strategies. However, on the whole multilinguals did not appear to have superior learning abilities over monolinguals. Nayak et al. (1990: 242) concluded that ‘one reason for the superior performance of the multilingual subjects is a greater flexibility in switching strategies.’

122

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

Numerous consecutive studies oriented at third and additional language acquisition converged at concluding that the variety and scope of strategies employed by bilinguals and multilinguals is connected with their metalinguistic awareness, which for instance, turned out to play a role both in the comprehension and production monitoring strategies (e.g. Thomas, 1992), where learners transferred elements of their knowledge about other languages to make more informed decisions with respect to the new language. Thomas (1992) argued that students’ prior linguistic experience influences the strategies which they subsequently adopt, their level of consciousness about which strategies may be effective and their ultimate success in learning a foreign language. Further on, Mißler (1999, 2000), found out that multilingual language learners outperformed monolinguals in the application and variety of learning strategies. In studies by Ó Laoire (2001) and Aronin and Ó Laoire (2004), Irish-English bilinguals used learning strategies more frequently than English monolinguals when learning German and French. Hufeisen (1998) concluded that multilinguals used learning strategies extensively in a variety of production and comprehension activities and they regarded these strategies as a key factor for their success in learning. All in all, it seems that multilinguals may not have any additional cognitive capacities for language learning, but are better at choosing the right strategy for the task.

Metalinguistic Awareness and Reliance on Cognate Vocabulary Strategic reliance on cumulative language experience Classroom studies on cognate learning (see Chapter 6) show that recognising and using cognates may involve a strategic behaviour which results from the individual learner’s metalinguistic awareness (e.g. Dressler et al., 2011; Nagy et al., 1993; Otwinowska-Kasztelanic, 2009). Jessner (1999: 204) states that ‘the search for similarities between languages can be seen as a part of the activities related to metalinguistic thinking in the learner.’ Studies based on the L2 classroom also emphasise the need to train learners who are less advanced in L2 to recognise cognates and capitalise on crosslinguistic similarity. Research shows that training to make conscious use of cognate words may enhance mastering L2 vocabulary because it increases learners’ metalinguistic awareness necessary to benefit consciously from cognateness (Haastrup, 1991; Swan, 1997; Jessner, 1999; Otwinowska-Kasztelanic, 2009; Ringbom, 2007). On the other hand, it seems that beyond a certain proficiency level in L2 and L3, adult multilinguals do not need training to recognise cognate vocabulary. Such increased linguistic awareness in multilinguals is described as an emergent property of multilingual proficiency, whose one component is crosslinguistic awareness focusing specifically on cross-language

Reliance on Le x ical CL I as a Strategic Behav iour

123

comparisons and contrasts (James & Garret, 1991; Jessner 2006). The main argumentation behind the beneficial impact of crosslinguistic awareness on the learning process revolves around the observation that when acquiring and using another language, learners intuitively, or consciously make use of every possible source of information already at their command (Ringbom, 2007). As discussed in the previous chapters, in particular they refer to the knowledge of their L1 and other foreign languages. So, crosslinguistic awareness refers to the learner’s tacit and explicit realisation that there are links between their language systems. As pointed out, crosslinguistic similarity is ‘an important variable in the use of learning strategies: how the learner tries to enhance the effectiveness of learning’ (Ringbom, 2007: 104). Also, ‘the more aware learners are of the similarities and differences between their mother tongue and the target language, the easier they will find it to adopt effective learning and production strategies’ (Swan, 1997: 178). So, we may assume that increased crosslinguistic awareness, being a part of metalinguistic awareness, predisposes multilinguals to recognise cognateness in an enhanced way. There is a link between the character of CLI and the enhanced level of metalinguistic awareness in multilingual subjects. This relationship appears especially prominent in the learners’ ability to adapt and apply familiar crosslinguistic strategies while learning another language. Jessner (1999) gives examples of such metalinguistic thinking, conscious language switching and conscious reference to cognates in the case of Italian-German bilinguals advanced in English as L3 who were supposed to think aloud while competing an academic writing task. The conscious switching and search for crosslinguistic similarity is illustrated by the example from the think-aloud protocol below: OK, this is proved, no this is sustained, #sostenere, sustained by the theory that all our 4000-6000 languages on earth, hmm are expected to be all the same for an external, how do you call it, *Beobachter, observer. (English → Italian → English → German → English). (Jessner, 199: 204)

Strategic reliance on lexical competence in the languages known The efficient receptive and productive strategies of multilinguals correlate with the previously developed lexical competences in other languages. For instance, in a multi-aspectual study on the acquisition of Romance languages by speakers of Germanic languages, Müller-Lancé (2003) examined the productive and receptive strategies of language learners. Productive strategies referred to their use of vocabulary-based or context-based techniques including transfer from other languages, while receptive strategies were connected with inferencing the meaning of words from various cues. The results showed that most receptive strategies were based on inferencing from the

124

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

previously acquired languages, but L1 turned to be of lesser importance. Also, productive strategies, i.e. the retrieval of words was supported by the knowledge of vocabulary from other languages. Other studies showed similar results concerning vocabulary: L3 learners consciously drew on their L2 when checking the correctness of a given word (De Angelis & Selinker, 2001), and their L2 served as lexical resource to overcome vocabulary deficits while producing L3 (Jessner, 2006; Tremblay, 2006). The results of these studies seem to imply that the development of a special kind of awareness available to students who have already mastered two languages provides the necessary conditions for a successful learning of further languages (Jessner, 2008). Research findings also support the assumption that the amount of language learning experience tends to facilitate further multilingual development. It seems that the higher the multilingual proficiency, the easier it is for them to consciously spot crosslinguistic similarity patterns, reflect on the nature of learning and transfer useful elements from language to language. Indeed, it has been demonstrated that the more proficient the language learner is in L2 and L3-Ln, the more experienced he/she is in searching for crosslinguistic similarity and that both factors of proficiency and multilinguality may interact. In a study already mentioned in Chapter 6, Berthele (2011) reports on a number of lexical inferencing tasks involving the recognition of cognates from an unknown, typologically close language from the Germanic or Romance family. In the study, the more multilingual a participant was, the more likely he/she was to correctly guess the meaning of the cognate word. Moreover, multilinguals advanced in at least two of their languages did significantly better at inferencing than less proficient multilinguals. Finally, multilinguals with high proficiency in two languages that were typologically close to the target performed better than all other groups. As concluded, ‘quality increases with increasingly proficient multilingual systems, and (. . .) the main process that enables multilinguals to be efficient in the tasks is (. . .) an inferencing process that exploits multlilinguals’ knowledge of what are likely and of what are unlikely correspondences across and within languages’ (Berthele, 2011: 216). The study demonstrated that recognising and taking advantage of cognates is clearly influenced by proficiency in L2 and L3-Ln, both at the level of vocabulary knowledge and the level of metalinguistic awareness. As has been shown, metalinguistic awareness grows with cumulative language learning experience. Multilinguals beyond a proficiency threshold in L2 and L3 should display higher metalinguistic awareness than bilinguals and should be able to spot more crosslinguistic similarities. Ecke (2014) draws our attention to the possible mechanism that influences vocabulary learning in L3/Ln which is based on the cumulative learning experience of the learner. He suggests that both crosslinguistic similarity and L2 status combine in lexical CLI since these effects rely on a cognitive mechanism that may be called parasitism in the mental lexicon (Hall, 2002),

Reliance on Le x ical CL I as a Strategic Behav iour

125

or parasitic vocabulary learning (Hall & Ecke, 2003). Learners universally search for, detect and use similarity between the known and the new lexical items, trying to integrate the new knowledge into the existing network of lexemes, lemmas, semantic representations and access routes between them (see Chapter 6 for the models of the mental lexicon and Chapter 5 for lexical CLI). On the one hand, the process of parasitism aids learning L3 words through establishing crosslinguistic identifications at the various levels and by basing on the existing processing routes. On the other hand, it also heavily constrains L3 vocabulary learning by establishing assumed parallels between words at the level of lexemes, lemmas and concepts, which may become fossilised (Ecke, 2014; Hall & Ecke, 2003). The Parasitic Model works as follows. First, form (lexemic) representation is established. The L3 word nodes (see the Distributed Features Model in Chapter 6) are activated and matched to the existing ‘host lexeme’ nodes in L2 or L1 if some threshold of similarity has been passed. Possible differences may be detected, or the formal connection becomes fossilised. It may, for example, result in overusing L1-L3 or L2-L3 false friends. Next, connections are built at the lemmatic level (syntactic frame) on the basis of the existing similarities to the L2 or L1 ‘host.’ They are later revised thanks to of contextual information and new connections are established between the lemmatic and conceptual information. If not, the lexemic (frame information) may become fossilised, and the word in L3 will be used in accordance with the L1 or L2 syntactic pattern. The final process, is the strengthening and automatisation of representations and access routes for all three words in L1, L2 and L3 (Ecke, 2014). The model merges the existing research into L3 lexical transfer with the research on the possible architecture of the mental lexicon.

Metalinguistic awareness and typologically distant languages When crosslinguistic correspondences are not evident, metalinguistic awareness in noticing cognates may be especially needed. This happens if the learner’s L1 and the target language are typologically distant because for such pairs of languages the number of close, similar crosslinguistic neighbours decreases, as discussed in Chapters 4 and 8 (Djikstra, 2003; Schepens et al., 2012). This is important to the extent that the actual typological similarity/ difference relations between languages reflected in the similarity of cognate forms may influence learner judgments concerning the languages involved (Odlin, 2002). If learners view typological distance subjectively, according to their individual perceptions (Kellerman, 1978, 1983, 1986), the smaller the psychotypological distance, the easier it is for them to recognise cognate forms as congruent between L1 and L2 (Ringbom, 2006, 2007). This, in turn, can facilitate acquisition of a target language which is typologically close to L1. The proficiency factor seems crucial at this point. At lower proficiency levels, learners of languages that are more distant typologically from their L1

126

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

may not perceive L1 and L2 as close enough to be similar and may not recognise cognates, unless they are identical, or almost identical with their L1 equivalents. Since psychotypological judgments are not fixed, but may be revised with growing L2 proficiency and metalinguistic awareness (Kellerman, 1979), more proficient and more multilingual learners may still view the languages in question as distant but are able to spot crosslinguistic similarities (De Angelis, 2007; Jessner, 2006). Research shows that also in typologically distant languages multilinguals with high proficiency in L2 and other languages, tend to display higher metalinguistic awareness of crosslinguistic similarity between the languages they know and consciously use cognate vocabulary as a learning strategy (Gabrys´-Barker & Otwinowska, 2013; Otwinowska-Kasztelanic, 2011a, 2011b). Metalinguistic awareness can be said to increase with the number of languages known to the speaker and his/her proficiency, which, in turn, aids multilinguals in acquiring new languages (Aronin & Singleton, 2012a; De Angelis, 2007; Jessner, 1999, 2006; Singleton & Aronin, 2007). Thus, we can assume that the ability to consciously benefit from cognates depends on both: L2 proficiency and on the multilingualism of the learner, which both add to his/her language awareness.

Attention, Awareness and Noticing in Language Learning Attention, awareness and noticing Language awareness is connected with attention and noticing phenomena, both of which are related to memory processes. Let us first focus on attention and memory. Attention is viewed as a mainly conscious process involving working memory, which is not the same as short-term memory (Ellis, 2008). Working memory ‘involves the temporary storage and manipulation of information that is necessary for the performance of a wide range of cognitive tasks’ and it comprises mechanisms for the storage of information and mechanisms for the executive control of information. Conversely, ‘short-term memory is used to describe static memory where information is held for a short period of time (less than 20 seconds)’ (Biedron´ & Szczepaniak, 2012: 290-291). Attention is defined as, the process that encodes language input, keeps it active in working and short-term memory, and retrieves it from long-term memory. (. . .) The focus of attention is a subset of short-term memory. (. . .) Long-term memory is where instances of encoded input are stored and assume (. . .) representational shape that recognition processes match to new instances of input in working memory during parsing and comprehension. (. . .)

Reliance on Le x ical CL I as a Strategic Behav iour

127

Attention, then, can be viewed as a process for which memory provides structure and constraint. (Robinson, 2003: 631) There is a network of intricate relations between memory, attention and noticing that have not been fully examined yet. Schmidt (1990) and Robinson (2002) present noticing as conscious recognition connected with holding information in short-term memory. They both argue that noticing is the condition necessary for learning, and in consequence, for storing information in the long-term memory. They distinguish four dimensions of consciousness necessary for SLA in the case of adults. The first is intention, which makes learning a deliberate activity. The second is attention, understood as detection and concentration on selected stimuli and necessary for learning, while it concerns input processing. The third manifestation of consciousness is awareness, associated with the learner’s knowledge, or subjective experience of registering a stimulus. Finally, there is control over performance, which participates in output processing. However, it still remains an open question if noticing and learning are possible without awareness (cf. Krashen, 1982). In Schmidt’s (1990, 1994a) terms, learning only happens as a result of noticing, which, in turn, results from attention and awareness. Also Robinson (2003: 635) argues that although attention cannot be regarded as an autonomous function, selection of linguistic input to learn is clearly an ‘aspect of attentional control, guided by the supervisory attentional system, and executive control mechanisms.’ Those mechanisms are involved in ‘task analysis, selection and control of cognitive and metacognitive strategies for performing the task, and in monitoring the effectiveness of strategies.’ Schmidt (1990, 1994a) and Robinson (2002, 2003) treat noticing connected with attentional control as indispensable in adult language learning. This strong position is often challenged because it undermines the existence of implicit learning (Al-Hejin, 2004; Krashen, 1982; Svalberg, 2007). Nevertheless, Al-Hejin’s (2004) meta-analysis of attention and awareness literature shows that attention is an important factor present in the process of learning. Research has been only able to demonstrate that less attention leads to less learning. As for the role of awareness in relation to learning, it has been pronounced beneficial rather than necessary. However, language awareness is positively correlated with working memory capacity and language aptitude (Biedron´, 2008; Biedron´ & Szczepaniak, 2012), so we can assume its importance for learning a language.

Explicit, declarative knowledge and proceduralised knowledge According to Schmidt (1990, 1994a), sheer noticing may not be enough for learning, since learning needs more effective and explicit declarative knowledge of language rules. Thus, noticing may be a condition necessary for the input to become intake in language learning, provided it is perceived as a personal reference (a subjective experience) and explicitly reflected upon.

128

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

As described by Paradis (2004, 2007), explicit, declarative knowledge and proceduralised knowledge depend on two kinds of memory processes and are supported by two different neural mechanisms. One supports the acquisition and use of skills, and the second is responsible for the learning and use of conscious knowledge. This distinction leads to distinguishing two language functions related to the two kinds of memory processes: implicit linguistic competence (i.e. fluency and accuracy of language use) and explicit metalinguistic knowledge (i.e. the knowledge ‘about’ language). Paradis (2004) argues that implicit memory allows automatic and involuntary processing, while explicit memory operates through controlled processes that undergo voluntary execution. Paradis further states that implicit linguistic competence is acquired incidentally, is stored implicitly, is used automatically, is subserved by procedural memory and is taskspecific. (. . .) Explicit knowledge, on the other hand, is learned consciously, is stored explicitly, is consciously controlled when used, is sustained by declarative memory, and is not modality specific. (Paradis, 2007: 23) Thus attention and noticing may be factors necessary for gaining explicit knowledge leading to language awareness and the strategic learning of a language. Attentional focus on an element of input allows for its entering into the working memory where the rehearsal process is taking place. Noticing is a consequence of this rehearsal, thanks to which information can be sustained by long-term, declarative memory. If an element to be learned must be consciously attended to, noticing may depend on proficiency and automaticity in that language. As Ellis (2008: 755–756) points out, neural and cognitive mechanisms involved in attention and memory have obvious implications for L2 learning: ‘linguistic input that is salient can elicit attention (. . .) but also a learner may fail to attend to novel features simply because they are not familiar.’ As discussed in Chapter 7, some learners indeed fail to notice such language features as crosslinguistic similarities, so cognate vocabulary often remains unnoticed by learners (Banta, 1981; Dressler et al., 2011; Kellerman, 1983; Lightbown & Libben, 1984; Nagy et al., 1993; Odlin, 1989; Schmitt, 1997; Singleton, 2006; Swan, 1997). Thus, conscious benefitting from cognates as a strategy for learning may be a matter of attention, noticing and awareness.

Crosslinguistic Lexical Similarities as Affordances The theory of affordances The discussion in the previous sections has shown that making use of crosslinguistic similarity, thanks to increased language awareness, is a strategic form of behaviour typical of more advanced L2 learners and those with more cumulative language learning experience. According to Singleton and

Reliance on Le x ical CL I as a Strategic Behav iour

129

Aronin (2007), multilinguals are better at noticing crosslinguistic similarity because they can capitalise upon ‘affordances’ offered by the multiple languages that they know (Aronin & Singleton, 2010, 2012a, 2012b; Singleton & Aronin, 2007), as discussed below. The theory of affordances is a theoretical construct stemming from perceptual psychology. Proposed originally by a psychologist, James Gibson (1977, 1979), the theory concerned the mutual relations between the organism and its environment in the area of visual perception. Contrary to other theories of the time, which viewed perception as a process where an organism processes the image formed by the stimuli in the environment, Gibson proposed that perception of the environment is also the perception of the self. According to his theory, the environment and the organism are mutually complimentary in such a way that the organism perceives its environment as a set of possibilities the environment provides, or affords. Thus, an affordance is ‘a specific combination of the properties of [the environment’s] substance and its surfaces taken with reference to the animal’ (Gibson, 1977: 67). In other words, affordances are the perceived opportunities for action provided for the observer by an environment. That is why affordances can be understood as the possibilities that an object, or environment offers (or appears to offer) to the organism for action, or functioning, or chances for the organism to fulfil its goals. The theory began to be applied to domains as diverse from psychology as industrial design, human–computer interaction, and finally language acquisition. Within the field of interaction design, Norman (1988, 1999) coined the term of ‘perceived affordance’, which links affordances not only to the physical capabilities of the acting agent, but also to his/her goals, plans, values, beliefs and past experience. As he puts it: ‘the term affordance refers to the perceived and actual properties of the thing, primarily those fundamental properties that determine just how the thing could possibly be used’ (Norman, 1999: 9). Effectively, Norman’s affordance ‘suggests’ how an object can be used and interacted with. As for connections between perceptual psychology and linguistics, Gibson’s construct had an impact on the development of a linguistic theory of situations (Barwise & Perry, 1983). Together with Norman’s perceived affordances, this theory has been used to point out that language, like any other environment, offers certain affordances to its users and plays an important role in communicating information about situational affordances (MacWhinney, 1999). When applied to the study of language acquisition, the theory of affordances sheds new light on the meaning and importance of awareness phenomena.

Awareness of affordances offered by language Segalowitz (2001: 15) notes that ‘affordances are important for learning, because it is only by being able to perceive affordances that an organism is able

130

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

to navigate its way around the environment successfully.’ According to him, a given language, like any physical environment, possesses affordances, and supports a particular set of constructions. However, these constructions are ‘available (. . .) if the speaker knows how to use them. These constructions afford the possibility of making certain messages but not others, and make some messages easier to communicate than others’ (Segalowitz, 2001: 15, emphasis added). The discussion on affordances is inherently connected with the role of awareness and noticing. Segalowitz (2001) claims that affordances are important for learning, on condition that the learner is aware of them, i.e. he/she can notice them and is ready to use them. So, for affordances to be perceived, observers must pay attention to the relevant sensory information: A situation provides a suitable niche only for those persons who are prepared to meet and use its affordances effectively. Those not properly tuned or prepared will in some way fail to perform effectively in the situation as given. (Snow, 1998: 107) In other words, language learners have at their disposal potential affordances connected with their language resources and their language-learning environments. Thus, learning or acquiring a language involves ‘attuning one’s attention system to perceive the communicative affordances provided by the linguistic environment’ (Segalowitz, 2001: 15-16). This framework ties up with Schmidt’s Noticing Hypothesis (1990, 1998, 2001), which states that language acquisition involves attention and noticing, as discussed above. Taking the discussion a step further, one can assume that without consciously noticing certain affordances available, the learner may not be able to use them. Van Lier (2000), who considers affordances with regards to actionbased language teaching and learning also emphasises the fact that affordances afford further actions, but do not cause or trigger them. This means that they can potentially be useful, but learners need to behave in a proactive way to take the chance that is offered to them. Following this thought, Tella and Harjanne stated that: Affordances speak a language of their own. Some actors can understand that language better than others. Others can be completely deaf to that language. (. . .) Linguistic affordances are for us to take advantage of, but they do not spontaneously engage us, unless we are active enough to notice them and proactive enough to start exploiting them. (Tella & Harjanne, 2007: 502)

Language affordances and cognate vocabulary Affordances are connected with the perception of certain opportunities and, as a result, for affordances to be perceived, the learner must be sensitive

Reliance on Le x ical CL I as a Strategic Behav iour

131

to the relevant information and attend to that information. Aronin and Singleton (2010, 2012a, 2012b) discuss language affordances, defined as ‘affordances through the realisation of which communication via a language or languages or the acquisition of language or languages is possible’ (Aronin & Singleton, 2012b: 318). Language affordances include social language affordances and individual language affordances. Social language affordances are those offered by a particular community at a specific time and with relation to licensing the use and acquisition of languages. Individual language affordances are those through the realisation of which an individual can interact with and make use of languages. Capitalising on individual language affordances stems from the fact that human beings, like other organisms, assess environmental stimuli to decide if they will enhance or hinder the fulfilling of their needs and goals. This involves assessing the emotional relevance of stimuli on the basis of their novelty, pleasantness and relevance to needs, and ability to cope. As pointed out by Schumann (1997) in his discussion of stimulus appraisal, language users and L2 learners judge phenomena within language and across languages and they intuitively decide how those phenomena may be relevant to enhance communication. The previous chapters have provided vast data on how crosslinguistic similarity (i.e. interlingual identifications) can enhance communication and language learning, so they may be assessed as relevant: ‘with specific reference to interlingual identifications, we can surmise that stimulus appraisal entails, inter alia, a judgement about [their] communicative utility’ (Odlin, 2006: 30). The existence of ‘interlingual identifications’ such as cognate words in the learner’s L1, L2 and Ln may enhance the process of language learning, when the learner is able to judge their communicative utility. If we accept that linguistic affordances involve the perception of opportunities associated with language learning and use, as well as with the perception of communicative utility, clearly, those who notice the crosslinguistic affordances will be able to utilise them. Noticing is connected with metalinguistic awareness which stems from learning experience while individual language affordances depend on language awareness: ‘the higher the level of language awareness is, the more effectively language-related possibilities are likely to be perceived and capitalised upon’ (Singleton & Aronin, 2007: 85). Consequently, affordances associated with language learning and use will only be available to those learners who are aware of them, i.e. those learners who have a higher awareness of crosslinguistic similarity. Learners’ affordances will depend on ‘their perception of the qualities of a new language and the amount of cross-linguistic and intralinguistic knowledge that they could mobilise when learning this new language’ (Dewaele, 2010: 106). This ties up with the suggestion that there are individual differences between learners in taking advantage of language affordances (Segalowitz, 1997, 2001). Learners may vary with regard to the flexibility and

132

Par t 3: Le x ical Crosslinguist ic Simil ar it y in Use

fluency required to deal with language, and their language performance may also be affected by sensitivity to environmental affordances, that is, the ability to adjust to the changing linguistic and non-linguistic context. Individual learner variation may include: abilities (intelligence, working memory and aptitude), propensities (learning style, motivation, anxiety, personality and willingness to communicate), and learner beliefs, actions and learning strategies (Dörnyei, 2006; Dörnyei & Skehan, 2003; Ellis, 2008). On the other hand, more advanced multilingual language learners seem special in terms of sets of individual learning features. Accomplished multilinguals have higher working memory capacity and higher language aptitude, as compared with ordinary L2 learners with a similar background (Biedron ´& Szczepaniak, 2012). More proficient multilinguals have automatised, fluent knowledge of several languages and enhanced abilities for adjusting to context. As discussed in the previous sections, this increases their metalinguistic awareness and noticing abilities, which have an impact on their abilities for strategic learning. All these abilities result from having more ‘experience’ in language learning and having more opportunities to interact with the environment. In turn, metalinguistic knowledge and awareness of multilinguals can counterbalance psychotypological judgments. As Singleton and Aronin (2007: 85) put it, ‘it is obvious that more potential affordances are at the disposal of multilinguals than of other language users. Self-evidently, multilinguals have larger overall linguistic repertoires than other language users.’ Research focusing on measuring language affordances is scarce, so the concept of enhanced language affordances in multilinguals is more of a theory that tries to explain the phenomena of crosslinguistic consultation and strategic learning in multilinguals. Apart from one study, evidence for the claim that multilinguals have more potential affordances at their disposal is only indirect (and it has been extensively discussed in the previous sections of this chapter). Dewaele (2010) contributed to the affordances approach in language acquisition by investigating the impact of the knowledge of other languages on self-perceived communication competence and communication anxiety in French. His study utilises the concept of cumulative language experience and affordances are operationalised as a cumulative score of typologically related languages. This allowed Dewaele to consider the combined effect of quantity and quality of specific affordances on communicative competence and communicative anxiety in FLA in French L1, L2, L3 and L4. In particular, Dewaele demonstrated the role of affordances, depending on the level of proficiency of the multilingual participant in typologically related, or unrelated languages. On the basis of a large-scale online survey, Dewaele (2010) proved that the knowledge of typologically related languages might be linked to stronger affordances than unrelated languages, but the affordances also depend on the frequency of language use, age of acquisition onset, context where the language is spoken and the number of languages known. In his study,

Reliance on Le x ical CL I as a Strategic Behav iour

133

it turned out that in high or low proficiency learners, the knowledge of other languages does not play a major role, but in learners with intermediate levels of proficiency, affordances add to their communicative competence and lowered levels of speaking anxiety. As concluded, ‘the notion of affordances is particularly useful because it allows the researcher to cast a wider net over the complex computation in an individual’s mind when that person judges how difficult the learning of a new language is going to be’ (Dewaele, 2010: 125). Having acknowledged that cognates, together with other crosslinguistic similarities, may constitute a set of affordances in learning a language, it is possible to say that multilinguals should also be advantaged over bilinguals in noticing such affordances. There should be a connection between individual language resources, language learning and language-using environments, language awareness and positive lexical CLI from L1 and other languages. This hypothesis will be investigated in more detail in the following chapters presenting research on the role of awareness in noticing and using cognate vocabulary by Polish learners of English.

Part 4 Investigating Lexical Crosslinguistic Similarity in Language Learning

11 Introduction: Researching the Awareness of Cognates in Polish Learners of English

Introduction: Aim and Scope of the Research Language learning and multiple language use are among the educational priorities in the European Union. Today, the knowledge of more than one language, or at least English, the European lingua franca, is a social must in many countries. Becoming bilingual, multilingual, or plurilingual is indispensable for taking an active part in education, culture and social life and for finding work (see Chapter 1 and 2). However, the goal of common multilingualism or plurilingualism is also a challenge for language teaching. Previous research shows that for bilinguals, and especially for multilinguals, cognate vocabulary between the languages they know is often a specific ‘trampoline’ for reaching higher levels of language knowledge because can be used as a specific learning strategy (see Chapter 10). As discussed earlier, cognate vocabulary between two languages, no matter whether genetic or non-genetic in origin (see Chapters 3 and 4), can be a fascinating topic of study in relation to language processing (Chapter 6) and acquisition (Chapter 7 and 9). Some language pairs abound in highly-similar cognate words, but in some other languages cognates are more difficult to find and recognise. This is often connected with language typology and the processes of borrowing and adaptation of loanwords (Chapters 3 and 4). If the language learner can recognise the similarity of cognate words with L1 or any other language he/she knows, the vocabulary-learning task becomes much easier (Chapter 5, 7 and 8). Thus, the awareness that cognates exist and the ability to notice them can considerably ease the language-learning burden (Chapters 7, 9 and 10). Based on all the assumptions above, the overall aim of the research presented in the book is to explore the role of cognates in language constellations 137

138

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

with Polish and English, which are typologically distant. The research answers the following basic questions: What measurable word-related and learner-related factors affect the awareness of cognates and CLI as a learning strategy? How can adult English language learners benefit from raising awareness of cognate vocabulary? Can raising awareness of cognates be used with learners younger than adults on their way to plurilingual language competence? The project examines whether Polish learners of English are aware of cognates, which properties of cognates modulate how learners notice them and whether the awareness of cognate vocabulary can help people learn and use English in an effective way. It is assumed that functionally, any formal orthographic similarity coupled with semantic similarity (cognates) may facilitate learning, while formal similarity alone may cause problems to the learner (false friends). Throughout the project I investigated bilinguals at various English proficiency levels and multilinguals with English as L2 and various constellations of L3-Ln languages. The participants of the project were young Polish adults and Polish teenage learners (14 years old). Younger learners were excluded since research has shown that conscious benefiting from cognates may be a matter of cognitive development and the linguistic maturity of the L1 language system (see Chapter 9). The research on Polish learners was inspired in the 1990s by Rusiecki’s (1980) idea of cognate teaching. Rusiecki first compiled a list of 200 core PolishEnglish ‘interlingual synonyms,’ as he called them, and suggested that those core cognate items could be used in teaching English to speakers of Polish in order to expand their active vocabulary and enhance their motivation to learn English. Although this idea has been strongly criticised even quite recently as ‘confusing for learners’ (Man´czak-Wohlfeld, 2006), research shows that activating cognates and raising metalinguistic awareness of their existence brings measureable effects on learners’ vocabulary (see Chapter 7 and 10). The project began at a given place and time: in Central Europe at the end of the 1990s after the fall of Communism, when the demand for learning English was very high (see Chapter 2 and the discussion of motivation to learn English and other languages). In 1997, as an enthusiastic young methodologist, I conducted the first research investigating the possibility of raising the awareness of cognate vocabulary in Polish learners of English (see Chapter 13). Run as an exploratory case study, the experiment was surprisingly successful. However, the main bulk of research presented here was spread over six years, i.e. between 2006 and 2012, when the demand for L2 English was still high, but multilingualism began to be seen as a strong asset. The entire project presented here consists of six different studies, which altogether involved nearly one thousand participants. Stemming from issues discussed in Chapters 8, 9 and 10, the research shows how various language-related and learner-related factors can have an impact on the learner’s conscious and unconscious benefitting from cognate vocabulary between English and Polish.

Researching the Awareness of Cognates in Polish Lear ners of English

139

Methods of Study The six studies examine a number of diverse issues, therefore, they also use different, quantitative and qualitative methodologies. Thus, the project fully utilises a mixed-method approach (Dörnyei, 2007) to studying the complex phenomenon of cognate noticing, recognition and use. Applying the mixed-methods approach is not accidental, but reflects the idea that quantitative and qualitative research can support and inform each other. It is assumed that when methods are well-suited to the research purposes, using both qualitative and quantitative methodology can give a more comprehensive and multidimensional picture of the issue under investigation. Using mixed methods, or the triangulation of research (Cohen et al., 2007; Dörnyei, 2007) allows for investigating complex phenomenon without going to the extremes of one, or another, methodology and presenting a fragmented picture as a result. Quantitative research (QUAN) is systematic, rigorous and involves precise measurement. For instance, in the case of language learning, results generalisable to other contexts allow us to state whether a particular experimental method is suitable for learners and whether the majority of learners at a given proficiency level tend to behave in a particular way, or use particular learning strategies. However, while quantitative, variable-driven analyses used in psycholinguistic and SLA studies provide us with a wide, generalisable view of the phenomenon, the statistical significance of results may not always help to understand the deeper reasons why a particular result has been obtained. Variable-driven analyses do not tell us much about the motivations of particular learners, and do not allow us to investigate outliers. On the other hand, qualitative research (QUAL) is more flexible, and is open to the details that may emerge during the process of investigation. Qualitative methods, explanatory in nature, are therefore, useful in approaching complex situations from various angles (Dörnyei, 2007). The case of making use of cognate vocabulary is definitely a complex situation when considering a multitude of language-related and learnerrelated factors at play. Thus, qualitative case studies combined with quantitative research let us use a more fine-grained approach to the problem and see real students, not just their scores.

Structure of the Research Project The conceptual structure of the project relies on three major areas of investigation: learner awareness of cognates, using cognates as a vocabularylearning strategy and using cognates as a motivational strategy in language learning, as summarised in Table 11.1. The six studies examining the three major areas employ various quantitative and qualitative methodologies: questionnaires, word translation, two

English language learners (n = 150) of varied L2 proficiency (intermediate, advanced) and varied degree of L3 knowledge (pre-intermediate vs. intermediate and above) English language learners (n = 8) total beginners

Study 2 (QUAN): pen and paper translation task involving cognates, false friends and ordinary words

Using cognates as a vocabulary learning strategy

AREA 2

Study 3 (QUAL): case study, longitudinal, classroom quasi-experiment

Learner-related factors in cognate awareness: – psychotypological distance – proficiency in L2 – multilingualism Word-related factors in cognate awareness: – orthographic similarity – word frequency Learner-related factors: – proficiency in L2 – multilingualism Word-related factors – orthographic similarity of cognate words Learner-related factors: – role of very low proficiency – ability to rely on cognateness when reading – use of cognates as a compensation strategy in learner output Teaching objectives: – leaner training in cognates as a strategy

English language learners (n = 512) of varied L2 proficiency (elementary, intermediate, advanced, accomplished multilinguals)

Study 1 (QUAN and QUAL): questionnaire survey on cognate awareness

AREA 1

Learner awareness of cognates

Factors investigated

Participants

Study methodology

Area of study

Table 11.1 The design of the project

140 Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Teenage multilingual learners of English, (n = 16) proficiency in L2 (upperintermediate), varied proficiency in L3

Teenage learners of English (n = 111) proficiency in L2 (preintermediate or upper-intermediate)

Study 6 (QUAN): short-term classroom quasiexperiment

Phase 1: Accomplished multilinguals (n = 83), i.e. learners of English as L2 (advanced) and learners of L3 (advanced) Phase 2: Advanced bilingual learners of English (n = 84)

Study 5 (QUAL): short-term classroom quasiexperiment, focus group

Notes: QUAL – qualitative, QUAN – quantitative methods.

Awareness of cognates as a motivational strategy

AREA 3

Part 2: longitudinal classroom quasiexperiment

Study 4 (QUAN and QUAL): Part 1: vocabulary learning strategies questionnaire

Learner-related factors: – age – level of bilingualism Teaching objectives: – training leaners in cognates as a strategy – gathering opinions on the method

Learner-related factors: – age – level of multilingualism Teaching objectives: – training leaners in cognates as a strategy – gathering opinions on the method

Learner-related factors: – vocabulary learning strategies employed by successful language learners Role of strategy training in: – noticing cognate vocabulary when reading change of vocabulary – use of cognates in learner output (writing, speaking) Teaching objectives: – learner training in cognates as a strategy

Researching the Awareness of Cognates in Polish Lear ners of English 141

142

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

longitudinal classroom experiments and two classroom-based case studies. Since the project was carried out over several years, some of the research was emergent in its nature: some unexpected results pointed to the possibility of studying the phenomenon of crosslinguistic lexical similarity from perspectives different than planned. Several findings from the studies presented in the following chapters have been partly published (Otwinowska-Kasztelanic, 2001, 2009, 2010, 2011a, 2011b; Otwinowska, 2012). However, I believe that presenting the full results in the new, more comprehensive form will provide readers with a more complete picture. Moreover, investigating the topic from such varied perspectives and with the use of both quantitative and qualitative methodologies will help to elucidate the complex mechanism of how cognateness ‘works’ for learners of a typologically unrelated language, and what role it might play depending on the age, proficiency and multilinguality of the learners.

12 Investigating the Awareness of Cognate Vocabulary: Polish Adult Learners of English

Introduction The overall aim of the research on Polish students of English presented in this chapter is to estimate how various word-related factors (see Chapter 8) and learner-related factors (see Chapter 9 and 10) may modify the awareness of cognate vocabulary. For the purpose of the research, cognate vocabulary is defined as words in English and Polish which have descended from a common parent word, were borrowed from English into Polish, or were borrowed independently by the two languages (see Chapter 3 and Chapter 4). As for learner-related factors, both studies investigate the role of proficiency in L2 and multilingualism (proficiency in L3/Ln) in cognate awareness. Study 1 additionally taps on learner beliefs concerning crosslinguistic similarity, psychotypological distance and affordances noticed and used by learners at the different L2 proficiency levels. The word-related factors in cognate awareness that are studied here involve the role of the orthographic similarity of English-Polish cognates, word frequency and word class in the ability to recall cognate words and to translate cognate words, false friends and ordinary words. Both studies are large-scale, quantitative investigations carried out in Warsaw, Poland. Study 1 presents results of a questionnaire carried out on young adults: students and learners of a private language school. Study 2 was a pen-and-paper translation task given to university students. The results of Study 1 were partly presented in OtwinowskaKasztelanic (2011a) and the results of Study 2 in Otwinowska and Szewczyk (2013, unpublished). 143

144

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Study 1: Focus on Language Learners. Cognate Awareness Questionnaire Study 1: Research design Aim, questions and hypotheses Study 1 focuses on the language learners and learner-related factors in cognate recognition and use. It assumes that language learners have certain potential affordances at their disposal and that learning languages successfully depends on the learner’s capacity to perceive and utilise the linguistic affordances embedded in the studying environment. For European languages, even as typologically distant as Polish and English, the presence of numerous cognate words may offer a set of affordances to the language learner, if he/ she is aware of their existence (see Chapter 10). However, as previous research shows, the awareness of cognates might depend on language typology and psychotypology, the language level of the learner and whether the learner is bilingual or multilingual. Study 1 on Polish learners of English took place in Warsaw, Poland in the years 2006–2008. Its primary aim was to determine the level of awareness of cognates depending on the students’ language learning experience. The awareness of cognates was operationalised as the ability to notice the role of crosslinguistic lexical similarity in learning English, knowledge concerning how many cognate words there are and the ability to enumerate some Polish-English cognates. The following research questions were posed: RQ1. Does the perception of crosslinguistic similarity and awareness of Polish–English cognate vocabulary depend on the L2 proficiency of the learners (elementary, intermediate, advanced)? RQ2. Does multilingualism enhance the perception of crosslinguistic similarity and the awareness of cognates? RQ3. Does the cumulative language learning experience influence the awareness of cognate vocabulary and the range of affordances available to learners? Here, it was assumed that the affordances available to the learners depend on their perception of the psychotypological distance between English and Polish, their level of English, and the number of languages they know. It was hypothesised that there is a positive relationship between the level of English and the awareness of cognates. It was also hypothesised that there is a positive relationship between multilingualism the presumed awareness of cognates. Additionally the following questions were posed in order to qualitatively analyse the student input to the questionnaire:

Invest igat ing the Awareness of Cognate Vocabul ar y

145

RQ4. Do the cognates enumerated by students at each level of language learning experience (L2 proficiency, multilingualism) differ in terms of word categories, types and tokens? RQ5. Do the types of cognates enumerated by students at each level of language learning experience differ in terms of word frequency? RQ6. Do the types of cognates enumerated by students at each level of language learning experience differ in terms of orthographic similarity to their Polish equivalents?

Participants Altogether 512 English language learners took part in the study. The respondents to the questionnaire were bilingual Polish learners of English, and the group of multilingual students proficient in at least three European languages. The number of participants in the four proficiency groups is graphically presented in Figure 12.1. The bilingual students of English displayed 3 different levels of English language proficiency. The group from now on called Bilingual Elementary (level A1/A2 CEFR according to the CEFR-compatible diagnostic tests) consisted of 95 beginners from a renowned private language school in Warsaw, all aged between 19 and 35. The group from now on called Bilingual Intermediate (level B1/B2 CEFR, assessed by CEFR-compatible university entrance exams) included 134 students from the University of Warsaw and a college of information technology, all aged between 19 and 24. The last bilingual group, from now on called Bilingual Advanced, comprised 200 firstyear students of the Institute of English Studies, at the University of Warsaw (level C1/C2, assessed by CEFR-compatible university entrance exams), aged 19 to 22. At the time when the study took place, teaching third languages was not obligatory in Polish schools, and most of the respondents from the bilingual group did not claim to know any language other than English, or assessed their proficiency in L3 as beginning/elementary. As a result, their L3

Figure 12.1 Number of participants depending on their level of L2 English

146

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

knowledge can be disregarded in this respect due to the L3 proficiency threshold (as discussed in Chapter 8). The group of multilingual students, from now on called Multilingual Advanced, included 83 students of the Institute of English Studies, University of Warsaw, who were aged 19 to 25 and were at least trilingual. Their native language was Polish, they were all advanced learners of English as L2 (C1/C2) and studied L3 languages (Spanish, French, Portuguese, Italian, Russian or German) at the advanced level (C1/ C2). Their L2 and L3 were assessed by CEFR-compatible university entrance exams. Moreover, most multilinguals learned various L4 or L5 languages (A1 to B2 in Spanish, Portuguese, Catalan, German, French, Italian). In their background questionnaires, none of them indicated knowing Latin or Greek. Since the Multilingual Advanced learners followed the same courses (practical English linguistics and literature) as the Bilingual Advanced group, the level of their L2 and their background was absolutely comparable with the Bilingual Advanced group.

Method and instrument The method of the research involved responding to a short questionnaire in Polish (see Appendix 2, p. 247) within a time limit of 10 minutes. The questionnaire consisted of seven items and a part gathering background information. The first question dealt with the psychotypological distance and the next three with learner perception of Polish-English crosslinguistic similarity. The fifth question asked the respondents to enumerate any five words ‘whose form and meaning are similar in English and Polish.’ So, the wording of the questionnaire items left no doubt about what cognates were. The last two questions concerned the awareness of the number of cognates between the two languages and the student’s ability to judge his/her own knowledge of cognates. Here, the respondents were asked to assess the number of cognates that they thought were available and that they believed to know. All the bilingual students filled in their questionnaires during their classes. While completing them, the respondents were not allowed to consult each other. The Multilingual Advanced students answered the questionnaires in class (52), or emailed their answers (31).

Method of analysis The answers to RQ1 and RQ2 were obtained from descriptive statistics, and Pearson’s Chi-square tests which were used to indicate differences between the choices in the groups of respondents. The three basic independent variables were the learners’ perception of the psychotypological distance between English and Polish, their level of English (L2 Proficiency: Bilingual Elementary, n = 95; Bilingual Intermediate, n = 134; Bilingual Advanced, n = 200), and the fact that they are Bilingual Advanced (n = 200), or Multilingual Advanced (n = 83). In order to obtain answers to RQ3, a new independent variable called Language Experience was created to rank the

Invest igat ing the Awareness of Cognate Vocabul ar y

147

four above groups in terms of their cumulative language learning experience. A set of Pearson’s Chi-square tests and an ANOVA were run to compare the distribution of the answers between the Language Experience groups and a Spearman’s rho correlation was used to assess the range of affordances available to the Language Experience groups. The respondents’ answers to Point 5 were analysed qualitatively in order to provide answers to RQ4, RQ5 and RQ6. This separate set of analyses was run to examine the role of Language Experience in the types and tokens of the cognate vocabulary enumerated. Here, type-token ratios of the cognates enumerated by each Language Experience group were calculated and compared. Additionally, 2 × 2 Chi square tests were used to indicate differences between the groups of respondents. The differences between the groups were also analysed in terms of word frequencies and orthographic similarity of the cognates, as indexed by Levenshtein distances (NLD, see Chapter 4). An ANOVA and the Kruskal-Wallis tests (non-parametric ANOVA) were run to analyse the differences in the frequency and NLD distributions of the cognates enumerated by the respondents.

