Claims, Changes and Challenges in Translation Studies
Benjamins Translation Library The Benjamins Translation Library aims to stimulate research and training in translation and interpreting studies. The Library provides a forum for a variety of approaches (which may sometimes be conflicting) in a socio-cultural, historical, theoretical, applied and pedagogical context. The Library includes scholarly works, reference works, post-graduate text books and readers in the English language.
EST Subseries The European Society for Translation Studies (EST) Subseries is a publication channel within the Library to optimize EST’s function as a forum for the translation and interpreting research community. It promotes new trends in research, gives more visibility to young scholars’ work, publicizes new research methods, makes available documents from EST, and reissues classical works in translation studies which do not exist in English or which are now out of print.
General editor Gideon Toury
Associate editor Miriam Shlesinger
Tel Aviv University
Bar Ilan University
Advisory board Marilyn Gaddis Rose
Zuzana Jettmarová
Juan C. Sager
Binghamton University
Charles University of Prague
UMIST Manchester
Yves Gambier
Werner Koller
Mary Snell-Hornby
Turku University
Bergen University
University of Vienna
Daniel Gile
Alet Kruger
Sonja Tirkkonen-Condit
Université Lumière Lyon 2 and ISIT Paris
UNISA
University of Joensuu
José Lambert
Lawrence Venuti
Ulrich Heid
Catholic University of Leuven
Temple University
University of Stuttgart
Franz Pöchhacker
Wolfram Wilss
Eva Hung
University of Vienna
University of Saarbrücken
Chinese University of Hong Kong
Rosa Rabadán
Judith Woodsworth
W. John Hutchins
University of León
Mt. Saint Vincent University Halifax
University of East Anglia
Roda Roberts
Sue Ellen Wright
University of Ottawa
Kent State University
Volume 50 (Volume 1 in the EST Subseries) Claims, Changes and Challenges in Translation Studies: Selected contributions from the EST Congress, Copenhagen 2001 Edited by Gyde Hansen, Kirsten Malmkjær and Daniel Gile
Claims, Changes and Challenges in Translation Studies Selected contributions from the EST Congress, Copenhagen 2001
Edited by
Gyde Hansen Copenhagen Business School
Kirsten Malmkjær Middlesex University
Daniel Gile Université Lumière Lyon 2
John Benjamins Publishing Company Amsterdam/Philadelphia
8
TM
The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences – Permanence of Paper for Printed Library Materials, ansi z39.48-1984.
Library of Congress Cataloging-in-Publication Data EST Congress (3rd : 2001 : Copenhagen, Denmark) Claims, changes and challenges in translation studies : selected contributions from the EST Congress, Copenhagen 2001 / edited by Gyde Hansen, Kirsten Malmkjær, Daniel Gile. p. cm. (Benjamins Translations Library, issn 0929–7316 ; v. 50) Contributions in English, French, German, and Spanish. Includes bibliographical references and indexes. 1. Translating and interpreting--Congresses. I. Hansen, Gyde. II. Malmkjær, Kirsten. III. Gile, Daniel. IV. Title. V. Series. P306.E76 2004 418’.02-dc22 isbn 90 272 1656 8 (Eur.) / 1 58811 509 7 (US) (Hb; alk. paper)
2004041064
© 2004 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa
"intro"> "che"> "tou"> "pal"> "sto"> "hel"> "kor"> "sch1"> "zab"> "pok"> "nob"> "mal">
Table of contents
Introduction Hypotheses about translation universals Andrew Chesterman Probabilistic explanations in Translation Studies: Universals — or a challenge to the very concept? Gideon Toury
vii 1
15
A thousand and one translations: Revisiting retranslation Outi Paloposki and Kaisa Koskinen
27
Creating “presence” in translation Radegundis Stolze
39
Ready-made language and translation Pál Heltai
51
Les attributs indirects en français et en danois: Différences typologiques et problèmes de traduction Hanne Korzen
73
Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft Michael Schreiber
83
Translating non-segmental features of textual communication: The case of metaphor within a binary-branch analysis Patrick Zabalbeascoa
99
Challenging the myth of native speaker competence in translation theory: The results of a questionnaire Nike K. Pokorn
113
Expectativas y evaluación en la traducción de folletos turísticos Marie-Louise Nobs
125
Censorship or error: Mary Howitt and a problem in descriptive TS Kirsten Malmkjær
141
< /R/TREARGET E FF
"sch2"> "mil"> "ris"> "dra"> "aze"> "doo"> "ahr"> "bar"> "dam"> "fen"> "rud"> "hal"> "gil"> "ni"> "ci"> "toc">
vi
Table of contents
Of holy goats and the NYPD: A study of language-based screen humour in translation Thorsten Schröter
157
The figure of the factory translator: University and professional domains in the translation profession John Milton
169
Migrating from translation to technical communication and usability Hanna Risku From raw data to knowledge representation: Methodologies for user-interactive acquisition and processing of multilingual terminology Barbara Dragsted and Benjamin Kjeldsen
181
197
The translator as a creative genius: Robert Schumann João Azenha Junior
209
Übersetzung zwischen Nationalismus und Internationalismus Luc van Doorslaer
219
Non-verbal phenomena in simultaneous interpreting: Causes and functions Barbara Ahrens
227
Simultaneous interpreting A-B vs. B-A from the interpreters’ standpoint Magdalena Bartlomiejczyk
239
Interpreters’ notes: On the choice of form and language Helle V. Dam
251
Expressing a well-founded fear: Interpreting in convention refugee hearings Sabine Fenton Cross-cultural dynamics in community interpreting. Troubleshooting Mette Rudvin The child in the middle: Agency and diplomacy in language brokering events Nigel Hall
263
271
285
The editorial process through the looking glass Daniel Gile and Gyde Hansen
297
Name index
307
Concept index
313
Introduction
