Multilingualism and Applied Comparative Linguistics (Volume II) [new ed.] 184718670X, 9781847186706

Citation preview

Multilingualism and Applied Comparative Linguistics (Volume II)

Multilingualism and Applied Comparative Linguistics (Volume II)

Edited by

Franks Boers, Jeroen Darquennes, Koen Kerremans and Rita Temmerman

Cambridge Scholars Publishing

Multilingualism and Applied Comparative Linguistics (Volume II), Edited by Franks Boers, Jeroen Darquennes, Koen Kerremans and Rita Temmerman This book first published 2008 by Cambridge Scholars Publishing 12 Back Chapman Street, Newcastle upon Tyne, NE6 2XX, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2008 by Franks Boers, Jeroen Darquennes, Koen Kerremans and Rita Temmerman 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): 1-84718-670-X, ISBN (13): 9781847186706

TABLE OF CONTENTS

LIST OF IMAGES .........................................................................................viii LIST OF TABLES........................................................................................... ix PREFACE...................................................................................................... xi ACKNOWLEDGEMENTS ............................................................................. xvii PART ONE: UNIVERSALITY AND CROSS-CULTURAL COMMUNICATION FIRE AND HOT LIQUIDS: ANGER IN HINDI AND ENGLISH VINEETA CHAND .................................................................................... 2 PLANNED LANGUAGES AND LANGUAGE PLANNING FEDERICO GOBBO ................................................................................. 24 PART TWO: BILINGUALISM, TRILINGUALISM AND LANGUAGE PROFICIENCY

A SOCIOCULTURAL PERSPECTIVE ON THE LANGUAGE PRACTICES OF TRILINGUAL FAMILIES IN ENGLAND AND GERMANY ANDREAS BRAUN ................................................................................. 44 NATIVE JUDGEMENTS OF L2 LEARNERS’ PRAGMATIC TRANSFER: THE CASE OF APOLOGIES

HORTÈNSIA CURELL I GOTOR AND MARIA SABATE I DALMAU ............ 64 DOES BILINGUALISM ACCELERATE LANGUAGE CHANGE? THE CASE OF SPANISH-CATALAN BILINGUALS AND COPULA CHOICE PEDRO GUIJARRO-FUENTES AND KIMBERLY GEESLIN ......................... 83

vi

Table of Contents

PART THREE: DISCOURSE ANALYSIS IN CROSS-CULTURAL COMMUNICATION RESEARCH POLITICAL JOURNALISM ACROSS THE LANGUAGE BORDER. COMMUNICATIVE BEHAVIOUR IN POLITICAL INTERVIEWS BY DUTCHAND FRENCH-SPEAKING JOURNALISTS WITH DUTCH- AND FRENCHSPEAKING POLITICIANS IN FEDERAL BELGIUM MARTINA TEMMERMAN AND DAVE SINARDET ................................... 110 THE AMBIGUOUS WE-FORM: INTERPRETATIVE CLUES IN PARALLEL TEXTS SYLVAIN DIELTJENS AND PRISCILLA HEYNDERICKX .......................... 139

PART FOUR: TERMINOLOGY AND SPECIALISED LANGUAGES IDENTIFYING CONCEPTUAL RELATIONS IN ENGLISH AND FRENCH SPECIALISED CORPORA

ELIZABETH MARSHMAN ..................................................................... 162 MULTILINGUAL TERMINOLOGICAL RESOURCES IN SEMANTIC WEB TECHNOLOGY KOEN KERREMANS ............................................................................. 186

THE PROCESS OF STABILIZATION OF TERMINOLOGICAL UNITS IN TEXTS CARME BACH MARTORELL ................................................................ 202 CONVERTING A MONOLINGUAL DATABASE INTO A MULTILINGUAL SPECIALIZED DICTIONARY

HEE SOOK BAE AND MARIE-CLAUDE L’HOMME................................ 225 PART FIVE: TRANSLATION STUDIES -ING CIRCUMSTANCE ADVERBIALS AND THEIR TRANSLATION INTO SPANISH: A CORPUS-BASED STUDY MARLÉN IZQUIERDO ........................................................................... 256 TEACHING ENGLISH-SPANISH CONTRASTIVE ANALYSIS THROUGH TRANSLATION ROSA RABADÁN AND BELÉN LABRADOR AND NOELIA RAMÓN ......... 278

Multilingualism and Applied Comparative Linguistics

vii

POLITENESS AND IMPOLITENESS IN THE SPANISH TRANSLATION OF THE CRUCIBLE: ASPECTS OF POWER AND GENDER Mª PILAR GONZÁLEZ VERA ................................................................ 302 A STUDY ON THE LEXICAL COHESION, MOOD AND MODALITY OF THREE ENGLISH TRANSLATIONS OF PERICLES’ FUNERAL ORATION EFTHYMIA KARAOUZA ....................................................................... 323 CONTRIBUTORS ........................................................................................ 345 SUBJECT INDEX ........................................................................................ 351

LIST OF IMAGES

1-1. Metaphor hierarchy: model 1 1-2. Metaphor hierarchy: model 2 1-3. Metaphor hierarchy: model 3 2-1. A taxonomy for planned languages 2-2. The two moments in the life of an interlanguage 5-1. Sample Questionnaire Item (translated to English) 6-1. Communities and Regions in federal Belgium 6-2. Question type continuum 7-1. Theoretically possible combinations 7-2. Table Dutch reference books 7-3. Parallel segmentation of the we-form in Dutch and French 8-1. POS classes of markers—Percentage of total in sample 8-2. POS of marker occurrences—Percentage of total in sample 9-1. A categorisation framework in FF POIROT 9-2. The Termontography workbench 11-1. Methodology for developing the Korean module 12-1. Distribution (%) of ‘subordinator + gerund-participle’ 12-2. Translational options per semantic function 14-1. Honorific axes (Brown and Levinson 1987, 181) 15-1. Halliday and Hasan’s classification of cohesive devices 15-2. Halliday’s (1994) categorisation of modal adjuncts

16 17 20 26 28 94 112 124 141 142 147 176 177 194 197 233 264 273 305 324 339

LIST OF TABLES

3-1. Typology of trilingual families in this survey 3-2. Parents’ language use with their children (by Type) 3-3. Type I parents’ commitment to pass on their cultural values 3-4. Type II parents’ commitment to pass on their cultural values 3-5. Type III parents’ commitment to pass on their cultural values 4-1. IFID realisations by native speakers and Catalan learners 4-2. Native-speaker ratings of native and non-native apologies 4-3. Mean native-speaker ratings of native and non-native apologies 5-1. Social variables under investigation 5-2. Summary of participant variables for speakers outside Catalonia 5-3. Summary of participant variables for speakers from Catalonia 5-4. Summary of the Linguistic Variables 5-5. Overall results 5-6. Estar allowed vs. estar not allowed 5-7. Results for regression analyses for both participant groups 6-1. Overview of the corpus 6-2. Average speaking time of interviewers and interviewees 6-3. Percentages of kinds of questions per interview type 6-4. Percentages of personal/non-personal questions 6-5. Percentages of shifts in footing by interviewer 7-1. Comparison Dutch and French we-forms 8-1. Relation occurrences and markers observed 8-2. Cognates observed in the results 8-3. Frequent English markers of the DISEASE–SUFFERER relation 8-4. Frequent French markers of the DISEASE–SUFFERER relation 8-5. Numbers of markers observed 8-6. Frequencies of observed markers 11-1. List of terms realized as actants 11-2. Contexts extracted from the corpus 11-3. Semantically related terms 11-4. Actantial structure and linguistic realizations of actants

49 51 52 54 57 70 72 73 91 92 93 95 96 97 97 114 115 124 128 132 146 166 167 171 171 174 175 228 229 230 231

11-5. Related terms for 낮1 ‘Web’ 11-6. Korean corpora 11-7. Examples of term candidates

232 234 236

x

List of Tables

11-8. Concordances for 놵녚뼍鲙 1 ‘save’ and 놵녚뼍鲙 2 ‘assign’ 11-9. Linguistic realizations of actants

239 240

11-10. Paradigmatically related terms of 늱ꭁ ‘address’

241

11-11. Syntagmatically related terms of 늱ꭁ ‘address’ 11-12. Korean entries and their equivalents

241 242

11-13. Verbalization with –뼍鲙 (Rom. -hada) 11-14. Intermediate table 12-1. Size of the P-ACTRES sample 12-2. Frequency rates of semantic categories 12-3. Spanish translational options for –ing circumstance adverbials 13-1. Composition of the part of CREA selected 13-2. Composition of the on-line part of The Bank of English 13-3. Contents of the English-Spanish Parallel Corpus P-ACTRES 13-4. Degree adverb fully and its translations into Spanish 13-5. Lexical item ‘kick’ and its translations into Spanish 14-1. Positive politeness vs. negative politeness 14-2. Culpeper’s positive and negative impoliteness output strategies 15-1. Superordinates and subordinates 15-2. The part and the whole 15-3. Mood and Modality 15-4. Pericles’ attitude towards the audience/degrees of obligation 15-5. Degree of commitment in Pericles’ statement 15-6. The ‘Scale of intensity’ (White 2006) 15-7. The three versions

245 248 261 262 267 282 283 285 286 289 307 310 326 329 332 333 337 340 342

PREFACE

In February 2006 the first international conference on Multilingualism and Applied Comparative Linguistics (MACL) was held in Brussels, Belgium. The present book is the second of two volumes containing a selection from the approximately 120 papers that were presented at that three-day event. The aim of the MACL conference was to bring together scholars from various branches of applied linguistics with a shared interest in crosslinguistic and cross-cultural communication. The conference thus fostered an exchange of knowledge and expertise among researchers from various disciplines, including educational linguistics, cultural linguistics, terminography, translation studies and studies of specialised languages. In this second volume we present the following issues: Part one. Part two. Part three. Part four. Part five.

Universality and cross-cultural communication Language practices, attitudes and proficiency Discourse analysis in cross-cultural communication research Terminology and specialised languages Translation studies

The first part of this volume covers two articles dealing with universality and cross-cultural communication. Vineeta Chand’s study is a cross-linguistic comparison of anger metaphors in Hindi and English. Although there is evidence for underlying conceptual similarities in the two languages, the study presents evidence against the notion that, on all levels, metaphoric paradigms are universal or cross-linguistically motivated by identical forces. Metaphor research thus clearly needs a method for discussing and quantifying variation at multiple levels. Introducing different types of variation and an explicit discussion of the metaphor hierarchy will permit a more fine-grained description of each language’s metaphors. This in turn will allow quantitative comparisons of multiple languages’ metaphors at various levels. The chapter by Federico Gobbo also deals with the notion of universality, but this time in the context of constructed international auxiliary languages or interlanguages such as Esperanto. Gobbo intends to

xii

Preface

show that although some issues in planning interlanguages are similar to those in natural language planning, some issues are very subject-specific. A comparison of interlanguages shows the important influence of the language creator’s mother tongue (i.e. the Bausani effect). This is particularly evident at the level of phonetics. The focus in Gobbo’s study is on the uses of the interlanguages (status planning) and on how users relate to them (acquisition planning). Specific attention is paid to Esperanto. The notion of language attitude, already appearing in Gobbo’s article, is further elaborated in the second part of this volume. Apart from language attitudes, the chapters that make up this second part also deal with language proficiency and practices. Andreas Braun aims to identify the sociocultural factors that influence parents’ language practices with their children in potentially trilingual families living in England and Germany, two predominantly monolingual European countries, which represent different geographical and sociolinguistic settings. His study, which is carried out on the basis of semi-structured interviews, endeavours to explore the cultural background of parents in trilingual families by focussing on the way in which they identify themselves with the cultural traditions and values that are attached to their native languages. This adds to the understanding of the reasons why some parents in trilingual families use all of their available languages with their children, while others do not. The chapter by Horténsia Curell I Gotor and Maria Sabaté I Dalmau primarily deals with the issue of language proficiency. In particular, Gotor and Dalmau examine (British) English native speakers’ assessment of Catalan learners’ apologies, elicited in various social situations (varying along the parameters of social distance and dominance). The study confirms their hypothesis that English native speakers generally rate learners’ apologies as less adequate than those produced by fellow natives. In order to help learners attain a better mastery of the target language, the authors conclude that it seems worth incorporating in second or foreign language teaching programmes a component that raises the learners’ awareness of cross-linguistic/cultural variation in speech act realisation. Pedro Guijarro-Fuentes and Kimberly Geeslin address the social and psycholinguistic factors that influence adult bilinguals’ use of Catalan and Spanish in Catalonia. Although these have been co-official languages since the introduction of democracy in 1978, there is a great deal of variation in individuals’ bilingualism. By studying the copula contrast of the Spanish in Catalonia, Guijarro-Fuentes and Geeslin’s study aims to explore whether speakers of Spanish in Catalonia select copulas differently from Spanish speakers in other regions of Spain. The study

Multilingualism and Applied Comparative Linguistics

xiii

shows that while the frequency of copula selection differs, the linguistic factors that predict that selection are largely the same. The study demonstrates that notions previously held about the acceleration of language change occuring in bilingual contexts can be expanded through additional research accounting for a wider range of variables that include, but are not limited to, the socio-political context, language attitudes and language identity. In the third part of this volume, language practices and language proficiency are also looked at, but this time from the point of view of discourse analysis. The aim of the study by Martina Temmerman and Dave Sinardet is to gain insight in political discourse by analysing news interviews with Belgian politicians in both their native and their nonnative tongue (Dutch/French) on the two public broadcasters, i.e. the Dutch-speaking VRT (Vlaamse Radio- en Televisieomroep) and the French-speaking RTBF (Radio-Télévision Belge de la Communauté Française). They investigate for instance whether Dutch-speaking and French-speaking journalists take a markedly different attitude towards the politicians they are interviewing. The differences they find in interviewing styles between the Dutch-speaking public broadcaster and its Frenchspeaking counterpart, tend to confirm earlier research. The results also confirm the current hypothesis on more general differences between the two public broadcasters. The interviews suggest that the French-speaking public broadcaster RTBF remains more strongly adherent to traditional public broadcasting culture than the Dutch-speaking VRT. In the chapter by Sylvain Dieltjens and Priscilla Heynderickx discourse analysis techniques are applied to business communication. The focus in this chapter is on the ambiguous we-form. Based on an analysis of source and translated texts (i.e. parallel texts), Dieltjens and Heynderickx demonstrate several ways in which the parallel texts help to interpret the we-form. Some of them are very straightforward (for example the use of a different pronoun), others are more complex (for example the implementation of an evasive strategy). They can indicate both a narrowing and an extension of the reference of the we-form. The study of political discourse and business discourse brings us to the fourth part of this volume, which covers chapters dealing with the study of specialised language and terminology in cross-cultural communication. Elizabeth Marshman presents some preliminary results of a research project focussing on the study of lexical knowledge patterns for identifying four non-hierarchical conceptual relations—cause–effect, association, process–patient and disease–sufferer—in corpora of medical texts in English and French. The goal of the project is to evaluate some of

xiv

Preface

the similarities and differences between the two languages that may affect the development and application of semi-automatic applications for the extraction of knowledge-rich contexts using lexical knowledge patterns. The data discussed in this article were generated by the analysis of two corpora composed mainly of research and review articles in the fields of breast cancer and heart disease, taken from specialized medical journals. That applications may benefit from terminological resources, is further elaborated in the chapter by Koen Kerremans. By discussing the results of two different research projects, the author demonstrates how multilingual terminological resources can play a role in semantic web technologies as dynamic entities. He reflects on application-oriented terminography as opposed to terminological dictionary production for human users and describes the Termontography methodology, a method which was worked out by the Centrum voor Vaktaal en Communicatie (a research centre of the Erasmus University College Brussels) for developing applicationoriented, ontologically-underpinned terminological resources. The method is supported by a software workbench, which is also presented in this chapter. Specialised discourses as dynamic entities is also discussed in the chapter by Carme Bach Martorell. The study focuses on a certain aspect taking place in the frame of texts: the study of the processes of creation and fixation of terminology in written texts, i.e. a process known as “stabilisation”. The aim of this chapter is to prove that specialised knowledge and the terminological units that dynamically convey it, are stabilized as a negotiation among the different groups of experts in each field. On the one hand, lexicographers and terminologists propose the terms which should be considered for the macrostructure of specialized dictionaries. On the other hand, professional groups stabilize and fix terminological units as well as their value within specialized discourses in a constant negotiation within previously established limits at a social and pragmatic level. The work presented in the chapter by Hee Sook Bae and Marie-Claude L’Homme is part of a larger project dealing with the conversion of a preexisting monolingual database into a multilingual dictionary. The general objective of the study is to expand a monolingual database containing French terms related to computing and the Internet to a multilingual dictionary in which several language modules are connected through a common model. The chapter deals more specifically with the construction and the adaptation required for a Korean module. The authors believe that this preliminary work on Korean will contribute to the development of a more general model for different languages.

Multilingualism and Applied Comparative Linguistics

xv

The final part in this volume deals with translation studies. The chapter of Marlén Izquierdo attempts to map out the resources frequently used by translators to convey the meanings expressed by –ing constructions functioning as circumstance adverbials. Starting from a corpus-based, contrastive analysis, the study tries to come up with an inventory of translational options which are produced in assumedly equivalent communicative contexts. Data have been taken from the ACTRES Parallel Corpus (henceforth P-ACTRES), a custom-made English-Spanish translation corpus compiled at the University of León. The intralinguistic analysis reveales that, on average, the English gerund-participle is frequently used to realize circumstance adverbials, which may express a great variety of semantic roles. The relevance of contrastive analysis for translators is further elaborated in the chapter by Rosa Rabadán, Belén Labrador and Noelia Ramón. The chapter sets out to present a course on corpus-based EnglishSpanish Contrastive Linguistics for Spanish students enrolled in an English degree. It is a compulsory third year course focusing on explicit instruction in L1-L2 differences. The course highlights the use of translation practice as a tool to grasp the functional semantic and pragmatic differences in a cross-linguistic study between two or more languages and uses corpora of two different types: bilingual comparable corpora and translation corpora. Informal tests show a marked difference between those students who had contrastive analysis training prior to the translation options offered to them and those who had not. Mª Pilar González Vera tries to show how the relay of power and gender relationships in a literary text is influenced by the process of translation. More specifically, the focus is on the translation of linguistic politeness and linguistic impoliteness strategies. The study of the two translations of The Crucible by Arthur Miller into Spanish, from different periods of time, one from the 50’s and another from the 90’s, aims to show how those target texts became a reflection of the process of change in interpersonal relationships within society. The comparative study shows how translations are subordinated to culture and ideologies. Apart from that, there is also a temporal dependence. Gender relationships as well as politeness are not static but they change with time. Therefore, the use of politeness strategies as well as the translation of those strategies, differ. In the final chapter Efthymia Karaouza discusses lexical cohesion, mood and modality deployed in three English translations of an Ancient Greek text remaining influential as part of the preamble of the European Constitution. The purpose of this study is to show that, although these

xvi

Preface

English versions render the same source text, the translators sometimes make different choices in terms of lexical cohesion, mood and modality. We hope you will enjoy reading this collection of research papers and get a taste of the cross-cultural spirit in which the MACL conference was held.

ACKNOWLEDGEMENTS

We would like to thank the following members of the MACL advisory board for reviewing papers that were submitted for publication in the present volume: Teresa Cabré (UILA Barcelona) Aline Remael (HIVTHogeschool Antwerpen), Marc Van Campenhoudt (ISTI Bruxelles), Sylvie Van Daele (Université de Montréal). We would again like to express our gratitude to all our co-members of the organising team of the MACL Conference and to all the volunteers (including the students of the Erasmus University College Brussels) who helped out so well during the three-day event.

PART ONE UNIVERSALITY AND CROSS-CULTURAL COMMUNICATION

CHAPTER ONE FIRE AND HOT LIQUIDS: ANGER IN HINDI AND ENGLISH VINEETA CHAND

Introduction This paper1 offers a comparison of anger metaphors in two languages (Hindi and English) and contributes to the growing body of work discussing the cross-linguistic manifestation of metaphors for the emotional state of anger. The conceptual metaphor theory (CMT) argues that certain metaphors, such as ANGER IS HEAT, are based on universally shared bodily experiences (Lakoff 1993). However, crosslinguistic work on anger metaphors has not presented unified evidence in favor of ANGER IS HEAT as an embodied, and hence, universal conceptualization of anger (Gevaert 2005). Divergent conceptualizations of anger may reflect and reify culturally bound differences or structural differences cross-linguistically. Expressions of anger are thus far unexplored in Hindi, an Indo-European language of northern India. This paper offers an initial foray into Hindi anger metaphors and metaphorical extensions and presents a cross-linguistic comparison with English metaphor data (Lakoff 1987 and 1993; Lakoff and Johnson 1980; Kövecses 1986; 2000a and 2000b). This research contributes to the CMT in two ways: first, it presents data from Hindi, contributing cross-linguistic metaphor data. Second, drawing from the similarities and differences at multiple levels within metaphor structure across Hindi and English, it highlights where culture can interface with language to structure metaphor systems, offering a new approach to discussing and comparing similarities and differences at multiple levels of the metaphor.

Multilingualism and Applied Comparative Linguistics

3

Past research Background on metaphors Until approximately 1978, metaphors were considered deviations from the literal whose main role was poetic and flowery; pleasing to the reader with images that help to maintain attention (Black 1962). However, current theories have reshaped the definition and role of metaphors in language and cognition. A growing body of linguists, cognitive scientists and psychologists now think that metaphors are ubiquitous in and pivotal to language: they work to mold our understanding of the surrounding world, while also reflecting our categorizations of the world (Gibbs 1994; Lakoff and Johnson 1980; Kövecses 1986; Yu 1998). Lakoff and Johnson (1980), considered the forbearers of the CMT, argue that metaphors are conceived of as a (partial) mapping between source and target domains in the form of target is source. The underlying metaphorical schema is referred to as the metaphor, and the actual instantiations of the mapping in language are considered metaphorical expressions. For example, the following are English examples of metaphorical expressions found in common speech, with the underlying metaphor in capitals: (1) Metaphorical Expression ĺ Metaphor a) He is getting hot under the collar. ĺ ANGER IS HEAT b) She spends her time unwisely. ĺ TIME IS MONEY (Lakoff 1993)

The target domains (anger, time) are mapped onto various source domains (heat, money), and a number of correspondences are then available to express and understand the target, given the entailments of the source domain. For example, the following shows a possible mapping from source onto target within an anger metaphor: (2) ANGER IS HOT FLUID IN A CONTAINER. (Kövecses 2002, 96) a) physical container ĺ angry person’s body b) top of the container ĺ rational self of the angry person c) hot fluid inside the container ĺ anger d) degree of fluid heat ĺ intensity of anger e) cause of increase in fluid heat ĺ cause of anger

Given the mappings associated with the metaphor ANGER IS HOT FLUID IN A CONTAINER, anger can be described using these terms.

4

Chapter One

Metaphorical expressions using entailments of the above mappings clearly show the transfer from source to target, as in 3 (a–i): (3) AN INCREASE IN ANGER IS A FLUID RISIN a) His pent-up anger welled up inside him. (Lakoff 1987) b) We got a rise out of him. (Lakoff 1987) c) My anger kept building up inside me. (Lakoff 1987) INTENSE ANGER PRODUCES STEAM d) She got all steamed up. (Lakoff 1987) e) I was fuming. (Lakoff 1987) f) Billy’s just blowing off steam. (Lakoff 1987) INTENSE ANGER PRODUCES PRESSURE g) I could barely contain my rage. (Lakoff 1987) h) He was bursting with anger. (Lakoff 1987) i) I could barely keep it in anymore (Lakoff 1987)

Lakoff and Johnson (1980) argue that some central metaphors are shaped by embodiment, the shared experience of being human and perceiving the world from this perspective (Lakoff 1987; Kövecses 2000), and are hence universal because they draw on and are universally based within the same core shared embodied experience. From this, they hypothesize that metaphors will be formed with a limited range of variation cross-linguistically: Lakoff argued that future cross-linguistic evidence would support this supposition. Heeding his challenge, the work of Yu (1998), Barcelona (2001), Kövecses (1986; 2000 and 2002), Kövecses and Palmer (1999) and Maalej (2004) on cross language comparisons of emotion metaphors (Chinese and English, Spanish and English, Hungarian, Chinese, Japanese, Wolof and English, Tagalog and English and Tunisian Arabic and English, respectively) as well as Alverson (1994) on cross-language and cross cultural conceptualizations of time in Sesotho, Mandarin Chinese, Hindi and English, MacArthur (2005) on conceptualizations of control and restraint in Spanish and English and Özçalıskan (2003) on conceptualizations of life, death and movement in Turkish have done much to strengthen the position of the CMT with respect to certain universal metaphors.

Problems with this model The underlying conceptual structure of certain metaphors is based in part on a shared embodiment, which, within the CMT is assumed to be

Multilingualism and Applied Comparative Linguistics

5

universal. However, newer cross-linguistic data shows that the majority of linguistic expressions vary from language to language and, it has been argued, from culture to culture (Emanation 1997; Yu 1998). As well, the evidence of universally shared emotional experiences (Lakoff 1997, Kövecses 2000) has been questioned (Gevaert 2005). Emanation (1997), examining Chagga, a Bantu language of Tanzania, argues that metaphors and embodiment cannot be conceived of as shaping or limiting culture; instead, the opposite must be true: metaphors are limited by, and are a subset of a larger cultural model. While offering a reasonable justification for culture’s role within CMT, Emanatian (1997) fails to offer a means of incorporating culture into models of metaphor structure. The expressions within each language are not haphazard, however: it is clear that these partial mappings adhere to a language internal structure, and a means of modeling this is necessary to advancing the CMT by integrating culture. Beyond the call for further work to test the validity and limits of the claims of the CMT in diverse languages (e.g. Lakoff and Johnson 1980; Kövecses 1986 and 1997; Gevaert 2005; Gibbs 1997), which has been heeded by research covering several languages and multiple metaphors, critical discourse regarding universality is needed. Much of the current metaphor research is focused on the overlap and similarity across metaphors, while comparative work focusing on the metaphoric expressions found across languages is sorely lacking, and, as the actual linguistic manifestation of metaphors, has something to offer to both a linguistic and cognitive assessment of metaphors cross-linguistically. Such work, I argue, would highlight areas problematic for the hypothesis of universal embodiment as a major constraint on certain conceptual metaphors. Within this, emotion tropes are a particularly rich area for metaphors, because of the abstract nature of emotions. There is also evidence that culture does interface with emotion (Gevaert 2005; Kövecses and Palmer 1999; Yu 1998). Minimally, it can be argued that further study and a focus on the metaphoric expression is necessary to justify the CMT position, while stronger “culturalists” argue that culture serves to create or structure our range of emotions (Quinn 1991), and evidence to this effect is waiting to be uncovered.

Anger Methods This data is from a variety of primary sources. The English expressions come from both introspection on my part as a native American English

6

Chapter One

speaker, and examples from the body of metaphor literature (cited as such within the text). The Hindi data, about 160 expressions in total, were collected during Spring 2004 from a variety of modern colloquial Hindi language guides and dictionaries (McGregor 1993; Tiwari and Kalia 1996), and were all verified individually by four native Hindi speakers from New Delhi, India.

English anger metaphors English has two dominant anger metaphors that are subsumed under the general metaphor of ANGER IS HEAT (Yu 1998, 50; with further support from Lakoff and Johnson 1980; Lakoff 1987; and Kövecses 1986). Yu (1998) posits that anger can be construed as a solid (4) or fluid (5): (4) ANGER IS FIRE a) Flames are coming out of his ears. b) You can see fire in his eyes. c) He was incensed over the remark. d) The fire in his heart couldn’t be contained. e) His angry remarks were blistering. (blister is the result of a burn) (5) ANGER IS A HOT FLUID IN A CONTAINER. a) He’s boiling over. b) He blew his top. c) He is stewing over that last comment. d) He’s seething. 2 e) He erupted with anger.

Citing Lakoff (1987), Yu (1998) argues that there is one dominant “most general” cultural metaphor for anger, that of ANGER IS HEAT. While this may be true, such claims would be better supported by evidence in this direction. Yu focuses on the above metaphors with no discussion of additional Chinese anger source domains, and other English anger source domains are under-theorized by Lakoff, who focuses on ANGER IS HEAT as both central and dominant (1993). Since these other source domains, their interactions, and their possible variable dominance crosslinguistically and cross-culturally have not been accounted for, they merit further discussion. Addressing this, I return to expressions originally mentioned by Lakoff and Johnson (1980) and incorporate with them further examples in order to explore a broader range of English anger

Multilingualism and Applied Comparative Linguistics

7

source domains, with the goal of understanding how well such mappings align across multiple target domains. (6) ANGER IS BAD WEATHER. a) He stormed around. b) He thundered his complaint. (7) ANGER IS DESTRUCTION a) He couldn’t hold it together. b) He fell apart. c) He’s spitting bullets. d) He went ballistic. (8) ANGER IS INSANITY a) He was insane! b) He went crazy when they lost the game. (9) ANGER IS AN OPPONENT a) He fought down his anger. b) He lost control of his anger and started shouting. c) He was wrestling with his anger. (10) EMOTIONS ARE A CONTAINER. a) He’s in a fury. b) He’s in a rage. (11) ANGER IS A WILD ANIMAL a) His anger was fierce. b) He bit the guy’s head off. c) He unleashed his anger on the innocent bystander. d) He tore that guy to pieces. e) He ripped him a new one.

8

Chapter One (12) ANGER IS DOWNWARD MOVEMENT ONTO A PATIENT a) The bulk of his anger was brought down on innocent bystanders. b) I was bowled over by his anger. c) He knocked them to the ground with his anger. d) Don’t take it out on me!

Target mapping overlaps The (potential) argument that the source domain of heat is primary for anger and is universally shared, is also problematic in that it fails to account for how the target domain of heat is also mapped onto other source domains. Love can also be conceptualized as heat, specifically fire: (13) LOVE IS FIRE a) She’s consumed with love. b) She’s burning up inside from her desire c) She’s hot/fiery with passion.

The metaphors LOVE IS FIRE and ANGER IS FIRE are difficult to conceptualize in tandem. Interpreting which metaphor is indexed requires contextual cues: the same phrase, as demonstrated in 14, indexes different metaphors based on the context: (14) a) Lisa hadn’t seen her husband in three months, but was more in love than ever, after their nightly phone calls. He was arriving tomorrow, and she was burning up inside. (passion or love) b) Lisa’s neighbor ran over her prized rosebush, laughing and unapologetic. Lisa was so angry, she was burning up inside. (anger)

What these two target domains share is that they are both strong emotions. Perhaps fire is not separately conceptualized as anger and love, but is instead, a conceptualization of strong emotions in general. Target mapping overlaps are relatively common: the target domain of insanity is also not limited to anger: (15) EXTREME HAPPINESS IS INSANITY a) She went crazy with happiness. b) When she learned the good news she was hysterical. c) She was out of her mind with joy at finding her children.

Multilingualism and Applied Comparative Linguistics

9

Additionally, in English, anger is not constrained to a hot liquid: (16) ANGER IS A LIQUID a) He was pissed (off). b) He was spitting anger.

While Lakoff (1987, 385) argues that 16a refers to liquid which is under pressure and hence hot, I am not convinced of this interpretation as there is no direct or implicitly assumed temperature reference. Also complicating the relationship between anger and liquids, the metaphor ANGER IS A POISON can be solid or liquid: (17) ANGER IS A POISON a) His words were poison. b) The venom of his anger cut into me deeply.

While pressure, fire and heat are strongly associated with anger, not all references that include such connotations refer to anger. The following expressions are all references to fire and heat which are not associated with anger: (18) a) He’s puffed up with pride. (pride) b) He’s full of hot air. (bragging, pride) c) He’s all fired up. (excited) d) He’s on fire. (attractive, excelling at something, excited) e) He’s smoking. (attractive, excelling at something) f) He’s hot. (attractive) g) He’s a hottie. (attractive) h) He sizzles. (attractive) i) He’s too hot to touch. (attractive) j) He’s all the rage. (popular) k) He’s a firebrand. (trouble or mischief maker)

While some of the past CMT literature may leave the impression that ANGER IS HEAT is both universal and embodied, this is problematic, given the target domain overlaps and the liquid examples which do not involve heat. The source domains associated with anger do have a commonality: poison, fire, hot liquids, bad weather, wild animals and downward forces are all dangerous. While this data has made clear-cut distinctions less feasible, the English anger data is clearly systematic. My

10

Chapter One

analysis demonstrates that many generalizations but fewer absolutes exist within the realm of anger metaphors. We will now compare Hindi data.

Hindi anger metaphors The metaphor with the most Hindi expressions is ANGER IS HEAT, manifesting as ANGER IS FIRE in 19 and 20, and as ANGER IS A HOT LIQUID in 21. Following Yu’s Mandarin examples (1998), Hindi data is presented in the following format: the first line is in the Hindi script, Devanagri, and in roman italics, the figurative meaning. The second line has a word-by-word literal translation followed by the same translation in English word order. (19) ANGER IS FIRE a) To burn within (on account of jealousy) Burning ember on to roll = To roll on burning ember To add fuel to the fire b) Fire in ghee (clarified butter) to cast = To cast ghee on fire c) To add fuel to the fire Burn-wound on salt apply = To put salt on a burn wound To add fuel to the fire, to fan the flames d) Fire to blow into flame = To aggravate a fire To be set on fire e) Fire apply = To set to fire f) To get enraged Heart in fire to happen = To feel fire in heart g) A firebrand or hot-headed person inciting violence Fire of effigy/puppet = An effigy/puppet driven by fire h) To fly into a rage A flame glowing to become = To become a glowing flame

In keeping with the mappings possible under the ANGER IS FIRE metaphor, increasing the fire would stand for an increase in anger, and a decrease in the fire stands for a lessening of the anger. Following this to its conclusion, completely putting out a fire is associated with a complete dispersal of anger.

Multilingualism and Applied Comparative Linguistics

11

(20) To become furious a) Anger flame to rise up = For the anger to flare up b) To put out a fire Cold make to do to other = To cool someone/something down c) For anger to vanish/disappear/cool down Anger wind be to go = For the anger to become wind

Anger is also conceptualized in a liquid state in Hindi, similar to English. The focus of the Hindi data is on the internal liquids of the body as boiling, or of the whole body as boiling. (21) ANGER IS HOT LIQUID a) To become murderously angry Blood to boil = For the blood to boil b) Bile to boil = For the bile to boil

To be enraged

To be boiling or to swell with emotion (as with rage, ride) c) To boil = For someone to boil

These metaphors associated with heat encompass only a fraction of the total expressions gathered: the remaining expressions, subsumed under the following metaphors, suggest that perhaps there is no dominant metaphor for anger in Hindi. It has been argued that expressions denoting anger which use liquids fall under the sub-metaphor anger is a hot liquid, which is from the general metaphor ANGER IS HEAT (Lakoff and Johnson 1980; Lakoff 1987; Kövecses 1986). However, the following Hindi data set disproves this by offering examples which do not have heat. This is not unique to Hindi. We have seen similar English metaphors (e.g. 16 and 17) also lacking reference to heat. (22) ANGER IS LIQUID a) To forgive and cool down, to shake off one’s anger Anger spit to make = To spit out one’s anger

12

Chapter One

b) To suppress one’s anger Anger to drink to go = To gulp down one’s anger To burst out in anger; for anger to rain down on others c) Rain to fall = To rain down on others d) To bear something in silence, to suppress one’s anger Speech drink to do = To swallow the matter

A (possible) sub-mapping of the above metaphor is ANGER IS A POISONOUS LIQUID. This is also found in English, in set 17. However, while the English poison source, metaphoric expressions are not limited to liquids, the Hindi expressions do not extend to non-liquids. (23) ANGER IS POISONOUS LIQUID a) To speak/write poisonous things Poison to spit out = To spit out poison

Anger is also conceptualized within the container metaphor, as a substance under pressure. Increasing the pressure on the container causes the anger to appear. This example also does not appear to use the ANGER IS HEAT basic level metaphor: (24) CONTAINER METAPHOR a) Ill-will or disaffection rising Heart in knot to make = To make a knot in the heart To bear anger quietly b) Chest on stone to stay/rest = To put a stone-slab on one’s chest

In 24a and b, the heart and chest respectively contain the anger. Sufficient pressure is offered to keep the anger under wraps in 24b. The next two sets show that anger is conceptualized as being outside of one’s bodily container in Hindi: (25) ANGER IS BEING OUTSIDE ONESELF a) To completely lose one’s temper Oneself of beside to become = To become beside of oneself

Multilingualism and Applied Comparative Linguistics

13

b) To hit the ceiling One in not remain = To not be able to contain oneself; to be beside oneself

There exist similar possibilities in English, with expressions like She was beside herself with rage. However, I was unable to come up with more English expressions within this schema, and, given this, argue that the range and number of expressions in Hindi can be tentatively considered broader. Similar to the English expression “a chip on one’s shoulder,” we find 26(a). Both metaphors imply that the anger is a visible substance, and both have the connotation of a ‘loose cannon’. In English, the focus is on the shoulders, which carry burdens and show one’s emotional state. Kövecses (1986) argues that He has a chip on his shoulder falls within ANGER IS A BURDEN, where responsibilities are understood as burdens (ibid., 27). This generalization is odd. Anger is not a responsibility to which one has a duty; it is, rather, the opposite. One’s responsibility is to control and contain anger. It is more appropriate to understand the phrase, and the Hindi equivalent in 26a, as expressions stemming from the conceptualization of anger being a physical presence: carrying this physical entity in a highly visible location, on one’s shoulder or one’s nose, is alerting others to the presence of the anger. It is not hidden, and hence it is easily accessible, through provocation. (26) BEING EASILY PROVOKED IS VISIBLE ANGER. a) To have a chip on one’s shoulder, easily provoked Nose on anger to become = To have anger on (one’s) nose/outside

Anger can also be conceptualized as being possessed by a supernatural power: (27) ANGER IS BEING POSSESSED BY A SPIRIT a) To be possessed by intense anger Head on demon to rise up to become = For a demon to ride on one’s head b) To become devilishly/destructively angry Anger of because spirit to become = To become a ghost because of anger

14

Chapter One

Interestingly, in English one can say She was possessed (with anger), which leaves unexpressed, but understood, the possessor as the devil within most Christian sects. This raises the question of why one language (or culture), within this data set, consistently expresses the possessor, while the other elides that agent. Initially, it suggests that within a monotheistic culture, one does not need to express who would possess one, resulting in anger. It is obviously the devil. Whereas, within a pantheistic culture, it becomes important to explicitly state the agent’s identity, given multiple possible agents. There is an association with anger as a physical force which can destabilize or knock over its objective. These Hindi examples highlight a downward force associated with anger. (28) ANGER IS DOWNWARD FORCE a) To look murderously angry Eye in blood down to come = For blood to descend in the eyes To burst out in anger; for anger to rain down on others b) Rain to fall = To rain down on others c) Pouring forth of abusive/angry speech Ember to rain/shower = To rain a shower of embers To burst out in fuming rage d) Fire to rain/shower = To rain a shower of fire e) To bring down anger (on someone) Anger to descend on = To descend the anger (on someone)

Hindi associates anger with being an animal, similar to English examples 11(a–h). Both allow expressions which focus on the defensive and offensive posturing of an animal, and allow for the attack and destruction of the victim, but only Hindi has expressions which focus on the next stage of an attack, consuming the victim. (29) ANGER IS AN ANIMAL a) To growl from anger Anger from growl = To growl in/from anger b) To react in an angry/ill tempered manner Food run/gallop = To run to bite/eat

Multilingualism and Applied Comparative Linguistics

15

Denotes a really angry outburst; to be really c) pissed Raw to eat = To eat raw

This section does show clear systematicity across English and Hindi metaphors. We will now explore in detail the areas of convergence and divergence.