Study 1: Quantitative results RQ1: Does the perception of crosslinguistic similarity and awareness of Polish-English cognate vocabulary depend on the L2 proficiency of the learners (elementary, intermediate, advanced)? In answering RQ1 we focus on Points 1, 2, 3 and 4 of the questionnaire (see Appendix 2, p. 247). First, let us estimate if the perception of psychotypological distance and similarity relations between the languages depends on the Proficiency of the learners. Psychotypological factors may play a major role in noticing crosslinguistic similarities and may also have manifold consequences for the learners’ ability to use the possible affordances offered by cognate vocabulary. In order to assess the psychotypological distance, the participants were asked: Are Polish and English closely related (Point 1 of the questionnaire)? It turned out that the vast majority of students (94–97% in each group), with no regard to their knowledge of English and the number of languages known, did not perceive the two languages as closely related. Since the answers concerning the perception of the typological distance did not differentiate the responses, the general factor of the psychotypological distance was excluded from further analysis. Although the psychotypological factors act against all proficiency groups, as discussed in Chapter 8, the detailed perception of crosslinguistic similarity may still depend on the language learning experience of the learner. Therefore, the consecutive analyses concerned the role of English proficiency in perceiving crosslinguistic similarity through analysing Points 2, 3 and 4 of the questionnaire. Conversely to Point 1, which assessed the

148

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

general perception of the psychotypological distance, the next three points of the questionnaire (Point 1, 2 and 3) focused on a more detailed perception of crosslinguistic similarity across the language systems of English and Polish. Here, we wanted to find out whether Proficiency in English can modify perceiving crosslinguistic similarity across the language systems and differentiate between the three Proficiency groups of respondents. As for Point 2: Which system of English is the most similar to Polish: grammar, vocabulary or phonetics? the participants had to choose one of the three options given. Over one third of Elementary (38.5%) and Intermediate (36%) students, as well as (45%) of the Advanced students pointed to lexical crosslinguistic similarity. However, the Pearson’s Chi-square run to compare the distribution of the answers across the Proficiency groups revealed no significant differences. The students’ responses are summarised in Figure 12.2. Concerning the next question question: Which systems of English are easiest for a Pole to master: grammar, vocabulary, phonology (Point 3)? around half the respondents in each group opted for vocabulary. The Pearson’s Chi-square test showed a significant difference between the three Bilingual Proficiency groups, χ2(6) = 20.619, p = 0.002. The differences between the Bilingual Elementary and the Bilingual Intermediate groups was not significant, but a significant difference was noted in the 2 × 2 Chi-square test between the Bilingual Intermediate and the Bilingual Advanced group, χ2(2) = 16.402, p = 0.001. Concerning the fourth question (Point 4): When is it easier to understand a written text: when you know grammar well/when you know vocabulary well? the vast majority of respondents (78–93%) pointed to the role of vocabulary in understanding texts, as presented in Figure 12.3. The Pearson’s Chi-square test showed a significant difference between the three Proficiency groups, χ2(4) = 16.592, p = 0.002, and the 2 × 2 Chi-square tests revealed significant differences between the Bilingual Elementary group and the Bilingual Intermediate group χ2(2) = 17.433, p = 0.000, as well as between the

Figure 12.2 Beliefs concerning the similarity of English and Polish systems

Invest igat ing the Awareness of Cognate Vocabul ar y

149

Figure 12.3 Beliefs concerning when it is easiest to understand an English text

Intermediate and the Advanced group, χ2(2) = 10.095, p = 0.006. The Bilingual Intermediate students valued grammar significantly less than the other groups. Summing up, the results for Points 1–4 of the questionnaire provide a positive answer to RQ1, concerning the impact of proficiency on the perception of crosslinguistic similarity. Although at the most general level all the proficiency groups perceive Polish and English as distant (Point 1 of the questionnaire), when we focus on the perception of the system similarity, there are differences between the three English proficiency groups of respondents. A vast majority of students believe that English vocabulary is relatively easy to master and that vocabulary is necessary to understand texts. However, the Advanced Bilingual students differ from the two lower proficiency groups since they believe more strongly that vocabulary in English is the most similar system to Polish. Concluding, the assessment of system similarity between Polish and English as well as the perception of similarity between the lexicons of Polish and English changes with the L2 proficiency, the possible threshold being between the Intermediate and the Advanced level (RQ1). RQ2: Does multilingualism enhance the perception of crosslinguistic similarity and the awareness of cognates? Let us now focus on the question whether being multilingual affects perceiving crosslinguistic similarity. As for Point 2 Which system of English is the most similar to Polish: grammar, vocabulary or phonetics? as many as 70% of the multilinguals decided that vocabulary was the most similar system of all (see Figure 12.2). The 2 × 2 Chi-square tests run to compare the groups showed a significant difference between the Advanced Bilingual and the Advanced Multilingual students with respect to this question, χ2(3) = 24.110,

150

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

p = 0.000. Advanced Multilinguals pointed to vocabulary significantly more often than Advanced Bilinguals. Since the Advanced Bilingual and the Advanced Multilingual groups did not differ in terms of their L2 proficiency and background, this suggests that multilingualism is an important factor in perceiving the similarity across the language systems. On the other hand, in the next two questionnaire items (Point 3 and 4) concerning the perception of difficulty in the mastery of the language systems, the differences in the distribution of the answers did not reveal significant differences between the Advanced Bilinguals and Advanced Multilinguals (see Figure 12.3). Concluding, multilingualism seems to enhance the perception of crosslinguistic similarity, hence the differences between the Advanced Bilinguals and the Advanced Multilingual groups of the same L2 proficiency and background (RQ2). The Advanced Multilingual respondents perceive crosslinguistic similarity in the area of vocabulary as more pronounced than the Advanced Bilingual respondents. However, they do not differ from the Advanced Bilinguals in their perceptions of difficulty in the mastery of the language systems. The issue of how English proficiency and the bilingualism/multilingualism factor affects noticing correspondences between the two lexicons, will be explored in more detail when discussing the answers to RQ3. RQ3. Does the cumulative language learning experience influence the awareness of cognate vocabulary and the range of affordances available to learners? Points 5-7 of the questionnaire helped to assess the awareness of cognates in relation to their language learning experience. It was assumed that the awareness of crosslinguistic lexical similarities, and hence, the range of affordances available to the learners, depends on the L2 proficiency, and the bilingualism or multilingualism of the respondents. In analysing the previous questionnaire points we could see that the Advanced Multilinguals were significantly more aware of the role of crosslinguistic lexical similarity in those questions which were sensitive to proficiency. Thus for the purpose of the consecutive analyses, a new one-dimension rank independent variable called Language Experience was created, which embraced both proficiency and bilingualism/multilingualism of the participants. The new variable ranked the experience with learning English as L2 and other languages beyond L2 in the following way: 1. Bilingual Elementary, 2. Bilingual Intermediate, 3. Bilingual Advanced, 4. Multilingual Advanced. The first step of the analysis involved assessing the impact of Language Experience on the type of cognates enumerated. In Point 5 of the questionnaire, the respondents were asked to enumerate any five words that are similar in form and meaning between English and Polish. Their answers were ranked with respect to the simplicity of the words enumerated: 1. simple,

Invest igat ing the Awareness of Cognate Vocabul ar y

151

Low-NLD borrowings bearing high formal similarity to Polish (e.g. radio, zebra, hamburger) and 2. more sophisticated words of Latin and Greek origin and higher NLD (e.g. intention, transcendental). It was assumed that enumerating those more sophisticated cognates required higher awareness of crosslinguistic similarity because such cognates are less similar to their Polish equivalents (see Chapter 4 for the NLD measure) and enumerating them also requires more metalinguistic awareness. For each participant group the percentages of simple and advanced cognates were calculated. The results are presented in Table 12.1 and Figure 12.4. It can be clearly seen from both Table 12.1 and Figure 12.4 that the percentage of simple cognates decreased with the learner L2 proficiency, while the percentage of the more advanced cognates increased steadily. The Pearson’s Chi-square test showed a significant difference between the four Language Experience groups with respect to the two types of cognates enumerated, as well as the lack of answer, χ2(6) = 75.888, p = 0.000. Also all the differences noted in the pairwise comparisons between the Language Table 12.1 The percentage of students who enumerated the different types of cognates Type of cognate

Bilingual Elementary

Bilingual Intermediate

Bilingual Advanced

Multilingual Advanced

simple cognates advanced cognates No answer

68% 8,50% 23%

65% 30% 5%

46.5% 35% 18.5%

35% 61.5% 3.5%

Figure 12.4 The percentage of students who enumerated the different types of cognates

152

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Experience were significant: for the Bilingual Elementary and Bilingual Intermediate groups, χ2(2) = 26.400, p = 0.000, for the Bilingual Intermediate and Bilingual Advanced groups, χ2(2) = 16.436, p = 0.000 and for the Advanced Bilingual and Advanced Multilingual groups, χ2(2) = 20.609, p = 0.000. Next, the two-tailed Spearman’s rho was used to correlate the participants’ Language Experience (the independent variable) with the types of cognates enumerated: simple and advanced (dependent variable). There was a significant positive correlation (albeit a modest one) between the Language Experience of participants and the types of cognates enumerated in this question, r = 0.279, p < 0.01. This means that the more Language Experience the participant has, the more advanced cognates he/she tends to enumerate. Let us now pass on to Points 6 and 7 of the questionnaire where the participants were asked to estimate the number of cognates. Obviously, there is no ‘correct’ answer to these questions, and the answers could not indicate whether the participants actually knew and used these cognates. However, Points 6 and 7 were supposed to indicate the level of the student’s metalinguistic awareness. It was assumed that the more he/she is aware of the number of cognates, the more likely he/she will be to benefit from them in language learning and use. In other words, the answers would indicate noticing the possible affordances offered by the cognate vocabulary. Point 6 of the questionnaire ‘How many words are there whose form and meaning are similar in English and Polish?’ estimated the respondents’ awareness of the number of English-Polish cognates available. The participants had to estimate the number of Polish-English cognates available within the following ranges: 20–50, 50–150, 150–500, 500–1000 and 1000–5000. The answers in the five categories were ranked 1–5, 5 being the highest. In order to compare the distribution of the answers between the Language Experience groups, a one-way ANOVA was calculated. It indicated significant differences between all the Language Experience groups, F(3508) = 30.092, p = 0.000. Additional t-tests run for pairwise comparisons indicated significant differences between the Bilingual Elementary and Bilingual Intermediate groups, t(227) = −5.503, p = 0.000, as well as between the Bilingual Advanced and Multilingual Advanced groups, t(281) = −4.761, p = 0.000. The t-test was not significant for the comparison between the Bilingual Intermediate and the Bilingual Advanced students. Next, the two-tailed Spearman’s rho was used to correlate the participants’ Language Experience (the independent variable) with the level of awareness indicated by the estimated number of cognates in Point 6 of the questionnaire (the dependent variable). The results showed a significant positive correlation between the Language Experience of the participants and the number of cognates estimated in this question, r = 0.359, p < 0.01. The more Language Experience the participant possessed, the larger number cognates he/she estimated to be available between English and Polish.

Invest igat ing the Awareness of Cognate Vocabul ar y

153

For the purpose of graphic presentation and description of the answers, the respondents’ answers were further reclassified and labelled: the votes for below 150 – Low Awareness, between 150 and 500 – Medium Awareness, and above 500 – High Awareness of crosslinguistic lexical similarity. As presented in Figure 12.5, over half the Bilingual Elementary respondents displayed Low Awareness of the existing cognates and nearly 20% of them either could not answer or did not want to answer this question at all. About one quarter of all the respondents showed Medium Awareness, while over half the Advanced Multilinguals proved to possess High Awareness of the cognates available. In Point 7 of the questionnaire, ‘How many cognates do YOU know?’, the participants again had to estimate the number of Polish-English cognates they believed to know. They could choose one of the following ranges: circa 10, circa 50, circa 100, circa 500, and more than 1000. The answers in the five categories were again ranked 1–5, 5 being the highest. In order to compare the distribution of the answers between the Language Experience groups, a Pearson’s Chi-square test was calculated, which indicated significant differences between all the groups, χ2(15) = 169.737, p = 0.000. The 2 × 2 Chisquare tests run for pairwise comparisons indicated significant differences between the Bilingual Elementary and Bilingual Intermediate groups, χ2(5) = 24.28, p = 0.000, between the Bilingual Intermediate and Bilingual Advanced students, χ2(5) = 26.589, p = 0.000, as well as the Bilingual Advanced and Multilingual Advanced groups, χ2(5) = 55.852, p = 0.000. Again, the two-tailed Spearman’s rho was used to correlate the participants’ Language Experience with the level of awareness indicated by the estimated number of cognates in Point 7 of the questionnaire. The results showed a significant positive correlation between the Language Experience available to participants and the number of cognates estimated in this question, r = 0.433, p < 0.01. The higher was the participant’s rank in the Language Experience, the larger number of Polish-English cognates he/she believed to know.

Figure 12.5 The respondents’ awareness of the number of English-Polish cognates

154

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Figure 12.6 The respondents’ awareness of the number of cognates in their lexicon

For the purpose of graphic presentation of the results in Figure 12.6 the answers were re-categorised and labelled: the votes for circa 10 or circa 50 – Low Awareness, the votes for circa 100 – Medium Awareness, and the votes for circa 500 or more than 1000 – High Awareness. As the figures present, about 75% of the Bilingual Elementary students, and over 80% of the Bilingual Intermediate students, as well as nearly 60% of the Bilingual Advanced students examined displayed Low Awareness. On the other hand, nearly 70% of Advanced Multilinguals showed Medium or High Awareness. In the last step of the analysis, the two-tailed Spearman’s rho was used to correlate the participants’ answers to Point 5 of the questionnaire (types of cognates enumerated) and Points 6 and 7 (the awareness of cognates). The answers to Point 5 of the questionnaire positively correlated with the answers to Point 6, r = 0.413, p < 0.01 and Point 7, r = 0.402, p < 0.01. The answers to Points 6 and 7 were also strongly correlated, r = 0.703, p < 0.01. The results clearly show that the participant’s ability to enumerate more advanced cognates grows together with awareness of the number cognates between English and Polish and the awareness of the cognates known. The above analyses allow us to answer RQ3, which asks whether the cumulative language learning experience influences the awareness of cognate vocabulary and the range of affordances available to learners. In accordance with the theoretical discussion in Chapter 10, it was assumed that the awareness of crosslinguistic lexical similarities, and hence, the range of affordances available to the learners, depends on their Language Experience (L2 proficiency, and bilingualism or multilingualism). While it is quite clear that the awareness of cognates rose steadily together with the growing L2 proficiency of the participants, from Elementary to Advanced, even the Advanced Bilingual students were not fully aware of the potential of the cognate vocabulary which they obviously knew and used, as word listing had proved. On the other hand, the Advanced Multilinguals, whose L2 level of English

Invest igat ing the Awareness of Cognate Vocabul ar y

155

and language background were fully comparable as those of the Advanced Bilingual students, revealed to have a much higher awareness of crosslinguistic similarities. The answers multilinguals gave to Points 2, 5, 6 and 7 of the questionnaire clearly stood out from the answers of the remaining groups of respondents. In accordance with the assumption that the range of affordances available to the learner is indicated by their awareness of crosslinguistic influences, we can claim that the range of affordances rose with the Language Experience. Thus, the Advanced Multilingual students had a widest range of affordances of all the students.

Study 1: A qualitative analysis of enumerated cognates RQ4. Do the cognates enumerated by students at each level of language learning experience differ in terms of word categories and type-token ratios? The qualitative analysis of cognates enumerated by language learners is explanatory in its nature and it aims at shedding more light on cognate knowledge in relation to the L2 proficiency level. For the purpose of the qualitative analysis let us come back to Question 5 of the questionnaire where learners were asked to list any five cognates that they could think of. Altogether, out of 512 respondents, 69 students left this questionnaire item blank, as summarised in Table 12.2. This either means that they could not cope with the task (probable in the case of the Bilingual Elementary students) or were not motivated to complete it (probable in the case of the Bilingual Advanced students). Interestingly, all Bilingual Elementary students provided their answers to Question 5 of the questionnaire in English, which means they enumerated only the English words from the Polish-English cognate pairs. Possibly, the Elementary learners enumerated only those words which they had remembered from their English language lessons. In the other groups, some students provided Polish-language equivalents of the cognate pair and some Englishlanguage equivalents, which again indicates higher awareness of the PolishEnglish cognate pairs. The words were analysed by the number of types Table 12.2 Total number of questionnaires analysed qualitatively

No answer Questionnaires answered

Bilingual Elementary n = 95

Bilingual Intermediate n = 134

Bilingual Advanced n = 200

Multilingual Advanced n = 83

22 73

7 127

37 163

3 80

156

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

(word-forms) of cognates, as well as the number of tokens (occurrences of word-forms) in the following manner: (1) All words enumerated by each group of respondents in English and in Polish were gathered into one single list per group. This list of individual words was treated as the list of tokens for each group (Bilingual Elementary, Bilingual Intermediate, Bilingual Advanced, Multilingual Advanced). (2) The tokens were analysed according to word categories (concrete nouns, abstract nouns, adjectives, adverbs). (3) Next, all types were extracted, i.e. a list of unique word forms within the total number of tokens. If both the English and the Polish variants of a cognate were enumerated (e.g. Eng. lamp, Pol. lampa), the case was treated as a single token for the cognate. (4) The type-token ratio (TTR) was calculated as a measure of lexical variation within the words enumerated by each group. The TTR (given in percentage) is the number of types divided by the number of tokens times 100. The measure indicates how rich or ‘lexically varied’ the vocabulary in each list was. Word categories. As for word categories in the sample, a vast majority of the cognates enumerated were nouns (88-93% of the sample), which holds for both types (Figure 12.7) and tokens (Figure 12.8). Next came adjectives,

Figure 12.7 Percentage of word categories in the word tokens enumerated by each group

Invest igat ing the Awareness of Cognate Vocabul ar y

157

Figure 12.8 Percentage of word categories in the word types enumerated by each group

but they constituted up to 10% of the sample, for types and less than 8% for tokens. The category represented least were verbs, which constituted below 3% of the tokens and the types. Although the high percentage of nouns is relatively steady for all proficiency levels, in the sample, there is an interesting decrease in the share of concrete nouns, as compared with abstract nouns. This can be seen in both analyses of word tokens and types. Focusing on the types enumerated, abstract nouns constitute only 12%, in the Bilingual Intermediate and Advanced groups it is 32% and 31%, respectively, while in the Multilingual Advanced group it is 40%. As shown by the Chi square 2 × 2 tests, the difference is significant for the Bilingual Elementary and the Bilingual Intermediate group, χ2(1) = 3.40, p < 0.04. The contrast in the use of concrete/ abstract noun types is not significant for the Bilingual Intermediate – Advanced pair and the Bilingual Advanced – Multilingual Advanced pair. These differences indicate that there is a considerable discrepancy in the share of abstract and concrete nouns between Bilingual Elementary and Bilingual Intermediate students, but it is scarce between Bilingual Intermediate and Bilingual Advanced students. However, the difference is again present between the Advanced and the Multilingual students. It is worth remembering that this last difference does not depend on the L2 proficiency level and background of the groups involved because they are the same. Let us now pass on to those words enumerated by the participants that were excluded from the analysis. As presented in Table 12.3, these were

158

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Table 12.3 Types of words excluded from the count of types and tokens Number of types

Bilingual Elementary

Bilingual Intermediate

Bilingual Advanced

Multilingual Advanced

False friends Partial false friends Recent unassimilated borrowings Semantic overlap, no formal similarity Acronyms Proper nouns and brand names

0 0 2 13

7 1 0 8

5 12 4 6

4 0 0 7

3 2

0 2

0 2

0 2

several false friends and some words semantically related to their Polish equivalents but not formally similar to them. Proper nouns (e.g. Hawaii, China, India, Stephen, Christ) and acronyms (e.g. OK, TV, DVD) were also excluded from the count of tokens and types. Interestingly, there were no false friends in the words enumerated by the Bilingual Elementary group, a small number of false friends in the words listed by Intermediate students (e.g. acceptation, lunatic, metro, offensive) and a small number of partial false friends in the words listed by the Bilingual Advanced students (hymn, impression, pension, rondo) and the Multilingual Advanced students (e.g. agonia, elaborate, testament). The number of false friends and partial false friends was relatively smaller in the Multilingual group (1.7%), and much higher in the Bilingual Intermediate and Advanced groups (4.9% and 5.5%, respectively). The number of words semantically, but not formally, related to their other language equivalents was highest in the Bilingual Elementary group and similar in the remaining groups. These words were always connected with the family (e.g. mother, brother, mum) or learning (e.g. school, lesson, English). This is interesting because it may point to the fact that strong bonds within a semantic field and full conceptual overlap of the words involved may have an impact on the perceived formal similarity of these words. Type-token ratios. Next, the type-token ratios for the words listed were calculated for each of the Language Experience groups. As presented in Figure 12.9 below, the type-token ratios of the words enumerated differed for the levels of Language Experience. The type-token ratios rose steadily with the levels of Language Experience, which illustrates the growing lexical variety in the wordlists by each group. The Bilingual Elementary learners had the lowest, while the Advanced Multilinguals had the highest type-token ratio of all. Again, the difference between the Advanced Multilinguals and the Bilingual Advanced students in this respect cannot be explained solely by their L2 proficiency level. To conclude, the answer to RQ4 is positive. The closer, qualitative analysis of the cognates enumerated by students at each level of Language

Invest igat ing the Awareness of Cognate Vocabul ar y

159

Figure 12.9 Type-token ratios for the cognates enumerated by each group

Experience reveals differences between the groups in the lexical variety of the lists. Also, the word categories listed by each Language Experience group differed. The majority of cognates listed were nouns, next came adjectives, while verbs were least numerous. However, the percentage of abstract nouns and adjectives was lowest in the Bilingual Elementary group and grew with the learners’ proficiency in L2. RQ5. Do the types of cognates enumerated by students at each level of language learning experience differ in terms of word frequency? Let us now focus only on the types of words listed. The types enumerated by each group of students were analysed in terms of word frequency. Since the log frequencies in all the groups had normal distribution, it was possible to use an ANOVA to compare the log-frequency of types enumerated across the four groups of students. The result is significant, F(3792) = 11.079, p = 0.000. However, t-tests only show differences in the frequency distributions between the Bilingual Elementary and the Intermediate groups, t(235) = 2.24, p < 0.05, as well as Intermediate and Advanced, t(486) = 3.26, p < 0.01. The difference in the frequency distribution between the Bilingual Advanced and the Multilingual Advanced group is not significant, which means that both these groups enumerated cognate types of similar frequencies. The histograms for the frequency distributions in each group are presented in Figure 12.10. RQ6. Do the types of cognates enumerated by students at each level of language learning experience differ in terms of orthographic similarity to their Polish equivalents?

160

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Figure 12.10 Histograms for word log frequency of types enumerated by all groups

The orthographic similarity between Polish-English cognates can be indexed by the LD, which counts the minimal number of substitutions, insertions, and deletions to edit one string into another of any length. The Normalised Levenshtein Distance (Otwinowska & Szewczyk, 2013) normalises the LD metric by the length of the two compared words (see Chapter 4 for detailed information on the LD and NLD measures). NLD takes values between 0 and 1, where score 0 means full formal overlap, while score 1 means the highest degree of difference. The histograms presenting the distribution of orthographic similarity of cognate types enumerated by all Language Experience groups are shown in Figure 12.11. As discussed in the section on quantitative results and presented in Figure 12.4, the lower the level of the respondents’ English, the simpler the words that enumerated. The Bilingual Elementary group filled in mainly one- or two-syllable words, which could mostly be found in course books for this level: concrete nouns, recent borrowings into Polish, or words which bore strong formal similarity to their Polish orthographic neighbours. This tendency is also illustrated by the distribution of the NLD for the PolishEnglish cognate pairs in Figure 12.11a. The formal similarity of the vast majority of cognates enumerated by Elementary students was reflected by their low NLD, which indicates a strong resemblance of the words enumerated to their Polish counterparts. The results were in the range of 0–0.1 (e.g. Eng.-Pl. radio – radio, Eng.-Pl. actor – aktor), 0.1–0.2 (e.g. Eng.-Pl. nose – nos; Eng.-Pl. lamp – lampa), or 0.2–0.3 (e.g. Eng.-Pl. telephone – telefon).

Invest igat ing the Awareness of Cognate Vocabul ar y

161

Figure 12.11 Histograms for orthographic similarity (indexed by NLD) of cognate types enumerated by all groups

The words listed by the Bilingual Intermediate students were a mixture of simple items similar to those enumerated by the Elementary students, with some more sophisticated and longer words of Latin origin, or internationalisms. This increase of more varied words is well illustrated by the NLD distribution presented in Figure 12.11b. Numerous respondents in the Bilingual Advanced and the Multilingual Advanced group enumerated quite sophisticated items stemming from Latin or Greek (e.g. hypocrisy, inquisition, spontaneous, transcendental). These words are longer, and the NLD to their Polish equivalents is higher, i.e. 0.4 (Eng.-Pl. spontaneous – spontaniczny), 0.6 (e.g. Eng.-Pl. hypocrisy – hipokryzja), 0.7 (e.g. Eng.-Pl. inquisition – inkwizycja). Noticing the correspondence between the English and Polish equivalents is also likely to rely on higher level of morphological awareness because such words are morphologically complex, which adds to their ‘word difficulty’ (cf. Laufer, 1997). There were no considerable differences between the types of words listed by the Bilingual Advanced and the Multilingual Advanced respondents (Figures 12.11c & 12.11d) and the share of high-NLD cognates among the types enumerated by these groups is similar, although Multilinguals tended to enumerate more cognates of high NLD. Since the NLD of the types enumerated by all groups did not have normal distribution, a non-parametric Kruskal-Wallis test was used to compare the groups. The test revealed that the group factor weakly affected the NLD of the cognates, H(3) = 8.3, p < 0.05. The follow-up comparisons of the mean ranks did not show significant differences between most groups, apart from

162

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

the comparison of the Bilingual Elementary and the Multilingual Advanced (rank difference 84.64, at the critical limit of 80.17, corrected for the number of tests). Concluding, the cognate types enumerated by the Language Experience groups were not very different orthographically from each other, except for the Elementary group, who tended to enumerate the simplest cognates and Multilinguals, whose cognates were most orthographically different from their Polish equivalents.

Study 1: Discussion and conclusion Language learners have certain potential affordances at their disposal (Singleton & Aronin, 2007), but they perceive and utilise the linguistic affordances embedded in the studying environment, if aware of their existence. Such a set of affordances may be offered by cognate words, even if cognates come from languages as distant typologically as Polish and English. Study 1 involved a large-scale questionnaire survey conducted on Polish learners of English in order to examine their awareness of Polish-English cognate vocabulary. On the basis of the literature on the subject (see Chapters 8–10) it was assumed that the awareness of cognates might depend on language typology and psychotypology, the level language proficiency of the learner and his/her bilingualism or multilingualism. As for results, Study 1 showed that a vast majority of all the respondents (94–95%) perceived Polish and English as typologically distant and that most bilingual respondents were not aware of the existence of cognates. Thus, the factor of psychotypology was excluded from further analyses because all language learners, with no regards to their language level and knowledge were equally disadvantaged in this respect. Since psychotypology was excluded from the analysis, we were left with only two factors: L2 proficiency and multilingualism. It was hypothesised that there is a relationship between the level of English and the awareness of cognates, as well as that there is a positive relationship between multilingualism and the awareness of cognates. As discussed in Chapter 10, when crosslinguistic correspondences are not evident, metalinguistic awareness in noticing cognates may be especially needed. It turned out that L2 proficiency of the learners (Bilingual Elementary, Bilingual Intermediate, Bilingual Advanced) had a lesser impact on the perception of crosslinguistic similarity and awareness of PolishEnglish cognate vocabulary than bilingualism/multilingualism of the respondents. The Multilingual Advanced students clearly noticed more correspondences between the two lexicons than the Bilingual Advanced students (RQ1, RQ2). This finding corresponds to claims concerning the enhanced metalinguistic awareness in advanced multilinguals in comparison to bilinguals (De Angelis, 2007; Jessner, 2006; Gabrys´-Barker &

Invest igat ing the Awareness of Cognate Vocabul ar y

163

Otwinowska, 2012). More proficient and more multilingual learners may view the languages in question as distant, but are still able to spot crosslinguistic similarities. However, the impact of the language learning experience on the awareness of crosslinguistic similarity needed further corroboration in the analysis of affordances. It was assumed that the level of the student’s metalinguistic awareness shown in the answers would indicate noticing the possible affordances offered by the cognate vocabulary. For this purpose a variable of Language Experience was established, which ranked all four groups of respondents on the basis of their cumulative language learning experience (from Bilingual Elementary to Advanced Multilingual). The analyses corroborated that the ability to notice cognate relationships depended on the learners’ Language Experience. The higher the level, the more cognates did the learners assume to know. Moreover, when asked to enumerate cognate words, the higher the Language Experience level, the longer and more sophisticated the cognate words could the students list. On the other hand, despite the growth of awareness connected with the L2 level, most bilingual students underestimated the number of cognates in their own lexicon. It was surprising to find that even Advanced Bilingual learners of English tended to be less aware of the cognates (which they obviously knew and used, as word listing had proved) in comparison with the Advanced Multilinguals. These results prove support for the first two research hypotheses. First, there is a positive relationship between the proficiency level of English and the awareness of crosslinguistic similarity, including cognates, which supports the claims that the increased linguistic awareness is an emergent property of bilingual and multilingual proficiency, whose component is crosslinguistic awareness focusing specifically on cross-language comparisons and contrasts (James & Garret, 1991; Jessner, 2006). If the awareness of cognates helps to capitalise upon affordances offered by cognate words, it is possible to claim that the higher the L2 level the more affordances the learners possess thanks to their heightened language awareness. Secondly, the study corroborates that there is a positive relationship between multilingualism and the awareness of Polish-English cognates (RQ2 and RQ3). Since the Multilingual Advanced students examined tended to be the most aware of Polish-English crosslinguistic similarity, we can claim that they are able to use a wider range of affordances offered by cognate vocabulary, as claimed by research into linguistic affordances (Aronin & Singleton, 2012a; Singleton & Aronin, 2007). The qualitative analysis of the word categories and type-token ratios of cognates listed by all groups rendered further support for the claim that Language Experience (both L2 Proficiency and Multilingualism) have an impact on cognate knowledge (RQ4). As for type-token ratios of the words listed by all the groups, they grew steadily with the L2 proficiency level. However, the type-token ratio for the Advanced Multilinguals was still

164

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

higher than that of Advanced Bilinguals. Since their L2 level and background was exactly the same as the Advanced Bilingual group, the difference cannot be explained by their language proficiency, but rather by their experience with language learning, noticing crosslinguistic similarity between various languages, and hence a wider range of language affordances available. An important qualitative finding (RQ4) pertained to the fact that the vast majority of cognates listed were nouns, which was true for both tokens and types. Next came adjectives, while verbs were least numerous. This may result from two facts: either nouns are more prototypical and easier to learn and recall than other word categories (e.g. Milton, 2009; Vidal, 2011), or cognate nouns may be easier to recognise than other word classes, due to a greater similarity to their native equivalents (see Chapter 4 for the similarity comparisons across word classes). Among the words enumerated by the groups there was an increase of abstract nouns with the growing Language Experience (Elementary, Intermediate, Advanced), which suggests a qualitative difference in the lexicon growing with the L2 level. The difference may, obviously, be caused by the English teaching programmes where abstract nouns begin to be introduced relatively late, around the intermediate level. However, it may also be caused by the awareness of cognate vocabulary which increases together with the Language Experience. Abstract words turn out to be more difficult to learn and retrieve than concrete words, which results from the lesser opportunity for their anchoring and retrieval, relative to concrete words (De Groot & Van Hell, 2005; Lemhöfer et al., 2008). Interestingly, this point of analysis ties up with the analysis of word frequency (RQ5) and the orthographic similarity (RQ6) of the cognates enumerated by students at each L2 level. The mean frequency of the types was reversely proportional to the growing L2 proficiency level. The orthographic similarity of the cognates enumerated grew steadily at each L2 level, the Elementary group having enumerated significantly more Low NLD words than others. These results are probably connected with two factors. First, we can assume that the vocabulary knowledge grows across the L2 proficiency groups, and the higher the proficiency, the more infrequent words the students know (Meara & Milton, 2003; Milton, 2009). Second, among the words enumerated by the students at the higher proficiency levels there is a growing share of abstract nouns and adjectives of Latin and Greek origin considered to be a feature of academic language used by proficient language users (Nagy & Townsend, 2012). Both Bilingual Advanced and the Multilingual Advanced students obviously engaged in academic discourse in English more than learners at the lower proficiency levels. Hence their heightened awareness and the ability to enumerate more ‘academic’ words in comparison with the lower level students. However, since the Bilingual Advanced and the Multilingual Advanced groups did not differ with respect to enumerating such words, we can conclude that it is L2 proficiency, rather than multilingualism which affects the cognate types enumerated by the students.

Invest igat ing the Awareness of Cognate Vocabul ar y

165

Overall, Study 1 revealed that cumulative language learning experience, including both L2 proficiency and the multilingualism of the study participants may affect their awareness and their knowledge of cognate vocabulary. That is why it is not only L2 proficiency which is traditionally taken into account in many SLA studies, but also the level of participants’ multilingualism must be accounted for when studying the impact of cognates on vocabulary learning. Both these factors, as well as word-related factors, are going to be further explored in Study 2 presented in the following sections.

Study 2: Focus on Cognate Words. Translation Task Study 2: Research design Aim, questions, hypotheses As discussed in the introduction to the research project, some of the research on cognate vocabulary was emergent in its nature. An important observation from Study 1 concerned the fact that learners at various L2 proficiency levels could enumerate cognate words with different degrees of orthographic similarity, but only advanced bilingual and multilingual students enumerated words whose NLD index was relatively high. This might mean that those cognates which are less similar to their native equivalents can only be noticed and learned later in the language-learning process. The primary focus of Study 2 was to examine whether cognates are easier to learn than other words. Secondly, it aimed to estimate if orthographic similarity between words from two typologically distant language can facilitate or hinder learning cognates. Finally, Study 2 was also meant to corroborate that the cumulative language learning experience has an effect on learning cognates. The following research questions were posed: RQ1. What is the likelihood that a particular word types (cognate, false friend, ordinary word) has been learned by students at the different English proficiency levels? RQ2. Does the degree of form-similarity of cognates modulate their learning by students at the different English proficiency levels? RQ3. Can the level of multilinguality modulate the learning of the word types (cognates, false friends, ordinary words) by students at the different English proficiency levels? For the purpose of the study, we operationalise the knowledge of a particular word (cognate and non-cognate) as the ability to translate this word from L2 English into L1 Polish with high confidence, which is based on Paribakht and Wesche’s (1993, 1997) 5-point Vocabulary Knowledge Scale. Here, we use the middle-point of the scale checking whether the learner can

166

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

translate the word confidently into L1. We hypothesise that cognates are more likely to have been learned, relative to matched ordinary words and false friends. We also hypothesise that both L2 proficiency level and multilinguality level should have an impact on word knowledge.

Participants The participants of the study were 150 Polish learners of English from the University of Warsaw, all between 20 and 25 years of age. The study took place in 2012, during students’ regular classes, but they were free to refuse taking part in it. Within the group tested, there were 95 students attending a general English course at B1 level (CEFR), and 55 first-year students of the Institute of English Studies attending a BA course in English grammar at the B2/B2+ level (CEFR). The level of their English proficiency was chosen for two reasons. First of all, we wanted to test abstract nouns and adjectives which begin to be taught around the B1 level on language courses. Secondly, basing on the results of Study 1, it could be assumed that students at levels lower than B1 might obtain very low scores, because the stimuli would be too difficult. Conversely, for highly advanced students, the stimuli might be too easy and ceiling effects might be obtained. All the participants completed a background questionnaire concerning their prior language-learning experience (see Appendix 3, p. 248). They were also asked to self-evaluate their knowledge of L3 and other languages beyond English (Ln). In fact, the questionnaires revealed that all participants were multilingual and they had studied from one to three languages other than English. This fact cannot be disregarded since other languages known beyond English may have exerted influence on cognate recognition. Chronologically, for some of them English was not their second, but their third language. However, for the sake of the statistical analyses carried out in Study 2 Polish will always be referred to as L1, English as L2, and any other language known best by the participant as L3. Admittedly, this is an oversimplification since the patterns of multilinguality are much more complex, both in terms of chronology of acquisition and in terms of the language levels achieved, as discussed in Chapter 1 (e.g. Cenoz, 2000; Hammarberg, 2010, 2014). However this way of coding the data allows us to systematically focus on the language tested (L2) and, at the same time, account for the participants’ multilinguality (L3).

Materials Study 2 used a list of 120 English words carefully chosen on the basis of several word-type criteria and matched in difficulty with the participants’ L2 level. The stimuli list included 40 cognates (20 nouns e.g. antagonism, loyalty and 20 adjectives e.g. popular, aesthetic); 40 false friends (20 nouns e.g. testament and 20 adjectives e.g. eventual) and 40 ordinary words (20 nouns e.g. embarrassment and 20 adjectives e.g. exhausted) extracted from a B1 level

Invest igat ing the Awareness of Cognate Vocabul ar y

167

course book.1 All the nouns used in the list were abstract nouns, as these turned out to be more difficult in Study 1. The cognates and false friends were controlled for the degree of similarity indexed by NLD between their orthographic L2 and L1 forms (see Study 1 and Chapter 4 for a detailed explanation). The 40 cognates used in the list of stimuli displayed full semantic overlap and various degrees of formal similarity indexed by NLD. Half the cognates were highly similar to their orthographic neighbours in Polish (Low NLD; NLD < 0.22) and half displayed a lower degree of formal similarity to their Polish equivalents (High NLD, NLD > 0.60). The 40 false friends, displayed no semantic overlap with their Polish counterparts, but bore high formal similarity (low NLD) to their orthographic neighbours in Polish. Finally, the 40 English ordinary words were not formally similar but displayed semantic similarity to their Polish equivalents. The choice of stimuli is summarised in Table 12.4. A full list of items together with the frequency and NLD criteria can be found in Appendix 4 (p. 250). All the stimuli had frequency below 30.0 per million (Subtlex-US). This range was determined by the frequency range of the English-Polish false friends available.

Method The participants were given a paper and pencil task which consisted in translating a randomised list of the 120 stimuli described above from L2

Table 12.4 Examples of stimuli (broken down by word category, frequency and by NLD in the case of cognates) Word Category Low NLD Cognates (high degree of formal similarity, semantic similarity)

High NLD Cognates (low degree of formal similarity, semantic similarity)

False Friends (formal similarity, no semantic similarity)

Ordinary Words (semantic similarity, no formal similarity)

Low frequency noun

antagonism (Pl. antagonism)

authenticity (Pl. autentycznos´c´)

manifestation (Pl. manifestacja = Eng. demonstration)

excursion (Pl. wycieczka)

High frequency noun

procedure (Pl. procedura)

loyalty (Pl. lojalnos´c´)

direction (Pl. dyrekcja = Eng. management)

judgement (Pl. osa˛d)

Low frequency adjective

modular (Pl. modularny)

mythical (Pl. mityczny)

consequent (Pl. konsekwentny = Eng. consistent)

amazed (Pl. zdumiony)

High frequency adjective

popular (Pl. popularny)

critical (Pl. krytyczny)

ordinary (Pl. ordynarny = Eng. vulgar)

urgent (Pl. pilny)

168

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

English to L1 Polish. The task lasted between 20 and 25 minutes, but no strict time limit was set. To avoid cheating, four versions of the test were created, each with a randomised order of items. Also, no two students with similar lists sat next to each other. The participants were asked to translate the lists of words into Polish without much thinking and were instructed to write a dash or leave the slot blank, if they could not translate the word. Next to each translation they also had to indicate the confidence level of their translation on a Likert scale (1 – I’m guessing; 2 – I think it might be so; 3 – I’m quite sure; 4 – I know for sure). It was assumed that participants might guess, basing on formal similarities to L1, which would obscure the measurement of L2 vocabulary knowledge. The confidence level was an additional measurement introduced to control for participants’ guessing strategies.

Method of analysis Word-related variables. We controlled two word-related variables. The first was a 3-level Word Type factor, including cognates, false friends and ordinary words, as described in the section on materials. The second was a 2-level cognate similarity factor. As defined in the previous section, it included the Low NLD cognates, similar to their Polish equivalents and the High NLD cognates, more different from their Polish equivalents. Learner-related variables. We also controlled two learner-related variables: of English proficiency level and multilingualism. First, we established that the labels of the English language courses (B1 and B2 levels CEFR) provided too vague and imprecise measure of participants’ L2 level. Thus, we used the correctness in translating the 40 ordinary words as an index of their proficiency. Effectively, the L2 proficiency level was operationalised as participants’ correctness in translating ordinary words. On this basis the participants were divided into four quartiles, denoting the percentage of correctly translated items, from the lowest to the highest English proficiency level, respectively: [0, 0.216] (0.216, 0.432] (0.432, 0.676] (0.676, 0.946]. Using this strategy allowed us to relate the number of known false friends and cognates to the number of known ordinary words. The second learner-related variable taken into consideration was multilingualism. In order to control this variable, we classified participants into two groups based on the declared proficiency in their best other language: Multilinguals-Low with best L3-Ln at A1/A2 level CEFR and MultilingualsHigh with best L3-Ln at B1 level CEFR and above. The reasons for such a division mainly concerned the learners’ linguistic and metalinguistic awareness resulting from L3-Ln proficiency (see Chapter 8 for a discussion of the issues). Coding and classification of data. Each translation of a word was coded as correct or incorrect. All missing responses were treated as incorrect.

Invest igat ing the Awareness of Cognate Vocabul ar y

169

After the initial analyses, data from seven outlier items were removed. Three false friends (adjectives) had an ambiguous translation, and their subordinate meaning allowed a translation different from the expected one. Three ordinary words and one cognate (all adjectives) were morphologically transparent, and therefore many participants translated them correctly despite their very low frequency. Correction for guessing. An important step of the analysis was correcting the data for guessing. Since the English cognates and false friends bore considerable formal similarity to their Polish orthographic neighbours, we assumed that the participants might guess their translation based on form. If the participant guessed, the result for cognates would not represent his/her knowledge of the word, and would be considerably higher than the real score. Moreover, we would not be able to differentiate between the translations that were guessed, and those based on knowledge. That is why in order to account for the possible guessing as a strategy, a correction for guessing was introduced. We decided that guessing would be best shown by false friend translations: in their case, if the participant guessed basing on form, the translation resulted in the incorrect answer. We assumed that when guessing participants would also provide lower confidence ratings than when translating words they knew. In order to establish the mean confidence rating of the guessed and known false friends, we classified all false friend translations into correct translations, guesses, and other errors. Next, for each participant we computed the mean confidence for the correctly translated false friends, and the mean confidence for the guessed false friends. Then we computed a confidence threshold, by averaging the mean confidence for the correct and guessed translations of false friends. From this point on, we assumed that all translations of cognates with confidence lower than threshold were guessed and we recoded them as incorrect. Using the same confidence threshold we also recoded the responses for translations of false friends and ordinary words. Statistical analysis. The statistical analyses involved a set of ANOVAs with items as a random effect, α decision level set at 0.05. The betweensubject variables were the learner-related factors, i.e. the English proficiency level and the level of multilinguality (Low vs High, as defined before). The within-subject variables included the word-related factors, i.e. word types (cognates, false friends and ordinary words), and cognate similarity (different High NLD cognates vs. similar Low NLD cognates). The dependent variable was translation performance on the different word types and on cognate similarity factors. Additional t-tests were used to determine contrasts between the variables. The Greenhouse and Geisser (1959) correction for violation of sphericity was applied to repeated measures analyses with more than one degree of freedom in the numerator. In such cases, the corrected p-value is reported (pGG).

