Within Language, Beyond Theories (Volume II) : Studies in Applied Linguistics [1 ed.] 9781443879804, 9781443874571

Citation preview

Within Language, Beyond Theories (Volume II)

Within Language, Beyond Theories (Volume II) Studies in Applied Linguistics Edited by

Anna Bloch-Rozmej and Karolina Drabikowska

Within Language, Beyond Theories (Volume II): Studies in Applied Linguistics Edited by Anna Bloch-Rozmej and Karolina Drabikowska Reviewed by Dr. hab. Bożena Cetnarowska (University of Silesia) Dr. hab. Bogusław Marek (John Paul II Catholic University of Lublin) This book first published 2015 Cambridge Scholars Publishing Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2015 by Anna Bloch-Rozmej, Karolina Drabikowska and contributors All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-7457-4 ISBN (13): 978-1-4438-7457-1

TABLE OF CONTENTS

Foreword .................................................................................................. viii Contributors .............................................................................................. xiii Part I: Language Learning and Teaching Chapter One ................................................................................................. 3 Extending Foreign Presence in English Phonetics Classes Anita Buczek-Zawiáa Chapter Two .............................................................................................. 23 Differences and Similarities in the Perception of German Vowels: The Case of Turkish and Polish Learners of German as a Foreign Language Katharina Nimz Chapter Three ............................................................................................ 34 Effects of Input Condition on Case Ending Processing in Initial Polish L2 Jacopo Saturno Chapter Four .............................................................................................. 55 Contemporary Learning Environments: A Call for Redefining Traditional Strategy Frameworks Kamila BurzyĔska Chapter Five .............................................................................................. 74 Testing Collocational Knowledge: The Question of Item Format Wojciech Malec Chapter Six ................................................................................................ 93 Conversational Convergence in an L2 Exam Setting Piotr Steinbrich

vi

Table of Contents

Chapter Seven.......................................................................................... 109 The Role of Pragmatics in Business English Courses: A Case Study ElĪbieta Jendrych Chapter Eight ........................................................................................... 123 Quality Issues in Bilingual Education Gail Taillefer Chapter Nine............................................................................................ 145 Learning Through CLIL in Polish Primary Education Artur StĊpniak Part II: Psycho- and Sociolinguistics Chapter Ten ............................................................................................. 157 Paradigm Acquisition and Conjugational Patterns in Hungarian Child Language Viola Baumann Chapter Eleven ........................................................................................ 177 Two Kinds of Onomatopoeias Maruszka Meinard Chapter Twelve ....................................................................................... 189 The Influence of Physical Attractiveness on Foreign Accent Rating Dominika ĩabiĔska Chapter Thirteen ...................................................................................... 203 Beyond Our Dreams, Within Our Means: Saving the Objects of Our Study Jerzy Warakomski Part III: Lexicography and Translation Studies Chapter Fourteen ..................................................................................... 217 Metaphor-Based Structure in a Dictionary of Idioms Jarosáaw WiliĔski

Within Language, Beyond Theories (Volume II)

vii

Chapter Fifteen ........................................................................................ 231 A Profile of Interdisciplinary Professional Dictionaries Published between 2007 and 2012 Agnieszka Rzepkowska Chapter Sixteen ....................................................................................... 252 Spanish Functionalism Viewed through Hispanic University Manuals and Dictionaries Cristina García González Chapter Seventeen ................................................................................... 274 A New Trend in the Translation of Polysemous English Words into Latvian: Semantic Hypertrophy of the Basic Equivalent Juris Baldunþiks Chapter Eighteen ..................................................................................... 289 Heinrich Heine’s “Lore-Ley” (1824): A Culturological and TranslationRelated Analysis of the Means of Expressivity Virginia Schulte Index ........................................................................................................ 309

FOREWORD

The monograph Within Language, Beyond Theories presents a collection of insightful studies pertaining to the most perplexing problems in the areas of theoretical and applied linguistics. Contributors offer accounts of new evidence drawn from a number of the world’s languages and analyses that surpass the limits of contemporary frameworks in search of more explanatorily adequate solutions to linguistic dilemmas. We delve into previously unexplored areas of linguistic reality, aiming to gain insight into the structure of the system and establish laws governing its inner organization. Importantly, linguists of different persuasions share the belief that our enhanced understanding of the grammar of language and its constituent modules will foster new advances in the novel application of the models proposed. Assisted by innovative ideas in corpus studies, translators and discourse researchers will be able to make invaluable contributions to the development of their fields. Volume Two, entitled Studies in Applied Linguistics, comprises eighteen chapters organized into three parts. The articles reflect current trends in scientific debate and offer solutions surpassing the boundaries of traditional approaches to applied linguistics. Authors focus on the most intriguing themes in the fields of language learning, language teaching, psycho- and sociolinguistics as well as lexicography and translation studies, thus pointing out directions for future inquiry and international scientific enterprise. Part I: Language Learning and Teaching Chapter One addresses the issues and challenges involved in teaching phonetics to non-L1-homogeneous classes. It discusses the phonetics and phonology of Polish, Spanish and Turkish, the knowledge of which is needed in order to predict the potential problems that a foreign student might encounter while learning English pronunciation. Moreover, it demonstrates how helpful such awareness can be in phonetics classes. In a similar vein, Chapter Two investigates the perception of German vowels by Polish and Turkish students. The means to establish the differences in vowel perception was an experiment involving Turkish and Polish students. The length of German vowels in nonsense words was

Within Language, Beyond Theories (Volume II)

ix

digitally manipulated so that the ability to distinguish between the short and long counterparts of same-quality vowels could be tested. As a result, the perceptive differences and similarities between these three languages were established, which also has a bearing on the teaching of phonetics. Chapter Three examines two types of instructions in language learning, that is, meaning- and form-based. An experiment was conducted involving 14 learners of Polish – a highly inflected language – and the acquisition of case endings was analysed. The study presents the error distributions at different stages of learning and the correlation between the type of instruction and the parameters: Ending, Word Order and Lexical Transparency. Strategies involved in language learning are also discussed in Chapter Four. However, a different perspective is taken since the study focuses on students’ use and preferences with regard to Computer Assisted Language Learning (CALL), in particular, learning, memorising, compensation, metacognitive, social and affective strategies. Moreover, the results of a survey conducted among 118 participants show correlations between some of the strategies. Chapter Five shifts the focus of Part I from language learning to testing. It is devoted to an essential part of test development, that is, task/item design. Item formats typologically categorised by Hudson (2002) as selected response and constructed response items are scrutinised with relation to their assessment effectiveness. Chapter Six explores the issue of conversational convergence among advanced non-native speakers. Its goal is to determine the ability of students at a high level of language proficiency (C2) to maintain dialogic relationships. The chapter presents an analysis of 23 conversations between Polish students of English gathered during their oral exams. What seems to distinguish their dialogues is the fact that their structure differs extensively from that of proficient users in discourse and lexicogrammatical terms. The chapter demonstrates how often such natural interaction dialogue features as repetition, backchannelling, tails, tags, and vague and hedged forms are utilised by advanced non-native speakers. Chapter Seven is concerned with the role of pragmatics in the business English course syllabus. Its primary aim is to demonstrate how pragmatic knowledge may contribute to the effectiveness of such courses with the focus on the skills that are necessary for future professionals. The chapter proposes specific elements that should be incorporated into business English courses because of their influence on students’ future performance.

x

Foreword

Chapters Eight and Nine offer interesting insights into Content and Language Integrated Learning (CLIL). Chapter Eight adresses the problem of quality in bilingual education (BE) with special emphasis put on three challenges, that is, national context and cultural self-knowledge, language policy and practices as well as the process of change management. It also stresses the role of teachers and institutions in creating a high-quality BE environment. Chapter Nine focuses on CLIL in Polish primary schools. The chapter provides a presentation of the goals of CLIL and its implementation with special attention on the effectiveness of the top-down and bottom-up approaches in the Polish educational environment. Part II: Psycho- and Sociolinguistics Chapter Ten takes up the theme extensively discussed in Part I, that is, language learning, however, it shifts its attention towards first language acquisition. Specifically, it is concerned with conjugation and declension in Hungarian. The system proposed in this study is a set of paradigmatic groups. The regular and irregular forms are here replaced by stronger and weaker features, which are activated when new input appears. The proposed system is supported by a Wug-test, containing both nouns and verbs and aiming at eliciting paradigmatic and derivational forms. Chapter Eleven takes a psycholinguistic perspective on onomatopoeias. It distinguishes two kinds: imitative and echoic ones, basing the categorisation on research in semantics, semiotics, phonetics, acoustics and neurology. Chapter Twelve investigates the impact that physical attractiveness has on the perception of a foreign accent. In order to establish the influence, an experiment was conducted whereby the attractiveness of speakers was manipulated according to the beauty stereotype. The results show what influence physical and vocal attractiveness has on male and female listeners. Chapter Thirteen explores the issue of the endangerment of the Vilamovian ethnolect. General problems such as reasons for language extinction and methods of revival are viewed from the perspective of a specific case, that is, a language spoken by a small community situated in the eastern part of Silesia.