Analysis of variation From this data, it is evident that the metaphors are systematic and different in their focus, yet, there are positive correlations across the two Indo-European languages. It is not possible to simply translate word-forword, but there is evidence for underlying conceptual similarities in Hindi and English: the dominant metaphors can be subsumed under the general metaphor ANGER IS HEAT with submappings onto fire and hot liquid present. However, this data has presented evidence against the notion that at every level, metaphoric paradigms are universal or cross-linguistically motivated by identical forces. Following from this, the CMT focus on and labeling of the basic-level metaphor ANGER IS HEAT as universal occludes discussion of the metaphoric expression, as well as intermediate levels of conceptualization. It does not follow, nor is it evident in the data, that universality at one level necessitates universality or even strong correlations, at other (lower) levels. A close look at this data, with the goal of describing the entire metaphor structure, requires both distinguishing between levels of the metaphor and between languages’ organization and recognizing the use of similar, identical or totally different mappings. With this goal in mind, the Lakovian approach is clearly insufficient. While “Lakoff and Kövecses suggest two ways of testing their claim that the ANGER IS HEAT conceptualization is embodied: [through] diachronic research and cross-linguistic research” (Gevaert 2005), the majority of earlier research (e.g. Alverson 1994; Lakoff and Johnson 1980; Lakoff and Kövecses 1987; Kövecses 2000a; Yu 1998) has focused on wholesale universal or language-specific judgments which disallow a nuanced discussion of shared or language-specific traits. This research has supported the notion of embodiment as a universal force constraining and motivating source domain choice and thus has ignored potential variability in the internal structure of the metaphor. However, quantitative metaphor research is rare (e.g. MacKay 1986), and diachronic research absent until Gevaert (2005). Gevaert’s diachronic work is particularly damning for the

16

Chapter One

CMT claim regarding embodiment as a universal motivation and constraint: examining anger expressions from Old and Middle English, she demonstrates that the conceptualization ANGER IS HEAT is a cultural product developed from Latin and the humoral doctrine. Metaphor research thus clearly needs, and lacks, a means of discussing and quantifying variation at multiple levels. Introducing different types of variation and explicit discussion of the metaphor hierarchy will permit a more fine grained description of each language’s metaphors. This in turn will allow quantitative comparisons of multiple languages’ metaphors at various levels. We need to begin by regularly construing metaphors within their broader hierarchy (Figure 1-1). Within this data-driven analysis, which forces one to examine the variation and similarity at the different levels of the metaphor, I argue variation manifests within and across the Hindi and English data in at least three ways: at the level of basic-level and sub-level metaphors we find instances of no overlap, at the level of sub-metaphors and the number of metaphorical expressions, we find different degrees of concordance, and across all levels, interacting with cultural and linguistic features of each language, we find variation in active zones. %DVLF/HYHO 0HWDSKRU

6XE/HYHO 0HWDSKRU

6XE/HYHO 0HWDSKRU

0HWDSKRULF H[SUHVVLRQV Figure 1-1. Metaphor hierarchy: model 1

Variation in overlap One way metaphors show variation is the presence or absence of a construction at one or more levels in one of the languages. Further evidence of this comes from Yu’s comparison of Mandarin and English (Yu 1998). He argues that English completely lacks the dominant Chinese

Multilingualism and Applied Comparative Linguistics

17

sub-metaphor ANGER IS HOT GAS IN A CONTAINER. Within this data, we see that English evidences, and Hindi lacks, the metaphors ANGER IS DESTRUCTION, ANGER IS INSANITY, ANGER IS BAD WEATHER, and ANGER IS AN OPPONENT. Present in Hindi, but not in English, are the sub-level metaphor ANGER IS BEING POSSESSED BY A SPIRIT.

Variation in concordance Within a metaphor, source domains have unequal distributions, which can be measured by quantifying metaphorical expressions associated with different subdomains. Examining the distribution of metaphorical expressions has descriptive power: it allows for discussion of the proportioning of metaphorical expressions relative to sub-level metaphors within a metaphor. This can be applied both within a language, to quantify the proportion of different metaphors and sub-level metaphors used to discuss a target domain like anger, and across languages, in a similar manner. Figure 1-2 is a generic example illustrating relatively unequal proportioning, with Sub-Level Metaphor B having a much higher concordance within Metaphor X than Sub-Level Metaphor A. %DVLF/HYHO 0HWDSKRU;

6XE/HYHO 0HWDSKRU$

6XE/HYHO 0HWDSKRU%

0HWDSKRULF H[SUHVVLRQV

Figure 1-2. Metaphor hierarchy: model 2

Source domains with cross-linguistic overlap (e.g. heat, a source domain for anger expressions in both Hindi and English), utilized within the languages to different degrees, can thus be compared. That is, while two languages may share a sub-level domain, (e.g. the domain of fire to

18

Chapter One

express anger), they may differ in the centrality of this source domain, as expressed by the range and quantity of metaphorical expressions. I propose that measuring the proportion of metaphorical expressions within a data set associated with each source domain is one way of quantifying the commonality or typicality of a source domain. I call this variation in source domain exploitation concordance. Within a language, different proportions of concordance are associated with particular source domains, and comparisons of a single source’s concordance levels across languages will also potentially demonstrate variation. One can thus compare the relative centrality of different source domains as a means of conceptualizing the target domain. Within the goal of quantifying metaphoric expressions, it is important to discuss the potential productivity of metaphors (Taub 1996). Instantiations of a conceptual metaphor may not form a finite ensemble3 of metaphoric expressions. While no work on metaphors has taken a corpus based approach to examining how productive these metaphors are (e.g. the frequency of “common” expressions versus the frequency of new and creatively constructed metaphoric expressions) I argue that it is reasonable to suppose that within a period of time and in a particular context there may be a number of metaphorical expressions more used than others. Unfortunately, there’s a lack of quantitative cross-linguistic synchronic work to use in demonstrate concordance. This data was originally gathered for the primary purpose of uncovering some stereotypical or central means of expressing anger in Hindi. Its differences from English motivated the need to reevaluate the Lakovian means of comparison and instead quantify concordances within a comparison. While I do not submit this data as an ideal example of appropriate data collection for quantitative purposes, given that it is on hand, it will be used to model how to make explicit source domain variation crosslinguistically. Thus, the following Hindi and American English concordances do not necessarily model proportions for naturally occurring speech or text, though these proportions may be verified as representative of Hindi and English if supported by further corpus-based work. Given these qualifications of the data set, I have limited my generalizations to approximate equality or heavy weighting for one domain over another, while more explicit quantification would be the next step. Further work in this vein would hence be based on, and accompany, quantitative corpusbased work. The preliminary results show that English has an equal concordance for the source domains of fire and hot liquid. One could argue that this shows the relative centrality of both source domains, however, Hindi

Multilingualism and Applied Comparative Linguistics

19

demonstrates an imbalance in the range and number of expressions, having a much higher concordance for fire than for hot liquids. At this exploratory point, it suggests that fire is a more central and dominant means of conceptualizing anger in Hindi. Hindi also demonstrates higher concordance with respect to the metaphors ANGER IS BEING OUTSIDE ONESELF and ANGER IS A LIQUID.

Variation in active zones Holding basic and sub-level metaphors constant, we also see variation in the actual linguistic expressions in two ways. The first is structural; the phrasing is not identical, due to variance in grammar, lexicon and syntax which is completely expected in cross-linguistic comparisons. Secondly, different aspects of the sub-level source domain can be systematically highlighted, and then expressed in metaphorical expressions, varying cross-linguistically. These variations have been argued in broad terms to be either another manifestation of culture’s influence on language or the opposite. The Sapir-Whorf hypothesis suggests that language shapes and constrains cognitive constructions. Regardless of which constrains the other, thus ‘primary,’ such variations is synchronically describable, and hence, comparable. Their description requires calling on another concept from Cognitive Grammar, the active zone phenomena: “[O]ften, when entities participate in a situation, only some facets of the participating entities are implicated.” (Taylor 2002, 110). Considered ubiquitous and first introduced by Langacker (1991), active zones capture how different facets of an entity are highlighted, brought into focus, or the only part possibly involved with respect to the other utterance components. We see how different facets of the house are highlighted in different contexts: (30) a) I painted the house. b) I fumigated the house. c) I sold the house.

In 30a, only the external part of the house was painted, in 30b, only the internal portion of the house was fumigated, and in 30c, a property boundary larger than the house itself is referenced. With respect to metaphors, the sub-level source domains (i.e. fire and hot liquid), while shared across the two languages, do not instantiate the exact same metaphorical expressions in each language. Additionally, certain

20

Chapter One

entailments, or facets of the source domain are not possible, e.g. Taylor’s example of how one never speaks of the ‘basement of a theory’ within the metaphor THEORIES ARE BUILDINGS (2002, 495; originally from Clausner and Croft (1997)). Extending Langacker’s concept, I argue that the active zone for a particular source domain can vary across languages, and is influenced by the associated language and culture (Figure 1-3). The source’s active zone highlights language- or culture-specific notions of the parts of the source domain most intimately involved, which then manifesting linguistically. The active zone of the source domain can vary, and demonstrates a third possible area for metaphor variation crosslinguistically. I demonstrate this with arrows flowing between the sub-level domain and the language and culture, interacting with the sub-level metaphor, to collaboratively produce the language specific metaphoric expressions. %DVLF/HYHO 0HWDSKRU

+LQGL /DQJXDJH  &XOWXUH

6XE/HYHO 0HWDSKRU

(QJOLVK /DQJXDJH  &XOWXUH

0HWDSKRULF H[SUHVVLRQV

Figure 1-3. Metaphor hierarchy: model 3

Variable active zones provide a means of accounting for metaphor and submetaphor level generalizations as well as metaphoric-expression-level differences. Sub-level metaphors are available universally: given that each language chooses some and not others, and also chooses to instantiate different aspects of shared source domains, this is tentative evidence that specific expressions instantiate sublevel metaphors according to individual languages’ active zones.

Multilingualism and Applied Comparative Linguistics

21

Conclusion The structure and instantiation of some metaphors and metaphoric expressions for anger in English and Hindi have been discussed here. While English and Hindi share the same basic level metaphorical conception of anger as heat, the source sub-mappings (or domains), entailments and metaphoric expressions vary across languages, and the metaphoric expressions focus on different aspects of heat. Closer examination has revealed that (at least) three types of variation should be considered in cross-linguistic metaphoric analyses: overlap, concordance and active zones. Within the ANGER IS HEAT metaphor, Hindi focuses on fire and hot liquids, with preliminary results suggesting that fire dominating hot liquids in Hindi, while English highlights both fire and liquids, with both having equal emphasis in the data. I have provided discussion on how best to approach and analyze metaphor data with respect to overlap, concordance and active zones, which should prove fruitful for future studies. Further refinements of this overall assessment would, of course, be dependant on much larger, corpusbased studies of the number and range of occurrences. The data conclusively show that metaphoric expressions vary considerably across the two languages. While some metaphors might be universal, it is also clear that there is wide variation in how the metaphors are expressed in these different languages, and which aspects of domains are used.

Notes 1. Versions of this paper were presented at SALA 25, LASSO 34 and the 2006 MACL conference, and helpful comments and critiques from all of these have helped my development of these ideas. I’d also like to thank an anonymous reviewer for the insightful comments and corrections. All remaining errors are, of course, my responsibility. 2. An anonymous reviewer asked about the lava being fire or a liquid. Polling native English speakers demonstrated an interesting conflict: lava was unanimously considered a liquid, but many would not define it without reference to fire—this does warrant further exploration. 3. Thank you to an anonymous reviewer for pointing this out.

References Alverson, Hoyt. 1994. Semantics and Experience: Universal Metaphors of Time in English, Mandarin, Hindi and Sesotho. Baltimore: Johns Hopkins University Press.

22

Chapter One

Barcelona, Antonio. 2001. On the systematic contrastive analysis of conceptual metaphors: case studies and proposed methodology. In Applied Cognitive Linguistics II: Language Pedagogy, ed. M. Pütz, S. Niemeier and R. Dirven, 117–146. Berlin: Mouton de Gruyter. Black, M. 1962. Models and Metaphors. Ithaca: Cornell University Press. Clausner, T. C., and W. Croft. 1997. Productivity and schematicity in metaphors. Cognitive Science 21(3): 247–282. Emanation, Michele. 1997. Congruence by Degree: On the relation between metaphor and cultural models. In Metaphor in Cognitive Linguistics. Selected Papers from the Fifth International Cognitive Linguistics Conference, ed. Raymond W. Gibbs and G. J. Steen, 205– 218. Amsterdam: John Benjamins Publishing Company. Gevaert, Caroline. 2005. The Anger is heat question. In Perspectives on Variation: Sociolinguistic, Historical, Comparative, ed. Nicole Delbecque, Johan van der Auwera and Dirk Geeraerts, 195–208. Berlin: Mouton de Gruyter. Gibbs, Raymond W. Jr. 1994. The Poetics of Mind: Figurative thought, language, and understanding. Cambridge: Cambridge University Press. —. 1997. Taking Metaphor out of our Heads and Putting it into the Cultural World. In Metaphor in Cognitive Linguistics. Selected Papers from the Fifth International Cognitive Linguistics Conference, ed. Raymond W. Gibbs and G. J. Steen, 145–166. Amsterdam: John Benjamins Publishing Company. Kövecses, Zoltán. 1986. Metaphors of Anger, Pride, and Love. A Lexical Approach to the Structure of Concepts. Philadelphia: John Benjamins Publishing Company. —. 1997. Metaphor: Does it constitute or reflect cultural models? In Metaphor in Cognitive Linguistics. Selected Papers from the Fifth International Cognitive Linguistics Conference, ed. Raymond W. Gibbs and G. J. Steen, 167–188. Amsterdam: John Benjamins Publishing Company. —. 2000(a). Metaphor and Emotion: Language, Culture, and Body in Human Feeling. Cambridge: Cambridge University Press. —. 2000(b). The Scope of Metaphor. In Metaphor and Metonymy at the Crossroads: A Cognitive Perspective, ed. Antonio Barcelona, 79–92. Berlin: Mouton de Gruyter. —. 2002. Metaphor: A Practical Introduction. Oxford: Oxford University Press. Kövecses, Zoltán, and Gary B. Palmer. 1999. Language and Emotion Concepts: What Experientalists and Social Constructionists Have in

Multilingualism and Applied Comparative Linguistics

23

Common. In Languages of Sentiment: Cultural Constructions of Emotional Substrates, ed. Gary Palmer and Debra Occhi, 237–262. Amsterdam: John Benjamins Publishing Company. Lakoff, George. 1987. Women, Fire and Dangerous Things. Chicago: University of Chicago Press. (with Zoltán Kövecses for Case Study 1). —. 1993. The Contemporary Theory of Metaphor. In Metaphor and Thought, ed. Andrew Ortony, 202–251. Cambridge: University Press. Lakoff, George, and Mark Johnson. 1980. Metaphors we live by. Chicago: University of Chicago Press. Langacker, Ronald. 1991. Concept, Image, and Symbol. The Cognitive Basis of Grammar. Berlin: Mouton de Gruyter. McGregor, R.S. 1993. The Oxford Hindi-English Dictionary. Oxford: Oxford University Press. Maalej, Zouhair. 2004. Figurative language in anger expressions in Tunisian Arabic: An extended view of embodiment. Metaphor and Symbol 19(1): 51–75. MacArthur, Fiona. 2005. The competent horseman in a horseless world: Observations on a conventional metaphor in Spanish and English. Metaphor and Symbol 20(1): 71–94. MacKay, Donald. 1986. Prototypicality among metaphors: On the relative frequency of personification and spatial metaphors in literature written for children versus adults. Metaphor and Symbol 1(2): 87–107. Özçalıskan, Seyda. 2003. ‘In a caravanserai with two doors, I am walking day and night’: Metaphors of death and life in Turkish. Cognitive Linguistics 14 (4): 281–320. Quinn, Naomi. 1991. The Cultural Basis of Metaphor. In Beyond Metaphor: The Theory of Tropes in Anthropology, ed. J.W. Fernandez, 56–91. Stanford: Stanford University Press. Taub, Sarah. 1996. How Productive are Metaphors? A Close Look at the Participation of a Few Verbs in the States are locations Metaphor (and Others). In Conceptual Structure, Discourse and Language, ed. Adele E. Goldberg, 449–462. Stanford: Center for the Study of Language and Information Publications. Taylor, John R. 2002. Cognitive Grammar. Oxford: Oxford University Press. Tiwari, Bhola Nath N., and Dvijendera Nath Kalia. 1996. How Do You Say It In English: A Dictionary of Hindi-English Idioms and Proverbs. Delhi: Aditya Prakashan. Yu, Ning. 1998. The Contemporary Theory of Metaphor: A Perspective from Chinese. Amsterdam: John Benjamins.

CHAPTER TWO PLANNED LANGUAGES AND LANGUAGE PLANNING: THE CONTRIBUTION OF INTERLINGUISTICS TO CROSS-CULTURAL COMMUNICATION FEDERICO GOBBO

A new science is developing, Interlinguistics–that branch of the science of language which deals with the structure and basic ideas of all languages with the view to the establishing of a norm for interlanguages, i.e. auxiliary languages destined for oral and written use between people who cannot make themselves understood by means of their mother tongues. —Otto Jespersen (1931)

Since the end of the 19th century up to the first half of the 20th century, the quest for “a norm for interlanguages” was a hotly debated issue among linguistic scholars and amateurs. In fact, about 1,000 language projects were proposed in that period, especially in Europe: the strongest effort for cross-cultural connection ever made (Albani-Buonarroti 1994). What can be inferred from it, in terms of language planning? Is there any relation or analogy between the issues encountered in natural language planning contexts (e.g. officialization, language revitalization, standardization) and interlanguage planning? Are there any general rules that can be deduced from the main historical developments of the above mentioned “quest”? Although Tauli (1968) considered interlinguistics as “the science of international language planning” (i.e. a branch of the science of language planning) very few linguists involved in language planning paid attention to interlinguistics until now. In this analysis of planned international language, the role of language amateurs will not be underestimated. Their importance had already been recognized by Wilhelm Ostwald in an influencing book about the pursuit of the proper interlanguage1. Language amateurs often attack problems

Multilingualism and Applied Comparative Linguistics

25

more bravely than professionals, i.e. linguists. Moreover, it should be emphasised that, during the period considered, modern language studies were relatively new, and interlinguistics was hardly considered reliable. In fact, in 1866 the Société linguistique de Paris decided to reject interlinguistics as a whole2. This paper intends to show how, while most issues encountered in planning interlanguages are similar to those met in natural language planning, some are very subject-specific. Furthermore, it wishes to point out some rules of planning an interlanguage that are of general relevance and interest in language planning as well as in cross-cultural communication. In the first section, a consistent taxonomy of planned languages will be presented that will allow us to analyse these languages within the framework of a language policy and planning, such as the one proposed by Hornberger (2006, 27-35). In the second section, the efforts directed toward the codification of the language system, in particular graphisation and standardization of the lexicon (corpus planning), will be investigated. In the third section, focus will be on the uses of the interlanguages (status planning) and on how users relate to them (acquisition planning). Specific attention will be paid to Esperanto, as its speakers are the widest speech community among interlinguistic communities. Conclusions will include a research agenda devised for the empirical study of planned languages and will highlight the novelties of this research3.

A taxonomy of planned languages The term 'interlinguistics' was coined by Jules Meysmans in 1911 (Blanke 1985) and became widely known after Otto Jespersen’s lecture at the Second International Conference of Linguistics in Geneve in 1931. In its broad sense, interlinguistics is mainly concerned with the study of every aspect of intercultural communication, including the roles, structures, ways of development and application of ethnic and planned languages intended as international means of communication (Schubert 1989). In a narrower sense, interlinguistics studies 'planned languages', originally a German term (Plansprachen) invented by Eugen Wüster in 1931 (Blanke 1985; Blanke 2006). A planned language can be described as a complete, self-enclosed system (a langue in Saussurian terms) which starts as written communication before being used in spoken communication4. This definition excludes the so-called a-priori languages, i.e. philosophical taxonomies of human knowledge proposed as 'perfect

26

Chapter Two

languages' for semantics, or codes of written communication published mainly in the 17th century (Yaguello 1991; Eco 1993). The formal languages used by logicians and computer scientists, such as first-order predicate calculus or computer languages as Algol, Pascal or Smalltalk are deemed as a-priori5. Bausani (1974) made an effort to compare planned languages on the basis of their structures and functions. His observations lead to a consistent and original taxonomy of planned languages, organized along two main axes: publicity and purpose (cf. Figure 2-1).

9RODSN

DX[LOLDULW\

SXEOLFODQJXDJHV

VHFUHWODQJXDJHV 7RONLHQ¶V ODQJXDJHV

,QWHUODQJXDJHV

-XVWIRUIXQ

.OLQJRQ

Figure 2-1. A taxonomy for planned languages

The publicity axis concerns the exposure of the language system. On the one hand, there are exoteric languages, i.e. languages whose form and meaning are explicitly secret, such as Balai-balaan, a language invented for religious purposes (Bausani 1974). On the other hand, there are public languages, i.e. languages whose grammar and vocabulary (sometimes even its phonetics6) are made widely available to the general public. The purpose axis concerns the original functions (the purposes that the language author originally had in mind) compared with the actual ones (the functions that the single interlanguage community members pursue). On the one hand, there are planned languages that are invented just for fun (e.g. for literary of fictional purposes), such as Tolkien's languages (e.g. Klocko 2002) and Klingon (see Okrand 1992 and Gobbo 2005b for more details). On the other hand, if the language planner intends to establish a community of practice, its members will use it mostly as a means of crosscultural communication, i.e. auxiliarity in Jespersen's terms. An interlanguage differs from a planned language in terms of its communication purposes and publicity. An interlanguage is developed for

Multilingualism and Applied Comparative Linguistics

27

international communication and is therefore meant for the public. Note that the terms 'interlanguage' and 'interlingua' have different meanings: (a) a language variety of L2 learners (i.e. with influences from L1 or overregularization traits); (b) a formal language in machine translation systems (Gobbo 2006); (c) ‘Interlingua’ is the original name of the planned language by Giuseppe Peano, better known as Latino Sine Flexione; (d) the name also refers to the planned language of the International Auxiliary Language Association (IALA), as published under the direction of Alexander Gode; (e) an international auxiliary language, i.e. a complete planned language system intended as an international means of communication (Jespersen 1931). If not noted otherwise, the term interlanguage in this last sense will be used from now on.

Corpus planning and interlanguages Corpus planning deals with both codification (language's form and linguistic aims) and elaboration (semi-linguistic aims) of a given language (Hornberger 2006). Interlinguistics is considered here as a special case of corpus planning, since the main function of interlanguages is crosscultural communication. In fact, unlike natural languages, an interlanguage always starts off as a written medium at a given time and has an identifiable creator. The first step in the analysis of a planned language is to detect the very moment the language gets published. Crucial issues as standardization, graphisation and reform become in fact very different after the “primitive contract” is settled. If in the case of natural languages the subscription of the primitive contract is conventional, while in the case of planned languages it is explicit, as seen in 1908-9 already by Ferdinand de Saussure, in his famous second course of general linguistics (Simone 1970, 42; my translation, from Italian): The primitive contract gets confused with everyday's language life. A system of signs as a language is received passively by the next generations. In any case the system of signs have the property to be transmitted by laws of their own, independent from the ones that settled the original contract (even if there is an explicit agreement, as in the case of Esperanto). The moment the contract is accepted, no one is the owner anymore. A language is like a goose brooded by a hen. After that moment, a language enters its semiological life, and it is impossible to get back.7

From the above text segment, we can infer a general rule about language evolution, nl. the primitive contract rule, which involves the fact that a human language, either a natural or a planned one, cannot be re-

28

Chapter Two

planned (i.e. touched in the language core–phonetics, morphology, syntax) once the “primitive contract” is settled, even by its author, as it created its own mystique of appartenence and permanence while entering its “semiological life”. As a consequence, a 'language project' (Blanke 1985) is an interlanguage before its publication (Figure 2-2). SULPLWLYHFRQWUDFW SURMHFWSKDVH

VHPLRORJLFDOOLIH QRPRUHUHIRUPV 

Figure 2-2. The two moments in the life of an interlanguage

In dealing with the distinctive features of planned languages, it should be noted that there are two crucial moments in language planning: before publication and after publication of a planned language. If anyone (even the interlanguage original author) intends to reform the language, i.e. to change some structural properties, after its publication, either a language shift to an existing interlanguage or a speech community split can occur. In this last case, a publication of a new language project usually derived from the original one. This is further illustrated by two case studies in the following subsections. We will first briefly discuss the language shift from Volapük to Esperanto. Afterwards, we turn to Ido, the planned language which was largely based on Esperanto. In both cases, the original speech community will be engaged in the defence of its interlanguage8.

The language shift Volapük-Esperanto The inextricable paradox of Volapük was the main cause of the language shift in favour of Esperanto. In fact, Volapük activists tried to reform the language after publication, i.e. to make a structural language planning change, although this was not successful, partly because of Schlyer's inflexibility (Blanke 2006). As a result, there had been a shift towards the use of Esperanto, which some German volapükists (who became esperantists afterwards), in particular the members of the Nürnberg club, helped to develop (Forster 1982).

Multilingualism and Applied Comparative Linguistics

29

Ido, the Ausbausprache of Esperanto In 1900, at the Exposition Universelle in Paris9, Louis Couturat, a mathematician and philosopher, launched a call for studies so as to find out the more suitable form of the “auxiliary international language”. Couturat wanted to create an autonomous normative language system with respect to Esperanto, despite retaining most of the properties of the latter. A text in Ido is deliberately intelligible for an Esperanto reader, so that Esperantists could pass to the new project with great facility. Ido should be considered as an Ausbausprache of Esperanto, as an Idist and an Esperantist may converse one another very readily, even if the respective communities of practice are quite separate. In fact, most devotees engaged in defence of Esperanto: the publication of the new interlanguage was significatively called the “Ido schism” (Forster 1982)10. What follows is a text fragment in Ido: La ideo pri mondo-literaturo, quan Herder e Goethe konceptis esence del vidpunto dil arto, ganis nun del vidpunto dil cienco mem plu granda importo. Nam del kozi, quin la homaro posedas komune, nula es tam vere universala ed internaciona kam la cienco. Or, omna komunikado e propagado dil cienco uzas la moyeno dil linguo, do la internacioneso dil cienco postulas nerezisteble la internacioneso dil linguo. Se ni konsideras, ke cadie sat multa ciencala verki, specale lernolibri, tradukesas aden dekedu o plu multa stranjera lingui, ni komprenas, qua enorma quanteso de laboro povus sparesar, se libri omnaloke sur la terglobo povus komprenesar tam generale, kam exemple muzikal noti o logaritmala tabeli.

This text fragment is a translation of the following English sample, taken from IALA's Comparative Texts (Comparative Studies, series A, part VI, division 1)11. The idea of a world literature, which Herder and Goethe conceived essentially from the point of view of art, has now gained even greater importance from the point of view of science. For, of the things that mankind possesses in common, nothing is so truly universal and international as science. Now all communication and propagation of science uses the means supplied by language, and so the internationality of science irresistibly demands the internationality of language. If we consider that today numerous scientific works, particularly textbooks, are translated into twelve or more foreign languages, then we understand what an immense quantity of labour could be saved, if everywhere on the globe books could be as generally understood as, for example, musical notes or tables of logarithms.

30

Chapter Two

Ido borrowed the part-of-speech tagging and the agglutinative strategy for derivation from Esperanto, even though the Ido derivation system is hard and complex compared to Esperanto's, as stated by the following example: martelo is 'hammer' in Esperanto and marteli is 'to hammer' as in English (or Italian: martell-o, martell-are). Conversely, Ido must verbalize the noun root with a bound morpheme: martel-ag-ar, 'hammer+NOUN-TOVERB+INF', as in French (marteau, marteler). Any slavic and mostly Germanic influences were cut off by Couturat.

Codification and the mélange/promotion dilemma What are the criteria in planning an interlanguage's form? As shown by the 'primitive contract rule', a language author should decide the language core (i.e. phonetics, morphology, syntax, lexicon) in the project phase, before publication. It is a general rule that decisions about planning are always taken by a single person, even in the case of Ido or IALA's Interlingua (see below), where committees were formed. It is quite obvious that the language's form will depend on the linguistic repertoire of its author. More precisely, the human languages known by the language author play the same role as substrata in the case of natural languages. These are called source languages of an interlanguage. What is the role of the mother tongues in planning an interlanguage? The comparison of major interlanguages, i.e. interlanguages that proved a language vitality and succeeded to establish a community of practice, showed that the more the language author plans next to the language core, the more important mother tongues become. This is particularly evident in phonetics. Bausani (1974; 1970) noticed that the phonetic space of an interlanguage cannot differ too much from the phonetic spaces of the mother tongues. It is worth noting that this phenomenon is valid for every planned language and it happens regardless of the purposes of the language itself, as it is mostly unconscious, even if the author is a linguistic professional. In this paper Bausani's observation is generalized and called the “Bausani effect”, as it is not limited to phonetics, but it is generally valid in other language layers, e.g. syntax and lexicon. As the effort in planning an interlanguage is a cross-cultural one, the language author faces a dilemma. In fact, he needs to choose the best structural features among his repertoire in order to plan an interlanguage which should be recognized as somehow familiar to the potential learners. The promotion of some features instead of others is deeply influenced by the Bausani effect. It is impossible to promote all features (it will coincide with an existing natural language) and at the same time it is impossible to

Multilingualism and Applied Comparative Linguistics

31

avoid a degree of mélange (see Latino Sine Flexione below). This rule is called the mélange/promotion dilemma, In the following case studies of Esperanto, Novial and Latino Sine Flexione, it is shown how the different language authors have tried to resolve this dilemma, and where the Bausani effect takes place.

The delicate equilibrium of Esperanto Zamenhof, the Esperanto author, tried to make sure that the linguistic mélange simplified the grammar of Esperanto, i.e. rules should be general and without exception. In this sense, Zamenhof chose a planning way similar to Schleyer's Volapük. Volapük has 28 phonemes, with eight vowels taken from German, and a phonematic writing system, i.e. every phoneme is written by one and only one grapheme, for simplicity's sake in learning. Esperanto has also 28 phonemes and a phonematic writing system, but only 5 vowels as in Croat or Czech. Both Volapük and Esperanto are clearly SVO languages. However, Volapük and Esperanto greatly differ. In particular, the Volapük morphology is very rich, e.g. every plural ends in -s, as in English, and there are four cases within a single declination, as in German, a proof of the Bausani effect. Moreover, a single verb root may have up to 505,440 different forms thanks to a lot of affixes. Esperanto, on the other hand, has an agglutinative morphology and almost no allomorphs, but the lexical assimilation rules do not hide their origins so strongly as in Schleyer's creation: a single verb root may have up to 6 basic forms (infinitive, present, past, future, conditional, imperative). Zamenhof made sure that some word roots would be recognized by speakers of Romance, Germanic and Slavic languages, so that nobody would perceive Esperanto as completely foreign at first sight. This is due to Zamenhof's perception of Esperanto as a kind of universal Yiddish: every speaker of a European language should find something familiar in it. That is why it has the consonants of Yiddish (Litvak, the socalled 'Lithuanian variety') and five vowels, as in mostly Sephardic languages. As he was a native speaker of Yiddish, the Bausani effect is clear12.

Novial and the failure of finding an Einbau Otto Jespersen became proficient in Ido and appreciated the work by De Wahl (1930). After careful studies, he proposed his own interlanguage project, called Novial (1928), an attempt to collect the best features of the

32

Chapter Two

major interlanguages of his time, i.e. Occidental, Ido and Esperanto. What follows is a translation in Novial of the previously mentioned English text fragment (cf. section ‘Ido, the Ausbausprache of Esperanto’): Li idee pri monde-literature, kel Herder e Goethe koncepted esentialim fro li vidpunctu del arte, ha nun ganat mem multim plu grand importanteso fro li vidpunctu del scientie. Den ek li coses kel li homaro posese comunim, nuli es tam verim general e international kam li scientie. Or omni comunico e mediatione del scientie usa li moyene del lingue, dunke li internationaleso del scientie demanda nonresistablim li internationaleso del lingue. Si nus considera ke disdi pluri sciential verkes, particularim lernolibres, es traductet en dekdu e plu multi stranjeri lingues, tand nus comprenda qui imensi quanteso de laboro povud bli sparat, si libres povud omnilok sur li globe bli comprendat tam generalim kam exemplim musical notes o tabeles de logaritmes.

Novial is more regular in structure than Occidental thanks to the affixes borrowed from Esperanto and Ido, e.g. bo- from Esperanto (bomatro, 'mother-in-law') and des- from Ido (desimportant, 'unimportant'). Novial is immediately readable by people educated in European-based cultures, in particular English. Jespersen greatly admired the English language and so he borrowed a lot of its structures. Nevertheless, Novial takes some features from German too, and from Scandinavian languages as well. Unfortunately, there are no studies of the Danish influences in Novial, Jespersen's mother tongue. After Jespersen's death (1943), no one used Novial until the web era. In fact, in 1998 a Novial reform was published, but without great success.

Latino Sine Flexione and the etymological fallacy Latino Sine Flexione was published in 1903 by Giuseppe Peano, a mathematician, who intended his interlanguage to be used mainly as the written medium of science. At that time, every man of culture knew classical Latin, so scholars should learn the interlanguage very quickly, only with a Latin vocabulary and some general rules of simplification. What follows is a translation of the text fragment (cf. previous section) in Latino Sine Flexione: Idea de literatura mundiale, que Herder et Goethe habe intellecto praecipue ex puncto de visu de arte, habe hodie acquisito, ex puncto de visu de scientia, sensu etiam majore. Nam, de commune possesiones de genere humano, nihil es tam generale et internationale quam scientia. Sed omne communicatione et propagatione de scientia ute auxilio de lingua, et ita

Multilingualism and Applied Comparative Linguistics

33

internationalitate de scientia postula in modo irresistibile internationalitate de lingua. Si nos considera, que hodie plure opere scientifico, in particulare tractatus, es translato in duodecim vel plus lingua extero, tunc nos cognosce quale immane mole de labore pote es praeservato, si libros, ubicumque in terra, pote es aequaliter intellecto in generale, sicut per exemplo notas musicale aut tabulas de logarithmos.

The derivation rules of nouns and adjectives take as the basic root the genitive forms, giving to the interlanguage an Italian flavour. Moreover, Peano introduced articles in the form illo, illa, which were absent in classical Latin (Bausani's effect). This was also noticed by Louis Couturat, who in a letter to his colleague pointed out the “idiotisme italien” into his interlanguage (Luciano and Roero 2005, 64). Without a defined grammar, users soon started introducing structures, usually from their mother tongues. But idiotisms were not the only problem of Latino Sine Flexione. Another great problem was the etymological fallacy, i.e. the misconception based on the idea that the actual meaning of the world is recognizable by its etymology. In fact, Latino Sine Flexione has a great degree of allomorphia due to Latin itself. This case study shows two general rules of interlanguage planning. First of all, the language author should not underestimate the role of grammar: if the limits of language variation are not clear before publication, every user will be free to insert structural features, e.g. tense markers, compound rules according to their mother tongues (Bausani's effect). Furthermore, Peano's Interlingua demonstrates that it is impossible to choose the right linguistic features from one source language only, in this case Latin. If so, the only viable choice is to try to do a language reclamation act as in the cases of Welsh or Hebrew. Theoretically Latin can be revitalized with its structural properties intact, even if some problems in such an effort will arise (see Gobbo 2005a).

Elaboration and the autonomy/integration dilemma The elaboration of an interlanguage is a very difficult task, as the language author faces a compelling dilemma: autonomy or integration? The historical developments of interlanguages show that there is a radical shift in interlanguage planning theory and practice: during the 19th century authors preferred autonomy, after the year 1905 they preferred integration. In fact, both Schleyer and Zamenhof tried to focus on linguistic productivity. Their interlanguages have phonematic writing systems, along Lautbild13, as the creation of new words should be simple and regular, and after a brief learning of some morphosyntactic rules and few basic roots a

34

Chapter Two

user should actually use the language. For instance, in Esperanto you only have one marker for quality, the suffix -eco, as in juneco, 'youth', and beleco, 'beauty'. Nevertheless, these compounds may appear strange for speakers of Western languages, and Esperanto was regularly accused of having “volapükisms” in its lexicon, as coinages are often unique, while later interlanguage preferred semantic or phonetic calques from source language. This led to a radical shift in interlanguage planning theory and practice during the 20th century (De Wahl 1930). In the 20th century, new interlanguages were formed only on the basis of Romance and Germanic languages, with a special attention to French and English, playing the role of superstrata. The focus changed from linguistic productivity to immediate readability by people educated on a Western-based culture and language basis, i.e. integration in a strictly Western sense. An immediate consequence of this change is the graphisation strategy. In fact, the writing strategies of new interlanguages are borrowed from English or French, which have a great respect for etymology, i.e. according to Schriftbild, but at the same time they lose phonematic correspondence. This is shown in Occidental and Interlingua.

Occidental between autonomy and integration In 1922, Edgard de Wahl published his own interlanguage called Occidental, a compromise between linguistic productivity and immediate readability. What follows is the translation of the English fragment in Occidental: Li idé pri mund-literature, quel Herder e Goethe hat conceptet esentialmen ex li vidpunctu del arte, ha nu ganat ancor mult plu grand importantie ex li vidpunctu del scientie. Nam de omni comun possedages del homanità niun es tam vermen general e international, quam scientie. Ma omni comunication e transmediation del scientie usa li medie del lingue. Do li internationalità del scientie ínresistibilmen postula li internationalità del lingue. Si noi considera, que hodie pluri sciential ovres, specialmen libres de aprension, trova se traductet in decidu o plu foren lingues, tande noi comprende quel immens quantità de labor on vell economisar, si on vell posser comprender libres partú sur li glob sam generalmen quam por exemples notes e tabelles de logaritmes.

Occidental is deeply influenced by French, in particular in the fields of graphisation and word order: li lingue international, un bon idé. Much of the work of de Wahl was about morphology, in particular derivation. He tried not to conceal the origins of stems as in Esperanto and Ido, keeping

Multilingualism and Applied Comparative Linguistics

35

at the same time a regular derivation. The result is less satisfying than the premises: not only the transformations from verb roots to nouns have six exceptions (e.g. mov-er, mot-, 'to move', 'motion') but also some derivation rules are accountable only in terms of French substratum (Bausani's effect). For instance from yun-al, 'young' you have yun-esse, 'youth' but from bell-al you have bell-ità and not *bell-esse. This is immediately reasonable for French people (jeunesse, beauté) but not for example for Italians (giovin-ezza, bell-ezza). Furthermore, in Occidental there are a lot of allomorphs, e.g. there are four allomorphs to express quality or abstract, e.g. liber-tà, propri-età, elegant-ie, polit-esse. For a comparison, in Esperanto and in Ido you have only one suffix: juneco, beleco and yuneso, beleso respectively. After de Wahl's death (1949) most supporters passed to IALA's Interlingua (interlanguage shift), while the rest changed the name into Interlingue. However, in the last years of the 20th century a small revival of Occidental/Interlingue started, thanks to the internet (Gobbo 2005b).

Interlingua and the quest of prototypical Interlingua was published by Gode in 1951. As we have seen in the case of Occidental, every interlanguage planned from Romance-base vocabulary can not avoid a certain degree of allomorphia, and Interlingua is no exception. As a specimen of Interlingua, the critique by Martinet published in Panorama (62, 1998) goes as follows: Io non voleva partir del latino ma trovar le parolas que le gente comprenderea al lectura directemente sin haber apprendite le lingua o forsan con un minimo de apprentissage. Un sorta de parve libretto esserea sufficente. Lo que on pote reprochar a Gode es haber automaticamente limitate le internationalitate al linguas romanic14.

Gode did not limit himself to Latin as Peano, but he considered some Romance languages (Italian, French, Spanish/Portuguese) and English as source languages, and German and Russian as control languages. Consider for instance the rule for assimilation: if a word root is graphically similar in four of the source languages, or in three source languages and in two control languages as well, it should enter Interlingua. If graphisation of a word differs too much among source languages, the Latin form is chosen. For instance, 'eye' in Interlingua is oculo, from Latin, after the following comparison in the source languages: in French it is oeil, in Italian occhio, in Spanish ojo, in Portoguese olho. Latin provides most words as well.

36

Chapter Two

Bausani's effect (Gode's L1 is American English) is evident in graphisation. For instance, let us compare the different graphisation strategies of English, Interlingua and Italian by some affixes derived from Greek very similar in orthography but different in meaning: phil-, from philein, 'to love', phyl- from phulon, 'tribe', and -phyllo, from phullon, 'leaf'. 'Philosophy' in Interlingua is philosophia, while 'phylogenesis' is phylogenesis, and 'chlorophyl' is chlorophylla. However, under the influence of Blair, Interlingua has an alternative alphabet which takes Italian as the model, and the results are quite similar, e.g. filosofia, filogenesis, clorofilla (Blair-Gode 1951). Nevertheless, this graphisation is hardly used by supporters, at least in their publications in the web. The influence of English also involves other levels. As a result, the meaning of a lot of words in Interlingua is the English one, and consequently there are a lot of so-called false friends with Romance languages. Finally, in respect of morphology and syntax, Interlingua is clearly a Western Romance language (e.g. it has the plural -s) although it borrowed some strategies from English and German too. Nevertheless, supporters prefer to present Interlingua in the official site as le latino moderne15.