170

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Study 2: Results RQ1. What is the likelihood that a particular word (cognate, false friend, ordinary word) has been learned by students at the different English proficiency levels? When measuring the translation performance, let us first only take into account the participants’ L2 proficiency level and the word types: cognates, false friends and ordinary words. We are interested whether any of these word types is more likely to have been learned earlier in the course of L2 acquisition. The results are graphically presented in Figure 12.12. All analyses presented below were run after the data correction for guessing. To measure the learners’ translation performance on the different word types an ANOVA with items as a random effect was run on all subjects with regards to the three-level word type factor and the English proficiency as the independent variable. The results indicate a significant main effect of word types, F(2101) = 5.6, p < 0.01. As showed by t-tests in pairwise comparisons, cognates are significantly easier to translate than ordinary words and false friends. False friends are also the most difficult to translate. The analysis also revealed a significant interaction between the word types and L2 proficiency: F(6303) = 10.5, pGG < 0.000001. Follow up analyses pointed to the following regularity: the lower the English level, the easier it is for the learner to benefit from cognateness relative to ordinary words. The effect is significant for the learners of the two lowest English proficiency levels. Here, it makes no sense to analyse the effect of proficiency on

Figure 12.12 The translation performance for word types (normal: ordinary words; cognate: cognates; false friend: false friends) by students of different English proficiency levels. Results after the correction for guessing

Invest igat ing the Awareness of Cognate Vocabul ar y

171

ordinary words, because this variable has been defined basing on the percentage of correct translations of ordinary words. Concluding, for the learners at the two lower L2 proficiency levels, the proportion of cognates that they know to ordinary words that they know is larger than at the two higher L2 proficiency levels. On the other hand, for all proficiency levels false friends are always significantly less known than ordinary words and cognates. Thus, false friends are the most difficult to learn in comparison to the remaining word types, while cognates tend to be the easiest, at least at the lower levels of proficiency. RQ2. Does the degree of form-similarity of cognates modulate their learning by students at the different English proficiency levels? Now, let us focus on the cognate similarity factor. In order to compare the participants’ translation performance on the High NLD and Low NLD cognates, an ANOVA was run using the 2-level cognate similarity factor and the English proficiency as the independent variables. The results do not indicate a significant difference in the translation accuracy of the two cognate types, F(1,37) < 1. They do show, however, a significant interaction between cognate similarity and L2 proficiency: F(3111) = 14.43; pGG < 0.00001. The t-tests used for pairwise comparisons indicate that the difference in the translation performance between High and Low NLD cognates is significant only for the lowest English proficiency Level. Learners with the lowest proficiency find High NLD cognates significantly more difficult to translate. The results are presented in Figure 12.13.

Figure 12.13 Translation performance for High and Low NLD cognates by learners at the different L2 proficiency levels. Results after the correction for guessing

172

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

In conclusion, high NLD makes cognate similarity less salient to learners at the lower proficiency levels. This result agrees with the finding from Study 1, which showed that at the lowest L2 level students know only those cognates which are more similar to their L1 equivalents. Possibly, this might be caused by the fact that learners at lower proficiency levels have not yet acquired morphological rules, which allow for noticing regularities governing the orthography-to-phonology correspondence. Many of the High-NLD cognates have much lower distance when one takes into account their phonological, rather than orthographic representations (see Chapter 6). Conversely, at higher proficiency levels, those morphological regularities become more evident, which diminishes the difference in noticing High and Low NLD cognates. RQ3. Does the level of Multilinguality modulate the learning of the word types (cognates, false friends, ordinary words) by students at the different English proficiency levels? In the last set of analyses we wanted to explore the influence of cumulative language learning experience on vocabulary knowledge. In order to check if multilingualism has an additional impact on the translation of the various word types, our subjects were further split into two levels: Multilinguals-Low and Multilinguals-High. It was assumed that the level of L3-Ln knowledge might influence the knowledge of cognates. The results are graphically presented in Figure 12.14. First, an ANOVA was run using the 2-level multilingualism factor and the English proficiency as the independent variables. Here, we will refer only to effects involving the multilingualism factor because the remaining effects

Figure 12.14 The translation performance for word types (normal: ordinary words; cognate: cognates; false friend: false friends) by students of different L2 proficiency levels at two levels of multilingualism (Low vs. High). Results after the correction for guessing

Invest igat ing the Awareness of Cognate Vocabul ar y

173

were already covered by the ANOVA run in the section focused on RQ1. The interaction between multilingualism and proficiency is highly significant, which shows that within the respective proficiency groups MultilingualHigh students are different from Multilingual-Low students, F(3303) = 17.21, pGG < 0.00000001. However, the t-test comparisons point only to the differences between the Multilingual-Low and Multilingual-High students at the highest English proficiency level and at the third English proficiency level. At the highest English proficiency level, Multilinguals-High are significantly better than Multilinguals-Low in translating all word types (p < 0.001). It seems that being an accomplished multilingual, proficient in three languages, significantly boosts vocabulary knowledge in L2. On the other hand, Multilinguals-High are significantly worse than Multilingual-Low at the third English proficiency level (p < 0.00001). When interpreting this result it is worth remembering that the data for the analysis run here had already been corrected for guessing: all results with confidence ratings lower than the threshold had been treated as incorrect (see Analysis, correction for guessing). Effectively, the current analysis was run only for those scores where our participants were confident in their translations. In other words, the result may indicate that Multilinguals-High, who are quite proficient in L2, felt confident enough to guess more. Since our analysis ‘punished’ such guessing, this group scored lower than Multilinguals-Low at the same English proficiency level. In the second step, an ANOVA with items as a random effect was run on all subjects with regards to the 2-level multilingualism factor, and the 3-level word type factor and the English proficiency as the independent variable. The interaction between the factors is significant, F(6303) = 3.92, pGG < 0.01, and it indicates differences in the translation of the different word types (cognates, ordinary words and false friends) by the Multilingual-Low and Multilingual-High students. Let us now compare the translation of cognates and ordinary words separately for Multilinguals-Low and Multilinguals-High, in order to check for which of the English proficiency levels the translation performance differs between the two groups of multilinguals. Pairwise comparisons with t-tests indicate that cognates are translated significantly better than ordinary words. For Multilinguals-Low, the difference is significant at the two lowest English proficiency levels (p < 0.01 and p < 0.05, respectively). For Multilinguals-High, the difference is significant at the same two lowest English proficiency levels (p < 0.000001 and p < 0.05, respectively). However, if we treat the p-value as the measure for the magnitude of the effect, we can clearly see that it is much larger for Multilinguals-High. In their case, at the lowest English proficiency level, cognates are much easier than ordinary words. Probably, when beginning to learn a new language, more proficient multilinguals largely benefit from cognateness in vocabulary learning. This difference levels off at the later stages of language learning. At the same time,

174

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

the performance for ordinary words does not differ significantly across the levels of multilingualism. To conclude, we can postulate a specific profile of those multilingual learners who are more proficient in L3-Ln. At the lowest English proficiency Levels, Multilinguals-High know more cognates than Multilinguals-Low because they are significantly better in providing cognate translations than Multilinguals-Low. This means that those multilinguals who are more advanced in L3-Ln are probably able to notice and take advantage of cognateness earlier in the language learning process.

Study 2: Discussion and Conclusions Due to the mixed results and scarceness of studies measuring cognate learning (see Chapter 7), the aim of Study 2 was threefold. First, it tested if Polish learners of English benefit from learning orthographically similar words (cognates), relative to ordinary words and false friends. Second, it tested whether the similarity of cognates can modulate the learning of cognates. Finally, it checked whether the learner-related factors of cumulative language learning experience, i.e. L2 proficiency and multilinguality Level (Low vs. High) have an impact on learning cognates. It was assumed for the purpose of this study that word knowledge can be operationalised as the ability to translate the word from English into Polish with high confidence (Paribakht & Wesche, 1993, 1997). To this end, the study participants had to translate lists of 120 English adjectives and abstract nouns into Polish stating their confidence level. The lists included 40 cognates, 40 false friends and 40 ordinary words. By measuring participants’ confidence level and performance in their translation of false friends we corrected for the possible guessing strategies in translating words that were orthographically similar to Polish. First and foremost, the results revealed that cognateness indeed does make words ‘special’ for learning (RQ1). When we do not ask participants for awareness-mediated answers, but measure their actual knowledge it turns out that cognates are easiest to learn, as compared to ordinary words, especially at lower proficiency levels. In the study learners at the two lowest proficiency levels were able to translate significantly more cognates than ordinary words, while false friends turned out to be the most difficult words types to translate, with no respect to the English proficiency level. This means that cognates are more likely to have been learned well than other words. If word learning occurs in the way described by the parasitic model (Ecke, 2014; Hall & Ecke, 2003), probably in the case of cognates it is easiest to establish lexemic-level links between L1 and L2 words that will be subsequently backed by the links with conceptual information. In Study 2 we also controlled the similarity of cognates to their Polish orthographic neighbours and how this similarity (indexed by the NLD)

Invest igat ing the Awareness of Cognate Vocabul ar y

175

modulates learning cognates. As discussed in the previous section, using such a measure is justified if assumed that language learners/users perceive the formal similarity of word pairs mainly through their congruent orthography. So, those learners who have achieved a minimal threshold level in their L2 knowledge will perceive certain correspondences between graphemes in both languages (c.f. Chapter 8 and the discussion of proficiency threshold in cognate activation). The results indicate a significant effect of the orthographic similarity of the L2 word to the L1 equivalent (RQ2), which is congruent with studies manipulating orthography in cognate facilitation (e.g. Font, 2001; Dijkstra et al., 2010), recognition (Nagy et al., 1993) and translation (Vanhove, 2014). We also realised that in the case of orthographically similar items the issue of ‘assumed similarity’ may come into play (Ringbom, 2007). If the form of a lexical item looks familiar, the learner tends to assume semantic similarity, although such assumptions may be wrong in the case of false friends (Ringbom & Jarvis, 2009). Therefore, we introduced a ‘correction for guessing’ calculated on the basis of false friend translations where the semantic similarity was merely assumed (see Method of analysis). After our ‘correction for guessing,’ the difference between the Low NLD cognates and the more different High NLD cognates was significant only for the learners at the lowest English proficiency level. These learners found the High NLD cognates more difficult to translate than the Low NLD ones. Probably, at higher proficiency levels, the learners have already established some learned associations (Jarvis, 2009; Ringbom & Jarvis, 2009) between the less similar cognates and their Polish equivalents, so both lexemic and semantic CLI can take place. Consequently, we can say that High NLD makes cognate words less recognisable for learners at lower L2 levels, who treat them as any other words they do not know and do not link them to the existing lexemic representations in their L1 lexicon (Ecke, 2014). Probably, their conscious, explicit metalinguistic knowledge (Bialystok, 1987) of morphology is still too low to notice the correspondence between the orthographic neighbours if they are blurred by the higher NLD. This is an important finding, which indirectly points to the connection between the degrees of cognate similarity and the notion of (psycho)typological distance (Kellerman, 1979, 1986, 1995). In Chapter 8 we discussed the fact that the typological distance between the learner’s L1 and the languages acquired is an important predictor of both visual recognition of cognates (Dijkstra, 2003) and transferability (Andersen, 1983; Odlin, 1989, 2003). As demonstrated by Schepens et al. (2012), if languages are typologically close, the orthographic forms of cognates are identical or highly similar: such languages share more Low NLD cognates. Within a typological family, even beginning language learners can easily notice and learn cognates thanks to the high congruence of their orthographic forms across languages. Faced with predominantly similar words, the learner judges the languages as

176

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

similar, so a small psychotypological distance facilitates learning (Ringbom, 2007). Conversely, for languages which are typologically distant, the number of Low NLD cognates decreases, relative to the number of High NLD cognates (Schepens et al., 2012). As indicated by our results, High NLD cognates are more difficult to notice and learn for less proficient learners. Consequently, the less proficient learner may not notice the less similar cognates between the languages and may judge the languages as considerably different, which will result in a large psychotypological distance ‘blocking’ transferability (De Angelis, 2007). Since psychotypological judgements change as learners gain experience with growing L2 proficiency and metalinguistic awareness (Kellerman, 1979), more experienced learners may still regard languages to be distant, but recognise crosslinguistic similarities (De Angelis, 2007; Jessner, 2006). Finally, Study 2 tested if learning the different word types (cognates, ordinary words and false friends) by the students at the different L2 proficiency levels is modulated by their level of multilinguality. Our results show that at the lowest English proficiency levels, Multilinguals-High are significantly better in providing cognate translations, and hence, know more cognates than Multilinguals-Low (RQ3). We may postulate that cumulative language learning experience helps to take advantage of cognateness earlier in the language learning process. This is probably due to the enhanced metalinguistic awareness resulting from the learning experience, which helps to recognise less similar cognates at lower levels of proficiency (De Angelis, 2007; Jessner, 2006; Aronin & Singleton, 2012a). Although learners tend to search for similarities between the languages at an early stage (cf. Ringbom, 2006, 2007), if their metalinguistic awareness is not high enough, crosslinguistic similarity to L1 may not help (c.f. Falk et al., 2013). Since our study divided learners according to the declared level of their best L3-Ln, it is also possible to establish a tentative threshold beyond which multilinguality facilitates vocabulary learning: Multilinguals High were those with L3-Ln at B1 (CEFR) or above. This finding will mostly pertain to constellations of languages that share lexical similarities. In such a case the lower-intermediate level (A2/B1) in L3 should be enough to for the awareness of cognate vocabulary to play a significant role in learning another language. This finding also corresponds with the results by Berthele (2011), whose analyses demonstrated that multilinguals with high proficiency in at least two of their languages were much better at inferencing meaning of unknown cognates than multilinguals with lower language proficiency: ‘multilinguals with high proficiency in many languages (. . .) can make use of their repertoire in the sense that it provides not only important lexical transfer bases, but also potential rules and regularities, i.e. conceptual knowledge and strategic know-how’ (Berthele, 2011: 200). Berthele’s results also point to a considerable amount of strategic behaviour of

Invest igat ing the Awareness of Cognate Vocabul ar y

177

multilinguals with higher proficiency who tend to use crosslinguistic similarity to enhance language learning. Such strategies will be explored in the following chapter.

Note (1) Words extracted from P. Kerr and C. Jones (2006) Straightforward Intermediate Student’s Book. Macmillan

13 Using Cognates as a Vocabulary Learning Strategy: Polish Adult Learners of English

Introduction Cognate vocabulary, which exists not only in typologically close languages but also across typological boundaries, may be important for the speed of language acquisition because it may quickly enhance the vocabulary knowledge of a language student. However, in order to make use of cognate words the student has to be aware of their existence. The awareness of cognates depends on language typology, which affects the formal features of L1-L2-L3 orthographic neighbours, as well as on learner characteristics, including individual L2 proficiency, learning strategies and multilingualism. As presented in Chapters 9 and 10, multilinguals tend to have greater linguistic and metalinguistic awareness of crosslinguistic similarity than bilinguals. They can also use it in a strategic way, as shown in Chapter 12, Study 2. The research presented in Chapter 13 explores the possibility of using cognate vocabulary as a learning strategy and to raise awareness of cognate vocabulary in emergent bilinguals and in advanced bilingual learners. Again, for the purpose of research, cognate vocabulary is defined as words in English and Polish which have descended from a common parent word, were borrowed from English into Polish, or were borrowed independently by the two languages (see Chapter 3 and Chapter 4). The two studies (Study 3, Study 4 Part 1 and Study 4 Part 2) presented in this chapter revolve around language learners. They focus on two extremes in terms of English language proficiency: adult total beginners and highly advanced bilingual students. Study 3 investigates the role of cognate awareness as a compensation strategy when communicating in English with very 178

Using Cognates as a Vocabul ar y Lear ning Strategy

179

limited language resources, as in the case of total beginners after 72 hours of tuition. Study 4 deals with the other end of the proficiency continuum: advanced bilingual and multilingual learners of English. The two phases of Study 4 are complimentary in the sense that the vocabulary-learning strategies of successful multilingual learners investigated in Part 1 of Study 4 are compared with the strategies of bilingual advanced learners at the beginning and at the end of a longitudinal classroom quasi-experiment in Part 2 of Study 4. However, the aims of Study 4 go beyond training learners to notice cognates as a strategy. The study also measures their performance on productive tasks to find out if the strategy works in practice. All studies are different in their scope and methodology: Study 3, which employs case-study methodology, is a longitudinal quasi-experiment on learners of a private language school in Warsaw. Part 1 of Study 4 takes the form of a survey gathering qualitative and quantitative data from advanced multilingual students. Part 2 of Study 4 is a longitudinal classroom quasiexperiment carried out over several semesters at the University of Warsaw. The results of Study 3 were partly discussed in Otwinowska-Kasztelanic (2001), and those of Study 4 in Otwinowska-Kasztelanic (2010, 2011b).

Study 3: Focus on Beginners of English. A Case Study Study 3: Research design Background, aim and questions The background of the research is important here. Study 3 was conducted in the second half of the 1990s, i.e. just a few years after the fall of Communism. At that moment in time Poland had opened towards Western Europe and it had also become obvious to many people that knowledge of English, the international language, was a necessary condition in doing business and contacting people worldwide. Thus, many adult Poles, who had never learned English before, began to study the language on courses in private language schools. They were highly motivated towards learning, which was comparable to the situation in other post-communist countries (see Dörnyei et al., 2006. The Hungarian situation, widely described in Chapter 2, was very similar to the situation in Poland). Some of those Poles who decided to study English for communicative purposes had never learned English at school due to different teaching programmes in the Communist times. In the main, they had learned Russian at school, but had never achieved higher proficiency levels, which resulted from the widespread reluctance of Poles towards Russian teaching (again, see Chapter 2 for an extensive overview of the situation). That is why the level of Russian in the case of those people was well below the communicativeness threshold. Besides,

180

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

as a language typologically distant from English, close to Polish but using different script, Russian could be neglected in the analysis. Study 3 focuses on the possibility to train cognate recognition and use in adult learners to help them learn faster and with lesser effort. In order to check the effectiveness of the method the study was conducted on total beginners, learning English for the first time. The research focuses on the role of very low L2 proficiency in benefitting from cognate strategy training, and more specifically on the learners’ ability to rely on cognateness when reading and on their use of cognates as a compensation strategy when speaking. Study 3 takes the form of a longitudinal, classroom experiment, but due to the small number of participants it is only a case study. The primary aim of Study 3 was to determine if raising the awareness of cognates may help beginners understand and produce English. The additional aims of the quasi-experiment were to provide beginners of English with a number of words and phrases which are easy to learn, and check if it is possible to practically activate the vocabulary that they know and use in Polish. It was assumed that activating cognate vocabulary, i.e. showing the students that while learning English they can rely on their crosslinguistic intuitions, would help them become more confident and ready to take risks in reading and using English in productive tasks. Thus, the awareness of cognates was operationalised as the sensitivity to English words which resembled their Polish orthographic neighbours. The following research questions were posed: RQ1. Does raising the awareness of cognates help learners write, read and understand written English more effectively? RQ2. Does raising the awareness of cognates help learners produce spoken English more fluently? RQ3. Does training in using cognates as a strategy affect the productive vocabulary of beginners of English?

Participants The participants of the study were 8 Polish-speaking females, between 26 and 40 years of age (M = 32). Initially, there were 12 students, but only eight students finished the course. All the participants were total beginners in English, which was an important assumption of the experiment. It is worth noting that the private language school where the quasi-experiment took place was careful in separating total beginners and false beginners of English on the basis of background questionnaires and oral interviews. Total and false beginners also followed different programmes tailored to their needs. The assignment of the students to the two different groups was random and depended entirely on the school procedures. Of the two total beginner groups formed by the school, one was then randomly called Experimental

Using Cognates as a Vocabul ar y Lear ning Strategy

181

and the other one Control. The groups were taught by two different teachers of similar experience, age and background who had agreed to participate in the experiment. All the students needed English for the purpose of work and travel. At the time of the study, the following background information was gathered about the students from the Experimental group: Student E1, 40 years old, is a housewife. She helps her husband in his small business and enjoys this work. Student E2, 26 years old, works in an office as a receptionist. She enjoys meeting people. Student E3, 32 years old, works in the marketing department of a travel magazine. She finds her work interesting and likes travelling. Student E4, 30 years old, works for a private television channel in the advertising department. She knows German (pre-intermediate level). The following background information was gathered about the students from the Control group: Student C1, 35 years old, and works in an office as an accountant. She enjoys this work. Student C2, 22 years old, and studies geography. She likes traveling. Student C3, 28 years old, works as an assistant in a bank. She needs English for her future career. Student C4, 37 years old, runs her own business. She enjoys meeting people.

Method and materials The method of the research involved a longitudinal classroom experiment. The quasi-experiment took place in one of Warsaw’s renowned private schools of English and lasted from November 1997 to early February 1998. The groups, Experimental and Control, received 72 hours of tuition, i.e. a two-hour class, three times a week, which was a standard course for total beginners offered by the school. Both groups followed the same main programme. First, they underwent a 12-hour introductory period based on a selection of materials (the same for both groups), and next they covered the first 12 units of the True to Life Elementary course book, by Cambridge University Press. The two groups were taught by two different female teachers in their late twenties, both with several years of experience in communicative language teaching and comparable teaching styles and skills. Both specialised in teaching beginning and elementary-level courses. While the Control group followed the basic programme accompanied by other exercises focusing on vocabulary practice, the Experimental group did additional 28 exercises and communicative tasks activating cognates which

182

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

were carefully incorporated into the programme to match the course-book topics. The cognates used for the exercises had been extracted from Oxford Elementary Learner’s Dictionary (1995) by Oxford University Press, or had been recent borrowings which had entered Polish from English in the 1990s. The lists of cognates were later incorporated into the cognate database described in Chapter 4. The exercises included a considerable number of relatively sophisticated terms and abstract concepts (e.g. optimistic, satisfaction, etc.) which are normally taught around B1 level and are considered to be far beyond the reach of beginners. An example of such an exercise is presented in Figure 13.1 below. For more examples see Appendix 5 (p. 253). The exercises were accompanied by detailed teacher’s notes. The teacher of the Experimental group had been instructed to make students aware of the similarities between the words and of certain spelling and pronunciation differences between the Polish and English words used in the exercises. While using the collection of exercises the teacher had to make the students aware that they could intuitively rely on the words they knew from Polish to boost their confidence and be more ready to take risks and become productive. The work of the groups was monitored throughout the course. The students had to write three tests, identical for both groups. Two were prepared by the experiment designer and one administered by the school. The tests prepared by the experiment designer consisted of close ended tasks (e.g. cloze, word ordering, sentence transformation) and open ended tasks that provided opportunities for using a lot of vocabulary (e.g. listing words for semantic categories, guided writing). In their final lessons, both groups had to write the same pen-and-paper translation exercise. The students were instructed to translate 15 English sentences into Polish while the time of their performance was measured. Out of 156 words in the sentences, over one third, i.e. 56, were easy, similar

Figure 13.1 Example exercise activating cognate vocabulary

Using Cognates as a Vocabul ar y Lear ning Strategy

183

cognates in the form of recent borrowings into Polish that were being extensively used in the press and advertising at that time. However, all the words were different from the ones introduced to the Experimental group (for sample sentences see Appendix 6, p. 256). Finally, each student underwent a structured oral test. After simple warm-up questions, they had to answer three sets of questions. The questions were topically connected with the material from the textbook but answering them required producing language above the students’ actual level. The first two were based on pictures shown by the examiner, the third was based on the textbook material covered during the course: (1) Can you tell me something about the man in the picture? What do you think about him? What do you think about his personality? (2) Can you tell me something about the girl in the picture? What do you think about her? What do you think about his personality? (3) In your book there is an exercise where people say what they think about the English and England. I want you to tell me what you think about Poles and Poland in the 1990s. The last question was repeated in Polish, if it was necessary. The questions, difficult for beginners, aimed at checking how the students would cope with such task difficulty, and whether the Experimental group would have a better chance of completing the difficult task. The oral tests were recorded with a non-surreptitious recorder, which could have been an additional stress factor for the students (fragments of tape scripts can be found in Appendix 7, p. 257).

Method of analysis Since it is a case study on a small sample, qualitative methods of material analysis were used. As for the participants’ final oral tests, type-token ratios were calculated in order to compare their vocabulary content. In the calculations both the word-level n-grams (unigram sequences) and the collocation-level n-grams (3-grams) were taken into account.

Study 3: Results RQ1. Does raising the awareness of cognates help learners write, read and understand written English more effectively? As far as written work is concerned, the three tests taken throughout the course did not show any considerable differences between the groups. As all participants from the Experimental and Control groups were highly motivated, they scored high, within the same range of 85–95%. The participants’

184

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

performance did not differ in the type of vocabulary enumerated in the openended listing tasks and guided writing tasks in the tests, although such tasks gave the Experimental group a unique opportunity to use some cognates. This might have been caused by the fact that learners treated the written tests as inherently connected with the course-book material and restricted themselves solely to the course-book vocabulary when competing the test tasks. In the translation exercise at the end of the course, the Experimental group proved both faster and more skilled in translating unknown, but familiar looking words. The results are graphically presented in Figure 13.2. While the time necessary for the Experimental group to finish the task ranged from 10 minutes to 10 minutes 45 seconds, the Control group needed more time: from 12 minutes to 12 minutes 25 seconds. In most cases, the Control group students did not translate the sentences accurately. Their Polish translation frequently did not render the meaning of the original English sentences, or contained only separate words and phrases. Moreover, the Control group students had problems with translating a number of expressions and even whole sentences, for example: personal assistant, yuppie, workaholic, playboy, top standard, Polish team, sponsor, dealers. All these words were recent borrowings into Polish (borrowed in the 1990s), some still unassimilated. The sentences which proved especially problematic were: 4, 5, 8, 9, 11, 14 and 15, although they contained a high number of cognate words, recent borrowings to Polish extensively used in the press and advertising: 4. He’s got a personal assistant and uses a pager and a notebook. 5. He is a typical yuppie: a total workaholic and not a playboy. 8. Peeling and lifting - top standard only. 9. I have an immobiliser in my new jeep. 11. I sent him the information by email.

Figure 13.2 Time (in minutes) needed to complete the translation exercise with cognates by subjects from the Experimental and Control groups

Using Cognates as a Vocabul ar y Lear ning Strategy

185

14. The Polish team was sponsored by one of the Peugeot dealers. 15. Amway products are promoted by regional representatives in network marketing. In these sentences two students from the Control group translated only single grammatical words or gave up translation altogether. The Experimental group, apart from the faster pace of work, did not have any problems with accuracy. All the Polish sentences were close approximations of the original English ones. What is more, only one person (Student E2) did not attempt to translate sentence 11 (in 1998 the internet in Poland was still in its infancy and most people did not send emails). All the other sentences were translated in full. This indicates that the Experimental group was more risk-taking when confronted with unknown, although familiarly looking English expressions embedded in a sentence context. The participants from the Experimental group became accustomed to relying on their intuition and making intelligent guesses, at least when confronted with written vocabulary. RQ2. Does raising the awareness of cognates help learners produce spoken English more fluently? At the end of the course, after 72 hours of tuition, each student underwent a structured oral test. On the whole, the oral tests in the Control group were shorter than in the Experimental group. They lasted up to 10 minutes, and the topics proved really difficult for the students. In the Experimental group, on the other hand, the oral tests were much longer, and lasted from 15–20 minutes (for longer fragments of two typescripts see Appendix 7, p. 257). An examination of the transcripts of the oral tests brought interesting results concerning the participants’ accuracy and fluency. As for accuracy, after only 72 hours of tuition, even though directed towards oral communication, all the students often spoke in strings of words not linked grammatically, and they also made numerous pronunciation mistakes. The participants from both groups made similar syntactic mistakes, typical of lower levels. These included the lack of the copula verb be in the progressive tenses, the lack of the morphological distinction of the third person singular, the use of the pronoun he interchangeably with she. They all often got lost for words or expressions and code-switched into Polish. However, as far as fluency is concerned, there were considerable discrepancies between the two groups, in favour of the Experimental group. They pertained to vocabulary choice and use, but also the ability to take risks in vocabulary use, and the overall creativity of the students. The differences in vocabulary use can be easily noticed in the case of the adjectives: while the Control group was limited to adjectives from the course book, the Experimental group tried to use Latin-based cognates.

186

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

In the case of the first two questions where participants had to provide a description, both groups made attempts at describing the people’s appearance, although one student from the Control group refused to perform the task. Apart from Students E3 and E4 who used cognate words not introduced in the course book (marked in bold), the remaining descriptions in the Control and the Experimental groups are similar (the examiner’s turns have been deleted in all the examples below): Control group describing appearance Student C1: yyy This man this is old. I . . . [pause] is pretty? yyy [hesitation] He is . . . He are yyy . . . brown eyes . . . yyy small nose, no, large nose [laughs] yyy small mouth and brown . . . black and white hair yyy She is rich . . . rich. yyy Watch is very expensive. [. . .] This girl is very pretty. yyy Is young yyy This is from Japan. . . . Japan, Vietnam . . . yyy is young . . .. no nie wiem co powiedzieć [I don’t know what to say]. Student C2: yyy He is yyy calm . . . very happy. . . Student C3: yyy . . . yyy He .. he is . . .he is . . . he is yyy old man. yyy . . . yyy He is pretty. Nie wiem czy moz˙na tak powiedzieć, przystojny [I don’t know if you can say it this way: handsome. yyy He is . . . yyy This he is yyy . . . yyy speaking. yyy . . . Experimental group describing appearance Student E1:

yhm . . . yyy . . . He . . . isn’t young, but he isn’t old. yyy . . . I think . . . he has . . . forty-five . . . no he has. He forty-five old. Student E2: yyy This is a mild a man in mild year? W s´rednim wieku [middle-aged]? Middle yyy he is eem . . . he is. . . He has dark yyy hair and brown eyes yyy and she . . . he yyy talk Student E3: Right. I think that he is an artist yyy because yyy . . . he yyy . . . mimika [facial expression]. . . It’s man about thirty . . . not thirty . . . forty-fifty years . . . old yyy He is . . . very . . .. yyy. . . . Z ta˛ mimika˛ [something with this facial expression] . . . maybe active, active, yes. yyy. . . yyy Him eyes is very . . . popla˛tałam sie˛ [I got it mixed up] His eyes is brown yyy hair short. . . and dark yyy little white. He dresses like yyy and yyy black and short a little blue. Student E4: This is man, this is old man, . . .yyy he . . . has . . . black hair and brown eyes. (. . .) For me he is actor, *direction, yyy artist. (. . .) She is yyy nice young girls yyy She . . . I think das she is . . . artist. For me singer . . . and . . . no, singers. yyy she is . . . shy. yyy . . . she has yyy nice nice eyes . . . That’s all. yyy

Using Cognates as a Vocabul ar y Lear ning Strategy

187

When asked about the personality of the people in the pictures, the students from the Control group were only able to use a limited number of adjectives introduced in the course book (kind, self-confident, calm, happy, shy, friendly, pretty, interesting, serious, intelligent): Control group describing personality Student C1:

yyy I . . . I think yyy this man is . . . kind . . . and I think . . .. and he . . .. talking . . . talking yyy in a . . . this. I think yyy he is pewny siebie [self-confident] . . .. w kaz˙dym razie [anyway] I think yyy cos´ zwia˛zanego z moda˛ [something connected with fashion] Student C2: Co ja o nim mys´le˛. . . [what I think about him] I . . . think . . . he . . . very . . . working. . . . He is . . . self-confident. . . . Student C3: He is . . . he is confident yyy but .. he . . .yyy but he’s friendly. yyy . . .. Nie wiem [I don’t know] yyy She is happy, she is pretty . . . yyy . . .yy She’s . . . friendly. yyy . . . yyy She’s . . . she’s live in yyy Japan. yyy No nie wiem, nie wiem [I don’t know, I don’t know]. Student C4: yyy This is . . . man . . . this . . . he is yyy . . .. interesting yyy intelligent yyy . . . yyy serious, but . . . but kind. yyy . . . yyy He is . . . he is . . . vet or . . . and old man. . . yyy. . . The students from the Experimental group used the above-mentioned words and expressions, but also struggled to put across more of the meaning which they had in mind. Although they also lacked vocabulary and had equally limited grammatical resources, they made attempts to use crosslinguistically similar adjectives not introduced in the course book for total beginners, and usually taught at higher levels on language courses (creative, artistic, optimistic, active, organised, concrete): Experimental group describing personality Student E1: Student E3:

He is . . . he’s intelligent . . . confident . . . confident . . . and creative yyy . . . he have family . . . yyy he’s work . . . is he’s work in . . . office, or theatre, or cinema, or film. Very busy man. yyy And this man like life. yyy . . .. yyy He is very quickly person . . . yyy I think that like people. And world yyy and he’s artistic man. . . . work an artist. I think . . .. work yyyy . . . I think that work in film. This girl . . . come from Asia, . . . yyy is very young girl, about . . . he’s about twenty, twenty-five years old. yyy he he . . . She is very yyy. . .. nice girl . . ... She have . . . Her eyes is love. No love. Miłos´ć [love] tylko us´miechnie˛ta [but laughing]. She’s very optimistic. Yes. yyy I don’t know . . . how her work. . . . Is very problem.

188

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Student E4: For me he is intelligent. . . . and . . . and active. For me he is actor, direction. yyy artist. yyy Very organised, and . . . yyy . . . nice yyy stanowczy, nie wiem. [decisive, I don’t know. She is yyy nice young girls yyy She . . . I think das she is . . . artist. For me singer . . . and . . . no, singers. yyy she is . . . shy. (. . .) For me . . . yyy . . . for me big . . . yyy bigger personality is this man . . . That yyy this is artist yyy this is . . . and this artist too and . . . yyy This is concrete [showing the man’s picture] yyy . . .yyy This girl is . . . young. This man . . . old yyy when . . . because because yyy this man is pewny siebie. Yes, confident. The third question asked at the test, which concerned opinions about Poland and Poles, usually caused serious problems because, as planned, it was far beyond the level of beginners after 72 hours of tuition. When reminded about the course book material the question was based on (a lesson about opinions people have about the English), Student C2 from the Control group said: To była najgorsza lekcja. (It was the worst lesson). Then, she totally refused to communicate. Other Control group students struggled but used the same limited vocabulary as in the case of descriptions and found it difficult to say much: Control group talking about Poland Student C1:

I think then Polen yyy are . . . yyy Polish . . .? Polish often watch TV. yyy They are yyy they they . . . like money. yyy . . . yyy . . . yyy haven’t hobbies. yyy She . . . not not go . . . not go . . . not go the yyy theatre theatre /tiatr/ yyy cinema. yyy . . .yyy Wszystko [That’s all]. Student C3: yyyy Polish . . . Polish is very beautiful . . . yyy Polish people . . . yyy people yyy people . . . people is yyy is very. . . very friendly. yyy . . . yyy gos´cinny? Nie wiem. [Hospitable? I don’t know]. Hospital. yyy and . . . yyy very nice. yyy . . . Very beautiful country [laughs] Student C4: Polish people are yyy . . . serious . . . friendly. yyy They are . . . yyy listen . . . yyy They have got . . . a lot problems. . . . yyy . . . yyy They have problems yyy with . . . with yyy . . . yyy work . . .. and childrens . . . Juz˙ [That’s all]. On the other hand, the Experimental group’s utterances, especially in the answers to the third question, sounded more advanced although grammatically they were comparable to those uttered by the Control group. This is due to the use of vocabulary which was more sophisticated in comparison with the Control group. Relying on their intuition, the Experimental group

Using Cognates as a Vocabul ar y Lear ning Strategy

189

students used cognate words and expressions, which they had not learned during the course, or seen in the exercises and tests (e.g. hooligan, generally, racism, anti-Semitism, transformation, communism). In this way, the Experimental group students were also able to touch upon a wider range of topics: Experimental group talking about Poland Student E1: For me . . .yyy impossible yyy . . . talk yyy generally . . . people in Poland, or people in England. Or people in country . . . yyy perha. . . no, because, because yyy people, people is . . . no people are yyy. . . very different . . . in all country. For example . . . I don’t yyy speak all yyy . . . people in Poland is . . . nice. (. . .) yyy People in Poland yyy is very different yyy are interes. . . interesting people and . . . boring people, yyy gentlemen . . . or hooligan. They . . . tidy or not tidy . . . yyy For me in Poland generally . . . yyy but is . . . it is not true for all . . . all people in Poland. People is . . . not . . . People isn’t tolerant . . . tolerant. People is . . . rasizm racism, people is . . .. anti-Semitism. yyy For example, in street . . . impossible is . . . it is . . .yyy go . . .yyy with clothes, clothes . . . different, strange. yyy Because all people watch. . . . All people . . . yyy when when . . .. yyy it . . . yyy . . .. All the people in Poland talk of politics politician. . . Polish yyy think . . . yyy then know . . . all . . . of politics, of situation . . . economic, of yyy neighbour country, of . . . war in Iraq. Student E2: I . . . I think yyy I think Po. . . Po. . . Poland Polish people people . . . Polish people is are very busy . . . yyy they are . . . yyy a lot work . . . a lot work yyy . . . They are a lot learn yyy . . . There are sometimes humour, sometimes humour, but . . . because yyy the . . . our our life is yyy very strong. . .. yyy . . . what else . . . Yes, they have yyy a lot problems with work with yyy money, with children yyy because yyy czasy the nineteen nine . . . is transformation time yyy they must, they must yyy live in new . . . w czasach . . . in new time. Student E3: Polish people is very change. yyy . . . someone. . . no someone, something . . . kiedys´ [some time ago] Sometimes ago yyy people was yyy . . . hard . . . hard znaczy [it means] . . . people yyy life hard. Now is a better. Very very a lot of people have a yyy business yyy have a more money. yyy And now is it’s very many people haven’t money yyy have . . . it’s life is very . . . hard, it’s very . . . a lot of change . . . between people. In life, in money.

190

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Student E4: yhm . . . the Polish people . . . people they are . . . yyy smiling yyy and chaotic and yyy. . . sometimes they are .. yyy . . .. niegrzeczny. . . ar. . . arrogant . . . arrogant. And yyy . . . in Poland . . . Polish people yyy they are confident. (. . .) Abroad tak [yes] abroad they are . . . are yyy small . . . sa˛ mali i nie daja˛ sobie rady [they are small and they cannot cope]. (. . .) Politics. . . . yyy I am . . . in Poland the people have . . . are different yyy . . . different sides yyy and I am I have yyy that is difficult problems . . . yyy in Poland is a different party yyy and I . . . Poland come in demo . . . *democration. OK .. I am . . . people have . . . people in Poland yyy aren’t . . . przyzwyczajeni [used to] used to . . . to yyy that situation. yyy a lot people remember . . . yyy they remember communism and . . . a lot people . . . remember think znaczy tak [so yes] think yyy that is good. As can be seen, the Experimental group coped much better, both in terms of the length of their utterances and the topics they touched upon. They also used abstract nouns, not normally taught at beginner courses, such as: racism, anti-Semitism, politics, *situation economic (economic situation), humour, transformation, business, party, *democration (democracy), communism. Although they lacked grammatical and lexical resources, they managed to convey more meaning than the control group. Let us now focus on the use of adjectives because this can be a good illustration of the differences between the groups. Apart from Student E1, who tended to use more adjectives than the rest, and Student C2, whose oral test was extremely short, the remaining students used similar numbers of adjectives introduced in the course book. This is presented in Figure 13.3.

Figure 13.3 The use of adjectives in the Experimental and Control group structured oral tests

Using Cognates as a Vocabul ar y Lear ning Strategy

191

The adjectives from the book used by both groups overlapped. The categories of the adjectives used included: • • • • •

Colour: black, blue, white, brown, dark, opinion: beautiful, pretty, expensive, rich, nice, busy, different, boring, age: old, young, size/proportion: small, little, bigger, short, large, personality: confident, happy, strong, friendly, kind, serious.

There were also two cognate adjectives for expressing opinion introduced in the course book: intelligent and interesting. However, apart from that, three students from the Experimental group used such adjectives as: creative, economic, tolerant, active, artistic, optimistic, arrogant, chaotic, concrete, organised. All of them were used as personality adjectives, or opinion adjectives. Since all these adjectives are also internationalisms, it seems that the Experimental group were more ready to take risks in drawing on crosslinguistic similarity, whether consciously or unconsciously. RQ3. Does training in using cognates as a strategy affect the productive vocabulary of beginners of English? In order to apply quantitative measures that would allow for a comparison of the two sets of oral tests, the entire texts of the tests from the Control and the Experimental group (excluding the tester’s comments and participant hesitations, pauses and Polish-language intrusions) were converted into a small-sized speech corpus. On this basis, it was possible to calculate the TTR, both for the level of words and the collocations. TTR is a measure used for comparing the number of different word-forms, i.e. word types, with the number of occurrences of those word-forms, i.e. word tokens. Usually there are fewer word types than tokens because word forms are repeated within the same text. Calculating TTR allows for establishing more objective comparisons of the speech produced by the two groups (Tognini-Bonelli, 2001). As presented in Table 13.1, the length of the speech produced in the oral tests by the Control and Experimental groups is mirrored by the number of tokens: the texts of the oral tests by the Experimental group contain more tokens (152–266), so they are longer than those of the Control group (15–117). Obviously, the number of tokens is a very rough measure. As we could see, the utterances in both groups contained numerous repetitions, often due to the speakers’ lack of fluency. However, also the number of types in the texts of the oral tests indicates that the Experimental group used more different word-forms (70–93) than the Control group (11–57). This makes the Experimental group’s speech more lexically varied. It could be expected that the primacy of the lexical variation in the Experimental group speech samples would be reflected in the TTR: the

192

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Word level n-grams (unigram sequences) TYPES

TOKENS

TTR

TYPES

TOKENS

TTR

E1

93

266

0.3

50

143

0.35

E2

70

152

0.5

24

60

0.40

E3

84

192

0.4

20

51

0.39

E4

85

230

0.4

35

89

0.39

C1

57

117

0.5

13

30

0.11

C2

11

15

0.7

1

2

0.07

C2

28

57

0.5

13

38

0.23

C4

28

70

0.4

3

7

0.04

Control

Group

Experimental

Table 13.1 Type-token rations at the word and collection level for the Experimental and Control group structured oral tests Collocation level n-grams (3-gram sequences)

higher the TTR, the more ‘lexically dense’ a text is. Contrary to expectations, the TTR for the Experimental group ranged between 0.3 and 0.5, while for the Control group between 0.4 and 0.7, which means that the Control group speech samples were more lexically dense. However, the TTR measure varies very widely in accordance with the length of the text. The conventional TTR is informative, if we compare equal-sized text segments. Because the Control group oral tests were much shorter, the higher TTR at the word level did not unambiguously show that their utterances were richer in vocabulary. That is why the conventional TTR at the word level (analysis of unigram sequences) was followed by the analysis of a 3-gram frequency, which is a measure at the level of collocations. In computational linguistics, an n-gram is a contiguous sequence of n items from a given sequence of text or speech. For the Experimental group, the 3-gram TTR (TTR for sequences of three consecutive words) oscillated between 0.35 and 0.4. On the other hand, the 3-gram TTR in the Control group was much lower: between 0.04 and 0.23, as presented in Table 13.1. Concluding, the analysis of 3-grams allowed for establishing that the Experimental group produced more ‘lexically dense’ utterances.