Within Language, Beyond Theories (Volume II)

xi

Part III: Lexicography and Translation Studies Chapter Fourteen is concerned with the practical application of the Conceptual Theory of Metaphor and Metonymy and Frame Semantics in lexicography. In particular it assesses the applicability of conceptual metaphor, metonymy and conventional knowledge to idiomatic expression and proposes a set of criteria for designing a dictionary of idioms. Chapters Fifteen and Sixteen aim at examining the contents of dictionaries that are already available. Chapter Fifteen offers an insight into interdisciplinary dictionaries published between 2007 and 2012. The profile emerging from this analysis comprises such general elements as subject, volume, terminographic technique, foreign equivalents, information about the corpus of texts and their main purpose as well as detailed features, namely volume, type of microstructure, level of symmetry, presence of indices or other attachments, metalanguage, a bibliography and information about the sources and the thematic scope. Chapter Sixteen also explores specialist dictionaries, however, its focus is on how manuals and dictionaries of linguistics make reference to Spanish functionalism. It establishes which basic tenets of Spanish functionalism, as the most fundamental pieces of information, should be included in entries and with respect to this it analyses the corpus of the most prestigious linguistic sources. Chapter Seventeen addresses the problem of the translation of polysemous English words into Latvian. The main focus here is on the semantic hypertrophy of the basic equivalent that seems to have a considerable influence on translational habits in present day Latvian. Chapter Eighteen investigates the translations of Heine’s “Lore-Ley.” It examines the linguistic means of expressivity and emotionality that contribute to the effect of the demonization of women in both the original and its Polish translations. They are recognized in the form itself, that is, bound language, its expressiveness, particular word choices as well as at the sound level. The present volume constitutes a means to report on research in progress that is being conducted by linguists of different persuasions, yet rooted in the field of applied linguistics. It captures the latest advances in this branch of language studies, illuminating both universal and languagespecific patterns and variety in the structure and use of natural languages. The ideas and solutions formulated on the pages of this book are intended to both enrich and stir scientific debate in applied linguistics. We wish to express our thanks to all the Contributors to this volume for sharing with us their innovative ideas and the results of their research findings. We are

xii

Foreword

also grateful to Maria Bloch-Trojnar, Konrad Klimkowski, ElĪbieta Sielanko-Byford and Nigel Byford for valuable comments on various aspects of this monograph. Anna Bloch-Rozmej and Karolina Drabikowska

CONTRIBUTORS 1.1.1. Juris Baldunþiks Ventspils University College, Latvia [email protected] Viola Baumann Hungarian Academy of Sciences, Budapest, Hungary [email protected] Anita Buczek-Zawiáa Pedagogical University of Cracow, Poland [email protected] Kamila BurzyĔska Maria Curie-Skáodowska University, Lublin, Poland [email protected] Cristina García Gonzáles Universidad de León, Spain [email protected] ElĪbieta Jendrych KoĨmiĔski University, Warsaw, Poland [email protected] Wojciech Malec John Paul II Catholic University of Lublin, Poland [email protected] Maruszka Meinard Paris 13 University, France [email protected]

xiv

Contributors

Katharina Nimz University of Potsdam, Germany / Newcastle University, United Kingdom [email protected] Agnieszka Rzepkowska Siedlce University of Natural Sciences and Humanities, Poland [email protected] Jacopo Saturno Università degli Studi di Bergamo, Italy [email protected] Virginia Schulte University of Warsaw, Poland [email protected] Piotr Steinbrich John Paul II Catholic University of Lublin, Poland [email protected] Artur StĊpniak John Paul II Catholic University of Lublin, Poland [email protected] Gail Taillefer Toulouse 1 Capitole University, France [email protected] Jerzy Warakomski Foreign Language Teacher Training College in Puáawy, Poland [email protected] Jarosáaw WiliĔski Siedlce University of Natural Sciences and Humanities, Poland [email protected] Dominika ĩabiĔska University of Silesia, Katowice, Poland [email protected]

PART I: LANGUAGE LEARNING AND TEACHING

CHAPTER ONE EXTENDING FOREIGN PRESENCE IN ENGLISH PHONETICS CLASSES ANITA BUCZEK-ZAWIàA

1. Introduction Whether one believes in the idea of the variety known as English as an International Language or not, the essential questions facing, among others, English pronunciation teachers are those related to diagnosing and prioritizing potential problems in pronunciation training. This concerns not only establishing problem areas in the prevailing English (foreign) – Polish (native) contrastive context, but it also needs to acknowledge the increasing presence of foreign students of English in practical phonetics and phonology classes in Poland, especially at the university level for philological courses. This presence is a direct consequence of the internationalization of university studies throughout Europe, including Poland. The consequence of all this is that we no longer have linguistically homogenous classes consisting of Polish Students of English (Philology), but more and more frequently we need to accommodate participants from other countries, with their tendencies to transfer their own sound systems onto English. Both Polish and foreign students are somewhat inhibited by their native articulatory habits and perceptive abilities, yet to different degrees and, possibly, in different areas. This chapter addresses these issues, trying to look for possible solutions or alternatives, based on the author’s experience with (mostly but not exclusively) Spanish and Turkish students of English spending one or two semesters in Poland and also of actually teaching pronunciation classes in the students' home institutions. Extending a teachers’ factual knowledge contrastively, resulting in a growing familiarity with the phonological systems of languages other than Polish or English, will

4

Chapter One

consistently contribute to not so much changing the curriculum dramatically but rather to varying its components and shifting its focus to include problem areas generated by the foreign students’ presence. The idea advocated here is, fundamentally, that of using this foreign presence to increase the phonetic/phonological awareness of all students and to actually facilitate the acquisition of good pronunciation patterns.

2. The target of practical phonetics courses Unlike participants in various extra-curricular language courses, students of English philology on the whole treat English professionally. They are studying it not only or not primarily for practical, communicative or utilitarian purposes, but mainly to use it as a medium in their (prospective/selected) professional career, be it as teachers or interpreters or in any other field. Thus, they simply do not fit into the category of target users of Jenifer Jenkins (2000) Lingua Franca Core, focused as it is on those features of English speech which are considered indispensable for communicative intelligibility between non-native users. Naturally, they still perceive and appreciate the value of English as a tool of international communication. In that context, actually less and less insensitivity among English philology students (and, thankfully, among an increasing number of university lecturers) towards the way their pronunciation deviates from an acceptable standard is observed. They are beginning to realize that being a competent, proficient language user entails gaining or developing that proficiency in all language components or planes. Thus, they begin to define their priorities not only in the fields of grammar or lexis, but they also want to be reasonably native-like in the area of speech, which is always a long-term process. Still, the study of the practical phonetics of a foreign language is commonly considered an unnecessary burden and a subject that puts unreasonable demands on both students and their teachers1 (BuczekZawiáa 2011). Whereas a conditio sine qua non of good pronunciation is an awareness of the normative regulations concerning a language’s orthophony, that is, an awareness of how exactly the target language ought to sound. This becomes particularly significant for those students who plan 1

The outline of the situation as well as the reasons behind it are elaborated on in Buczek-Zawiáa (2011). Baran-àucarz (2006) adds here the inadequate amount of time allocated to pronunciation training in the course curriculum for philology, ranging between 60 to 120 hours at most.

Extending Foreign Presence in English Phonetics Classes

5

on having a professional career in areas connected with talking: teaching or interpreting. As professionals they ought to possess all the necessary competences and be good models to be emulated by their own students or colleagues (Baran-àucarz 2006, 7). The goal of a nearly-accent-free oral competence may be unattainable to all but to a good number of L2 acquirers it will prove possible. Therefore, phonetics must be viewed as an important element of a quality education (Buczek-Zawiáa 2011, 326). Sometimes the trainees are not fully aware of their own pronunciation deficiency and need to have their attention drawn to any deviations. They form a fairly homogeneous group in that respect, being rather more tolerant of errors than can be accepted. That is true for those whose first language is Polish, that is, our regular students, as well as for those with other mother-tongue backgrounds, like those from Spain or Turkey. However, through studying practical phonetics, all of them become much more critical evaluators of their own performance. Minimally, then, for students of English Philology, the curriculum of a course in practical phonetics ought to incorporate: x Explicit training in segmental phonetics, both for the vocalic and the consonantal inventories and with significant vacillations or (positional) variants. x Instruction in suprasegmental features, including stress, rhythm and intonation. x Familiarizing students with aspects of connected speech, such as assimilations, elisions and liason. x The rudiments of phonetic/phonological theory that would facilitate the acquisition of good speech habits and progress in pronunciation. Needless to say, this content is frequently subject to time/space limitations. Nonetheless, the attainment of a commendable level of target language proficiency must involve learning the correct pronunciation of segments, words and whole utterances. Native-like pronunciation – alongside structural accuracy, fluency, sociolinguistic and pragmatic competence, knowledge of the realia etc. – is a logical objective of advanced foreign language learning, particularly expected of prospective teachers. Trainees whose grammar and vocabulary are native-like but whose foreign accent is significantly marked are, sadly, a much more typical end-product of teacher preparation programs.