Status and acquisition planning of interlanguages In general, the failure of planned interlanguages compared to the ethnic interlanguage of the present era (English as a global language) is due to quasi-extralinguistic factors, in particular in status and acquisition. Even though, in absolute terms, Esperanto is a linguistic minority (the most realistic estimate about the actual speakers of Esperanto is 50,000-100,000 people), Esperanto is clearly the winner of the “interlanguage war” that took place in the 20th century. This is verifiable by the frequency and popularity of international congresses, or by the number of Wikipedia pages written by interlanguage enthusiasts in their favourite planned language. The Esperanto community is at least 10 times bigger than Ido's and Interlingua's, the only planned interlanguages which have maintained some vitality since their publication until today. Moreover, estimates show that about 60% of interlinguistics literature is written in planned languages–among them, 95% in Esperanto (Blanke 2004). Finally, a key factor of the Esperanto success is the language policy of Zamenhof: absolute freedom of use and at the same time absolute closeness to structural reforms, at least since 1905. As it is possible to date precisely the moment where the language vitality starts in the case of interlanguages, they can be considered as a

Multilingualism and Applied Comparative Linguistics

37

laboratory, a test case of hypotheses about the evolution of natural languages, in particular about the relation between language and culture. Is it possible for a language to create its own culture ex nihilo? The history of interlanguages, and Esperanto in particular, shows that an interlanguage can create a culture, if language vitality and vigour can be achieved. Moreover, also the reverse is true: it is impossible for a living language not to have a culture even if it was planned with purely auxiliary purposes, i.e. without taking care of language identity aspects. Otherwise the Esperanto speech community could not have survived two world wars, in spite of the persecutions by Nazis and Stalinists.

A research agenda for empirical Esperantic studies Unfortunately, there are few empirically-based studies, except perhaps on the subfield of Esperanto as a family language16. For example, empirical observations show that overregularization phenomena are very rare in Esperanto, although not totally absent (Corsetti 2004). Even though Esperanto is attested as a first language in a family context, it is important to note that there is no monolingual speaker of more than three years old and Esperanto is never the strongest language in plurilingual families. This is due to the lack of a Sprachraum, i.e. a proper language space where it is also learned at school. Versteegh (1993) compares Esperanto to the situation of Hebrew at the end of the 19th century, while Lindstedt (2006) writes that they are both test cases of first language acquisition with restricted input, if we take Hebrew in a historical perspective, i.e. before the foundation of the first kindergarten. In reviewing the successful experiences in language revitalization, Tsunoda (2005) underlines that Modern Hebrew, Welsh and Strait Salish showed evident structural changes in phonology, grammar and lexicon while they were acquired as a first language by children. Nothing similar has ever happened to Esperanto or any other interlanguage.

Conclusions Two phenomena occurring in interlinguistics and generally valid also in language planning were pointed out in this paper and were verified following the historical developments of the discipline itself. Firstly, it was stated that planned languages are designed by a single person (even though a committee may be settled) and, after publication, they can not be planned again (see de Saussure's “primitive contract rule”).

38

Chapter Two

Secondly, it was shown how the linguistic repertoires act as substrata and how the author's mother tongue(s), in particular, deeply influence the structure of every layer of the interlanguage: from phonetics and graphisation up to word order and lexicon (see Bausani's effect). Thirdly, since the analytical models of language planning were applied here to interlanguages, it was demonstrated that such attempts as the Einbau efforts (i.e. merging existing interlanguages in a new one) are doomed, and what can happen is either a language shift (VolapükEsperanto, Occidental-Interlingua) or a speech community split through the launch of an Ausbau language (Esperanto-Ido). Lastly, it was made evident that interlanguages–in particular Esperanto–are means to cross-cultural communication as they proved that language and culture are intertwined. Therefore even auxiliary languages act as identity markers.

Notes 1. Blanke (1985; 2006) analysed this book, published in Leipzig, 1910 under the title Die Forderung des Tages. 2. Statuts, Art. 2: “La Société n'admet aucune communication concernant, soit l'origine du langage~ soit la création d'une langue universelle.” The reference was an a-priori system such as Solresol (see Yaguello 1991, 46–47). As work languages in MACL include French and German, I will not translate quotations in these languages. Vice versa for texts originally in interlanguages. 3. I wish to acknowledge Jeroen Darquennes who gave me very interesting feedback, and the anonymous reviewers who helped me to clarify my ideas and improve the organization of this paper. I should also thank Detlev Blanke for his bibliographic support and general advice. 4. Terms as 'artificial language' or lingua franca are intentionally avoided here, since they are used in too many different contexts, and they are often sources of confusion. 5. From the point of view of theoretical linguistics, planned languages are fully human languages, being non-natural without necessarily being unnatural, since they are acquired or acquirable as a normal part of the process of maturation and socialization (Lyons 2006). Consequently, they will be scrutinized by linguistic level, from language core to language use, that is to say phonetics, morphology, syntax, semantics, pragmatics. 6. The comparison between the interlanguages published during the first half of the 20th century and the new ones published thereafter via web sites showed that the most important and original solutions to interlanguage planning problems were proposed in the Golden Age period of interlanguages (Gobbo 2005a), that is from Schleyer's Volapük (1879) to IALA's Interlingua (1951). 7. This passage is excerpted from a critical edition of the course. The text was collected by Robert Godel using students' notebooks. Neither interlinguistics nor

Multilingualism and Applied Comparative Linguistics

39

language planning appreciated the explicit reference to Esperanto, even if René de Saussure, mathematician and Ferdinand's brother, was involved in the early Esperanto movement (Künzli 2001). 8. As already pointed out, once used, an interlanguage creates its own mistique of appartenance as any other human language. 9. Note that in the same year, in Paris, David Hilbert outlined the famous 23 mathematical problems at the International Congress of Mathematicians. 10. Since 1907, a speech community split occurred, instead of a language shift. A previous attempt to reform Esperanto occurred in 1894 when a ballot about reforming Esperanto was taken, based on a proposal by Zamenhof himself: 157 votes against reforms, 107 votes for (Dietterle 1983, 198). Many commentators argue that Zamenhof deliberately wanted reformists to lose. In a letter to his friend Javal written in 1906, Zamenhof wrote (my translation, from Esperanto): “Regarding your opinion about the so-called volapükisms I don’t agree with you at all. You know, that in 1894 I tried by myself to cut off every constructed word, but afterwards I convinced myself this would have been a big mistake... I ask to you never take suggestions from linguists, confront yourself only with people who have a feeling for philology and a lot of practice in Esperanto... who sincerely love the language and want to see in it something eternal, living, powerful (Kanzi 1980).” 11. I wish to thank James Chandler, an Ido expert, who included this text in his web site. 12. Zamenhof himself was an Ashkenazi influenced by the Haskalah (Jewish Enlightenment) and freemasonry, and Esperanto was the first step of a greater project of reform of religions on a common, neutral non-ethnic base, in order to unify humankind as a single family (Gishron 1986, Holzhaus 1969, Maimon 1978). In 1905 Zamenhof decided to let the language free for every use, i.e. its political and religious ideas did not influence directly Esperanto language vitality (Dietterle 1983). For a comparison, note that Schleyer said that Volapük was given to him by God, so it was untouchable. 13. Lautbild = phonematic writing, i.e. the principle is "one grapheme, one phoneme". In these languages you never have more-than-one grapheme for one phoneme in their writing system strategy. 14. My translation: “I did not want to start from Latin but to find words that people understand immediately without having learned the language or maybe with a bit of study. A kind of small handbook might be enough. What you might reproach to Gode is to have automatically limited internationality to Romance languages.” 15. See http://www.interlingua.com for updates. 16. For example there is only one corpus-based description of the language (Gledhill 2000), where there are some cues about evolution. For example, Zamenhof and the pioneers wrote lingvo internacia, 'international language', based on the French model, while in contemporary Esperanto people write internacia lingvo, based on the English one.

40

Chapter Two

References Primary Sources Bausani, Alessandro. 1970. Geheim- und Universalsprachen: Entwicklung und Typologie. Stuttgart: Verlag W. Kohlhammer GmbH. —. 1974. Le lingue inventate: linguaggi artificiali, linguaggi segreti, linguaggi universali. Roma: Astrolabio-Ubaldini. Blanke, Detlev. 1985. Internationale Plansprachen. Ein Einfürung. Berlin: Akademie Verlag. —. 2004. Interlinguistics and Esperanto Studies: Paths to the scholarly literature. Rotterdam: Universala Esperanto-Asocio. —. 2006. Interlinguistische Beiträge. Zum Wesen und zur Funktion internationaler Plansprachen. Frankfurt am Main: Peter Lang. De Wahl, Edgard. 1930. Evolution: Du principies, du directiones. Cosmoglotta. http://www.geocities.com/athens/parthenon/1222/occart.html Eco, Umberto. 1993. La ricerca della lingua perfetta nella cultura europea. Bari: Laterza. Gledhill, Christopher. 2000 Esperanto. A corpus based description. München: Lincom Europa. Gobbo, Federico. 2005a. The digital way to spread conlangs. Language @t work: 45–53. —. 2005b. The European Union's Need for an International Auxiliary Language. Journal of Universal Language 6: 1–28. Hornberger, Nancy H. 2006. Frameworks and Models in Language Policy and Planning. In An Introduction to Language Policy: Theory and Method, ed. Thomas Ricento, 24–41. London: Blackwell. Jespersen, O. 1931. Interlinguistics. International Communication. http://www.geocities.com/Athens/Forum/5037/IL.html Okrand, M. 1992. The Klingon Dictionary. New York: Pocket Books. Schubert, Klaus. 1989. Interlinguistics: Aspects of the Science of Planned Languages. Berlin: Mouton-de Gruyter. Tsunoda, Tanaku. 2005. Language Endangerment and Language Revitalization. Berlin: Mouton-de Gruyter.

Secondary Sources Albani, Paolo, and Berlinghiero Buonarroti. 1994. Ága Magéra Difúra. Dizionario delle lingua immaginarie. Bologna: Zanichelli.

Multilingualism and Applied Comparative Linguistics

41

Blair, Hugh E., and Alexander Gode. 1951. Interlingua-English, a Dictionary of the International Language. New York: F. Ungar P. Co. Corsetti, Renato. 2004 Regularizing the regular: The phenomenon of overregularization in Esperanto-speaking children. Language Problems and Language Planning 28: 261–282. Dietterle, Joh., ed. Ludwig Lejzer Zamenhof. Originala verkaro: antaǎparoloj, gazetartikoloj, traktaƵoj, paroladoj, leteroj, poemoj. Osaka: Oriental-libro, 1983. Forster, Peter G. 1982. The Esperanto Movement. Berlin: Mouton-de Gruyter. Gishron, Jeremi. 1986. Lingvo kaj Religio. Studo pri la frua esperantismo kun speciala atento al L.L. Zamenhof. Jerusalem: Sivron. Gobbo, Frederico. 2006. L'esperanto e la traduzione automatica: Storia, risultati e prospettive esperantologiche dell'approccio statistico. In Loquentes linguis: Linguistic and Oriental Studies in Honour of Fabrizio A. Pennacchietti, ed. Mauro Tosco, Pier G. Borbone and Alessandro Mengozzi, 319-327. Wiesbaden: Harassowitz Verlag. Holzhaus, Adolf. 1969. Doktoro kaj lingvo Esperanto. Helsinki: Fondumo Esperanto. Kanzi, Itô. 1980. Klopodis organizi, sed vane! Plena Verkaro Zamenhofa, kajero 7. Kioto: Ludovikito. Klocko, Edouard. 2002. Lingue degli Hobbit ,dei Nani, degli Orchi: Dizionari, grammatiche, storie. Rome: Tre Editori. Künzli, Andy. 2001. René de Saussure (1868-1943). Tragika sed grava esperantologo kaj interlingvisto el Svislando. http://esperantic.org/librejo/dbstudoj/35_Kuenzli.htm Large, Andrew. 1985. The Artificial Language Movement. Oxford: Blackwell. Lindstedt, Jouko. 2006. Native Esperanto as a Test Case for Natural Language, SKY Journal of Linguistics 19: 47–55. Luciano, Erika, and Claudia Silvia Roero. 2005. Giuseppe Peano – Louis Couturat. Carteggio (1896-1914). Firenze: Leo. S. Olschki. Lyons, John. 2006. Natural language and universal gammar. Essays in linguistic theory. Volume 1. Cambridge: Cambridge University Press. Maimon, N.Z. 1978. La kadžita vivo de Zamenhof – Originalaj studoj. Tokio: Japana Esperanto-Instituto. Ostwald, Wilhelm. 1910. Die internationale Hilfssprache. In Die Forderung des Tages, ed. Ostwald Wilhelm, 437-512. Leipzig: Akademische Verlagsgesellschaft. Pennacchietti, Fabrizio A. 1987. L'internazionalità dell'esperanto e il carattere degli elementi indoeuropei in esso. In La comunicazione

42

Chapter Two

internazionale tra politica e glottodidattica, ed. Chiti-Batelli, Andrea. Milan: Marzorati. Simone, Raffaele, ed. 1970. Ferdinand De Saussure. Introduzione al secondo corso di linguistica generale (1908 – 1909). Rome: Astrolabio. Tauli, Valter. 1968. Introduction to a theory of language planning. Almqvist, and Wiksells: Uppsala. Versteegh, Kees. 1993. Esperanto as a first language: language acquisition with a restricted input. Linguistics 31: 539–555. Yaguello, Marina. 1991. Lunatic lovers of languages. London: The Athlone Press.

PART TWO BILINGUALISM, TRILINGUALISM AND LANGUAGE PROFICIENCY

CHAPTER THREE A SOCIOCULTURAL PERSPECTIVE ON THE LANGUAGE PRACTICES OF TRILINGUAL FAMILIES IN ENGLAND AND GERMANY ANDREAS BRAUN

Introduction This study aims to identify the sociocultural factors that may influence parents’ language practices with their children in potentially trilingual families living in England and Germany. By drawing on theoretical concepts in multilingualism (e.g. Herdina and Jessner 2002) and multiculturalism (e.g. Edwards 1994) this study endeavours to explore the cultural background of parents in trilingual families in the way in which they identify themselves with the cultural traditions and values that are attached to their native languages. This may add to the understanding of the reasons why some parents in trilingual families use all of their available native language with their children, while others do not. Although past research has documented a link between culture and language (e.g. Birken-Silverman 2004; Olshtain and Nissim-Amitai 2004), it is unclear whether the language practices of trilingual families are affected by their cultural background or vice versa. First, this study describes and categorises trilingual families according to their language background, i.e. the native languages each parent spoke (Table 3-1). This, it is believed, will provide the foundation for the second part of this paper, aiming at an investigation of parents’ commitment whether or not to pass on their native cultural traditions to their children that were attached to their native languages. In order to meet the objectives of this study, parents of 70 trilingual families living in England and Germany were interviewed. The participants were encouraged to comment on their own language and cultural practices with their children at home

Multilingualism and Applied Comparative Linguistics

45

and in the wider community leading to the following two research questions: -

In what ways do parents in trilingual families define their cultural background and how may this affect their language practices with their children? Are parents’ language background variations reflected in different cultural practices in their families?

Brief Literature Review Terminological aspects of trilingualism Trilingualism is a relatively new academic research field, related to disciplines such as sociolinguistics, second and third language acquisition and cross-cultural communication. Ó Laoire (2006, 1) states: Although there has been a growing corpus of research recently on third language acquisition, we are still far from understanding all the details of the process of acquisition. Many factors interact in third language acquisition, e.g. attitude, motivation, typological factors, L2 factors, and earlier learning experience.

Other researchers (e.g. Barron-Hauwaert 2003; Etxeberria 2004) criticize that trilingual studies have often been carried out within the framework of bilingualism. Besides the similarities, however, there are also a number of differences between the sociocultural and linguistic dynamics in trilingual families and those in bilingual families (Herdina and Jessner 2002; Hoffmann and Ytsma 2004). For example, Herdina and Jessner (2002) have proposed a dynamic model of multilingualism (DMM), stating that the acquisition of more than two languages leads to the development of new cognitive and metalinguistic skills. In support of the DMM, Lasagabaster (1998) found that trilinguals who were highly competent in all three languages obtained better results in a metalinguistic awareness test than those whose competence across languages was uneven. Similarly, Aronin and Ó Laoire (2004) argue that trilingualism as opposed to bilingualism is more complex, with qualitative rather than quantitative differences as suggested in some case studies (Barnes 2005; Tokuhama-Espinosa 2000 and 2003). Little research has been done on the sociocultural effects of using three or more languages within the family

46

Chapter Three

and in the community in a European context as compared to using one or two languages. Many bilingual theories simply cannot be transferred to trilinguals. Trilingualism is unusual because the three languages (or cultures) cannot be ‘balanced’ or equal, as in bilinguals. One (or two) languages are always at risk of becoming underused or passive (Barron-Hauwaert 2003, 131).

Languages in a multicultural context Oksaar describes multiculturalism as “the ways of a people”. “Multiculturalism of a person is realized in his ability to act here and now according to the requirements and rules of the cultures” (Oksaar 1983, 20). Edwards (1994, 176) adds: “The existence of many languages in the world implies the existence of many cultures”. Two or three languages usually fulfil different roles depending on sociocultural factors. The mother tongue belongs to the personal and individual environment, whereas the language of the wider community is part of the official and sociocultural one. Language is a component of culture along with other entities like, for example, values, beliefs and norms; language is a product of culture, transmitted from one generation to the next in the socialisation process; it also moulds culture, that is to say, our cultural representations are shaped by languages (Hamers and Blanc 2003, 199).

Moreover, Tarmo (2006) suggests that effective communication in two or more languages can only proceed without hindrance when the language user is aware of the cultural implications. For instance, when a multilingual person talks to somebody in Language B but applies cultural values from Language A or C then miscommunication or even communication breakdown can occur as the speaker may unintentionally appear strange, ignorant, arrogant or even rude because cultural conventions and linguistic items are not observed. Finally, Clyne (1982) uses the term “Language Ecology” to assess the factors and institutions promoting the maintenance of the home languages and cultures in a multilingual and multicultural setting. In this line, Cummins (2003) and Maneva (2004) refer to the importance of the use of languages in educational institutions (formal and informal), influencing the way in which the identity and language development of multilingual children proceeds. This depends on the learning environment either in school, community or at home in relation to the values and cultural beliefs practiced in each setting. Thus, it is anticipated that the sociocultural and

Multilingualism and Applied Comparative Linguistics

47

linguistic input from both the community and the parents could play a significant role in the language practices of trilingual families, and interact in that process in complex ways.

Methodology Semi-structured interviews This survey was conducted in England and Germany, two predominantly monolingual European countries, which represent different geographical and sociolinguistic settings. England is part of Great Britain, a former colonial power with English being the official language of many domains worldwide, such as media, politics, science and business (Cenoz and Jessner 2000). Germany is situated in the heart of Europe, yet the German language only has a limited international status (Ammon 2003). It was anticipated that these differences would have implications for the decisions parents in trilingual families make about which languages to use with their children. The survey was carried out using semi-structured interviews, linked to the expectation that the participants’ viewpoints were more likely to be expressed freely in an interview situation that had openended questions rather than in a questionnaire or a fully structured interview (Flick 1998). All in all, 70 interviews were conducted and transcribed, 35 in England and 35 in Germany. Each interview lasted between 40 and 60 minutes. The interview schedule was separated into three main sections: a) general background, such as age, number of children and parents’ profession, b) parents’ cultural background, linguistic proficiency and communication strategies and c) parents’ reflections upon their children’s language command and preferences. Children’s language proficiency was not measured as the emphasis was placed on parents and their decision whether or not to use their native languages with their children. The parents were also asked if they wanted to pass on their cultural heritage to their children and whether they thought there was a link between language and culture in their trilingual families.

Terminology Where possible, the data analysis in this study is described with everyday terms to capture the function being assigned to each trilingual family. However, in the course of this report, certain terms and abbreviations are used, such as “Community Language” (CL), describing the language that is spoken in the wider community and neighbourhood

48

Chapter Three

where the trilingual families lived. In addition, “Native Language” (NL) refers to one, two or three of the parents’ languages acquired natively at a young age (Cenoz and Jessner 2000). It can also include the CL, depending on the families’ types of language background (Table 3-1). Lastly, the term “Home Language” (HL) refers to both parents’ NLs as a unit, which may include the CL if one or both parents spoke it natively.

Participants For the purpose of this survey, only parents were interviewed as informants about their own and their children’s language practices, because they were in the position to decide whether or not to provide a context for their children to develop three languages. In some trilingual families, however, older children answered questions as well if their parents wished so. The participating parents had a language situation that could provide their children with two native home languages in addition to the community language, where between community and home three languages were available to use. The survey had a preponderance of participants with high socio-economic status. As a requirement of this survey, children were aged 15 years or younger. The volunteering parents in England were represented by 28 nationalities who spoke 25 different native languages. The largest number encountered had British nationality, held by 24 parents usually in combination with an additional nationality. Other large groups were Finnish (13), German (5) and Russian (5). In the 35 interviews conducted in Germany, the participants shared 32 nationalities, speaking 28 different native languages, with German nationality as the largest group (16), followed by Italian (8) and USAmerican (7). The volunteers were recruited through a specially designed webpage (21) (Braun 2003), international schools (19), personal contacts (12) and the snowballing technique (18), which is based on community networks where a number of individuals are located and then asked for contact details of other people who meet the research requirements (cf. Flick 1998).

Ethics This study strictly followed the Code of Conduct and Ethical Guidelines of the British Psychological Society (BPS 2000). All parents providing information were guaranteed anonymity, confidentiality and that participation was voluntary. Interviewees’ names have been disguised in the report of this study in order to assure complete anonymity. Participants

Multilingualism and Applied Comparative Linguistics

49

were also informed about the aspects of the research investigation that might have influenced their willingness to participate, e.g. length and scope of interviews. Although this was a survey of parents, a few of them wanted their children to answer some questions, which was done only in the presence of at least one parent.

Discussion of the Results Typology of trilingual families The first objective of this paper is to categorize trilingual families according to their types of language background, living in two officially monolingual countries, England and Germany. The typology of this survey, which refers to different language constellations, is an attempt to predict parents’ language choices with their children. In the course of this study, three groups of trilingual families were identified. As shown in Table 3-1, the majority of participants were categorised in Type I and Type II, while Type III was only represented by a relatively small number of cases. Type I describes trilingual families where each parent spoke one different NL. Neither of the parents spoke the CL as a NL. For instance, in one family the father spoke German as his NL, while the mother spoke Finnish. Both parents moved to England as adults, which created a third language opportunity for their children. 13 families living in England and 11 in Germany were categorised for Type I (Table 3-1). Table 3-1. Typology of trilingual families in this survey Type Description ENG GER Total I The parents each 13 11 24 spoke one different NL. (NL ≠ CL) II One or both 13 18 31 parents spoke two NLs. (NL can be CL) III ‘Others’ – 9 6 15 Amalgamation of different backgrounds. (NL can be CL) Total 35 35 70 Abbr. CL – Community Language / NL – Native Language

50

Chapter Three

As outlined in Table 3-1, Type II refers to families where one or both of the parents spoke two NLs. This can include families, where one parent spoke one NL and the other one two NLs. One of them spoke the CL as a NL. This is shown, for instance, in a family living in Germany where the mother spoke Italian and English natively, while the father spoke the CL, German, as a NL. Type II may also refer to families where both parents were bilingual with the same linguistic background. In this case neither of them spoke the CL natively. For example, in one family both parents grew up with two NLs, Russian and Mongolian. Once they had moved to Germany, a third language was available to them and their children. Both parents were already adults when they moved to their host country, where they learnt the CL as a non-NL. In England, 13 families were classified as Type II and in Germany 18 families, which makes it the highest number of cases encountered in this study (Table 3-1). Finally, Type III is an amalgamation of families with different backgrounds and language combinations, which were represented by the smallest number of cases in this study (Table 3-1). Some of these families were difficult to define in terms of their linguistic and cultural backgrounds. Type III may refer to a family consisting of a trilingual parent who shared the CL as a NL with her/his monolingual partner, as shown in the example of the following family. The mother spoke Finnish, Swedish and English as NLs, while her husband spoke only English. Type III may also refer to any other parental language constellation that did not fit within the description of this typology. This group is listed under “Others” (Table 3-1), including families with the potential of more than three NLs or where English was employed as a third and non-native language. All in all, there were nine Type III families in England and six in Germany (Table 3-1). Through thematic analysis (Flick 1998) of the interview transcripts, it was discovered that Type I parents in England and Germany were more likely to employ all their NLs with their children than parents with other types of language background (Table 3-2). For instance, 9/11 couples with a Type I background in Germany and 10/13 couples in England raised their children trilingually, by using all their NLs with their children on a regular basis (priority was given to speaking and listening skills). In contrast, the majority of Type II families in England (10/13) and Germany (10/18) raised their children bilingually, as they discontinued one of their NLs. And finally, most Type III families in England (8/9) only used the CL, English, with their children, while in Germany 5/6 Type III families employed English as a third and non-NL (Table 3-2).

Multilingualism and Applied Comparative Linguistics

51

Table 3-2. Parents’ language use with their children (by Type) Description Type I Type II Type III ENG GER ENG GER ENG GER All NLs used regularly 0 9 0 2 0 1 (triling.) Trilingualism with 2 2 3 4 0 0 difficulties One NL discontinued 1 0 10 10 1 0 (biling.) Only CL used (monoling.) 0 0 0 1 8 0 English used as 3rd and 0 0 0 1 0 5 non-NL Total 13 11 13 18 9 6 Abbreviations CL-Community Language / NL-Native Language

Language as a cultural marker? (Type I Families) This section examines Type I families’ use of three possible languages in relation to the parents’ cultural background and whether they wished to pass on their cultural heritage to their children. By looking at all trilingual families in this study, it was found that each of the Type I parents in England and Germany only spoke one NL, which most of them associated with their respective cultural heritage (Table 3-3). This group of parents grew up monolingually, in societies which usually had one main language, embedded in cultural traditions and historical and national values. Once the parents had moved to England or Germany, a new culture and language were at their disposal. A Greek mother living in England commented: Mother: I think it [language use1] was not a choice, it just happened like this. It happened that I am Greek and he [husband] is French, so we had the two languages. We take it as a matter of fact… it is our situation. I don't give up my language.

In previous studies, Olshtain and Nissim-Amitai (2004) argue that language is a symbol of a person’s cultural and/or ethnic background, which plays a vital role in intergroup relations when cultures and languages are in contact. Hamers and Blanc (2003) support a less firm position with the observation that language only plays a key role when it is a core component of a culture as there are other markers, such as religion, ethnicity or cultural traditions. The data analysis indicates that most Type I parents wanted to pass on their cultural values to their children as shown

52

Chapter Three

in Table 3-3. In England, 11/13 mothers and 9/13 fathers regarded it as important that their children would learn some of their cultural values or that they would feel in some way related to their cultural background. A French mother commented: “It’s important that I pass on my culture to the children, the French culture”. There was only one mother who did not see it as important that her children would relate themselves to her cultural heritage. Four fathers and one mother did not comment on this issue, either they were not available for the interview or they simply agreed mutually with their partners, without stating an explicit view (Table 3-3). In Germany, 6/11 mothers and 6/11 fathers wished to pass on their cultural traditions to their children. Only two mothers and one father did not see it as a necessity to teach any cultural values to their children: “For me it’s not so important because I don’t personally have a cultural belonging”. Three mothers and four fathers could not comment, mostly because they were unavailable for the interview (Table 3-3). Table 3-3. Type I parents’ commitment to pass on their cultural values England Germany Mother Father Mother Father Important 11 9 6 6 Unimportant 1 0 2 1 Unavailable 1 4 3 4 Total 13 13 11 11

The findings summarised in Table 3-3 show that most Type I parents in England and Germany regarded it as important to pass on their native cultural values to their children. When Type I parents were describing their cultural values, they generally referred to the following markers: language, family values, food, religious ceremonies or clothing. The fact that so many Type I parents wished to pass on their cultural heritage to their children, might be related to their decision to use their HLs (Table 32). A Romanian father explained: “Of course I would like her [daughter] to learn about it [cultural values] … I think language is half of the culture”. Another example is given by a US-American mother married to her Italian husband: Mother: For me speaking English keeps alive my Manhattan, the beaches of Florida, the Grand Canyon, Hollywood, the great vastness of America…For my husband speaking his language goes together with having a long delicious meal of pasta, speaking of politics, his Siena and Florence, corruption, chaos, artistic excellence, opera, the pope … Language and cultural identity for us just belong together.

Multilingualism and Applied Comparative Linguistics

53

Further quotations show that ten Type I parents in England and seven in Germany reported a direct link in their thinking between cultural traditions and language in their families. They wanted their children to learn about their cultural values and customs, which seemed to them to include the home languages as well. Besides their cultural background, Type I parents also commented on their high competence in their NLs, which had a positive effect on their trilingual language choice for their children. At the same time, 6/13 Type I couples in England did not use the CL with their children because it was not their “mother tongue”. In addition, 7/11 Type I couples in Germany referred to their limited proficiency in the CL, as a factor in their decision not to use it with their children, as pointed out by a Portuguese mother: “We can’t speak perfect German, which would not be nice if the children learnt broken German from us”. Therefore, the findings suggest that the language practices of most Type I families were partly influenced by their cultural background in conjunction with a variety of sociolinguistic factors, such as language competence or the “mother tongue effect”.

Cultural practices and language use (Type II Families) In contrast, the description of Type II families’ cultural backgrounds seems to be slightly more complex than the portrayal of Type I families, because one or both parents in each family spoke two NLs. Recent studies suggest (e.g. Aronin and Ó Laoire 2004; Barron-Hauwaert 2003) that the cultural context or sociolinguistic environment plays an important role in determining multilinguality. This section looks at the sociocultural context of trilingual Type II families and whether it had implications for their language use with their children. The findings indicate that most mothers in Type II families in England (10/13) and Germany (11/18) regarded it as important to pass on parts of their cultural values and traditions to their children. There were also 8/13 fathers in England and 10/18 in Germany who wanted their children to learn about their cultural values and customs (Table 3-4). Most of these parents gave general statements about their cultural customs, such as “food, discipline, eating together in the evening” as well as “personal values”. The following table (Table 3-4) provides a statistical summary of Type II parents’ comments on their commitment whether or not to pass on their cultural traditions and values to their children.

54

Chapter Three

Table 3-4. Type II parents’ commitment to pass on their cultural values England Germany Mother Father Mother Father Important 10 8 11 10 Unclear 2 2 3 2 Unavailable 1 3 4 6 Total 13 13 18 18

The figure above indicates that most Type II parents in England and Germany wanted to pass on their cultural values and customs to their children. However, not all of the parents commented on cultural issues, in particular the fathers in Germany. Either it was unclear or they were absent at the time of interviewing. Amongst the parents who did comment were two bilingual fathers and four mothers in England as well as three bilingual fathers and five mothers in Germany who specifically referred to more than one cultural tradition when they described their cultural background. These 14 parents also wanted to pass on their multiple cultural traditions to their children, even if most of these parents dropped one of the associated NLs (Table 3-2). A Spanish/French mother commented: “So, I want to pass to him [son] the French and Spanish [cultural traditions] … But, ah, at home we don’t speak French”. It could be argued, therefore, that these 14 Type II parents’ cultural backgrounds did not necessarily correspond to their NL use, because they may have been committed to passing on two or three cultural traditions to their children, but not necessarily all the associated NLs. Another example is given by a bilingual US-English/Italian mother, who spoke only American English with her children, while Italian, her second NL was described as “not useful”. At the same time, she said: “I’ve always considered myself as Italian. I perceive myself as Italian”. When the mother was asked if she wanted to pass on her cultural values, she replied: Mother: The children should recognise that papa is German and they are German and mum is Italian American. I don’t consider myself 100% American because I really grew up with these two cultures as very important in our family … I am not saying they [children] have to wear their German, American [and] Italian culture on their sleeves but they are aware—I don’t think we make a conscious effort and yet they [children] are very much aware that they do come from these three different backgrounds.

Here, the mother regarded her two cultural traditions as “very important” in her family, yet she only used English with her children. In addition, in three cases the history of the country where the participants

Multilingualism and Applied Comparative Linguistics

55

came from was bilingual such as in Burkina Faso. The father spoke French and Mòoré as NLs. Although the father did not identify with the French cultural tradition, he only used French with his child. Father: I feel as an African from Burkina Faso. It was natural and I feel good when I speak French. It is better for the future when you speak French or English and when the child speaks French she can integrate more easily in Africa ...

This would afford some further support for the assumption that some Type II parents’ available NLs were not necessarily used in conjunction with their cultural traditions, but instead for practical reasons, language competence and language status, as pointed out by other researchers (e.g. Olshtain and Nissim-Amitai 2004). What is more, Skutnabb-Kangas (1984) asserts that bilingual speakers identify themselves as bilingual with two languages and/or two cultures if exposed to both cultures and languages from the beginning of childhood. This may be equally true for Type II parents in trilingual families. For example, five Type II parents in England and Germany found it challenging to describe their cultural background. A Spanish/British father commented: “I’m more European than anything else…But what do I feel … deep down I feel more Spanish than English”. However, he needed to be encouraged by his wife to use Spanish as he preferred speaking the CL, English, with his children. This applies to the other four Type II parents as well, such as the Croatian/German father who mainly used the CL, German, with his child, despite his comments: “I would never regard myself as German”. At the same time this father said: “I don’t like speaking Croatian, I must say I can express myself much better in German”. As argued in the previous paragraph, language status and language competence seemed to be core reasons for the father’s use of the CL rather than his cultural background. In fact, when bilingual Type II parents were asked to describe their language and cultural practices, most of them in England (92%) and Germany (86%) did not link their language use with their different cultural traditions. Even if so many Type II families wished to pass on parts of their cultural traditions and values to their children (Table 3-4), they did not usually include all the associated languages, as the majority of bilingual parents actually dropped one of their available NLs (Table 3-2). Generally, the informants commented that the beginning of school or nursery made it harder for many of them to maintain the home languages, as pointed out by a bilingual Greek/British father in England: “When they [children] started going to school, especially the older one, I [father] started speaking a lot more English to him [son]”. In a different family in

56

Chapter Three

Germany, a Cypriot/S-African mother explained her feelings on the point: “ID [daughter] used to speak all three languages pretty okay before she started kindergarten. After that, life is not the same”. The beginning of nursery and school marked an important milestone in these families’ language practices in the way in which the parents had much less time with their children, who were overwhelmed by new vocabulary from different language domains taught in schools (c.f. Cummins 2003; Maneva 2004). Therefore, language became a stronger issue in terms of educational choices because the HL input was less than the CL input. This imbalance led many couples in Type II families in England and Germany either to rethink their HL strategies or to discontinue one of their NLs (Table 3-2). Therefore, it could be argued in the light of the presented data analysis that the language choices of most bilingual Type II parents were not significantly influenced by their commitment to pass on their cultural values to their children. Instead, it emerged that Type II parents’ language command, language status and the beginning of school may have been of similar if not greater significance in determining their distinctive language practices with their children.

Acculturation and the use of English (Type III Families) As outlined in Table 3-2, families with a Type III background in England were much more likely to use only the CL with their children (8/9) than they were in Germany (0/6). The following discussion delineates Type III parents’ specific language choices in relation to their cultural background. Aronin and Ó Laoire (2004) assert that the cultural context or sociolinguistic environment plays a distinctive role in specifications and structure of multilinguality. As explained earlier, the cultural background of parents in Type I families were more influential on their language choices with their children than in Type II families. This chapter looks at the cultural backgrounds of Type III parents as a possible determiner for their different language practices between families living in England and their German counterparts. Table 3-5 summarises Type III parents’ comments on their commitment whether or not to pass on their cultural values to their children.

Multilingualism and Applied Comparative Linguistics

57

Table 3-5. Type III parents’ commitment to pass on their cultural values England Germany Mother Father Mother Father Important 5 4 4 4 Unclear 2 2 0 0 Unavailable 2 3 2 2 Total 9 9 6 6

Table 3-5 indicates that in England 5/9 mothers and 4/9 fathers regarded it as important to pass on their cultural values and traditions to their children. In general, these values incorporated certain food, religious ceremonies, clothing or respecting older people. However, it seems that language was not included as a cultural marker because the parents in eight out of nine Type III families raised their children monolingually by dropping the associated NLs (Table 3-2). For example, in one family the parents spoke three NLs between them, Finnish (mother) and Punjabi and Urdu (father) as well as the CL English. The father regarded himself as “Asian, Asian Christian”, while the mother rejected the English cultural tradition quite strongly. “I have never considered myself English and I never will be”. Despite the rejection of English cultural values, the parents in this family only spoke English with their children. The description of the cultural background of the next family was also complex. When the father was asked about his family’s cultural background, he replied: “Chinese and [the] same for my wife”. In terms of their linguistic background, however, the father referred to four languages: “We are fluent in English, French, mother tongue is Pidgin French [Mauritian] and as having a Chinese background we also speak Chinese”. According to the father, the family were practicing Chinese and Mauritian customs and traditions, yet they only used English with their children. Similarly, the father in another family described his cultural background “as a Chinese” and “Buddhist”, yet he grew up in Malaysia: “You have to learn Malaysian culture. It’s a very cosmopolitan culture. It’s a mix, because in Malaysia you have over 55% Malay and 30% Chinese about”. The father spoke a number of languages and dialects, though the distinction was intricate as the following quotation shows: Father: My mother tongue is Chinese. In China there are different dialects … so I speak Cantonese. The official language in China is Mandarin but then you come from different provinces, they have their own dialect … I think being more from China and Malaysia; we have to learn to speak Malay … so English is a [school] subject.

58

Chapter Three

Despite his rich linguistic and cultural background, the father only spoke English with his children as did most of the other Type III parents in England (Table 3-2). It was found that the parents who spoke English as a non-native language with their children had been living in England for over 20 years. This could imply an “intercultural migration process” (Horenczyk 2000, 14), in which some bilingual or trilingual Type III parents may have developed a cultural background that incorporated two or more cultural traditions, including English cultural values. A FinnishSwedish-English speaking mother commented: “I see myself as European. Somebody specifies it then I would say I am half and half, English and Finnish”. A British/Pakistani father adds: “I think I am a mixture of Asian and British”. Although the parents referred to themselves as having more than one cultural tradition, they only used one language, English, with their children. Four parents also referred to the English cultural tradition when they described their children’s cultural background. “They [children] see themselves as English more than we do. They were born here [England] and they think of themselves as English”. Generally, these parents believed that it was not convenient or that their children would get confused if they had to learn multiple languages. A Chinese father stated: Father: We thought we can always ask them [children] to learn [a second language] when they have an established language first so they can learn better after that … So they would not get so confused.

Turning to Type III parents in Germany, their language practices were distinctively different from the English sample, in the way in which all six families raised their children trilingually. However, five families employed English as a third non-native language (Table 3-2) through multilingual programmes, partly because English was regarded as an important language for them. In terms of their cultural background, Table 3-5 shows that 4/6 mothers and 4/6 fathers wanted to pass on parts of their cultural values to their children. Two non-native English speaking parents even preferred the English language and the associated cultural values to one of their own native traditions and NLs, as pointed out by a Czech/Dutch father: Father: I would say that we probably identify more with that [English], so probably English first, Dutch second [and] German third … I consciously made the choice which would be the most useful for them [children] in the future. Czech is not spoken by many people.