Study 3: Discussion and Conclusions According to Ringbom (2007), if formal and semantic similarity of words in L1 and the target language exist, the learner does not need to make an effort to store it in the mental lexicon, but only makes a mental note that ‘this word in a similar form works in L2, too.’ This, in turn, positively influences CLI from L1 and facilitates learning. This is naturally the case of

Using Cognates as a Vocabul ar y Lear ning Strategy

193

languages that are in ‘similarity relations’, i.e. languages typologically similar. However, Polish and English remain in ‘contrast/difference’ relations (Ringbom, 2006, 2007). The beginning Polish learner will encounter numerous problems in learning English, and figuring out how this language works will take a lot of time. We also know that some learners, especially at lower proficiency levels, may have considerable difficulties noticing the formal correspondence of words (Odlin, 1989; Rubin, 1987; Swan, 1997). Since several studies indicated that the possibility to train noticing crosslinguistic similarity is beneficial (e.g. Banta, 1981; Dressler et al., 2011; Tréville, 1996), Study 3 aimed at finding out, whether it is possible to train cognate awareness as a learning strategy in low-level students of a typologically distant language to facilitate recognising and taking advantage of crosslinguistic similarities. The longitudinal study involved two small groups of adult beginners of English (Experimental and Control) in their first 72 hours of tuition. It was assumed that the learner should notice crosslinguistic similarity on the basis of formally similar, or identical individual words. Therefore, instead of additional exercises consolidating the course book vocabulary, the Experimental group was given 28 exercises activating Polish-English cognates. The exercises were meant to help students acknowledge the existence of cognates and thanks to that become more autonomous in their language learning. It was also assumed that the exercises activating cognates would give adult learners of English a feeling of success and achievement and in this way enhance their motivation. The teacher’s role was to provide the learners with vocabulary exercises, but also create a friendly environment, necessary for practice. Since the similarity of the target lexical form is perceived first, while functional or semantic similarity later in the process (Jarvis, 2009; Ringbom, 2007), the teacher was also supposed to explicitly point to formal and semantic correspondences between the cognates used in the exercises and their native equivalents. The teacher facilitated cognate activation by helping learners discover spelling and pronunciation differences. In such non-threatening, but illuminating way the teacher enhanced learners’ metalinguistic awareness, and especially the analysis of knowledge (Bialystok, 1987), missing at the early stages of learning a typologically distant language. Although the written tests administered throughout the course did not show any differences between the Control and Experimental groups, the differences occurred in two course-final exercises: the translation task and the structured oral test. Both tasks were purposefully above the actual level of the beginners. The translation sentences contained a high percentage (over one third) of words that had not been introduced in the course. However, these words were easy, form-similar cognates in the form of recent borrowings into Polish that were extensively used in the press and advertising. The time of the translation was measured. In the exercise, the Experimental group did better, in comparison with the Control group: the former’s translations were c.a. 2 minutes faster and they were also more accurate. This result

194

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

indicates that the Experimental group started taking for granted that formal similarity may help in learning. Thanks to that, they became more ready to take risks, and thus faster in their translations (RQ1). The oral tests show that the Experimental-group students used a much wider range of words, including infrequent abstract nouns (e.g. transformation, communism, racism), and a wider range of adjectives (e.g. creative, arrogant), which are normally taught at much higher levels. Most probably, the exercises on cognates not only provided the students with a number of words and phrases which were easy to learn, but also increased their metalinguistic knowledge and activated the strategic reliance on vocabulary known from L1 (see Chapter 9 for the role of L1 metalinguistic knowledge as discussed by Falk et al., 2013). The students from the Experimental group became more confident in their use of English because they had learned how to rely on their intuitions concerning crosslinguistic similarity (RQ2 and RQ3). An obvious limitation of the study lies in the fact that the extensive use of transfer of form and meaning (Jarvis, 2009; Ringbom, 2001) may also have lead the learners to some false assumptions. These can result in overusing cognates, using them in inappropriate contexts and activating false friends. If such false assumptions concerning form and meaning relationship are not revised, this may lead to fossilisation in vocabulary use (Ecke, 2014). Another limitation is the fact that the two groups were taught by two different teachers, albeit of similar skills and qualifications. It may well be that a particular teaching style had some impact on participants’ overall fluency. Finally, the small sample in the study does not allow for drawing more generalisable conclusions. Importantly, however, the study was conducted on total beginners to English, with practically no knowledge of another language (see Chapter 2 for more background), so the chance that the results were confounded by students’ previous knowledge of the L2 language are minimal. Also at that moment (second half of the 1990s) the influence of the internet and cable TV on incidental learning was much lower than today, because both these media were not widely used at that time. Thus, it may be concluded that raising awareness of cognates has a positive influence on learner performance, even at a very low level and in the case of languages distant typologically.

Study 4: Focus on Advanced Learners of English Study 4: Research design Aim, questions and hypotheses Study 4 on Polish advanced learners of English took place in the Institute of English Studies at the University of Warsaw, Poland in the years

Using Cognates as a Vocabul ar y Lear ning Strategy

195

2006–2008. It zooms in on another end of the language learning spectrum, namely on advanced learners of English. Seemingly, it would not make much sense to raise awareness of cognates as far as advanced learners of English are concerned because their metalinguistic awareness following from their proficiency level should be very high. However, as demonstrated in Study 1 (Chapter 12), even at the advanced level there are discrepancies in the awareness of lexical crosslinguistic similarity between students who are only bilingual and those who are accomplished multilinguals with at least two foreign languages above the B2 level. Polish advanced learners attending courses in the Institute of English Studies at the University of Warsaw are either students regular students attending courses in linguistics and British/American literature and culture who may be considered bilingual, or come from Interdisciplinary Cultural and Linguistic Studies and are accomplished multilinguals: they are advanced in English as their L2 and in at least two other foreign languages. Thus, in order to study both groups, the research consisted of two parts. Part 1 examined the vocabulary learning strategies of accomplished multilinguals. Part 2 checked whether there were differences in the vocabulary learning strategies between bilinguals and multilinguals with the same English language level and the same background. Then a longitudinal experimental classroom research followed, which was meant to establish whether strategy training, paired with raising awareness of English-Polish cognate vocabulary, would change the bilingual students’ strategies of vocabulary learning and use. All in all, Study 4 aimed to investigate the impact of raising awareness of cognates on student performance in receptive and productive tasks, as well as to observe potential changes in advanced learners’ vocabulary learning strategies. The following research questions were posed: RQ1. Do Advanced Multilingual learners of English consciously use crosslinguistic similarity as a learning strategy? RQ2. Do Advanced Bilingual and Multilingual learners use similar vocabulary learning strategies? RQ3. Does training change the vocabulary learning strategies of Advanced Bilinguals and brings them closer to the ones used by Advanced Multilinguals? RQ4. Does raising awareness of Polish-English cognates have an impact on recognising cognates when reading a text? RQ5. Does raising awareness of Polish-English cognates have an impact on cognate use in written language? RQ6. Does raising awareness of Polish-English cognates have an impact on cognate use in spoken language? Here, it is hypothesised that raising awareness of Polish-English cognates vocabulary should have a positive impact on using cognates as a learning

196

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

strategy in the case of advanced bilingual learners. It is also hypothesised that raising awareness of cognates should increase the ease of noticing cognates in texts, and using them in language production (written and spoken) whenever appropriate.

Participants The study involved 165 bilingual advanced learners of English as L2 (C1/C2) with Polish as L1, aged 19 to 25. The sample included 83 multilingual students (the same as in Study 1, Chapter 13), from now on called Multilinguals, and 84 bilingual students (different from those who took part in the survey presented in Study 1, Chapter 13), from now on called Bilinguals. The Bilinguals were 6 groups of students from the Institute of English Studies at Warsaw University, following a general English course aiming at vocabulary improvement. At the time when the study took place, teaching third languages was not obligatory in Polish schools, and the bilingual group either did not claim to know any language other than English, or assessed their proficiency in L3 as elementary. That is why they are treated here as bilingual: it was assumed that their L3 knowledge can be disregarded in this respect due to the threshold in proficiency needed for cognate activation (as discussed in Chapter 9). Also, the L2 status effects to occur the languages would need to be used relatively frequently and recently, which was not the case. The Bilingual participants’ awareness of cognates was measured before the course with the same questionnaire as presented in Study 1 (see Appendix 1, p. 246). The participants’ questionnaire results did not show significant differences to the ones obtained by the Advanced Bilinguals in Study 1. Thus, it was assumed that the participants might benefit from raising awareness of cognate vocabulary. The Advanced Multilinguals (the same as in Study 1, Chapter 13), who were at least trilingual, were advanced in their L3 languages (C1/C2 in Spanish, French, Portuguese, Italian, Russian or German) and most of them learned various L4 or L5 languages (A1 to B2 in Spanish, Portuguese, Catalan, German, French, Italian). However, in their background questionnaires, none of them indicated knowing Latin or Greek. The Multilinguals followed the same English courses as the advanced group. So, the level of L2 English of Multilinguals and their background was absolutely comparable with the Advanced Bilingual group. However, in comparison with the bilingual students such accomplished multilinguals (c.f. Biedron´, 2008; Biedron´ & Szczepaniak, 2012) can be considered more successful language learners in terms learning strategies they employ, including vocabulary learning strategies (see Chapter 10). Thus, it is possible to use them as a gauge for measuring students with less language learning experience. This will be done in Study 4.

Using Cognates as a Vocabul ar y Lear ning Strategy

197

Method and procedure Part 1: Multilinguals Part 1 of the research involved only the Multilinguals (n = 83). It investigated the role of multilingualism in noticing the enhancing factor of cognate vocabulary in language learning. In a separate set of questions added to the questionnaire in Study 1 (Chapter 13) the Multilinguals were asked to comment on the similarities between the European languages they knew. They had to answer four open-ended questions (formulated in Polish) concerning their perception of similarities between the European languages they knew: (1) Do you notice any similarities between the European languages you know? (2) Which language systems (lexis, phonology, grammar) of the languages you know are the most similar? (3) Do the similarities help you in language learning? (4) What helps most? Additional space was left for comments on any regularities they noticed concerning the lexical similarities between languages. The Multilinguals were also asked to fill in a closed-ended questionnaire concerning their vocabulary learning strategies which was based on a list in Chamot (1987: 77). The questionnaire can be found in Appendix 8 (p. 260). They were asked to choose up to seven strategies they found the most useful when learning vocabulary. In the questionnaire, each strategy was accompanied by its definition, so the students clearly understood what was meant by each strategy name. Part 2: Bilinguals Part 2 of the research involved only the Advanced Bilinguals (n = 84). The study in the longitudinal quasi-experimental design lasted 1 semester, i.e. 30 hours of instruction (1.5 per week). Initially, an oral placement test took place, which tested students’ general communication skills and pronunciation. The placement was administered by the Institute and was common to all first year students. Next, the students who successfully passed were randomly assigned to different groups. The quasi-experiment was carried out on six groups of students, each consisting of 12 to 14 participants and taught by the same teacher (the experiment designer). The entire quasi-experiment was spread over three semesters because the experiment designer taught two groups each semester. Out of the two groups taught in the semester, one was randomly assigned to be Experimental and the other one Control. Thus, the Experimental group 1 and the Control group 1 participated in the study in the first semester of the experiment, the Experimental group 2 and the Control group 2 in the second semester, while the Experimental group 3 and the Control group 3 in the third semester of the experiment.

198

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Basic course for the Experimental and Control groups. All groups, Experimental and Control, underwent a course called Integrated Skills. The aim of the course, which was topic and function based, was to provide intensive linguistic practice in communication skills. The course additionally focused on introducing, revising and systematising relevant topical vocabulary and functional expressions in communicatively close-to-real contexts. The course material covered the following broad vocabulary areas: relationships, describing people, housing, food, sports and holidays. The intensive vocabulary study component was based on selected units from The Heinemann English ELT Wordbuilder, and was connected with the topical areas mentioned above. Experimental intervention. The experimental intervention consisted of two components: vocabulary learning strategy training and cognate awareness-raising: 1. Vocabulary learning strategy training. The importance of vocabulary learning strategies was explained to both the Experimental and the Control groups at the beginning of the course. During the course all the students were informed about various vocabulary learning strategies, as enumerated by Chamot (1987: 77), i.e. note-taking, imagery, translation, inferencing, repetition, contextualisation, elaboration, grouping, auditory representation, directed physical response, deduction, resourcing, transfer and recombination. The strategy training was incorporated in the course programme in such a way that several methods of vocabulary study were pointed to within each new vocabulary topic introduced. Effectively, strategy training took place during 13 out of the 15 classes. 2. Cognate awareness raising. All groups, Experimental and Control, were additionally presented with examples of cognate vocabulary gathered into thematic groups. All the groups also took part in short, simple tasks activating cognates, similar to the ones described in Study 3. These included word mapping, listing, grouping, matching, contextualisation, recombination and communicative activities (see Appendix 5 for exaples, p. 253). However, during the cognate activation tasks only the Experimental groups were specially sensitised to the existence of crosslinguistic similarity and they were made aware that certain vocabulary areas (e.g. personality adjectives) abound in cognates deriving from Latin and Greek common to Polish and English. On each occasion of introducing the cognate-oriented tasks, i.e. eight times throughout the course, the Experimental groups were informed that using cognates might be a useful strategy. Importantly, in the Control groups, the same tasks on cognates were treated as unimportant warm-ups and time fillers. This made the whole teaching procedure nearly the same for both groups, only with the difference in the level of students’ awareness-raising of crosslinguistic similarity.

Using Cognates as a Vocabul ar y Lear ning Strategy

199

Measurements 1. Pre-test of vocabulary learning strategies. At the beginning of the course, students from each group were given a background questionnaire including a part which concerned their vocabulary learning strategies (see Appendix 8, p. 260). 2. Recognition and production of cognates. During Part 2, the recognition and production of cognates was measured at two stages: reading and writing after the first intervention (half of the course), and speaking after the second intervention (end of the course). a) Reading. The students’ recognition of cognates was assessed on the basis of a reading task. The students were given a text of 830 words, which included 70 cognates of mean NLD = 0.52.1 Their task was to scan the text and underline the cognates within the time limit of 6 minutes. This time limit was enough for advanced students to read a text of this length. b) Writing. The students’ production of cognates was measured by a writing task that took place in the middle of the course during a class on describing personality. It involved producing a short paragraph comparing two people and it was the final step of a series of vocabulary study and communicative activities involving the use of personality adjectives (which included numerous cognate adjectives originating from Latin). The students from the Experimental groups had been sensitised to cognates during the communicative activities and vocabulary study, whereas Control groups had been exposed to the same vocabulary, but not sensitised to cognates (see Appendix 5, sample exercise 3, p. 254). c) Speaking. Production in speaking was measured after the second intervention phase, at the end of the course. The task was a formal oral test performed in pairs in front of the teacher (the tester). The task was a role-play and the topics involved describing personality, clothes, interior design, and travel and holidays. A non-surreptitious microphone was used and the speakers were always informed that they are being recorded. These circumstances, together with the fact that the grade for the oral task added to the final grade for the course, created high anxiety-provoking conditions. It was hoped that in such conditions using cognates as a coping strategy may be revealed. 3. Post-test of vocabulary learning strategies. At the end of the course, students from each group were again given a questionnaire concerning their vocabulary learning strategies, the same as at the beginning of the course (Appendix 8, p. 260). The entire experimental procedure is summarised in Table 13.2.

Pre-test (week 1) INTERVENTION (week 2–6) INTERVENTION (week 2–6) Midterm measurement Receptive (week 7) Midterm measurement Productive (week 8) INTERVENTION (week 9–14) INTERVENTION (week 8–14) End-of term measurement Productive (week 15) Post-test (week 15)

Experimental and Control

Experimental and Control

Experimental and Control

Experimental

Experimental and Control

Experimental and Control

Experimental and Control

Experimental

Experimental and Control

Stage of Experiment

Target group

Vocabulary learning strategies questionnaire

Speaking – discussing a topic on the basis of clues given

Vocabulary learning strategy training + Activities on cognates Raising awareness of cognates

To train vocabulary learnings strategies and provide both groups with cognate input To examine the impact of cognate awareness-arising To compare using cognates as a compensation strategy when speaking in a stressful situation To compare with results obtained from accomplished Multilinguals

To compare using cognates as a compensation strategy in written texts

Writing a short personality description

Reading a text of 800 words and underlining cognates

To compare with results obtained from accomplished multilinguals To train vocabulary learnings strategies and provide both groups with cognate input To examine the impact of cognate awareness-arising To compare recognising cognates

Justification

Vocabulary learning strategies questionnaire Vocabulary learning strategy training + Activities on cognates Raising awareness of cognates

Description

Table 13.2 Experiemental procedure outline in advanced Experimental and Control groups

200 Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Using Cognates as a Vocabul ar y Lear ning Strategy

201

Method of analysis To answer RQ1, the questionnaires provided by Multilinguals were analysed with the use of qualitative methods and descriptive statistics. For the purpose of answering questions RQ2 and RQ3 all the results from Experimental groups 1, 2 and 3 as well as from Control groups 1, 2 and 3 were conflated. The answers to RQ2 and RQ3 were obtained from descriptive statistics and 2 × 2 Pearson Chi-square tests, α decision level set at 0.05, which were used to indicate differences between the Experimental and the Control bilingual groups and Multilinguals at the beginning and at the end of the course. The independent variable was the level of awareness raising of cognates in the groups. In order to obtain answers to RQ4, RQ5 and RQ6, the noticing and production of cognates were analysed by t-tests used to indicate contrasts between groups. Additionally, the written texts produced by students were analysed in terms of type-token ratios (TTR) and the use of cognate types and tokens for each word category (adjectives, nouns, verbs and adverbs). The oral tests were transcribed, and also analysed qualitatively in terms of type-token ratios and numbers of cognates used.

Study 4, Part 1: Qualitative results RQ1. Do Advanced Multilingual learners of English consciously use crosslinguistic similarity as a learning strategy? As for the answers to the four questions concerning their perception of similarities between the European languages they knew (Do you notice any similarities between the European languages you know? Which language systems [lexis, phonology, grammar] of the languages you know are the most similar? Do the similarities help you in language learning? What helps most?), responses from 83 students were analysed. A great majority of the Multilingual respondents (89%) admitted noticing considerable similarities between the European languages they used. Of the systems (grammar, lexis or phonology), 68% of the respondents opted for vocabulary as the most similar across European languages, 31% chose grammar and only 1% phonology. The Multilinguals strongly agreed that crosslinguistic similarities helped them with language learning (95% positive answers) and strongly pointed to lexical similarities (64%) as the most helpful. Interestingly enough, they did not see phonological similarities as relevant for language learning. The Multilinguals were also encouraged to comment on the regularities concerning the lexical similarities between European languages. Altogether, additional commentaries from 77 participants were obtained, with only one student claiming not to notice any regularity. The comments, all in Polish, were often long and sophisticated: they revealed the students’ considerable

202

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

linguistic and metalinguistic awareness, although many of the respondents did not use metalanguage when commenting. To save space, the comments are presented only in their translated versions, for example: Student M1 (English C1/C2, French C1, Spanish C1, Dutch C1): A lot of words, especially those of higher register originate from Latin and have a similar form across languages. It is also possible to see relations between languages that result from the history of the language and the country, e.g. the influence of French on Dutch (mostly in Flanders). Apart from that, specialist vocabulary in many domains has a similar form/pronunciation because it derives from ancient languages, or borrows English expressions, modifying them to agree with the phonetics of the language. Student M4 (English C1/C2, Russian C1, Italian A2): In European languages there are a lot of words deriving from Latin and Greek. They sound and are written practically in the same way across languages. Also business and IT vocabulary is borrowed and used in original forms. Student M12 (English C1/C2, French C2, Portuguese C1): It is those words which derive from Latin that are especially similar. There are also many borrowings which are adjusted to the spelling or pronunciation of a given language. Apart from that, in European languages there is more and more English vocabulary to do with computers and technology, which are often used in their original form. Personally, I notice phonological regularities which let me read books or newspapers in Spanish. I have never learned that language but I know its phonetic rules and I can mentally replace the Spanish prefix with its Portuguese equivalent which allows me to understand the text. Altogether, over half the Multilinguals (54.5% of the respondents) pointed to the influence of Latin and Greek on the European languages. A considerable number mentioned common roots of languages, language families and morphological similarities (40%), as well as the historical processes of language change, including the process of borrowing (18%). A considerable number among the Multilinguals commented on similarities in specialist vocabulary (22%). Their answers are graphically presented in Figure 13.4. Four students additionally commented on the learning process and stylistic issues: Im lepiej zna sie˛ je˛zyk, tym wie˛cej podobien´stw sie˛ dostrzega (The better you know the language, the more similarities you notice). Zauwaz˙yłam, ˙ze zwie˛kszaja˛c swój zasób słów je˛zyka polskiego poszerzam równiez˙ zakres rozumianych słów je˛zyków obcych (I have noticed that when

Using Cognates as a Vocabul ar y Lear ning Strategy

203

Figure 13.4 The distribution of comments provided by Multilinguals concerning lexical similarities between European languages

expanding my Polish vocabulary I also increase the range of foreign words I understand). Według mnie, najwie˛cej podobnych słów jest typowych dla je˛zyka formalnego, bardziej eleganckiego niz˙ codzienny (According to me, the largest number of cognate words are those typical of formal language, more elegant than everyday language). Łatwiej przyswaja sie˛ kolejne słowa nowego je˛zyka, gdy wyste˛powały podobne we wczes´niej poznanym je˛zyku. (It is easier to acquire words in the new language if there are similar words in the language you already know). The comments presented above strongly indicate that the Multilinguals, who are proficient in several languages, acknowledge the existence of crosslinguistic similarities in the area of lexis and try to use the potential affordances such similarities offer. The Multilinguals were additionally asked to fill in a questionnaire, based on Chamot (1987: 77), concerning their vocabulary learning strategies. Thanks to the wording of the questionnaire the students could clearly understand what was meant by each strategy name (see Appendix 8, p. 260). Out of the entire group, 52 Multilinguals responded to the questionnaire. The Multilingual respondents’ top five choices were: inferencing, i.e. guessing from context (73.1%), transfer, i.e. searching for L1 similarities (59.6%), translation (55.8%), note-taking (51.9%), and contextualisation, i.e. using a new word in a meaningful context (51.9%). We can notice that transfer, involving the conscious reliance on cognate vocabulary, was the second most important learning strategy in the sample and was chosen by nearly two-thirds of the Multilinguals. These answers left no doubt that a lot of accomplished Multilinguals examined here perceived searching for crosslinguistic similarities in the area of lexis as an important and useful strategy in language learning.

204

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Study 4, Part 2: Quantitative results RQ2. Do Advanced Bilingual and Multilingual learners use similar vocabulary learning strategies? RQ3. Does training change the vocabulary learning strategies of Advanced Bilinguals and brings them closer to the ones used by Advanced Multilinguals? The vocabulary learning strategies of the Bilingual participants of the quasi-experiment were examined at the beginning and at the end of the semester via the same questionnaire (Appendix 8, p. 260). Students were asked to choose up to seven strategies they found most useful when learning vocabulary. The strategy choices were compared between the Bilingual Control and the Experimental groups and also compared with the responses from the 52 Multilinguals, i.e. with successful language learners who consciously employed the strategy of searching for crosslinguistic similarities. Here, we will not report on all the results, but will focus only on the most important differences in the between the Multilinguals and the Bilingual groups, as well as the most visible changes in the Bilingual groups after the experiment. Initially, the five most highly valued vocabulary learning strategies in the Experimental and Control groups were note-taking (chosen by 65.9% and 59.5% students, respectively), imagery (56.1% and 54%, respectively), translation (51.2% and 54%), inferencing (43.9% and 45.2%) and repetition, i.e. rote learning (41.5% and 40.5%). Importantly, the pre-test showed that in both Experimental and Control groups, the strategy of transfer, i.e. looking for L1-L2 similarities, was one of the least commonly employed strategies. On the other hand, when choosing the most important learning strategies Multilinguals strongly relied on cognate-based strategies. Figure 13.5 graphically presents the vocabulary learning strategies, as indicated by the Experimental and Control groups at the beginning of the course (pre-test) and compared with the Multilinguals. The vocabulary learning strategies in the Experimental and the Control groups were examined again at the end of the course (post-test). The results for all strategies at the end of the course in comparison with Multilinguals are presented in Figure 13.6. After the strategy training, an increase was noted in the use of contextualisation (chosen by 65.9% students in the Experimental and 58.5% in the Control group) and grouping (73.2% and 68.3%, respectively). There was a decrease in repetition (chosen by 17.1% and 19.5% students in each group, respectively). There was also a considerable increase in the use of transfer, now mentioned as useful by nearly half the learners from the Experimental group (43.9%), and by one quarter of the Control group (22%). The Pearson 2 × 2 Chi-square tests were calculated to compare each of the top five strategies chosen by Multilinguals with the respective strategy

Using Cognates as a Vocabul ar y Lear ning Strategy

205

Figure 13.5 Pre-test: The choice of most useful vocabulary learning strategies by the Experimental and Control groups at the beginning of the semester, compared with accomplished Multilinguals

Figure 13.6 Post-test: The choice of most useful vocabulary learning strategies by the Experimental and Control groups at the beginning of the semester, compared with accomplished Multilinguals

206

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

in the Experimental and in the Control groups at the beginning of the course (pre-test) and at the end of the course (post-test), as presented in Table 13.3. The pre-test results showed significant differences in the strategy choice between Multilinguals and Bilinguals for transfer in the Experimental group, χ2(1) = 15.91, p = 0.000, and in the Control group, χ2(1) = 27.03, p = 0.000. The difference between Multilinguals and Bilinguals was also significant for inferencing in the Experimental group, χ2(1) = 8.14, p = 0.0043, and in the Control group, χ2 = 6.91, p = 0.008. In the Experimental and the Control groups, the Chi-square values were not significant for translation, contextualisation and note-taking. The 2 × 2 Chi-square tests were again calculated to compare the strategies chosen at the end of the course (post-test) by the Experimental and the Control groups with the top five strategies chosen by Multilinguals. Most of the Chi-square values were not significant. Importantly, there was no significant difference in the choice of transfer between the Multilinguals and the Experimental groups, χ2(1) = 2.27, p = 0.132. However, the difference for transfer between Multilinguals and the Control groups in the post-test was still significant, χ2(1) = 13.27, p = 0.000. This means that raising awareness of cognate vocabulary in the Experimental group brought the students significantly closer to Multilinguals with respect to their approach to crosslinguistic lexical similarity. On the other hand, in the Control groups, the mere presentation of cognates in exercises which was not supported by awareness raising did not bring Control group students closer to Multilinguals with respect to acknowledging the role of cognates in vocabulary learning. It may be concluded that strategy training influenced the Advanced Bilingual students’ vocabulary learning strategies. After strategy training, Table 13.3 The top five strategies chosen by the Experimental and Control groups as compared with the strategies chosen by the Multilinguals. Pre-test and post-test results. Asterisks (*) indicate significant differences between a given Bilingual group and Multilinguals Strategies

Inferencing Transfer Translation Contextualisation Note-taking

Multilinguals (n = 52)

73.1% 59.6% 55.8% 51.9% 51.9%

**p < 0.01, ***p < 0.001

Bilingual Experimental Group

Bilingual Control Group

Pre-test (n = 41)

Post-test (n = 41)

Pre-test (n = 41)

Post-test (n = 41)

43.9%** 17.1%*** 51.2% 36.6% 65.9%

56.1% 43.9% 56.1% 65.9% 53.7%

45.2%** 7.1%*** 40.5% 35.7% 59.5%

56.1% 22.0%*** 56.1% 58.5% 65.9%

Using Cognates as a Vocabul ar y Lear ning Strategy

207

four of the most popular strategies in the Experimental and the Control groups resembled those of the accomplished Multilingual learners. However, transfer was still valued less by the Control group students than the Experimental group students who had undergone cognate awareness raising. Thus, it may be stated that awareness raising, which was the only difference between the teaching in the groups, helped the Experimental group acknowledge the role of cognates in learning vocabulary. RQ4. Does raising awareness of Polish-English cognates have an impact on recognising cognates when reading a text? The students’ recognition of cognates was assessed on the basis of a reading task in the middle of the course, after the first experimental intervention. The students were given a text of 830 words, which included 70 cognates, mostly of high NLD, i.e. not highly similar to their Polish equivalents (see Chapter 4 for a discussion of NLD). The students’ task was to scan the text and underline the cognates within the time limit of 6 minutes, which was enough for an advanced learner to complete the task. The recognition of cognates by the groups is graphically presented in Figure 13.7. Altogether, 74 students completed the task, 36 in the Experimental group and 38 in the Control group. Although there was a lot of individual variation within the groups, the mean number of cognates recognised in the text was higher for the Experimental group (M = 32.53, Md = 30.5, SD = 14.11) than for the Control group (M = 21.05, Md = 21.5, SD = 10.20). The t-test run to compare the means showed a significant difference between the Experimental and the Control groups, t(72) = 4.023, p = 0.000. The result indicates that already after the first experimental intervention, the Experimental group students,

Figure 13.7 Mean number of cognates noticed in text by the Experimental and Control groups

208

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

who had undergone cognate awareness raising, were able to notice and underline more cognates when reading the text, in comparison to the Control group. RQ5. Does raising awareness of Polish-English cognates have an impact on cognate use in written language? The students’ production of cognates was measured by means of a writing task in the middle of the course after the first experimental intervention. The writing task took place during a class devoted to describing personality. It involved producing a short paragraph comparing personalities of two people and it was the final step of a series of vocabulary study and communicative activities practising the use of personality adjectives (which included numerous cognate adjectives originating from Latin). The students from the Experimental group had been sensitised to the existence of cognates among the adjectives to be learned, whereas the Control group had been exposed to the same vocabulary and given the same task (see Appendix 5, sample exercise 3, p. 254), but not sensitised to cognates. Altogether, 74 students completed the task. Sample answers from the Experimental group and Control groups can be found below, original spelling and grammar. The PolishEnglish cognates used are marked in bold, cognate adjectives in italics. Experimental: Student 2E1: In my opinion Mr A would be the best boss for Dolores. At the first look he seemed very conservative and traditional, but I have to say that there is something in his face which makes me think that he’s a really nice, older guy. From the picture I can see that he is bit smiling. Boss A is demanding but friendly and tolerant in a way that he understands that the young assistant can be lost at the beginning. He’s the kind of boss who is no longer interested in ‘short skirts’, but rather he’s searching for an energetic, outgoing young assistant who will have opposite features of character that he has. They both would make a good working couple: sensible, mature, creative and imaginative. Control: Student 3C3: From my point of view the best for Ally’s boss is Mr A. First of all he looks very trustworthy. She could count on him. Secondly he seems to be friendly and kind so probably would treat her gently. What is more, he looks as if he were not a snob or boastful person who do whatever he wants. He would respect her. I suppose he is quiet reasonable and patient so he solve the problems without

Using Cognates as a Vocabul ar y Lear ning Strategy

209

shouting and arguments. Also he seems to be wellorganised person who keeps all documents in one place and tells his secretary about all his appointments. It may help her with her work. Finally he might be very spontaneous and cheerful what may make the time they spent together in office more funny and sunny. In the task, we were interested to find out if the texts produced by the Experimental and the Control groups differed in the overall cognate use and in the use of adjectives (cognate and non-cognate), which was the focus of the class. As for the overall cognate use in the texts, the Experimental group produced more cognates (M = 6.55, Md = 6, SD = 3.04) in their texts than the Control group (M = 4.81, Md = 4.5, SD = 2.33). As indicated by the t-test, the use of cognates differed significantly between the Experimental and Control groups, t(72) = 2.766, p = 0.007. The comparison of the means is presented in Figure 13.8. As the entire class was devoted to the use of adjectives, another comparison was made to determine for use of cognate adjectives between the groups. The Experimental group used more cognate adjectives (M = 3.64, SD = 1.88) than the Control group (M = 2.53, SD = 1.60). The t-test result also indicated that the difference between the groups was significant, t(21) = 2.73, p = 0.008. This means that the Experimental group used more cognate adjectives in their writing than the Control group. In the next step, a more detailed linguistic analysis of the texts was performed. Corpora of texts produced by each group (Experimental 1, 2, 3 and Control 1, 2, 3) were analysed in terms of their type-token ratios. It can be seen that when compared to the text length, cognate tokens were used

Figure 13.8 Mean number of cognates produced in texts by the Experimental and Control groups

210

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Table 13.4 Cognate tokens used in the writing task by the Experimental and Control groups Group

Experimental 1 Experimental 2 Experimental 3 Control 1 Control 2 Control 3

Number of texts

tokens

types

cognate tokens

Type-token Percentage of ratio cognates in tokens

11 12 15 14 14 8

1007 1033 1640 936 1516 895

322 296 417 257 433 297

68 70 115 54 88 41

0.32 0.28 0.25 0.27 0.28 0.33

6,8% 7,0% 7,2% 5,9% 5,8% 4,7%

systematically more often in the Experimental than in the Control groups, as presented in Table 13.4. Next, percentages of cognate adjectives, nouns, verbs and adverbs were calculated for the number of tokens (i.e. all word-forms) within each category produced by each group. The results are presented in Table 13.5. Out of all the word categories, only the use of adjectives between the Experimental and the Control groups was systematically different in the case of adjective tokens. Whereas in the Experimental groups cognate adjectives constituted 31–33% of all adjective tokens, in the Control groups these were only 21–23% of the adjective tokens used. A more detailed analysis of the use of adjectives found in Table 13.6. To conclude, the results discussed above point to the fact that raising awareness of Polish-English cognates had an impact on cognate use in the written language samples produced by the Experimental group students. The use of cognates produced in their texts was higher than in the Control group, who had not taken part in the awareness-raising tasks. Also the use Table 13.5 Percentage of cognate tokens used in the writing task by the Experimental and Control groups Group

Experimental 1 Experimental 2 Experimental 3 Control 1 Control 2 Control 3

Cognate adjectives in all adjective tokens used

adverbs in all adverb tokens used

nouns in all noun tokens used

verbs in all verb tokens used

32% 31% 33% 23% 22% 21%

23% 12% 0 21% 3% 6%

16% 19% 20% 20% 14% 19%

8% 6% 9% 7% 6% 4%

Using Cognates as a Vocabul ar y Lear ning Strategy

211

Table 13.6 Adjectives used in the writing task by the Experimental and Control groups Adjectives in total

Cognate adjectives

Group

types

tokens

types

tokens

% of adjective tokens used

Experimental 1 Experimental 2 Experimental 3 Control 1 Control 2 Control 3 Experimental TOTAL Control TOTAL

70 64 87 51 99 78 221 228

106 114 175 88 170 121 395 379

19 19 30 10 21 12 68 43

34 36 57 21 38 22 127 81

32% 31% 33% 23% 22% 21% 32% 21%

of cognate adjectives, which were the focus of the lesson when written performance was measured, was systematically higher in the Experimental group. We may assume that Experimental group students found using cognate adjectives easier when faced with a written task that followed the awareness-raising activities. RQ6. Does raising awareness of Polish-English cognates have an impact on cognate use in spoken language? As for the speaking task, altogether, 84 students performed the task, from the Experimental group (n = 43) and the Control group. (n = 41). They had to take a formal oral test that was performed in pairs in front of the teacher/ examiner. The task was a pair-work negotiation activity, where students were assigned topics and roles to perform. Six different topics were used, which involved describing personality, clothes, interior design, travel and holidays. All the topics were supported by visual clues that students were supposed to refer to. A non-surreptitious microphone was used and the speakers were always informed that they were being recorded. The oral tests were transcribed, and analysed quantitatively and qualitatively. The number of running words and type-token ratios were calculated for the cognates used. Table 13.7 below shows the results for both groups. It is worth noting that the Experimental group always had a higher type-token ratio for the cognates they used and a higher percentage of cognates among all tokens in comparison to the Control group. Table 13.7 Cognates used in the speaking task by the Experimental and Control groups Group

Experimental Control

Total number of running words

Number of cognates used Tokens

Types

Type-token ratio

11,880 12,160

419 328

268 192

0.64 0.58

Cognates among all tokens 3.53% 2.69%

212

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

While the Experimental group produced together 11,880 running words (tokens) of spoken text, including 419 tokens of cognate use, the Control group produced 12,160 running words of text and 328 tokens of cognate use. This gives 3.53% of cognates in the total of the running words for the Experimental group, and 2.69% for the Control group. When speaking, the Experimental group students used cognate tokens on average more often (M = 9.93, SD = 4.62) than the Control group students (M = 7.73, SD = 4.14). The t-test indicates a significant difference between the groups in this respect, t(82) = 2.29, p = 0.024. The Experimental group also used significantly more cognate types (M = 7.77, SD = 3.52) than the students in the Control group (M = 5.76, SD = 2.99), t(82) = 2.81, p = 0.006. In the speaking task, there was a lot of individual variation in the length of the utterances, the type of vocabulary used and the use of cognates. The differences probably resulted from the varied topics of the tasks, the students’ spoken fluency, anxiety, and beliefs concerning the task difficulty. The utterances of the Experimental group consisted of 170–460 running words (tokens) and the utterances of the Control group of 181–496 running words. That is why it was decided to compare the equal samples of the first 150 words uttered by each student. Of interest may also be a tendency in the use of cognates observed when analysing the even 150-word samples of the final speaking task. In the first 150 words uttered, the Experimental group students often resorted to cognates. This tendency is illustrated by the samples from task beginning (two students taking part in the interaction in each sample): Student 1E1:

Student 2E1:

Student 13E1:

So, according to me. . . I think that those two people are a bit extravagant. As though they are trying to hide. . . what. . . who they are, so they are not showing. . . themselves. In my opinion. . . On the contrary, I can’t agree here. I would say that they are rather creative individuals and. . . also extravagant, which is definitely not a negative feature. So they are. . . By wearing the kind of clothes they wear, they just want to express what they have inside. (first 80 words of a conversation of 442 words)

So, I consider this person very tidy and. . . She seems to take great effort in personal hygiene, in these matters. She seems to have a. . . big burning ego. A person like this often goes to the disco, flirts around and chats with men. I think she may be a bit addicted to trying to play like that, and. . . Student 14E1: I... I don’t think so. She looks like a creative person. I think that she’s just conscious of her beauty. And... She

Using Cognates as a Vocabul ar y Lear ning Strategy

Student 13E1: Student 14E1:

Student 5E2:

213

looks... she looks like... I think she’s an independent person. But just look at that gap. . . between the skirt and those leather or latex boots. . . high-heeled boots. It’s quite provocative and. . . Oh I don’t think so. Maybe in some. . . eastern countries like Saudi Arabia this could be shocking, but in our culture it’s it’s is normal. (first 136 words of a conversation of 556 words) So what appeared to me after looking at this portrait or picture is a fact that. . . an unfortunate fact that . . . her appearance and the way she is dressed and her makeup perfectly reflects her personality because she she she appears to me as someone absolutely . . . who has nothing to . . . who has nothing to offer. She’s wearing fashionable clothes, fashionable jewellery, and and . . . and a haircut. I think it’s quite sad ehm. . . In most cases beauty does not come along with intelligence and . . . and some kind of . . . some kind of internal values. Or. . . If I call . . . could call this somehow metaphorically. What would you say? (first 111 words opening a conversation)

The number of cognate tokens uttered in the first 150 words by each participant was calculated and compared between the groups. The comparison revealed that the Experimental group students tended to use more cognates in the first 150 words uttered (M = 5.93, SD = 3.42) than the Control group students (M = 4.07, SD = 2.73). The t-test indicated that the difference between the Experimental group and the Control group was statistically significant, t(82) = 2.738, p = 0.008. Although the differences are significant, these data must be treated with caution. This is mainly due to the fact that the speech samples obtained are hardly homogeneous. However, it seems that the students who were more aware of cognates tended to rely on them, consciously or not, at the initial stages of a stressful and difficult oral task. We may very tentatively conclude that the students from the Experimental group were inclined to use cognate vocabulary as a strategy for coping with anxiety and task difficulty.

Study 4: Discussion and Conclusions Although objective crosslinguistic similarities are constant over time, the perception of such similarities changes as the learner’s language experience and proficiency increase (Kellerman, 1979). Research shows that multilinguals with considerable language learning experience have an enhanced perception of crosslinguistic similarity and treat searching for such similarity as

214

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

a strategy in vocabulary acquisition (e.g. Berthele, 2011). In Part 1 of Study 4 accomplished multilinguals with L1 Polish clearly admitted that the conscious search for crosslinguistic similarities at the level of lexis, or morphology, was one of their most important ways to learn new vocabulary (RQ1). This finding is a congruent with earlier research showing heightened metalinguistic awareness in multilinguals advanced in several languages (e.g. Aronin & Singleton, 2012a; Berthele, 2011; De Angelis, 2007; Gabrys´-Barker & Otwinowska, 2013; Jessner, 1999). Part 2 of Study 4 was inspired by the results of Study 1 (Chapter 13) and the finding that the awareness of cognate vocabulary often differed considerably between the Advanced Multilingual and Advanced Bilingual English learners of comparable L2 knowledge and background. Therefore, Part 2 of Study 4 on Polish advanced learners of English aimed at showing that in the case of such learners raising awareness of cognates brings them closer to accomplished multilinguals, for whom a conscious reliance on crosslinguistic influences between languages is a learning strategy. Thus, Part 2 of Study 4 involved strategy training of Polish Advanced Bilingual learners attending first year courses in the Institute of English Studies at the University of Warsaw, randomly assigned to three Experimental and three Control groups. After the pre-test, we established significant differences between the top five vocabulary learning strategies chosen by Multilinguals and by the Experimental and Control group students, with searching for crosslinguistic similarities largely neglected by all the bilingual students (RQ2). However, in the post-test at the end of the course, the strategies chosen by the Experimental group did not differ significantly from the top five strategies chosen by the Multilinguals. The strategy choice by the Control group still differed significantly in the case of searching for crosslinguistic similarities (RQ3). Since all the students were similarly trained in the vocabulary learning strategies and used the same exercises on cognates, but only the Experimental group’s attention was drawn to the crosslinguistic similarity of the words to be learned, we can conclude that it was awareness raising, rather than the mere strategy training that changed the bilingual students’ strategies of vocabulary learning. This finding has important implications, which stem from Schmidt’s Noticing Hypothesis (1990, 1998, 2001) stating that language acquisition involves attention and noticing. However, noticing alone may not be enough for learning, while learning needs more effective and explicit declarative knowledge of language rules. The Advanced Bilingual Control group obviously noticed cognates because they worked on the same exercises as the Experimental group. However, as emphasised by Schmidt (1990, 1998), for input to become intake noticing should be aided by explicit reflection and perceived as a subjective experience. This ties up with the notion of explicit, declarative knowledge by Paradis (2004, 2007). The results of the present study possibly indicate that raising awareness of crosslinguistic similarity between typologically different L1 and Ln requires a component of explicit

Using Cognates as a Vocabul ar y Lear ning Strategy

215

metalinguistic knowledge tapping onto explicit memory processes (Paradis, 2004). The finding also corresponds with a recent result by Falk et al. (2013), who show that crosslinguistic similarity is noticed easier if the learner has a high level of explicit metalinguistic knowledge of L1. Study 4 also aimed to investigate the role of raising awareness of cognates on student performance in receptive and productive tasks. The results of a reading and writing task in the middle of the course show that raising awareness significantly increased the ease of noticing cognates in texts (RQ4), as well as using them in written language production (RQ5). Taken together, the results for the Experimental groups were significantly higher than those for the Control groups as far as noticing cognates in a texts were concerned, as well as using more cognates in writing. As for the use of cognates in speech, the Experimental group used more cognates when performing the oral task, as compared to the Control group. Since the students’ utterances largely varied in length, equal 150-word samples from the beginning of utterances were compared. Here, the Experimental group revealed an interesting tendency to use more cognate words at the beginning of their utterances. Since the oral task was a stressful experience, we may very cautiously conclude that the Experimental group students used cognates as a way of coping with anxiety during the oral production task. In accordance with Dörnyei (2006) and Tseng et al. (2006), this may be regarded as an example of selfregulating behaviour. Dörnyei (2006: 61) states that ‘the important thing about proactive strategic learners is not necessarily the exact nature of the strategies, tactics or techniques they apply, but rather the fact that they do apply them.’ He proposes shifting the focus from the product of learning, such as the ability to use certain strategies, to the process of self-regulation. The results of the research presented in Study 3 and Study 4 may be interpreted in terms of both: product and process. The product of both studies was the change in thestudents’ vocabulary use and acknowledging the role of crosslinguistic influences. However, when analysing the speech samples of the Experimental groups in Study 3 and Study 4 we could also notice the actual process of using cognate words as a strategy. The fact that the Experimental group students used cognates when coping with difficult oral tasks shows that raising awareness of crosslinguistic similarity may lead students to employing some self-regulating behaviour in language production. To conclude, it seems that raising awareness of cognates as a learning strategy worked both for beginning monolingual and advanced bilingual learners: after training, both beginners and advanced students recognised similarities between Polish and English words and used them in production.