6

Chapter One

3. Instructor’s competence in a non-L1-homogenous group After Poland’s accession to the European Union and the acceptance of the Bologna Process standards for higher education, it became possible for students of Polish universities to go and study abroad for a semester or two in partner institutions within the Erasmus exchange program. The traffic, however, was/is not one-way: an increasing number of students from various European countries come to spend part of their course of studies in Poland. This is chiefly observed in Neo-philological Departments, where the visitors find courses running parallel to those that they experience in home institutions. Characteristically, these foreign students try to choose as many courses in the module of Practical Language Skills as possible, and that includes classes in pronunciation. Observations over the years indicate that among young Europeans two nations seem to be particularly mobile and willing to come and study in Poland: the Spaniards and the Turks. Occasionally, Italians and Lithuanians crop up in larger quantities. This situation creates new challenges for instructors, especially pronunciation trainers. The curriculum and, accordingly, the techniques of teaching and practicing pronunciation skills, have been designed with Polish students in mind, taking into account their characteristic problem areas and making frequent excursions to whatever mother-tongue knowledge and skills can offer. Certain characteristic types of error occur and they may often be grounded in habits transferred from the first language. Such occurrences are by no means restricted to interferences from Polish. Spanish and Turkish students exhibit their own crosslinguistic influences on English. Yet, it appears that it is here that at least a basic familiarity with the students’ native phonologies, treated contrastively with the English targets can be of help. One has to realize, however, that even the most carefully constructed comparisons between two linguistic systems cannot in principle prevent interlingual interference, especially between closely related languages. Such a transfer of acquired skills or knowledge can be positive, when production of acceptable L2 utterances is facilitated by the compatibility of linguistic behaviour in the first and foreign language; or negative, which manifests itself in the errors that learners make (Papierz 2009, 53-54). Characteristically, linguistic confrontation focuses on the important differences, while treating the similarities as redundancies which do not contribute in a significant way to effective language acquisition. The idea of comparing and contrasting languages which are genetically and typologically close may appear a rather vacuous activity, since one can expect more similarities than differences between them. The important

Extending Foreign Presence in English Phonetics Classes

7

aspect is the direction in which the comparison is made, as the language that constitutes the basis for such a juxtaposition will in a way be more privileged in its description (Papierz 2009, 50). The popular belief seems to be that since mistakes resulting from transfer arise where the two systems of languages are similar but not identical, they are most common in the interlanguage of people whose L1 is closely related to English (mostly within the Indo-European context), whereas speakers of unrelated languages (Chinese, Arabic) actually have fewer problems with transfer effects but more with those connected to the intrinsic difficulty of certain aspects of English (Swan and Smith 2001, xi). When attention is shifted to the phonetic-phonological plane, one has to recognize the more confrontative bias in the analysis. Phonetically, we examine the articulatory, physiological properties of segments in the compared systems as well as the specificity of prosodic means (such as stress, quantity, intonation). The differences at this level can be rather dramatic. In contrastive phonology the basic unit of analysis is the phoneme as a unit of sound structure serving the purpose of perceiving, recognizing and distinguishing between the existing units of meaning, that is morphemes. Such an analysis of phonological systems may proceed in five steps: defining the phonemic inventories in the compared languages, identifying the equivalents between the two phonemic systems, listing the existing variants, establishing distributional restrictions for phonemes and allophones and defining the frequency of occurrence of each phonemic opposition in the compared languages (Papierz 2009, 51). It would seem logical to assume that the instructor who has the minimum realization of what may be expected from all of his/her students in terms of predictable deviations from the standard is in a favourable position, since they can now identify new priorities or re-organize the course organization in terms of focus, emphasis and technical repertoire. This stand is further corroborated by a statement from Sobkowiak (2004, 19-20): learners of differing linguistic backgrounds acquire English differently; they have their specific problems (and successes), their particular paths of progress, their characteristic errors. A Pole needs different advice and guidance in English than other learners. Some knowledge of what types of errors are characteristically Polish (Spanish, Turkish, Russian etc.), and why, is thus a necessary background for offering remedies.

8

Chapter One

4. Polish, Spanish and Turkish pronunciation learners A closer inspection of the phonological and phonetic facts related to Polish, Spanish (of the mainland European variety) and Turkish reveals certain striking similarities between these systems. However, the differences between them and English are frequently equally surprising. Let us give a brief overview of those facts.

4.1. Polish Apparently, according to ĝpiewak and GoáĊbiowska (2001, 162), the greatest challenge for Polish speakers of English is connected with the inconsistencies between the spelling and the pronunciation of English, since Poles expect every letter in a word to be pronounced, a fact so much characteristic of their own language. The problem is compounded by the flexible stress placement rules in English, especially when compared to the regular penultimate accent in Polish words (with only several, again regular, exceptions). These two factors are most clearly observable when one notes the consistent use of the full, unreduced forms of all words in the speech of Polish learners of English: no vowel reduction is found in unstressed syllables, they tend to receive full, frequently spellinginfluenced articulation. Polish has six oral monophthongs and two nasal vowels. With no length contrast between the vowels, considerable difficulty distinguishing tense vs. lax vowels (bit/beat) is characteristic of Polish users of English (Rogerson-Revell 2011, 286). What follows directly from the above is the fact that numerous instances of phonetic substitution are evidenced in the speech of individuals, for instance [æ] may be confused with [e] or [‫ ]ݞ‬and actually the Polish open-mid vowel [‫ ]ܭ‬may substitute for many distinct vowels. ĝpiewak and GoáĊbiowska (2001, 164) state in a straightforward manner that open vowels are the single most difficult area for Polish speakers, both as regards hearing the differences between them and producing accurate vowel sounds.

In the opinion of many, mistakes in the production of English vowels by Polish learners are considerably more difficult to prevent and eradicate than the mistakes made in the articulation of consonants (Miatluk, Szymaniuk and Turáaj 2008, 29). In an attempt to identify possible or actual problem areas resulting from phonetic interference from the learners’ native language, we need to add two more elements: the non-

Extending Foreign Presence in English Phonetics Classes

9

existence of the so-called “weak vowels” [‫ ]ۑ‬and [ܼ] in Polish, hence the replacement with a full vowel (ĝpiewak and GoáĊbiowska 2001, 164), and the absence of diphthongs in the native vocabulary, with the related tendency to articulate the second element of English diphthongs in a glidal fashion as [j] and [w] (Miatluk, Szymaniuk and Turáaj 2008, 30, 41). Polish has a large and complex consonant system with many place of articulation oppositions. In perception and comprehension, their English (near)counterparts do not present major difficulties for learners, however, when producing them certain persistent deviations can be noticed. Among them is the non-aspiration of voiceless plosives in English since the initial [p, t, k] are not aspirated in Polish (Rogerson-Revell 2011). Additionally, since [ș] and [ð] do not exist in Polish, the first segment may be substituted with [t], [s] or [f], the latter by [d], [z] or [v]. They will always be a problem (at least until reaching a certain substantial level of proficiency) as predicted by the classic assertion of Contrastive Analysis: the degree of difficulty of learning any Target Language (henceforth TL) structure should be equivalent to the degree to which this structure is different from Native Language (henceforth NL). Thus, according to this reasoning, maximum pronunciation difficulty should be represented by learning an entirely new TL contrast (Eckman and Iverson 1997, 188).