Multilingualism and Applied Comparative Linguistics

59

Another example is given by a French/Italian/Moroccan mother: “We want our children to learn the best of all three [cultural traditions] and more”. In terms of languages, however, the mother dropped Italian and Arabic by using only French with her children plus English and German as non-native languages through bilingual schools. The parents in these two families wanted to pass on three languages and three or more cultural traditions to their children, including the English cultural values and English language. The other Type III parents only wanted their children to learn English as a world language without any reference to English culture. Besides the status of the English language, Cenoz and Jessner (2000) point out that the context of multilingual acquisition plays a similar role in a polyglot’s language choices. For example, all six Type III families in Germany had different trilingual situations, though they needed to be categorised in one group because of their relatively low number of individual cases (Table 3-1). Despite the differences in their cultural and linguistic backgrounds, 8/12 Type III parents in Germany were committed to passing on their home cultural values to their children (Table 3-5), but not necessarily all associated native languages (Table 3-2). The parents in two families even dropped two home languages, yet some of the associated cultural customs were practiced in the families as pointed out by a French-Italian-Arabic mother: “I have three nationalities … As a result the cultural habits of the family are very mixed”. In addition, the majority of Type III parents in England (83%) and in Germany (83%) did not link their cultural practices with their language choices. Therefore, it could be concluded that the cultural background of Type III parents in England and Germany only fractionally influenced their language choices with their children. Of more importance seemed to be the status of the English language and the families’ variations in their language background.

Conclusions To sum up, the quantitative data on whether parents thought it was important to pass on their cultural heritage to their children (Table 3-3, 3-4 and 3-5) did not differ significantly across the three family groups. Attention was paid to parents’ different cultural backgrounds in the way in which they affected their language use with their children. With reference to the reported comments made by the informants, it could be concluded that the majority of parents across all types of language background were committed to passing on certain parts of their cultural traditions and values

60

Chapter Three

to their children. It seemed, however, that Type I parents reported a more confident picture of their own cultural background, which most of them associated with their NLs, used with their children. This group of parents grew up in societies which usually had one dominant language and cultural tradition, as explained by a French father: “Because our idea is [that] language is a culture, so we don't want [our] children to lose their part for themselves and for ourselves”. Most of these parents wanted their children to learn about their cultural values and customs, which also included the language as a cultural marker (Table 3-3). A Spanish mother commented: “Well, I think if you take out the language out of the country then it is a bit difficult to bring the culture with it”. Therefore, Type I parents in England and Germany tended to be more persistent in using their available native languages with their children than parents with other types of language background (Table 3-2). These parents consistently referred to their high competence in their NLs and to their cultural heritage, as important factors to use their NLs with their children. In contrast, the cultural description of Type II and Type III families proved convoluted, because one or both parents in each family generally referred to two or three cultural traditions. A Croatian/German father elaborated on this issue: “I wouldn’t see myself as German or Croatian, but rather as Croatian-German, a mix … I am between two worlds”. Although most Type II and Type III parents in England and Germany wanted to pass on certain parts of their native cultural values to their children (Table 3-4 and 3-5), they did not usually include the associated NLs (Table 3-2). Bilingual parents with a Type II background often dropped one of their NLs once the children became older and started nursery or school. Furthermore, most parents with a Type III background in England only used English with their children, because they thought it was more convenient or they did not want to confuse their children. In Germany, Type III parents employed English as a third and non-native language through multilingual schooling. Therefore, Type III was the only group of trilingual families in this study that differed quite considerably in their language practices between England and Germany (Table 3-2). However, it seemed that the difference was related to their specific language constellations rather than to the situation in the two countries. Finally, the majority of bilingual and trilingual parents did not explicitly link their different cultural traditions with their language use. This may lead to the conclusion that the cultural background of Type II and Type III families were less influential on their language practices with their children than Type I families, though in conjunction with other

Multilingualism and Applied Comparative Linguistics

61

factors, such as language proficiency, schooling, language status and the importance of English as a world language.

Notes 1. Anaphoric reference to clarify a pronoun or a missing grammatical utterance

References Ammon, U. 2003. The international standing of the German language. In Languages in a Globalising World, ed. J. Maurais and A. M. Morries, 231–249. Cambridge: Cambridge University Press. Aronin, L., and M. Ó Laoire. 2004. Exploring Multilingualism in Cultural Contexts: Towards a Notion of Multilinguality. In Trilingualism in Family, School and Community, ed. C. Hoffmann and J. Ytsma, 11–30. Clevedon: Multilingual Matters. Barnes, J. D. 2005. Early Trilingualism: the development of communicative competence in English alongside Basque and Spanish. Paper presented at the 4th Internationaler Kongress über den Erwerb von Drittsprachen und die Mehrsprachigkeit, September 8–10, in Freiburg, Schweiz. Barron-Hauwaert, S. 2003. Trilingualism. A Study of Children Growing Up with Three Languages. In The Multilingual Mind: Issues Discussed By, For, and About People Living with Many Languages, ed. T. Tokuhama-Espinosa, 129–150. United States of America: Praeger Publishers. Braun, A. 2003. Trilingual Issues. http://www.trilingualism.org/. Birken-Silverman, G. 2004. Language Crossing Among Adolescents in a Multiethnic City Area in Germany. In Trilingualism in Family, School and Community, ed. C. Hoffmann and J. Ytsma, 75–100. Clevedon: Multilingual Matters. BPS. 2000. British Psychological Society. http://www.bps.org.uk/. Cenoz, J., and U. Jessner, ed. 2000. English in Europe. The Acquisition of a Third Language. Clevedon: Multilingual Matters. Clyne, M. 1982. Multilingual Australia. Melbourne: River Seine Publications. Cummins, J. 2003. The Ethics of Doublethink: Language Rights and the Bilingual Education Debate. http://www.iteachilearn. com/cummins/researchbildebate.html/. Edwards, J. 1994. Multilingualism. London: Routledge.

62

Chapter Three

Etxeberria, F. 2004. Trilinguals at Four? Early Trilingual Education in the Basque Country. In Trilingualism in Family, School and Community, ed. C. Hoffmann and J. Ytsma, 185–201. Clevedon: Multilingual Matters. Flick. 1998. An introduction to qualitative research. London: Sage. Hamers, J. F. and M. H. A. Blanc. 2003. Bilinguality and Bilingualism. Cambridge: Cambridge University Press. Herdina, P., and U. Jessner. 2002. A Dynamic Model of Multilingualism: Perspectives of Change in Psycholinguistics. Clevedon: Multilingual Matters. Hoffmann, C., and J. Ytsma, ed. 2004. Trilingualism in Family, School and Community. Clevedon: Multilingual Matters. Horenczyk, G. 2000. Conflicted identities: Acculturation attitudes and immigrants. In Language, Identity and Immigration, ed. E. Olshtain and G. Horenczyk, 13–30. Jerusalem: The Hebrew University Magnes Press. Lasagabaster, D. 1998. The Threshold Hypothesis Applied to Three Languages in Contact at School. International Journal of Bilingual Education and Bilingualism 1(2): 119–133. Maneva, B. 2004. 'Maman, je suis polyglotte!': A Case Study of Multilingual Language Acquisition from 0 to 5 Years. The International Journal of Multilingualism 1(2): 109–122. Ó Laoire, M. 2006. Issues in instructed third language acquisition involving German as a first or second foreign language: Introduction. Zeitschrift für Interkulturellen Fremdsprachenunterricht 11(1): 1. Olshtain, E., and F. Nissim-Amitai. 2004. Being Trilingual or Multilingual: Is there a Price to Pay? In Trilingualism in Family, School and Community, ed. C. Hoffmann and J. Ytsma, 30–50. Clevedon: Multilingual Matters. Oksaar, E. 1983. Multilingualism and multiculturalism from the linguist’s point of view. In Multicultural and multilingual education in immigrant countries, ed. T. Husén and S. Opper, 17–36. Oxford: Pergamon Press. Skutnabb-Kangas. 1984. Bilingualism or not: The education of minorities. Clevedon. Avon: Multilingual Matters. Tarmo, A. 2006. "Just keep talking, I missed your point". Strategy Choice in Lingua-Franca Interaction. Paper presented at the Conference on Multilingualism and Applied Comparative Linguistics, February 8–10, in Brussels, Belgium. Tokuhama-Espinosa, T. 2000. Raising Multilingual Children: Foreign Language Acquisition and children. London: Bergin and Garvey.

Multilingualism and Applied Comparative Linguistics

63

Tokuhama-Espinosa, T., ed. 2003. The multilingual mind: Issues discussed by, for, and about people living with many languages. Westport, CT and London: Praeger.

CHAPTER FOUR NATIVE JUDGEMENTS OF L2 LEARNERS’ PRAGMATIC TRANSFER: THE CASE OF APOLOGIES HORTÈNSIA CURELL I GOTOR AND MARIA SABATÉ I DALMAU

Introduction Recent studies on cross-cultural, intercultural and interlanguage pragmatics in a multidisciplinary range of fields (migration movements, multilingual business and L2 language teaching contexts) seem to have emerged hand in hand with a growing awareness of the prevalence of multilingual societies and internationalisation processes, and with the use of English as the world’s language of communication in that context (Boxer 1996; House 2006). Questions about the L2 standards that learners should aim at have been addressed, and it has been agreed among researchers that a degree of convergence with native-speaker norms that is sufficient to enable effective communication is a more realistic goal for the vast majority of non-native speakers (of English) than a full convergence (Kasper and Schmidt 1996; Kasper 1992; Barron 2003). This consensus has brought in its wake a considerable amount of research on native speakers’ assessment of non-native production of several speech acts (Murphy and Neu 1996; Bergman and Kasper 1993; Walters 1979; for apologies). In the present chapter, we examine English native speakers’ appreciation of the ways in which Catalan learners of English express the speech act of apologising in their target language In a cross-cultural study carried out by Curell and Sabaté (in press), it was demonstrated that the expression of apologies by learners of (British) English tended to differ from native speakers’ (a) in the number of illocutionary force indicating devices (IFIDs) per situation, (b) in the

Multilingualism and Applied Comparative Linguistics

65

degree of IFID intensification and (c) in the number of strategies employed per situation. The aim of the present study is to examine whether British native speakers of English actually notice these discrepancies from the nativespeaker “norm”. This may help us estimate whether the transfer of sociopragmatic patterns that are typical of the learners’ culture and L1 may jeopardise effective communication in the L2 (Clyne et al. 1991, 299) and may even fuel native interlocutors’ (negative) stereotyping of nonnatives’ perceived inclination towards “rudeness” or “over-politeness”. In our study, we will examine English native speakers’ assessment of L2 sociopragmatic performance by proficient Catalan learners of English in their production of L2 apologies in different situations (varying along the parameters of social distance and dominance), taking into account (a) the number of IFIDs and (b) the average number of strategies per situation. We hypothesise that English native speakers will tend to consider learners’ apologies less adequate than those of natives. Before we report the actual study, it may be useful to briefly discuss the nature of the speech act of apologising. Apologies are a subclass of Expressives, which “express (…) mental states of the speaker such as joy, approbation or discontent, which are important in our social forms of behavior” (Vanderveken 1990, 213). When one apologises, one expresses “sorrow or regret for something with the preparatory conditions that one is responsible for it and that it is bad for the hearer” (Vanderveken 1991, 163). An apology is “a speech act intended to provide support for the H (hearer) who is actually or potentially malaffected by a violation x” (Olshtain 1989, 156). It is a redressive act with the aim of maintaining social contract and restoring harmony between the interactants potentially put at risk by the speaker’s acts. Goffman (1967) includes it within the category “remedial work”, and Leech (1983) considers it a convivial speech act. It is face-threatening for the offender (since it implies the acceptance that something wrong has been done, intentionally or not), and face-saving for the offended (since it is made explicit that s/he has been harmed by the speaker’s actions). It has already been established that apologies vary cross-culturally (Barron 2003; Blum-Kulka et al. 1989; Coulmas 1981). Every culture has its own apologising behaviour, which tends to be highly ritualistic. The variation ranges from the actual form of the apology (I’m sorry, I apologise, excuse me) to the situations in which an apology is expected. As Clyne et al. (1991) encountered in naturally-occurring apologies among different interlanguages, apology performance can be realised differently across cultures in three different ways: formulaically,

66

Chapter Four

creatively, and with or without the use of honorifics. In terms of politeness, depending on the culture, one might lose face by (a) apologising, (b) not apologising, (c) not apologising in the appropriate way and (d) not successfully soliciting a particular reassurance (Clyne et al. 1991, 258). An apology typically consists of one or more of the following strategies, which, altogether, form the apology speech act set (Blum-Kulka et al. 1989, 289–294): -

-

-

-

-

Explicit or direct apology (IFID): IFIDs are explicit apologies, that is, they carry the illocutionary force of an apology and are the actual performative forms. They are coded by languagespecific realisations. They are formulaic, routinised expressions in which the speaker makes the apology explicit. Examples1: (I’m) sorry, I apologize, my apologies. Taking on responsibility: the speaker explicitly expresses responsibility for the offence which created the need to apologise, that is, s/he accepts to lose face. It is intended to appease the hearer. There are several subcategories, ranging from strong self-humbling to a complete denial of responsibility. Examples: I forgot to bring your book in with me today, I am late, That was completely my fault. Explanation or account: the speaker tries to self-justify by explaining the reason of the offence as caused by external (noncontrollable) factors. Examples: My phone wouldn’t stop ringing, I was held in an unexpected meeting. Offer of repair: the speaker makes a proposal to compensate for the damage or inconvenience, in contexts where the offence can be redressed. Examples: I’ll have the correct dish brought out to you immediately, I shall return it to you first thing tomorrow morning. Promise of forbearance: the speaker promises that the situation that caused the offence will not be repeated again. Examples: It won’t happen again, I shall try harder to be on time next time.

In every apology one or more of these strategies is used, together with forms that intensify the apology or downgrade the offence: -

Intensification can be IFID-internal (e.g. I’m very/terribly/incredibly sorry, I do apologise), or external, by expressing concern for the hearer, for instance (Are you alright?), or a combination of both.

Multilingualism and Applied Comparative Linguistics

-

67

Downgrading, on the other hand, consists in minimising the offence or the harm caused to the hearer and mitigating the impact caused be the socially unacceptable act, e.g. please, unfortunately.

Method Participants The non-native participants whose English (L2) apologies informed the study were 26 Catalan students of English Studies from the Universitat Autònoma de Barcelona, in their final two years, mean age 22, who had been studying English for twelve years on average and had spent a mean of two months and a week in an English-speaking country. Their level of proficiency in English corresponded to that required in the Cambridge Proficiency Exam. The English L1 apologies were collected from 26 native (British) students of Modern Languages from the University of Durham (United Kingdom) aged between 19 and 22. Both groups’ responses to a background questionnaire showed they shared comparable educational backgrounds, and were equally familiar with the kinds of situations we would present to them in the actual study (see below). A second group of native speakers were asked to evaluate the appropriateness of given apologies. These were 11 students of the universities of Durham and Manchester, aged between 19 and 23.

Materials and procedure In the first stage of the study, the Catalan participants and the first group of native speakers were asked to complete a questionnaire consisting of 52 open-ended Discourse Completion Tasks, without hearer response (Blum-Kulka et al. 1989). They were invited to write down as much as they wished and were also given the possibility to opt out (if they felt that in a specific situation they would not say anything).2 The questionnaire, which is partly shown in Appendix A, presented the respondents with eight apology situations, varying along the parameters of social distance and dominance, which have already been validated by several researchers (Cohen and Olshtain 1981; Trosborg 1995; Mir 1992; Márquez Reiter 2001), first used in the CCSARP3. These eight situations comprise all the possible combinations of the two social parameters just mentioned: social distance (+SD, -SD) and dominance (x>y; xy), hearer dominant (xy). A student forgetting to return a book to the teacher (–SD, xy). A waiter bringing the wrong dish to a customer (+SD, xy) 2.23 2.37 S2—book (-SD, xy) 1.59 2.73 S4—waiter (+SD, x, , , , 뼍鲙 (Rom. –hada)>, , , , , , , , , and . 11. For more information on lexico-semantic criteria which can be used to identify terms, refer to L’Homme (2004a). In L’Homme and Bae (2006), the application of these criteria to term selection and sense distinction in the field of computing is described in detail. 12. In these three examples, the first two Korean nouns are realized as adjectives in English and only the third is realized in English as a noun form. 13. In Korean, the adjective is a syntactically predicative unit. It is why the adjective ꟹ鲙 is translated into be empty in English. Bae (2006) gives more explanations on the functioning of Korean adjectives in general and specialized languages.

References Apresjan, Ju. D., I.M. Boguslavsky, L.L. Iomdin, and L.L. Tsinman. 2002. Lexical Functions in NLP: Possible Uses. In Computational Linguistics for the New Millennium: Divergence or Synergy? Festschrift in Honour of Peter Hellwig on the occasion of his 60th Birthday, ed. M. Klenner and H. Visser, 55–72. Heidelberg: Peter Lang. Bae, H.S. 2006. Termes adjectivaux en corpus médical coréen : repérage et désambiguïsation. Terminology 12(1): 19–50. Binon, J., S. Verlinde, J. Van Dyck, J., and A. Bertels. 2000. Dictionnaire d’apprentissage du français des affaires. Dictionnaire de compréhension et de production de la langue des affaires. Paris: Didier. Chang, S.J. 1996. Korean. Amsterdam/Philadelphia: John Benjamins. Drouin, P. 2003. Term Extraction Using Non-Technical Corpora as a Point of Leverage. Terminology 9(1): 99–117. Drouin, P., and Bae, H.S. 2005. Korean Term Extraction in the Medical Domain by Corpus Comparison. In Terminology and Content Development, ed. B.N. Madsen and H.E. Thomsen, 349–361. Copenhagen: Litera.

252

Chapter Eleven

Han, Y.K. et al. 2000. KAIST (Manual of KAIST Morpho-syntactic Analyzer). Daejeon. Jousse, A., and M. Bouveret. 2003. Lexical Functions to Represent Derivational Relations in Specialized Dictionaries. Terminology 9(1): 71–98. Lee, I.S., H.B. Im. 1983. (Grammar of Korean language). Seoul: Hagyeonsa. Lee, S.O. 1999. 霢꽩넍 ꩡ鶎·뻱鶎 霡ꓭ 꾥霡 (Research on Korean Causative and Passive Phrases). Seoul: Jipmundang. L’Homme, M.C. 2004a. La terminologie: principes et techniques. Montréal : Presses de l'Université de Montréal. —. 2004b. Sélection des termes dans un dictionnaire d’informatique: comparaison de corpus et critères lexico-sémantiques. In Proceedings of the 11th EURALEX International Congress, ed. Geoffrey Williams and Sandra Vessier, 583–593. Lorient: Université de Bretagne-Sud. —. 2005. Sur la notion de ‘terme’. Meta 50(4): 112–132. —. dir. 2006. DiCoInfo, le dictionnaire fondamental de l’informatique et de l’Internet. http://olst.ling.umontreal.ca/dicoinfo/. L’Homme, M.C., H.S. Bae. 2006. A Methodology for developing multilingual resources for terminology. Article presented at the International Conference on Language Resources and Evaluation, May 22–28, in Genoa, Italy. Lemay, Ch., M.C. L’Homme, and P. Drouin. 2005. Two Methods for extracting specific single-word terms from specialized corpora: Experimentation and Evaluation. International Journal of Corpus Linguistics 10(2): 227–255. Mel’þuk, I. 1974. Opyt teorii lingvisticheskix modelej “Smysl ļ Tekst” (A Theory of Meaning ļ Text Linguistic Models). Moscow: Nauka. —. 1996. Lexical Functions: A Tool for the Description of Lexical Relations in the Lexicon. In Lexical Functions in Lexicography and Natural Language Processing, ed. Leo Wanner, 37–102. Amsterdam/Philadelphia: John Benjamins. Mel’þuk, I. A. Clas, and A. Polguère. 1995. Introduction à la lexicologie explicative et combinatoire. Paris: Duculot. Mel’þuk, I. et al. 1984–1999. Dictionnaire explicatif et combinatoire du français contemporain. Recherches lexico-sémantiques 1-IV. Montréal: Presses de l’Université de Montréal. Mel’þuk, I., and A.K. Zholkovskij. 1984. Tolkovo-kombinatornyj slovar’ sovremennogo russkogo jazyka (An Explanatory combinatorial Dictionary of the Contemporary Russian Language). Wiener Slawistischer Almanach Sonderband 14.

Multilingualism and Applied Comparative Linguistics

253

Office québécois de la langue française. 2006. Le grand dictionnaire terminologique. http://www.granddictionnaire.com/btml/fra/r_motclef/index1024_1.asp Polguère, A. 2003. Lexicologie et sémantique lexicale: notions fondamentales. Montréal: Presses de l’Université de Montréal. (Korean Language Society), ed. 1987. (Korean orthography). Seoul: Korean Language Society.

PART FIVE TRANSLATION STUDIES

CHAPTER TWELVE -ING CIRCUMSTANCE ADVERBIALS AND THEIR TRANSLATION INTO SPANISH: A CORPUS-BASED STUDY. MARLÉN IZQUIERDO

Introduction Translation corpora have sometimes been marginalised in contrastive analysis in the belief that translated language does not provide reliable data for describing naturally-occurring language use. Such a statement is only partly true. It is true that translations are not reliable sources for describing native, spontaneous, usage of the language, but translations actually provide authentic data of a different, yet real, usage of the language, naturally occurring to make interlinguistic communication possible. Adopting a functional view of language as a means for communication, the study of translated language is thus necessary so as to understand language use in general (Baker 1993; Toury 1995). Traditionally, the English –ing form or gerund-participle (Huddleston and Pullum 2002) has been associated with the Spanish gerund (cantando). However, they do not express the same array of grammatical functions or semantic meanings, which diminishes the degree of correspondence between these units. Being aware of the real relationship between them as well as of the other Spanish resources to which the English gerund-participle is equivalent is essential for translators to do their job with some guarantee of acceptability and correctness, and for evaluators to assess the quality of translations. Within the framework of Translation Studies (TS) and focusing on its applications for language contrast, this paper attempts to map out the resources frequently used by translators to convey the meanings expressed by –ing constructions functioning as circumstance adverbials. This is a corpus-based, descriptive way to come up with an inventory of translational options which are produced in assumedly equivalent

Multilingualism and Applied Comparative Linguistics

257

communicative contexts. Likewise, I will comment on the quality of the translations by checking if the semantic content of the original text stays the same in the translation. A parallel-corpus-based methodology is the tool to select, describe and contrast the linguistic units under analysis.

Language use in contrast In an attempt at analysing and improving interlinguistic communication, Contrastive Analysis1 (CA) has settled down as a leading discipline for describing languages in contrast. In this paper I would like to draw attention to the possibility of putting language contrast under the scope of TS as well. After all, the borders between these two disciplines, CA and TS, are more and more blurred (Rabadán 1991), and cooperation of experts in each field has turned out a successful and promising approach to describing language use, as well as improving translator training and translation assessment. When combining these two disciplines, contrastive linguists are mainly concerned with contrasting linguistic phenomena which are potentially problem-triggers for translation. Therefore, their analysis focuses on establishing interlinguistic correspondences which show the similarities and–especially–differences in use between them. From the point of view of TS, translators may benefit enormously from such correspondences since these can be adopted as translational norms2 that help translators acquire language acceptability and correctness in their task. In such an approach, the direction of research implies departing from CA towards the field of TS. On the other hand, this paper attempts to show that an alternative direction of research is feasible. Researchers may depart from the analysis of translated language when interested in contrasting forms which are considered functional equivalents but still represent a problem or difficulty for the translator. Such is the case of the English –ing verbal ending or gerund-participle (Huddleston and Pullum 2002), which can display a great variety of grammatical functions and convey various meanings. CA and TS have always been related somehow but a real combination of both of them for the same piece of interlinguistic research has remained a daunting task for a long time. If this is possible now, it is partly due to the development of Corpus Linguistics, as Granger et al. (2003, 19) state: Researchers in CL [Contrastive Linguistics] and TS have come to rely on corpora to verify, refine or clarify theories that hitherto had had little or no empirical support and to achieve a higher degree of descriptive adequacy.

258

Chapter Twelve At the same time however, they do not always use exactly the same types of corpus and do not have the same research objectives.

Concerning translation, both as product and as process, corpora of translated texts are widely used as a reliable source of data on the nature of translation (Baker 1993). Since the activity of translation involves the relation of at least two different linguistic codes, which are to serve similar communication needs in different contexts, it is necessary to know them both independently as well as in contact. This is one reason why CA cannot be detached from TS or vice versa. In turn, it implies that parallel corpora may serve as a useful tool for contrastive analyses as well, being a source of translational options or, in other words, interlinguistic correspondences. Moreover, translation corpora in general are a source of data from which new exploratory ideas on the target language might arise (Mauranen 2002), insights which contrastive researchers risk missing by concentrating on individual languages alone. However, there is no doubt the source language and text might interfere into the translation, questioning the idiomaticity of the target text. In other words, the source language might have a negative influence on the translator’s task, whereby they misuse a given resource in the target language or produce an unacceptable text. In other words, there might be “translationese”. Therefore, it is advisable to verify the degree of acceptability of the translational options with which the original resource is contrasted, which can be done by means of a monolingual translation corpus consisting of spontaneous use of the target language and translations in the same language. This procedure will serve to further evaluate the degree of equivalence of the options contrasted. One way in which “translationese” may take shape is the overuse of a translational option, rendering an unidiomatic expression in the translation. As a matter of fact, due to the assumed similarity between the English –ing form and the Spanish gerund (–ndo) and the long-lasting overgeneralized correspondence between both forms, the Spanish resource is likely to be used massively, in detriment of other possible resources.

The problem: -ing circumstance adverbials The –ing ending is “a particularly ambiguous ending in English grammar” (LGSWE3 1999: 67), which is partly due to its wide and diverse functionality acquired throughout the evolution of the English language. Traditionally, the –ing form has been associated with two grammatical categories: the present participle and the gerund, but recent studies (Huddleston and Pullum 2002) have claimed that this distinction is a

Multilingualism and Applied Comparative Linguistics

259

complication in the language that is no longer justified in functional terms, coining the term gerund-participle, which refers to its wide functionality completely and implies that the English resource covers “the ground of both gerunds and present participles in other languages (ibid. 1120)”. This is the term adopted in this paper. As a non-finite verbal form4 (gerund-participle), this ending can function as subordinate clause, dependent on a finite verb phrase (VP), such as adverbial clauses, among other functions (complement subordinate clauses, relative clauses, etc.). Adverbials are not less tricky or ambiguous than the gerund-participle is, which doubles the degree of difficulty for translations. An adverbial is a clause element that adds further information to the proposition of the clause by commenting upon the circumstances of the situation or the attitude and/or style of the speaker towards their reporting. The former are widely referred to as “adjuncts” and the latter as “disjuncts” (Quirk et al. 1985). Another function of the adverbials is that of discourse connectors or “conjuncts” (ibid.), which link sentences within the discourse establishing a semantic relation between them (summative, additive, etc.). When adverbials take shape as non-finite –ing clauses, it is the expression of circumstances that prevails over the other two, which also occur but at a lower rate (Biber et al. 1999; Izquierdo 2006). Circumstance adverbials or adjuncts, using the terminology by Quirk et al., have a wide range of semantic roles which are not always easy to classify due to overlapping of meanings. Sometimes, the existence of a subordinator helps identify the meaning implied, but not all circumstance adverbials are introduced by a subordinator. This is the reason why deciding whether we have a non-finite –ing circumstance clause or not and, especially, what meaning it conveys is not always clear-cut. Hence, some of the expressions analysed in this case study may be open to discussion. For the present paper I have followed three main criteria: a) I understand that an –ing circumstance adverbial says something about the time, manner, cause, condition, etc. of the activity and/or state expressed by the main verb of the proposition it belongs to; e.g. In the past, this responsibility was given to graduate trainees after they had spent a few years working in head office. b) the –ing circumstance adverbial is governed by the VP of the main clause, rather than by a smaller element as in the example. That is, I have been careful in distinguishing –ing circumstance adverbials from –ing modifiers of other clause elements; e.g. This became our regular way of working

260

Chapter Twelve c) the –ing adjuncts identified may be introduced or not by a subordinator. Whenever there is one, I have made sure that this is not a particle required by the main verb (prepositional complement as in “…prestige of celebrated outlaws does not prevent their fame from being rather short-lived,…” ) or any other element of the clause, but part of the adverbial, which helps clarify its semantic role (see example). e.g. Three years later, when Ivor was thirteen, he took another route by working in a shop.

It seems that –ing circumstance adverbials are inherently ambiguous and, in certain co-texts, indeterminate5, as in the following example where it is not clear whether the adverbial expresses time or cause-effect. e.g. Nevertheless, in stipulating that a republican constitution "is the original basis of every kind of civil constitution" (Kant [1795] 1923, 435; [1795] 1994, 100), and in linking peace among states to their internal constitutions, Kant paved the way from Westphalian to a liberal understanding of sovereignty.

Indeterminacy poses a problem for the translator, who is to paraphrase the semantic content of the original text into the target language. In order to comply with the requirement of functional equivalence between the original text and its translation, the translator ought to have a sound knowledge of the options the target language offers to render the meanings found in the original text, as much as he ought to use them carefully. The question, thus, arises: would the Spanish translator always use a gerund to convey any of the various meanings expressed by –ing circumstance adverbials? Or, are there other ways of realizing such meanings than the gerund? If so, which are these options? One way to find an answer is by analysing what translators do, what translational options are offered. Once we know what data to contrast the gerund-participle with in Spanish, there is material to establish an inventory of translational options from which translators might benefit, as long as it has been verified whether these resources are typical native resources of Spanish to convey the original meaning to be transferred. But not only can an assumed equivalent resource in the target language misguide translators as is the case in Spanish, the lack of such an equivalent might be problematic as well, as happens in Norwegian (Behrens 1999) or German (Doherty 1999). These two authors have analysed the translation of –ing adverbials into Norwegian and German respectively adopting a semantic model in order to identify which means are used in each language to paraphrase the semantics of the original. The translation of English –ing adverbials is indeed a challenging task, no matter what the target language is.

Multilingualism and Applied Comparative Linguistics

261

The data This paper describes a corpus-based analysis of over 200 –ing circumstance adverbials. Data have been taken from the ACTRES6 Parallel Corpus (henceforth P-ACTRES), a custom-made English-Spanish translation corpus compiled at the University of León. P-ACTRES contains nearly 2.5 million words evenly distributed into five sub-corpora representing fiction literature, non-fiction writing (popular science books mainly), press editorials, magazine articles and miscellanea. For this study a smaller sample comprising over fifty thousand words in original English along with the corresponding translations into Castilian Spanish has been extracted from the popular science books sub-corpus. Table 12-1. Size of the P-ACTRES sample SAMPLE English Spanish

58,694 65,087 123,781

The study The study has been carried out according to a descriptive and contrastive methodology designed in three different stages: selection, description and contrast. The method is partly assisted by software and partly carried out manually.

Selection The parallel corpus used is grammatically tagged, which allows the users to specify their search for linguistic input in part-of-speech categories7. However, the ambiguous nature of the –ing ending, which may represent three different grammatical categories, and the fact that the tagger is not 100%8 accurate made me consider two different possibilities in order to identify as many input forms as possible. I attempted a first search in the CWB9 browser to retrieve all the words tagged as “gerundpresent participle”. This search provided 829 hits in the English original half of the sample. In a second trial, I entered “any verb ending in *ing” as the key word, which retrieved 893 hits. Even though the difference in number of concordances to be analysed between these two searches is not very big, I obviously selected the larger amount of data in an attempt at guaranteeing a relative degree of representativeness of my sample. Out of

262

Chapter Twelve

these 893 –ing constructions, 218 forms (24.4%) were manually classified as circumstance adverbials, discarding finite verbal constructions (example 1) or relative clauses (example 2) among other functions, including different types of adverbials (example 3). 1. (Finite VP) She was beginning to succumb to a ghastly fate that had already destroyed several close relatives before her eyes, a fate she had hoped would pass her by. 2. (Relative clause) To grasp the nature of energy, we need to understand two very important features concerning events and processes in the world. 3. (Style disjunct) Rather, following the Kantian tradition of cosmopolitan federalism, I will underscore the significance of membership within bounded communities and defend the need for "democratic attachments" that may not be directed toward existing nation-state structures alone.

These 218 –ing adverbials were first described intralinguistically and then contrasted with their translation from a functional point of view.

Description All the -ing adverbials were manually analysed in context so as to specify their semantic roles. This qualitative analysis yielded ten different semantic functions realized by a gerund-participle preceded or not by a subordinator. Table 12-2 below shows the frequency rates of each semantic function expressed by original –ing adverbials. Table 12-2. Frequency rates of semantic categories Semantic function Occurences MEANS 42 MANNER 35 REASON-CAUSE 34 RESULT-EFFECT-CONSEQUENCE 28 TIME 26 PURPOSE 21 SUPPLEMENTIVE 20 CONCESSION 6 CONDITION 4 PLACE 2 TOTAL 218

Percentage 19.2% 16.1% 15.6% 12.8% 12 % 9.6% 9.1% 2.8% 1.8% 1% 100

Multilingualism and Applied Comparative Linguistics

263

The semantic classification of the –ing adverbials on our sample has been done following various reference grammars such as Quirk et al. (1985) and Biber et al. (1999), although in an attempt at making up a taxonomy which unites their different criteria in a simpler way. In relation to this, instead of distinguishing means from manner, as Quirk et al. (1985) do, I have preferred a broader label, namely process, which clearly implies both semantic functions (manner and means); these meanings are so close to each other and there is so much overlap that they all somehow refer to the procedural development of an event or circumstance: in which way, by what means, with which instrument, etc., expressing a similar process relationship between the adverbial and the main verb of the proposition. 4. Carnot had developed his ideas by using simple arithmetic and had not drawn on the rigour of mathematics to express his ideas in a more modern and compelling idiom.

The function labelled as elaboration refers to so-called supplementive adverbials (Quirk et al. 1985; Biber et al. 1999) whose semantic content is not clear due to the various interpretations possible, their most clear effect being that of elaborating on the main event, hence the rather broad label of “elaboration”. Let us consider example 5; does the adverbial convey time or cause-effect? 5. Nevertheless, in stipulating that a republican constitution "is the original basis of every kind of civil constitution" (Kant [1795] 1923, 435; [1795] 1994, 100), and in linking peace among states to their internal constitutions, Kant paved the way from Westphalian to a liberal understanding of sovereignty.

This semantic function is especially interesting from the point of view of TS, since it represents a challenge for the translator, who needs to figure out the most acceptable meaning within this specific context. In turn, it foresees a wider degree of variety, subjectivity and idiosyncrasy among the translational options. The label “place” is rather broad and, as a matter of fact, the instances identified hint more at source and distance than they do at place-position. As mentioned, I have considered constructions with and without a subordinator. So, as part of the intralinguistic analysis carried out, it has also been investigated whether there are certain patterns frequently used so as to convey a specific meaning. The results obtained show that the gerund-participle is usually preceded by certain subordinators for

264

Chapter Twelve

conveying some semantic functions (e.g. concession or process) whereas some others are mostly realized through an –ing construction which does not have any subordinator (e.g. elaboration, result). Figure 12-1 below illustrates the use of a subordinator before the gerund-participle in each semantic function. - ing realization per semantic function Subordinator + 120 100 80 60 40

100 100

PLACE CONCESSION

91,4 89

MANNER 72,7 69,2

MEANS PURPOSE

50

TIME 36,3 35

CONDITION 21,4

20 0

CAUSE SUPPLEMENTIVE RESULT

Figure 12-1. Distribution (%) of ‘subordinator + gerund-participle’

Place and concession are both realized through a “subordinator + gerund-participle” construction in all their occurrences. They use the subordinators frequently associated with the semantics implied, such as ‘though, despite or whereas” for concession (example 6) or ‘from’ for place. 6. Even Lampião's most loyal lieutenant, the "blond devil" Coriseo, though remaining sentimentally attached to his old chief, quarrelled with him and took his friends and followers away to form a separate band.

In other cases, there are recurrent patterns in semantic categories such as “by + gerund-participle” to convey means and manner meanings with 88% and 54.2% of frequency rate respectively. 7. Furthermore, with fruit flies, you can increase levels of genetic variation by subjecting individuals to X rays, or to other mutagenic agents: such options, happily, are not available in dealing with humans.

Next there is ‘without’ introducing 28.2% of the –ing adverbials expressing manner.

Multilingualism and Applied Comparative Linguistics

265

8. With respect to the second principle, Kant suggests that to deny the foreigner and the stranger the claim to enjoy the land and its resources, when this can be done peacefully and without endangering the life and welfare of original inhabitants, would be unjust.

Purpose adverbials resort to “in + gerund-participle”’ in 8 out of the whole 26 occurrences (30.7%), with “for” and “to” appearing in other examples (19.2% and 11.5% respectively). 9. Much effort was expended during the nineteenth century in trying to measure the "mechanical equivalent of heat ", the work that could be obtained from a given quantity of heat, and effectively finding a conversion factor from calories to ergs.

As is typical in the English language, “after” and “before” are higher on the scale to convey time meaning. In the sample here analysed, these subordinators account for 34.6% of the total number of occurrences followed by “while” and “when”, also central in English to express time, with 15.3% rate of appearance (example 7) respectively. 10. While completing a doctorate in psychology at the University of Michigan—her thesis fittingly concerned the psychology of being at risk— she found herself increasingly involved in the affairs of the foundation.

Likewise, “when” and “if” introduce half of the –ing conditional adverbials. 11. This principle obliges signatory states not to forcibly return refugees and asylum seekers to their countries of origin if doing so would pose a clear danger to their lives and freedom.

Most causal-reason meanings are realized by a gerund-participle alone. However, out of those cases where there is a subordinator (12) “for” is the most frequent one, occurring on 9 occasions (26.4%). 12. He who deserves most credit for wasting his time in such an exceptionally fruitful manner is James Joule (1818-89).

Among the constructions conveying supplementive meanings, I have found “in” and “by” introducing some adverbials, but the tendency is not to use a subordinator as shown by example 14. 13. Nevertheless, in stipulating that a republican constitution "is the original basis of every kind of civil constitution" (Kant [1795] 1923, 435;

266

Chapter Twelve [1795] 1994, 100), and in linking peace among states to their internal constitutions, Kant paved the way from Westphalian to a liberal understanding of sovereignty. 14. Starting out in the dawn of the RFLP era, Gusella had his work cut out for him.

It seems as if the lack of a concrete subordinator tended to broaden the semantic indeterminacy in these constructions. In relation to the adverbials conveying result, “thus” and “thereby” are the predominant devices representing 66.6% of all the realizations with a subordinator, although the preferred pattern for the whole batch selected is to have the gerund-participle on its own (example 16). 15. The perversion of the modern state from being an instrument of law into one of lawless discretion in the service of the nation was completed when states began to practice massive denaturalizations against unwanted minorities, thus creating millions of refugees, deported aliens, and stateless peoples across border. 16. …a band of armed Christians attacked a caravan from Mostar near Nevesinye, killing seven Moslem carter.

After having described the source language units, I contrasted them with the translations in the target language to identify recurrent patterns for the expression of the original meanings. The contrastive study was facilitated by the alignment option in P-ACTRES10.

Contrast Almost all the –ing circumstance adverbials have a corresponding unit in the target language. Only three cases have been fully omitted, there being no translation. Two of them belong to the semantic category of process and the other one to the group of adverbials conveying result. In undertaking this interlinguistic analysis I was somewhat biased regarding the extent to which Spanish translators have resorted to the Spanish gerund to paraphrase the meaning of the original expression. Much to my surprise, the results obtained have shown that clause subordination is more frequently used than the Spanish gerund, although, overall, this resource should be considered a central one in translating circumstance meanings. The present study has shown that there is a wider range of expressive means, some of which have a frequency of use higher than that of the gerund in certain semantic categories. Table 12-3 shows

Multilingualism and Applied Comparative Linguistics

267

the various translational options found in the parallel sample under analysis and their frequency of occurrence. Table 12-3. Spanish translational options for –ing circumstance adverbials Translational option Occurrences Percentage gerund 72 33% non-finite subordinate clause 51 23.4% Finite subordinate clause 47 21.5% pp

19

8.7%

finite simple clause coordination omission Modulation total

18 7 3 1 218

8.2% 3.2% 1.5% 0.5% 100

As expected, the G is one of the most frequently used resources in the translation of –ing circumstance adverbials, since expressing adverbial meanings is its central function in original Spanish (Izquierdo, ongoing). However, the G is particularly typical in the realization of certain semantic functions, i.e. it does not express any adverbial meaning with the same typicality, and, in addition, there are other resources capable of expressing adverbial meanings in Spanish. This explains why there are other options competing with it as a translational equivalent of English –ing adverbials, such as clause subordination, which is actually more frequently used if considering non-finite and finite subordinate clauses together.