Note (1) http://www.onestopenglish.com/skills/reading/pdf-content/reading-lessons-tattooworksheet-1-upper-intermediate/147276.article

14 Awareness of Cognates as a Motivational Strategy: The Age Factor

Introduction As demonstrated in Chapter 13, raising awareness of cognate vocabulary has an impact on the receptive and productive language performance of both beginning and advanced learners of English. The awareness of cognates is also important for the development of vocabulary learning strategies. However, due to various typological and psychotypological factors, as well as learner characteristics (see Chapter 8, 9 and 10), cognate vocabulary may not be salient to the learner, so students need to be made aware of its existence. The two studies presented in this chapter investigate the role of the age factor in raising cognate awareness. Younger learners are different from adult learners: although some may possess a relatively high language command, their explicit knowledge and metalinguistic awareness is often lower than that of adults (Bialystok, 1987). Moreover, recent research on agerelated differences in recognising cognates indicates that adolescents are still worse than adults in recognising less similar cognates, even from related languages (Vanhove, 2014). Therefore, investigating teenagers’ awareness of cognates would be an interesting follow-up to studies on adults. The language learners studied here are much younger than the adults investigated in the previous four studies: they are teenagers in their second grade of the Polish lower-secondary school (gimnazjum). The studies focus on intermediate-level and elementary level students, who are 14-15 years of age and as teenagers are considered a difficult and demanding group to teach. It is worth noting that due to its popularity English is nearly by default the L2 for Polish teenagers (EURYDICE, 2012b). They begin learning the language in the first classes of primary school. In the lower-secondary school they start learning a compulsory L3, which is almost always German or French (see Chapter 2 for more information on teaching languages in Poland). 216

Awareness of Cognates as a Mot ivat ional Strategy

217

However, Polish schools differ a lot with respect to language teaching: in the majority of state schools the level of languages taught is rather low. Learners’ English proficiency is expected to reach the pre-intermediate (A2/A2+) level towards the end of the lower-secondary school, when students are 16 (EURYDICE, 2012b). There are also schools (private and public) which often organise additional school entrance exams to select students on the basis of their skills and language abilities. In these schools the level of language teaching is much higher and the students leave the lower-secondary school with English at the intermediate level (B1/B1+) at least, and one or two L3 languages (A2/B1). The studies in this chapter deal with both types of schools. The research presented in Chapter 14 also explores the possibility of using cognate vocabulary as a motivational strategy. A study by Gabrys´Barker and Otwinowska (2012) indicated that for Polish successful multilingual learners noticing crosslinguistic similarities is a strong motivating factor in additional language acquisition. Thus, we can assume that raising awareness of cognates can also be used to raise learners’ motivation. It is expected that raising students’ awareness of cognates should motivate them to become more autonomous in learning vocabulary and will help them adapt learning strategies based on crosslinguistic similarities. Study 5, which employs a qualitative methodology, is a classroom quasi-experiment on learners of a small private school in a suburban area of Warsaw. It explores the awareness of cognates in teenage students and treats them as a focus group to investigate the role of cognate awareness as a motivational strategy when teaching English to teenagers. Study 6, which is a large scale quantitative investigation, deals with teenagers from an average state school in a medium-sized town in central Poland and compares their results to teenagers from a renowned state school in a large city. The results of Study 5 were partly discussed in Otwinowska-Kasztelanic (2012).

Study 5: Focus on Teenage Learners of English Study 5: Research design Background, aim and questions Since age is an important factor in transferability (cf. Jarvis & Pavlenko, 2007), the aim of Study 5 was to assess whether awareness-raising tasks could be used for learners younger than adults. The study was also meant to qualitatively examine teenagers’ attitude towards using crosslinguistic similarity in the English language classroom. For the purpose of the study, the following research questions were asked: RQ1. Does exposure to and activation of English-Polish cognates help teenage learners recognise cognates more effectively while reading an English text?

218

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

RQ2. Is there any change in the awareness of cognates and their role at the end of the course? RQ3. What are the learners’ attitudes towards the awareness-raising tasks and cognate use at the end of the course?

Participants The participants of the small-scale study were 14 Polish teenagers (mean age 14.6 years) learning English since the first class of primary school, i.e. from the age of 7. They were attending the second grade of a private lowersecondary school in a prestigious suburban location near Warsaw, Poland and all came from middle-class families of high socio-economic status (SES). Their level of English (B1+) had been tested orally and in the written form by the school at the beginning of the school year. The students learned in two parallel groups, but their assignment to the groups had been random (they had been divided according to alphabetical order). Both groups used the same course book (Activate B1+ by Pearson), but were taught by two different teachers. For the purpose of the study, the groups were randomly appointed as Experimental and Control, so there were 8 students in the Experimental group and 6 in the Control group. All students learned German or French as their L3 at school and had reached at least the pre-intermediate level in one of the languages (A2 and above). Depending on the additional languages they had taken up outside school, the Experimental group learners differed in their knowledge of L3-Ln languages and their in cumulative learning experience: Student E1 (female): English B1+, French B1, Spanish B1 Student E2 (female): English B1+, Spanish A2, German B1 Student E3 (female): English B1+, French A2 Student E4 (male): English B1+, German A2 Student E5 (female): English B1+, German A2 Student E6 (female): English B1+, Spanish A1, German A2 Student E7 (male): English B1+, German B2 Student E8 (male): English B1+, French A2

Method and materials, analysis The study, which took place in spring 2010, lasted for three months. A mixed-methods approach was used. It involved a quasi-experiment with a pre-test and a post-test in the Experimental and Control groups (some statistics were performed to obtain approximate knowledge on quantitative results), and qualitative methods, such as an opinion questionnaire and a guided discussion in the Experimental group after the post-test. First, before the pre-test, cognates were defined to teenagers as those English words whose form and meaning was similar to their Polish equivalents. In the pre-test, participants in both groups were asked to fill in a Polishlanguage version of Questionnaire 1 (see Study 1, p. 144 and Appendix 1,

Awareness of Cognates as a Mot ivat ional Strategy

219

p. 246) to assess their awareness of Polish-English cognates. Next, both groups were given Reading 1, an unfamiliar English text1 of 414 running words, which included 76 cognates. Their task was to read it and underline as many cognates as possible within a time limit of 5 minutes. It was assumed that the time limit would be enough to scan a text of this length. The cognates were of low NLD (mean NLD = 0.35), so they were easy to notice. Throughout the course (for three months), the Experimental group students were sensitised to the existence of cognates and were given activities for raising awareness of cognate vocabulary (listing, word mapping, grouping, contextualisation, matching and recombination, as well as communicative activities), which were carefully interwoven into the course topics. The Control group were given other vocabulary tasks that did not involve cognates. In the post-test (at the end of the course), the students’ recognition of cognates was again assessed on the basis of Reading 2, which consisted of an English text2 of 830 words with 70 cognates (mean NLD = 0.52), the same as in Study 4. Again, they had to scan the text to underline cognates within a time limit of eight minutes. The teenagers were given more time to read the text than the advanced students in Study 4 because their level of English was lower. Since the small-scale study was meant to be more qualitative than quantitative in nature, emphasis was placed on investigating the learners’ subjective points of view on the method. Thus, the Experimental group were given Questionnaire 2 (originally in Polish, see Appendix 9, p. 261) in order to gather their opinions about the use of cognates in ELT. The teenagers had to choose those sentences that best described their attitudes towards the exercises: (1) The exercises helped me learn vocabulary faster. (2) The exercises helped me notice that there are similar words in both languages. (3) Earlier, I did not notice such words. (4) The exercises did not change my way of learning vocabulary. (5) I have not noticed any specific exercises dealing with cognates. (6) Now, I try to use cognates more often when writing in English. (7) Now, I try to use cognates more often when speaking English. (8) Now, I notice cognates more often when reading in English. (9) I stopped checking the meaning of the words which look similar to Polish. (10) I do not see the point in introducing such exercises. The Experimental group learners were also asked to write comments concerning the tasks and to assess their own knowledge of cognates. After that, the Experimental group were engaged in a class discussion on the role of cognate vocabulary. As for the analysis of cognate recognition, due to the small sample size, means and independent sample non-parametric Mann-Whitney U-tests

220

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

were calculated. Qualitative methods were used for the second part of the study.

Study 5: Results RQ1. Does exposure to and activation of English-Polish cognates help teenage learners recognise cognates more effectively while reading an English text? The pre-test consisted in recognising cognates in a written text (Reading 1). Although in absolute numbers the mean for recognising cognates the Experimental group (M = 38.75, SD = 13.81) was lower than in the Control group (M = 48.00, SD = 12.81), the independent samples Mann-Whitney U-tests did not show significant differences between the groups U(13) = 29.0, p = 0.573. Thus, it can be assumed that the two groups did not differ in terms of recognising cognates at the beginning of the study. The post-test took place roughly after three months from the beginning of the experimental intervention. Reading 2 consisted in scanning an English text of 830 words and to underline 70 cognates. The time limit was set to 8 minutes, since the text was twice as long as and more difficult than the first one. The cognates in Reading 2 were also of higher LD than in Reading 1. In the Experimental group, 7 out of 8 students managed to recognise over half the cognates in the text. Also, the mean for the Experimental group (M = 45.75, SD = 12.88) was much higher than for the Control group (M = 28.33, SD = 11.62). The independent samples MannWhitney U-test indicated a significant difference between the results for both groups U(13) = 6.00, p = 0.02. This shows that the Experimental group students were better than the Control group at recognising cognates in the post-test. RQ2. Is there any change in the awareness of cognates and their role at the end of the course? Similarly to the procedure described in Study 1 in Chapter 10, the participants’ answers to questions 6 and 7 of Questionnaire 1 (Appendix 1, p. 246), together with the pre-test, were categorised and labelled: the votes for below 150 cognates – Low Awareness, between 150 and 500 – Medium Awareness, and above 500 – High Awareness of the existing Polish-English cognates. When asked about the number of words whose form and meaning were similar in English and Polish, the teenagers in both groups tended to choose the answers between 50 and 100, rarely between 100 and 500 cognates, which indicates Low or Medium awareness of cognates. The fact that teenage learners show to know fewer cognates than the advanced group in Study 4 is probably related to their general vocabulary size, which is

Awareness of Cognates as a Mot ivat ional Strategy

221

connected with their age and L2 proficiency. Interestingly, there were three students whose knowledge of L3-Ln languages (other than English) was higher than the rest. It is these students who actually displayed higher awareness when answering about the number of cognates available. However, none of the teenage learners displayed High Awareness of cognates in their own vocabulary (question 7), as neither of them believed to know more than 500 such words, (see Table 14.1). Together with the post-test, the learners were again directly asked about their beliefs concerning the number of Polish-English cognates available and how many such words they knew. They were also asked to comment on their learning experiences, and on their knowledge of cognate vocabulary in general. In their comments, originally written in Polish, most learners appreciated the fact of being aware of similarities between languages, for example: Student E5 (L3 German, A2): I think it’s useful and it helps to learn the language. Such exercises help to notice that. (. . .) But I think there are more such words between English and German, because most German words were borrowed from English. Student E6 (L3 German, A2; L4 Spanish, A2): In my opinion noticing similarities between languages is useful especially when teaching young learners or when learning specialised language, e.g. business, where there are lots of words borrowed from other languages. I myself notice more similarities between the second and the third language I’m learning. Student E8 (L3 French, A2): I think that such exercises make us aware that there are lots of such words. The more languages we learn, the more similarities we notice, that help us learn further languages. I think that there are many such words. I know lots of them. Table 14.1 Students’ awareness of Polish-English cognates before and after the training Student

Student E1 Student E2 Student E3 Student E4 Student E5 Student E6 Student E7 Student E8

Level of awareness before the training

after the training

Medium Medium Low Low Low Low Medium Medium

High High Medium Medium Medium Medium High High

222

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

What is quite interesting, the higher the level of the learners’ L3-Ln languages, the more precise were their comments. Some comments involved deepened metalinguistic awareness of crosslinguistic similarity, e.g.: Student E7 (L3 German, B2): I don’t know what those exercises have changed, but now I notice similarities between some English and Polish words. (. . .) I think there are about 5000 such words, but I probably know, when I think about it, around 1000. Student E1 (L3 Spanish B1, L4 French B1): I don’t think that such exercises are good for me as an individual because I have used this method for about 4 years, i.e. since I started learning Spanish. When you learn several languages, especially from the same group, it is easier to learn through similarities. I can recommend it to all having problems with remembering words. There are thousands of similar words because once Latin was the international language. But I cannot say how many I know. It all happens when I find them in the text. I can know fewer words from such a range as “introvert” etc. It can also be noticed that in the case of all the learners from the Experimental group their awareness of cognates rose. The change of the learners’ awareness levels as indicated in Questionnaire 1 and as indicated by their comments is presented in Table 14.1. RQ3. What are the learners’ attitudes towards the awareness-raising tasks and cognate use at the end of the course? Questionnaire 2 (see Appendix 9, p. 261), added to the post-test, was meant to survey the learners’ awareness of cognates and attitudes towards the method. As the first task the learners had to choose statements in Questionnaire 2 (originally in Polish) to indicate what they thought of the awareness-raising exercises and of using cognate vocabulary in learning English. Although none of the teenagers indicated that the exercises helped them learn vocabulary faster, they did not negate the sense of introducing such exercises, either. Their answers are summarised in Figure 14.1. On the whole, the learners considered raising awareness of cognates as a positive and useful experience. Most of them decided that the exercises helped them notice crosslinguistic lexical similarities in general, and in particular, notice them in reading texts. Two students claimed they had never noticed cognate vocabulary before. What is interesting, five out of eight students claimed that their ways of learning vocabulary did not change under the influence of the exercises, which is somewhat contrary to what they expressed more explicitly in their free comments, as discussed below.

Awareness of Cognates as a Mot ivat ional Strategy

223

Figure 14.1 The Experimental group attitudes to the exercises (Questionnaire 2). The numbers on the scale indicate the number of students’ answers (n = 8)

The last step of the study was a moderated discussion about the role of crosslinguistic similarity run after filling in Questionnaire 2. The learners were asked to say whatever came to their minds in connection with cognate vocabulary. They were encouraged to state their opinions freely concerning whether such vocabulary can be useful in learning English and other languages to teenagers. The whole-group discussion was run in English towards the end of the lesson. It was recorded and the learners’ ideas were additionally noted on the board in the form of a mind-map to stimulate the exchange and to organise ideas. Below, there are some fragments of the teenagers’ comments in the discussion. Examiner: Student E1: Examiner: Student E1: Student E7: Student E5: Student E3: Student E1: Student E2:

What comes to your mind when you think of those words, those cognates? Useful for learning foreign languages. Do they help you in learning English, or other languages? They help in noticing connections between languages. The most important thing is to notice them first. They are easier to understand. They are harder words, from Latin, used in the past years. The people who travelled brought those words to our language. I can see connections with Spanish and French And German. For example the word ‘homonim’ ‘homonym’ and ‘Homonymus.’ There are more connections between Polish and French, Polish and German than Polish and English. Also between English and French.

When reading the teenagers’ answers we can clearly see that they had already understood that crosslinguistic similarity could help in language

224

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

learning. As their attention had been drawn to cognates, they could also notice similarity across the different languages they were learning. The students even tried to comment on some historical processes governing its emergence. Further on, the learners argued about the process of learning vocabulary and the usefulness of awareness-raising activities. Examiner: Student E8: Examiner: Student E1:

Student E5: Student E2: Student E7: Student E3: Examiner: Student E1: Examiner: Student E8: Examiner: Student E2:

So what do you think about such exercises, about showing learners that such words exist? I think it’s a good idea, but we need longer sessions, (...) Like this they are hard to remember. Do we really have to remember them? But . . . we don’t have to remember them! They are at the back of our heads. For example I can never remember the word for ‘wiosła’ [‘oars’, in Polish], but I don’t have to remember those cognate words. You can see them on the labels. These words are also around us in shops, when we use the net, on portals like Facebook. We don’t have to remember them. When we study, they will help us understand biology, chemistry. . . . No, not really. Rather history. We can use them when we are abroad and we are talking to other teenagers, our peers, and they help us understand them. So, how many such words are there? Fifty? (. . .) More. There are thousands of those words. There are over 3000 of such words between English and Polish. Do you know them all? It’s difficult to say. How can we know a word? If we can use it, we know it actively. If we only understand it we know it . . . Passive. Passively. We know lots of them passively.

The discussion shows that the teenagers from the focus group have discovered the potential of cognates and are ready to use the affordances offed by such crosslinguistic similarity.

Study 5: Discussion and Conclusions The main bulk of research on cognates has been carried out on adult language learners and users. Age and cognitive development of learners has been indicated as an important factor in hindering transferability (Jarvis & Pavlenko, 2007). Thus, the aim of Study 5 was to assess whether awarenessraising tasks could be used for learners younger than adults. The study was also meant to qualitatively examine teenagers’ attitude towards using

Awareness of Cognates as a Mot ivat ional Strategy

225

crosslinguistic similarity in the English language classroom because it was not entirely obvious if such learners, due to their lower metalinguistic awareness would be able to benefit from raising awareness of crosslinguistic similarities. For this purpose we concentrated on a small focus group of 14-year-olds with a relatively high English proficiency. In Study 5, not only their results, but also their attitude towards the tasks was important. The results of Study 5 clearly show that awareness-raising tasks can also be used with learners younger than adults, which corroborates the findings by Dressler et al. (2011). The activation of English-Polish cognate vocabulary helped teenagers recognise cognates more effectively while reading an English text, as compared to the Control group (RQ1). What is more, the teenagers’ awareness of cognate vocabulary grew, as shown by their answers to questionnaires at the beginning of the study and by their commentaries at the end of the study. In all cases the awareness rose from low to medium, or from medium to high. Although five out of eight learners claimed the activities did not change their way of learning vocabulary, the teenagers’ awareness of cognates was higher at the end of the quasi-experiment, which was indicated by their comments (RQ2). Another conclusion from the study is that the teenagers’ attitude towards using crosslinguistic similarity in language learning was positive. At the end of the study, the students also displayed quite enhanced awareness of crosslinguistic lexical similarity and its role, both in their written comments and in the discussion (RQ3). What is of special interest, the comments of the teenagers from the focus group revealed that also in the case of younger learners higher proficiency in L3-Ln leads to enhanced metalinguistic awareness, similarly to more mature learners (Aronin & Singleton, 2012a; De Angelis, 2007; Jessner, 1999, 2006; Singleton & Aronin, 2007). The higher the level of the learners’ L3/Ln languages was, the more likely they were to notice the usefulness of cognate vocabulary. Multilingual teenage learners with at least intermediate knowledge of L3 tended to be more aware of the role of crosslinguistic similarity than their peers with lower proficiency in L3-Ln languages. Such awareness may enhance language learning through evoking positive attitudes and motivation to learn languages other than English. Thus, an important conclusion of the study, stemming from the learners’ comments, is that the teaching of L3 should start as early as possible, probably in the upper-primary school. The teaching of L3 should also be carried out to a level higher than elementary. The main limitation of Study 5 is its small size and the SES of the learners. The teenagers in the research came from quite specific backgrounds (their parents sent them to a private school), which probably means that their families were rather well-off and appreciated the value of education. In such families, although English is perceived as the natural option, or a part of the obligatory, ‘ought-to-have package’ of the teenager (c.f. Dörnyei, 2006; Dörnyei, et al., 2006), other languages like Italian and Spanish gradually

226

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

became of interest for the most motivated learners and their parents. These teenagers and their familaies understand that a constellation of the local Polish language, global English and some other language (French, Spanish, Italian, German, or Portuguese) may be an important asset for a professional career in adulthood and a serious advantage on the job market. Admittedly, the SES of the learners, as well as the motivational factors disussed above that may have confound the results, making them less generalisable to an average population. We will try to overcome these limitations in Study 6.

Study 6: Motivating Teenagers. A Large-scale Study Study 6: Research design Background, aim and questions Study 5 was very limited in its scope, and size. The teenagers in the research came from middle-class families of high SES. Also their English proficiency was higher than the average for their age (EURYDICE, 2012b). Thus, the main aim of Study 6 was to validate whether awareness-raising tasks used with the teenagers from more privileged backgrounds could also be used with average Polish teenage learners. To this end, a larger-scale quasiexperiment was conducted on teenagers whose English proficiency was A1+/A2, and who learned English in an typical state school in a small town. We also needed to compare the results of teenagers in cognate recognition with results from a group of teenagers whose SES and language knowledge matched those of the focus group in Study 5. However, for practical reasons enrolling a larger number of teenagers attending small private schools was not possible. Thus, the results of the quasi-experiment on A1+/A2 learners were compared with the results achieved by multilingual teenagers of the same age, from a renowned large state school running bilingual education programme. Their L2-L3 profile and SES resembled the profile of the multilingual learners in Study 5 and the English-language level of these teenagers was B1+, similar to the level of the learners in Study 5. The following research question was formed: RQ1. Does exposure to and activation of English-Polish cognates help teenage learners recognise cognates more effectively while reading an English text? It was hypothesised that the activation of cognate vocabulary would increase the low-level teenagers’ ability to recognise cognates in texts and would bring the results closer to those of teenage multilinguals. Study 6 was also meant to examine the attitude of average teenagers towards using

Awareness of Cognates as a Mot ivat ional Strategy

227

crosslinguistic similarity in the English language classroom and to find out about the opinions of the teachers using the method. Thus, another question was formulated: RQ2. What are the learners’ and teachers’ attitudes toward the awareness-raising tasks and cognate use at the end of the course?

Participants and schools The participants of the study were 111 Polish teenagers from the second grade of lower-secondary school (mean age 14.65 years) learning English in two different schools. Participants from School A. Fifty-five students (mean age 14.5) from School A were the Experimental group examined in the study. School A is a large state school in a medium-sized town in north-central Poland (c.a. 55,000 inhabitants), 120 km from the capital. The school is a typical lowersecondary for teenagers from the region. It runs several parallel classes at each of the three grades. The research included second-grade students from five parallel A2-level (pre-intermediate) groups taught by five different teachers. In the secondary school, the students also start learning L3 languages (German or Russian), but in the second grade they do not achieve levels higher than A1. Participants from School B. The results of students from School A were compared with the results of 56 students (mean age 14.9) from School B, a bilingual secondary school in a large city in northern Poland (250,000 inhabitants). School B, ranked one of the best state secondary schools in the country, runs a bilingual Polish-English programme at the B1+/B2 level (upper-intermediate) and offers classes in L3 languages (German, French, Russian, Spanish) at the B1/B1+ level. Effectively, the students from School B are multilingual and their language knowledge is comparable to that of the most gifted learners from the focus group examined in Study 5. Students from three classes of the second grade in School B took part in the study.

Method and materials The study, which took place in the autumn of 2010, lasted for three months. Similarly to Study 5, a mixed methods approach was used. It involved a quasi-experiment with a pre-test and a post-test in the Experimental group from School A and a comparison of the test results with the scores from the Control group of multilingual teenagers from School B. Preparatory stage – instructing teachers. The experimental procedure looked as follows. Before the pre-test, teachers in School A had been instructed how to introduce the topic of cognates and how to define cognate words to learners. They received written instructions concerning the method, the classroom procedures and the stages of the quasi-experiment.

228

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Pre-test. In the pre-test, participants in School A were asked to fill in a Polish-language version of Questionnaire 1 (see Study 1, p. 144 and Appendix 1, p. 246) to assess their awareness of Polish-English cognates. Next, they were given Reading 1, an unfamiliar English text of 414 running words, which included 76 cognates, the same as in Study 5. The cognates were of low NLD (mean NLD = 0.35), so they were easy to notice. The students’ task was to read the text and underline as many cognates as possible within a time limit of 6 minutes. The same Reading 1 task, with the same instructions were given to learners in School B treated as a Multilingual Reference group. Experimental intervention. For one month, the students from School A (the Experimental group) were sensitised to the existence of cognates by their teachers. They were also given activities for raising awareness of cognate vocabulary (listing, word mapping, grouping, contextualisation, matching and recombination, as well as communicative activities), which were served as a supplementation of their course topics. School B students followed their own programme. Post-test. The post-test again involved the students’ recognition of cognates assessed on the basis of Reading 2, which consisted of an English text of 830 words with 70 cognates (the same as in Study 4 and Study 5). The cognates in the text were of higher NLD (mean NLD = 0.52), so they were more difficult to notice. Again, the students from School A and School B had to scan the text to underline cognates within a time limit of 8 minutes. The time limit was set to 8 minutes, since the text was more difficult than the first one and twice as long. Study 6 was also meant to gather more opinions on the method from the teenagers. Thus, the Experimental group was given Questionnaire 2 in order to gather their opinions about the use of cognates in ELT. The teenagers had to choose those sentences that best described their attitudes and reactions towards the method, the exercises and cognates in general (see Methods in Study 5 for the questionnaire points). After the experiment, the teachers of the Experimental group were asked to write their comments concerning the tasks, to assess the method and students’ reactions to the awareness-raising exercises. The teacher’s questionnaire can be found in Appendix 10 (p. 262).

Analysis For the analysis of cognate recognition, means, SD and dependent and independent sample t-tests were calculated for the results in School B. Since the data in the Experimental group (School A) were not normally distributed, non-parametric tests were used: the Mann-Whitney U-test to compare the pre-test and the post-test scores between School A and School B, and the scores within the Experimental group from School A. The Kruskal-Wallis H-test was calculated to compare the cognate recognition scores between the groups in School A taught by five teachers.

Awareness of Cognates as a Mot ivat ional Strategy

229

Study 6: Results RQ1. Does exposure to and activation of English-Polish cognates help teenage learners recognise cognates more effectively while reading an English text? Fifty-five learners from School A and 56 learners from School B took part in the pre-test (Reading 1). The text contained mostly similar low NLD cognates, but still the mean for recognising cognates in the Experimental group from School A was low (M = 20.80, SD = 13.89). The mean for recognising cognates in the Multilingual Reference group from School B was much higher (M = 34.16, SD = 14.06) and very close to the means of the Experimental group in Study 5. A non-parametric test was used to compare the groups because the data from School A were not normally distributed. The Mann-Whitney U-test showed a significant difference between the groups from School A and School B, U(109) = 653.5, p = 0.000). Of interest is also the distribution of the scores in the two schools, as presented in Figure 14.2 below. It was decided to divide the learners in the two schools into Low-Scorers (below 1/3 cognates recognised) and HighScorers (above 1/3 cognates recognised). The reason for such a division (below and above 1/3 of cognates) pertained to the fact that teenagers in Study 5 and Study 6 on average recognised fewer cognates than the advanced students in Study 4. For instance, only 4 students from School A achieved scores higher than 1/2 of cognates recognised in Reading 1. Thus, a different grouping (e.g. below and above 1/2 of cognates) would result in very uneven groups and difficulties in statistical comparison. In School A, only there was only 27% of High-Scorers, i.e. learners who recognised over 1/3 of the cognates in Reading 1. On the other hand, in the

Figure 14.2 The percentage of learners from each school who recognised cognates in Reading 1 (pre-test)

230

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Multilingual Reference group from School B there was only 30% of LowScorers. The results of the High-Scorers in School A were still lower than in School B (M = 38, SD = 14.19, and M = 41.15, SD = 9.38, respectively). The mean results of the Low-Scorers in School A were also lower than the mean results in School B (M = 14.35, SD = 6.7, and M = 18.11, SD = 7.88, respectively). Fifty-one learners from School A and 54 learners from School B took part in the post-test in the form of Reading 2, which was administered after three months of the experimental treatment in School A. The mean scores of the Experimental group students from School A for the post-test (M = 25.63, SD = 12.77) were higher than for the pre-test (M = 20.80, SD = 13.89). Since cognate recognition scores in School A did not have normal distribution, non-parametric tests were used. First, the Kruskal-Wallis test was run to compare the scores in order to check if the person of the teacher had any impact on Reading 2 result. Importantly, the score in cognate recognition did not depend on the group and teacher, since the Kruskal-Wallis test did not indicate any significant difference in the scores of students taught by the five different teachers H(4) = 5.364, p = 0.252. The difference in the means could not be directly compared between the pre-test and the post-test because the number and difficulty of the cognates to be recognised in Reading 1 and Reading 2 was different. However, in the post-test, School B was again treated as the Multilingual Reference group. The mean scores in School B students’ Reading 1 (M = 34.16, SD = 14.06) and Reading 2 (M = 33.69, SD = 14.26) were nearly the same. In order to compare the difference in the cognate recognition between School A and school B in the post-test the Mann-Whitney U-test was conducted. The results reveal that the difference between the schools was significant U(105) = 898.5, p = 0.002, which means that School A students were still worse than School B students in recognising cognates. On the other hand, of interest can also be the comparison between the Low-Scorers (those who recognised less than 1/3 of the cognates) and High-Scorers (those who recognised over 1/3 of the cognates) from School A and School B, as presented in Figure 14.3. Although the mean score of High-Scorers from School A was not much higher in Reading 2 than in Reading 1 (M = 38.8, SD = 9.25), in the more difficult Reading 2 the percentage of High-Scorers in School A increased to 40%, as compared with the 27% for the easy Reading 1. The mean score for the Low-Scorers in School A increased (M = 17.12, SD = 5.04) and the score was more evenly spread across the sample. The lowest score in the post-test in School A was 8, while in the pre-test there were four students with the 0 score. The Mann-Whitney U-test was conducted to compare the means between the Low-Scorers from Schools A and B. It showed that the differences between the groups were not significant, U(55) = 2.33, p = 0.812. Similarly, another Mann-Whitney U-test was conducted to compare the

Awareness of Cognates as a Mot ivat ional Strategy

231

Figure 14.3 The percentage of learners from each school who recognised cognates in Reading 2 (post-test)

means of the High-Scorers from Schools A and B. The differences between the groups were not significant, U(46) = 7.37, p = 0.35, either. These results indicate that within the Low-Scorers and the High-Scorers categories, School A teenagers’ recognition of cognates did not differ substantially from that of the Multilingual Reference group from school B. In School A the recognition of cognates improved over time between Reading 1 and Reading 2, most probably thanks to the experimental intervention. RQ2. What are the learners’ and teachers’ attitudes toward the awareness-raising tasks and cognate use at the end of the course? In parallel with Study 5, Questionnaire 2 (Appendix 9, p. 261) was added to the post-test in order to survey the learners’ awareness of cognates and attitudes towards the method. The learners had to choose statements in Questionnaire 2 (originally in Polish) to indicate what they thought of awareness-raising exercises and of using cognate vocabulary in learning English. Importantly, the teenage learners’ attitudes towards the awareness-raising exercises were strongly positive. Over half of the teenagers (51%) indicated that the exercises helped them learn vocabulary faster, and helped them notice similar words across languages (54.9%). Although only less than one quarter of the learners (23.5%) indicated they had not noticed such words earlier, now over half the learners claimed the exercises helped them notice such words when reading in English (51%). They also claimed trying to use cognates when writing in English (35.5%) and when speaking English (23.5%), and to have stopped checking the meaning of words which look similar to Polish (37%.3). A small minority negated the sense of introducing such exercises (13%): these were five Low-Scorers, but also two High-Scorers. Four of the learners were taught by the same teacher, so possibly the

232

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

teacher’s attitude might have had an impact on the students’ claims. The answers to Questionnaire 2 answers are summarised in Figure 14.4. On the basis of these descriptive results, we can tentatively conclude that positive attitudes towards the method prevailed. The learners from School A considered raising awareness of cognates as a positive and useful experience, rather than a negative experience. As for the teachers, the five teachers who conducted the experimental classes were asked to comment on the method and on learners’ reactions (for the questionnaire see Appendix 10, p. 262). Importantly, none of the teachers presented negative attitudes towards the method, or claimed not to see the point in introducing such exercises. None of them stated that the exercises were boring and unpleasant for the learners. However, they did not see them as particularly motivating, either. Three teachers confirmed that the students liked the method, although at the beginning they had been reluctant towards the exercises. On the other hand, all five teachers agreed that the awareness-raising method helped learners notice crosslinguistic similarity, that they stopped asking about the meaning of formally similar words and that they learned vocabulary faster. Four out of five teachers also declared that learners noticed cognates more often when reading texts, and one teacher claimed that they tried to use cognates when writing and speaking English. Incidentally, this was the teacher whose group scored highest on the posttest. However, the teacher himself/herself was not a significant predictor of the Reading 2 score, as shown by the regression analysis. Four teachers agreed that they would use the method in the future, and three claimed to have used a similar method before. The comments clearly indicate that both the teenagers from School A and their teachers discovered the potential of cognates and were positive about the usefulness of awareness-raising activities concerning crosslinguistic similarity.

Figure 14.4 The Experimental group’s (School A) attitudes revealed in Questionnaire 2. The numbers on the scale indicate the number of students’ answers (n = 51)

Awareness of Cognates as a Mot ivat ional Strategy

233

Study 6: Discussion and Conclusions In accordance with the qualitative results of Study 5, awareness-raising tasks could be used with teenagers. However, Study 5 was conducted in a private school, with teenagers of similar, high SES (their parents could afford a private school) and their English L2 proficiency was higher than average in Poland. Some of them also had considerable cumulative language learning experience with L3-Ln languages. Thus, the aim of Study 6 was to corroborate the findings of Study 5 on a larger group of teenagers, more representative for the population of 14–15 year olds in Poland (EURYDICE, 2012b) and check whether such learners would benefit from awareness-raising tasks equally to the group in Study 5. To this end, School A in a medium-sized town was chosen, with learners of average SES and English proficiency at the A1+/A2 level. School B was chosen in order to compare the results of learners from School A with learners similar to the group in Study 5 in terms of SES, level of L2 English and knowledge of other languages. School B runs bilingual Polish-English programmes (L2 at B1+/B2) and the pupils there are practically multilingual (L3 at B1). These learners served as the Multilingual Reference group, against which the results of the learners from School A could be compared. As could be expected, the pre-test (Reading 1) showed a huge discrepancy in the ability to recognise cognates between learners from School A and School B. Probably, this result is connected with the general vocabulary size of the learners from School A and School B. However, after the experimental intervention in School A the results of the post-test indicate change in cognate recognition. In both schools there were obvious differences between learners in taking advantage of language affordances (Segalowitz, 1997, 2001), so for the analysis the teenagers were divided into Low-Scorers and HighScorers in cognate recognition. Although the post-test results do not show a dramatic increase in terms of the mean scores for cognate recognition in entire School A, the comparison of the pre-test and the post-test between the Low-Scorers (less than 1/3 of the cognates recognised) and High-Scorers (over 1/3 of the cognates recognised) from School A and B, the differences are not significant. This indicates that the results of the Experimental group students from School A in cognate recognition in texts began approaching those of the Multilingual Reference group from School B. Having noted that in Reading 2 (post-test), the text was twice as long as in Reading 1 (pre-test) and the cognates were less similar than in the first text, we can conclude that the activation of English-Polish cognate vocabulary helped teenagers recognise cognates more effectively while reading an English text (RQ1). Most probably, raising their awareness of cognates increased their metalinguistic awareness and helped in establishing learned associations (Jarvis, 2009) between L1 and L2, allowing more lexical CLI from Polish.

234

Par t 4: Invest igat ing Le x ical Crosslinguist ic Simil ar it y in L anguage Lear ning

Study 6 examined not only the experimental results but also attitudes of both teachers’ and their students towards the tasks. The teenagers’ attitudes were strongly positive, as revealed by their comments and their distribution: over half of them believed that the exercises helped them learn vocabulary faster, and helped them notice similar words across languages and only some negated the sense of introducing such exercises. Also the teachers were positive that the method helped their students learn vocabulary better and recognise cognates. The teachers were eager to use the method again and to draw their learners’ attention to crosslinguistic similarities. (RQ2). As pointed out by Dewaele (2010), the perception of features of a new language and the perception of crosslinguistic knowledge that students can mobilise when learning a new language has an impact on the crosslinguistic affordances these learners will notice and strategically use when learning. In Study 6, both the test results and the comments revealed that in the case of average Polish teenagers awareness-raising helps then notice the usefulness of cognate vocabulary. In accordance with Dewaele (2010), we can conclude that drawing learners’ attention to crosslinguistic similarities and raising their metalinguistic awareness of L1 and L2 may influence their learning of other languages. Let us hope that also for teenagers from less privileged backgrounds learning about cognates will open up an array of social and individual language affordances (Aronin & Singleton, 2012b: 318) necessary to succeed in plurilingual environments and on the job market. Hopefully, thanks to its size and scope, Study 6 overcomes the limitation of Study 5. The main implication of the two studies is such that it is absolutely possible to introduce the notion of crosslinguistic similarity and to raise awareness of such similarities in the case of teenagers. Indirectly, Studies 5 and 6 also show the possible direction of changes in education. At the national level, apart from using international textbooks, it is necessary to prepare teaching programmes and materials which take into account learners’ L1 background and which focus on crosslinguistic comparisons. As shown in Studies 5 and 6 they can be quite successful even at the lowersecondary school level.

Notes (1) http://cottenceau1.free.fr/lab/Guardian/headscarves2.pdf (2) http://www.onestopenglish.com/skills/reading/pdf-content/reading-lessons-tattooworksheet-1-upper-intermediate/147276.article

15 Towards Plurilingual Language Teaching with English: Conclusions and Pedagogical Implications

A Summary of the Research Aims The research presented in this book stemmed from two basic assumptions. The first concerned the dominant role of English in the multilingual world. The second concerned the role of crosslinguistic lexical similarity awareness in multilingual language learning within and across typological language families. Before presenting the final conclusions from the research and implications for language teaching, let me briefly summarise these assumptions. Today, being bilingual, multilingual or plurilingual clearly becomes indispensable for taking an active part in education, in the job market, culture and social life. However, multilingualism and plurilingualism should be presented in terms of language constellations (DLC), rather than individual languages known to the speaker (Aronin, 2006; Aronin & Singleton, 2012a; Singleton & Aronin, 2007). Those constellations more than often include English, the current international and European lingua franca. Within language constellations English has a special rolewith reference to participation in various public fora. Whereas national and minority languages suffice at the country level, on the international forum, both real and virtual, ‘English has already become the very linguistic means to give speakers, especially of lesser-used languages, their voice within a (. . .) public discourse’ (Breidbach, 2003: 20). Therefore, proficiency in this language is still of major importance for participating in international communication, especially for speakers of such lesser-used languages as Polish. English, taught from the first classes of primary school, becomes a crucial part of Polish speakers’ multilingual 235

236

Cognate Vocabul ar y in L anguage Acquisit ion and Use

language constellations. Since achieving the goal of multilingualism is challenging for learners and teachers, my research project aimed to find out whether language teaching could be enhanced. As for the second assumption above, previous studies have shown that for bilinguals, and especially for multilinguals, cognate vocabulary between the languages they know becomes an important trigger for reaching higher levels of language knowledge (see Chapter 9). As often pointed out, ‘the more aware learners are of the similarities and differences between their mother tongue and the target language, the easier they will find it to adopt effective learning and production strategies’ (Swan, 1997: 178). A lot has been said about making students aware of similarities between the cognate vocabularies found within typological families: Germanic, Slavonic, Romance (e.g. Hufeisen & Neuner, 2004; Hufeisen & Marx, 2007). However, English, as a Germanic language, thanks to its historical development, shares hundreds of vocabulary items not only with Germanic languages, but also with languages from other groups. Thus, the aim of the research presented in the book was to explore the topic of cognate vocabulary between typologically distant English and Polish. It was meant to examine how Polish learners of English are aware of cognates, and whether raising awareness of cognates can help them learn and use English in an effective way. The entire project, which embraced six different studies involving nearly one thousand participants, revolved around three basic questions: What word-related and learner-related factors affect the awareness of cognates? How can adult English language learners benefit from raising awareness of cognate vocabulary? Can raising awareness of cognates be used with learners younger than adults? The answers to these questions are summarised in the sections below.

Factors that Affect the Awareness of Cognates The first two studies on Polish students of English (Chapter 12) dealt with the awareness and knowledge of cognates across the L2 proficiency spectrum and the role of cumulative language learning experience in enhancing this awareness. I assumed that the awareness of cognates cannot depend solely on the learner, but it will also be shaped by the features of words that the learner encounters. Thus, the essential conclusions from the first two studies pertain to the types of factors that modify the awareness of cognate vocabulary.

Word-related factors: Frequency and orthographic similarity of Polish-English cognates As for the word-related factors in cognate awareness, in both studies we can distinguish the effects of word frequency and the orthographic similarity

Towards Plur ilingual L anguage Teaching w ith English

237

of cognates. First, word frequency plays a role in the ability to recall cognate words together with L2 proficiency of learners. As Study 1 reveals, learners at lower L2 proficiency levels can enumerate only more frequent cognate words, probably because cognates of higher frequency are acquired earlier in the learning process. The ability to recall and enumerate less frequent words grows together with the L2 proficiency. This finding is also in line with research concerning vocabulary acquisition discussed in Chapter 8. Of interest is also the effect of orthographic similarity, as indexed by the NLD (see Chapter 4). Importantly, although NLD is an artificial measure, it bears psychological reality: studies reveal that it helps to simulate learners’ reactions to cognate vocabulary. As shown in Study 1, at lower levels learners can enumerate only those cognates which are more similar to their Polish equivalents. Probably, this is because L2 learners acquire cognates of low NLD earlier in the learning process. Study 2 has demonstrated that NLD can also help predict the performance of students in a translation task. Learners at lower levels of L2 proficiency translated the similar (Low NLD) cognates easier than the less similar ones (High NLD). Effectively, High NLD cognates are less recognisable because dissimilar orthography makes them more difficult to notice and learners at lower proficiency levels treat them as any other words they do not know. Thus, orthographic similarity affects the noticing and learning of cognate vocabulary: the lower the NLD, the more aware the learners are of the actual formal correspondence between the L2 and L1 orthographic neighbours and the easier they can learn the word.

Learner-related factors: Proficiency in L2 and L3-Ln As for the learner-related factors which interact with the word-related factors, it turns out that both: proficiency in L2 and multilingualism (proficiency in L3-Ln) play a role in cognate awareness. First, L2 proficiency is crucial in cognate knowledge: the higher the proficiency, the more cognates a learner knows and the greater the awareness of their number. On the contrary, the lower the level of English, the lower the awareness of lexical similarities across languages. Also, as demonstrated in Study 2, together with a growing proficiency in L2, the learner is likely to know cognates that are less similar to their native equivalents. On the other hand, L2 proficiency itself cannot explain the heightened awareness of cognates that multilinguals display and the qualitative differences between the cognates enumerated by advanced bilingual and multilingual students in Study 1. Multilingual students tend to be more aware of cognate vocabulary than bilinguals of the same L2 proficiency and background. The cognates they enumerate when asked to do so differ qualitatively from those enumerated by bilinguals: there are more adjectives and abstract nouns. Also, the type-token rations of their cognate list is higher in comparison to advanced bilinguals. This all points to certain qualitative

238

Cognate Vocabul ar y in L anguage Acquisit ion and Use

differences in the cognate knowledge and use of advanced multilinguals, as compared to advanced bilinguals. We can conclude that accomplished multilinguals, thanks to their cumulative language learning experience, more readily acknowledge the existence of crosslinguistic similarities. They also have greater metalinguistic awareness of cognates across languages (hence the use of more abstract nouns deriving from Latin and Greek) and can use the potential affordances such similarities offer. The quantitative results of the translation study (Study 2) reveal that multilinguals more proficient in L3 learn cognates earlier in the language learning process than other words. At higher levels they tend to employ guessing, or inferencing strategies more often: they can translate the cognate word, but are not entirely sure of its meaning. These results point to the conclusion that cumulative language learning experience does not guarantee higher vocabulary knowledge in a given language, or more correctness, but it helps in noticing relationships between cognates across languages. It also encourages the strategic guessing of meaning in the case of unknown words formally similar to the known words in other languages.