Another problem sound is [ƾ], typically mispronounced as [ƾg] word medially or [ƾk] domain finally. The velar nasal in standard Polish functions only as a positional variant of [n]. The palato-alveolar English fricatives and affricates [‫ݕ‬, ‫ݤ‬, ‫ݹ‬, ‫ ]ݶ‬are regularly articulated by Polish speakers as alveolars, which gives them a too harsh quality to the English ear (ĝpiewak and GoáĊbiowska 2001, 165). And because of palatalised consonants in Polish, learners are frequently unable not to palatalize English consonants in a similar context, e.g., before a front vowel as in animal, modelled on nigdy. Word final voiced stops, fricatives and affricates are regularly devoiced (Rogerson-Revell 2011) and that leads to potential confusion between pairs like miss – Ms; eaten – (Martin) Eden or bed – bet. Polish permits complex consonantal clusters and in clusters voicing agreement is observed between obstruents or obstruents and sonorants. This feature is transferred onto English, resulting in un-English combinations such as [‘br‫ܧ‬:tk‫ܤ‬:st] instead of [‘br‫ܧ‬:dk‫ܤ‬:st] or [dܼz’mܼs] for [dܼs’mܼs] (Miatluk et al. 2008, 60-61). As mentioned, word stress, predominantly, is placed on the penultimate syllable – which may cause problems. Sobkowiak (2004, 21) considers errors resulting from word-accent misplacement to be rather

Chapter One

10

serious, not only because they reflect what he terms a “deeply rooted Poglish habit,” but also because they trigger vowel substitution and mispronunciations affecting whole sets of words across the language, whenever the appropriate context arises. Polish falls towards the “syllabletimed” end of the stress vs. syllable-timed continuum, which makes it difficult to acquire and copy the stress-timed rhythm of English. Finally, among the global pronunciation errors which contribute to a foreign accent in the English of Poles is the (initial at least) inability of Polish speakers to link words in connected speech: they do not link final consonants to initial vowels in neighbouring words as in an apron, more oranges further on, they also do not use “linking glides” [j] or [w], as in my eye or two eggs, as these are non-existent in Polish (ĝpiewak and GoáĊbiowska 2001, 163).

4.2. Spanish Coe (2001, 91) reports that speakers of Spanish find English pronunciation harder than speakers of most other European languages. There are many regional and national varieties, generally mutually intelligible, consequently most of the difficulties mentioned here are common across varieties.2 The Spanish and English consonant systems show many similarities, but the vowel systems and sentence phonetics differ significantly and as such can cause problems for Spaniards learning English. The major difficulties are summarized in Coe (2001, 91):    

Difficulties in recognizing and using English vowels Strong devoicing of final voiced consonants Even sentence rhythm, without the typical prominence of English, making understanding difficult for English listeners Narrower range of pitch [...] producing a bored effect

2 The list is compiled on the basis of several distinct sources, most of them devoted to Spanish English contexts (Coe 2001; Rogerson-Revell 2001; Mott 2005; Buczek-Zawiáa and Okas 2008). Some are descriptions of the phonetics and phonology of Spanish (Haáadkiewicz-Grzelak 2006; Nowikow 2012) and are believed to offer vital insights in the area of possible cross-linguistic influence. However, some inconsistencies have been observed in some of these accounts, these will be pointed out when necessary.

Extending Foreign Presence in English Phonetics Classes

11

These difficulties are deeply rooted in the Spanish phonological system. For example, the language possesses the most frequently encountered system of basic vowels /a/, /u/, /i/, /e/ and /o/. HaáadkiewiczGrzelak (2006, 38) refers to it as the “essence of simplicity,” especially in comparison to the seven short /ܼ, e, æ, ‫ݞ‬, ‫ܥ‬, ‫ݜ‬, ‫ۑ‬/ and five long /i:, u:, ‫ܤ‬:, ‫ܧ‬:, ‫ܮ‬:/ vowels of English. The Spanish vowels are higher, tenser and shorter than their English counterparts (Mott 2005, 250). Where length is part of the difference, speakers encounter considerable difficulties in the perception and production of such contrasts. In English, vowel reduction is a phenomenon inducing various phonotactic changes, whereas in Spanish even the vowels in absolute final positions retain their full quality, hence vowel deletion or vowel reduction of any kind are totally absent in Spanish (Haáadkiewicz-Grzelak 2006, 41). The phonotactic restriction that bans some of the short vowels in English from occurring word-finally is also problematic for students. The Spanish vowels preserve their original quality in all contexts and are not, as in English, reduced to schwa when unstressed. Also, they are free to occur in closed or open syllables, word initially or finally. This already points to the prospective difficulties in assimilating the system of English (Buczek-Zawiáa and Okas 2008, 14344). And because schwa [‫ ]ۑ‬does not exist, unstressed vowels are pronounced simply with the “written”, “spelling” vowel (Coe 2001, 94). According to Pamela Rogerson-Revell (2011), considerable confusion is demonstrated by speakers in several individual segments: [e] and [æ] can be a problem, with [e] frequently used for both. Since [‫ ]ݞ‬does not exist in Spanish, it is taken over by other vowels, typically [‫ܤ‬:] or the diphthong [‫ۑ‬u] (must sounds like mast or most). Both [aܼ] and [eܼ] diphthongs occur in Spanish, yet there may be spelling interference, so that main [meܼn] may be pronounced [maܼn]3. As regards the consonants, the bilabial fricative [ȕ], as in Cuba, the voiced velar fricative [Ȗ], as in Diego, and the palatal [Ȝ], as in calle, are a Spanish idiosyncrasy and are absent from languages such as Polish or English. As for the other consonants, very close equivalents can be found in English and they are therefore articulated with no serious difficulty. A feature that Spanish shares with Polish, but which creates problems in English is the devoicing of word-final consonants in indigenous words. Spanish actually allows a total lenition of a stop consonant in this position 3 Students uniformly underline (in personal interviews) that they tend to pronounce everything they see and normally the way they see it written. Hence the erroneous forms with the past tense –ed ending, which is normally pronounced everywhere as a syllabic variant, so that a word like laughed has two syllables [‘l‫ܤ‬:fed].

12

Chapter One

and bans word-final plosives in clusters. As for the voiceless plosive series [p, t, k], there are conflicting messages in the relevant sources: Haáadkiewicz-Grzelak (2006, 24) explicitly stresses that “the [...] crucial difference is the lack of aspiration of Spanish voiceless stops” (this is corroborated in Mott 2005), whereas Rogerson-Revell (2011) speaks of “little aspiration for /p t k/” without specifying what she means by that. An additional difficulty is connected with the regular lenition – which in this case means spirantization – of intervocalic voiced plosives into corresponding fricatives.4 Mott (2005) warns further that voiced fricatives such as [v] or [z] may only appear as positional variants of voiced stops (e.g., between vowels or voiced segments), and the voiceless [s], [f] and [ș] are remnants of a once very rich set. This constitutes immediate problem areas, given the role the voiced alveolar fricative [z] plays in the inflectional system of English. The lacking [‫ ]ݕ‬spirant and its voiced counterpart [‫ ]ݤ‬do not occur even as positional allophones. The (inter)dental [ș] has a reasonably wide distribution. Its voiced partner, however, appears only as a result of voice assimilation whenever it is followed by a voiced consonant (Haáadkiewicz-Grzelak 2006, 28), it sometimes may sound like a [d] for a beginner Spanish learner of English. As for affricates, European Spanish has only [‫]ݹ‬, “with obvious consequences for the learners” (Coe 2001, 93). All this results in confusion and mispronunciations such as Sue/chew for shoe and pleasure realized as pletcher, plesser or plesher. One more segmental problem is connected with the confusion between [b] and [v]: in Spanish there is a sound which is a sort of combination of the two sounds, it is realised as a [b] in the initial position (berry for very) but as a continuant intervocalically. Mott (2005, 260) notices that the initial clusters of two consonants of Spanish are similar to those in English, although they are generally less frequent, except for the fact that Spanish speakers do not have an [s]+ consonant combination. Here, a vowel is inserted, a prothetic [e], applying to loanwords like estructura, estadistica (Haáadkiewicz-Grzelak 2006, 41). The same strategy of introducing a vowel in this position is regularly transferred into English: learners will tend to insert a vowel before the “s”: escream, a good eschool (Rogerson-Revell 2011). In terms of phonological processes, voicing seems to play a significant part in Spanish, it is, however, always the regressive type of assimilation. Progressive voicing is completely absent in Spanish. This situation causes 4 The precise situation and contexts are outlined comprehensively in HaáadkiewiczGrzelak (2006, 25-26).

Extending Foreign Presence in English Phonetics Classes

13

problems in its own right: voicing segments in word medial clusters, also in compound structures, and an inability to progressively voice some of the inflectional endings of English (-es and –ed). Similarly, final consonants in English are sometimes difficult for foreigners. For instance, Spaniards have difficulty with the final [d] in words like card, heard or field. The problem lies not in a gap in the consonant inventory, as the phoneme /d/ exists in Spanish, but in the fact that Spanish has few final consonants and [d] is not one of them. The example of field illustrates one more difficulty, this one arising due to the non-occurrence of groups of consonants word-finally in Spanish, such sequential problems are usually resolved via elision (Mott 2005, 148).