Translational options The GERUND is, by far, the most recurrent option for translating process –ing adverbials, which is due to the fact that this is one of the central semantic functions of Spanish G constructions (along with elaboration and time), function which has been inherited from the ablative case in Latin. 17. To achieve that work, an engine must be used (including the foodpowered muscle-engines of animal bodies), and for the engine to be viable, entropy must be generated by dispersing energy into the environment 17T. Para conseguir dicho trabajo hay que utilizar una máquina (incluyendo las máquinas-músculo de los cuerpos animales, cuya fuente de combustible son los alimentos), y para que la máquina sea viable hay que generar entropía dispersando energía en el entorno.

268

Chapter Twelve

Likewise, there is a great deal of G constructions in the translation of – ing adverbials of result. The use of G to express this meaning is interesting because it seems to challenge traditional grammar according to which the G is bound up with progressive aspect and simultaneity in time, which is the reason why referring to circumstances or events that occur after that of the main verb would be ungrammatical or unacceptable11. However, real data show that the G is frequently used to convey result, a meaning which logically implies time-sequence or posteriority (example 18). 18. The cells lining the tubes of the lungs can't clear out the mucus in which bacteria thrive, producing pulmonary infections. 18T. Las células alineadas en las vías pulmonares no pueden deshacerse del moco, que resulta ser un caldo de cultivo perfecto para las bacterias, produciendo infecciones pulmonares.

The G is also fairly recurrent in the translation of cause adverbials, ranking second on the list. 19. Nevertheless, despite the controversy, most Icelanders were in favor of the company, viewing it as a means of combining a noble mission— fighting genetic disease—with the happy prospect of serious money swelling the country's small economy. 19T. Sin embargo, y a pesar de la controversia, la mayor parte de los islandeses se pusieron del lado de la empresa, viéndola como el medio que permitiría combinar una noble misión, como es la lucha contra las enfermedades genéticas, con la perspectiva feliz de que cantidades importantes de dinero enriqueciesen la reducida economía del país.

I have observed that few time adverbials have been translated into Spanish by a G. This fact is a little surprising since expressing time is one of the central meanings conveyed by G constructions (along with elaboration and process). This must be due to the fact that the majority of time –ing adverbials are subordinator-headed, which triggers subordinate clauses in Spanish, for the G is never introduced by a conjunction or preposition when it realises time. As for their semantic connotations, the few G identified convey simultaneity as shown in example 20. 20. It is important to respect the claims of diverse democratic communities, including their distinctive cultural, legal, and constitutional selfunderstandings, while strengthening their commitments to emerging norms of cosmopolitical justice.

Multilingualism and Applied Comparative Linguistics

269

20T. Es importante respetar las reivindicaciones de diversas comunidades democráticas, incluyendo sus autocomprensiones distintivas en materia cultural, legal y constitucional, fortaleciendo a la vez su compromiso con normas emergentes de justicia cosmopolítica

The G is rarely used for translating purpose adverbials, which makes sense considering that this is a rather marginal function of original G constructions. Moreover, such a semantic realization is lexically constraint, i.e. in those G constructions which convey purpose this meaning lies on the lexical nature of the base verb than on the construction as a whole, as happens with “buscando” [looking for] in the following example: 21. Esto mismo sucede con los ciervos -la mayoría procedentes de repoblación- en muchos valles y que suelen salir de sus territorios buscando campos trabajados con comida fresca.

Concerning SUBORDINATION, non-finite subordinate clauses are more frequent than finite subordinates. Overall, this option is acceptable for translating any semantic function. Non-finite subordinate clauses marked by por + infinitive or al + infinitive are preferred by the translators to render causal meanings (example 22). As for finite subordinate clauses, most of them are introduced by the common Spanish causal conjunctions such as debido a, ya que, or pues. 22. Others criticize the Universal Declaration for not endorsing "institutional cosmopolitanism," and for upholding an "interstatal" rather than a truly cosmopolitan international order […]. 22T. Otros critican la Declaración Universal por no avalar el "cosmopolitismo internacional" y por sostener un orden "interestatal" en vez de un orden verdaderamente cosmopolita internacional [...].

Subordination is also the most frequent option used to translate time adverbials. The clauses are introduced by después de + infinitive or antes de + infinitive and cuando + finite verb or mientras + finite verb (mainly a finite form conjugated in the imperfect past). These Spanish subordinators are the semantic equivalents of those used in original English. 21. After qualifying as a physician at Columbia University, Huntington returned…

270

Chapter Twelve 21T. Después de terminar Medicina en la Universidad de Columbia, Huntington trabajó…

Likewise, most translations resort to para + infinitive to render purpose –ing adverbials into Spanish. 22. The energy we have extracted flows through the mechanism for converting heat into work (the piston and the cylinder in an actual steam engine) and enters the cold sink. 22T. La energía que hemos extraído fluye por el mecanismo (el pistón y el cilindro, en el caso de una máquina de vapor real) para convertir el calor en trabajo, y entra en el recipiente frío.

As already stated, elaboration adverbials do not entail specific semantic content, which in principle foresees a greater variety within the translated material, which is actually the case as illustrated in figure 12-2. Considering the possible semantic functions of the original batch, I identified in most of them binary interpretations of cause-time, time-result, time-place. The bulk of the translations express time, which in most cases is realized by means of a subordinate clause with después de [after] or cuando [when]. 24. Recognizing his own people as a nearly perfect population for genetic research, he devised a project to link the genealogical and medical records to create a database for gene-hunting. 24T. Cuando se percató de que su propia gente constituía la población casi perfecta para la investigación genética, diseñó un proyecto que permitiera vincular las historias clínicas y los datos genealógicos con el fin de crear una base de datos que sirviera para localizar genes.

Most English –ing adverbials of concession prefer a concessive subordinator which somehow clarifies the meaning of the construction. Similarly, in the Spanish translations a subordinate clause introduced by the corresponding concessive conjunction seems a better option since the G alone sounds more ambiguous as can be seen in example 25, where it is sólo which reinforces the contrast or opposite meanings implied by the combination of marginar [marginalise] and reconocer [acknowledge]. 25. Every democratic demos has disenfranchised some, while recognizing only certain individuals as full members.

Multilingualism and Applied Comparative Linguistics

271

25T. Cada demos democrático ha marginado a algunos, reconociendo sólo a ciertos individuos como miembros plenos.

Finally, subordination is the only option found amongst the translations of conditional (example 26) and locative –ing adverbials. 26. There is no obligation to shelter the other when doing so would endanger one's own life and limb. 26T. No hay obligación de dar refugio al otro cuando hacerlo pudiera poner en peligro la propia vida y seguridad.

By contrast, subordination ranks second on the list of translational options of process and result –ing adverbials. In the former, the recurrent pattern is sin + infinitive or sin que + subjunctive (example 27). 27. Clausius's remark applies to a process that is not interfered with in any way, a process that can occur without us having to drive it. 27T. El comentario de Clausius se aplica a un proceso en el que no interfiere nada, un proceso que pueda ocurrir sin que nosotros lo activemos.

The high frequency of this subordinator may be correlated with that of “without”, its functional and referential equivalent in English, and the most frequent subordinator introducing the –ing process adverbials of the original corpus sample. In the translation of result –ing adverbials there is a high frequency of relative clauses with lo que, which reinforces the cause-effect link implied in the result meaning of these adverbials (example 28). 28. The net effect is the sum of a small increase and a large decrease, giving a decrease overall. 28T. El efecto neto es la suma de un aumento pequeño y una reducción grande, lo que da una reducción general.

Some other translational options such as PREPOSITIONAL PHRASES (PP) or finite SIMPLE CLAUSES are fairly recurrent as well, although they are not common to all semantic functions. PPs are especially recurrent when translating purpose adverbials, mainly through “para + NP” (example 29).

272

Chapter Twelve 29. …that express the learning process of the nations of this world in dealing with the horrors of this century 29T. que expresan el proceso de aprendizaje de las naciones de este mundo para el manejo de los horrores de este siglo:

Concerning Finite Simple Clauses, they are a central option for translating elaboration as well as time –ing adverbials. Curiously enough, most clauses which translate elaboration adverbials carry phrases of time reference such as en cuanto [as soon as], una vez [once] etc. 30. Eli Lilly beat the rest of the world, announcing what they rather coyly called a "strong candidate" for the BRCA1 gene. 30T. Eli Lilly se impusieron al resto del mundo cuando informaron que tenían, lo que tímidamente denominaron, " el candidato con más posibilidades " de ser el gen del BRCA1.

Finite Simple Clauses are also used for translating those time adverbials which are not overtly temporal, that is, those in which there is not a time subordinator. The translation, however, includes either a time expression or another phrase with time reference (en una ocasión [on one occasion], una vez [once], etc.) (see example 31). 31. Driving with my father through a wooded road leading from East Hampton to Amagansett, we suddenly came upon two women both bowing, twisting, grimacing. 31T. En una ocasión iba con mi padre en el coche por la carretera arbolada que va desde East Hampton a Amagansett y, de repente, nos encontramos con dos mujeres arqueadas que se retorcían y hacían muecas.

Amongst the least recurrent options, there are a few instances of COORDINATION, frequently as a translation of result and elaboration – ing adverbials. There are also isolated cases among the translations of time and process –ing constructions. 32. she has made of them an extended family, greeting them each time "without theater, without simulation, without pose. 32T. para Nancy todos ellos son como una familia numerosa y siempre les saluda " sin afectación, sin fingimiento y sin pose alguna

Multilingualism and Applied Comparative Linguistics

273

Last, I have identified one case of MODULATION realised by a verbal periphrasis. This shift consists in translating the original adverbial into a periphrasis in Spanish which merges the VP of the original and the adverbial into just one unit. This option, in turn, conveys a different point of view whereby a sense of action in progress is more obvious than in the original, which seems to establish the starting point of an action. 33. While the gene hunters pushed ahead trying to isolate the genes involved in Huntington disease and DMD, a revolution of a quieter kind was taking place in the offices of clinical geneticists. 33T. Mientras que los cazadores de genes seguían intentando aislar los genes causantes de la enfermedad de Huntington y de la DMD, en los despachos de los genetistas clínicos se estaba llevando a cabo una revolución pacífica.

Figure 12-2 illustrates the range of translational options provided by the Spanish translators to paraphrase the various meanings of the original –ing adverbial constructions on our sample. 100%

Modulation Coordination

80%

Simple Clause

60%

Omission Fin Sub

40%

PP 20%

NonF Sub Gerund

Pu rp os e El ab or ati on Co nc es sio n

m e Ti

Re su lt

Ca us e

Pr oc e

ss

0%

Figure 12-2. Translational options per semantic function

Going back to the question posited earlier, we have observed that translators do not only resort to the G as a translational option of English – ing adverbials. An ongoing, large-scale contrastive project based on a comparable corpus (Izquierdo, ongoing) has revealed that Spanish G

274

Chapter Twelve

constructions do convey all the semantic functions that the G on our translated sample expresses, so, we are at a fairly safe position to argue that, qualitatively, the G is not misused. Moreover, the frequency of use is somewhat similar, for example with the G as a marginal option for translating purpose. We have observed, though, a lower frequency of use as an expressive means of time –ing adverbials, which is nevertheless explained by the fact that the original constructions are all subordinatorheaded, a better equivalent of which are subordinate clauses in Spanish.

Conclusions The intralinguistic analysis carried out on our sample has revealed that, on average, the English gerund-participle is frequently used to realize circumstance adverbials, which may express a great variety of semantic roles. This kind of adverbial realization seems to prefer the existence of an introductory subordinator, which in some cases makes the identification of the semantics conveyed by the construction clearer. What’s more, the recurrence of certain particles before the gerund-participle are well established in the language as typical patterns for the expression of certain meanings, for example, “by + gerund-participle” to express process, or “if + gerund-participle” as a conditional adverbial. The cross-linguistic analysis has revealed that the English gerundparticiple and the Spanish G are equivalent resources which share the realization of certain circumstantial meanings. Syntactically as well as semantically, they do share certain functions. Nonetheless, the G is not the only functional equivalent, for the gerund-participle functioning as a circumstance adverbial is sometimes preferably associated with some other resources in Spanish, such as subordinate clauses or PP. Assessing the acceptability of the translational options observed would require intralinguistic contrast between translated and native Spanish, which remains to be done, though. As a matter of fact, along with the selection, description and contrast stages of the method adopted in this analysis, a fourth one could be added, namely verification. This fourth stage would consist of using a monolingual corpus as a control corpus in order to contrast translator use of the Spanish gerund with native usage of this resource. In this way, the researcher could well verify to what extent the translational options identified resemble the typical usage of the language or whether they are affected by the distribution and frequency of the original gerund-participle. I have not included this stage in this paper because the set objective was to come up with a preliminary inventory of translational options which

Multilingualism and Applied Comparative Linguistics

275

unveils the relationship between the English –ing ending and the Spanish language. Moreover, the sample analysed is not representative enough, so a process of verification would not have taken us very much further. Overall, the translation keeps the semantics of the original, except those shifts into simple or coordinate clauses, where the circumstantial meaning is lost. It is remarkable that, whereas some concordances are ambiguous in English, their translation is meaning-defined as happens with the elaboration category. As for the possible applications of the results obtained, it is the field of TS where I foresee the most applications, and more precisely translator training, since the options established could serve as guides for translators. A further extension of application would cover foreign language teaching, because the contrastive results shed light on the functionality of the gerund-participle in English and also on the various ways of relating these to Spanish. Not only are the results derived from such a study useful for translation purposes, but also the linguistic descriptive process itself. In order to classify the various adverbial –ing constructions in semantic terms, the researcher has to interpret them in context, interpretation which must be done systematically, following research criteria as the ones mentioned in section three. Such a stage in the description is essential in any translation task as well, for without it the translator might not get at the meaning to be conveyed. Thus, the corpus-based study here described could well do as a guideline for translators. It is believed that the study provides them with translation strategies derived from the descriptive-contrastive analysis of parallel concordances. And it might as well help them approach the original constructions from a semantic point of view. To conclude, this paper has attempted to map out the translator’s behaviour when facing –ing circumstance adverbials. The size of the sample is not representative enough of the whole of the English language, so larger quantities of data are required. Nonetheless, the outcome obtained sheds light on the one-to-many possibilities implied in the translation process of the English gerund-participle into Spanish.

Notes 1. Currently, the term Contrastive Linguistics is very frequently used instead of Contrastive Analysis. In this paper I will talk about CA to avoid confusion between CL standing for Contrastive Linguistics and CL standing for Corpus Linguistcs. However, by CA I imply Contrastive Linguistics. 2. It is necessary to make clear that the term “norm” in translation is not understood in the same sense as Toury, for whom translation norms are

276

Chapter Twelve

extralinguistic factors influencing the translation task. The actual sense here is that of being strategies observed in translated texts and which can do as guidelines for translators. 3. Longman Grammar of Spoken and Written English (Biber et al. 1999). 4. By definition, the English –ing ending is a non-finite form of the verb which neither expresses tense or personal reference nor functions as the head of a verb phrase, even less as the nucleus of a sentence. However, when preceded by a conjugated form of the auxiliary ‘to be’, the resulting construction (the so-called progressive construction) does have referential and tense meanings. On its own, the –ing form cannot function as the head VP of an independent sentence, although it may appear in subordinate sentences. 5. By indeterminacy I refer to the fact that one given –ing construction might seem to convey, simultaneously, more than one meaning (Kortmann 1991). 6. This is the Spanish acronym for English-Spanish Contrastive Analysis and Translation Studies. (Rabadán 2002). The ACTRES Parallel Corpus has been compiled by the ACTRES research group entitled at the University of León (Spain). This corpus is one of the most fined tools developed for research and translation and language teaching applications. 7. I have used the TreeTagger POS (part-of-speech) tagging system developed by the University of Stuttgart to annotate our P-ACTRES. http://www.ims.unistuttgart.de/projekte/corplex/TreeTagger/DecisionTreeTagger.html 8. Tagging has scored 94.6% accuracy rate. 9. Corpus WorkBench. 10. Alignment has been done using a new version of the Translation Corpus Aligner developed by Knut Hofland and Øsytein Reigem at the University of Bergen (Norway), with the support of the Multilingua Marie-Curie Program. 11. This phenomenon is referred to as ‘gerundio de posterioridad’, marginalised as ungrammatical and as poor usage in Spanish. However, performance runs counter to normativity since this gerund is used more and more often. Whether this is an influence of translated language or not needs to be verified by comparing translations with original texts of the same language.

References Baker, Mona. 1993. Corpus Linguistics and Translation Studies. Implications and Applications. In Text and Technology. In Honour of John Sinclair, ed. Monica Baker, G. Francis and E. Tognini-Bonelli, 233–250. Philadelphia-Amsterdam: John Benjamins. Behrens, Bergljot. 1999. A Dynamic Semantic Approach to Translation Assessment: ING-Participial Adjuncts and Their Translation into Norwegian. In Sprachspezifische Aspekte der Informationsverteilung, ed. Monika Doherty, 99–111. Berlin: Akademie.

Multilingualism and Applied Comparative Linguistics

277

Biber, Douglas, S. Johansson, G. Leech, S. Conrad, and E. Finegan. 1999. Longman Grammar of Spoken and Written English. London: Longman. Chesterman, Andrew. 1998. Contrastive Functional Analysis. Amsterdam/Philadelphia: John Benjamins. Doherty, Monika. 1999. The Grammatical Perspective of –ing Adverbials and their Translation into German. In Out of corpora, ed. Hasselgard and Oksefjell, 269–282. Amsterdam: Rodopi. Fernández Lagunilla, Mª Luisa. 1999. Las construcciones de gerundio. In Gramática Descriptiva de la Lengua Española, ed. Ignacio Bosque and Violeta Demonte, 3443–3503. Madrid: Espasa Calpe. Granger, Silviane, J. Lerot, and S. Petch-Tyson. 2003. Corpus-Based Approaches to Contrastive Linguistics and Translation Studies. Amsterdam/ Nueva York: Rodopi. Huddleston, Rodney, and G. K. Pullum. 2002. The Cambridge Grammar of the English Language. Cambridge: Cambridge University Press. Izquierdo, M. 2006. Análisis contrastivo y traducción al español de la forma –ing verbal inglesa. Ma. Diss., University of León. Kant, E. 1795. Perpetual Peace: A Philosophical Sketch. Mount Holyoke. http://www.mtholyoke.edu/acad/intrel/kant/kant1.htm Kortmann, B. 1991. Free Adjuncts and Absolutes in English. Problems of control and interpretation. London and NY: Routledge. Mauranen, Anna. 2002. Will ‘translationese’ ruin a contrastive study?. Languages in Contrast 2(2): 161–185. Quirk, R., S. Greenbaum, and J. Svartvik. 1985. A Comprehensive Grammar of the English Language. London/New York: Longman. Rabadán, Rosa. 1991. Equivalencia y traducción. Problemática de la equivalencia translémica inglés-español. León: University of León. —. 2002. Análisis Contrastivo y traducción inglés-español: El programa ACTRES. In Nuevas Perspectivas de los Estudios de Traducción, ed. Jose Mª Bravo and Fernández Purificación, 35–55. Valladolid: University of Valladolid. Shuttleworth, M., and C. Moira. 1997. Dictionary of Translation Studies. Manchester: St. Jerome. Toury, G. 1995. Descriptive Translation Studies and Beyond. Amsterdam/Philadelphia: John Benjamins.

CHAPTER THIRTEEN TEACHING ENGLISH-SPANISH CONTRASTIVE ANALYSIS THROUGH TRANSLATION ROSA RABADÁN, BELÉN LABRADOR AND NOELIA RAMÓN

Introduction Corpus-based work has brought the benefits of empirical research to a number of cross-linguistic activities including advanced language learning and translation. These benefits are mainly concerned with the direct access to real language, both original and translated; the possibility of processing a large volume of descriptive data and the relative ease of accessibility to these data. This paper sets out to present a course on corpus-based EnglishSpanish Contrastive Linguistics for Spanish students enrolled in an English degree. It is a compulsory third year course focusing on explicit instruction in L1-L2 differences. The contents and objectives of this course will be described here with the help of illustrative examples. Why teach contrastive analysis through translation? The course outlined in this paper highlights the use of translation practice as a tool to grasp the functional semantic and pragmatic differences in a crosslinguistic study between two or more languages and uses corpora of two different types: bilingual comparable corpora and translation corpora. The main aims of this ‘contrast-through-translation’ course are to offer students the opportunity a) to raise their awareness of interlingual differences and the problems they pose (Clark and Ivanic 1995, 168–185) b) to explore the semantic and pragmatic cross-linguistic relationships between English and Spanish at different levels of analysis, and c) to improve both their understanding of the source language (English) and their accuracy in the production of the target language (Spanish).

Multilingualism and Applied Comparative Linguistics

279

The main features of the approach to language analysis to be employed in the subsequent contrastive study are outlined in the preliminary chapters.1 The text is taken as the basic linguistic unit, and language is described from the perspective of meaning in context. The method followed in contrasting two languages is explained and the concept of tertium comparationis is presented and discussed in detail. This concept is understood as the measure against which cross-linguistic similarity or difference is considered (Krzeszowski 1990). Throughout the course a set of semantic-pragmatic cross-linguistic labels fulfils this role: e.g. ‘progressive meaning’, which does not necessarily involve a specific form in Spanish, although it does in English (A staggering twenty-two percent of all the children in the U.S. are living in poverty/ Un asombroso 22 % de los niños de EE UU viven en la pobreza)2. In the final part of the preliminaries, the students are given a brief introduction to corpus linguistics from the perspective of CA (Contrastive Analysis). Large monolingual reference corpora are presented, both in English (Cobuild, BNC) and in Spanish (CREA), as well as comparable and parallel corpora in the two working languages. The parallel corpus of choice is P-ACTRES, which is an open corpus being compiled at the University of León. Students work in class with the online versions of the Cobuild, BNC, the CREA and P-ACTRES. They are shown how to use the various search possibilities of each of the corpora by means of directed tasks. A top-down design has been employed to cover most linguistic levels, from macrolinguistics (contrastive rhetoric) to microlinguistic analysis (morphosyntax, phraseology and lexis). Contrastive work is preceded by a functional description of the linguistic possibilities in the two languages for the category/semantic area being examined (Bondarko 1991) and it focuses on areas that are known to be problem triggers in cross-linguistic transfer at all the three levels: rhetorical, grammatical and lexical. In the case of macrolinguistics, authentic textual materials are used to identify and compare textual conventions in English and Spanish. As for microlinguistics, specific materials have been designed making use of authentic language samples extracted from Cobuild’s Bank of English, CREA, and from the English-Spanish parallel corpus P-ACTRES. These exercises consist of concordance lines or microtexts specifically selected from one of our three corpora. The aim is to expose students to these texts where they have to identify a) the meanings conveyed by certain grammatical structures, or b) the various formal resources employed to realise a particular meaning in English. Next, students are asked to discuss the translational solutions available in Spanish for the actualisation of

280

Chapter Thirteen

those particular meanings, and finally they are required to render those microtexts into acceptable idiomatic Spanish.

Methodology The methodology developed over the years is based on the use of a fairly standard procedure, the selection of an appropriate tertium comparationis and the use of an analytical ‘tool kit’ that includes comparable and parallel corpora as sources of empirical data. The procedure comprises three stages: description, juxtaposition and contrast (Krzeszowski 1990). The first describes the linguistic phenomenon being analysed in each of the languages independently, the second yields the map of possible cross-linguistic equivalences as well as the interlinguistic ‘gaps’3, and the third supplies the differences between the one-to-many choices, both formally similar and dissimilar in the TL. In the first and second phases we use comparable corpora, and parallel corpus data are also used in the final contrast phase. Both types of data— comparable and parallel—can be used profitably in different ways provided the same tertium comparationis is used throughout the analysis independently of the status of the data. This is one of the key issues in any contrastive study, as it has to do with what is to be compared. Do we pick up formal correspondents and work from there? Do we rather start from semantic equivalence and compare formally dissimilar items? Or do we go for chunks of language that do the same job in the same context? Although different possibilities have been discussed, among others by Catford (1965) and Krzeszowski (1984), the most widely accepted tertium comparationis for CA is translation equivalence (Halliday et al. 1964; Chesterman 1998). It is generally accepted that translation is always present either implicitly or explicitly in cross-linguistic contrastive analysis, as a means or as an end. Independently of whether we favour an inclusive definition or a partial one of ‘language that is doing the same job in the same context’ across languages, translation equivalence seems to be our best option for CA (James 1980). In our corpus-based approach this means that we need to find crosslinguistic labels at the rhetorical, grammatical and lexical levels that will comply with the requirement of ‘doing the same job’ in English and in Spanish texts. Our choices have been cross-linguistic labels which stand for what is shared in terms of global meaning: the textual function, the semantic, pragmatic or lexical meaning (see ‘Sample applications’ section). These are applied at the descriptive stage to comparable materials

Multilingualism and Applied Comparative Linguistics

281

in both languages separately and also to parallel corpus data so as to obtain the ‘translation options’ Spanish offers for one given problem area in English. The ‘tool kit’ in our procedure includes, as mentioned before, comparable and parallel corpora. Comparable corpora are “original texts in two or more languages which share certain features (typically, texts are from the same genre or subject domain)” (Salkie 1997, 305) whereas parallel corpora are said to be composed of both original texts and their translations; they are even called “the Rosetta stone of modern corpus linguistics” (McEnery and Oakes 2006, 211). Baker (1995, 234) uses the term ‘comparable’ to define corpora which are composed of original and translated texts in the same language. On the other hand, the term ‘parallel’ has been used instead of ‘comparable’—“parallel texts, i.e. authentic material in the two languages, selecting texts that are maximally equivalent on as many levels as possible” (Louhivaara 1998, 145–157). Nowadays the terminology most widely accepted is ‘comparable’ to refer to original texts in two different languages—in order to avoid ambiguity, Zanettin (1998, 616–629) calls them “bilingual comparable corpora”—and ‘parallel’ for ‘translation corpora’. The relevance of our chosen corpora for our applied purposes and their contribution as well as the advantages of using them are discussed in the following section.

The tools Corpora provide us with the authentic materials needed to offer students both data to support regularities in the use of language and also to examine particular translation possibilities for a given utterance in the source language, that is, corpora have the advantage of being “able to cater both for the linguist’s love of generalities and for the translator’s love of the instance”(Malmkjaer 1998, 534–541). That is the reason why corpora are used as the main tool in our applied study. By working on corpus-based exercises of a variety of types, students can become aware of the differences, learn to solve concrete translation problems and get a grasp on the general tendencies that characterize the use of English as compared with the use of Spanish, a kind of ‘contrast-through-translation’ course.

282

Chapter Thirteen

Monolingual corpora as “sources” for comparable corpora Large monolingual reference corpora can be used both on their own and in combination with others, as ‘source corpora’ for the building of bilingual comparable corpora (Zanettin 1998). The monolingual reference corpora of our choice are Cobuild’s Bank of English and CREA.4 Their role is to supply microtextual concondance lines in English and in Spanish, respectively. CREA is the Reference Corpus of Contemporary Spanish built by the Spanish Royal Academy. As of May 2006 it comprises 160 million words from texts written in American Spanish and in European Spanish (approximately 50% each). It includes materials from a variety of registers: books, press, miscellanea and transcriptions of oral recordings. The chronological span goes from 1975 up to the present and the interface offers an interactive feature that allows users to define the period and the size of the corpus to be used to their own convenience. Table 13-1 below shows the composition of the part of CREA that we use for our purposes – contemporary Spanish as used in Spain for the period 1990-2005. Access to the corpus is free. Table 13-1. Composition of the part of CREA selected CREA (Spain) (1990-2005) Number of words Books 25,943,651 Press 28,049,721 Miscellanea 2,454,697 Oral 2,530,072 TOTAL 58,978,141

Concordances from CREA allow us to illustrate the real use of Spanish when we are dealing with some English-Spanish contrastive topics in class, such as particular areas of non-equivalence. In Task 4, for instance, there is a preliminary stage where concordance lines from CREA are used to exemplify the use of the Spanish imperfect. The Bank of English, launched in 1991 by Collins and the University of Birmingham, is a corpus of modern English language that contains 524 million words from different sources. The online version–Cobuild Direct– offers nearly 58 million words. Throughout this paper Bank of English corpus means the restricted online version. It includes written texts come from newspapers, magazines, fiction and non-fiction books, brochures, reports, and websites. Spoken material comes from television and radio broadcasts, meetings, interviews, discussions, and conversations. A

Multilingualism and Applied Comparative Linguistics

283

subscription is needed for access to the only part of the corpus that is available online. Table 13-2 shows the composition of this part, which we use to collect data for our teaching applications. Table 13-2. Composition of the on-line part of The Bank of English Bank of English Number of words Books 10,980,698 Press 27,076,057 Miscellanea 4,349,064 Oral 15,011,670 TOTAL 57,417,489

Tasks 4 to 10 are based on microtexts which consist of concordance lines from The Bank of English specifically selected on the basis of the difficulties that some of the words or expressions they contain cause in their translations into Spanish. There are two basic perspectives to approach this type of microtextual task—the functional approach (Tasks 6, 7, 8, 10) and the formal approach (Tasks 4, 5, 9). Although there is a general agreement on the terminology, comparability still remains a controversial issue; not all authors agree upon the requirements two separate monolingual corpora need to fulfill in order to be considered comparable. It is true that CREA and The Bank of English have not been compiled as a comparable corpus ad hoc: they are completely independent of each other. For our purpose we define comparable corpora as “a collection of texts composed independently in the respective languages and put together on the basis of similarity of content, domain and communicative function” (Zanettin 1998, 617). CREA and The Bank of English are compatible in terms of quantity and quality as can be seen in Tables 13-1 and 13-2; both the total number of words and the genres included are quite similar. They are both large monitor corpora composed of a wide variety of textual types, genres, subjects, registers and media representative of their respective languages. The advantages of comparable corpora with respect to parallel corpora are, first of all, that they are much easier to obtain. It also makes sense to rely upon utterances produced by native speakers when describing linguistic uses in a contrastive way, hence the use of a corpus that contains original texts in two languages without any of the translation effects, whether induced by the source text–the so-called “translationese” features of translated texts (Baker 1996) or caused by universal translation strategies. This does not mean that translated texts are less entitled to belong to the target language; it is simply that they are considered different from original texts. Curiously enough, translated texts have been

284

Chapter Thirteen

even proved to contain more target language phrases, structures, and so on particularly characteristic of the target language than original texts, what has been called “the Eliza Doolittle phenomenon” (Malmkjaer 1998, 535).5 On the other hand, the main drawback of comparable corpora is the fact that it is difficult and time-consuming to identify equivalence relations between elements in the two languages, something which poses no problems when using parallel corpora, where the textual chunks have been aligned on the basis of translation equivalence. The use of comparable corpora is not shown in any particular task in this course but underlies many of them. Data gathered from contrastive ‘comparable-corpus-based research’ was used to empirically identify problematic areas and therefore select appropriate concordance lines for tasks 4-10. For instance, sentence 1 in task 7 has been specifically chosen because it contains a quantifier intensified by a premodifier: ‘a great many’ and a previous English-Spanish contrastive study based on the comparable corpus CREA-Bank of English (Labrador 2005) showed the frequent use of this premodifier in English as opposed to Spanish, which tends to use suffixes or other lexemes to express the intensification of quantity in general.

P-ACTRES: The English-Spanish Parallel Corpus Even though some authors have highlighted the importance of using comparable corpora for contrastive analysis (Ramón 2002; Granger et al. 2003; McEnery and Xiao 2006), our applied study has shown that this is not enough and that useful insights may also be gathered from a parallel corpus in various ways. A parallel corpus includes original texts in one language and the corresponding translation in one or more other languages. The ACTRES (Spanish acronym of “English-Spanish Contrastive Analysis and Translation”) research group has been involved in the compilation of an English-Spanish parallel corpus for several years now. For the time being, this parallel corpus is unidirectional, i.e. it is made up of original texts in English with the corresponding translations into Spanish, and is known as P-ACTRES (Parallel ACTRES). It includes written material from a variety of different registers (fiction, non-fiction, press and miscellanea) published originally both in English and in Spanish in the year 2000 or later, thus representing the contemporary stage of the language. All the materials have been published, and although it is assumed that there has been a previous editing and revision process, a simple informal test is run on each candidate textual pair to guarantee the ‘threshold quality level’ of the translations.6 The P-ACTRES corpus is still

Multilingualism and Applied Comparative Linguistics

285

under construction and aims at reaching a size of several million words. The version currently available has been aligned at sentence level using a modified version of the Translation Corpus Aligner developed for the English-Norwegian Parallel Corpus (Hofland and Johansson 1998), and can be searched with the Corpus Workbench Browser (CWB). The English half of the corpus is also tagged for part-of-speech, which enables us to carry out more refined searches7. At present, P-ACTRES contains over one million three hundred thousand words distributed between the two languages and across registers as follows:8 Table 13-3. Contents of the English-Spanish Parallel Corpus P-ACTRES MATERIALS ENGLISH SPANISH Number of words Books 482,201 509,910 992,111 Newspapers 85,441 94,912 180,353 Magazines 61,721 69,218 130,939 Miscellanea 40,178 49,026 89,204 TOTAL 669,541 723,066 1,392,607

The P-ACTRES corpus is being set up with several aims in mind, including both research and teaching purposes in the field of contrastive analysis and translation between English and Spanish. Practical applications derived from the research include teaching materials (FLT and translator training), lexicographic resources and other tools for use in the translation process or in translation quality assessment (TQA). The use of parallel corpora in teaching translation-oriented language contrasts and in translator training has been investigated by a number of authors (Ebeling 1998; Bowker et al. 1998; Botley et al. 2000; Véronis 2000; Zanettin 2000 and 2001). Translation corpora have been found useful primarily for studying the translation process and product. The main advantage of a translation corpus is the fact that it is an extensive and accurate repertoire of strategies deployed by previous translators, provided the quality of the translations can be guaranteed. A parallel corpus supplies lists of translation equivalents for potentially problematic linguistic items in the source language (SL) that have already been dealt with and transferred into the target language (TL) by professional translators. The high number of authentic translation solutions offered is an excellent source of information for translators and translator trainees, since parallel corpora contain information that is often not included in bilingual dictionaries. In other words, parallel corpora provide information that is not available elsewhere. An additional advantage of this type of database is that a parallel corpus constitutes a translation memory per se and can therefore be fed into specific computer-aided translation software

286

Chapter Thirteen

(TRADOS, Déjà Vu, etc.). Finally, most existing parallel corpora are specialized corpora, i.e. they are relatively small and contain texts belonging to a very specific subject matter.9 In these cases, parallel corpora are particularly useful, since they improve the terminological and phraseological accuracy of the text in the TL (Williams 1996). We argue in this paper that general-language parallel corpora can also be considered helpful tools in teaching language contrast through the analysis of professional translations. The P-ACTRES corpus has a number of advantages and is currently being used in different ways to build applications for the teaching of English-Spanish contrastive analysis: 1) parallel concordances for polysemous lexical items in English and their corresponding translations into Spanish can easily be extracted and edited, so that students can identify the different translational options taken and set up lists of resources available in the TL. They can further evaluate this list and complete it if they consider that there are additional possibilities that have not been exploited, or correct it if they identify translation errors; (see Task 2); 2) fill-in-the-gap exercises may be constructed on the basis of PACTRES by leaving blanks in the SL and providing the corresponding translations. Students would then be asked to fill in the gap with one or more English words that could have led to those particular translations. In Table 13-4, the degree adverb fully has been omitted. Table 13-4. Degree adverb fully and its translations into Spanish The transition to a ___ oxic world La transición hacia un mundo appears to have been protracted. plenamente óxico parece haber sido (EKAH1E.s524) larga. (EKAH1S.s501) As discussed more ____ in chapter 11, Como se discute con mayor detalle en oceans rich in oxygen from top to el capítulo 11, hasta casi el final del bottom may not have developed until eón Proterozoico no debieron de the Proterozoic Eon was almost over. haberse desarrollado océanos ricos en (EKAH1E.s525) oxígeno desde la superficie hasta el fondo. (EKAH1S.s502) Scientists didn't ___ appreciate either Los científicos de la época no podían the genealogical relationships we apreciar plenamente las relaciones know from the Tree of Life or the genealógicas del árbol de la vida que time relationships among ancient hoy conocemos, ni tampoco las rocks. (EKAH1E.s576) relaciones temporales entre rocas antiguas. (EKAH1S.s552)

Multilingualism and Applied Comparative Linguistics There was a great column which was ____ three times the reach of a man's arm in thickness... (EPJ1E.s144) On the croup of the horse and on the head and round about there were ____ ten nests of herons, who nested there every year. (EPJ1E.s148)

287

Había una gran columna cuyo grosor era tres veces la longitud del brazo de un hombre... (EPJ1S.s136) En la grupa del caballo y en la cabeza y alrededor de ella había un total de diez nidos de garzas, que anidaban allí todo el año. (EPJ1S.s140)

3) work-in-progress based on P-ACTRES has provided the research group with examples of translationese, i.e. unidiomatic translations due to the influence of the SL. Students are given these examples and are asked to provide an assessment of the translation and a justification of their judgement. In example (1) below we can see one of these instances, where the most idiomatic translational option in Spanish would have been the lexical verb in the past tense and imperfect aspect, but where the influence of the source text can be seen in the superfluous use of the verb 'poder'. Example 1: But he couldn't remember what came next. (HP1.470) No podía recordar cómo seguía. (HP1T.475)

To conclude this section, we agree with McEnery and Xiao (2006) in considering that parallel corpora cannot be the sole source of data for teaching contrastive analysis, but should be complemented with the use of comparable corpora. Our work has shown that using a combination of both comparable and parallel corpora and exploiting them for different contrastive and translation purposes is the best procedure to profit from the advantages of both.