Benefits from Raising Awareness of Cognates The conclusions from the next two studies on Polish students of English (Chapter 13) refer to who and how can benefit from the awareness of cognate vocabulary in the language learning process. As Ringbom (2007: 104) points out, cross-linguistic similarity is ‘an important variable in the use of learning strategies: how the learner tries to enhance the effectiveness of learning.’ When asked explicitly if they notice regularities concerning similarities between European languages, accomplished multilinguals can point to language families, historical processes of borrowing and the influence of Latin and Greek on the vocabulary of contemporary languages (see Study 4). They also add comments such as: ‘The better you know the language the more similarities you notice,’ or ‘I have noticed that if my Polish vocabulary expands, I understand a wider range of foreign words’, which reveal heightened metalinguistic awareness. Such multilinguals can use the potential offered by cognates to a greater extent, probably is thanks to their higher metalinguistic awareness.

Cognate activation as a product of training My main question to be answered in Studies 3 and 4 was, if such ‘multilingual/plurilingual’ awareness of Polish-English cognates can also be trained in bilingual learners. For instance, Jessner (1999) claims that learners’ metalinguistic awareness can be increased through explicit teaching of similarities between languages. Thus, the next two classroom-oriented studies (Chapter 13) were devoted to examining the effects of training bilingual learners’

Towards Plur ilingual L anguage Teaching w ith English

239

sensitivity to crosslinguistic similarities through the use of awareness-raising tasks. They focused on young adult learners from two extreme ends of the language learning spectrum, namely the total beginners and advanced learners of English. In both cases, the teachers used short exercises activating cognates that supplemented the course material. In the case of the total beginners in Study 3 it turned out that the exercises not only provided them with a number of words and phrases which were easy to learn, but also activated the strategy of reliance on cognate vocabulary from their L1. The students from the experimental group became more ready to take risks in the use of English because they had learned to rely on their crosslinguistic intuitions. This could be noticed to the fullest in the oral tests at the end of the 72 hour course, where the experimental group clearly outperformed the control group students both in terms of the length, lexical variety and in the type-token ratios of their utterances, as calculated at the collocation level. Wheas the control group often gave up answering challenging questions, the experiemental group learners used a range of nouns and adjectives that they had never been taught during the course. The study was conducted in 1997, so they could not have learned the words from the internet or foreign channels of cable TV, as these were hardly available in Poland at that time. All these results, although admittedly obtained via a case study, point to the fact that raising awareness of cognates enhances positive transfer from L1 even in languages as diverse typologically as Polish and English. Awareness-raising activities and strategy training concerning cognate vocabulary, coupled with the teacher’s comments, explicitly built up students’ metalinguistic awareness of cognates, which in turn helped them become more autonomous in their vocabulary use. The results indicate that it is worth teaching about crosslinguistic similarity and sensitising learners to the existence of cognates even at very low L2 levels.

Change in the vocabulary learning strategies as a product of training In the case of the advanced group examined in Study 4, exposure to cognates and activation of English-Polish cognate vocabulary had a different purpose. It aimed to help students from the experimental group appreciate searching for cognateness as a vocabulary learning strategy. Indeed, as reported in questionnaires at the end of the course, the training influenced the advanced students’ vocabulary learning strategies, bringing them closer to those of successful accomplished multilingual learners, which was a clear product of the experiment. The experimental group in Study 4 not only appreciated the role of cognates, but also more often made use of the affordances offered by activating their cognate vocabulary. As compared to the control group, they were better at recognising cognates while reading an English text and used a different vocabulary range, especially adjectives, in the writing and speaking tasks

240

Cognate Vocabul ar y in L anguage Acquisit ion and Use

assigned. Since both: control and experimental groups were presented with the same lexical material, the results indicate that the mere presentation of cognates (not supported by awareness raising) is not enough to affect acknowledging the role of cognates as a learning strategy. The results of the present study possibly indicate that raising awareness of crosslinguistic similarity between typologically different L1 and Ln requires a component of explicit metalinguistic knowledge tapping onto explicit memory processes.

Cognate use in the process of self-regulation Of special interest might be the use of cognates in performing the oral task, which was difficult for most of the students and provoked anxiety. The data reveal that the experimental group students tended to rely on cognate vocabulary at the initial stages of the oral task. Obviously, the use of cognates in speaking was an individual feature. However, since the differences between the experimental and the control groups in the use of cognates in the initial stages of the oral test are significant, we may tentatively conclude that the experimental group students used cognate vocabulary as a self-reguletory strategy for coping with difficulty of the task. Self-regulation, aided by metacognition, is a process where learners develop constructive behaviour that affects their learning. Self-regulated learners are aware of their academic strengths and have a repertoire of strategies they can apply. Thus, when interpreting the behaviour of the experimental group in the oral task we can say to have possibly captured instances of the self-regulation process in action. Self-regulated learners control not only their goal-oriented actions, but also their learning environments. They can exert this control by directing and regulating their own actions toward their goals and by adapting to the changes of the learning environment. As discussed in Chapter 8, the use of affordances relies on the interaction between the organism and its environment. If so, self-regulated language learners possess more affordances. By controlling their language environments and adapting their actions to the learning environment they can better interact with the language they are learning. The self-regulated learner has a better metacognitive understanding of the language, and thanks to this metalinguistic awareness he/she can better exploit the potential of crosslinguistic similarities to direct actions towards success in learning. Concluding, it seems that making learners aware of crosslinguistic similarities helps them to use the potential of cognate affordances and indulge in the process of self-regulation.

The Age Factor in Raising Awareness of Cognates All the studies discussed in the previous sections were carried out with adult learners. Further research explored the possibility of using cognate

Towards Plur ilingual L anguage Teaching w ith English

241

vocabulary as a motivational strategy with learners younger than adults (Chapter 14). Two studies investigated the role of the age factor in raising cognate awareness in the case of 14- and 15-year-olds at the intermediate and elementary level. The lower-secondary learners also differed with respect to the language learning conditions (a small private and a large state school) and with respect to their knowledge of languages beyond L2.

Multilingualism aids awareness of teenagers The results of Study 5 and 6 indicate that cognate awareness-raising tasks can be used for learners younger than adults, whose abilties to recognise crosslinguistic similarity are lower than those of adults (see Chapter 9). In both studies, run in the focus group from a private school and a large group in a state school, the exposure to and activation of English-Polish cognate vocabulary helped teenagers recognise cognates more effectively while reading an English text at the end of the course when compared to their results at the beginning of the course. In Study 5, at the end of the classrom experiment, those learners whose L3 was at least intermediate tended to be more aware of the role of crosslinguistic similarity than their peers without this language knowledge, which again points to the role of their enhanced metalinguistic awareness. To conclude, also in the case of teenage learners multilingualism increases the chance that they will make use of cognate vocabulary when learning.

Crosslinguistic similarity as a motivational strategy The results of Study 6 showed that raising awareness of cognates was also successful when used with teenagers whose proficiency in English was elementary. The pre-test and post-test results of high- and low-achievers were compared with the results of high- and low-achievers from a multilingual reference group of teenage learners of higher English proficiency. The comaprison showed that the activation of English-Polish cognate vocabulary levelled off some of the differences between the groups of high- and lowachieving teenagers from the experiemental and the reference groups. However, Study 6 examined not only the results of awareness-raising but also the attitudes of teenagers and teachers towards the tasks. The method of activating cognate vocabulary was stimulating and motivating for all concerned. Their attitudes were strongly positive: both the teachers and learners claimed that exercises on cognates helped them notice similar words across languages and learn vocabulary faster. To conclude, both the test results and the comments revealed that awareness-raising helps teenagers notice the usefulness of cognate vocabulary. Considering the results, it is absolutely possible and advisable to introduce lower-secondary school learners to the notion of crosslinguistic similarity and to raise their awareness of such similarities.

242

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Implications for Teaching and Syllabus Design: The Way to Plurilingual Education The conclusions from studies presented in this book fit well into language policy devised by the Council of Europe and the European Union, which is centred upon promoting plurilingualism (Council of Europe, 2001; European Commission, 2005). In accordance with the policy of the Council of Europe expressed in Recommendation (98)6, member states should develop learners’ skills in more than one language, increase the number of languages taught in the school systems; encourage flexibility in syllabus design and ensure possibilities for life-long language learning. Such recommendations, coupled with studies on multilingualism, have led to the emergence of concepts of using a methodology that would be different from the monolingual one in order to bring people closer to the goal of individual multilingualism. These approaches aim to teach ‘subsequent foreign languages (. . .) in a manner different from that used for the first foreign language in order to tap the potential already developed through the teaching of the mother tongue and the first foreign language’ (Hufeisen, 2004: 7). The methodology accounts for the fact that L2 and L3-Ln are not learned in isolation from the languages already known to the learner, but make use of the learner’s experience in life and their cognitive abilities. They also take into account learners’ language awareness, knowledge about their learning styles and strategies, and their actual learning experience, including emotional factors (e.g. motivation, attitudes). According to Hufeisen and Neuner (2004), such an approach lays foundations for basic individual plurilingualism. Thus, ‘tertiary language didactics’, as it is called, refers to learners’ previous experiences, using the qualitative difference to advantage in L3-Ln teaching. This leads to pointing to similarities and differences between languages, which can foster the process of language acquisition. According to the concept, when several languages are learned, the learner does not begin at the basic level in each case, but rather continually extends the existing language knowledge. The goal is never to achieve the ideal of ‘near nativeness’ in each of the languages. Instead, the level of competence and the language profile in each language learned can be different. The principles of plurilingual education agree with the CEFR (Council of Europe, 2001) promoting ‘partial competences’, which are limited abilities of some kind within a given language, as well as ‘transversal competences,’ denoting knowledge or skills transferable across languages and used for various purposes. The Council of Europe’s Guide (Beacco & Byram, 2002, 2007) recommends that member states should ensure curricular coordination of language teaching, which involves planning the sequence of introducing languages to the school curriculum and linking their teaching. Moreover, it

Towards Plur ilingual L anguage Teaching w ith English

243

recommends coherence in teaching, which in turn, involves making use of mutual comprehension of related languages. Unfortunately, apart from projects led by universities (e.g. EuroCom), ECML in Graz (e.g. Language Educator Awareness, Ja-Ling: The Gateway to Languages, Plurilingual whole school curricula), or local translanguaging initiatives in the US (e.g. García, 2009) , it is hard to find such crosslinguistic and plurilingual approaches in the language teaching world. This is probably a result of the negative side of globalisation and the myth of the effectiveness of teaching languages as L2 with monolingually prepared materials. Course books, especially those for ELT, are often published to be sold worldwide. Therefore, as Dörnyei and Skehan (2003) emphasise, publishers necessarily disregard individual learning differences, assuming that learners are similar: since differences are anti-egalitarian, it is hard, or ‘unpublishable’ to account for them in a course book. This view may well be extended to more individualised approaches to accounting for learners’ L1 and L2 backgrounds in language teaching. Apart from some most obvious internationalisms (e.g. radio, computer, hamburger) and false friends, crosslinguistic similarities and differences are mostly neglected in language teaching materials. This usually results in the lack of awareness of how to compare languages be for learning purposes. That is why we should postulate introducing awareness raising to the teaching of foreign languages in those cases where formal similarities between languages occur. It is most probably not feasible to fight with the globalised language teaching industry. However, the studies presented in this book point to the possible directions of changes in education. At the national level, language teaching curricula may provide guidelines emphasising the role of crosslinguistic comparisons and language repertoires, while school syllabuses should be supplemented with materials and activities which take into account learners’ L1 background and crosslinguistic similarities with L2, or L3. A good example of such a practice are tasks that raise awareness of cognates described in the previous chapters and in the appendices. As demonstrated, such awareness-raising tasks can be successful not only with adults, whose metalinguistic skills are higher, but also with teenagers at a lower-secondary school level. Importantly, the studies presented here show that multilingual learners with at least intermediate knowledge of L3 tend to be more aware of the role of cross-linguistic similarity, which enhances language learning. Consequently, it does not suffice to teach one foreign language (L2) at school to a higher level. The teaching of L3 should begin in the upper-primary, because it guarantees achieving higher levels when ending secondary education. Ideally, banks of successful crosslinguistic activities for various languages could be published online for free use. Since psychotypological factors are important, tasks should not only focus on typologically close languages (e.g. EuroCom), but also on typologically distant languages, where noticing

244

Cognate Vocabul ar y in L anguage Acquisit ion and Use

cognate vocabulary is more problematic for the learner due to the lesser similarity of crosslinguistic orthographic neighbours.

A Final Comment: Multilingual Constellations with English Much has been said about the dominating role of English, its expansion and the way English jeopardises other languages. Truly, English is by far the most widely taught language in Europe (EURYDICE, 2012). Statistics reveal that at the primary level 60% of learners take English as a foreign language, while at the secondary level, where two languages are taught, nearly 90% of learners take English, while only 20% choose French or German. The predominance of English in Europe is linked to its high prestige: according to Eurobarometer Report 243, 77% of EU citizens believe that children should learn English as their first foreign language, and English is number one in almost all countries polled. Consequently, English in Europe is usually acquired as a second language (L2) or a third language (L3). That is why, most probably, there is no point in fighting with the spread and popularity of the language, but rather it might be better to use English as a bridge between languages in Dominant Language Constellations (LDC). I do not think that in the European context multilingual constellations (DLC) with English endanger one’s identity. Conversely, since English is also the language most widely used for international communication, politics, business, science and higher education, such multilingual constellations help to retain the uniqueness of the original identity and support its further development. Since proficiency in English is of major importance to European citizenship, plurilingualism in Europe should entail a good knowledge of this language. As noted by Breidbach: the teaching of English with the aim of high proficiency can and should be welcomed as long as the individual and collective language rights of the learners are being protected. The lack of plurilingualism, which includes deficient competence in English, might do as much harm as the devaluation of linguistic diversity in terms of democratic participation. (. . .) Sustainable cultural and political inclusion, which can lead to opportunities of participation in multilingual Europe, requires a holistic language education policy inclusive of English and linguistic diversity. (Breibach, 2003: 19–22) Along similar lines, Seidlhofer (2007) claims that languages can and should serve different functions, from official language to community language, from regional to global. She stresses strongly that adapting English to European needs does not mean teaching it and learning it badly with the

Towards Plur ilingual L anguage Teaching w ith English

245

‘anything goes’ attitude. She advocates conscious decisions to assign to English its role of a lingua franca as an important step towards making plurilingualism a more attainable target for every European citizen. As often emphasised throughout this book, English indeed can serve as a ‘bridge’ between various languages in a multilingual constellation of the individual. Due to its vocabulary, borrowed from and into other languages, English bears strong lexical similarities to many European languages. Historically, some borrowings penetrated to English and other languages in a parallel way from Latin, French and German. Some other resulted from various stages of complex contact, i.e. from English through the other languages or from other languages via English. For the last decades, the expansion of English and its privileged cultural and socio-economic position has resulted in extensive lexical borrowing into many languages. Therefore, it is not difficult to notice that many of the internationalisms shared with English are also quite common in formal styles and registers used by educated European speakers in their L1, no matter which major language group this L1 belongs to. Through raising awareness of crosslinguistic similarity learners obtain quick access to numerous cognate lexical items, very similar to their L1 equivalents. Thanks to that they may become more efficient when learning and using languages. Having learned about crosslinguistic similarities between their L1 and English learners may, in turn, become more motivated to learn other languages. All in all, it is well justified to introduce the notion of crosslinguistic similarity to secondary school learners by creating additional activities and tasks supplementing their English language lessons. Through English the activities will draw learners’ attention to crosslinguistic similarities with other languages. Obviously, plurilingual and crosslinguistic awareness of learners depends on their teachers’ knowledge of several languages. Only multilingual teachers advanced in their L3-Ln are able to notice and utilise crosslinguistic similarities and show a positive attitude towards using crosslinguistic similarity in the English language classroom. Thus, probably an obligatory part of teacher training should be the study of more than one language to at least intermediate level. Teacher training should also involve activities helping teachers to notice crosslinguistic similarities in the area of grammar and lexis and use them in the classroom.

246

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Appendix 1 English-Polish identical cognates (Words in the database with NLD = 0) absurd agent alarm album algebra alibi alligator aluminium anagram anorak apartheid arena argument atlas atom badminton banjo bank bar barman baron barter beret bestseller biceps bikini bingo blond boa bonus boom brandy burger bus cargo casting cement chaos character cholera cholesterol collie confetti copyright curry cylinder dealer delta demon

diagram Diesel domino drink dynamo echo element email embargo emigrant fan farmer fauna film fjord flirt flora fort foyer fragment front gang gangster gin gladiator golf gospel graffiti grill guru gyros hamburger happening hardware hit hobby holding hologram horror hotel idea idiom idol igloo import incognito instrument internet jaguar

jazz joystick judo jumbo-jet karate kebab khaki kidnapping koala Koran lady laptop laser leasing leopard lifting limit lobby logo lord lycra major malaria mango mania manicure marina marketing mascara masochism matador medal media melon menu metal minimum minister minus modem monitoring motel motor motto multimedia musical mutant napalm narrator

neon nostalgia nylon ocean opera operator organ ornament palmtop panda panorama paracetamol park parka partner party patchwork patio patrol patron peeling pelican penis penthouse pepperoni pigment pilot ping-pong pizza plan planetarium plaster plus poker pop pop-art popcorn port problem profit program propaganda protest pub punk puzzle quiz radar radio

ranking rap record referendum reggae region remake reporter reprint ring risotto robot rock Rottweiler rugby rum safari saga salami salmonella salon sanatorium sari sauna science-fiction senator sentiment sequel sherry show slalom slang slogan smog snob snowboard software solo sonata spaniel sponsor sport spray sprint squash standard start status stereo

steward student studio super superman supermarket symbol symptom system talent talk-show tampon tango telegram temperament terminal terror test top topless tornado tramp transfer transport trauma trend T-shirt tundra tunnel turban tweed twist ultimatum uniform utopia voucher weekend western whisky wigwam windsurfing yuppie zebra zero zoo

Towards Plur ilingual L anguage Teaching w ith English

247

Appendix 2 Questionnaire 1 (originally in Polish) used in Study 1 and Studies 4–6

Please, choose 1 answer. 1. Are Polish and English closely related? a) yes b) no 2. Which systems of English are the most similar to Polish? a) grammar b) vocabulary c) phonology 3. Which systems of English are the easiest for a Pole to master? a) grammar b) vocabulary c) phonology 4. When is it easier to understand a written text? a) When you know grammar well. b) When you understand vocabulary well. 5. List 5 cognates (i.e. words whose form and meaning are similar in English and Polish). .............................................................. .............................................................. .............................................................. .............................................................. .............................................................. 6. How many such cognate words are there? a) 20–50 b) 50–150 c) 150–500 d) 500–1000

e) 1000–5000

7. How many such words do YOU know? a) ca 10 b) ca 50 c) ca 100 d) ca 500

e) more than 1000

Age below 25 25–40 40–55

above 55

Gender female male English level beginner elementary

upper-intermediate

advanced

Education primary

student

higher

secondary

248

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Appendix 3 Background questionnaire used in Study 2 (fragment). Questions G-M were repeated for L2, L3, L4 and L5. LANGUAGE BACKGROUND QUESTIONNAIRE A. Age: . . ... B. Gender: female/male C. What do you do? (study/work) .................................................................................................................................. Where are you from? (country and city) .................................................................................................................................. This part of the questionnaire concern your knowledge of foreign languages – this information is of crucial importance to my research! Please complete it to the best of your ability. D. What is your native language (L1) ................................................... ? E. How many foreign languages do you know? (give a number) ............. F. List all your languages in order of learning/acquisition. L2 ............................................................ L3 ............................................................ L4 ............................................................ L5 ............................................................ L6 ............................................................ First Foreign Language (L2) Information on proficiency in L2. In the list below, circle the level of your proficiency in the four skills: Writing: A1 A2 B1 B2 C1 C2 Listening: A1 A2 B1 B2 C1 C2 Reading: A1 A2 B1 B2 C1 C2 Speaking: A1 A2 B1 B2 C1 C2 GL2. How long have you been learning this language? .................................................................................................................................. HL2. Where did you learn this language (e.g. public school, language school, private lessons) ..................................................................................................................................

Towards Plur ilingual L anguage Teaching w ith English

249

I L2. Do you use this language in everyday life? In what situations? (e.g. work, university, school, talking with friends or family, etc.) .................................................................................................................................. JL2. Do you like to speak this language? If yes, than in what situations and why? .................................................................................................................................. KL2. Are you anxious (or stressed) while speaking this language? If yes, then in what situations and why? .................................................................................................................................. LL2. What do you find easy about this language? (e.g. vocabulary, grammar, pronunciation) .................................................................................................................................. ML2. What do you find difficult about this language? (e.g. vocabulary, grammar, pronunciation) .................................................................................................................................. Thank you for filling in the questionnaire

250

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Appendix 4 Lists of stimuli for Study 2 COGNATE Nouns English

Polish

English frequency per million

LD raw

NLD

antagonism stereotype procedure nonsense infrastructure impulse scandal electorate preference strategy bureaucracy eccentricity ethics explosion crisis assertiveness authenticity loyalty sovereignty privacy

antagonizm stereotyp procedura nonsens infrastruktura impuls skandal elektorat preferncje strategia biurokracja ekscentrycznos´c´ etyka eksplozja kryzys asertywnos´c´ autentycznos´c´ lojalnos´c´ suwerennos´c´ prywatnos´c´

0.21 0.70 16.13 28.47 1.17 5.27 8.11 0.17 2.33 9.45 0.50 0.19 5.35 16.49 16.64 0.17 1.15 11.66 0.56 15.17

1 1 1 1 2 1 1 2 2 2 7 10 4 6 4 9 9 7 9 9

0.1 0.1 0.111 0.125 0.143 0.143 0.143 0.2 0.2 0.222 0.636 0.666 0.666 0.666 0.666 0.692 0.692 0.777 0.818 0.9

English

Polish

English frequency per million

LD raw

NLD

planetary parliamentary arbitrary modular federal popular

planetarny parlamentarny arbitralny modularny federalny popularny

1.07 0.23 1.01 0.156 23.00 23.07

10 2 2 2 2 2

0.1 0.153 0.2 0.222 0.222 0.222

Cognate Adjectives

Towards Plur ilingual L anguage Teaching w ith English

sentimental brutal stoic aesthetic complicated qualified ethical paralysing fascinating critical registered paralysing mythical pagan

sentymentalny brutalny stoicki estetyczny skomplikowany wykwalifikowany etyczny paraliz˙uja˛cy fascynuja˛cy krytyczny zarejestrowany paraliz˙uja˛cy mityczny pogan´ski

FALSE FRIENDS nouns

7.03 7.35 0.41 0.66 27.15 8.25 2.72 0.098 15.07 12.25 10.13 0.09 0.88 1.35

3 2 2 6 8 10 5 9 7 6 10 9 6 6

251

0.23 0.25 0.285 0.6 0.615 0.666 0.714 0.75 0.636 0.666 0.71 0.75 0.75 0.75

FALSE FRIENDS adjectives

English

Polish

English English frequency per million

revision mandate accord manifestation oppression pension prospect hazard testament diversion expertise lecture conduct obligation audition rent revenge smoking direction data

powtórka 0.91 upowaz˙nienie 0.74 porozumienie 0.81 przejaw 1.04 ucisk 1.49 emerytura 1.61 szansa 1.75 ryzyko 1.93 s´wiadectwo 2.35 objazd 3.1 znawstwo 3.98 wykład 5.24 zachowanie 5.55 zobowia˛zanie 6.75 przesłuchanie 6.91 czynsz 8.64 zemsta 9.52 palenie 13.08 kierunek 24.08 dane 25.61

consequent genial eventual suggestive obscure transparent sympathetic complex fatal prominent elaborate hazardous hysterical invalid premier corresponding civil actual ordinary pathetic

Polish

English frequency per million

nastepuja˛cy przyjazny ostateczny dwuznaczny niejasny przezroczysty współczuja˛cy złoz˙ony s´miertelny wyrazisty skomplikowany niebespieczny bardzo s´mieszny niewaz˙ny najwaz˙niejszy zgodny uprzejmy faktyczny zwyczajny z˙ałosny

0.1 0.2 0.43 0.51 1.29 1.8 3.8 6.56 7.1 2.55 2.38 1.84 7.1 0.43 1.25 0.28 15.94 17.33 19.08 24.2

252

Cognate Vocabul ar y in L anguage Acquisit ion and Use

CONTROL nouns

CONTROL adjectives

English

Polish

English English frequency per million

Polish

English frequency per million

achievement challenge complaint contribution custom disability thrill earnings eyesight frontier embarrassment income influence judgement minority redundancy friendliness sacrifice tension excursion

osia˛gnie˛cie wyzwanie zaz˙alenie wkład zwyczaj niepełnosprawnos´c´ podnieta zarobki wzrok granica zawstydzenie

3.49 9.93 10.16 4.37 6.2 1.98 0.06 0.94 1.94 3.27 4.84

amazed annoying crowded dreadful exhausted flexible gorgeous parental polite reliable sincere

zdumiony denerwyja˛cy zatłoczony okropny wyczerpany elastyczny cudowny rodzicielski grzeczny godny zaufania szczery

2.56 5.56 2.98 8 6.42 2.67 24.06 2.06 13.94 5.9 5.98

dochód wpływ osa˛d mniejszos´c´ nadmiar przyjacielskos´c´ pos´wie˛cenie napie˛cie wycieczka

7.41 7.39 16.45 2.22 0.63 0.12 8.95 8.55 0.67

sparkling stunning suitable superb superficial talkative unforgetteable unpaid urgent

błyszcza˛cy niesamowity odpowiedni najlepszy powierzchowny rozmowny niezapomniany niepłatny pilny

1 2.05 4.84 2.88 2.9 0.9 0.94 1 14.86

Towards Plur ilingual L anguage Teaching w ith English

253

Appendix 5 Examples of exercises for raising awareness of cognate vocabulary used in Studies 3–6 (illustrations by A. Fijałkowska) SAMPLE EXERCISE 1 Divide these features into the positive and negative ones. Then work in a pair and see if your partner divided them in the same way. Podziel poniz˙sze cechy na pozytywne i negatywne. Potem pracuja˛c w parze sprawdz´, czy twój partner podzielił je tak samo. Example: A: I think ’active’ is a positive word/feature. B: I think so too./I don’t think so. active, delicate, nervous, romantic, critical, ignorant, ambitious, natural, passive, elegant, tolerant, punctual, attractive, emotional, liberal, extravagant, intelligent, creative, optimistic, fanatical, arrogant, pessimistic, assertive, consequent

SAMPLE EXERCISE 2 Who wears these clothes at work? Write what you think and then tell your partner. Kto nosi te ubrania w pracy? Napisz, co mys´lisz, a potem powiedz swojemu koledze. Example: I think a bus driver wears a T-shirt and jeans. Jobs a professional pilot, an effective businessman, a discreet stewardess, a disorganized archaeologist, a competent doctor, a depressed cashier, an extravagant computer programmer, an extroverted actor, a frustrated geologist, a blond secretary, a delicate ballerina, a policeman, a corrupt politician, an enthusiastic skiing instructor, an emancipated baby sitter, an eccentric fashion designer Clothes a T-shirt a bikini

254

Cognate Vocabul ar y in L anguage Acquisit ion and Use

an elegant uniform a mini skirt a tweed jacket and a tie a mohair sweater a mask a thermo-active anorak leggings with lycra authentic Levi’s jeans khaki shorts and sandals a cashmere shawl SAMPLE EXERCISE 3 Part A Work in pairs. Write what you think about the girl in the picture. Pracujcie w parach. Napiszcie, co mys´licie o dziewczynie na obrazku. 1. What’s her name?.............................................. 2. How old is she?.................................................. 3. What does she do?............................................. 4. What does she like to do?................................. 5. What does she like to eat and drink?............. 6. What is she like?................................................

Part B The girl in the picture wants to become a secretary. Each of the three men can become her boss. Choose the words to describe each of them. You can add other adjectives. Dziewczyna z obrazka chce zostać sekretarka˛. Kaz˙dy z trzech me˛˙zczyzn moz˙e zostać jej szefem. Wybierz słowa najlepiej opisuja˛ce kaz˙dego z nich. Moz˙esz dodać inne przymiotniki. active, ambitious, attractive, creative, critical, cynical, delicate, elegant, emotional, extravagant, intelligent, ironical, liberal, natural, nervous, optimistic, old, original, passive, young, professional, practical, romantic, sarcastic, tolerant, traditional

Towards Plur ilingual L anguage Teaching w ith English

255

Mr. A Mr. B Mr. C

Part C Choose the best boss for the girl in Part A. Then discuss your choice with a partner. Use the words from Part B. Wybierz najodpowiedniejszego szefa dla dziewczyny z Cze˛´sci A. Przedyskutuj swój wybór z kolega˛, uz˙ywaja˛c słów z Cze˛s´ci B. Example: A: I think Mr. A is the best He is intelligent and she is intelligent, too. B: But I think he is too . . . . for her. Mr. B is the best A: But he is not . . . enough for, her and I think. . .

256

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Appendix 6 Translation exercise used in Study 3 (Original instruction in Polish) Translate the following sentences as fast as you can. Do not use a dictionary. 1. I like TV: I watch quizzes, talk-shows, horrors, thrillers and other films. 2. This is a reprint of his 1920 bestseller. 3. John is marketing manager of a consulting firm. 4. He’s got a personal assistant and uses a pager and a notebook. 5. He is a typical yuppie: a total workaholic and not a playboy. 6. At weekends he occasionally goes to a fitness club or goes windsurfing to relax. 7. Promotion! We offer new generation hypoallergenic cosmetics to create your new image. 8. Peeling and lifting - top standard only. 9. I have an immobiliser in my new jeep. 10. I’ve got a lot of cassettes and compact discs with hard rock and metal music. 11. I sent him the information by email. 12. Feminists don’t read love-stories: they are professionals interested in business and money. 13. My favourite sports are football, basketball and snowboarding. 14. The Polish team was sponsored by one of the Peugeot dealers. 15. Amway products are promoted by regional representatives in network marketing

Towards Plur ilingual L anguage Teaching w ith English

257

Appendix 7 Fragments of structured oral tests with total beginners after 72 hours of teaching, Study 3 E1 - Experimental group student, I – examiner To save space, examiner’s phatic utterances (I: Yhm.) have been deleted Interview 1 in the Experimental Group I: E1: I: E1: E1: E1: I: E1: I: E1: I: E1: I: E1: E1: I: E1: I: E1: I:

E1: E1:

I’d like you to have a look here. I’ve got, I’ve got a photo for you. OK.? I would like you to tell me everything about this photo now, OK.? [showing a picture of a man] yhm . . . yyy . . . he . . . isn’t young, but he isn’t old. yyy . . . I think . . . he has . . . forty-five . . . no he has. He forty-five old. OK. He is . . . he’s intelligent . . . confident confident . . . and creative yyy . . . he have family . . . yyy he’s work . . . is he’s work in . . . office, or theatre, or cinema, or film. yyy . . . he is . . .. Nie wiem. [I don’t know] [laughs] Well what about his personality, perhaps? Maybe he is . . . nice [laughs] [laughs] OK. An have a look at this girl. [showing picture of a woman] . . . He is Japanese She. . . She is Japanese . . . no she is from Singapore I see [laughs]. And? yyy She is beautiful . . . young and very nice. He is happy. . . . She is happy. . . . She is a student. yhm. Personality perhaps? . . . He is . . . she is . . . shy. [laughs] OK. [laughs] Right. And now. Perhaps the last question, OK? Do you remember this lesson? [showing a page from the course book] Yes, OK. Here some people were talking about the English. ‘What do you like about the English?’ OK? And now. I want you to tell me: what do you think about Polish people. OK? What do you like? What don’t you like? OK. And about Poland now. OK? Yeah. eee . . . I I like yyy . . . I like . . . For me . . .yyy impossible yyy . . . talk yyy generally . . . of people in Poland, or people in England. Or people in country . . . yyy perha. . . no, because, because yyy people, people is . . . no people are yyy. . . very different . . . in all country

258

E1:

E1: E1: E1: E1: E1: I: E1: E1: E1:

Cognate Vocabul ar y in L anguage Acquisit ion and Use

yyy For example . . . I don’t yyy speak all yyy . . . people in Poland is . . . nice. Because . . . not all people . . . in Poland is nice. yyy . . . yyy for example, I don’t . . . talk . . . yyy then people in Poland is . . . ti.. tidy . . . or intidy. Because non all people . . . have . . . done . . . things. . . . yyy people in Poland yyy is very different yyy are interes. . . interesting people and . . . boring people, yyy gentlemen . . .or hooligan. They. . . tidy or not tidy . . . yyy For me in Poland generally . . . yyy but is . . . it is not true for all . . . all people in Poland. People is . . . not to. . . People isn’t tolerant . . . tolerant. yyy People is . . . rasizm? racism, people is . . .. anti-Semitism, yyy For example, in street . . . impossible is . . . it is . . .yyy go . . .yyy with clothes, clothes . . . different, strange. yyy because all people watch. OK. Yes! . . . All people . . . yyy when when . . .. yyy it . . . yyy . . .. All the people in Poland talk of politics politician . . . Only all . . . Polish? Polish yyy think . . . yyy then know . . . all . . . of politics, of situation . . . economic, of yyy neighbour country, of . . . war in Iraq

C1 - Control group student, I – examiner To save space, examiner’s phatic utterances (I: Yhm.) have been deleted Interview 1 in the Control Group I: C1: C1: C1: C1: C1: C1: I: C1: C1: I:

Now, can you have a look at this picture? OK? Can you tell me something about this man in the picture? [showing a picture of a man] yyy This man this is old I . . . is prety? yyy he is . . . he are yyy . . . brown eyes . . . yyy small nose, no, large nose [laughs] yyy small mouth and brown . . . black and white hair yyy she is rich . . . rich yyy watch is very expensive Aha, OK. What do you think about this man? About his personality perhaps? Character? yyy I . . . I think yyy this man is . . . kind . . . and I think . . .. and he . . . talking . . . talking yyy in a. . . this I think yyy he is pewny siebie [self-confident] . . . w kaz˙dym razie [anyway] I think yyy cos´ zwia˛zanego z moda˛ [something connected with fashion]. Aha. OK. [laughs] And now have a look at this girl. [showing a picture of a girl]

Towards Plur ilingual L anguage Teaching w ith English

C1: C1: C1: I: C1: C1: I:

C1: I: C1: I: C1: I: C1: I: C1: I.

C1: I: C1: C1: I: C1: C1: C1: C1: C1: I: C1: C1: I:

259

This girl is very pretty. yyy Is young yyy This is from Japan. . . . Japan, Vietnam. . . . yyy is young . . .. no nie wiem co powiedzieć. What do you . . . What do you think about her? Her personality perhaps? She . . . yyy teacher. She is a teacher. yyy is a model. . . OK, OK. Right. yhm Thank you very much, and just last question OK? Here, in your book [showing a page from the course book] . . . eem you had a chapter about what people, what other people think about the English. yhm OK? And now, I want you to . . . you remember this thing? What do you think about the English? No problem. Right? And now I want to ask you. What do you think about Polish people, OK? And Poland in the nineteen nineties. Right? Poland now in the last. . . Problems? Well, whatever you want to say, want to tell me about Poland and the Polish people. yyy It can be about problems. Like . . . English. . . money.. and house yyy. . .yyy To było: co inni mys´la˛ anglikach, tak. Co ty mys´lisz o Polakach i o Polsce w latach dziewie˛ćdziesia˛tych. [this is what you had: what others think about the English. What do you think about Poles and Poland in the nineteen nineties] Ale z perspektywy swojej, tak?[But from my perspective, yes?] Yes, yhm. I think then Polen yyy are . . . yyy Polish. . .? Yes, Polish. Polish often watch TV. yyy They are yyy they they . . . like money. yyy . . . yyy . . . yyy haven’t hobbies yyy she . . . not not go not go . . . not go the yyy theater [pronounced /tiater/] yyy Theatre? Theatre yyy cinema yyy . . .yyy Wszystko [That’s all] OK. Thank you.

260

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Appendix 8 Vocabulary learning strategies questionnaire used in Study 4 (based on Chamot 1987:77). VOCABULARY LEARNING STRATEGIES There are various strategies for learning and activating vocabulary. Which of these do YOU find the most useful? Please choose up to seven. • • • • • • • • • • • • • •

Contextualization – using a new word in a meaningful context (e.g. when working with a partner) Grouping – ordering, classifying, labelling groups of words (on handouts or small pieces of paper) Transfer – looking for similarities between English and Polish words (or basing on words whose form and meaning are similar in other languages you know) Directed Physical Response – relating words to physical actions, performing Translation – using Polish to understand an English word (translating lists of words) Auditory representation – remembering a sound you associate with the word Note-taking – writing information down Deduction – applying rules (e.g. morphological) to understand English words Recombination – combining words to construct sentences and paragraphs Imagery – relating new words to visual concepts, (e.g. using photographs and pictures) Inferencing – guessing the meaning of a new word from the context Elaboration – relating new words to other concepts in memory (e.g. using word maps) Repetition – imitating a language model Resourcing – defining or expanding a definition through the use of English books

Do you use any other strategies?

Towards Plur ilingual L anguage Teaching w ith English

261

Appendix 9 Questionnaire 2 used in Study 5 and 6 (originally in Polish) QUESTIONNAIRE 2 Over the last few weeks you have been doing exercises which activate the vocabulary that is similar in Polish and English. Try to evaluate how useful are such exercises. You can choose more than one answer.  1 The exercises helped me learn vocabulary faster  2 The exercises helped me notice that there are similar words in both languages.  3 Earlier, I did not notice such words.  4 The exercises did not change my way of learning vocabulary.  5 I have not noticed any specific exercises dealing with cognates.  6 Now, I try to use cognates more often when writing in English.  7 Now, I try to use cognates more often when speaking English.  8 Now, I notice cognates more often when reading in English.  9 I stopped checking the meaning of the words which look similar to Polish.  10 I do not see the point in introducing such exercises. My comments:

262

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Appendix 10 Teacher’s questionnaire used in Study 5 and 6 (originally in Polish) TEACHER’S QUESTIONNAIRE Over the last few weeks you have been asked to introduce exercises which activate the vocabulary that is similar in Polish and English. Try to evaluate how useful are such exercises for your learners. You can choose more than one answer. 1 I did not notice such words earlier. 2 I used this vocabulary teaching method earlier. 3 I liked the method and I am willing to use it in the future. 4 I did not like the method. I do not see the point in introducing such exercises. 5 My students liked the method. 6 The method was motivating for my students. 7 My students did not like the method. 8 Throughout the whole period my students thought the exercises were boring and useless. 9 At the beginning, my students thought the exercises were boring, but later enjoyed doing them. 10 The activities did not change much in my students’ way of learning vocabulary. 11 The exercises helped my students notice that there are similar words in both languages. 12 The exercises helped my students learn vocabulary faster. 13 My students tried to use cognates more often when writing in English. 14 My students tried to use cognates more often when speaking English. 15 My students tried to use cognates more often when reading in English. 16 My students asked fewer questions about the meaning of the words which look similar to Polish. My comments:

References

Akulenko, V.V. (1958) Ob internacionalnych slovach v sovremennom russkom jazyke. [About international words in the sovereign Russian language]. Charkov: Izdatel’stvo Charkovskogo Universiteta. Al-Hejin, B. (2004) Attention and awareness: Evidence from cognitive and second language research. Teachers College, Columbia University Working Papers in TESOL and Applied Linguistics 4 (1), 1–22. Andersen, R. (1983) Transfer to somewhere. In S. Gass and L. Selinker (eds) Language Transfer in Language Learning (pp. 177–201). Rowley, MA: Newbury House. Anderson, J.R. (1983) The Architecture of Cognition. Cambridge, MA: Harvard University Press. Anderson, J.R. and Lebiere, C. (1998) The Atomic Components of Thought. Mahwah, NJ: Lawrence Erlbaum Associates. Ard, J. and Homburg, T. (1983) Verification of language transfer. In S. Gass and L. Selinker (eds) Language Transfer in Language Learning (pp. 157–176). Rowley, MA: Newbury House. Armon-Lotem, S. (2012) Introduction: Bilingual children with SLI − the nature of the problem. Bilingualism: Language and Cognition 15 (1), 1–4. Aronin, L. (2005) Theoretical perspectives of trilingual education. International Journal of the Sociology of Language 171, 7–22. Aronin, L. (2006) Dominant Language Constellations: An approach to multilingualism studies. In M. Ó Laoire (ed.) Multilingualism in Educational Settings (pp. 140–159). Schneider Verlag Hohengehren. Aronin, L. and Ó Laoire, M. (2004) Exploring multilingualism in cultural contexts: Toward a notion of multilinguality. In I.C. Hoffmann and J. Ytsma (eds) Trilingualism in Family, School and Community (pp. 11–29). Clevedon: Multilingual Matters. Aronin, L. and Hufeisen, B. (2009) The Exploration of Multilingualism. Amsterdam/ Philadelphia: John Benjamins Publishing Company. Aronin, L. and Singleton, D. (2008) Multilingualism as a new linguistic dispensation. International Journal of Multilingualism 5 (1), 1–16. Aronin, L. and Singleton, D. (2012a) Multilingualism. Amsterdam: John Benjamins Publishing Company. Aronin, L. and Singleton, D. (2012b) Affordances theory in multilingualism studies. Studies in Second Language Learning and Teaching (3), 311–331. Atkinson, D. (1999) TESOL and culture. TESOL Quarterly 33 (4), 625–654. Baetens Beardsmore, H. (1986) Bilingualism: Basic Principles. Clevedon: Multilingual Matters. Baker, C. (1988) Key Issues in Bilingualism and Bilingual Education. Clevedon: Multilingual Matters. Banta, F. (1981) Teaching German vocabulary: The use of English cognates and common loan words. Modern Language Journal 65, 129–136. Bardel, C. and Falk, Y. (2007) The role of the second language in third language acquisition: The case of Germanic syntax. Second Language Research 23, 459–84. 263

264

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Bardel, C. and Falk, Y. (2012) Procedural distinction. In J.C Amaro, S. Flynn and J. Rothman (eds) Third Language Acquisition in Adulthood (pp. 61–78). Amsterdam: John Benjamins Publishing. Barwise, J. and Perry, J. (1983) Situations and Attitudes. Cambridge: MIT-Bradford. Beacco, J.-C. and Byram, M. (2002) Guide for the Development of Language Education Policies in Europe: From Linguistic Diversity to Plurilingual Education. Strasbourg: Council of Europe. Beacco, J.-C. and Byram, M. (2007) Data and methods for the development of language education policies. Part 2 of the Guide for the Development of Language Education Policies in Europe. Strasbourg: Council of Europe. Beijering, K., Gooskens, C. and Heeringa, W. (2008) Predicting intelligibility and perceived linguistic distances by means of the Levenshtein algorithm. Linguistics in the Netherlands 15, 13–24. Ben-Zeev, S. (1977) Mechanisms by which childhood bilingualism affects understanding of language and cognitive structures. In P.A. Hornby (ed.) Bilingualism: Psychological, Social, and Educational Implications (pp. 29–55). New York: Academic. Berndt, R.S., Reggia, J.A and Mitchum, C.C. (1987) Empirically derived probabilities for grapheme-to-phoneme correspondences in English. Behavior Research Methods. Instruments and Computers 19 (1) 1–9. Berthele, R. (2011) On abduction in receptive multilingualism. Evidence from cognate guessing tasks. Applied Linguistics Review (2), 191–220. Bialystok, E. (1987) Influences of bilingualism on metalinguistic development. Second Language Research 3, 154–166. Bialystok, E. (1988) Levels of bilingualism and levels of linguistic awareness. Developmental Psychology 24, 560–567. Bialystok, E. (1991) Metalinguistic dimensions of bilingual language proficiency. In E. Bialystok (ed.) Language Processing in Bilingual Children (pp. 113–140) Cambridge: Cambridge University Press. Bialystok, E. (2001) Bilingualism in Development, Language, Literacy and Cognition. Cambridge: Cambridge University Press. Bialystok, E., Craik, F.I.M. and Freedman, M. (2007) Bilingualism as a protection against the onset of symptoms of dementia. Neuropsychologia 45, 459–464. Bialystok, E., Craik, F.I.M., Green, D.W. and Gollan, T.H. (2009) Bilingual minds. Psychological Science in the Public Interest 10 (3), 89–129. Biedron´, A (2008) Teoria s´wiadomego postrzegania (noticing) Intencja i s´wiadomos´ć w akwizycji drugiego je˛zyka. In M. Pawlak (ed.) Autonomia w nauce je˛zyka obcego–co osia˛gne˛lis´my i doka˛d zmierzamy? Wydział Pedagogiczno-Artystyczny UAM w Poznaniu and Pan´stwowa Wyz˙sza Szkoła Zawodowa w Koninie. Biedron´ , A. and Szczepaniak, A. (2012) Working memory and short-term memory abilities in accomplished multilinguals. The Modern Language Journal 96 (2), 290–306. Bley-Vroman, R. (1983) The comparative fallacy in interlanguage studies: The case of systematicity. Language Learning 33, 1–17. Bloomfield, L. (1933) Language. New York: Holt. Bobb, S.C. and Wodniecka, Z. (2013) Language switching in picture naming: What asymmetric switch costs (do not) tell us about inhibition in bilingual speech planning. Journal of Cognitive Psychology 25 (5), 568–585. Braun, P. (1985) Internationalisms – Identical Vocabularies in European Languages. Paper presented at a Conference on Vernacular Languages for Modern Societies. See http:// www.eric.ed.gov (accessed May 2011). Braun, P. (1986) Die Deutsche Sprache in Europäischen Vergleich. Muttersprache 96, 330–344. Braun, P. (1989) Internationalisms: Identical vocabularies in European languages. In F. Coulmas (ed.) Language Adaptation (pp. 158–167). Cambridge: Cambridge University Press.