4.3. Turkish Turkish uses fundamentally the same alphabet as English (since 1928), yet its orthographic system, which to a large extent employs one-to-one lettersound correspondence, can cause interference with the English pronunciation (Rogerson-Revell 2011). It has eight vowels and no diphthongs. Vowels in Turkish tend to be shorter than in English and in some contexts vowels are elided (city may turn into stee). Since no quantitative contrast between vowels exists in this system, learners exhibit distinct difficulty distinguishing between tense vs. lax vowels. For Turkish speakers [æ] may be confused with [‫ ]ݞ‬or [‫ܤ‬:] (Rogerson-Revell 2011); additionally it may be substituted with [e], resulting in forms like set for sat (Thomson 2001, 215). It is a sound that “plagues Turkish-speaking learners” (Thomson 2001, 215). [‫ ]ܥ‬and [‫ܧ‬:], as in not and nought are also normally confused, both may be replaced with a vowel closer to [‫ܤ‬:] (Rogerson-Revell 2011) or diphthongized to [o‫]ݜ‬ (Thomson 2001, 215). Confusion is also observed between [u]/[‫ ]ݜ‬and [eܼ] may be confused with [e] (mate/met) (Rogerson-Revell 2011). Additionally, the final element of the English diphthongs closing in [ܼ], namely [aܼ, eܼ, ‫]ܼܧ‬, may become devoiced and resemble more the German palatal spirant [ç] (Thomson 2001, 215). The English schwa [‫ ]ۑ‬finds its nearest equivalent in the tenser and higher [Õ] of Turkish. However, under the influence of orthography, speakers often produce the full vowel quality for the unstressed one, as in [ܼnkonvܼnient] or [eddi‫ݕ‬onal]. Turkish has twenty consonants and shares many of them with English. Among those which are not shared are [ș] and [ð], which do not occur in Turkish and are typically substituted with an over-aspirated [t] and [d] (Thomson 2001, 216). Strangely, [v] and [w] may be confused in perception and production, especially in the neighbourhood of back

14

Chapter One

vowels. The difficulty is compounded by the fact that the Turkish alphabet does not have the letter “w”, thus loanwords containing this sound are spelled with a “v”, contributing to the confusion. /ƾ/ is found only before [k] or [g], therefore these stops are never elided after the velar nasal in English. Turkish speakers tend to devoice [b d g ‫ ]ݶ‬at the end of words and syllables, they sound like their voiceless counterparts. They will similarly not make the vowel lengthening necessary before final voiced sounds (Rogerson-Revell 2011). Turkish shares many final consonant clusters with English. However, if the final consonant is a voiced plosive it may be problematic (bulb > bulp). Initial consonant clusters are not allowed in Turkish, while clusters of more than three consonants in any position are unusual. The most problematic initial clusters are those beginning with [s], where vowel insertion is a common strategy (step > istep or sitep) (Rogerson-Revell 2011). In Turkish, loanwords from English opt for the initial easing vowel (istasyon “station”), while when speaking English the tendency is to insert the vowel after the first consonant: siprink for spring or filute for flute (Thomson 2011, 216). As with English, word stress is mobile in Turkish, so word stress should not be too problematic except where Turkish patterns are different to English ones. For example, stress is typically on the final syllable in many Turkish words, including nouns, pronouns, verbs, conjunctions and adverbs. In comparison, English nouns, main verbs and adjectives are front-stressed (Rogerson-Revell 2011). There is no schwa so the vowel reduction which is necessary to produce an English rhythm is not present and the rhythmic pattern is generally much more even because of the tendency to pronounce each syllable clearly. This is also transferred onto English. To sum up, what transpires from the above outline is that certain problems are more likely to recur than others in pronunciation training. Moreover, one can extrapolate certain areas of difficulty that are common to speakers from different native language backgrounds. The table below illustrates the commonalities and differences between Polish, Spanish and Turkish speakers of English.

Extending Foreign Presence in English Phonetics Classes

15

Aspiration of initial plosives [p t k]

[ș] [ð]

[v] [w]

[s] [z]

[‫]ݤ[ ]ݕ‬

[‫]ݶ[ ]ݹ‬

[r]

Consonant clusters s(C)C

Tense/lax vowels contrast

[‫]ۑ‬

Word stress

Rhythm & intonation

Polish

¥

¥

¥

x

¥

¥

¥

¥

x

¥

¥

¥

¥

Spanish

¥

x

x

¥

¥

¥

¥

¥

¥

¥

¥

x

¥

Turkish

¥

¥

¥

¥

¥

x

x

x

¥

¥

¥

x

¥

Language

Final voiced/voiceless stops distinguished

Table 1-1. Overview of potential pronunciation problems by language group

A cursory look through the data in the table shows that, interestingly, native speakers of these very different languages share many pitfalls when trying to master the way English ought to be spoken. It should therefore be relatively easy to combine them into one instructional group. It turns out, however, not to be that simple.

5. How to reconcile the same with the different? Parrot-fashion repetition of foreign-sounding new sounds for familiar letters, rolling them around your mouth, exercising sounds that you did not know were possible, incredulous of the idea that they could deliver meaning is what characterizes a student’s experience of many pronunciation training classes. The idea advocated in this chapter is that significantly better results can be achieved when the students develop an awareness of their own problem areas as well as some of their underlying reasons. Escudero (2005, 180) proposes that the role of the input is more important than an age-related lack of cognitive plasticity. This is because rich L2 input can overrule the reduced level of plasticity in adult learners. One prerequisite to a successful L2 pronunciation class is to provide sufficiently rich input, both in terms of the model on which L2 sound perception can be systematically based and the practice activities accompanying it. Escudero (2005, 180-81) underlines as well that throughout their learning, L2 students will have different perception

16

Chapter One

grammars for their two languages. In the case of Spanish learners of English, where [i] and [ܼ] contrast for example, the new length distinction will not be used when perceiving L1 vowels because this distinction was created in the L2 perception system to optimally cope with the L2 production environment. Consequently, drawing on the rigorous phonetic and phonological description of L1 and the target L2 provided leads to predicting the initial state for L2 acquisition, i.e., the perceptual system that learners will initially use in their L2 (Escudero 2005, 308-309). In linguistically non-homogenous groups, mutual observation and interaction with each other allow students to develop a sort of metaawareness not only of their own deficiencies in the pronunciation of English but also of the problems and difficulties experienced by their foreign classmates. In that respect all of them become more phoneticallyconscious and able to recognize, but also frequently reproduce, foreign sounding articulations. They probably regard it as a sort of fitness training for their speech organs. That in turn helps them to self-monitor their own production. That is a worthy aim in itself. Apart from worthy aims, there are also the course requirements, which specify, again rigorously, the objectives for the students to attain. It is also stated that any student participating in the course is to achieve the specific goals written into its program, naturally, to a different degree. The issue here is how to convince all of them that these goals are attainable and how to guide them in order to assist them in their learning. It comes as a revelation to the instructor as well that some of the aspects of English speech so far have been taken for granted and not paid particular attention to. Additionally, whenever it is deemed helpful, we draw on what the students are familiar with in their native language, which, until comparatively recently, meant Polish only. To offer but one example, the notoriously problematic English (E) [ܼ] vowel5 is regularly introduced via contrasting it with the Polish (P) [ܺ]: the students’ attention is drawn to pairs of words like P. syn – E. sin; P. dym – E. dim; P. tym – E. Tim; P. typ – E. tip etc. At first, it is often difficult to hear sounds that are not identical to those of our first language. Until one is capable of perceiving the difference properly, any attempt at reproducing them will be futile, that is why the first priority is to train students to perceive such differences. Thus, they are made sensitive to the 5

A vowel of practically identical quality is found in Lithuanian. Students of English there do not have to spend weeks trying to get it right, they also grab the quantity/quality distinction between [ܼ] and [i:] of English straightaway. As such, they constitute useful peer-models for their Polish classmates to follow.