Teaching applications The various translation-based applications that have been designed, tested and used so far fall within four categories and address different students’ needs: a) low-quality translations from English into Spanish to be rewritten by students. The aim is to develop the students’ awareness of cross-linguistic differences and the problems this causes. TASK 1: The following text has been taken from the Spanish edition of Newsweek. Read for instances of interlanguage, give reasons why those choices are not acceptable in Spanish, and try to point out the English

288

Chapter Thirteen

constructions/lexical items they are likely to derive from. Finally, try to rewrite the text in idiomatic Spanish. LOS INVISIBLES REFUGIADOS [Los (afectados) invisibles]10 Desde que viví un desastre, adquirí una nueva comprensión de las dificultades de los desposeídos. [Tras verme envuelta en un incendio, comprendo mejor los problemas de los afectados] POR JOAN R. CALLAHAN Mientras me apresuraba hacia la línea rápida, sintiendo una carga más pesada de la que pudiesen sugerir los 12 artículos que llevaba, una anciana se las arregló para expresar simultáneamente su alarma, confusión y simpatía. “Quiere pasar primero, querida?”, me dijo. Le eché una mirada a mi carrito de compras; probablemente ella asumía que yo vivía en él. Mi ropa estaba sucia y sudorosa, mi nariz había sangrado no hacía mucho y mis brazos estaban cruelmente amoratados. Sólo Dios sabe qué tipo de expresión tenía mi rostro. Para ella, el único misterio que quedaba era el por qué yo necesitaba cosas tales como yogur, ropa interior y Oil of Olay. [Cuando me acercaba, abatida, a la caja rápida (con 12 artículos), una anciana, confundida y alarmada por mi aspecto, me dijo con amabilidad: “Pase, pase”. Miré el carrito. Seguro que la señora pensaba que era una “sin techo”. Tenía la ropa sucia y sudada, los brazos llenos de cardenales y me había sangrado la nariz hacía poco. Prefiero no pensar en la cara que tendría. Para la anciana, no tenía mayor sentido que yo comprara yogur, ropa interior y crema hidratante] b) parallel texts of original English materials and their corresponding translations into Spanish to identify differences in the use of lexical, phraseological and terminological units. The aim is to add up to the students’ resource pool so as to help improve understanding of the SL and production in the TL. TASK 2: “Support verb constructions” are used in both languages but not always in the same cases. There are several possibilities: a) the same resource and the same verb are used in the two languages, e.g. ‘do the shopping’—hacer la compra, b) the same resource is used but a different verb is chosen, e.g. ‘take pictures’, hacer fotos, c) different resources are used, e.g. ‘click’ (fully semantic verb) – hacer clic (support verb

Multilingualism and Applied Comparative Linguistics

289

construction). Comment on the translation solutions adopted in the following examples for the lexical item ‘kick’. Table 13-5. Lexical item ‘kick’ and its translations into Spanish "Oh, let's kick her, just this once, - Oh, vamos a darle una patada, sólo "Ron whispered in Harry's ear, but una vez—murmuró Ron en el oído de Harry shook his head. (FRJK3E.s257) Harry, que negó con la cabeza. (FRJK3S.s263) I lie and kick my legs until the boots Me tumbo y pataleo hasta que las fly from my heels. (FWS1E.s451) botas vuelan de mis talones. (FWS1S.s439) We can indeed feel the force exerted De hecho, podemos sentir la fuerza on a ball when we strike it with a bat aplicada a una pelota cuando la or kick it with our foot. (EAP1E.s74) golpeamos con un bate o le damos un puntapié. (EAP1S.s84) We should kick the whole evil pack ¡Deberíamos echar a patadas de of them out of our country! " nuestro país a toda esa jauría! (FFK2E.s160) (FFK2S.s156)

c) comparable texts to identify differences in rhetorical and textual conventions in English and Spanish (death notices, recipes, abstracts, ads, job offers, CVs., etc.). The aim is to develop the students’ awareness of differences in rhetorical structure in texts serving the same function. TASK 3: a) Starting from text A, identify text type, text structure and text function. Do the same with text B. b) Following the procedure used in class, identify text patterns, syntactic patterns and culture-bound areas for each text. Establish typical rhetoric differences for this text form between English and Spanish texts. c) Put forward tentative guidelines for translation English-Spanish and translate text A using your newly acquired contrastive knowledge. Text A: SPICED BEEF SERVES 6 1.8 kg (4lb) salted rolled silverside 1 medium onion, skinned and sliced 4 medium carrots, sliced 1 small turnip, peeled and sliced

290

Chapter Thirteen

8 cloves 100 g (4 oz) dark soft brown sugar 2.5 ml (1/2 tsp) ground cinnamon juice of 1 lemon 1) If necessary, soak the meat in cold water for several hours or overnight, then rinse. Tie up the meat to form a neat joint and put in a large saucepan or flameproof casserole with the vegetables. 2) Cover with water and bring slowly to the boil. Skim the surface, cover and simmer for 3-4 hours, until tender. Leave to cool completely in the liquid for 3-4 hours. 3) Drain the meat well, then put into a roasting tin and stick the cloves into the fat. Mix together the remaining ingredients and spread over the meat. 4) Bake at 180ºC (350ºF) mark 4 for 45 minutes to 1 hour, until tender, basting from time to time. Serve hot or cold. (The Dairy Book of British Food) Text B: MARMITAKO Para 4 personas: 650 gramos de bonito.- 1 kilo de patatas.- ½ kilo de tomate.- 3 pimientos verdes.- 1 pimiento rojo. – 1 cebolla grande. – 3 dientes de ajo.Un trozo de guindilla.- 12 cucharadas grandes de aceite. – 1 litro de agua.Sal. En una cazuela se fríe en el aceite la cebolla y los dientes de ajo bien picados. Cuando éstos empiecen a dorar se añaden los pimientos verdes y el rojo, y los tomates pelados, todo ello bien picado. Cuando esté hecho este sofrito, se agregan las patatas troceadas y el agua, dejando que hierva con el puchero tapado. Cuando veamos que las patatas están casi hechas y blandas, se añade el bonito, que antes habremos limpiado, quitado la piel y las espinas y cortado en trozos más bien pequeños. Si por la cocción se ha evaporado mucha agua y ha quedado algo seco, se añade más agua templada.

Multilingualism and Applied Comparative Linguistics

291

Cuando esté cocido el bonito, que suele tardar aproximadamente unos 10 minutos, se sazona con sal y algo de guindilla roja u otro picante, en la cantidad que uno desee. El caldo debe quedar un poquito espeso. Este plato debe ser caldoso. Se sirve. (J. D. de Echevarría, Cocina vasca tradicional) d) English microtexts extracted from Cobuild’s Bank of English or from the English subcorpus of P-ACTRES to be analysed from a contrastive perspective and subsequently translated into naturalsounding Spanish. Our sample applications will concentrate on this task type (see ‘Sample applications’ section below). The aims are to raise the students’ awareness of the differences, facilitate the acquisition of useful conceptual tools, develop well-grounded and effective work habits, boost resourcefulness and adaptiveness and improve ST understanding and TT production. The concordance lines or microtexts are specifically selected from the English reference corpus to illustrate the different problem areas. Their length is determined by translational considerations. There are two basic perspectives when approaching this type of microtextual task: the functional and the formal approach. In the functional approach students are given microtexts in which they have to identify and analyse the various formal resources that actualise a particular function, e.g. characterisation (a list of complex noun phrases is provided where students have to identify all elements that characterise the head noun). Then students are asked to discuss the translational solutions available in Spanish for the actualisation of that particular semantic and pragmatic function; and finally students should choose the most appropriate option in each case and render those microtexts into acceptable idiomatic Spanish. Some forms in English present problems for our Spanish-speaking students due to their multifunctional nature, e.g. –ing forms. In order to increase their awareness of one-to-many correspondences between form and function, the formal approach is preferred and students are given exercises where they have to identify and analyse the meanings conveyed by one particular formal structure in different co-texts. Next, the translational solutions available in Spanish are discussed to decide upon the most appropriate translation in the microtexts given.

292

Chapter Thirteen

Sample Applications Verbal Contrasts English and Spanish differ considerably in their tense system, particularly in the way meaning functions are associated with different forms. The way modal meanings are expressed in both languages is also a contrastive issue (English modal verbs vs mood in Spanish). The main focus is on the following areas: the English Simple Past/Spanish pretérito and Spanish imperfecto contrast (Rabadán 2005a), the English Progressive forms/Spanish Simple and periphrastic forms contrast (Rabadán 2005b) and the de-centering of the English future and the options in Spanish (Rabadán, forthcoming). Our sample application focuses on the obligatory choice between pretérito and imperfecto when rendering a Simple Past into Spanish. The English form can convey absolute, anaphoric, habit, or hypothetical meanings. The Spanish preterite means exclusively absolute past, the rest of functions being covered by the imperfect. Besides, the Spanish imperfect is also capable to convey the progressive, as in (1): 1) Los guardas jurados llamaron a la policía municipal y ésta, ante el cariz que tomaba la noche, pidió ayuda a la Guardia Civil. [The security guards phoned the local police and, seeing how things were starting to look, these asked the Guardia Civil to come and help] and irrealis, as in (2): 2) En un principio, el programa iba a ser presentado por cuatro mujeres -de ahí su título-(…) Finalmente, acudirán al magazine dos colaboradoras que irán variando semanalmente….. [Initially the show was to be conducted by four women – hence its name- (…) In the end, the magazine will be presented by two different females each week …] which are typically expressed by other means in English. TASK 4: The Simple Past in English may convey ‘absolute/anaphoric past’, ‘habit’, ‘hypothetical’, etc. In Spanish the pretérito generally conveys ‘absolute past’ while the imperfecto offers a wide range of semantic functions, the most salient being ‘anaphoric past’. In the following microtexts (a) identify the meaning of the past forms; (b) decide

Multilingualism and Applied Comparative Linguistics

293

upon a translational solution into Spanish, and (c) try to render your text into acceptable Spanish. 3) Every one of the millions of television viewers around the world who saw the video of them clubbing and kicking Rodney King knows that. Meaning: [absolute past] Translation solution: Spanish pretérito. [Lo saben todos y cada uno de los millones de telespectadores que en todo el mundo vieron el vídeo en el que golpeaban y daban patadas a Rodney King. 4) People who knew him saw him as a friendly braggart, someone always having a joke, and a family man. Meaning: [anaphoric past] Translation solution: Spanish imperfecto [La gente que le conocía le consideraba un fanfarrón simpático, siempre dispuesto a una broma y dedicado a su familia]. 5) Such letters always meant disciplinary trouble Cartas como aquella significaban problemas disciplinarios indefectiblemente. Meaning: [habit] Translation solution: Spanish imperfecto. Spanish does not have a class of modal verbs, at least not in the same way English or German do. Modality can be expressed in different ways: periphrastic constructions, tenses both in the indicative and the subjunctive, adverbs, prepositions, etc., very much as in other languages, but with a radically different distribution. Class work focuses on discriminating meaning functions cross-linguistically and assigning typical translation solutions to each. This is done by means of semantic labels, which used as a tertium comparationis, help students interpret the microtext and provide a meaningful and correct translation into Spanish, as shown in task 6 below. Besides, modality is an area in which the techniques of either using low-quality translations from English into Spanish to be rewritten by students (see task 1) or having them work with parallel texts of original English materials and their corresponding translations into Spanish have proven to be extremely beneficial as they successfully raise students’ awareness of the differences, as in 6) 'Could be useful,' he said to Harry and Hermione as they went into dinner. 'Shall we go?

294

Chapter Thirteen

Podría ser útil-les dijo a Harry y Hermione cuando se dirigían a cenar-. ¿Vamos? TASK 5: MUST is a carrier of various modal meanings. Frequent functions of must are ‘obligation’; ‘command’ or ‘logical deduction’, another manifestation of necessity. Spanish offers a range of translation possibilities depending on meaning: haber que, deber or tener que when we are dealing with the meanings of ‘obligation’ and ‘command’, and haber/deber de when the meaning is ‘logical deduction’. In the following microtexts (a) characterize the meaning of MUST; (b) decide upon a translational solution into Spanish and explain why, and (c) try to render your text into acceptable Spanish. 7) You must be John’s wife Meaning: [logical deduction] Translation solution: deber de + infinitive. Debes de ser la esposa de John. 8) Official rules dictate that ministers must declare all gifts valued at more than £ 140. If they want to keep them they must fund the difference with their own money, while the rest are held for safekeeping. Meaning: [obligation] Translation solution: haber de + infinitive; tener que + inf; deber + inf. [El reglamento establece que los ministros han de declarar todos los regalos de valor superior a £ 140. Si quieren quedárselos tienen que abonar la diferencia de su bolsillo, si no se guardan a buen recaudo].

Actualization of Nouns Some of the differences between English and Spanish that are explained in detail in class are: a) The structure of the NP (more premodification in English vs. more postmodification in Spanish, possible combination of predeterminers, determiners and postdeterminers in English, e.g. ‘all eight of my’ + noun), b) zero article in English with plural count nouns and non-count nouns that whenever they express universal reference, as opposed to the use of definite article in Spanish, e.g. ‘dogs are nice pets’ and c) more frequent use of possessive determiners, especially with parts of the body and clothes, in comparison with the combined use of object personal pronouns and definite articles in Spanish in those co-texts.

Multilingualism and Applied Comparative Linguistics

295

TASK 6: Identify resources that actualise nouns in the following utterances, comment on their function in the given context and think of a suitable Spanish version. Remember that sometimes typical determiners and pronouns have other functions apart from actualising. 9) She laughs shrilly, slapping her thigh. [Se ríe de forma estridente, dándose palmadas en el muslo]. 10) All eight of my great-grandparents lived in the city. [Mis ocho bisabuelos vivían en la ciudad]. 11) A favourite of mine is a lemon sponge cake covered with lemon syrup. [Uno de mis favoritos es el bizcocho de limón cubierto con sirope de limón].

Quantification Some of the differences between English and Spanish that are problem-triggers are: a) the premodification of indefinite quantifiers (especially of those expressing proportional quantification) which is very frequent in English and rare in Spanish that tends to use suffixes to intensify or attenuate quantities, b) the use of partitives and collectives, the choice of which depend on the noun they are followed by , and c) affixative quantification which exists in both languages, but in Spanish it is much more frequent and there is a wider range of affixes available. TASK 7: Identify quantifying resources in the following microtexts and explain how they contribute to the overall meaning. Remember that sometimes quantifiers do not express quantification whereas some words that are not usually considered quantifiers can do so. Suggest suitable translations into Spanish. 12) It was applicable to a great many of us. [Se podia aplicar a muchos de nosotros] 13) A glaze of melted butter, a squeeze of lemon and a tiny pinch of nutmeg are more than enough. [Una capa de mantequilla derretida, un chorrito de zumo de limón y una pizquita de nuez moscada son más que suficiente]

Characterisation Some of the differences between English and Spanish characterising adjectives and adverbs that are explained in detail in class are: (1) the

296

Chapter Thirteen

different position in attributive function: in English, generally before the noun, in Spanish, generally after the noun, (2) the use of nouns as adjectives in English, whereas Spanish prefers prepositional phrases with de or relative clauses, (3) the common use of present and past participles in English that often correspond to adjectives, prepositional phrases or relative clauses in Spanish, (4) the frequent accumulation of adjectives in front of a noun in English, whereas in Spanish periphrastic resources and a greater variety of structures are more appropriate, and (5) the nonequivalence of –ly adverbs in English and –mente adverbs in Spanish based on pragmatic considerations, where other resources must be used instead, e.g. prepositional phrases or clauses. In order to raise the students’awareness of these differences between English and Spanish, assignments based on corpus-based contrastive materials, such as tasks 8 and 9 below are regularly given: TASK 8: In the following NPs locate the head noun and all the individual characterising resources that modify it. Translate the examples into Spanish taking care not to accumulate too many similar resources in the same NP. 14) a pronounced, sweet and sour cherry flavour [un sabor a cereza intenso y agridulce] 15) an unpasteurised organic cow’s milk cheese [un queso de vaca orgánico sin pasteurizar] 16) ten home-produced, low-budget films [diez películas de producción casera y bajo presupuesto] TASK 9: In the following examples, provide the translational options available in Spanish for the underlined adverbs. 17) a wildly inappropriate attempt at humour [un intento de ser gracioso del todo fuera de lugar] 18) used to hearing spectacularly oddball stories [acostumbrado a oír historias disparatadas]

Degree Some of the differences between English and Spanish that create crosslinguistic problemsare: a) Spanish students find it difficult sometimes to recognise the function of degree whenever it is expressed by –ly adverbs or other resources, especially lexical items originally negative in nature,

Multilingualism and Applied Comparative Linguistics

297

which intensify the quality or quantity of the head of the phrase: e.g. terribly, awful, a hell of a lot, etc., and b) another typical difficulty is to identify the function of degree in idiomatic expressions in English, e.g. wide awake, fast asleep, bone idle, etc., which can greatly affect the way the textual material is interpreted and/or transferred into Spanish. TASK 10: Identify grading resources which typically have other meanings apart from degree and suggest suitable translations in Spanish for the given utterances. 19) Many men hit their partners when they are stone cold sober.[Muchos hombres maltratan a sus parejas cuando están completamente sobrios.] 20) Ten thousand homes were badly damaged by the storm. [La tormenta causó destrozos muy importantes en diez mil casas.]

Conclusions Teaching CA through corpus-based translation applications has proven to have numerous advantages, which can be tentatively summarized as follows: -

-

It can be an effective awareness-raising tool as it highlights the expressive differences between both languages and the possible ways to bridge them. It seems to bring (linguistic) conceptual tools to the students’ resource pool and to help them develop adaptive and effective work habits. There is the perception that it has an impact on the students’ understanding and information processing of the source language (English), as it seems to make a difference in improving their skills in this area and ultimately in the quality of the target text. It seems to have a positive impact on the students’ production in the target language, as accuracy and naturalness show improvement rates.

Informal tests carried out on end of term results for four years running show a marked difference between those students who had CA training previous to the translation options offered to them and those who had not. Improvement levels were checked by giving these two different groups of

298

Chapter Thirteen

students a poor translation into Spanish and asking them to i) first identify mistakes, non-acceptable expressions and low idiomatic choices in Spanish, and ii) to provide an alternative text justifying the changes made (see Task 1 above used as an awareness raising tool). So far, we are able to report in the following areas: better understanding, more effective work routines and, above all, a general improvement in both their language analysis skills and their production in both languages. Work to produce more refined teaching tools is already underway.

Acknowledgement Research for this paper has been funded by the Ministry of Science and Education of Spain and ERDF through project Contrastive Analysis and Translation English-Spanish. Applications. (ACTRES). [Ref. HUM200501215].

Notes 1. As a background to the actual translation-based contrastive analysis between English and Spanish, the students are also provided with an introduction to the field of Contrastive Linguistics by commenting upon the relationships of CA with linguistics, psychology, FLT and translation. Some of the topics approached include a) language universals and linguistic relativism, b) bilingualism, diglossia, code-switching, pidgins and creoles, c) communicative competence, d) Transfer Theory and the concept of interlanguage, e) Error Analysis. Special care is devoted to distinguish between CA and other disciplines related to languages in contact or involving two or more languages. 2. All bilingual examples come from P-ACTRES. 3. Throughout this paper ‘interlinguistic gaps’ refers to blank spaces in the semantic/ pragmatic contrastive grid that result from juxtaposing the resources available in each language for encoding a certain function. 4. http://www.rae.es/. Students are also exposed to other English language corpora such as the BNC as an illustrative example of what a monolingual corpus in English looks like, but this particular corpus is not used for the extraction of concordances. For Spanish the choice is CREA. 5. On the frequency of use of typical features of the target language in a target language text, see also Tirkkonen-Condit (2004). 6. See Mauranen (1999). 7. We are grateful to Knut Hofland for his co-operation in the setting up of the P– ACTRES parallel corpus. 8. The figures may vary according to the date each application has been built and to adjustments of the contents to make P-ACTRES as ‘back-comparable’ as possible to CREA. The reason for these adjustments is to facilitate the comparison

Multilingualism and Applied Comparative Linguistics

299

of translated and non-translated Spanish using the same corpora used for the crosslinguistic contrast. 9. On the representativeness of small corpora and their uses, see Ghadessy et al. (2001). 10. Translations given between square brackets are our own and have not been published; otherwise they have been extracted from P-ACTRES.

References Baker, Mona. 1995. Corpora in Translation Studies: An Overview and Suggestions for Future Research. Target 7: 223–245. —. 1996. Corpus-based Translation Studies: The Challenges that Lie ahead. In Terminology, LSP and Translation. Studies in Language Engineering in Honour of Juan C. Sager, ed. Harold Sommers, 175– 186. Amsterdam/Philadelphia: Benjamins. Bondarko, Alexander. 1991. Functional Grammar. A Field Approach. Trans. I.S. Chulaki. Amsterdam/Philadelphia: John Benjamins. Botley, P. Simon, Anthony M. McEnery, and Andrew Wilson, ed. 2000. Multilingual Corpora in Teaching and Research. Amsterdam: Rodopi. Bowker, Lynne, Michael Cronin, Dorothy Kenny, and Jennifer Pearson, ed. 1998. Unity in Diversity? Current Trends in Translation Studies. Manchester, UK: St. Jerome. Catford, J.C. 1965. A Linguistic Theory of Translation. Oxford: Oxford University Press. Chesterman, Andrew. 1998. Contrastive Functional Analysis. Amsterdam/Philadelphia: John Benjamins. Clark, Romy, and Roz Ivanic. 1995. Consciousness-raising about the Writing Process. In Language Awareness in the Classroom, ed. Carl James and Peter Garrett, 168–185. London/New York: Longman. Ebeling, Jarle. 1998. Contrastive Linguistics, Translation, and Parallel Corpora. Meta 43: 602–615. Ghadessy, Moshen, Alex Henry, and Robert L. Roseberry, ed. 2001. Small Corpus Studies and ELT. Theory and Practice. Amsterdam/Philadelphia: John Benjamins. Granger, Sylvianne, Jacques Lerot, and Stephanie Petch-Tyson, ed. 2003. Corpus-based Approaches to Contrastive Linguistics and Translation Studies. Amsterdam: Rodopi. Halliday, M.A.K., Angus McIntosh, and Peter D. Strevens. 1964. The Linguistic Sciences and Language Teaching. London: Longmans, 1964. Hatim, Basil. 1997. Communication across Cultures. Translation Theory and Contrastive Text Linguistics. Exeter: University of Exeter Press.

300

Chapter Thirteen

Hofland, Knut and Stig Johansson. 1998. The Translation Corpus Aligner: A Program for Automatic Alignment of Parallel Texts. In Corpora and Cross-linguistic Research: Theory, Method and Case Studies, ed. Stig Johansson and Signe Oksefjell, 87–100. Amsterdam: Rodopi. James, Carl. 1980. Contrastive Analysis. London: Longman. Krzeszowski, Tomasz P. 1984. Tertium comparationis. In Contrastive Linguistics. Prospects and Problems, ed. Jacek Fisiak, 301–312. Berlin/New York: Mouton. —. 1990. Contrasting Languages. The Scope of Contrastive Linguistics. Berlin: Mouton. Labrador, Belén. 2005. Estudio contrastivo de la cuantificación inglésespañol. León: Universidad de León. Louhivaara, Sakari. 1998. Aspects of Adverb Position in some Finish and English Texts. In Compare or Contrast? Current Issues in CrossLanguage Research, ed. W.R Cooper, 145–157. Tampere: University of Tampere. Malmkjaer, Kirsten. 1998. Love Thy Neighbour: Will Parallel Corpora Endear Linguists to Translators?. Meta 43: 534–541. Mauranen, Anna. 1999. Will ‘Translationese’ Ruin a Contrastive Study? Languages in Contrast 2: 161–185. McEnery, Anthony, and Z. Xiao. 2006. Parallel and Comparable Corpora: What are They Up To? In Incorporating Corpora: Translation and the Linguist, ed. G. James and Gunilla Anderman. Clevedon: Multilingual Matters. McEnery, Tony, and Michael Oakes. 1996. Sentence Alignment in the CRATER Project. In Using Corpora for Language Research. Studies in Honour of Geoffrey Leech, ed. Jenny Thomas and Mick Short, 211– 231. London/New York: Longman. Rabadán, Rosa. 2005a. The Applicability of Description: Empirical Research and Translation Tools. Special issue of Revista Canaria de Estudios Ingleses 51 : 51–70. —. 2005b. Proactive Description for Useful Applications: Researching Language Options for Better Translation Practice. Meta 50(4): 1–13. —. forthcoming. Translating the ‘Predictive’ and ‘Hypothetical’ Meanings English-Spanish. Meta. Ramón, Noelia. 2002. Contrastive linguistics and translation studies interconnected: the corpus-based approach. Linguistica Antverpiensia. New Series 1: 393–406. Salkie, Raphael M. Naturalness and Contrastive Linguistics. 1997. In Proceedings of PALC ’97, ed. B. Lewandowska-Tomaszczyk and P.J.Melia, 297–312. Lodz: University of Lodz.

Multilingualism and Applied Comparative Linguistics

301

Tirkkonen-Condit, Sonja. 2004. Unique Items--- Over- or Underrepresented in Translated Language?. In Translation Universals. Do They Exist?, ed. Anna Mauranen and Pekka Kujamäki, 177–184. Amsterdam/Philadelphia: John Benjamins. Véronis, Jean, ed. 2000. Parallel Text Processing. Alignment and Use of Translation Corpora. Dordrecht: Kluwer. Williams, Ian A. 1996. A translator’s reference needs: Dictionaries or parallel corpora?. Target 8: 277–299. Zanettin, Federico. 1998. Bilingual Comparable Corpora and the Training of Translators. Meta 43: 616–629. —. 2000. Parallel Corpora in Translation Studies. In Intercultural Faultlines. Research Models in Translation Studies I: Textual and Cognitive Aspects, ed. Maeve Olohan, 105–118. Manchester, UK: St. Jerome. —. 2001. Swimming in Words: Corpora, Translation, and Language Learning. In Learning with Corpora, ed. Guy Aston, 177–197. Bologna: CLUEB.

CHAPTER FOURTEEN POLITENESS AND IMPOLITENESS IN THE SPANISH TRANSLATION OF THE CRUCIBLE: ASPECTS OF POWER AND GENDER Mª PILAR GONZÁLEZ VERA

Introduction Over the centuries, translations have contributed to the processing and exchange of information both within and across cultural boundaries. In the twentieth century Translation Studies has evolved with the changes and development of Western society and has been a reflection of them. Thus, the study and practice of translation has become in many cases an exploration of power relationships within textual practice that reflect power structures within the wider cultural context. The present analysis tries to show how the relay of power and gender relationships in a literary text is influenced by the process of translation. More specifically, the focus will be on the translation of linguistic politeness and linguistic impoliteness strategies. Politeness is an everyday term that embraces, and is conditioned by, a wide variety of issues. It is a phenomenon which occurs in all cultures but in different ways. This is the reason why polite acts cannot be separated from their cultural and social background. The aim of this paper is to study linguistic politeness and impoliteness, as influenced by power and gender, in translations. To do so, a theoretical framework on linguistic politeness, cross cultural diversity in linguistic politeness, linguistic impoliteness and an analysis of The Crucible and its translation into Spanish will be drawn up, and finally some conclusions will be drawn.

Multilingualism and Applied Comparative Linguistics

303

Linguistic Politeness Linguistic politeness is regarded as the set of strategies or verbal habits which are set as a norm for particular communities. This social dimension is commented on by House (1999, 54) when she states: “[Politeness is] a sociocultural phenomenon, roughly to be defined as showing, or appearing to show, consideration of others. Politeness can thus be seen as one of the basic social guidelines for human interaction”. The relationship between language and the social dimension has to do with the functions of language. Among the multiple functions of language, Holmes (1995, 3) points out the referential and the affective function. She (1995, 3) defines these functions in the following terms: “The referential function of language is its function in conveying information, facts, or content. The affective function refers to the use of language to convey feelings and reflect social relationships”. Therefore, in spite of being referential, the interpretation of an utterance depends on the context in which it takes place and it may also disclose the social relationship between the addresser and the addressee. The social norm perspective, the Face-saving view and the conversational-contract view have different approaches to linguistic politeness. These three approaches take a wider view of those factors which enter into the consideration of linguistic politeness. The first one, the social norm perspective, deals with the idea of linguistic politeness as a social factor. The second approach adopts Brown and Levinson’s proposal of the concepts of ‘face’ and ‘FTA’ (Face Threatening Act) in relation to the speaker’s (S) and hearer’s (H) relationship. Lastly, the third approach presents linguistic politeness as a contract for the development of the conversation that is accepted by a social group.

The Social norm perspective The social-norm perspective traditionally refers to politeness as the action of being respectful with the set of social and behavioural rules adopted by a society. The social norm view is associated with the concept of deference. Deference is related to speech styles, formality and the T/V subsystem. The forms of address, T (tu) and V (vous), are forms in pronominal systems that are used as markers of the interdependence between power, distance, and dominance. Address forms are connected with the restrictions of dominance and social distance in speech communities. The connection between participants’ behaviour, roles and situations within the social dimension is established by Spencer-Oatey

304

Chapter Fourteen

who proposes the following sets of factors which influence conversations (1996, 16): Behaviour factors:

1. Superordination 2. Association 3. Intimacy

Role factors:

1. Status and differential power 2. Formal and academic roles 3. Intimacy

Situation factors:

1. Informality 2. Constraint

In accordance with it, a speaker–who performs a specific role in the relationship–behaves in a particular way in a concrete situation. The rules which determine the participants’ behaviour are dictated by society and more specifically by those who are in power as Toledano Buendía (2003, 155) states: “Las normas son las convenciones sociales a través de las cuales el poder regula lo que se considera correcto y aceptable bajo determinadas circunstancias” (The norms are social conventions by which power regulates what is correct and acceptable under certain circumstances). Apart from these sets of social factors, gender relationships have to be taken into account in the study of the distribution of power. Although traditionally relationships between men and women have been on an equal basis, it has to be taken into consideration that societies represent hegemonic ideologies by means of gender-binary opposition. With reference to it, Coates (1997, 126) says that: “We as actors actively engaged in the construction of our social worlds inevitably perform gender in our daily interactions as either ‘being a woman’ or ‘being a man’”. Gender is undoubtedly an influential agent in the composition of social power distribution. In order to show power relationships, society has generated some strategies that highlight the distance between participants. Thus, the differentiation between levels of power in the social sphere is marked by means of forms of address, that is to say, realisations of politic behaviour and linguistic politeness like the honorific systems. Brown and Levinson (1987, 180) make reference to Comrie (1976) who argues that there are three main types of honorifics, which are classified depending on the axis on which the systems are built: the speaker – addressee axis: the relation of speaker to hearer (addressee honorifics)

Multilingualism and Applied Comparative Linguistics

305

the speaker – referent axis: the relation of speaker to things or persons referred to (referent honorifics) the speaker – bystander axis: the relation of speaker (or hearer) to ‘bystanders’ or overhearers (bystander honorifics).

Brown and Levinson (1987, 181) add another axis to those proposed by Comrie. Then the honorific axes are: Referent Addressee Speaker Bystander Setting Figure 14-1. Honorific axes (Brown and Levinson 1987, 181)

The addition of the Speaker-Setting axis is explained by the necessity of taking into account the background where the act occurs. Thus, consideration is given to social roles in a particular community, conditioned by setting and activity type. On the other hand, the speakeraddressee axis includes a recurrent instrument for giving deference, that is, the T / V systems, where the Speaker expresses respect to the Hearer with references to him/her. In short, whereas a correct assessment of the relevant socio-cultural factors and a deliberately over-politic use of address terms lead to polite situations, an incorrect judgement of those factors, whether deliberate or not, may lead to interpretations of non-politic behaviour.

The Face-saving view Brown and Levinson introduced the terms face and FTA (Face Threatening Act). The concept of face “derives from Goffman (1967) and from the English folk terms “losing face” (being humiliated) and “saving face” (being saved from humiliation)” (Brown and Levinson 1987, 61). The authors divided face into positive face and negative face depending on whether face is approved, damaged, maintained or enhanced in interaction

306

Chapter Fourteen

with others. Positive and negative politeness are attempts to satisfy the addressee’s positive or negative face desires. Holmes (1995, 5) refers to positive and negative face in the following terms: Politeness involves showing concern for two different kinds of face needs: first, negative face needs or the need not to be imposed upon; and secondly, positive face needs, the need to be liked and admired. Behaviour which avoids imposing on others (or avoids ‘threatening their face’) is described as evidence of negative politeness, while sociable behaviour expressing warmth towards an addressee is positive politeness behaviour (Brown and Levinson 1987). According to this approach, any utterance which could be interpreted as making a demand or intruding on another person’s autonomy can be regarded as a potential face-threatening act.

Moreover, Vázquez (1994) points out that “everyone is vulnerable to face-loss”, which makes participants in conversation eager not to lose face; nevertheless, “certain acts such as requests, criticisms and complaints, intrinsically threaten face” (Vázquez 1994). Those acts are known as face-threatening acts. Vázquez (1994) proposes these three variables which assess the FTA: -

the social distance (D), of speaker and hearer (a symmetric relation), the relative power (P) of speaker and hearer (an asymmetric relation) and the absolute ranking (R) of impositions within a particular culture.

Actors think they share a degree of mutual knowledge about what the variables D, P and R represent in a given context, but these are no more than the actors’ mutual assumptions, not sociologists’ estimations of the same. However, these factors may be found in all cultures so translation should take them into consideration. Furthermore, translators have to respect the strategies for doing FTAs, although they have to adapt them to the target culture in many cases. The reason for this adaptation is that different cultures show different preferences for strategies for doing FTAs. This is the case of Anglo-Saxon and Spanish cultures, where the former prefers negative politeness strategies, whereas the latter opts for positive politeness strategies, as Vázquez (1994) and Hickey (1991) point out. All in all, positive politeness involves closeness and affiliation whereas negative politeness involves distance and formality. Moreover, positive and negative politeness apply to the majority of linguistic devices used in

Multilingualism and Applied Comparative Linguistics

307

interactions, as the following table shows (Brown and Levinson 1987, 101–210): Table 14-1. Positive politeness vs. negative politeness Positive Politeness Negative Politeness 1. Claim common ground 1. Be indirect 2. Convey that S and H are 2. Don’t presume / assume cooperators 3. Fulfil H’s want for some X 3. Don’t coerce H 4. Communicate S’s want to not impinge on H. S communicates that any infringement of H’s territory is recognised as such and is not undertaken lightly. There are two basic ways that this can be done: a) Apologize for the infringement; and b) Convey a reluctance on the part of S to impose on H. S conveys that it is not his own wish to impose on H but someone else’s. 5. Redress other wants of H’s. A strategy of negative politeness consists in offering partial compensation for the face threat in the FTA by redressing some particular other wants of H’s. A basic strategy that this can be done is to acknowledge that in doing some FTA that imposes on H, one has incurred a debt.

The Conversational-contract view: Criticism to Brown and Levinson Fraser (1990) puts forward an alternative approach to the study of politeness which is called the conversational-contract view. According to Fraser when participants take part in a conversation, they are conscious of a set of rules (a contract) which have to be followed in order not to alter the atmosphere. Fraser (1990, 232) points out: each party brings an understanding of some initial set of rights and obligations that will determine […] what the participants can expect from the other(s). […] because of a change in the context, there is always the possibility for a renegotiation of the conversational contract: the two parties may readjust just what rights and what obligations they hold towards each other.

308

Chapter Fourteen

As Fraser suggests, the terms of the contract depend on the previous context of situation, on the distribution of power, the type of relationships or the participant’s aims during the interaction. Fraser maintains that complaints with the set rules is a sign of politeness. This idea is linked with Watts’ politic behaviour (1992). Both, Fraser and Watts, highlight that unmarked situations correspond with polite behaviour, whereas when the conversational contract is violated participants are aware of it and impolite exchange takes place. Fraser (1990) points out that the dimensions on which participants establish rights and obligations vary from those presented by Brown and Levinson (1987). These variations are: -

-

Some terms of a conversational contract may be imposed by convention, such as taking turns to speak, using a mutually intelligible language or speaking sufficiently loudly. Other terms are imposed by social institutions applicable to the interaction. An example of this will be observed in the analysis of the corpus, where a witness in court is expected to speak only when questioned. Finally, there are terms which are determined by previous encounters or the particulars of the situation. They may be determined by factors such as the status, the power, and the role of each speaker.

Lastly, Fraser (1999, 223) points out the main difference between Brown and Levinson’s approach and the conversational-contract view. He states that for the former being polite involves making the hearer not ‘feel bad’, whereas the latter, the conversational-contract view, involves getting on with the task in hand in light of the terms and conditions of the conversational-contract.

Cross-cultural diversity in language use: Systems of Linguistic Politeness The interest of the study of cross-cultural diversity in language use is to show how the most subtle differences in the pragmatic features of a linguistic variety are sufficient to provoke mismatches in perceived politeness, as Labov (1978) points out. Since power, distance and the weight of an imposition are different from culture to culture, the cultural context where politeness is embedded and the value systems are highly important for enabling speakers and hearers to produce and interpret

Multilingualism and Applied Comparative Linguistics

309

politeness. Cross-cultural communication shows some differences in ways of interacting politely. One of the instances in which Spanish linguistic politeness is expressed is by means of morphology. Spanish morphology is more flexible than the English and presents a wider diversity of forms for expressing politeness linguistically. Spanish verb forms are explicitly marked with suffixes selected for the degree of politeness. The English language system does not have the Spanish T/V subsystem. This linguistic device, consisting of choosing between T and V second person pronouns, strengthens social distance and power. Subordinated individuals are forced to use the V pronoun in interaction with superiors. In comparison with the use of V pronoun, T pronoun is reserved for equal relations. In a different way, English resorts to modal verbs such as would or could in order to soften directives. Apart from morphological devices, both cultures, the Anglo-Saxon and the Spanish, offer the possibility of using honorary titles. In this way, the speaker uses either the respectful or the honorary form, in order to express respect. Overall, Hickey (1991) and Vázquez Orta (1994) reach the conclusion that Spanish and English cultural systems have basic differences in the conception of what politeness means, and, consequently, in its linguistics manifestations. Thus, Spanish and English systems of linguistic politeness differ in that the former is predominantly positive whereas the latter is predominantly negative.

Impoliteness After commenting on linguistic politeness the introduction of its opposite is essential for the subsequent analysis, as the corpus presents examples of both linguistic politeness and impoliteness. Impoliteness is any face-threatening act in which the individual makes no attempt to soften it. People are supposed to behave and communicate in a polite way. However, there are circumstances where impoliteness imposes on politeness. This behaviour may be regarded as an attempt to break the rules and the status quo. Culpeper (1996, 357–358) lists positive and negative impoliteness output strategies. These are presented in the following table:

310

Chapter Fourteen

Table 14-2. Culpeper’s positive and negative impoliteness output strategies Positive Impoliteness output Negative Impoliteness output Strategies Strategies Ignore, snub the other Frighten Exclude the other from an activity Condescend, scorn or ridicule Disassociate from the other Invade the other’s space Be disinterested, unconcerned, Explicitly associate the other with a unsympathetic negative aspect Use inappropriate identity markers Put the other’s indebtedness on record Use obscure or secretive language Seek disagreement Make the other feel uncomfortable Use taboo words

The role of impoliteness in plays is crucial for the development of the plot. With regard to the function of impoliteness in drama dialogues Culpeper (1996, 364) suggests that: “impolite behaviour, either as a result of social disharmony or as the cause of it, does much to further the development of character and plot”. Balance is disturbed by impoliteness, which provokes a situation of imbalance. Culpeper (1996, 364) highlights courts as typical places where harmony is not maintained: “The courtroom provides a socially respectable and legitimate form of verbal aggression”. The fact that the principal setting of the play under analysis is a court and the overall atmosphere is of a continuous judgement marks the need to incorporate the focus on impoliteness.

Arthur Miller’s The Crucible: the play When Arthur Miller wrote The Crucible not only did he explore the madness of the Salem witch trials, but he also portrayed people from a community who were distinguished by the differences in their relationships. The Crucible portrays the awful witchcraft that took place in Salem Village (Massachusetts) in 1692. A group of adolescent girls (from eleven to twenty years old) became subject to strange and alarming fits after having heard the fantastic tales told by a West Indian slave, Tituba. The doctor believed that the daughter of the Reverend was bewitched. Nineyear-old Betty Parris and her cousin Abigail Williams, made the initial accusations because they were afraid of being punished for being witches. Immediately after, a court was convened to hear the charges of witchcraft, and those who were accused were pressured and incriminated themselves and other neighbours.

Multilingualism and Applied Comparative Linguistics

311

The story of The Crucible deals with the witch-hunting of women. Against this discrimination there was an uprising of women who aimed to improve their current situation; and this was one of the initial “rebellions” undertaken by women. As a consequence of this, women’s search for power evolved. The improvement of women’s social position is shown when translations from different periods of time are compared and a different usage of linguistic politeness is found, as will be seen in this study.

The Corpus: The Crucible and its Spanish translations There were three main reasons for the selection of the corpus for this analysis. Firstly, the fact that The Crucible is a famous play with a range of repercussions. A second reason for the selection is the main issue dealt with in the play, that is to say, the establishment, maintenance and modification of personal relationships and the ways in which these are made or not made explicit in language. Consequently, interpersonal pragmatics is the key point for the analysis of politeness and impoliteness. Thirdly, the play is full of passages in which characters seek to impose their will or have their “face” approved by the others. These face threatening acts materialise in realizations of politeness and impoliteness. Two translations have been used as target texts. The earlier one was done by Jacobo and Mario Muchnik and it was published by Compañía General Fabril Editora in 1955 in Buenos Aires. The title was translated as Las brujas de Salem. Drama en cuatro actos. The other translation was done by José Luis López Muñoz in 1997 for Tusquets Editores (Madrid). The title in this case was Las brujas de Salem. Drama, y El crisol. Guión cinematográfico basado en la obra de teatro. The dates of the translations of The Crucible are very significant for the analysis. In the 50’s Spanish culture was more traditional and conservative, in comparison with the source culture, North American, which was more progressive and not so constrained as the target culture. The earlier translation is set within a period of time where censorship was at work. At that time the target culture, Spanish, was characterised by the strong influence of religion and a consolidated patriarchal system. In accordance with those ways of thinking, translations were subordinated to a social censorship that did not accept ideas which opposed theirs. In order to avoid conflict, translators had to soften or omit those features which could alter the established situation and characters’ profiles depending on the socio-political framework. However, after four decades, the 1997 translation is framed within a modified society. This new society does not

312

Chapter Fourteen

seem to be so religious and, for instance, is more used to the occurrence of marital infidelities. On the contrary, they are more conscious about the role of women in society. Women are actively working outside the home and have invaded the man’s world. This reality is shown in the translation. Moreover, men are allowed to show their feelings, but they are still in the dominant position and they have to demonstrate their power in relation to the others. Furthermore, the ill-treatment to women becomes an increasingly discussed issue in the 90’s. Likewise, women’s situation is taken more seriously into account, and this is also shown in the more recent translation.