References

265

Braun, P., Schaeder, B. and Volmert, J. (1990) (eds) Internationalismen. Studien zur interlingualen Lexikologie und Lexikographie. Tübingen. Breidbach, S. (2003) Plurilingualism, Democratic Citizenship in Europe and the Role of English. University of Bremen Language Policy Division. Strasbourg: Council of Europe. Brenders, P., Van Hell, J.G. and Dijkstra, A. (2011) Word recognition in child second language learners: Evidence from cognates and false friends. Journal of Experimental Child Psychology 109, 383–396. Brohy, C. (2001) Generic and/or specific advantages of bilingualism in a dynamic plurilingual situation: The case of French as official L3 in the school of Samedan (Switzerland). International Journal of Bilingual Education and Bilingualism 4 (1), 38–49. Brutt-Griffler, J. (2002) World English. A Study of its Development. Clevedon: Multilingual Matters. Buttler, D. (1981) Tendencje rozwojowe w zasobie słownym powojennej polszczyzny. In H. Kurkowska (ed.) Współczesna polszczyzna. Wybór zagadnien´, Warszawa: PWN. Byram, M. and Feng, A.W. (2004) Culture and language learning: Teaching, research and scholarship. Language Teaching 37, 149–168. Cenoz, J. (2000) Research on multilingual acquisition. In J. Cenoz and U. Jessner (eds) English in Europe. The Acquisition of a Third Language (pp. 39–53). Clevedon: Multilingual Matters. Cenoz, J. (2001) The effect of linguistic distance, L2 status and age on cross-linguistic influence in third language acquisition. In J. Cenoz, B. Hufeisen and U. Jessner (eds) Cross-linguistic Influence in Third Language Acquisition: Psycholinguistic Perspectives (pp. 8–20). Clevedon: Multilingual Matters. Cenoz, J. (2003) The intercultural style hypothesis: L1 and L2 interaction in requesting behaviour. In V. Cook (ed.) Effects of the Second Language on the First (pp. 62–80). Clevedon: Multilingual Matters. Cenoz, J. and Valencia, J. (1994) Additive trilingualism: Evidence from the Basque Country. Applied Psycholinguistics 15, 197–209. Cenoz, J. and Genesee, F. (1998) Beyond Bilingualism: Multilingualism and Multilingual Education. Clevedon: Multilingual Matters. Cenoz, J. and Genesse, F. (1998) Psycholinguistic perspectives on multilingualism and multilingual education. In J. Cenoz and F. Genesse (eds), Beyond Bilingualism: Multilingualism and Multilingual Education (pp. 16–32). Clevedon: Multilingual Matters. Cenoz, J. and Jessner, U. (eds) (2000) English in Europe. The Acquisition of a Third Language. Clevedon: Multilingual Matters. Cenoz, J., Hufeisen, B. and Jessner, U. (eds) (2001) Crosslinguistic Influence in Third Language Acquisition: Psycholinguistic Perspectives. Clevedon: Multilingual Matters. Chamot, A.U. (1987) The learning strategies of ESL students. In A. Wenden and J. Rubin (eds) Learner Strategies in Language Learning (pp. 71–83). New Jersey: Prentice Hall. Chomsky N. (1959) Review of Skinner’s Verbal Behavior. Language 35, 26–58. Chomsky, N. (1965) Aspects of the Theory of Syntax. Cambridge, MA: MIT Press. Chomsky, N. (1986) Knowledge of Language: Its Nature, Origin and Use. New York: Praeger. Christ, H. (2001) Wie das Postulat der Erziehung zur Mehrsprachigkeit den Fremdsprachenunterricht insgesamt verändert. Fachverband Moderne Fremdsprachen, Landesverband Niedersachsen, Mitteilungsblatt 2, 2–9. Cies´licka, A. (2000) The effect of language proficiency and L2 vocabulary learning strategies on patterns of bilingual lexical processing. Poznan Studies in Contemporary Linguistics 36, 27–53. Cies´licka, A. and Ekert, A. (2009) Investigating asymmetry in lexical processing with proficient Polish-English bilinguals: The lexical decision task. In J. Arabski and A. Wojtaszek (eds) Multidisciplinary Perspectives on Second Language Acquisition and Foreign Language Learning (pp. 241–253). Katowice: Wydawnictwo Uniwersytetu S´la˛skiego, Oficyna Wydawnicza WW.

266

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Cohen, L., Manion, L. Morrison, K. (2007) Research Methods in Education (6th edn). London: Routledge. Commission of the European Communities (2008) COM 566 Final – Multilingualism: An Asset for Europe and a Shared Commitment. Brussels. Comrie, B. (1981) Typology and universals. In B. Comrie (ed.) Language Universals and Linguistic Typology (pp. 30–35). Chicago: University of Chicago Press. Cook, V. (1991) The poverty-of-the-stimulus argument and multicompetence. Second Language Research 7 2, 103–117. Cook, V. (1992) Evidence for multi-competence. Language Learning 42 (4), 557–591. Cook, V. (1997) Monolingual bias in second language acquisition research. Revista Canaria de Estudios Ingleses 34, 35–50. Cook, V. (1999) Going beyond the native speaker in language teaching. TESOL Quarterly 33 (2), 185-209. Cook, V.J. (2002) Background to the L2 user. In V. Cook (ed.) Portraits of the L2 User (pp. 1–27). Clevedon: Multilingual Matters. Cook, V. (2003) (ed.) Effects of the Second Language on the First. Clevedon: Multilingual Matters. Cook, V. (2007) The goals of ELT: Reproducing native-speakers or promoting multicompetence among second language users? In J. Cummins and C. Davison (eds) International Handbook of English Language Teaching (pp. 237–248). New York: Springer. Cook, V.J., Iarossi, E. Stellakis, N. and Tokumaru, Y. (2003) Effects of the second language on the syntactic processing of the first language. In V. Cook (ed.) Effects of the Second Language on the First (pp. 193–213). Clevedon: Multilingual Matters. Costa, A., Caramazza, A. and Sebastian-Galles, N. (2000) The cognate facilitation effect: Implications for models of lexical access. Journal of Experimental Psychology: Learning, Memory, and Cognition 26 (5), 1283–1296. Costa, A., Santesteban, M. and Ivanova, I. (2006) How do highly-proficient bilinguals control their lexicalization process? Inhibitory and language-specific selection mechanisms are both functional. Journal of Experimental Psychology: Learning, Memory, and Cognition 32, 1057–1074. Council of Europe (2001) Common European Framework of Reference for Languages: Learning, Teaching, Assessment. Cambridge: Cambridge University Press. Council of Europe (2007) From Linguistic Diversity to Plurilingual Education: Guide for the Development of Language Policies in Europe. Strasbourg: Language Policy Division. Crawley, A. (1995) Oxford Elementary Learner’s Dictionary. Oxford: Oxford University Press. Crystal, D. (1997a) The Cambridge Encyclopedia of Language (2nd edn). Cambridge: Cambridge University Press. Crystal, D. (1997b) English as a Global Language. Cambridge: Cambridge University Press. Crystal, D. (2001) A language revolution: From reluctance to renaissance. See http:// www.davidcrystal.com/DC_articles/Langdeath8.pdf (accessed March 2010). Crystal, D. (2003) A Dictionary of Linguistics and Phonetics. Oxford: Blackwell. Cummins, J. (ed.) (2000) Language, Power, and Pedagogy: Bilingual Children in the Crossfire. Clevedon: Multilingual Matters. Cummins, J. and Davison, C. (eds) (2007) International Handbook of English Language Teaching. New York: Springer. Cummins, J. and Davison, C. (2007) The global scope and politics of ELT: Critiquing current policies and programs. In J. Cummins and C. Davison (eds) International Handbook of English Language Teaching (pp. 3–11). New York: Springer. Cunningham, T.H. and Graham, C.R. (2000) Increasing native English vocabulary recognition through Spanish immersion: Cognate transfer from foreign to first language. Journal of Educational Psychology 92 (1) 37–49.

References

267

Davies, N. (2005) God’s Playground: A History of Poland in Two Volumes (Vol. 2). Oxford: Oxford University Press. De Angelis, G. (2005) Interlanguage transfer of function words. Language Learning 55, 379–414 De Angelis, G. (2007) Third or Additional Language Acquisition. Clevedon: Multilingual Matters. De Angelis, G. and Selinker, L. (2001) Interlanguage transfer and competing linguistic systems. In J. Cenoz, B. Hufeisen and U. Jessner (eds) Crosslinguistic Influence in Third Language Acquisition: Psycholinguistic Perspectives (pp. 42–58). Clevedon: Multilingual Matters. de Bot, K. (2004) The multilingual lexicon: Modelling selection and control. International Journal of Multilingualism 1 (1), 17–32. de Bot, K. and Kroll, J. (2010) Psycholinguistics. In N. Schmitt (ed.) An Introduction to Applied Linguistics (pp. 124–142). London: Hodder Education. De Courtenay, J.N.B. (1911) W sprawie «antysemityzmu postepowego» [On so called ‘progressive antisemitism’]. Kraków: Skład Główny w Ksia˛garni G. Gebethnera i Spółki. See http://www.newworldencyclopedia.org (accessed December 2010). De Groot, A.M.B. (1992) Determinants of word translation. Journal of Experimental Psychology: Learning, Memory, and Cognition 18 (5), 1001–1018. De Groot, A.M.B. (1993) Word type effect in bilingual processing tasks: Support for a mixed representational system. In R. Schreuder and B. Weltens (eds) The Bilingual Lexicon (pp. 27–51). Amsterdam: John Benjamins Publishing Co. De Groot, A.M.B. and Comijs, H. (1995) Translation recognition and translation production: Comparing a new and an old tool in the study of bilingualism. Language Learning 45 (3), 467–509. De Groot, A.M.B. and Keijzer, R. (2000) What is hard to learn is easy to forget: The roles of word concreteness, cognate status, and word frequency in foreign language vocabulary learning and forgetting. Language Learning 50, 1–56. De Groot, A.M.B. and Van Hell, J. (2005) The learning of foreign language vocabulary. In J.F. Kroll and A.M.B. De Groot (eds) Handbook of Bilingulism. Psycholinguistic Approaches (pp. 9–29). Oxford: Oxford University Press. De Groot, A.M.B., Dannenburg, L. and Hell, J. (1994) Forward and backward translation by bilinguals. Journal of Memory and Language 33, 600–629. Denison, D. and Hogg, R. (2006) Overview. In R. Hogg and D. Denison (eds) A History of the English Language (pp. 1–42.) Cambridge: Cambridge University Press. De Saussure, F. (1916) Cours de linguistique générale. Payot. Dewaele, J-M. (1998) Lexical inventions: French interlanguage as L2 versus L3. Applied Linguistics Journal 19 (4), 471–490. Dewaele, J-M. (2001) Activation or inhibition? The interaction of L1, L2 and L3 on the language mode continuum. In J. Cenoz, B. Hufeisen and U. Jessner (eds) Crosslinguistic Influence in Third Language Acquisition: Psycholinguistic Perspectives (pp. 69–89). Clevedon: Multilingual Matters. Dewaele, J-M. (2008) The emotional weight of I love you in multilinguals’ languages. Journal of Pragmatics 40 (10), 1753–1780 Dewaele, J-M. (2010) Multilingualism and affordances: Variation in self-perceived communicative competence and communicative anxiety in French L1, L2, L3 and L4. IRAL 48, 105–129. Dewaele, J-M. and Pavlenko, A. (2003) Productivity and lexical diversity in native and non-native speech: A study of cross-cultural effects. In V. Cook (ed.) Effects of the Second Language on the First (pp. 120–141). Clevedon: Multilingual Matters. Dijkstra, T. (2003) Lexical processing in bilinguals and multilinguals. In J. Cenoz, B. Hufeisen and U. Jessner (eds) The Multilingual Lexicon (pp. 11–25). Dordrecht: Kluwer.

268

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Dijkstra, A. (2005) Bilingual visual word recognition and lexical access. In J.F. Kroll and A. De Groot (eds) Handbook of Bilingualism: Psycholinguistic Approaches (pp. 178–201). Oxford: Oxford University Press. Dijkstra, A. and Van Heuven, W.J.B. (2002) The architecture of the bilingual word recognition system: From identification to decision. Bilingualism: Language and Cognition 5, 175–197. Dijkstra, A., Van Jaarsveld, H. and Ten Brinke, S. (1998) Interlingual homograph recognition: Effects of task demands and language intermixing. Bilingualism: Language and Cognition 1, 51–66. Dijkstra, A., Grainger, J. and Van Heuven, W.J.B. (1999) Recognition of cognates and interlingual homographs: The neglected role of phonology. Journal of Memory and Language 41, 496–518. Dijkstra, A., Miwa, K., Brummelhuis, B., Sappelli, M. and Baayen, H. (2010) How crosslanguage similarity and task demands affect cognate recognition. Journal of Memory and Language 62, 284–301. Dörnyei, Z. (2006) Individual differences in second language acquisition. In K. BardoviHarling and Z. Dörnyei (eds) Themes in SLA Research. AILA Review 19, 42–68. Dörnyei, Z. (2007) Research Methods in Applied Linguistics: Quantitative, Qualitative, and Mixed Methodologies. Oxford: Oxford University Press. Dörnyei, Z. and Skehan, P. (2003) Individual differences in second language learning. In C. Doughty and M. Long (eds) The Handbook of Second Language Acquisition (pp. 589– 630). Oxford: Blackwell. Dörnyei, Z., Csizer, K. and Nemeth, N. (2006) Motivation, Language Attitudes and Globalisation: A Hungarian Perspective. Clevedon: Multilingual Matters. Doughty, C. and Long, M. (2003) The scope of inquiry and goals of SLA. In C. Doughty and M. Long (eds) The Handbook of Second Language Acquisition (pp. 3–16). Oxford: Blackwell. Dressler, C., Carlo, M.S., Snow, C., August, D. and White, C. (2011) Spanish speaking students’ use of cognate knowledge to infer the meaning of English words. Bilingualism: Language and Cognition 14, 243–255. Dubisz, S. (2005) Je˛zyk – Historia – Kultura. Warszawa: Wydział Polonistyki Uniwersytetu Warszawskiego. Dulay, H. and Burt, M. (1974) Natural sequences in child second language acquisition. Language Learning 24, 37–53. Duškova, L. (1984) Similarity – an aid or hindrance in foreign language learning? Folia Linguistica 18, 103–115. Ecke, P. (2014) Parasitic vocabulary acquisition, cross-linguistic influence, and lexical retrieval in multilinguals. Bilingualism: Language and Cognition, 1–18. Edwards, J. (1994) Multilingualism. London: Routledge. Edwards, J. (2004) Foundations of bilingualism. In T.K. Bhatia and W.C. Richtie (eds) The Handbook of Bilingualism (pp. 7–31). Blackwell Publishing. Ellis, N. (1995) The psychology of foreign language vocabulary acquisition: Implications for CALL. Computer Assisted Language Learning 8(2/3), 103–128. Ellis, N.C. and Beaton, A. (1993) Psycholinguistic determinants of foreign language vocabulary learning. Language Learning 43, 559–617. Ellis, R. (2008) The Study of Second Language Acquisition (2nd edn). Oxford: Oxford University Press. EuroCom See http://www.eurocomcenter.com [online]. (accessed 03 May 2009). European Commission (2005) A New Framework Strategy for Multilingualism. See http:// eacea.ec.europa.eu/education/eurydice/index_en.php (accessed 15 November 2013). European Comission (2006) Eurobarometer Report 243: Europeans and their Languages. See http:// ec.europa.eu/public_opinion/archives/ebs/ebs_243_en.pdf (accessed 15 November 2013).

References

269

EURYDICE (2012a) See http://eacea.ec.europa.eu/ressources/eurydice/pdf/0_ integral/095EN.pdf (accessed 07 December 2013). EURYDICE (2012b) The System of Education in Poland. Warsaw: FRSE. Ewert, A. (2009) L2 users’ L1. Poznan´: Pracowania wydawnicza IFA UAM. Falk, Y. and Bardel, C. (2011) Object pronouns in German L3 syntax: Evidence for the L2 status factor. Second Language Research 27 (1), 59–82. Falk, Y., Lindqvist, C. and Bardel, C. (2013) The role of L1 explicit metalinguistic knowledge in L3 oral production at the initial state. Bilingualism: Language and Cognition, 1–9. Firth, A. (1996) The discursive accomplishments of normality: On lingua franca English and conversation analysis. Journal of Pragmatics 26, 237–259. Fisiak, J. (1961) Zjawisko depluralizacji niektórych rzeczowników angielskich zapoz˙yczonych przez je˛zyk polski. Je˛zyk Polski XLI, 138–139. Fisiak, J. (1962) Złoz˙ony kontakt je˛zykowy w procesie zapoz˙yczania z je˛zyka angielskiego do polskiego. Je˛zyk Polski XLII, 286–294. Fisiak, J. (1963) Kategoria rodzaju rzeczowników zapoz˙yczonych z je˛zyka angielskiego. Rozprawy Komisji Je˛zykowej ŁTN, IX, 63–68. Fisiak, J. (1964) English sports terms in modem Polish. English Studies XLV, 230–236. Fisiak, J. (1968) Phonemics of English loanwords in Polish. Biuletyn fonograficzny IX, 69–79. Poznan´: PTN. Fisiak, J. (1970) The semantics of English loanwords in Polish. Studia Anglica Posnaniensia 1, 41–49. Fisiak, J. (1993) An Outline History of English. vol. 1 External History. Poznan´: Kantor Wydawniczy SAWW. Flippula, M. (2010) Contact and the early history of English. In Hickery R. (ed.) The Handbook of Language Contact (pp. 432–453). Oxford: Wiley-Blackwell. Font, N. (2001) Rôle de la langue dans l’accès au lexique chez les bilingues: Influence de la proximité orthographique et sé mantique interlangue sur la reconnaissance visuelle de mots. Unpublished doctoral thesis, Universite’ Paul Valery, Montpellier, France. Franceschini, R. (2009) The genesis and development of research in multilingualism: Perspectives for future research. In L. Aronin and B. Hufeisen (eds) The Exploration of Multilingualism (pp. 27–61). Amsterdam/Philadelphia: John Benjamins Publishing Company. Gabrys´-Barker, D. (2005) Aspects of Multilingual Storage, Processing and Retrieval. Katowice: Wydawnictwo Uniwesytetu S´la˛skiego. Gabrys´-Barker, D. (2012) (ed.) Crosslinguistic Influences in Multilingual Language Acquisition. Berlin/Heidelberg: Springer. Gabrys´-Barker, D. and Otwinowska, A. (2012) Multilingual learning stories: Threshold, stability and change. In A. Otwinowska and G. De Angelis (eds) Social and Affective Factors in Multilingualism Research. International Journal of Multilingualism 9 (4), 367–384. García, O. (2009) Bilingual Education in the 21st Century: A Global Perspective. New York: Blackwell/Wiley. Gardner, R.C. (1985) Social Psychology and Second language Learning: The Role of Attitudes and Motivation. London: Edward Arnold Publishers. Gass, S. (1988) Second language acquisition and linguistic theory: The role of language transfer. In S. Flynn and O’Neil (eds) Linguistic Theory in Second Language Acquisition (pp. 384–403). Dodrecht: Kluwer. Gass, S.M. and Selinker, L. (eds) (1992) Language Transfer in Language Learning: Revised edition. Amsterdam/Philadelphia: John Benjamins Publishing. Gass, S., Svetics, I. and Lemelin, S. (2003) Differential effects of attention. Language Learning 53 (3), 497–545.

270

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Gibson, J.J. (1977) The theory of affordances. In R.E. Shaw and J. Bransford (eds) Perceiving, Acting, and Knowing. Hillsdale, NJ: Lawrence Erlbaum Associates. Gibson, J.J. (1979) The Ecological Approach to Visual Perception. Boston: Houghton Mifflin. Gibson, M. and Hufeisen, B. (2003) Investigating the role of prior foreign language knowlege. In J. Cenoz, B. Hufeisen and U. Jessner (eds) The Multilingual Lexicon (pp. 87–101). Dordrecht: Kluwer. Gooskens, C. (2007) The contribution of linguistic factors to the intelligibility of closely related languages. Journal of Multilingual and Multicultural Development 28 (6), 445–467. Gooskens, C. and Van Bezooijen, R. (2006) Mutual comprehensibility of written Afrikaans and Dutch: Symmetrical or asymmetrical? Literary and Linguistic Computing 21 (4), 543–557. Gooskens, C., Kürschner, S. and Van Bezooijen, R. (2011) Intelligibility of Standard German and Low German to speakers of Dutch. Dialectologia. Special Issue II, 35–63. Grabowska, I. (1972) Nowe zapoz ˙yczenia angielskie w je˛zyku polskim. Prace filologiczne 23, 221–233. Graddol, D. (1997) The Future of English? London: The British Council. Graddol, D. (2006) English Next. London: British Council. Green, D. (1986) Control, activation and resource: A framework and a model for the control of speech in bilinguals. Brain and Language 27, 210–223. Greenhouse, G.W. and Geisser S. (1959) On methods of analysis of profile data. Psychometrika 24 (2), 95–112. Grin, F. (2002) Using Language Economics and Education Economics in Language Education Policy. Guide for the Development of Language Policies in Europe. From Linguistic Diversity to Plurilingual Education. Strasbourg: Language Policy Division, Council of Europe. Grosjean, F. (1982) Life with Two Languages: And Introduction to Bilingualism. Cambridge, MA: Harvard University Press. Grosjean, F. (1985) The bilingual as a competent but specific speaker-hearer. Journal of Multilingual and Multicultural Development 6, 467–477. Grosjean, F. (1989) Neurolinguists beware! The bilingual is not two monolinguals in one person. Brain and Language 36, 3–15. Grosjean, F. (1997) Processing mixed language: Issues, findings and models. In A. De Groot and J. Kroll (eds) Tutorials in Bilingualism: Psycholinguistic Perspectives (pp. 225– 254). Hillsdale, NJ: Erlbaum. Grosjean, F. (1998) Studying bilinguals: Methodological and conceptual issues. Bilingualism: Language and Cognition 1, 131-149. Grosjean, F. (2001) The bilingual’s language modes. In J. Nicol (ed.) One Mind, Two Languages: Bilingual Language Processing (pp.1–25). Oxford: Blackwell. Grosejan, F. (2008) Studying Bilinguals. Oxford: Oxford University Press Grucza, F. (2001) Origins and developments of applied linguistics in Poland. In E.F.K. Koerner and A.J. Szwedek (eds) Towards a History of Linguistics in Poland: From the Early Beginnings to the End of the 20th Century (pp. 53–100). Amsterdam and Philadelphia: John Benjamins Publishing Company. Haastrup, K. (1991) Lexical Inferencing Procedures or Talking about Words: A Book about Receptive Procedures in Foreign Language Learning with Special Reference to English. Tübingen: Gunter Narr. Hakuta, K. (1986) Mirror of Language. The Debate on Bilingualism. New York: Basic Books. Hakuta, K. and Diaz, R.M. (1985) The relationship between degree of bilingualism and cognitive ability: A critical discussion and some new longitudinal data. In K.E. Nelson (ed.) Children’s Language (pp. 319–344). Hillside: L. Erlbaum. Hall, C. and Ecke, P. (2003) Parasitism as a default mechanism in L3 vocabulary acquisition. In J. Cenoz, U. Jessner and B. Hufeisen (eds) The Multilingual Lexicon (pp. 71–85). Dordrecht: Kluwer Academic Publishers

References

271

Hall, C.J., Newbrand, D., Ecke, P., Sperr, U., Marchand, V. and Hayes, L. (2009) Learners’ implicit assumptions about syntactic frames in new L3 words: The role of cognates, typological proximity, and L2 status. Language Learning 59 (1), 153–202. Hamers, J.F. and Blanc, M. (1989) Bilinguality and Bilingualism. Cambridge: Cambridge University Press. Hamers, J.F. and Blanc, M. (2000) Bilinguality and Bilingualism (2nd edn). Cambridge: Cambridge University Press. Hammarberg, B. (2001) Roles of L1 and L2 in L3 production and acquisition. In J. Cenoz. B. Hufeisen and U. Jessner (eds) Crosslinguistic Influence in Third Language Acquisition: Psycholinguistic Perspectives (pp. 21–41). Clevedon: Multilingual Matters, Hammarberg, B. (2010) The languages of the multilingual: Some conceptual and terminological issues. IRAL, International Review of Applied Linguistics in Language Teaching 48, 91–104. Hammarberg, B. (2014) Problems in defining the concepts of L1, L2 and L3. In A. Otwinowska and G. De Angelis (eds) Teaching and Learning in Multilingual Contexts: Sociolinguistic and Educational Perspectives (pp. 3–18). Bristol: Multilingual Matters. Haugen, E (1950) The analysis of linguistic borrowing. Language 26, 210–231. Haugen, E. (1953) The Norwegian Language in America. Philadelphia: University of Pennsylvania Press. Haugen, E. (1968) Bilingualism in the Americas: A Bibliography and Research Guide. Montgomery: University of Alabama Press. Herdina, P. and Jessner, U. (2002) A Dynamic Model of Multilingualism: Perspectives of Change in Psycholinguistics. Clevedon: Multilingual Matters. Hickey, R. (2010) Language contact: Reconsideration and reassessment. In R. Hickey (ed.) The Handbook of Language Contact (pp 1–27). Oxford: Wiley-Blackwell. Hockett, C.F. (1958) A Course in Modern Linguistics. NY: Macmillan. Hoffmann, C. (2000) The spread of English and the growth of multilingualism with English in Europe. In J. Cenoz and U. Jessner (eds) English in Europe. The Acquisition of a Third Language. Clevedon: Multilingual Matters. Hoffmann, C. (2001) Towards a description of trilingual competence. International Journal of Bilingualism 5 (1), 1–17. Hoffmann, C. and Ytsma, J. (2004) Trilingualism in Family, School and Community. Clevedon: Multilingual Matters. Holliday, A. (1999) Small cultures. Applied Linguistics 20 (2), 237–264. Holmes, J. and Ramos, G.R. (1993) False friends and reckless guessers: Observing cognates recognition strategies. In L. Huckin, M. Haynes and J. Coady (eds) Second Language Reading and Vocabulary Learning (pp. 86–108). Norwood, New Jersey: Ablex Publishing Corporation. Hornberger, N.H. (2002) Multilingual language policies and the continua of biliteracy: An ecological approach. Language Policy 1, 27–51. Horst, M. and Meara, P. (1999) Test of a model for predicting second language lexical growth through reading. The Canadian Modern Language Review 56, 308–328. House, J. (2003) English as a lingua franca: A threat to multilingualism? Journal of Sociolinguistics 7/4, 556–578. Hufeisen, B. (1998) L3-Stand der Forschung – Was bleibt zu tun? In B. Hufeisen and B. Lindemann (eds) Tertiärsprachen. Theorien, Modelle, Methoden (pp. 169–183). Tübingen: Stauffenburg. Hufeisen, B. (2004) A brief introduction to the linguistic foundations. In B. Hufeisen and G. Neuner (eds) Tertiary Language Learning – German after English (pp. 7–13). Strasbourg: Council of Europe Publishing. Hufeisen, B. and Marx, N. (2007) EuroComGerm – Die Sieben Siebe: Germanische Sprachen lesen lernen. Aachen: Shaker Verlag. Hufeisen, B. and Neuner, G. (2004) Tertiary Language Learning – German after English. Strasbourg: Council of Europe Publishing.

272

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Hulstijn, J. (2001) Intentional and incidental second language vocabulary learning: A reappraisal of elaboration, rehearsal and automaticity. In P. Robinson (ed.) Cognition and Second Language Instruction (pp. 258–286). Cambridge: Cambridge University Press. Hymes, D.H. (1971) On Communicative Competence. Philadelphia: University of Pennsylvania Press. Iwan, K. (1972) Nauczanie je˛zyków obcych nowoz˙ytnych w Polsce w latach 1919-1939. Koncepcje organizacyjno-programowe. Poznan ´ UAM. James, A.R. (2000) English as a European Lingua Franca: Current realities and existing dichotomies. In J. Cenoz and U. Jessner (eds.) English in Europe. The Acquisition of a Third Language. Clevedon: Multilingual Matters. James, A.R. (2005) The challenges of the lingua franca: English in the world and types of variety. In C. Gnutzmann and F. Intemann (eds) The Globalisation of English and the English Language Classroom (pp. 133–144). Tübingen: Gunter Narr Verlag. James, C. and Garrett, P. (1991) Language Awareness in the Classroom. Harlow: Longman. Jarvis, S. (2003) Probing the limits of L2 effects in the L1: A case study. In V. Cook (ed.) Effects of the Second Language on the First (pp. 81–102). Clevedon: Multilingual Matters. Jarvis, S. (2009) Lexical transfer. In A. Pavlenko (ed.) The Bilingual Mental Lexicon. Interdisciplinary Approaches (pp. 99–124). Bristol: Multilingual Matters. Jarvis, S. (2011) Conceptual transfer: Crosslinguistic effects in categorization and construal. Bilingualism: Language and Cognition 14 (1), 1–8. Jarvis, S. and Pavlenko, A. (2007) Crosslinguistic Influence in Language and Cognition. London: Routledge. Jenkins, J. (2006) Points of view and blind spots: ELF and SLA. International Journal of Applied Linguistics 16 (2), 137–162. Jenkins, J. (2007) English as a Lingua Franca: Attitude and Identity. Oxford: Oxford University Press. Jespersen, O. (1928) An International Language. London: Allen and Unwin. Jessner, U. (1999) Metalinguistic awareness in multilinguals: Cognitive aspects of third language learning. Language Awareness 8, 201–209. Jessner, U. (2006) Linguistic Awareness in Multilinguals: English as a Third Language. Edinburgh: Edinburgh University Press. Jessner, U. (2008a) Language awareness in multilinguals: Theoretical trends. In J. Cenoz and N.H. Hornberger (eds) Encyclopedia of Language and Education, 2nd Edition, Volume 6: Knowledge about Language (pp. 1–13). New York: Springer. Jessner, U. (2008b) Teaching third languages: Findings, trends and challenges. Language Teaching 41, 15-56. Jessner, U. and Cenoz, J. (2007) Teaching English as a third language. In J. Cummins and C. Davison (eds) International Handbook of English Language Teaching (pp. 155–168). New York: Springer. Kachru, B.B. (1985) Standards, codification and sociolinguistic realism: The English language in the Outer Circle. In R. Quirk and H.G. Widdowson (eds) English in the World: Teaching and Learning the Language and Literatures (pp. 11–30). Cambridge: Cambridge University Press. Kachru, B.B. (1991) Liberation linguistics and the Quirk concern. English Today 7 (1), 3–13. Kachru, B.B (1999) Asian Englishes: Contexts, constructs, and creativity. Keynote address, the 12th World Congress of the International Association of Applied Linguistics, Tokyo. Kachru, B.B. and Nelson, C.L. (1996) World Englishes. In S. McKay and N. Hornberger (eds) Sociolinguistics and Language Teaching (pp. 71–99). Cambridge: Cambridge University Press.

References

273

Kastovsky, D. (2006) Vocabulary. In R. Hogg and D. Denison (eds) A History of the English Language (pp. 199–270). Cambridge: Cambridge University Press. Kecskes, I. and Papp, T. (2003) How to demonstrate the conceptual effect of the L2 on L1? In V. Cook (ed.) Effects of the Second Language on the First (pp. 247–267). Clevedon: Multilingual Matters. Kellerman, E. (1977) Towards a characterization of the strategy of transfer in second language learning. International Studies Bulletin 2, 58–145. Kellerman, E. (1978) Giving learners a break: Native language intuitions as a source of predictions about transferability. Working Papers on Bilingualism 15, 59–92. Kellerman, E. (1979) Transfer and non-transfer: Where we are now. Studies in Second Language Acquisition 2 (1), 37–57. Kellerman, E. (1983) Now you see it, now you don’t. In S. Gass and L. Selinker (eds) Language Transfer in Language Learning (pp. 112–134). Rowley, MA: Newbury House. Kellerman, E. (1986) An eye for an eye: Cross-linguistic constraints on the development of the L2 lexicon. In E. Kellerman and M. Sharwood-Smith (eds) Cross-linguistic Influence in Second Language Acquisition (pp. 35–48). Oxford: Pergamon. Kellerman, E. (1987) Aspects of Transferability in Second Language Acquisition. Chapter 1: Crosslinguistic Influence: A Review. Unpublished manuscript. University of Nijmegen. Kellerman, E. (1995) Crosslinguistic influence: Transfer to nowhere? Annual Review of Applied Linguistics 15, 125–150. Kellerman, E. and Sharwood Smith, M. (eds) (1986) Crosslinguistic Influence in Second Language Acquisition. Oxford: Pergamon. Kemp, C. (2009) Defining multilingualism. In L. Aronin and B. Hufeisen (eds) The Exploration of Multilingualism (pp. 11–26). Amsterdam/Philadelphia: John Benjamins Publishing Company. Klein, C. (2007) The valuation of plurilingual competences in an open european labour market. International Journal of Multilingualism 4 (4), 262–281. Klein, H.G. Meißner F. and Zybatow L. EuroCom – European Intercomprehension. See http://www.eurocomresearch.net/lit/ECengl-Innsbruck.htm (accessed 4 March 2013). Klemensiewicz, Z. (1961) Historia je˛zyka polskiego (Vol. 1). Warszawa: PWN. Klemensiewicz, Z. (1965) Historia je˛zyka polskiego (Vol. 2). Warszawa: PWN. Knapp, K. and Meierkord, Ch. (2002) Approaching lingua franca communication. In K. Knapp and C. Meierkord (eds) Lingua Franca Communication (pp. 3–15). Frankfurt/ Main: Peter Lang. Kolwa, A. (2001) Internationalism im Wortschatz der Politik. Interlexikologische Studien zum Wortschatz der Politik in neuen EU-Amtssprachen sowie im Russischen und Türkischen. Frankfurt am Main: Peter Lang. Komorowska, H. (2004) Polityka je˛zykowa – warianty rozwia˛zan´ w krajach europejskich, Je˛zyki obce w szkole 2, 38–45. Komorowska, H. (2007a) Polska polityka je˛zykowa na tle innych krajów Unii Europejskiej. In H. Komorowska (ed.) Nauczanie je˛zyków obcych – Polska a Europa (pp. 13–55). Warszawa: Academica. Komorowska, H. (2007b) Metodyka nauczania je˛zyków obcych w Polsce (1957–2007). Warszawa: CODN. Komorowska, H. (2010) Approaches to multilingualism in Europe – Shared values and controversial issues. In ISET Working Paper 15. London: London Metropolitan University. See http://www.londonmet.ac.uk/londonmet/fms/MRSite/Research/iset/ Working%20Paper%20Series/WP%2015.pdf (accessed 1 October 2014). Komorowska, H. (2014) Analyzing linguistic landscapes. A diachronic study of multilingualism in Poland. In A. Otwinowska and G. De Angelis (eds) Teaching and Learning

274

Cognate Vocabul ar y in L anguage Acquisit ion and Use

in Multilingual Contexts: Sociolinguistic and Educational Perspectives (pp. 19–31). Bristol: Multilingual Matters. Koneczna, H. (1936) Wyrazy angielskie w je˛zyku polskim. Poradnik Je˛zykowy 9, 161–170. Krashen, S. (1982) Principles and Practice in Second Language Acquisition. Oxford: Pergamon Press. Kroll, J.F. (1993) Accessing conceptual representations for words in a second language. In R. Schreuder and B. Weltens (eds) The Bilingual Lexicon (pp. 53–81). Amsterdam: John Benjamins. Kroll, J.F. (2008) Juggling two languages in one mind. Psychological Science Agenda SCIENCE BRIEFS. See http://www.apa.org/science/about/psa/2008/01/kroll.aspx (accessed January 2014). Kroll, J.F. and Stewart, E. (1994) Category interference in translation and picture naming: Evidence for asymmetric connections between bilingual memory representations. Journal of Memory and Language 33, 149–174. Kroll, J.F. and De Groot, A.M.B. (1997) Lexical and conceptual memory in the bilingual: Mapping form to meaning in two languages. In A.M.B. De Groot and J.F. Kroll (eds) Tutorials in Bilingualism (pp. 169–199). Mahwah: Lawrence Erlbaum. Kroll, J.F. and Bialystok, E. (2013) Understanding the consequences of bilingualism for language processing and cognition. Journal of Cognitive Psychology 25 (5), 497–514. Kroll, J.F., Michael, E. and Sankaranarayanan, A. (1998) A model of bilingual representation and its implications for second language acquisition. In A.F. Healy and L.E. Bourne (eds) Foreign Language Learning: Psycholinguistic Studies on Training and Retention (pp. 365–395). Mahwah, NJ: Erlbaum. Kroll, J.F., Bobb, S.C. and Wodniecka, Z. (2006) Language selectivity is the exception, not the rule: Arguments against a fixed locus of language selection in bilingual speech. Bilingualism: Language and Cognition 9, 119–135. Kroll, J.F., Van Hell, J.G., Tokowicz, N. and Green, D.W. (2010) The revised hierarchical model: A critical review and assessment. Bilingualism: Language and Cognition 13, 373–381. Kweon, S. and Kim, H (2008) Beyond raw frequency: Incidental vocabulary acquisition in extensive reading. Reading in a Foreign Language 20 (2), 191–215. Labov, W. (1969) The Study of Nonstandard English. Washington, DC: National Council of Teachers of English. Lado, R. (1957) Linguistics Across Cultures. Ann Arbor: The University of Michigan Press. Larsen-Freeman, D. and DeCarrico, J. (2010) Grammar. In N. Schmitt (ed.) An Introduction to Applied Linguistics (pp. 18–33). London: Hodder Education. Laufer, B. (1990) Ease and difficulty in vocabulary learning: Some teaching implications. Foreign Language Annals 23, 147–156. Laufer, B. (1997) What’s in a word that makes it hard or easy: Some intralexical factors that affect the learning of words. In N. Schmitt and M. McCarthy (eds) Vocabulary, Description, Acquisition and Pedagogy (pp. 140–155). Cambridge: CUP. Laufer, B. (2003) The influence of L2 on L1 collocational knowledge and on L1 lexical diversity in written expression. In V. Cook (ed.) Effects of the Second Language on the First (pp. 19–31). Clevedon: Multilingual Matters. Lemhöfer, K. and Dijkstra, T. (2004) Recognizing cognates and interlexical homographs: Effects of code similarity in language specific and generalized lexical decision. Memory and Cognition 32, 533–550. Lemhöfer, K., Dijkstra, T., Schriefers, H., Baayen R.H, Grainger, J. and Zwitserlood, P. (2008) Native language influences on word recognition in a second language: A megastudy. Journal of Experimental Psychology: Learning, Memory, and Cognition 34 (1), 12–31. Leung, C., Harris, R. and Rampton, C. (1997) The idealized native-speaker, reified ethnicities, and classroom realities. TESOL Quarterly 31, 543–560.

References

275

Lightbown, P.M. and Libben, G. (1984) The recognition and use of cognates by L2 learners. In R.W. Anderson (ed.) Second Languages (pp. 393–417). Rowley, Mass: Newbury House. Lotto, L. and De Groot, A.M.B. (1998) Effects of learning method and word type on acquiring vocabulary in an unfamiliar language. Language Learning 48, 31–69. Mackey, W.F. (1970) The description of bilingualism. In J. Fishman (ed.) Readings in the Sociology of Language (pp. 554–584). The Hague: Mouton. Mackey, W.F. (2000) The description of bilingualism. In L. Wei (ed.) The Bilingualism Reader (pp. 26–54). Routledge: London and New York. Mackiewicz, W. (2002) Plurilingualism in the European Knowledge Society, speech delivered at the Conference on lingue e produzione del sapere, Swiss Academy of Humanities and Social Sciences, Università Della Svizerra Italiana, 14 June 2002. See http:// www.celelc.org/docs/speech_final_website_1.doc (accessed 30 April 2009). Maćkiewicz, J. (1984) Co to sa˛ tzw. Internacjonalizmy. Je˛zyk Polski LXIV, 176–184. Maćkiewicz, J. (1988) Internacjonalizmy a je˛zykowy obraz s´wiata. Poradnik Jézykowy 7, 488–496. Maćkiewicz, J. (1992) Wyrazy mie˛dzynarodowe a kształtowanie sie˛ europejskiej ligi słownikowej. In J. Maćkiewicz and J. Siatkowski (eds) Je˛zyk a Kultura. Kontakty je˛zyka polskiego z innymi je˛zykami na tle kontaktów kulturowych (pp. 145–153). Wrocław: Wiedza o Kulturze. Maćkiewicz, J. and Siatkowski, J. (eds) (1992) Je˛zyk a kultura. Kontakty je˛zyka polskiego z innymi je˛zykami na tle kontaktów kulturowych. Wrocław: Wiedza o Kulturze. Macnamara, J. (1967) The bilingual’s linguistic performance: A psychological overview. Journal of Social Issues 23, 59–77. MacWhinney, B. (1999) The emergence of language from embodiment. In B. MacWhinney (ed.) The Emergence of Language (pp. 213–256). Mahwah, NJ: Lawrence Erlbaum. Malabonga, V., Kenyon, D.M., Carlo, M. August D. and Louguit M. (2008) Development of a cognate awareness measure for Spanish-speaking English language learners. Language Testing 25 (4), 495–519. Man´czak-Wohlfeld, E. (1994) Angielskie elementy leksykalne w je˛zyku polskim. Kraków: Universitas. Man´czak-Wohlfeld, E. (1995) Tendencje rozwojowe współczesnych zapoz˙yczen´ angielskich w je˛zyku polskim. Kraków: Universitas. Man´czak-Wohlfeld, E. (2006) Language contact vs. foreign and second language acquisition. In J. Arabski (ed.) Cross-linguistic Influences in the Second Language Lexicon (pp. 37–46). Clevedon: Multilingual Matters. Mansour, G. (1993) Multilingualism and Nation Building. Clevedon: Multilingual Matters. Markowski, A. (1992) Nowsze zapoz˙yczenia w polszczyz´ nie: Anglicyzmy gramatyczne i leksykalne. Poradnik Je˛zykowy 3, 237–241. McArthur, T. (ed.) (1992) Oxford Companion to the English Language. Oxford: Oxford University Press. McKay, S. (2002) Teaching English as an International Language. Oxford: Oxford University Press. Meara, P. (1992) ELF Vocabulary Tests. University College Swansea: Centre for Applied Linguistics Studies. Meara, P. (1993) The bilingual lexicon and the teaching of vocabulary. In R. Schreuder and B. Weltens (eds) The Bilinual Lexicon (pp. 279–298). Amsterdam: John Benjamins. Meara, P. and Milton, J. (2003) X_lex, The Swansea Levels Test. Newbury: Express. Milton, J. (2009) Measuring Second Language Vocabulary Acquisition. Bristol: Multilingual Matters. Milton, J. and Daller, H.M. (2007) The interface between theory and learning in vocabulary acquisition. Paper presented at the EuroSLA conference, Newcastle.