Extending Foreign Presence in English Phonetics Classes

17

perceptible difference in the quality of the two vocalic units, and then they are asked to say the Polish words with an English vowel. To make it attractive, the words are expanded into little phrases which are also to be articulated to sound “English”: to nie mój syn, siwy dym, tym samym, ciemny typ.6 When it comes to disyllabic pairs like P. syty – E. city (jestem syty) or P. czyli – E. chilly (czyli co) the so-called “happy” [i] is modelled on a line from a pre-WW2-song performed by Hanka Ordonówna: “[...] i tylko w oczy stĊsknione popatrz” or “MiáoĞü Ci wszystko wybaczy.” This sort of practice is obviously only directed at Polish students. With the increasingly regular presence of non-Polish students in such classes, certain goals have to be re-defined and practices re-invented. It is not only that suddenly different types of error begin to crop up, but also that certain difficulties turn out not to be difficulties at all. Relevant cases in point are the notorious vowel insertion in front of s+C(C) sequences by Spanish and Turkish speakers (We are estudents from Espain and we want to estay in this class to estudy more espeech), compared with the ease Spanish speakers have with the English (voiceless) interdental [ș]. The prothetic [e] in clusters is frowned upon, not only by the instructor, but, by common consent, also by the classmates who act as peer-reviewers, pretending not to tolerate or understand those utterances (“a matress made of a straw (=estraw)? Just one?”). The [ș] competence of Spanish students is used in a similar fashion: they serve as models demonstrating its target, ideal quality using words from their own language, like cervesa [șerðeșa] “beer” or zapato [șapato] “shoe.” Such demonstrations turn out to be infinitely more attractive and memorable than even the most meticulous instruction offered by the tutor. In class pair-practice, foreign students serve as critical peer-reviewers of their partners’ mispronunciations. On other occasions they need to be shown that certain sounds are familiar though perhaps not as speech sounds. The students are then asked to imitate sounds characteristic of some animals or natural phenomena. Among these, the sound of the wind blowing through leaves in the autumn is smuggled: a prolonged [‫ݕݕݕݕ‬...] whistle. Following this gymnastics of the speech apparatus the students practice words with the [‫ ]ݕ‬segment in the initial position, which is easier for them to grasp and reproduce there than finally or even medially: shoe, shine, show, shoulder to be followed by the [‫ݕ‬r-] initial words like shrew, shrink, shrine. Once the voiceless spirant is assimilated, the time comes to acquire the voiced [‫]ݤ‬. 6

Some of the ideas listed here were originally developed by Piotr Okas, a longtime pronunciation trainer at the English Department of the Pedagogical University of Cracow, to whom I am deeply indebted for sharing them with me.

18

Chapter One

Collaboration between classmates is always a desirable feature of any class and it is manifested when, first, Polish students notice not only the [e] (or [i])-insertion in front of clusters or other cluster simplifications, but also when it comes to first learning their partners’ names: it takes time to learn how to pronounce Spanish Begonia, Ruben or Virginia. Once Polish speakers become aware of the lenitions regularly produced by Spaniards (and transferred onto English), they begin to exercise peer-monitoring and correction, stigmatizing errors such as to go or a game produced with the fricative in place of a stop [t‫ ]ݜۑܵ ۑ‬and [‫ܵ ۑ‬eܼm]. The instruction often heard on the part of a Polish or a Turkish student towards a Spanish partner is that they need to properly close or open their mouth/lips when saying words with the easily confused [b] and [v], frequently realised as [ȕ], such as very (berry?), robin, habit or volleyball. An activity which works well in that context, both to make Spanish students aware of what they are doing wrong and to expand the general phonetic proficiency of their Polish and Turkish classmates is connected with transcribing certain Spanish words with an English accent (and [p, t, k] aspiration marked) but without the spirantizations so characteristic of Spanish (Mott 2005, 264): Madrid, Argentina, bolero, rodeo, seniorita, patio, universidad etc. A reverse activity, that is transcribing English words with the characteristic Spanish features proves equally successful: leader, stress, Barbara Streisand, football etc. It thus becomes immediately clear that this sort of class activity can only be obtained under a more cognitive approach, in which learners are taught to be aware of their speech and to self-monitor. Such an approach is successful, because it appeals to the intellectual and cognitive component of learning (Eckman and Iverson 1997, 202). Brian Mott (2005, 246) states an obvious problem in the following fashion: When a person speaks a foreign language, they tend to transfer their own sound system and produce sounds which they have been accustomed to making since they first learned to speak.

Put differently, our pronunciation of a foreign language is inhibited by our native articulatory habits. And yet, not all pronunciation problems are of the same degree of importance. Some sounds in a system may be relatively infrequent in occurrence and in fact be substitutable by others. Such sounds (like [‫ ]ݤ‬in English) are believed to have a low functional load. Consequently, less attention needs to be paid to them than to other areas of difficulty in the English phonological system (Mott 2005, 248).

Extending Foreign Presence in English Phonetics Classes

19

One problem area, common to all the groups of students discussed here, is the problem with word final obstruent devoicing. In class it appears relatively early in the course, when the inflectional “-es” ending is introduced in its three variants and three functions. The difficulty is compounded by the non-occurrence of the [z] segment in Spanish, a feature that every student from Spain immediately proclaims, alongside the non-existence of the voiced [ð], [‫ ]ݤ‬and the problematic [v]. This, however, is only partially true. This impediment to learning is only potential, since these sounds appear as positional variants of [s], [ș] or [d] after vowels before a voiced segment or through voice assimilation, e.g., in words like isla “island”, desdo “from”, Immaculada, juzgar “to judge”, Israel. This readiness for voice assimilation can be immediately used, first to establish [z] as a regular feature of their pronunciation and then to eradicate errors in word-final devoicing. It is true, on the other hand, that both Turkish and Spanish speakers have to consciously learn the rules for progressive voice assimilation obtaining with the “-es” and later the “-ed” endings. It is insufficient to simply explain which variant appears where, it must be complemented by practice in choral, pair and individual repetitions of not so much individual words, but rather word complexes uttered with no pause: another problem common to students from Poland or Turkey, who are very happy to articulate each word separately. Therefore, they drill chunks of words, such as: stars are shining, lies a lot, sees me, loves me, legs break, things go wrong, seems right, misses me, guns and roses, breathes deeply, watches whatever and similar ones. Thus, when these are produced as combined groups, the relevant segment is no longer final and can be pronounced as voiced by students with different NL backgrounds. A related phenomenon, namely vowel clipping in syllables closed by voiceless segments as opposed to full length vowels in those ending in a voiced one, can be helpful here. Initially, all students, when practising pairs like eaten – Eden, Miss – Ms, hit – hid, feet – feed, concentrate more on marking the difference in the relative length of the vowel as a concomitant of the presence or absence of voicing in the following consonant. Since they do it in short dialogue exchanges (Student A: Did you say bit? – Student B: No, I didn’t say bit. I said bid.), their pair-partners can monitor their accuracy. Only after some time do the requirements change from just clearly underlining the length to additionally maintaining the voicing where required. Finally, since back-chained drilling is a regular feature of a practical phonetics class, some of these drills can incorporate problem segments in an almost tongue-twisting manner. In addition to practising the rhythm of English with its reduced syllables, they tackle the difficulties some

20

Chapter One

students may have, while for others they will be plain fun. Spanish and Turkish students especially will benefit from combinations such as William will want to win the war like Wellington or Vivian ventured to revenge Veronica in Venice and Vienna. For Polish students a challenge was presented in the form of nonsensical sentences like Sosny rosną nisko a sasanki wysoko or Zuzanna zaraziáa zebrĊ w zoo, where their job was to produce [ș] where they saw [s] and [ð] instead of [z]. In addition to the drill sentences proposed by the instructor, students themselves thrive on inventing and producing their own sentences, the crazier the better and the more effective. Examples of such student creations include the following: Heather’s thoroughly-clothed body was thrown up thirty-three times to celebrate her birthday healthily (Spanish students for Poles and Turks); Such occasional delusions shouldn’t overshadow our finishing time for leisure and pleasure (Polish speakers for their Spanish classmates).

6. Conclusions The examples presented above hopefully demonstrate that it is possible, at least to some extent, to build the presence of non-Polish students into the classes of a course in practical phonetics, and also to build from it and to use this presence to the general advantage of all the students. Trying to successfully guide the students in their practical efforts, we need to resort to measures that can do justice to their needs and necessities facing them. In order to be able to prevent and also find remedies for pronunciation errors it makes sense for the instructor to know their causes, and to be able to hear and diagnose them appropriately when they crop up. The instructions and exercises aimed at eradicating errors must be maximally simple and comprehensible, it appears pointless to offer elaborate descriptions of the relevant tongue positions and vocal folds states for students at any level. An important cause of pronunciation errors appears to be native language phonetic interference. Apparently, it is not the absence of equivalent structures that is most difficult to repair, but the stable and durable near correspondences between the given elements which promotes interference. As Miatluk, Szymaniuk and Turáaj (2008, 22) claim: it is difficult enough to assimilate foreign sounds that are alien to the native sound system, but it is more difficult still to assimilate foreign sounds which have some common features with the native sounds. And the more such common features there are, the more difficult it is to obtain efficiency in pronunciation.