A Proposal of methodology: Relationships between the characters In order to explore the power and gender relationships and their linguistic expression in sample cases of politeness and impoliteness, a comparison will be made between men’s and women’s discourses. This division is based on initial feminist theories which set up a binary opposition between men and women. Nevertheless feminist theories developed and these two groups merged sometimes as will be shown in the examples. Therefore, The Crucible presents two big groups of characters with some internal divisions in relation to power distribution within Puritan society. On the one hand, the peak in the pyramidal structure is occupied by the religious authorities (Parris and Hale); on a second level the public sphere that is controlled by men; and the base of the pyramid is formed by women and lastly by servants/slaves. The starting point for the study of the translations of The Crucible will be the analysis of the character of Elizabeth. Elizabeth’s behaviour towards religious authority contrasts with her behaviour towards her husband, John Proctor. Her performance in intimacy is different from her performance in the public sphere. This change is closely related to contextual situations which determine the type of discourse. From a sociolinguistic approach and focusing on Joos’ (1967) classification of conversations, Elizabeth’s and John’s conversations belong to the category of “intimate”, which means that participants are not subjected to formality constraints. The lack of restrictions allows Elizabeth to show her determination and superiority in her relationship with her husband. Some examples of her dominant role in the relationship are found in:

Multilingualism and Applied Comparative Linguistics

313

Example 1 ST (1953, 60) Elizabeth: Then go and tell her she’s a whore. TT1 (1955, 84) Elizabeth: ¡Ve, entonces, y dile que es una ramera! TT2 (1997, 75) Elizabeth: Entonces ve a decirle que es una ramera.

The communication of emotion in marriage has to be borne in mind. Marriage is an arrangement in which both the man and the woman are involved in a close interpersonal relationship, but where gender is also a highly remarkable factor. Politeness and impoliteness play a significant role in the communication of emotions. Elizabeth’s and John’s close relationship is marked by conflict. The love affair between Abigail and John Proctor brought about a tense atmosphere in John and Elizabeth’s marital life. From the impoliteness devices set out by Culpeper (1996) a case of positive impoliteness is found in the example. The positive impoliteness resource is the use of taboo words: whore in the source text and ramera in both target texts. Elizabeth is referring to Abigail when she describes her as a prostitute, since Abigail had a love affair with her husband. Women were accused of sexual promiscuity that constituted a kind of threat to men, because it represented a way of subordinating men. This is the reason why Puritans did not like any sign of women’s sexual power. The term used in the original is a swearword and is especially inappropriate in women’s speech, but not so much in men’s discourse. Nevertheless, her usage of that word brings Elizabeth closer to men’s behaviour and consequently to power. Furthermore, Elizabeth’s purpose is to offend and insult Abigail, and impoliteness functions as a tool of offence as Holmes (1995, 4) states: “‘politeness’ describes behaviour which is somewhat formal and distancing, where the intention is not to intrude or impose […] Being polite means expressing respect towards the person you are talking to and avoiding offending them”. On the other hand, the translation as ramera is a softened version of whore in the original. Whereas in the 1955 translation it would be admissible because of the social perception of women at the time, this option did not have the same effect in the 90s. This choice relies on a question of gender as can be seen when John Proctor uses the same word to insult Abigail: Example 2 ST (1953, 97) John Proctor: How do you call Heaven! Whore! Whore! TT1 (1955, 144): ¡Cómo te atreves a clamar al Cielo! ¡Ramera! Ramera! TT2 (1997, 121): ¡Cómo te atreves a clamar al cielo! ¡Puta, más que puta!

314

Chapter Fourteen Example 3 ST (1953, 98) John Proctor: My wife is innocent, except she knew a whore when she saw one! TT1 (1955, 146): ¡Mi mujer es inocente, solo que reconocía a una ramera cuando la veía! TT2 (1997, 122): ¡Mi esposa es inocente, toda su culpa fue reconocer a una puta cuando la vio!

Both 2 and 3 are examples of impoliteness through the use of the swearword whore. John Proctor attacks Abigail’s promiscuity when he calls her “whore”. In example 3 John sets a comparison between his wife and Abigail. While Abigail is described as a whore, Elizabeth, John’s wife, is “innocent”, as he states. The contrast shown by John’s utterance backs up the relationship which has already been discussed between promiscuity in female bad characters, and good mothers and traditional women. On account of this, translations are given: “mujer” in the 1955 and “esposa” in the 1997 target text. Whereas the second term is more respectful, the first one deals with women as possessions. In that way, translations and their implicatures are conditioned by the socio-political context. As in the 50’s women had no economic independence and they were submissive to men, they were called “woman”. By comparison, in the 90’s women were more autonomous and men started to refer to them including deference nuances. On the other hand, the word whore is translated in two ways: “ramera” in 1955 and “puta” in 1997. Rude language tended to be softened in the 50’s in Spain. But in the 90’s naturalism was in fashion. Writers tended to show daily life accurately without disguising any detail in spite of the fact that portrayals were too harsh. In this aim to reflect reality, swearwords were not censored, as they were present in everyday language. Apart from that, it is remarkable that the same word is used by Elizabeth to insult Abigail, as was seen in example 1. However, the translation of the term differs depending on the gender of the speaker. Although both examples, 1 and 2, have resorted to the same word in the source text to insult the same woman, Abigail, this word has been translated differently. The translation changes depending on who utters it. John as a man is allowed to refer to Abigail in such terms. On the contrary, Elizabeth, as a woman, is restricted in the way she expresses herself. In the 1955 publication, there is no difference in the translation of whore in John and Elizabeth’s interventions, due to reasons of censorship; and the old form ramera is shared by male and female speakers. In that decade Spanish translations were subject to censorship and a swearword such as puta, which would be

Multilingualism and Applied Comparative Linguistics

315

more appropriate, was not easily accepted, so that a euphemism had to be used (ramera). Nevertheless, in 1997 there was no censorship and different translations are still found in the target texts. In the 1997 target text whore is translated as ramera in example 1 and as puta in example 2. The choice relies on the gender of the speaker. Puta, the swearword with all its power and harshness, is uttered by a man (Proctor), whereas ramera, the old-fashioned word, which softens its original meaning, is uttered by a woman. Women, therefore, do not have enough power to use such a term as puta, but men are allowed to use it. With regard to Abigail’s discourse, first of all, attention has to be paid to Hymes’ (1972) sociolinguistic notion of scene. The setting is in a Puritan background. She belongs to a community, which implies the adoption of certain standards, as Mills (2004, 2) points out, quoting Eckert and McConnell-Ginet (1998, 490): A community of practice is an aggregate of people who come together around mutual engagement in some common endeavour. Ways of doing, ways of talking, beliefs, values, power relations – in short, practices – emerge in the course of their joint activity around that endeavour.

According to Mills and Eckert and McConnell-Ginet, members of a community accept a mutual engagement in which they undertake to follow a set of specific rules of behaviour. This compromise works as a contract and one of the terms involved deals with conversational precepts. Conversational-contract constrains the behaviour of the participants in a verbal interaction. Regarding this contract, Watts (1992, 59) states that “if both (or all) participants abide by the rules of that contract, they are said to be acting “politely”. Violation of those rules results in impolite behaviour (Fraser and Nolen 1981, 96)”. A distinctive feature of Abigail’s character is her continuous attempt to impose her own will. By nature Abigail belongs to the female group and should follow female-speech patterns. However, she diverts from the norm and she does not behave as she is supposed to do. She is considered a kind of revolutionary in the light of feminist theories. In that sense and following Fraser and Nolen (1981), since Abigail breaks the rules, her behaviour is consequently found impolite. Holmes (1995) suggests that distribution of power in societies is one of the three agents for explaining men and women’s interaction. Power has a binary distribution in The Crucible, that is, members who move in the public sphere, and those from the domestic sphere. However, Abigail tries to cross the boundary between them. The leap from one group (the female) to the other (the male) is shown by her using male-speech resources.

316

Chapter Fourteen

An example of transgression is found in her way of dealing with sexual matters and referring to her body without embarrassment: Example 4 ST (1953, 29) Abigail: I know how you clutched my back behind your house and sweated like a stallion whenever I come near! Or did I dream that? It’s she put me out, you cannot pretend it were you. I saw your face when she put me out, you cannot pretend it were you. I saw your face when she put me out TT1 (1955, 37): Lo que sé es cómo me estrechabas en los fondos de tu casa, y sudabas como un caballo cada vez que me acercaba. ¿O es que lo he soñado? Quien me echó fue ella, no puedes simular que fuiste tú. Te vi el rostro cuando ella me echó TT2 (1997, 36): ¡Sé que me abrazaste por la espalda detrás de tu casa y que sudabas como un semental cada vez que me acercaba! ¿O es que eso lo he soñado? Fue ella quien me puso de patitas en la calle, no finjas que fuiste tú. Vi la cara que pusiste cuando me echó

Example 4 shows Abigail’s description of her sexual relation with John Proctor. The way in which she relates the facts is more typical of men than of women. She talks like a man and this cross-gender attitude is regarded as impolite. In respect to that, Mills (2004, 5), referring to Walsh (2000), states that “women often use styles of speech in their interventions in the public sphere which are coded as masculine, but they run the risk of being judged as transgressive or abnormal for engaging in them”. Two points emphasise this “male speech”: the image of John Proctor as a stallion and the commentary on “who put her out”. Regarding the first issue, there are two versions for stallion: “caballo”–horse–in TT1, and “semental”–stallion–in the 1997 target text. In the first case there has been a loosening of the source term, stallion; a neutral form has been preferred. This choice avoids the sexual connotation found in the source text, whereas the 1997 target text keeps the essence of the word. Stallions are stud horses, those which are destined for breeding. John Proctor behaves sexually as an animal with Abigail, in accordance with her depiction of the scene, and that behaviour loses strength in the 1955 translation, where the neutral word horse is used. In that sense TT1 is an example of negative impoliteness. By contrast, TT2 prefers a literal translation of the term and thus an explicit association of John Proctor with a negative aspect (behaving as an animal) is made. Consequently, TT2 opts for positive impoliteness, since the sexual connotations of stallion are not “disguised”, and the inappropriate identity marker from the original is maintained.

Multilingualism and Applied Comparative Linguistics

317

Secondly, Abigail is open to John about the fact that it was John’s wife and not he who put her out. Abigail uses the negative impoliteness device of fear as a weapon to make John lose face. She ridicules (another case of negative impoliteness) John, because he is less powerful than his wife. She is the one who makes decisions and the one who is indeed in power. This accusation makes John lose face. The effect is achieved in both translations, but the way of expressing it is different. In TT1, the statement is a literal translation of the source text, while TT2 enriches the source text. The latter translation provides explicitly a colloquial nuance which portrays Abigail as a woman who tackles man’s power through insolent and discourteous manners. Moreover, those women who do not follow the femininity patterns are identified with evilness and even with violence. The Crucible collects a great quantity of examples where Abigail or teenagers who support her turn to violence. Some cases are: Example 5 ST (1953, 26) Abigail: Betty? Now, Betty, dear, wake up now. It’s Abigail. I’ll beat you, Betty! TT1 (1955, 33): Betty. Vamos querida, Betty, despierta ya. Es Abigail. ¡Betty, voy a pegarte! TT2 (1997, 33): ¿Betty? Vamos, Betty, cariño, despiértate ya. ¡Te voy a dar una paliza!

And Example 6 ST (1953, 26) Abigail: […] I saw Indians smash my dear parents’ heads on the pillow next to mine, and I have seen some reddish work done at night TT1 (1955, 34): […] He visto cómo, sobre la almohada junto a la mía, los indios destrozaban las cabezas de mis pobres padres, y he visto algunas otras sangrientas faenas realizadas en la noche, TT2 (1997, 32): […] Vi cómo los indios aplastaban la cabeza de mis padres sobre la almohada, a mi lado, y también he visto otros horrores nocturnos con mucha sangre;

Abigail threatens to hurt Betty physically. Physical mistreatment is more commonly associated with men than with women, but Abigail resorts to force as men would. Regardless of the harshness reflected in the source text about Abigail’s violent reaction, there is a great contrast between the 1955 translation and the 1997 one. In TT1 Abigail’s conduct

318

Chapter Fourteen

is handled as if it were a sample of childish behaviour. In order to elude female brutality, the expression “I’ll beat you” is loosened. In contrast, it has to be noticed that in the 90’s, people were already extremely concerned with gender violence. Woman battering became a usual topic in twentieth century social conversations. This is the reason why the translator opts for “te voy a dar una paliza”. It expresses explicitly the act itself and “dar una paliza” is the colloquial expression that becomes popular in the 90’s to refer to physical ill-treatment. In a similar way, in example 6, Abigail includes cruel references such as the image of seeing Indians “smash my dear parents’ heads”. The verb “smash” is softened in the 1955 translation, whereas in 1997 it holds the barbarity and inhumanity of the action. The reason for it is again a question regarding the social concern in the 90’s for violent acts. Finally, with reference to authorities, it is a distinctive feature in The Crucible that Miller unfurls the plot within a Puritan community. In addition, the fact of belonging to the Puritan community sets an obvious distance between religious leaders and the rest of the community. Distribution of power may be understood as men’s physical strength but also as men’s superiority in issues such as social prestige or economic matters. Those men to whom the community has to give deferential treatment, deal with mature women with no mark of deference, as if they were girls or teenagers. Although information about hierarchical social dimension is encoded indirectly, as Brown and Levinson (1987) say as well as Haas (1964) and Dixon (1972), a direct encoding of social category with reference to gender is perceived in the next example: Example 7 ST (1953, 66) Hale (to Elizabeth): And you, woman? TT1 (1955, 93): ¿Y tú, mujer? TT2 (1997, 83): ¿Y usted, señora?

Hale, one of the men in power within the community, addresses Elizabeth with no sign of deference other than the word “woman”. That term is very vague and can be used for any kind of woman without alluding to her social or age group. It may be regarded as a contemptuous manifestation of Hale towards Elizabeth because of her gender. TT1 preserves this effect, not only does it keep a literal translation of “woman” for “mujer”, but it also translates “you” as “tú”. The personal pronoun for the second person, “tú”, does not imply deference, but closeness or superiority of Speaker (S) in relation with Hearer (H). By contrast, the

Multilingualism and Applied Comparative Linguistics

319

1997 translation prefers “señora” for “woman” and “usted” for “you”. Firstly, “señora” implies distance between S and H. It is a negative politeness strategy and it suggests formality. Moreover, “usted” is the V form for addressing the H showing formality and consideration for him; so this translation is not as accurate as the first one, in relation to the source text.

Conclusion From the comparative analysis of the translations of The Crucible one can conclude that the most useful concept of linguistic politeness applied in translation studies must be a general one, which covers on the one hand those concepts of linguistic politeness regarded as relevant for translation, and on the other hand, which takes into account a broader perspective of language functions, specifically the interpersonal. The function of a text may be seen in a wider, social context, that is, how a text manifests the structure and functioning of a particular society. Similarly, translations may also have an effect on the target culture. Such an effect may result from the portrayal of the source culture that a translation presents to the target readers. Furthermore, if the target text is translated after a period of time in which society has suffered many changes, translations may become a vehicle for showing the development of society. The study of the two translations of The Crucible into Spanish, from different periods of time, one from the 50’s and another from the 90’s, has aimed to show how those target texts became a reflection of the process of change in interpersonal relationships within society. For the analysis of the Spanish translations, the study of linguistic politeness and impoliteness, cross-cultural diversity in language and gender and power relationships has been necessary. After the analysis of the two translations, it can be seen how translations are subordinated to culture and to current ideology. There is also a temporal dependence. Gender relationships as well as politeness are not static but they change with time. Therefore, the use of politeness strategies as well as the translation of those strategies, differ due to temporal reasons among other factors. With regard to social changes, it can be seen that the 50’s society is based on a binary distribution between men and women. This duality comes from the natural differentiation gender provides, but also from manmade causes: the distribution of social power. This dual situation is also reflected in literature as Cantor (1986, 69) points out:

320

Chapter Fourteen We contend that no genre is “realistic”. Rather, through stories, a fictionalised representation of our social structure and social relations are presented. These fictionalised representations provide a mirror of the world, showing how power is allocated in society and how dominance and submission are idealized.

This distribution of power where men hold the control and power in society and women are subordinated to men is noticed in the earlier translation (1955). There, the hegemonic position in society is held by men and, consequently, women tend to use deferential terms of address, honorary titles and V pronoun to address men, whereas men do not usually refer to women by means of titles or through negative politeness and impoliteness devices. Men use the T pronoun form to speak to women, regardless of women’s age. Therefore, the choice of linguist politeness strategies relies on gender relationships. However, the distribution of power on the basis of gender starts to be blurred in the 90’s, after feminist ideas became widespread in the 60’s. It is in the 1997 translation where impoliteness becomes significant. The importance of impoliteness for the analysis has to do with its relationship with gender and the frequency with which it was applied, especially in this TT2. In comparison with recent translations, those done in the 90’s, the earlier ones did not present any proper impolite feature. Moreover, if rudeness or impoliteness has been found towards another participant in the 1955 target text, there are two reasons which may explain its use: a) the speaker is a man addressing women; or b) all the participants are men, but the man who addresses the others impolitely is the one who holds the authority in the group. The language of the 1955 target text becomes artificial in comparison with the one from the 90’s, which uses up-to-date language. That is the case of the translation of those terms which make reference to Abigail and her relationship with John Proctor, which keep the harshness from the source text in the 90’s, whereas the 1955 target text prefers softening them. Furthermore, examples of positive and negative linguistic politeness are found throughout the play. Positive politeness is used for minimizing the distance between participants and it is preferred among women who are friends. By contrast, negative politeness avoids intruding, and so emphasizes the social distance between people, so this is preferred to address men. Nevertheless, due to social and cultural reasons, whereas

Multilingualism and Applied Comparative Linguistics

321

English seems to prefer negative politeness devices, Spanish translations tend to use more positive politeness devices, as has been observed. It is also remarkable that in the 1997 translation violent references are softened in the majority of examples. In my view, the reason for this softening may be that the society from the 90’s started to be aware of gender violence. In conclusion, gender dominates the choice of politeness strategies and any change a society suffers is reflected in language use. Consequently, it can be said that the analysis of 1955 and 1997 translations has demonstrated that translation becomes a vivid and dynamic manifestation of the structure and functioning of the society in which it is produced.

References Brown, P., and S. Levinson. 1987. Politeness. Some universals in language usage. Cambridge: Cambridge University Press. Cantor, M. G, and S. Pingree. 1986. The Soap Opera. Beverly Hills / London / New Delhi: Sage Publications. Coates, J. 1997. One-at-a-Time: The Organization of Men’s Talk. In Language and Masculinity, ed. S. Johnson and H.M. Ulrike, 107–129. Oxford: Blackwell. Comrie, B. 1976. Linguistic politeness axes: speaker-addressee, speakerreferent, speaker-bystander. Pragmatics Microfiche 1.7: A3. Culpeper, J. 1996. Towards an anatomy of impoliteness. Journal of Pragmatics 25: 349–367. Dixon, R.M.W. 1972. The Dyirbal language of North Queensland. Cambridge: Cambridge University. Eckert, P., and S. McConnell-Ginet. 1998. Communities of practice: where language, gender and power all live. In Language and Gender: A Reader, ed. J. Coates, 484–494. Oxford: Blackwell. Fraser, B. 1990. Perspectives on Politeness. Journal of Pragmatics 14: 219–236. Fraser, B., and W. Nolen. 1981. The association of deference with linguistic form. International Journal of the Sociology of Language 27: 93–109. Goffman, E. 1967. Interaction Ritual: Essays on Face to Face Behaviour. New York: Anchor. Haas, M. 1964. Men’s and women’s speech in Koasati. In Language in culture and society: A Reader in Linguistics and Anthropology, ed. Dell Hymes, 228–233. New York / Evanston / London: Harper & Row.

322

Chapter Fourteen

Hickey, L. 1991. Comparatively polite people in Spain and Britain. ACIS 4(2): 2–6. Holmes, J. 1995. Women, men and politeness. London: Longman. House, J. 1999. Politeness and translation. In The Pragmatics of translation, ed. L. Hitckey, 54–71. Clevendon: Multilingual Matters. Hymes, D. 1997. Models of the interaction of language and social life. In Analysing Casual Conversation, ed. S. Eggins, and D. Slade, 35–71. London: Cassell. Joos, M. 1967. The Five Clocks. New York: HBJ Harbinger. Labov, W. 1978. Modes of mitigation and politeness. In A pluralistic nation: the language issue, ed. M.A. Lowrie, and N.F. Conklin, 382– 389. Rowley, Mass.: Newbury House. Miller, A. 1968 [1953]. The Crucible. A Play in Four Acts. London: Penguin Books. —. 1955. Las brujas de Salem. Drama en cuatro actos. Trans. Jacobo and Mario Muchnik. Buenos Aires: Compañía General Fabril Editora. —. 1997. Las brujas de Salem. Drama, y El crisol. Guión cinematográfico basado en la obra de teatro. Trans. José Luis López Muñoz. Madrid: Tusquets. Mills, S. 2004. Rethinking Politeness, Impoliteness and Gender Identity. http://www.lboro.ac.uk/departments/ea/politeness/Genderand Politeness.htm Toledano Buendia, C. 2003. La traducción de la Obscenidad. Santa Cruz de Tenerife: La Página. Spencer-Oatey, H. 1996. Recogsidering power and distance. Journal of Pragmatics 26: 1–24. Vázquez Orta, I. 1994. Politeness phenomena in English and Spanish: a pragmatic approach. Article presented at the XII AESLA Congress, April 20–22, In Barcelona, Spain. Walsh, C. 2000. Gender and mediatised discourse. PhD diss., Sheffield Hallam University. Watts, R. 1992. Linguistic politeness and politic verbal behaviour: reconsidering claims for universality. In Politeness in Language: Studies in its History, Theory and Practice, ed. R. Watts, I. Sachiko, E. Konrad, 155–199. Berlin / New York: Mouton de Gruyter.

CHAPTER FIFTEEN A STUDY ON THE LEXICAL COHESION, MOOD AND MODALITY OF THREE ENGLISH TRANSLATIONS OF PERICLES’ FUNERAL ORATION EFTHYMIA KARAOUZA

Introduction This paper discusses lexical cohesion, mood and modality deployed in three English translations of an Ancient Greek text remaining influential as part of the preamble of the European Constitution. The source text is Pericles’ Funeral Speech (§35-46). It derives from the second book of Thucydides’ History of the Peloponnesian War. This speech was written in 431 BC, the first year of the Peloponnesian War (431-404 BC) between Athens and Sparta. At that time, Pericles, the leader of Athens, was asked to deliver the Funeral Oration for the Athenian soldiers and sailors who had lost their lives during the various small actions of the first year of the war. The English translations which were written by Dale (1863), Crawley (1910) and Warner (1972) have stood the test of time and have been reprinted many times. The purpose of this study is to show that, although these English versions render the same source text, the translators sometimes make different choices in terms of lexical cohesion, mood and modality. Even the two translations, which would be expected to share many similarities, namely Crawley’s and Warner’s, since Warner was influenced by Crawley’s version (cf. France 2000) appear to make dissimilar choices in some cases.

324

Chapter Fifteen

Lexical Cohesion Lexical cohesion entails the frequent use of the same content word or of related words conveying a sense of unity in a text (Toolan 1998, 30). According to Halliday and Hasan (1976, 274–288), this kind of cohesion refers to the semantic relation that can exist between lexical items and encompasses two broad sub-classes: reiteration (residing on the paradigmatic axis) and collocation (residing on the syntagmatic axis). Reiteration involves the repetition of a lexical item, the use of general words, the employment of synonyms, and pairs of superordinates and subordinates. Collocation, namely the cohesion achieved through the association of lexical items that regularly co-occur, incorporates pairs of opposites, as well as pairs of words belonging to the same ordered or unordered series (cf. Carter 1998, 80–83; Hoey 1991, 6–10; also Karaouza 2005 and 2007 for a detailed analysis of Halliday and Hasan’s model). The following figure provides a summary of Halliday and Hasan’s classification (see Figure 15-1): Repetition of a lexical item General nouns REITERATION Synonyms LEXICAL COHESION

Superordinates/Subordinates Opposites COLLOCATION Words drawn from the same ordered or unordered series

Figure 15-1. Halliday and Hasan’s classification of cohesive devices

On the grounds of their analysis of collocation Halliday and Hasan (1976, 285–286) refer to the cohesive effect deriving from the occurrence in proximity of pairs such as laugh and joke, blade and sharp, garden and dig, in addition to the subcategories shown above. However, they admit that the meaning relation between such pairs of words that occur in proximity cannot be classified easily in systematic semantic terms. Hoey (1991, 8) makes a similar point by saying that intuition, which is not a reliable guide, is used to label such pairings and this muddles the study of

Multilingualism and Applied Comparative Linguistics

325

collocation. Carter (1998, 81) generalises this claim and suggests that it is difficult to define the nature of collocational relations in any systematic way. With this in mind, reference to the cohesive effect generated by pairs of words like laugh and joke is avoided in this paper and the analysis is confined to the scrutiny of the two subcategories of collocation that can be classified in systematic semantic terms: opposites and words drawn from the same ordered or unordered series.

Reiteration The repetition of the same word and the use of synonyms Even when the translations use the same word, there are differences. These derive from the personification of an inanimate noun or from the pre-modifiers that appear with the word repeated. Firstly, the repetition of the lexical item hope in the three translations provides a good example of the reappearance of the same word. In paragraph 42, hope is repeated in Dale’s, Crawley’s and Warner’s version. Although the three texts repeat this word, only Warner personifies hope: hands of Hope. Secondly, the repetition of country in paragraph 36 is remarkable; Dale utilises the country, Crawley repeats the country, the mother country, and Warner translates a free country. Crawley uses the pre-modifier mother while Warner uses the pre-modifier free. Thus, Crawley and Warner emphasise different aspects of the country of Athens; Crawley shows the strong relation between Athenians and their country by means of a metaphor (mother country), while Warner underscores the freedom that characterised the citizens of Athens. Hence, even when the same word is repeated in all three translations, the different pre-modifiers illuminate dissimilar aspects of this word and consequently different meaning is conveyed. Also, there are differences as regards synonyms. The following examples from paragraphs 39 and 42 corroborate this argument. In paragraph 39, Dale translates enemies and does not employ synonyms in order to refer to Sparta, which is the antagonist of Athens. Crawley uses antagonists and enemy, and Warner opponents and enemies. In paragraph 42, Dale uses the synonyms conceiving, minded, and regarding to underpin that the sacrifice of the men who fell in the battle was a conscious act, marked by thought, will, design, and perception. Crawley employs the synonyms reckoning and thought. The third translation does not make use of synonyms; Warner utilises only thought. On the whole, two points should be highlighted: i) the translators may use

326

Chapter Fifteen

or avoid synonyms and ii) when the translators use synonyms they employ different lexical items.

Superordinates and subordinates With reference to Athens, all three translations use the superordinate empire which refers to the empire of Athens in paragraph 36. However, Dale translates the great empire as the superordinate of the country and then country as the superordinate of city. Alternatively, Crawley employs the superordinate the empire and then its subordinate country. Lastly, Warner utilises our state as a subordinate of the empire. Table 15-1 illuminates these diverse lexical choices: Table 15-1. Superordinates and subordinates Dale Crawley the great empire the empire the country the country our city -

Warner the empire our state -

Additionally, the three translators provide diverse superordinates of democracy. In paragraph 37, Dale translates form of government as the superordinate of democracy. Warner regards our system of government and constitution as the superordinates of democracy while Crawley gives the pair constitution and democracy. Thus, there are dissimilarities in terms of superordinates and subordinates.

General Nouns Halliday and Hasan (1976, 274–277) use the term ‘general noun’ to refer to a set of nouns such as ‘human noun’, ‘place noun’, and ‘fact noun’. These have generalised reference within the major noun classes. In terms of the three translations, there are dissimilarities in the choice of ‘general nouns’. For instance, Crawley uses the general noun man denoting human for paragraph 37, while Warner translates people. Also, the translators may avoid ‘general nouns’. For instance, Dale does not utilise general nouns included in the category of human for the translation of paragraph 37.

Multilingualism and Applied Comparative Linguistics

327

Collocation In this subclass of lexical cohesion, opposites and words drawn from the same ordered or unordered series are incorporated. As regards the latter, the two recognisable semantic relations in which these pairs of words stand to one another are i) the relation between a part and the whole and ii) the relation between the parts of the whole.

Opposites Thucydides deploys many pairs of words that are opposites. For example, words against deeds, and private against public. Notably, the translators use different words in order to convey these opposite meanings. On the one hand, the pair many/few appears in Dale’s and Crawley’s translations of paragraph 37. Warner, on the other hand, gives the pair minority/the whole people. For paragraph 40, Dale translates domestic/public, the pair private/public occurs in Crawley’s translation and finally Warner writes own affairs/affairs of the state. Moreover, paragraph 43 provides another good example of the diverse lexical choices that the translators make for expressing opposite meanings. A sentence with two pairs of opposites, gave/received and collectively/individually, occurs in Dale’s translation. For the same sentence, the opposites offering/received and in common/individually appear in Crawley’s translation. Finally, Warner uses the pairs of opposites gave/won and to her [the city] and to all of us/own selves What needs to be added at this point is the absence in any translation of a pair of opposites occurring in all the other translations. The pair of opposites word and deed appears in paragraph 46 in Dale’s and Crawley’s translations. This is a frequent pair of opposites in Thucydides’ work. Still Warner does not employ this pair. Therefore, the three translations are sometimes different not only in respect of the words that they use to convey opposite meanings, but also in terms of the translator’s choice to use or avoid a pair of opposites. Furthermore, the pairs of opposites are frequently underscored by parallelism, which is a repeated structure filled with new elements, according to Beaugrande and Dressler (1981, 49). In addition, Galperin (1971, 207) points out that parallelism is used as a technical means in building up antithesis. For example, the structure verb subject + anaphoric sentence + might is reiterated in paragraph 35 in order to emphasise the antithesis between acquainted and unacquainted in the sentences the

328

Chapter Fifteen

hearer who is acquainted with…might and while he who is unacquainted with […] might of Dale’s translation. For the same sentence, Crawley uses: the friend who is familiar […] may/he who is a stranger […] may. The syntax verb subject + anaphoric sentence + may is repeated to emphasise the opposite meanings strange and familiar. Similarly, Warner repeats verb subject + anaphoric sentence + may in the sentences the man who knows […] may /others who do not know […] may. Nonetheless, all translations do not always deploy parallelism for emphasising the same pairs of opposites. In the sentences collectively they gave her their lives, individually they received (§43) the structure adverb + they + verb is replicated and thereby the two pairs of opposites, namely collectively/individually and gave/received are emphasised by Dale. For the same sentences, Crawley does not employ parallelism: his phrases are offering of their lives in common /individually received. Similarly, Warner does not reproduce the same structure in the sentences they gave their lives, to her [the city] and to all of us and for their own selves they won praises conveying the opposite meanings gave/won, and to her and to all of us/for their own selves. In a nutshell, the three translators use different words to render opposite meanings. Also, they do not always employ parallelism in order to articulate opposite meanings.

Words drawn from the same ordered/unordered series Halliday and Hasan (1976, 285) consider pairs of words that are associated as ‘the part to the whole’ or ‘as part to the part’ a subcategory of ‘words drawn from the same ordered/unordered series’ (see Figure 151). Differences in rendering these relations were found in the three translations. In particular, Dale, Warner and Crawley employ different lexical items in order to refer to the Athenian assembly and its parts. Dale uses assembly, citizens and strangers. Crawley renders assemblage and citizens and foreigners. Warner writes assembly and citizens and foreigners as the parts of it. Table 15-2 displays these differences:

Multilingualism and Applied Comparative Linguistics Table 15-2. The part and the whole The whole Dale Assembly Crawley Assemblage Warner Assembly

329

The parts citizens, strangers citizens, foreigners citizens, foreigners

Effects of the different choices in lexical cohesion The preceding analysis raises another issue that should be taken into consideration; what might be the results of the different choices in lexical cohesion? One possible result might be different meanings. This issue will be addressed henceforward. The translations of paragraph 42 and especially the three versions of its last sentence, which has been called by Kennedy (1963, 156) “one of the most emotional and perhaps the finest sentence in Greek prose” can assist me with illustrating that the different choices that the translators make in terms of lexical cohesion can lead to diverse meanings. The three translations are listed below: Dale (1848): And after the shortest crisis, when at the very height of their fortune, were taken away from their glory rather than their fear. Crawley (1874): And after one brief moment, while at the summit of their fortune, escaped, not from their fear, but from their glory. Warner (1954): and in a small moment of time, the climax of their lives, a culmination of glory, not of fear, were swept away from us.

In Dale’s and Crawley’s translations, glory and fear are the causes that led to death, since they are incorporated in the prepositional phrases from their glory rather than their fear (in Dale’s translation) and not from their fear, but from their glory (in Crawley’s translation) denoting cause. Warner makes different choices regarding lexical cohesion by utilising the synonyms climax and culmination that do not appear in the other versions. By means of this reiteration, glory and fear occur in the prepositional phrase post-modifying culmination. In this case, glory and fear are employed to denote the soldiers’ feelings at the moment in which they gave their lives and are not considered the reasons that led to the death of soldiers (like in Dale’s and Crawley’s versions). In addition, it should be noted that although it is reasonable to interpret the forms given by Dale and Crawley as causal, their formulations are obliquely stated. Actually, readers are given the impression that all three translations were trying to convey that the heroes acted in “a state of glory and not of fear”.

330

Chapter Fifteen

However, that is clearly stated only in Warner’s version. In the other two translations it is indirectly stated and readers can infer that glory and fear are also the cause that led to the soldiers’ death. The listing of the following translations can also justify the meaning differentiations analysed above and attributed to the different choices in lexical cohesion. Jowett (1881): and in an instant, at the height of their fortune, they passed away from the scene, not of their fear, but of their glory. Smith (1919): and in the brief instant ordered by fate, at the crowning moment not of fear but of glory, they passed away. Marchant (1961): at the crisis, not of their fear, but of their glory, they passed away. Burn (1966): at the very culmination of glory, rather than of terror, passed away.

In Jowett’s translation fear and glory are incorporated in the prepositional phrases denoting cause, similarly to Dale’s and Crawley’s translations. Notably, Jowett, Dale and Crawley wrote their translations in the same period, namely the 19th century and their chronological relation may account for their resemblance. Smith and Burn do not regard glory and fear as the reasons which led to death but solely as the soldiers’ feelings at the moment of their death since they include the words fear and glory in the prepositional phrase which is coupled with the words crowning moment, crisis and culmination. Thus, these translations share the same meaning with Warner’s translation also written in the 20th century. There is only one translation whose meaning is ambiguous, that of Marchant. In this case, the prepositional phrase not of their fear, but of their glory can be interpreted in two ways: firstly, as a post-modifier of the word crisis and consequently as a description of the soldiers’ feelings at the moment of their death (similarly to Smith’s, Burn’s and Crawley’s translation), and secondly as a prepositional phrase expressing the reasons that caused death (similarly to Dale’s, Crawley’s and Jowett’s versions). The translations listed above confirm that the last sentence of paragraph 42 was interpreted differently in the 19th and 20th centuries. This is also obvious in the translations used for this analysis; Dale and Crawley on the one hand, who wrote their translations in the 19th century, appear to convey a slightly different meaning from Warner who wrote his translation in the 20th century.

Multilingualism and Applied Comparative Linguistics

331

Additionally, in this section on lexical cohesion, it has been argued that the use of different pre-modifiers even when the same word is repeated and the employment or the absence of synonyms can affect the meaning conveyed to the reader. To be more precise, the use of different premodifiers for the word country, which is repeated in all translations, illuminates different aspects and characteristics of Athens. Warner underscores the freedom that describes the Athenian citizens (via the premodifier free) while Crawley refers to the strong relation between the citizens of Athens and their country since Athenians call their country mother country (consult section Opposites). Thus, even when all three translations repeat the same word, they can convey different meanings. Furthermore, Crawley and Dale render the thoughtfulness of the soldiers who fell in the battle via the repetition of synonyms (Dale uses conceiving, minded, and Crawley translates regarding reckoning and thought) denoting consideration of the situation in paragraph 42. Warner, however, does not convey this thoughtfulness since he does not repeat the verba sentiendi. In conclusion, the different choices that the translators make in terms of lexical cohesion can produce different meanings. This issue is also highlighted in Venuti’s work. Venuti (2000, 299), however, widens the scope of his study to all the cohesive markers, that is to say he does not confine his argument in lexical cohesion, and believes that shifts in the cohesive markers of the translations lead to shifts in explicitness and in text meaning as well.

Mood and modality In the previous sections, it has been argued that sometimes the three translators make different choices in lexical cohesion. That is to say that in some cases the three versions are dissimilar in terms of the textual function of language, which refers to the creation of relevance to context, according to Halliday’s model of the language (meta)functions (1994, 35– 36). In this section, the three versions will be examined in respect of another function, named interpersonal. This involves the speaker’s/writer’s statements, questions, demands and offers as well as his/her opinions expressed in these speech functions. More specifically, interpersonal meaning deals with both mood and modality. At this point, it might be useful to draw a distinction between the two, although there are some problems concerning the terminology, the nature of the mood and the way it relates to modality. Mood on the one hand can be regarded as the grammatical realisation of statements,

332

Chapter Fifteen

questions, demands and offers in the main sentence types, namely declarative, interrogative and imperative sentences (Fairclough 2003, 105). That is to say, mood refers to the relation between the speaker/writer who gives commands or offers ‘goods and services’ (or information) and the listener/reader who receives these commands or offers of ‘goods and services’ (or information), according to Halliday (1994, 69). Modality on the other hand is defined as “the speaker’s judgement of the probabilities, or the obligations, involved in what he is saying” (ibid., 75) or simply as the “speaker’s opinion or attitude” (Lyons 1977, 452). Thus, modality denotes the speaker’s/writer’s attitude towards his/her statements, questions, demands and offers and is expressed by modal verbs and adverbs, evaluative verbs, adjectives and adverbs (mood and modality are further illuminated in Table 15-3). It can be inferred from the above analysis of mood and modality that i) if speakers/writers make different choices in mood then they can either demand or give ‘goods and services’ (or information) to the listeners/readers and ii) if they make diverse modality choices then they will express disparate opinions and attitudes towards their propositions. With reference to the three versions, it will be shown that Pericles seems to adopt dissimilar attitudes towards his propositions as well as towards his audience by virtue of the different choices in mood and in modality. Table 15-3. Mood and Modality Realises

Relates

Mood

statements, questions, demands, offers

speaker/writer to listener/reader

Modality

speaker’s / writer’s attitude

speaker/writer to propositions

Expressed by declarative, interrogative and imperative sentences modal verbs modal adverbs evaluative verbs, adjectives and adverbs

Mood In this section it will be shown that the three versions appear to be dissimilar in terms of mood, namely the way (demand or offer) by which the speaker is related to the listener. An example provides the end of paragraph 46, where the relationship between Pericles and the Athenian citizens is portrayed. For instance, Dale uses the imperative depart, Crawley translates you may depart and Warner you must depart. Thus, two

Multilingualism and Applied Comparative Linguistics

333

different sentence types are used: imperative by Dale and declaratives by the other two translators. Pericles in Dale demands ‘goods and services’ since he utilises the imperative, which is the strongest type of directive, the most confident and direct, frequently employed by a superior when ordering an inferior (Palmer 1986, 29). Pericles therefore appears to have a relationship of unequal power with his audience. On the contrary, in Crawley’s and Warner’s translations, Pericles appears to use declarative sentences. In other words, he offers information and does not demand ‘goods and services’ (Halliday 1994, 69). Thus, in the last sentence of paragraph 46 Pericles seems to share a relationship of equal power with his audience. The above analysis shows that there are dissimilarities in terms of mood. These differences can be further illuminated by the reference to another feature of the declarative sentences, namely modality. More specifically, modal verbs (must and may) are employed in the declarative sentences in order to show that the information offered deals with obligation and in particular different degree of obligation; Must expresses high degree of compulsion while may is more polite and denotes a lower degree of obligation (Halliday 1994, 75). Dale therefore seems to regard Pericles as the leader of the Athenians who has full authority over the audience and subsequently he can order it. Warner appears not to consider that Pericles has so much authority so as to order the Athenians but he can still require them to depart, while in Crawley’s translation Pericles seems not to require or demand the audience’s departure but to give his permission for them to leave. Thus, the obligation expressed in the three translations reaches different degrees of compulsion, and shows different attitude on the part of Pericles towards his audience. Furthermore, this graduation of obligation bears similarities to White’s (2006) argument that another aspect of modality, namely evaluation, comes on a ‘scale of intensity’ since evaluative verbs, adjectives and adverbs gather in semantic sets of terms which extend from low to high intensity as in the set of like/love/adore (consult Table 15-4). Table 15-4. Pericles’ attitude towards the audience/degrees of obligation Pericles’ relation to the audience Unequal Equal Order Requirement Permission Depart (Dale) You must depart (Warner) You may depart (Crawley)

334

Chapter Fifteen

Modality In this section, the analysis will focus on the modality in the three versions and especially on the modality categories in which the three translators made dissimilar choices and consequently conveyed different attitudes on the part of Pericles. These categories are: modality denoting obligation, probability and wishes, modality showing usuality and finally modality expressing evaluation.