276

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Meißner, F., Meissner C., Klein H. and Stegmann T.D. (2004) EuroComRom: Les sept tamis: Lire toutes les langues romanes dès le départ; avec une esquisse de la didactique de l´eurocompréhension. Aachen: Shaker, Ed. EuroCom. Mesthrie, R. (2010) Sociolinguistics and sociology of language. In B. Spolsky and F.M. Hult (eds) The Handbook of Educational Linguistics (pp. 66–82). Singapore: Wiley-Blackwell. Miller, J. (1983) Many Voices. Bilingualism, Culture and Education. London: Routledge. Mißler, B. (1999) Fremdsprachenlernerfahrungen und Lernstrategien. Eine empirische Untersuchung. Tübingen: Stauffenburg. Moz˙ejko, Z. (2009) Z badan´ nad kształceniem zawodowym i je˛zykowym przyszłych nauczycieli je˛zyka angielskiego. In Komorowska-Janowska, H. (ed.) Kształcenie je˛zykowe w szkolnictwie wyz˙szym (pp. 255– 270). Warszawa: Wydawnictwo SWPS Academica. Müller-Lancé, J. (2003) A strategy model of multilingual learning. In J. Cenoz, U. Jessner and B. Hufeisen (eds) The Multilingual Lexicon (pp. 117–132). Dordrecht: Kluwer Academic Publishers. Murphy, Sh. (2003) Second language transfer during third language acquisition. Working Papers in TESOL Applied Linguistics 3 (1), 1–21. Myers-Scotton, C. (2005) Multiple Voices. Oxford: Blackwell. Nagy, W.E., García, G.E., Durgunoglu, A.Y. and Hancin-Bhatt, B. (1993) Spanish-English bilingual students’ use of cognates in English reading. Journal of Reading Behavior 25, 241–259. Nagy, W. and Townsend, D. (2012) Words as tools: Learning academic vocabulary as language acquisition. Reading Research Quarterly 47(1), 91–108. Naiman, N., Fröhlich, M., Stern H. and Todesco A. (1978) The Good Language Learner. Clevedon: Multilingual Matters. Nation, P. and Meara, P. (2002) Vocabulary. In N. Schmitt (ed.) An Introduction to Applied Linguistics (pp. 35–53). London: Arnold. Nayak, N., Hansen, N., Krueger, N. and McLaughlin, B. (1990) Language-learning strategies in monolingual and multilingual adults. Language Learning 40 (2), 221–244. Norman, D.A. (1988) The Psychology of Everyday Things. New York: Basic Books. Norman, D.A. (1999) Affordances, conventions and design. Interactions 6 (3), 38–43. Ó Laoire, M. (2001) Balanced bilingual and L1-dominant learners of L3 in Ireland: A case study. In J. Cenoz, B. Hufeisen and U. Jessner (eds) Looking Beyond Second Language Acquisition: Studies in Tri and Multilingualism (pp. 153–160). Tübingen: Stauffenburg. Ó Laoire, M. and Aronin, L. (2004) Exploring multilingualism in cultural contexts: Towards a notion of multilinguality. In C. Hoffmann and J. Ytsma (eds) Trilingualism in Family, School and Community (pp. 11–29). Clevedon: Multilingual Matters. Odlin, T. (1989) Language Transfer. Cambridge: CUP. Odlin, T. (2002) Language transfer and cross-linguistic studies: Relativism, universalism, and the native language. In R.B. Kaplan (ed.) The Oxford Handbook of Applied Linguistics (pp. 253–261). Oxford: Oxford University Press. Odlin, T. (2003) Cross-linguistic influence. In C. Doughty and M. Long (eds) The Handbook of Second Language Acquisition (pp. 436–486). Oxford: Balckwell. Odlin, T. (2005) Crosslinguistic influence and conceptual transfer: What are the concepts? Annual Review of Applied Linguistics 25, 3–25. Odlin, T. (2006) Could a contrastive analysis ever be complete. In J. Arabski (ed.) Crosslinguistic Influences in the Second Language Lexicon (pp. 22–35). Clevedon: Multilingual Matters. Odlin, T. (2008) Conceptual transfer and meaning extensions. In P. Robinson and N. Ellis (eds) Handbook of Cognitive Linguistics and Second Language Acquisition (pp. 306–340). New York: Routledge.

References

277

Odlin, T. (2009) Transfer and code-switching: Separate territories, but common concerns on the border. In L. Isurin, D. Winford and K. de Bot (eds) Multidisciplinary Approaches to Code Switching (pp. 337–358). Amsterdam: John Benjamins. O’Malley, J.M. and Chamot, A. (1990) Learning Strategies in Second Language Acquisition. Cambridge: Cambridge University Press. Otwinowska, A. (2012) Plurilingualism and Polish teenage learners of English. Linguarum Arena 3, 37–52. Otwinowska-Kasztelanic, A. (2000a) A Study of the Lexico-semantic and Grammatical Influence of English on the Polish of the Younger Generation of Poles (19–35 years of age). Warszawa: Wydawnictwo Akademickie Dialog. Otwinowska-Kasztelanic, A. (2000b) Korpus je˛zyka mówionego młodego pokolenia Polaków (19-35 lat). Warszawa: Wydawnictwo Akademickie Dialog. Otwinowska-Kasztelanic, A. (2001) Latent bilingualism and language learning. In J. Arabski (ed.) Insights into Foreign Language Acquisition and Teaching (pp. 106–120). Katowice: Wydawnictwo Uniwersytetu S´la˛skiego. Otwinowska-Kasztelanic, A. (2007a) Na drodze ku róz ˙noje˛zycznos´ci: Podnoszenie s´wiadomos´ci je˛zykowej przy nauczaniu słownictwa na przykładzie je˛zyka angielskiego. In A. Niz˙egorodcew and M. Jodłowiec (eds) Dydaktyka je˛zyków obcych na pocza˛tku XXI wieku (pp. 99–112). Kraków: Wydawnictwo Uniwersytetu Jagiellon´skiego. Otwinowska-Kasztelanic, A. (2007b) Positive transfer and motivation in teaching cognate vocabulary. In J. Arabski (ed.) Challenging Tasks for Psycholinguistics in the New Century (pp. 613–621). Katowice: Oficyna Wydawnicza WW. Otwinowska-Kasztelanic, A. (2009) Raising awareness of cognate vocabulary as a strategy in teaching English to Polish adults. Innovation in Language Learning and Teaching 3 (2), 131–148. Otwinowska-Kasztelanic, A. (2010) Language awareness in using cognate vocabulary: The case of Polish advanced students of English in the light of the theory of affordances. In J. Arabski and A. Wojtaszek (eds) Neurolinguistic and Psycholinguistic Perspectives on SLA (pp. 175–190). Bristol: Multilingual Matters. Otwinowska-Kasztelanic, A. (2011a) Awareness and affordances. Multilinguals versus bilinguals and their perceptions of cognates. In G. De Angelis and J-M. Dewaele (eds) New Trends in Crosslinguistic Influence and Multilingualism Research (pp. 1–18). Bristol: Multilingual Matters. Otwinowska-Kasztelanic, A. (2011b) Awareness of cognate vocabulary and vocabulary learning strategies of Polish multilingual and bilingual advanced learners of English. In J. Arabski and A. Wojtaszek (eds) Individual Learner Differences in SLA (pp. 110– 126). Bristol: Multilingual Matters. Otwinowska, A. and Szewczyk J. (2013) Is similar always better? Factors in bilingual learning of cognates. Poster presented at the conference Architectures and Mechanisms of Language Processing. Marseille: Universite Aix-Marseille. Oxford, R. (1990a) Language Learning Strategies: What Every Teacher Should Know. New York: Newbury House. Oxford, R. (1990b) Styles, strategies, and aptitude: Connections for language learning. In T.S. Parry and C.W. Stansfield (eds) Language Aptitude Reconsidered (pp. 67–125). Englewood Cliffs, NJ: Prentice Hall. Oxford, R. (1992/1993) Language learning strategies in a nutshell: Update and ESL suggestions. TESOL Journal 2 (2), 18–22. Oxford, R. (2011) Teaching and Researching Language Learning Strategies. London: Pearson. Paradis, M. (2004) A Neurolinguistic Theory of Bilingualism. Amsterdam/Philadelphia: John Benjamins Publishing. Paradis, M. (2007) Why single word experiments do not address language representation. In J. Arabski (ed.) Challenging Tasks for Psycholinguistics in the New Century (pp. 22–31). Katowice: Oficyna Wydawnicza WW.

278

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Paribakht, T.S. and Wesche, M.B. (1993) Reading comprehension and second language development in a comprehension-based ESL program. TESL Canada journal 11 (1), 09–29. Paribakht, T.S. and Wesche, M. (1997) Vocabulary enhancement activities and reading for meaning in second language vocabulary acquisition. In J Coady and T. Huckin (eds) Second Language Vocabulary Acquisition: A Rationale for Pedagogy (pp. 174–200). New York, Cambridge University Press. Pavlenko, A. (1999) New approaches to concepts in bilingual memory. Bilingualism: Language and Cognition 2 (3), 209–230. Pavlenko, A. (2003) ‘I feel clumsy speaking Russian’: L2 influence on L1 in narratives of Russian L2 users of English. In V. Cook (ed.) Effects of the Second Language on the First (pp. 32–61). Clevedon: Multilingual Matters. Pavlenko, A. (2009a) Conceptual representation in the bilingual lexicon and second language vocabulary learning. In A. Pavlenko (ed.) The Bilingual Mental Lexicon (pp. 125–160). Bristol: Multilingual Matters. Pavlenko, A. (2009b) Multilingualism and Democracy. A keynote speech delivered at the Sixth International Conference on Third Language Acquisition and Multilingualism, September 2009, Free University of Bozen-Bolzano. Pavlenko, A. and Jarvis, S. (2002) Bidirectional transfer. Applied Linguistics 23 (2), 190–214. Peal, E. and W. Lambert (1962) The relation of bilingualism to intelligence. Psychological Monographs 76, 1–23. Pennycook, A. (2007) ELT and colonialism. In J. Cummins and C. Davison (eds) International Handbook of English Language Teaching (pp. 13–24). New York: Springer. Phillipson, R. (1992) Linguistic Imperialism. Oxford: Oxford University Press. Phillipson, R (2007) English, no longer a foreign language in Europe? In J. Cummins and C. Davison (eds) International Handbook of English Language Teaching (pp. 123–137). New York: Springer. Pisarek, W. (2007) The Polish Language. Warsaw: The Council for the Polish Language Poarch, G.J. and Van Hell, J.G. (2012) Cross-language activation in children’s speech production: Evidence from second language learners, bilinguals, and trilinguals. Journal of Experimental Child Psychology 111, 419–438. Potter, M.C., So, K.F., Eckardt, B.V. and Feldman, L.B. (1984) Lexical and conceptual representation in beginning and proficient bilinguals. Journal of Verbal Learning and Verbal Behavior 23 (1), 23–38. Poulisse, N. (1999) Slips of the Tongue: Speech Errors in First and Second Language Production. Amsterdam/Philadelphia: John Benjamins Publishing. Prodromou, L. (2006) Defining the ‘successful bilingual speakers’ of English. In R. Rubdy and M. Saraceni (eds) English in the World: Global Rules, Global Roles (pp. 51–69). London, New York: Continuum. Pyles, T and Algeo, J. (1982) The Origins and the Development of the English Language. Harcourt Barace Jovanovich Inc: New York Ricciardelli (1992) Bilingualism and cognitive development in relation to threshold theory. Journal of Psycholinguistic Research 21 (4), 301–316. Ringbom, H. (1986) Crosslinguistic influence and foreign language learning process. In E. Kellermann and M. Sharwood Smith (eds) Crosslinguistic Influence in Second Language Acquisition (pp. 150–162). New York and Oxford: Pergamon. Ringbom, H. (1987) The Role of the First Language in Foreign Language Learning. Clevedon: Multilingual Matters. Ringbom, H. (2001) Lexical transfer in L3 production. In J. Cenoz., B. Hufeisen and U. Jessner (eds) Cross-linguistic Influence in Third Language Acquisition: Psycholinguistic Perspectives (pp. 59–68). Clevedon: Multilingual Matters. Ringbom, H. (2006) The importance of different types of similarity in transfer studies. In J. Arabski (ed.) Cross-linguistic Influences in the Second Language Lexicon (pp. 36–45). Clevedon: Multilingual Matters.

References

279

Ringbom, H. (2007) Cross-linguistic Similarity in Foreign Language Learning. Clevedon: Multilingual Matters. Ringbom, H. and Jarvis, S. (2009) The importance of cross-linguistic similarity in foreign language learning. In M.H. Long and C.J. Doughty (eds) The Handbook of Language Teaching (pp. 107–118). Blackwell Publishing. Robinson, P. (ed.) (2002) Individual Differences and Instructed Language Learning. Amsterdam/Philadelphia: John Benjamins Publishing. Robinson, P. (2003) Attention and memory during SLA. In C. Doughty and M. Long (eds) The Handbook of Second Language Acquisition (pp. 631–678). Oxford: Blackwell Romaine, S. (1989) Bilingualism. Oxford: Blackwell. Rubin, J. (1987) Learner strategies: Theoretical assumptions, research history and typology. In A. Wenden and J. Rubin (eds) Learner Strategies in Language Learning (pp. 15–30). New Jersey: Prentice Hall. Rudolf, K.F. (2003) Fałszywi przyjaciele. False friends. Słownik angielsko-polski polsko-angielski wyrazów zdradliwych. Kraków: Wydawnictwo Literackie. Rusiecki, J. (1980) Latent bilingualism. Papers and Studies in Contrastive Linguistics 12, 81–98. Rusiecki, J. (2002) Friends true and false. A contrastive approach to the lexicon. In J. Arabski (ed.) Time for Words (pp. 71–81). Frankfurt-am-Main: Peter Lang. Rybicka, H. (1976) Losy wyrazów obcych w je˛zyku polskim. Warszawa: PWN. Sandhofer, C. and Smith, L.B. (2007) Learning adjectives in the real world: How learning nouns impedes learning adjectives. Language Learning and Development 3 (3), 233–267. Sawicka, A. (1995) Norma a problem tak zwanych zapoz ˙yczen´. Poradnik Je˛zykowy 9–10, 78–86. Schepens, J., Dijkstra, T. and Grootjen, F. (2012) Distributions of cognates in Europe as based on Levenshtein distance. Bilingualism: Language and Cognition 15 (1), 157–166. Schmidt, R. (1990) The role of consciousness in second language learning. Applied Linguistics 11, 129–158. Schmidt, R.W. (1994a) Deconstructing consciousness in search of useful definitions for applied linguistics. AILA Review 11, 11-26. Schmidt, R.W. (1994b) Implicit learning and the cognitive unconscious: Of artificial grammars and SLA. In N.C. Ellis (ed.) Implicit and Explicit Learning of Languages (pp. 165–209). San Diego, CA: Academic Press. Schmidt, R. (1998) The centrality of attention in SLA. University of Hawaii Working Papers in ESL 16, 1–34. Schmidt, R. (2001) Attention. In P. Robinson (ed.) Cognition and Second Language Instruction (pp. 3–32). Cambridge: Cambridge University Press. Schmitt, N. (1997) Vocabulary learning strategies. In N. Schmitt and M. McCarthy (eds) Vocabulary. Description, Acquisition and Pedagogy (pp. 199–227). Cambridge: Cambridge University Press. Schmitt, N. and Zimmerman, C.B. (2002) Derivative word forms: What do learners know? TESOL Quarterly 36 (2), 145–171. Schramm, E. (2008) Dzieje nauki je˛zyka angielskiego i innych je˛zyków nowoz˙ytnych w Polsce w okresie zaborów (od schyłku I do narodzin II Rzeczpospolitej. Warszawa: Fraszka Edukacyjna. Schumann, J. (1997) The Neurobiology of Affect in Language. Malden, MA: Blackwell. Schwartz, A.I. and Kroll, J.F. (2006) Bilingual lexical activation in sentence context. Journal of Memory and Language 55 (2), 197–212. Segalowitz, N. (1997) Individual differences in second language acquisition. In A. De Groot and J.F. Kroll (eds) Tutorials in Bilingualism: Psycholinguistic Perspectives (pp. 85–112). Mahwah, NJ: Lawrence Erlbaum Associates. Segalowitz, N. (2001) On the evolving connections between psychology and linguistics. Annual Review of Applied Linguistics 21, 3–22.

280 Cognate Vocabul ar y in L anguage Acquisit ion and Use

Segalowitz, N. (2003) Automaticity and second language acquisition. In C. Doughty and M. Long (eds) The Handbook of Second Language Acquisition (pp. 382–408). Oxford: Blackwell. Segalowitz, N. (2007) Access fluidity, attention control, and the acquisition of fluency in a second language. TESOL Quarterly 41 (1), 181–186. Seidlhofer, B. (2001) Closing a conceptual gap: The case for a description of English as a Lingua Franca. International Journal of Applied Linguistics 11, 133–158. Seidlhofer, B. (2002) The shape of things to come? Some basic questions about English as a lingua franca. In K. Knapp and C. Meierkord (eds) Lingua Franca Communication (pp. 269–302). Frankfurt/Main: Peter Lang. Seidlhofer, B. (2005) Standard future, or half-baked quackery? Descriptive and pedagogic bearings on the globalisation of English. In C. Gnutzmann and F. Intemann (eds) The Globalisation of English and the English Language Classroom (pp. 159–173). Tübingen: Gunter Narr Verlag. Seidlhofer, B. (2007) Common property: English as a lingua franca in Europe. In J. Cummins and C. Davison (eds) International Handbook of English Language Teaching (pp. 137–153). New York: Springer. Se˛kowska E. (1993) Funkcjonowanie nowszych zapoz ˙yczen´ angielskich we współczesnej polszczyz´nie. Wybrane zagadnienia. Poradnik Je˛zykowy 5, 242–248. Selinker, L. (1983) Language transfer. In S. Gass and L. Selinker (eds) Language Transfer in Language Learning (pp. 33–68). Rowley, MA: Newbury House. Sharwood Smith, M. and Kellerman, E. (1986) Crosslinguistic influence in second language: An introduction. In E. Kellerman and M. Sharwood Smith (eds) Crosslinguistic Influence in Second Language Acquisition (pp. 1–9). Oxford: Pergamon. Sheen, R. (1977) ‘Faux amis’ for Francophones learning EFL. SPEAQ Journal 3 (1–2), 95–103. Siegel, J. (2003) Social context. In C. Doughty and M. Long (eds) The Handbook of Second Language Acquisition (pp. 178–223). Oxford: Balckwell. Singleton, D. (1987) Mother and other tongue influences on learner French: A case study. Studies in Second Language Acquisition 9 (3), 327–345. Singleton, D. (1999) Exploring the Second Language Mental Lexicon. Cambridge: Cambridge University Press. Singleton, D. (2003) Perspectives on the multilingual lexicon: A critical synthesis. In J. Cenoz, B. Hufeisen and U. Jessner (eds) The Multilingual Lexicon (pp. 167–176). Dordrecht: Kluwer. Singleton, D. (2006) Lexical transfer: Interlexical or intralexical? In J. Arabski (ed.) Crosslinguistic Influences in the Second Language Lexicon (pp. 130–143). Clevedon: Multilingual Matters. Singleton, D. and Aronin, L. (2007a) Multiple language learning in the light of the theory of affordances. Innovation in Language Learning and Teaching 1 (1), 83–96. Singleton, D. and Aronin, L. (2007b) The role of English in multilingual world. In J. Arabski, D. Gabrys´-Barker and A. Łyda (eds) PASE Papers 2007 vol. 1: Studies in Language Methodology of Teaching Foreign Language (pp. 11–21). Katowice: University of Silesia. Sjöholm, K. (1976) A comparison of the test results in grammar and vocabulary between Finnish- and Swedish-speaking applicants for English. In H. Ringbom and R. Palmberg (eds) Errors made by Finns and Swedish-speaking Finns in the Learning of English (pp. 54–137). Åbo: Åbo Academy. Skutnabb-Kangas, T. (1984) Bilingualism or Not – The Education of Minorities. Clevedon: Multilingual Matters. Skutnabb-Kangas, T. (1995) Multilingualism and the education of minority children. In O. García and C. Baker (eds) Policy and Practice in Bilingual Education: Extending the Foundations (pp. 40–62). Clevedon: Multilingual Matters.

References

281

Slobin, D.I. (1996) From ‘thought and language’ to ‘thinking for speaking’. In J. Gumperz and S. Levinson (eds) Rethinking Linguistic Relativity (pp. 70–96). Cambridge: Cambridge University Press. Smith, L.B., Gasser, M. and Sandhofer, C.M. (1997) Learning to talk about the properties of objects: A network model of the development of dimensions. In R. L. Goldstone, P.G. Schyns and D.L. Medin (eds) Perceptual Learning, The Psychology of Learning and Motivation (pp. 219–255). San Diego, CA: Academic Press. Snow, R.E. (1998) Abilities as aptitudes and achievements in learning situations. In J.J. McArdle and R.W. Woodcock (eds) Human Cognitive Abilities in Theory and Practice (pp. 93–112). Mahwah, NJ: Lawrence Erlbaum Associates. Svalberg, A. (2007) Language awareness and language learning. ELT Journal 40, 287–308. Swan, M. (1997) The influence of the mother tongue on second language vocabulary acquisition and use. In N. Schmitt and M. McCarthy (eds) Vocabulary. Description, Acquisition and Pedagogy (pp. 156–180). Cambridge: Cambridge University Press. Sweet, H. (1899/1972) The Practical Study of Languages. London: OUP Szpila, G. (2005) An English-Polish Dictionary of False Friends. Kraków: Egis. Szubko-Sitarek, W. (2011) Cognate facilitation effects in trilingual word recognition. Studies in Second Language Learning and Teaching 1 (2), 189–208. Tella, S. and Harjanne, P. (2007) Can we afford any more affordances? Foreign language education specific reflections. In K. Teoksessa, K. Merenluoto, A. Virta and P. Carpelan (eds) Opettajankoulutuksen muuttuvat rakenteet. Ainedidaktinen symposium 9.2.2007. Turun opettajankoulutuslaitos (pp. 500–506). See http://www.helsinki. fi/~tella/ads07.pdf (accessed 20 March 2008). Thomas, J. (1988) The role played by metalinguistic awareness in second and third language learning. Journal of Multilingual and Multicultural Development 9, 235–246. Thomas, J. (1992) Metalinguistic awareness in second- and third-language learning. In R. Harris (ed.) Cognitive Processing in Bilinguals (pp. 531–545). Amsterdam: North Holland. Thomason, S.G. (2001) Language Contact. An Introduction. Edinburgh: Edinburgh University Press. Thomason, S. (2010) Contact explanations in linguistics. In R. Hickey (ed.) The Handbook of Language Contact (pp 31–47). Oxford: Wiley-Blackwell. Thomason, S.G. and Kaufman, T. (1988) Language Contact, Creolization, and Genetic Linguistics. Berkeley: University of California Press. Thompson, G. (1952) Child Psychology. Boston: Houghton Mifflin. Tognini-Bonelli, E. (2001) Corpus Linguistics at Work (Vol. 6). John Benjamins Publishing. Tokowicz, N., Kroll, J.F., De Groot, A.M.B. and Van Hell, J.G. (2002) Number-oftranslation norms for Dutch–English translation pairs: A new tool for examining language production. Behavior Research Methods, Instruments and Computers 34 (3), 435–451. Trask, R.L. (1994) Language Change. London: Routlege. Traxler, M.J. (2012) Introduction to Psycholinguistics: Understanding Language Science. Oxford: Wiley-Blackwell. Tremblay, M-C. (2006) Crosslinguistic influence in third language acquisition: The role of L2 proficiency and L2 exposure. CLO/OPL Ottawa 34, 109–119. Tseng, W., Dörnyei Z. and Schmitt, N. (2006) A new approach to assessing strategic learning: The case of self-regulation in vocabulary acquisition. Applied Linguistics 27 (1), 78–102. Van Avermaet, P. and Pulinx, R. (2012) The role of language in the integration policies. In H. Komorowska and J. Zaja˛c (eds) Multilingual Competences for Professional and Social Success in Europe (pp. 267–278). Warszawa: FRSE. Van Gelderen, E. (2006) A History of the English Language. Philadelphia: John Benjamins.

282

Cognate Vocabul ar y in L anguage Acquisit ion and Use

Van Hell, J. (2005) The influence of sentence context constraint on cognate effects in lexical decision and translation. In J. Cohen, K.T. McAlister, K. Rolstad, J. MacSwan (eds) Proceedings of the 4th International Symposium on Bilingualism (pp. 2297–2309). Somerville, MA: Cascadilla Press. Van Hell, J.G. and De Groot, A.M.B. (1998) Conceptual representation in bilingual memory: Effects of concreteness and cognate status in word association. Bilingualism: Language and Cognition 1 (03), 193–211. Van Hell, J. and Dijkstra, T. (2002) Foreign language knowledge can influence native language performance in exclusively native contexts. Psychonomic Bulletin and Review 9, 780–789. Van Hell, J.G. and Tanner, D. (2012) Second language proficiency and cross-language lexical activation. Language Learning 62 (2), 148–171. Vanhove, J. (2014) Receptive multilingualism across the lifespan. Cognitive and linguistic factors in cognate guessing. PhD thesis. University of Fribourg (Switzerland). Van Lier, L. (2000) From input to affordance: Social-interactive learning from an ecological perspective. In J.P. Lantolf (ed.) Sociocultural Theory and Second Language Learning (pp. 245–259). Oxford: Oxford University Press. VanPatten, B., Williams, J. and Rott, S. (2004) Form-meaning connections in second language acquisition. In B. VanPatten, J. Williams and S. Rott (eds) Form-meaning Connections in Second Language Acquisition (pp. 1–26). Mahwah, NJ: Erlbaum. Vidal, K. (2011) A comparison of the effects of reading and listening on incidental vocabulary acquisition. Language Learning 61 (1), 219–258. Vildomec, V. (1963) Multilingualism. Leyden: A. W. Sythoff. Walczak, B. (1994) Wpływy obce w powojennej polszczyz´nie. In S. Gajda and Z. Adamiszyn (eds) Przemiany współczesnej polszczyzny. Materiały konferencji naukowej. Opole: Instytut Filologii Polskiej. Wandruszka, M (1979) Die Mahrsprachingkeit des Menschen. Stuttgart: Kohlhammer. Waszakowa, K. (1994) Tendencje rozwojowe w słowotwórstwie polszczyzny kon´ca XX wieku. In S. Gajda and Z. Adamiszyn (eds) Przemiany współczesnej polszczyzny. Materiały konferencji naukowej. Opole: Instytut Filologii Polskiej. Waszakowa, K. (1995) Dynamika zmian w zasobie leksykalnym najnowszej polszczyzny. Poradnik Je˛zykowy (3), 2–12. Waszakowa, K. (2005) Przejawy internacjonalizacji w słowotwórstwie współczesnej polszczyzny. Warszawa: Wydawnictwa Uniwersytetu Warszawskiego. Wei, L. (2000) Dimensions of bilingualism. In L. Wei (ed.) The Bilingualism Reader (pp. 3–25). London and New York: Routledge. Weinreich, U. (1953) Languages in Contact. The Hague: Mouton and Co. Publishers. Weinreich, U. (1964) Languages in Contact: Findings and Problems. The Hague: Mouton. Wełna J. (1977) Deceptive words. Papers and Studies in Contrastive Linguistics 7, 73–84. Wenden, A. and Rubin, J. (1987) Learner Strategies in Language Learning. Englewood Cliffs, NJ: Prentice Hall. Whitley, M.S. (2002) Spanish/English Contrasts: A Course in Spanish Linguistics. Washington, DC: Georgetown University Press. Whorf, B.L., Lee, P., Levinson, S.C. and Carroll, J.B. (2012) Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf. Cambridge, MA: MIT Press. Widdowson, H.G. (1997) EIL, ESL, EFL: Global issues and local interests. World Englishes 16 (1), 135–146. Wierzbicka, A. (1988) The Semantics of Grammar. Amsterdam: John Benjamins. Wilczyn´ska, W. (2007) Wieloje˛zycznos´ć – przegla˛d problematyki w uje˛ciu dydaktycznym. In M. Jodłowiec and A. Niz˙egorodcew (eds) Dydaktyka je˛zyków obcych na pocza˛ tku XXI wieku (pp. 27– 40). Krakow: Wydawnictwo Uniwersytetu Jagiellon´skiego.

References

283

Williams, S. and Hammarberg, B. (1998) Language switches in L3 production: Implications for a polyglot speaking model. Applied Linguistics 19, 295–333. Winford, D. (2003) An Introduction to Contact Linguistics. Oxford: Blackwell Publishing. Winford, D. (2010) Contact and borrowing. In R. Hickery (ed.) An Introduction to Contact Linguistics (pp. 170–187). Oxford: Blackwell Publishing. Witaszek-Samborska, M. (1992) Wyrazy obcego pochodzenia we współczesnej polszczyz´nie (na podstawie słowników frekwencyjnych). Poznan´: Wydawnictwo Nakom. Witaszek-Samborska, M. (1993) Zapoz˙yczenia z róz˙nych je˛zyków we współczesnej polszczyz´nie. Poznan´: Wydawnictwo Poznan´skiego TPN. Wodniecka, Z., Craik, F.I.M., Luo, L. and Bialystok, E. (2010) Does bilingualism help memory? Competing effects of verbal ability and executive control. International Journal of Bilingual Education and Bilingualism 13, 575–595. Wrembel, M., Gut, U. and Grit, M. (2010) Phonetics/phonology in third language acquisition (Special Issue on L3 phonology). International Journal of Multilingualism 7 (1), 1–3. Yelland, G.W., Pollard J. and Mercuri, A. (1993) The metalinguistic benefits of limited contact with a second language. Applied Psycholinguistics 14, 423–444. Zabawa, M. (2007) Language contact and foreign language teaching. In J. Arabski (ed.) On Foreign Language Acquisition and Effective Learning (pp. 89–102). Katowice: Wydawnictwo Uniwersytetu S´laskiego. Zabawa, M. (2008) English-Polish language contact and its influence on the semantics of Polish. In A. Ka˛tny (ed.) Kontakty je˛zykowe i kulturowe w Europie/Sprach- und Kulturkontakte in Europa (pp. 154–164). Gdan´sk: Wydawnictwo Uniwersytetu Gdan´skiego. Zabawa, M. (2009) The use of English elements in Polish commercials and press advertisements: Success or failure? Linguistica Silesiana 30, 185–194. Zabawa, M. (2012) English Lexical and Semantic Loans in Informal Spoken Polish. Katowice: Wydawnictwo Uniwersytetu S´la˛skiego. Zybatow, L. and Zybatow, G. (2002) Die Eurocom-Strategie als Weg zur europäischen Mehrsprachigkeit: EuroComSlav. In D. Rutke (ed.) Europäische Mehrsprachigkeit: Analysen – Konzepte – Dokumente (pp. 65–96). Aachen: Shaker.

Index

affordance, 120, 128–133, 143, 144, 147, 150, 154, 155, 162–164, 203, 224, 233–234, 238, 239, 240 Aronin, Larissa, 4, 14–15, 20–24, 118, 122, 126, 129–132, 162–163, 176, 214, 225, 234, 235 assumed similarity, 64, 66, 70 attitudes towards bi-/multilingualism, 12, 18, 19–20, 23–28, 32–34

Common European Framework of Reference (CEFR), 16, 145–146, 166, 168, 176, 242 conceptual transfer, 69–71 contact-induced language change, 31–32, 46, 59, 202 Contrastive Analysis Hypothesis (CAH), 60–61, 66, 84, 91, 94 Cook, Vivian, 4–10, 17, 19, 62, 115 crosslinguistic influences (CLI), xiii, 36, 59–63, 66–67, 70–71, 93–95, 97, 102–105, 112–114, 123–125, 133, 138, 155, 175, 193, 214–215, 233 crosslinguistic similarity, xi–xiii, 31, 42, 43, 48, 52, 59, 60, 63–64, 67, 71, 88, 90–92, 94–96, 104, 106, 118, 120, 122–124, 126, 128–129, 131, 143– 144, 146–151, 162–164, 176,177, 178, 191, 193–195, 198, 201, 214–215, 217, 221, 223–225, 227, 232–234, 239–241, 245 cumulative (language) learning experience, 106, 109, 114, 117, 120, 124, 131, 144–145, 147, 150, 152, 154, 155, 159, 163, 165, 166, 172, 174, 176, 196, 214, 218, 221, 233, 236, 238, 242

Baetens Beardsmore, Hugo, 8–10, 14, 116 Berthele, Raphael, 47, 51, 103, 124, 176, 214 BIA+ Model, 79 Bialystok, Ellen, 115–116, 118, 121, 175, 193, 216 bilinguality, 5, 11, 14 Bloomfield, Leonard, 6–7, 14, 19 borrowing, 12, 27, 31–36, 37–41, 44, 45, 59, 61, 65, 68, 84, 91, 104, 137, 151, 158, 160, 182, 183–184, 193, 202, 238, 245 borrowing transfer, 59 Cenoz, Jasone, 11–14, 62, 104, 116, 166 Chamot, Anna, 120–121, 197–198, 203, 260 code-mixing, 59 code-switching, 23, 32, 59–60 cognate (vocabulary/words), xi–xiii, 43–46, 47–56, 60, 64–68, 72–82, 83–93, 94–96, 98–101, 103, 105, 106–108, 111, 114, 116, 118, 120, 124–126, 128, 130–131, 133, 137–142, 143, 145–165, 167–172, 174–177, 178–185, 189, 191, 193–201, 203–204, 206–212, 214–215, 216–234, 236–245 cognate advantage, 76–78, 81, 100 cognate facilitation effect, xiii, 76–78, 90, 95, 100, 106–109, 116, 175

De Angelis, Gessica, 10–11, 62–63, 95, 104, 112–113, 121, 124, 126, 162, 176, 214, 225 de Bot, Kees, 6, 10, 44, 46–47, 68, 109 De Groot, Anette, 44, 62, 75–78, 80–82, 95–100, 164 deceptive cognate, 46–47, 48, 50 degree of (cognate) similarity, 50–53, 96, 167 Dewaele, Jean-Marc, 11, 62, 68, 99, 102–103, 113–114, 131–133, 234 284

Inde x

Dijkstra, Ton, 44, 46, 50–51, 73, 76–79, 95–102, 106, 108, 115, 117, 175 Dominant Language Constellations, 24, 137, 235–236, 244 Dörnyei, Zoltan, 23–24, 27, 91, 121, 132, 139, 179, 215, 225–226, 243 Dynamic Model of Multilingualism, 9 Ecke, Peter, 66, 113, 124–125, 174–175, 194 English as a Lingua Franca (ELF), 21–23, 137, 235, 245 false friend, xii, 42, 46–50, 60, 64, 66–68, 72, 76–79, 83, 88, 91–92, 101–102, 107, 125, 138, 140, 143, 158, 165–172, 174–176, 194, 243 Fisiak, Jacek, 37–39, 41 Gabrys´-Barker, Danuta, 11, 28, 62, 117, 128, 162, 214, 217 Gibson, James, 129 Gooskens, Charlotte, 84 Green, David, 99 Grosjean, Francois, 6, 8–9, 19, 99, 114 Hall, Christopher, 43, 68, 103, 113, 124–125, 174 Hammarberg, Björn, 12–13, 62, 68, 104, 112, 114, 166 Haugen, Einar, 7, 10–11, 14, 33–36, 46 Hufeisen, Britta, 16, 21, 84, 90–91, 102, 104–105, 113, 122, 236, 242 individual bilingualism, xi–xii, 5–6, 7–8, 10–13, 16–17, 18–20, 32, 80, 73, 84, 115–116, 118, 141, 150, 154, 162 monolingualism, 4–6 multilingualism, xi–xii, 4, 6–7, 9, 10, 11–13, 14, 15–17, 18–19, 23, 24, 25, 28, 60, 62, 84, 116, 120, 126, 138, 140–141, 143–145, 149–150, 154, 162–165, 168, 172–174, 178, 197, 235–237, 241–242 intercomprehension, 24, 84, 90–91 interference, 8, 31, 59–61, 68, 70 interlingual homograph, 46, 72, 76–78 internationalism, 12, 37–38, 41, 43–46, 84, 161, 191, 243 item transfer, 64–66

285

Jarvis, Scott, 19, 43, 47–48, 60, 62–64, 66–71, 82, 83, 93, 97, 103, 112, 175, 193–194, 217, 224, 233 Jessner, Ulrike, 4, 9–12, 14–15, 19, 21, 23, 62, 116–118, 121–124, 126, 162–163, 176, 214, 225, 238 Kaufman, Terrence, 31, 33–34, 59, 61 Kellerman, Eric, 61, 63, 71, 87, 93, 95, 102–104, 106, 125–126, 128, 175–176, 213 Komorowska, Hanna, 15, 24–26 Kroll, Judith, 6, 10, 62, 74, 76–77, 80–81, 98, 100, 109, 115 L2 status, 103, 106, 112–114, 124, 196 Lado, Robert, 46, 50, 60, 84, 91 language awareness, 106, 117, 126–128, 131, 133, 163, 242 language contact, 12–13, 26, 31–34, 36, 42, 43, 83 language prestige, 12, 18, 26, 32, 37, 42, 55, 244 language typology, 43, 52, 63, 71, 101–102, 114, 137, 144, 162, 178 language user, 3–5, 17, 20, 32, 44, 49, 51–52, 62, 67, 69, 71–72, 75, 80, 99, 105, 110, 114–115, 120, 131–132, 164 learning strategy, 89, 120, 126, 137–140, 178, 193, 195, 198–201, 203–204, 206, 213–215, 239–240 Lemhöfer, Kristin, 44, 46, 50, 76, 78, 96, 98, 116, 164 lemmatic transfer, 67–68 Levenshtein Distance/LD (Normalised Levenshtein Distance/NLD), 50–54, 140, 147, 151, 160–161, 164, 165, 167–168, 171, 172, 174–176, 199, 207, 219–220, 228–229, 237 lexemic transfer, 67–68, 82 lexical borrowing, 34, 40–41, 244 lexical borrowings into English, 38–39 lexical borrowings into Polish, 39–41 loanword, 35, 38, 41, 45, 55, 137 Man ´ czak-Wohlfeld, Elz ˙bieta, 26, 41, 138 Meara, Paul, 85, 87, 91, 97, 111, 164 mental lexicon, xiii, 67–68, 72–74, 76, 78–81, 105, 106, 109–110, 124–125, 192

286

Cognate Vocabul ar y in L anguage Acquisit ion and Use

metalinguistic awareness, 106, 117–119, 120–126, 131–132, 138, 151–152, 162–163, 168, 176, 178, 193, 195, 202, 214, 216, 221, 225, 234, 238–241 Mixed-Representational Model, 75, 77 monolingual bias in SLA, 19 Multicompetence, 8–9, 62 multilingual language acquisition (MLA), xii–xiii, 12–13, 60 multilingualism with English, xii, 23, 26 multilinguality, 11, 14, 16, 18, 28, 92, 124, 142, 165–166, 169, 172, 174, 176 Nation, Paul, 85, 87 native speaker, 3–5, 7, 9–10, 19, 22, 23, 88–89, 113, 115 noticing, 91, 93, 102, 104, 106, 120, 125–133, 137–138, 144, 163, 174–176, 179, 193, 197, 201–203, 208, 215, 219, 221–225, 228, 231–232, 234, 237–238, 241, 245 Odlin, Terence, 45–46, 59–64, 66, 70–71, 83–84, 87, 93, 102, 104, 112, 125, 128, 131, 175, 193 orthographic similarity (of cognates), 50–53, 72, 79, 95–96, 135, 140, 143, 145, 147, 159–161 Oxford, Rebecca, 120–121

Schepens, Job, 44, 50–52, 64, 87, 101, 125, 175–176 Schmidt, Richard, 127, 214 Schmitt, Norbert, 44, 87, 97, 128 second language acquisition (SLA), xi, 12–13, 83, 85, 91–92, 96–97, 112–116, 127, 139, 165 Segalowitz, Norman, 110–111,129–131 Seidlhofer, Barbara, 22, 244 self-regulation (self-regulated learning), 121, 215, 240, 164–165, 175, 240 semantic loan, 35, 38 Sharwood Smith, Michael, 61 Singleton, David, 4, 15, 20, 22, 24, 87–88, 97, 100–101, 104–105, 109, 112, 118, 126, 128–129, 131–132, 162–163,176, 214, 225, 234, 235 societal bilingualism, 5, 13–16, 18, 20, 23, 32, 34, 39–40 monolingualism, xi, 5, 20, 26 multilingualism, 13–16, 18–19, 20–21, 23–25, 28, 34, 242 spread of English, 12, 22–23 substratum transfer, 59 system transfer, 64–66

Paradis, Michel, 113, 120, 128, 214–215 Pavlenko, Aneta, 19, 26, 43, 49–50, 60, 62–63, 67, 69–71, 93, 97, 103, 112, 217, 224 perceived similarity, 64, 66 plurilingualism, 4, 10, 15–17, 21, 137–138, 235, 238, 242–245 positive transfer, 60, 65–66, 86, 239 psychotypology (psychotypological factors/distance), 63, 71, 88, 101, 103–105, 106, 112–113, 125, 132, 140, 143–144, 146–148, 162, 176, 216, 243

Thomason, Sarah, 31–34, 59, 61 transfer, xiii, 23, 32, 35, 59–71, 82, 86, 90–91, 97, 102, 104–105, 112–114, 122–125, 176, 194, 198, 203–204, 206–207, 239 transferability, xiii, 91, 93, 95–98, 102–104, 112, 175–176, 217, 224 translation equivalent, 44–45, 49, 51, 77, 79, 81, 96, 98 translation experiment, 74, 96

Revised Hierarchical Model (RHM), 74–77, 109 Ringbom, Hakan, 10, 36, 45, 47–48, 50, 52, 62–67, 83, 102–103, 104, 112, 116, 122–123, 125, 175–176, 192–193, 238 Rusiecki, Jan, 36, 41, 46–47, 50, 138

Van Hell, Janet, 44, 62, 77–78, 80–82, 96, 98, 100, 106–110, 115–116 Van Heuven, Walter, 78–79 Vanhove, Jan, 50, 84, 96, 119, 175, 216 Weinreich, Ulrich, 7, 10, 14, 33–34, 60