Extending Foreign Presence in English Phonetics Classes

21

In the light of the visible resurgence over the last few years in the attention that has been devoted to the teaching of pronunciation in second language instruction, it is perhaps recommendable to investigate innovative and hopefully more effective intervention strategies. It is the belief of the present author that one such approach is to use the presence of foreign, i.e. non-Polish students in the class to the maximum, both in terms of introducing guidance and practice in areas previously not included and which is specific to groups of particular foreign learners and also of utilizing the foreign learners’ competence with some elements of pronunciation to the general benefit and increased competence of Polish students. The fundamental aim then is an overall improvement in the learners’ pronunciation of English with an accompanying awareness of all matters related to sound perception and production.

References Baran-àucarz, Maágorzata. 2006. “Prosto w Oczy – Fonetyka jako „Michaáek” na Studiach Filologicznych?” In Dydaktyka Fonetyki JĊzyka Obcego w Polsce, edited by Wáodzimierz Sobkowiak, and Ewa Waniek-Klimczak, 7-17. Konin: PaĔstwowa WyĪsza Szkoáa Zawodowa w Koninie. Buczek-Zawiáa, Anita. 2011. “Phonetics – Is It an Unwanted Study?” In Wspóáczesne Paradygmaty w Literaturze, JĊzykoznawstwie, Translatoryce, Pedagogice i Kulturoznawstwie w KontekĞcie Interdyscyplinarnym, edited by Andrzej KryĔski, Stanisáaw àupiĔski, and Maria Urbaniec, 325-34. CzĊstochowa: Wydawnictwo Educator. Buczek-Zawiáa, Anita, and Piotr Okas. 2008. “Second Language Interference: Does it Feed or Does It Bleed?” In New Trends in English Teacher Education, edited by Ignacio Ramos Gay, A. Jesus Moya Guijarro, and Jose Ingacio Albentosa Hernandez, 135-46. Cuenca: Ediciones de la Universidad de Castilla-La Mancha. Coe, Norman. 2001. “Speakers of Spanish and Catalan.” In Learner English. A Teacher’s Guide to Interference and Other Problems, edited by Michael Swan, and Bernard Smith, 90-112. Cambridge: Cambridge University Press. Eckman, Fred R., and Gregory K. Iverson. 1997. “Structure Preservation in Interlanguage Phonology.” In Focus on Phonological Acquisition, edited by S.J. Hannahs, and Martha Young-Scholten, 183-205. Amsterdam: John Benjamins.

22

Chapter One

Escudero, Paola. 2005. Linguistic Perception and Second Language Acquisition. Explaining the Attainment of Optimal Phonological Categorization. Utrecht: Landelijke Onderzoekschool Taalwetenschap. Haáadkiewicz-Grzelak, Maágorzata. 2006. Phonotactic Preferences of English and Spanish. Warszawa: Sygnatura. Agencja Wydawnicza. Jenkins, Jennifer. 2000. The Phonology of English as an International Language. Oxford: Oxford University Press. Miatluk, Anna., Dorota Szymaniuk, and Olga Turáaj. 2008. Contrastive Phonetics of English and Polish. Phonetic Interference. Biaáystok: Trans Humana. Mott, Brian. 2005. English Phonetics and Phonology for Spanish Speakers. Barcelona: Publicacions i Edicions Universitat de Barcelona. Nowikow, Wiaczesáaw. 2012. Fonetyka HiszpaĔska. Warszawa: Wydawnictwo Naukowe PWN. Papierz, Maryla. 2009. “Gramatyka Kontrastywna a Nauczanie JĊzyków Obcych.” In Nauczanie JĊzyków Pokrewnych, edited by Jacek Baluch, and Maryla Papierz, 47-56. Kraków: Wydawnictwo „scriptum.” Rogerson-Revell, Pamela. 2011. English Phonology and Pronunciation Teaching. London: Continuum. Sobkowiak, Wáodzimierz. 2004. English Phonetics for Poles. PoznaĔ: Wydawnictwo PoznaĔskie. Swan, Michael, and Bernard Smith. 2001. Introduction to Learner English. A Teacher’s Guide to Interference and Other Problems, by Michael Swan, and Bernard Smith, ix-xi. Cambridge: Cambridge University Press. ĝpiewak, Grzegorz, and Justyna GoáĊbiowska. 2001. “Polish Speakers” In Learner English. A Teacher’s Guide to Interference and Other Problems, edited by Michael Swan, and Bernard Smith, 162-78. Cambridge: Cambridge University Press. Thomson, Ian. 2001. “Turkish Speakers” In Learner English. A Teacher’s Guide to Interference and Other Problems, edited by Michael Swan, and Bernard Smith, 214-26. Cambridge: Cambridge University Press.

CHAPTER TWO DIFFERENCES AND SIMILARITIES IN THE PERCEPTION OF GERMAN VOWELS: THE CASE OF TURKISH AND POLISH LEARNERS OF GERMAN AS A FOREIGN LANGUAGE KATHARINA NIMZ

1. Introduction Learning a foreign language is not an easy task. Not only do we have to learn new words, new grammar, and differences in its pragmatic use, we also have to learn how to pronounce the new language, i.e. we have to master its phonetic and phonological units. The difficulty in producing foreign sounds is assumed to be related to how well we perceive them (Flege 1995), which is why much research in the field of second language (L2) phonetics and phonology is concerned with L2 sound perception (Munro and Bohn 2007). While English is the most studied L2 in the field (Hayes-Harb 2012), a few recent studies have looked at the perception of German vowels by L2 learners as well (Altmann, Berger and Braun 2012; Darcy and Krüger 2012). The present study contributes to this research by investigating the perception of (manipulated) German vowels both by Turkish and Polish German-as-a-Foreign-Language (GFL) learners.1

1

I would like to thank Ghada Khattab, Adamantios Gafos and the Erasmus Mundus PhD Programme International Doctorate for Experimental Approaches to Language and Brain (IDEALAB) for their support and insightful feedback. I further thank the schools, teachers and participants in Poland, Turkey and Germany, who made this research possible.

24

Chapter Two

2. Brief background: L2 phonetics and phonology The field of experimental L2 phonetics and phonology is a relatively young field of research, with dedicated anthologies being published in the 90s and later (Strange 1995; James and Leather 1997; Bohn and Munro 2007; Edwards and Zampini 2008). However, even before this recent interest in L2 speech, there were hypotheses put forward more than half a century ago which are still relevant in the field today. As early as 1939, Trubetzkoy had described our native (L1) sound system as a “phonological sieve” in that it filters out those properties of the L2 speech signal that are not relevant to our L1 system, consequently leading to mistakes in the perception and the production of foreign sounds. Related to this idea is a prominent hypothesis by Lado (1957). His contrastive analysis hypothesis states that difficulties in learning a language can be predicted on the basis of a systematic comparison between the (sound) systems of the learner’s native and second language. He assumes that sounds (and other linguistic units) which are similar to the learner's native language will be easy to learn, while those elements which are different will be more difficult. It has been noted though that, firstly, the definition of “similar” is not as straightforward as it might seem, and that, secondly, not all L2 speech difficulties may be explained through structural comparisons between the L1 and L2 sound systems alone, but are tied to various other variables such as age, L1 and L2 input, foreign language instruction, and others as well (Bohn 1995; Munro and Bohn 2007). In his famous Speech Learning Model (SLM), Flege (1995) accounts for findings which have shown that not all foreign sounds are necessarily difficult for L2 learners to acquire. An important concept in this context is that it is not the different sounds (as postulated by Lado 1957) which are difficult for L2 learners, but the similar ones. Flege assumes that the mechanisms used in L1 phonetic category formation remain intact over the learner’s life span and can be applied to L2 learning, that is, it is possible that learners can build new L2 phonetic categories and, with that, perceive and produce L2 segments like a native speaker. This new category formation becomes more likely the younger learners are and the more perceptually different the closest L1 and L2 sounds are. If sounds are too similar, new category formation is blocked, as so called “equivalence classification” will take place, that is, new sounds are falsely processed as L1 sounds. Even though the concept of similarity is a difficult one – i.e.

Differences and Similarities in the Perception of German Vowels

25

what are reliable and objective measures for classifying two sounds as similar or different?2 – a brief contrastive analysis of the sound systems concerned is a necessary step in describing the issue under scrutiny even if no precise predictions may be formed in terms of the SLM.