Modality denoting obligation, probability, and wishes Before proceeding to the discussion of modality expressing obligation, probability and wishes, reference to the terminology adopted by several researchers as regards these categories of modality is needed, so as to clarify and at the same time justify the terms that are used for the scrutiny of the three versions. Actually, modality expressing the degree of obligation attaching to the performance of certain actions has been already mentioned. This modality is called deontic according to Simpson’s model (1993, 46–49) and is one of the three types of modality. The other two are epistemic and boulomaic. Epistemic denotes the speaker’s confidence or lack of confidence in the truth of the proposition expressed and boulomaic indicates the speaker’s/writer’s desires and wishes. This model bears similarities to the modality system of other researchers. For instance, before Simpson, as early as 1976, Hofmann refers to epistemic modality but avoids distinguishing between boulomaic and deontic modality. He incorporates these two kinds of modality in one category, namely ‘root modality’, expressing duties, along with wishes and desires. Ten years later, in 1986, Palmer argued that there are three types of modality: deontic, epistemic and dynamic. Hence, Palmer utilises a different term from that of Simpson for the modality expressing wishes: dynamic modality. Additionally, he includes in this category modality denoting ability. In this study, Simpson’s and Palmer’s categories are deployed. Epistemic, deontic and boulomaic modality are frequently expressed by the use of modal verbs and adverbs or other devices denoting probability, obligation as well as willingness similarly to the modal verbs. In this study, we analyse the expression of obligation and probability in the next section. After that, we turn to the denotation of wishes.

Multilingualism and Applied Comparative Linguistics

335

Obligation and probability via modal verbs In the section about Mood, the contribution of the modal verbs may and must was integral for corroborating the argument about the ‘scale’ of obligation used in the imperative and declarative sentences. At this point modal verbs will be the focus of my analysis. Modal verbs are also known as modal auxiliaries and are a small group of verbs, which can appear to the left of the ‘main’ or ‘content’ verbal word, according to Toolan (1998, 48) or in the grammatical ‘mood’ according to Halliday’s analysis of the clause parts (1994, 71–75). Can, could, may, might, must, should, will, would and ought to can be used to express probability, obligation and willingness. The focus will firstly be modal verbs denoting only obligation and probability and used by the three translators in order to convey Pericles’ attitudes towards i) the praises given to the dead in the battle, ii) the Athenian qualities, iii) the future actions of the survivors of the battle and iv) the sons and the brothers of the dead. At the beginning of the speech (§36), Pericles communicates his attitude towards the tradition of paying homage to the soldiers who died in the battle. The translations are given below: Dale: for it is just, and becoming too at the same time, that on such an occasion the honor of being thus mentioned should be paid them. Crawley: it is both just and proper that they should have the honour of the first mention on an occasion like the present. Warner: I shall begin by speaking about our ancestors, since it is only right and proper on such an occasion to pay them the honour of recalling what they did.

Dale and Crawley employ deontic modality denoting high degree of obligation (Halliday 1994, 75): should be paid and they should have honour respectively. Hence, Pericles appears to feel obliged to honour the dead. Warner, however does not position Pericles towards his speech in that way, since deontic modality is not utilised. Shall which is unmodalised with pure future reference (cf. the verb will in Toolan 1998, 49; Palmer 1986, 62) is employed instead. In Warner’s version therefore Pericles seems to agree with the praise of the dead and proceeds to his eulogy. Additionally, Pericles refers to the points in which Athens is worthy of admiration (§39). Deontic modality expressing high degree of obligation (Halliday 1994, 75) is employed in Crawley’s and Warner’s translations:

336

Chapter Fifteen

We have to attend, and we do not have to spend respectively. In these translations the speaker estimates that the Athenians are obliged to act in a certain way. Deontic modality is not deployed in Dale’s translation for the description of the Athenian qualities. The readers may thereby infer that Pericles’ opinion is that the Athenians act in an admirable way in their own free will and not because they are obliged. In paragraph 41, Pericles conveys his opinion about the future actions of the survivors of the battle. For the translation of this paragraph, Warner writes that everyone of us who survive them should be willing to undergo hardships and employs deontic modality (should), expressing high degree of obligation (ibid.) along with dynamic (Palmer 1986, 102) or boulomaic modality (Simpson 1993, 48). It can be inferred from Warner’s translation that Pericles requires the survivors to undergo hardships for Athens and attempts to form their wills. Nevertheless, this is not the case for the other two translations. Dale writes may well be willing and Crawley may every one of their survivors be ready to suffer. Therefore epistemic and not deontic modality is coupled with modality denoting willingness. Thereby, Pericles does not require that the survivors sacrifice themselves for Athens; he considers it probable yet not obligatory. Thus, Pericles appears to have different attitude towards the future actions of the survivors in the three versions. Furthermore, the Athenian leader does not confine his speech to actions that should be undertaken by the survivors of the battle but broadens it in scope to position himself towards the future actions of the relatives of the dead soldiers. In paragraph 45, Pericles seems to remind the sons and the brothers of the dead about their duties by utilising the modal verb must expressing high degree of obligation (Halliday 1994, 75) in Dale’s and Crawley’s translations. This is not the case for Warner’s translation, where deontic modality is not used. Therein Pericles sympathises with the relatives and understands the difficulties that they experience rather than emphasising their duties. All the above examples denote that Pericles appears in three versions to adopt dissimilar attitudes towards some of his propositions due to different translators’ choices of modal verbs expressing deontic and epistemic modality.

Modality denoting wishes Warner and Dale use boulomaic or dynamic modality for the translation of paragraph 36. Dale translates as I do not wish to enlarge and Warner employs modality denoting Pericles’ wishes twice: I have no wish

Multilingualism and Applied Comparative Linguistics

337

to make a long speech and what I want to do is. By dint of this use of ‘volitive’ verbs (Palmer 1986, 152) Pericles appears to express solely wishes and intentions and consequently avoids committing himself to his statement that he won’t enlarge on the Athenian military achievements. In Crawley’s version nevertheless Pericles is portrayed as certain about this decision by the use of shall denoting real and expected future events to which Pericles appears committed. In a nutshell, Pericles’ commitment to his decision not to dwell upon the military accomplishments of the Athenians seems to graduate in the three translations. In other words, it comes on a ‘scale of intensity’. This graduation is shown in Table 15-5. Table 15-5. Degree of commitment in Pericles’ statement Degree of commitment Translations High I shall pass it by (real and expected future events) (Crawley) I do not wish to enlarge (Dale) Low (intention, wish) I have no wish / what I want to do is (Warner)

Modality denoting usuality: Modal adverbs This section deals with modal adverbs expressing usuality. Halliday (1994, 81–82) as well as Butt and his colleagues (2000, 113–115) include modal adverbs in the category named ‘mood adjuncts’ since they regard them as a part of what they call ‘mood’ or ‘mood block’. However, the use of the word ‘mood’ raises problems since it has already been said that mood refers to declarative, imperative and interrogative sentences. Modal adverbs express the speaker’s attitude towards the offers or demands of ‘services and goods’ (or information) and consequently belong to modality and not to mood. Due to this, the term modal adverbs is adopted, instead of ‘mood adjuncts’ that would perplex the study of modality (consult also the section Modality denoting evaluation: evaluative devices for the ‘mood adjuncts’). To return to the three English versions of the Funeral Oration, in paragraphs 39 and 40 Pericles is concerned with the description of the Athenian qualities. For these paragraphs two different choices in terms of modality expressing usuality were made by the translators: i) in some cases, adverbs denoting usuality and therefore portraying Pericles as confident about the permanent character of the Athenian capacities are utilised, ii) while in other cases, these are avoided and consequently

338

Chapter Fifteen

Pericles does not appear so certain that Athenians always act in the same commendable way. To be more precise, Dale and Crawley employ the modal verb never in paragraph 39. By this means, Pericles is absolute that Athenians always show respect to strangers and never prevent them from learning things about Athens. This is not the case for Warner’s version; there never is avoided and Pericles does not emphasise the permanent character of this Athenian attitude. The three translations are provided below: Dale: We throw our city open to all, and never, by the expulsion of strangers, exclude anyone from either learning or observing things. Crawley: We throw open our city to the world, and never by alien acts exclude foreigners from any opportunity of learning or observing; Warner: Our city is open to the world, and we have no periodical deportations in order to prevent people observing or finding out secrets

A similar instance occurs in paragraph 40, where Crawley alone deploys modal adverbs denoting usuality: again, usually, never. Thereby it can be inferred that Pericles claims that the Athenian braveness is everlasting. The other two translations do not appear to share this view since modal adverbs denoting usuality are not employed.

Modality denoting evaluation: evaluative devices Verbs, adjectives and adverbs expressing the text-speaker’s stance towards the material he/she relates fall under the definition of evaluative devices according to Toolan (1998, 56). As regards evaluative adverbs, it had been already pointed out in an earlier study carried out by Halliday (1994, 81–83) that some adverbs, which he had named ‘comment adjuncts’ and incorporated in ‘modal adjuncts’ along with ‘mood adjuncts’ express evaluation. Although Halliday admits that there is no clear line between ‘comment’ and ‘mood adjuncts’ he holds ‘that comment adjuncts are less closely tied to the grammar of the mood’ than ‘mood adjuncts’ are. We share this view and for this reason we have examined modal adverbs (or ‘mood adjuncts’) denoting usuality separately, in section Modality denoting usuality: Modal adverbs because the modal adverbs (or ‘mood’ adjuncts) usually and never are more closely tied to the grammatical mood than the ‘comment adjunct’ perfectly utilised for the following analysis of the evaluative adverbs and adjectives. The distinction drawn by Halliday

Multilingualism and Applied Comparative Linguistics

339

between ‘mood’ and ‘comment adjuncts’ is further illuminated in Figure 15-2: mood adjuncts modal adjuncts comment adjuncts Figure 15-2. Halliday’s (1994) categorisation of modal adjuncts

Modality denoting evaluation Evaluative verbs are deployed in the three versions a number of times in order to convey Pericles’ opinion. At the beginning of paragraph 41, the three translators use three different verbs in order to convey Pericles’ attitude towards the statement which is considered to be one of the most essential in Thucydides’ work; that Athens is the school of Hellas (Harding 1973, 56). For this sentence, Warner translates I declare, Crawley I doubt and Dale I say. Palmer (1986, 142–145) considers that the verbs say and declare (used by Warner and Crawley) express strong assertion and total positive commitment to the proposition in contrast to the verb doubt (selected by Dale) which is non-assertive denoting partial negative commitment. Consequently, in Dale’s version Pericles is positioned differently from the other translations; he is only partially committed to his statement that Athens is the school of Hellas. Additionally, the three translators convey different attitudes on the part of Pericles towards the soldiers and their actions in the battle (§42) by the use of different evaluative verbs. Dale translates resolving, Crawley determined and Warner thought. Determined denotes commitment on the part of the verb subject (the dead soldiers) since it is clearly indicated that action will take place. Resolving expresses commitment, similarly to determined but at the same time denotes deliberation (cf. the verb decide in Palmer 1986, 130). The verb thought, however, selected by Warner conveys only the deliberation of the soldiers’ actions and not their determination. Thus, the three translators position Pericles differently towards the soldiers’ determination in the battle by choosing dissimilar verbs to describe the soldiers’ thoughts. Hence, Pericles sometimes appears to be positioned differently towards his propositions due to the dissimilar choices of evaluative verbs. At this point of the analysis, we will be concerned with the different choices that the three translators made in terms of evaluative adverbs and

340

Chapter Fifteen

adjectives. My argument this time will be twofold: firstly, in some cases Pericles seems to evade evaluative adverbs and adjectives denoting his personal positive evaluation of Athens and secondly when evaluative adverbs and adjectives are used in the three translations these can come on ‘a scale of intensity’ and consequently denote different attitudes as regards the degree of the support that Pericles appears to offer to the city of Athens. I will corroborate my first claim by referring to the evaluative adverbs and adjectives utilised in paragraph 36. There, Pericles refers to the ancestors’ accomplishments in order to show to Athenians how glorious their past is. He thereby encourages them for the future battles. Although Pericles appears in the three translations very confident about the achievements of the past, he seems to employ his personal evaluation to reinforce this confidence only in two versions: in Warner’s translation certainly, perfectly well and great are used to show Pericles’ opinion. Alternatively, Dale translates great and painful. In these two translations, Pericles’ attitude towards Athens appears to be significantly supportive. On the contrary, in Crawley’s translation Pericles describes the achievements of the ancestors in order to give confidence to Athenians but at the same time avoids evaluating them. He thereby expresses a more neutral attitude than in the other two translations. The second claim will be consolidated by the examination of the evaluative adverbs and adjectives employed in the translation of paragraph 40. In particular, Crawley writes worthy of admiration, singular. Alternatively, Warner translates unique and Dale translates in a remarkable degree. Thus, Athens is evaluated as worthy and remarkable and then as unique and singular. That is to say Pericles’ attitude varies in the three translations. Table 15-6. The ‘Scale of intensity’ (White 2006) Scale of intensity Translations Medium Worthy (Crawley) High Unique (Warner)

Remarkable (Dale) Singular (Crawley)

The above analysis clarifies that the different choices made by the three translators in the realm of modality called evaluative adverbs and adjectives have as a result the different positioning of Pericles towards the Athenians. These different attitudes can be seen from two perspectives: firstly, from the different degree of positive evaluation and secondly from Pericles’ choice not to use evaluative adverbs and adjectives and confine himself to the description of the Athenian capacities without adding his own evaluation. In this last case, Pericles is depicted as distanced from his

Multilingualism and Applied Comparative Linguistics

341

own propositions attempting to give the impression to his hearers that what he describes is the truth and not his personal estimations and probably exaggerations about the city that he leads, namely Athens. With this section the argumentation about the diverse choices that the three translators make as regards mood and modality has been completed. We have shown that the different choices made in mood reflect different relations—of equal or unequal power—between Pericles and his audience while the dissimilar choices in modality reveal dissimilar attitudes on the part of Pericles towards his propositions. These findings become more valuable if we take into consideration that Crawley’s translation was expected to share similarities with Warner’s, since Warner admits that he was influenced by Crawley’s version. It has been shown that the two versions usually (except for the use of modality denoting obligation in paragraph 39 and of modality denoting wishes in paragraph 36) appear to be dissimilar and consequently they convey diverse attitudes on the part of Pericles towards his propositions.

Conclusion The purpose of this study was to analyse the differences in the lexical cohesion as well as in the mood and modality of the three English translations of Pericles’ Funeral Oration. The conclusions drawn from this examination will be outlined in this final section. In the first section, it has been shown that although the three translators utilise the same source text, differences occasionally appear in their lexical cohesion (textual function). The three translators have the choice to utilise or avoid lexical cohesion for rendering same parts of the source text. When all three translators employ lexical cohesion, there are dissimilarities in the use of i) the same word ii) superordinates/subordinates, iii) general nouns, iv) synonyms, v) opposites and vi) words drawn from the same ordered or unordered series. Additionally, it was shown that although Warner and Crawley were expected to share many similarities in terms of lexical choices since Warner had admitted that he was influenced by Crawley’s translation, they sometimes make dissimilar lexical choices. The differences in the lexical cohesion could entail dissimilarities in meaning. Actually, the diverse translations of the last sentence of paragraph 42 justified this claim. More specifically, it was illustrated that the translators’ decision to utilise or avoid synonyms resulted in two different meanings. Additionally, it was shown that these differences in lexical choices and consequently in meanings are due to the two different interpretations given to the Ancient Greek text in the 19th and 20th century.

342

Chapter Fifteen

In particular, Dale and Crawley (19th century) had interpreted the source text differently form Warner (20th century). Moreover, it has been shown that different choices of pre-modifiers coupled with words repeated, as well as the repetition or evasion of synonyms can also lead to different meanings. The analysis of lexical cohesion was followed by the scrutiny of mood and modality. Dissimilar mood choices were made at the end of paragraph 46 and it was also argued that the translators made different choices in modality denoting obligation, probability, wishes, usuality as well as evaluation. As regards the two versions that were expected to share similarities—Crawley’s and Warner’s—, although they appear to make the same choices in mood at the end of paragraph 46, they usually make diverse choices in terms of modality (except for the choices of modality denoting obligation (§39), and modality denoting wishes (§36)). In addition, it has been shown that by virtue of the diverse choices in mood and modality, Pericles appears to adopt dissimilar attitudes to his propositions and audience. More specifically, in Warner’s and Crawley’s translations Pericles seems to be less imperial than in Dale’s version. Additionally, the Athenian army is obliquely characterised by Pericles as somewhat less heroic in Warner’s version. Finally, Pericles is portrayed as slightly less supportive of the city of Athens in Crawley’s translation (consult Table 15-7). Table 15-7. The three versions Translations Dale Pericles Slightly more Pericles as a leader imperial Pericles’ attitude Slightly more towards the heroic Athenian army Pericles’ attitude towards the Slightly more proposition that supportive Athens is the school of Hellas

Crawley

Warner

Slightly less imperial

Slightly less imperial

Slightly more heroic

Slightly less heroic

Slightly less supportive

Slightly more supportive

These dissimilarities can be attributed to the diverse ways in which the three translators interpreted the source text. This conclusion bears similarities with the argument that the dissimilar lexical choices are due to different interpretations.

Multilingualism and Applied Comparative Linguistics

343

Acknowledgement This article has been inspired during my MA studies at the University of Birmingham. Many thanks are due to Professor Michael Toolan for his invaluable comments on previous drafts.

References Beaugrande de, R. and W. Dressler. 1981. Introduction to Text Linguistics. London: Longman. Burn, A.R. 1966. Pericles and Athens. New York: Crowell-Collier. Butt D., R. Fahey, S. Feez, S. Spinks, and C. Yallop. 2000. Using Functional Grammar: An explorer’s Guide. (2nd edition) Sydney: Macquarie University. Carter, R. 1998. Vocabulary: Applied Linguistic Perspectives. London: Routledge. Crawley, R. 1910. History of the Peloponnesian war done into English by Richard Crawley. London: J.M. Dent. Dale, H. 1863. Thucydides: The history of the Peloponnesian War: a new and literal version, from the text of Arnold. New York: Harper. (Harper's classical library). Fairclough, N. 2003. Analysing discourse: Textual Analysis for social research. London: Routledge. France, P., ed.. 2000. The Oxford Guide to Literature in English Translation. Oxford: Oxford University Press. Galperin, I.R. 1971. Stylistics. Moscow: Higher School Publishing House. Halliday, M.A.K. 1994. Functional Grammar. (2nd edition) London: Arnold. Halliday, M.A.K., and Hasan, R. 1976. Cohesion in English. London: Longman. Harding, H.F. 1973. The speeches of Thucydides, with a general introduction and introductions of the main speeches and military Harangues. Lawrence: Coronado Press. Hoey, M. 1991. Patterns of Lexis in Text. Oxford: Oxford University Press. Hofmann, T.R. 1976. Past tense replacement and the modal system. In Syntax and Semantics 7. Notes from the linguistic underground, ed. J.D. Mc Cawey, 85–100. New York: Academic Press. Jowett, B. 1881. Thucydides translated into English with introduction, marginal analysis, notes and indices. Oxford: Clarendon Press.

344

Chapter Fifteen

Ȁaraouza, E. 2005. The cohesive use of the definite article in Ancient Greek prose texts of the 5th and 4th centuries BC. Article presented at the 7th International Conference on Greek Linguistics, September 8– 10, in York, UK. —. 2007. Cohesion and Text Structure in Attic Greek prose. Ph.D diss. University of Birmingham. Kennedy, G.A. 1963. The Art of Persuasion in Greece. Princeton: Princeton University Press. Lyons, J. 1977. Semantics. Cambridge: Cambridge University Press. Marchant, E.C. 1961. Thucydides Book II. London: Macmillan and Co. Palmer, F.R 1986. Mood and Modality. London: Cambridge University Press. Simpson, P. 1993. Language, Ideology and point of view. London: Routledge. Smith, C.F. 1919. Thucydides with an English translation by C.Foster Smith. History of the Peloponnesian War. Vol.I Books I-II. London: Heinemann. Toolan, M.J. 1998. Language in Literature. London: Arnold. Venuti, L. 2000. The translation Studies Reader. London: Routledge. Warner, R. 1972. History of the Peloponnesian war translated by Rex Warner with an introduction and notes by M.I. Finley. Harmondsworth: Penguin Books. (Penguin Classics). White, P. 2006. An introductory tour through appraisal theory. http://www.grammatics.com/appraisal/.

CONTRIBUTORS

Hee Sook Bae is an instructor of terminology and terminography at the University of Montreal. She currently researches the application of the French semantic annotation theoretical model to Korean with the OLST (Observatoire de linguistique Sens-Texte). She is interested in terminology, lexicology and lexical semantic networks in specific domains. She worked at Delphes Technology Int. as a linguist on the French content modeling for a knowledge based search engine. She also participated in the construction of CoreNet as an invited professor at KorTerm/KAIST. She has published in journals such as Terminology and Language and Information. Andreas Braun works as a research assistant within the Faculty of Health and Social Sciences at the University of Bedfordshire, UK. He has a background in linguistics, EFL and cultural studies. His main research interests incorporate English and German language studies, multiculturalism and multilingualism with the focus on second and third language acquisition and language maintenance. His doctoral thesis concerned parents’ reports on their use of languages and cultural traditions with their children in trilingual families. Vineeta Chand is a PhD candidate in Linguistics at the University of California at Davis. Her research interests include sociocultural linguistics, cognitive grammar, language disorders and issues surrounding crossdialect intelligibility in India. Her dissertation research focuses on English in the postcolonial setting of India, examining the relationship between structural variation and social ideologies, to better understand and link processes and features of dialect emergence and identity formation within the multilingual milleux of New Delhi. Maria Sabaté I Dalmau has conducted research on intercultural communication, pragmatics, and the acquisition of politeness systems in foreign languages in multilingual contexts. In the field of ethnography, she investigates language practices in Catalan and Spanish, as well as identity and languages ideologies in immigration contexts in Catalonia. She has been a junior lecturer at the Universitat Autònoma de Barcelona, and has

346

Contributors

an MA in Linguistic Anthropology from the University of Toronto. She is currently a PhD candidate in the English Philology Department at the UAB. She has participated in various congresses and has published her research nationally and internationally. Sylvain Dieltjens teaches French and business communication at the Department of Applied Linguistics of the Lessius University College (Antwerp–Belgium) and is head of the board of the communication agency TxT-Ibis nv, Kortenberg (www.txt-ibis.be). His research focuses on business communication, mainly on linguistic and strategic aspects of bilingual (Dutch/French) internal business documents. He has published several articles in journals such as Ad Rem, Document Design, The Journal of Technical Writing & Communication, Onze Taal and Tekst[blad] and presented papers on similar subjects at conferences organized by Anéla, the Association for Business Communication, Cerlis, Document Design, VIOT, etc. Pedro Guijarro-Fuentes is Senior Lecturer in Spanish and Head of the Spanish Section at the University of Plymouth where he teaches Spanish and area studies. His research interests are in the interdisciplinary field of Spanish Applied Linguistics, SLA, sociolinguistics, psycholinguistics, and bilingualism. His main current research projects include the study of copula verbs ser/estar and the acquisition of spatial prepositions by L2 learners of both English and Spanish. He has authored numerous articles and book chapters, and his research has appeared in different international, refereed journals such Language Learning, International Journal of Applied Linguistics, Spanish in Context and Hispanic Research Journal. Kimberly L. Geeslin is an associate professor and the Director of Hispanic Linguistics at Indiana University where she teaches advanced undergraduate and graduate courses in second language acquisition and Hispanic linguistics. Her research focuses on Spanish as a second language and the intersection of second language acquisition and sociolinguistics. She has published a monograph with Lincom Europa press and her research articles have appeared in Studies in Second Language Acquisition, Language Learning, Hispania, Spanish in Context, Linguistics and Revista espanola de linguistica aplicada as well as in edited volumes published by Cascadilla Press, John Benjamins, Peter Lang and Arco Libros.

Multilingualism and Applied Comparative Linguistics

347

Federico Gobbo teaches computing, philosophy, interlinguistics and esperantology at the University of Insubria (Varese, Italy). He is book review editor of Language Problems, Language Planning for Europe. His research interests lie both in computer science and linguistics. He is currently writing his PhD major dissertation in mathematical linguistics. Hortènsia Curell I Gotor is Senior Lecturer in English linguistics at the Universitat Autònoma de Barcelona, where she teaches a variety of subjects: English Descriptive Grammar, Functional Grammar and Contrastive Linguistics to undergraduates, and Language and Cognition in the PhD programme. She has worked in contrastive linguistics for many years. She has also studied the temporal interpretation of narrative texts, both in English and cross-linguistically. She is currently working on the cognitive basis of linguistic structure, researching the correlation between differences in language structures and underlying categorization patterns. Priscilla Heynderickx teaches Dutch and business communication at the Department of Applied Linguistics of the Lessius University College (Antwerp–Belgium) and is chief editor of Ad Rem, a journal of business communication (http://www.vvvz.be/AdRem). Her research focuses on business communication, mainly on linguistic and strategic aspects of bilingual (Dutch/French) internal business documents. She has published several articles in journals such as Ad Rem, Document Design, The Journal of Technical Writing & Communication, Onze Taal and Tekst[blad] and presented papers on similar subjects at conferences organized by Anéla, the Association for Business Communication, Cerlis, Document Design, VIOT, etc. Marlén Izquierdo teaches ESP at the University of Cantabria, Spain. Her research interests include English-Spanish contrastive linguistics and translation, following a corpus-based, functional approach. She has greatly contributed to the compilation of P-ACTRES, one of the first EnglishSpanish Parallel Corpora compiled for cross-linguistic research on grammatical and rhetorical issues. She is an active member of the ACTRES Research Group (http://actres.unileon.es/). Marlén Izquierdo has participated in various international conferences, done research in foreign institutions with renowned scholars and published in journals of applied linguistics and humanities.

348

Contributors

Efthymia Karaouza is a systemic linguist (PhD, MA Literary Linguistics), with an extensive background in classics (BA Classics). Her research interests involve the application of text linguistics upon Ancient Greek prose, historiography, and philosophy from the 5th century BCE onwards. Her lecturing experience involves teaching undergraduate modules such as English linguistics (UCE Birmingham), Ancient Greek Language and culture, Modern Greek Language, Research Methods in Linguistics, Textual Analysis, Narrative Analysis, and Visual Analysis (University of Birmingham). She has published in collective series of monographs and in other conference proceedings. Koen Kerremans is a PhD researcher at the 'Centrum voor Vaktaal en Communicatie', a research centre at the Erasmus University College Brussels (Department of Applied Linguistics). His main research interests range from ontologies and knowledge management to translation, special languages and terminology. His work has appeared in journals such as Terminology and Linguistica Antverpiensia. Belén Labrador teaches EFL and English lexico-grammar at the University of León (Spain). Her research interests include corpus-based linguistics, contrastive and translation studies and their application to the teaching of English lexico-grammar and translation. She has published in journals such as Meta, Babel, Target, Literary and Linguistic Computing and Languages in Contrast as well as in collective volumes published by the University of León. Marie-Claude L’Homme is full professor at the Department of Linguistics and Translation at the University of Montreal. She teaches terminology and computer tools for translators. She is a member of the Observatoire de Linguistique Sens-Texte (OLST), a multidisciplinary research group that develops theoretical models and resources in lexicology and terminology. Her research interests include terminology, lexical semantics, and computational terminology. Marie-Claude L'Homme is also co-editor of the journal Terminology. She is the author of textbooks in terminology and computer-assisted translation and has published in journals such as Terminology, Cahiers de lexicologie, International Journal of Lexicography. Elizabeth Marshman is an Assistant Professor at the University of Ottawa School of Translation and Interpretation and a member of the Observatoire de linguistique Sens-Texte. She received her doctorate from

Multilingualism and Applied Comparative Linguistics

349

the Département de linguistique et de traduction of the Université de Montréal in 2007, after obtaining her M.A. and B.A. in translation from the University of Ottawa (1997, 2002). Her doctoral research focused on the interlinguistic comparison of conceptual relation markers in medical texts. Her research interests also include computer tools for corpus analysis, terminology, terminography and translation. Carme Bach Martorell is a professor of Catalan Linguistics at the University of Pompeu Fabra and a researcher of the consolidated research group IULATERM at Pompeu Fabra University (Barcelona, Spain). Her research focuses on General and Specialized Discourse Analysis, Lexicography and Corpus Linguistics. Her publications deal with connectives, reformulation and its importance in the process of specialized discourse construction and in extraction of semantic information of specialized discourse. Her PhD thesis about reformulation markers includes a lexicographical implementation prototype for these units. Rosa Rabadán teaches at the Department of Modern Languages at the University of León in Spain. Her research interests range from translation theory and linguistic applications to corpus-based grammar and lexicology. She is a member of the international research team ACTRES (Contrastive Analysis and Translation English-Spanish) and has published in journals such as Meta, Babel, Target, Languages in Contrast, International Journal of Corpus Linguistics and Literary and Linguistic Computing. Presently she focuses on theoretical issues and the design of linguistic applications for translating and Translation Quality Assessment. Noelia Ramón teaches at the Department of Modern Languages at the University of León in Spain. Her research interests range from translation theory and linguistic applications to corpus-based grammar and lexicology. She is a member of the international research team ACTRES (Contrastive Analysis and Translation English-Spanish) and has published in journals such as Meta, Babel, Target, Languages in Contrast, International Journal of Corpus Linguistics and Literary and Linguistic Computing. Presently she focuses on theoretical issues and the design of linguistic applications for translating and Translation Quality Assessment. Dave Sinardet is a post-doctoral researcher (funded by the National Fund for Scientific Research, FWO) and lecturer at the Faculty of Political and Social Sciences of the University of Antwerp. He is also visiting lecturer at the Facultés Universitaires Saint-Louis in Brussels. His research

350

Contributors

mainly concentrates on national identity, nationalism and federalism in plural societies (particularly applied to the Belgian context), as well as on the role of political and media discourse in these processes. His PhD dealt with the mutual mediarepresentation of Belgium' s two main language communities and more generally with Flemish and French-speaking media's role in the Belgian federal model. Martina Temmerman is co-ordinator of the Master in Journalism at the Erasmus University College Brussels and lecturer in professional communication at the University of Antwerp. She teaches discourse analysis and conversation analysis, as well as writing courses. Her main research interest goes to the linguistic and stylistic analysis of journalistic interviews and (opinion) articles. Mª Pilar González Vera teaches English language at the Education Faculty Zaragoza (Spain). Her research interests include audiovisual translation, literary translation, cultural linguistics and children’s literature. She has been involved in a research programme which investigated the reception of Swift’s works in Spain and their translations into Spanish. She has published in journals such as RILA.

SUBJECT INDEX addressee axis, 304 Algol, 26 assertive question, 121 Ausbausprache, 29, 32 auxiliary language, 24, 27, 29, 38 bilingualism, 45, 83, 87, 88, 89, 90, 100, 103, 144, 298 bystander axis, 305 cascade question, 118, 125, 137 categorisation framework, 193, 194, 195, 196, 198, 199 chi-square, 96 circumstance adverbial, 256, 258, 259, 260, 261, 262, 266, 267, 274, 275 cognate, 162, 166, 167, 168, 169, 170, 172, 176, 177, 179, 181, 182 Cognitive Grammar, 19 collocation, 229, 248, 324 communication, 25, 26, 29, 38, 46, 47, 65, 83, 102, 127, 134, 139, 144, 150, 181, 206, 256, 258, 309, 313 Communicative Theory of Terminology, 202 community language, 47, 49, 51 comparable corpus, 273, 278, 280, 281, 282, 283, 284, 287 complex question, 118, 137 conceptualization, 2, 8, 13, 15, 202, 203, 204, 205, 216 concordance, 16, 17, 18, 21, 166, 196, 228, 232, 235, 238, 239, 261, 275, 279, 282, 283, 284, 286, 291, 298

concordancer, 235 Contrastive Analysis, 257, 275, 276, 278, 279, 284, 298 conversation analysis, 110, 114 conversational-contract view, 303, 307, 308 coordination, 267 copula, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 101, 103, 104 Corpus Linguistics, 257 corpus planning, 25, 27 Corpus Workbench Browser, 285 cross-cultural communication, 25, 26, 27, 38, 45 cross-linguistic metaphoric analysis, 21 diachronic research, 15 dictionary, 6, 187, 198, 206, 207, 213, 219, 225, 226, 228, 231, 232, 242, 246, 249, 285 Dictionnaire fondamental de l’informatique et de l’Internet, 225, 226, 227, 228, 230, 238, 249 Einbau, 31, 38 embodiment, 4, 15 Esperanto, 25, 27, 28, 29, 30, 31, 32, 34, 36, 37, 38, 39 etymology, 33, 34 exemplification, 214 exoteric language, 26 Explanatory and Combinatorial Lexicology, 227, 250 Face Threatening Act, 303, 305 Finite Simple Clause, 272

352

Subject index

foreign language teaching, 75, 275 Gerund, 256, 257, 258, 259, 260, 261, 262, 263, 264, 265, 266, 267, 274, 275, 276 grammar, 19, 26, 31, 33, 37, 71, 74, 258, 268, 338 grammatical knowledge pattern, 173 home language, 46, 48, 53, 55, 59 honorific axis, 305 Ido, 28, 29, 30, 31, 32, 34, 36, 38, 39 –ing adverbial, 260, 262, 263, 264, 267, 268, 270, 271, 272, 273 interlanguages, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 65, 287, 298 Interlingua, 27, 30, 33, 34, 35, 36, 38, 194 interlinguistic communication, 256, 257 interlinguistics, 24, 25, 36, 37, 38 Klingon, 26 knowledge pattern, 162, 163, 164, 165, 166, 173, 175, 180, 181 knowledge-rich context, 162, 164, 170, 172, 179, 180, 181 language acquisition, 37, 45, 94 language change, 102, 103, 104 language competence, 53, 55 Language Ecology, 46 language planning, 24, 25, 28, 37, 38, 39 language practice, 44, 45, 47, 48, 53, 56, 58, 60 language proficiency, 47, 61

language shift, 28, 38, 39, 83, 90, 100 language status, 55, 56, 61 Latino Sine Flexione, 27, 31, 32, 33 lexical cohesion, 323, 327, 329, 330, 331, 341, 342 lexical function, 227, 229, 231, 232 lexical knowledge pattern, 162, 164, 165, 173, 178, 179 lexical semantics, 226 lexical unit, 164, 166, 202, 203, 218, 227, 229, 230, 247, 250 lexicon, 19, 25, 30, 34, 37, 38 lift model, 144, 146 Macrolinguistics, 279 Meaning-Text Theory, 227 mélange/promotion dilemma, 30, 31 metaphor, 2, 3, 4, 6, 8, 9, 10, 11, 12, 13, 15, 16, 17, 18, 19, 20, 21, 325 metaphorical expression, 3, 5, 12, 15, 16, 17, 18, 19, 20, 21 modality, 293, 323, 331, 332, 333, 334, 335, 336, 337, 340, 341, 342 mood, 292, 323, 331, 332, 333, 335, 337, 338, 339, 341, 342 morphology, 28, 30, 31, 34, 36, 38, 309 multiculturalism, 44, 46 native language, 48, 49, 51, 91 Novial, 31, 32 Occidental, 32, 34, 35, 38 Ontology, 164, 187, 188, 191, 195, 197, 199 open question, 118, 122, 125, 137

Multilingualism and Applied Comparative Linguistics

P-ACTRES, 261, 266, 276, 279, 284, 285, 286, 287, 291, 298, 299 paralinguistic knowledge pattern, 164 parallel corpus, 258, 261, 279, 280, 281, 283, 284, 285, 287, 298 paraphrasing question, 119 Pascal, 26 phonetics, 28, 30, 38 planned languages, 25, 26, 27, 28, 30, 36, 37, 38 polysemy, 182, 225, 238, 246, 286 pragmatics, 38, 114, 311 primitive contract rule, 27, 30, 37 pronoun, 61, 110, 126, 127, 141, 145, 146, 149, 150, 151, 152, 153, 154, 157, 158, 176, 294, 295, 309, 318, 320 public language, 26 qualitative analysis, 262 questionnaire, 47, 67, 68, 69, 71, 86, 87, 93 referent axis, 305 reformulation, 204, 208, 209, 210, 211, 212, 213, 214, 216, 217, 220 regression analysis, 89, 96, 97, 98 Reiteration, 324, 329 Sapir-Whorf hypothesis, 19 second language acquisition, 85 securities fraud, 188 semantic role, 228, 259, 260, 262, 274 Semantic Web, 186, 187, 188 semantics, 26, 38, 187, 260, 264, 274, 275

353

Sephardic languages, 31 Smalltalk, 26 social distance, 65, 67, 68, 73, 303, 306, 309, 320 social dominance, 73 social norm perspective, 303 sociolinguistic factor, 53 sociolinguistic research, 84, 86, 87, 100 sociolinguistic variation, 85, 90 sociolinguistics, 45, 90 source domain, 3, 6, 8, 9, 15, 17, 18, 19, 20 source language, 30, 33, 34, 35, 258, 266, 278, 281, 285, 297 source text, 144, 283, 287, 313, 314, 316, 317, 319, 320, 323, 341, 342 speaker-addressee axis, 305 speaker-setting axis, 305 specialized discourse, 202, 203, 205, 209, 210, 215, 216, 217, 219 speech act, 64, 65, 66, 68, 69, 70, 73, 75, 143, 153 speech community, 25, 28, 37, 38, 39, 303 stabilization, 202, 203, 204, 205, 206, 207, 208, 211, 220 standardization, 24, 25, 27 sub-level domain, 17, 20 sub-level metaphor, 16, 17, 19, 20 synonym, 230, 238, 241, 324, 325, 329, 331, 341 syntax, 19, 28, 30, 36, 38, 328 target domain, 3, 7, 8, 9, 17, 18 target language, 64, 75, 258, 260, 266, 278, 283, 285, 297, 298

354

Subject index

target text, 258, 297, 311, 313, 314, 315, 316, 319, 320 taxonomy, 25, 26, 263 term candidate, 235, 236, 237, 249 term extractor, 232, 235 terminological unit, 202, 203, 206, 210, 211, 212, 213, 214, 215, 216, 219, 288 terminology, 164, 165, 181, 187, 188, 190, 191, 192, 193, 194, 195, 196, 198, 199, 202, 204, 205, 206, 207, 208, 209, 210, 212, 220, 227, 237, 243, 259, 281, 283, 331, 334 terminology analysis, 187, 195 terminology management system, 196 Termontography, 187, 194, 195, 196, 197 Termontography workbench, 196, 197 TermoStat, 166, 235, 236, 237, 250

tertium comparationis, 279, 280, 293 third language acquisition, 45 translated language, 256, 257, 276 translation corpus, 258, 261, 278, 281, 285 Translation Corpus Aligner, 276, 285 Translation Studies, 256, 276, 302 translationese, 258, 283, 287 trilingualism, 45 universality, 5, 15, 145 variation, 4, 15, 16, 18, 19, 20, 21, 33, 65, 68, 69, 75, 83, 84, 85, 87, 90, 92, 103, 173, 174, 176, 178, 180, 202, 204, 210, 211, 219, 264 VAT carousel fraud, 188 vocabulary, 26, 32, 35, 56, 163, 186 Volapük, 28, 31, 38, 39 WordSmith, 165, 182