3. The German, Polish and Turkish vowel systems Even though the vowel charts produced for the handbook and journal of the IPA are articulatory approximations and are therefore not suitable for precise conclusions about how phonetically similar or different the vowels of the three languages of interest really are, they do allow for a brief contrastive analysis on a broader phonological level. The vowels which were investigated in the perception experiment to be described below were the German point vowel pairs /i:/-/Õ/, /a/-/a:/ and /u:/-/‫ݜ‬/ (circled). (Henceforth, for the sake of convenience, these pairs may be referred to as “i-pair”, “a-pair” and “u-pair”).

Fig. 2-1. German (left) (Kohler 1999, 86), Polish (middle) (Jassem 2003, 105) and Turkish (right) (Zimmer and Orgun 1999, 154) vowel systems as depicted in the Handbook and Journal of the IPA. Vowels of interest to the study have been circled.

3.1. Vowel quality As can be deduced from the left-hand chart, the German low vowels /a/ and /a:/ are assumed to be identical in vowel quality, as they are depicted with the same vowel symbol at the same position in the vowel space. Long tense /i:/ and long tense /u:/ however, differ from lax /Õ/ and /‫ݜ‬/ in that their tongue position is more peripheral, that is, /i:/ is articulated with a 2

Bohn (2002) proposes direct assessment of similarity in perceptual studies, perceptual confusion studies and studies of interlingual identification, combined with goodness-of-fit ratings.

26

Chapter Two

tongue position that is much higher and more forward than for /Õ/, while /u:/ is articulated further back and higher than /‫ݜ‬/. Due to the fact that German contrasts short lax and long tense vowels such as, for example, /‫ܧ‬/ and /o:/ in Pollen [p‫ܧ‬lԥn] “pollen” versus Polen [po:lԥn] “Poland”, German has approximately twice as many vowels as Polish and Turkish. Where German has at least two vowels in the high front vowel space (i-pair), the high back vowel space (u-pair) and the low central vowel space (a-pair), Polish and Turkish each have only one point vowel, which in the IPA charts are depicted with the same tense vowel symbols /i/, /u/ and /a/. Still, from the charts one cannot draw the conclusion that the phonetic quality of the Polish and Turkish vowels is therefore the same or that the quality of these vowels is the same as the German tense vowels. As far as Turkish is concerned, comparable acoustic data (first and second formant values – F1 and F2) of German and Turkish vowels exist which suggest that the vowel quality of Turkish /i/ is almost identical to that of German /i:/, in contrast to German /Õ/, which seems relatively distant from the Turkish point vowel counterpart. Turkish /u/, however, seems to be equally similar to German /u:/ as it is to German /‫ݜ‬/ (as it is in between the two German vowel qualities) and Turkish /a/ is equally far away from German /a/ and /a:/ (Nimz forth.). Hentschel (1986) compares acoustic data for Polish and German, where similar conclusions can be drawn, except that Polish /u/ seems to be almost identical to German /u:/, while it is relatively distant from German /‫ݜ‬/ in the F1-F2 acoustic space, which is an important difference compared to the Turkish-German acoustic comparison.

3.2. Vowel quantity German lax vowels are generally short, while German tense vowels are long. In the German IPA chart above, vowel length is not explicitly marked by means of the diacritic [:]; however, it is the case that, in stressed syllables, German tense vowels are on average about twice as long as German lax vowels (Antoniadis and Strube 1984). Early experiments on German vowel perception (Weiss 1974; Sendlmeier 1981) have come to the conclusion that – for German native speakers – vowel length plays a dominant role in distinguishing the low central vowel pair, while vowel quality is more important for the mid to high vowel pairs. Still, in most phonological descriptions, German is described as a language with contrastive vowel length (Meinhold and Stock 1982; Ternes 2012).

Differences and Similarities in the Perception of German Vowels

27

Turkish, on the other hand, is generally described as a language for which vowel length is not contrastive (Kabak 2004). However, there are two occasions in which long vowels may occur in the language, which is why one could assume that Turkish has “phonetic” vowel length (in contrast to “phonological” vowel length): Firstly, Turkish borrows a considerable amount of words from Arabic, a language, which has phonological vowel length contrast. In these borrowed words, long vowels are preserved in Turkish pronunciation (e.g., saat [sa:t], engl. “hour”). Secondly, the Turkish voiced velar fricative /ܵ/ (in orthography represented as , “yumu‫܈‬ak g”) is not pronounced in Turkish, resulting in a compensatory lengthening of the preceding vowel (for example in çi÷ [t‫ݕ‬i:] (“landslide “), or tu÷ [tu:] (“flag“), see Kirchner 1999). Polish has neither long vowels on a phonetic nor on a phonological level (Tworek 2012), which is why many GFL researchers predict that Polish speakers will have problems acquiring German vowel length (Morciniec 1990; Slembek 1995 or Müller 2005).

4. Background of study In the following, two sets of data will be presented of which one is concerned with data from Turkish GFL learners (Nimz 2011; forth.) and the other from Polish GFL learners. In an earlier study, Nimz (2011) conducted the following perception experiment with Turkish learners of GFL. The research question was whether quantity and/or quality distinctions in the German a-, i- and upairs are problematic for Turkish learners acquiring GFL. The method of investigation was a simple discrimination task with manipulated nonsense words (such as [bu:p] versus [b‫ݜ‬p]), which either differed in length (short versus long), quality (tense versus lax), or both (unmanipulated vowels, a condition therefore called proto). The manipulation was done in PRAAT (Boersma and Weenink 2008); short vowels were lengthened to the length of the average long vowel and long vowels were shortened to the length of the average short vowel, and then matched in pairs according to the experimental condition. The Turkish learners and an age-matched German native speaker control group listened to the respective word pairs and had to judge whether the two words that they heard were the same or different.3 The accuracy data (see Fig. 2-2 below) show that Turkish GFL learners were not significantly worse than German native speakers in 3

For details of stimuli manipulation and experimental design, see Nimz (forth.).

28

Chapter Two

distinguishing vowels based on their length, while quality differences (in the u-pair) were difficult for the learners to distinguish. There were two possible explanations for the results in this first study: a) Turkish learners can hear quantity differences as well as German native speakers because duration cues are always easy to access whether or not learners have experience with them in their native language (which would support the desensitization hypothesis by Bohn 1995)4, or b) Turkish learners can hear the quantity differences well because they have experience with it on a phonetic level in their native language (which would support the feature hypothesis by McAllister, Flege and Piske 2002).5 The same experiment was also conducted with a comparable group of Polish GFL learners at a Polish high school in Lublin in order to investigate which of the above mentioned explanations may better explain the Turkish results. Like the learners from Nimz (2011), the Polish learners were around the age of 18 years old and about to take the Deutsches Sprachdiplom II (German Language Certificate of the Education Ministers Conference), which tests language competence on a medium-advanced level. All of the students (both Turkish learners from the earlier study and the new Polish group) had received extensive GFL lessons for at least three years. 21 Polish GFL learners were tested in the follow-up study, of which three were excluded from the statistical analysis

4

In Bohn’s study German, Spanish and Mandarin native speakers had to identify English vowels on a bet to bat (and beat to bit) continuum as bet or bat (beat to bit), which were manipulated in duration and spectral features in equal steps. Interestingly, Spanish and Mandarin participants – just like the German ones – relied much more on durational differences when identifying the stimuli than native English speakers, which led Bohn to the conclusion that durational differences are always easy to access whether or not learners have experience with them in their native language (desensitization hypothesis). 5 McAllister, Flege and Piske (2002) investigated the perception and production of Swedish quantity distinctions by Estonian, English and Spanish learners of L2 Swedish by means of an identification task with real Swedish words and nonwords, which were created by replacing long segments by short ones and vice versa. They found that Estonian speakers were better at identifying the test words correctly than the English and Spanish speakers (and the English were better than the Spanish speakers), which led the researchers to formulate the feature hypothesis, which states that L2 features (such as duration) not used to signal phonological contrast in the L1 (for example in Spanish and partly in English) will be difficult to perceive for L2 learners.

Differences and Similarities in the Perception of German Vowels

29

because they had made too many errors in the filler condition (where both stimuli were exactly the same).

5. Results While in the earlier study with Turkish GFL learners and German native speakers significant differences in the perception of German vowels could only be found within the u-pair for the quality condition (in which vowels differed only in their vowel quality), the results for the Polish GFL learners were very different

*

* *

*

Fig. 2-2. Correct responses of German, Polish and Turkish speakers

Because the data were not normally distributed, non-parametric Wilcox tests were conducted to test for group differences. Since there were three groups, which were tested in three conditions for three different vowel pairs (3x3x3), the alpha level was adjusted according to the Bonferroni correction, hence Į was set at 0.002 (0.05/27). The tests revealed four significant group differences: In the length condition for the a-pair, Polish learners and German native speakers differed significantly (p