Since its foundation in Vienna in 1992, the European Society for Translation Studies (EST) has held three Congresses: in 1995 (Prague, Czech Republic), 1998 (Granada, Spain) and 2001 (Copenhagen, Denmark). The respective proceedings of the first two (Snell-Hornby et al. 1995, Chesterman et al. 1998) show a lively and diversified interest in many aspects of translation and interpreting, with papers on topics ranging from the links between translation and modern democracy to translation workplace procedures through terminology policies, markers of ideology in translation, aspects of universal grammar, creativity in legal translation, twentieth-century Chinese ideas of translation, the ethics of translation, a case study of simultaneous interpreting on television, Schleiermacher and modern translation theories, working memory and simultaneous interpreting, or gender stereotypes in Bible translation, to mention just a few examples. As the Translation Studies community grows and evolves, the range of its interests widens. EST Congresses, as opposed to focused, thematic seminars, offer translation and interpretation scholars the opportunity to meet and present their work, and their proceedings offer a sample of the diverse research interests of the EST community. This is also reflected in the articles published in this collection. At the third International Congress of the European Society for Translation Studies, held at the Copenhagen Business School, Denmark, in August 2001, participants continued to work on known themes, but also broached less familiar topics, including a new look at a personality from the world of music as a translator, the translation of screen humour, and the behaviour of children as language mediators. The first three papers, by Chesterman, Toury, and Paloposki and Koskinen, address the theoretical concept of Translation Universals, which has attracted a great deal of attention in Translation Studies recently, especially in conjunction with increased awareness of the possibilities offered by automatic corpus processing. Chesterman’s point of departure is a set of categories of hypotheses, which he proposes to use as tools for the description of historical thinking about translation universals. He claims that predictive hypotheses can be found underlying prescriptive statements made in the past. He also evokes “universal” criticism, a direct consequence of the prescriptive approach, before moving on to contemporary descriptive research and its descriptive universals, which he characterizes as either S-universals, which concern the processing of the source text in the translation, or
viii Introduction
T-universals, which concern the way translators process the target language. He offers a list of examples of potential S- and T-universals for further consideration. Toury focuses on methodological questions. He addresses the issue of the level at which a search for universals could be carried out (not too local, not too global), stresses the non-deterministic nature of candidates for universality, in particular in view of the existence of multiple causation in translational phenomena, and discusses in some detail the potential formats of propositions which could be tested. A particularly welcome question in this paper is how the relevance and power of variables involved in causation could be determined. The issue is central in engineering, physics and mathematics, but is rarely mentioned in Translation Studies, in spite of its importance in applied research. Paloposki and Koskinen discuss one candidate for Universal status, namely the hypothesis that retranslations are less domesticating than first translations. This seems intuitively reasonable, insofar as a “foreign” element initially introduced into a target culture is best assimilated if adapted, whereas in a retranslation, the need is less conspicuous. However, Paloposki and Koskinen look at actual cases of translation and retranslation of foreign literature into Finnish and find some counterevidence. Beyond the challenge to the hypothesis per se, and the conclusion that different factors may be involved, this paper is a demonstration of the need to do extensive empirical testing of candidate-Universals. Stolze’s ideas lie in the realm of hermeneutic language philosophy. She suggests that the translator’s knowledge base is activated by the source text. Comprehension of a source text and the ability to “create presence in translation” depend on the translator’s knowledge base and its adequacy. The translator’s task is to “create presence” for target-text readers, just as in acting on a theatre stage, an idea is “made present” for an audience. Translation is an assignment yet to be fulfilled, an open process towards an optimal solution. It is a cognitive event: as soon as the message is understood, target-language words and phrases appear on the cognitive scene on which the message is “present”. Several authors in this collection focus their attention on linguistic issues. Heltai makes the interesting point that language production involves many readymade multi-word units. He suggests that technical translation is associated with more use of such units than literary translation, especially under high time-pressure conditions, that translation under time pressure is actually impossible without the use of ready-made phrases, and that inappropriate use of multi-word units may be one of the characteristics of translationese. One of his conclusions is that the translator’s phraseological competence should include appropriate mastery of ready-made phrases in different registers and in both source language and target language. Translation is not an autonomous text production process, however, and can be viewed as constrained by relationships between the source language and the target language. Korzen establishes rules for the translation of free adjuncts between
Introduction
French and Danish. The problem arises against the background of important language-typological differences between Romance and Scandinavian languages. Free adjuncts are statistically much more frequent in French than in Danish. Korzen divides the construction into more or less descriptive and more or less relational free adjuncts and shows that only the more descriptive adjuncts can be translated directly into Danish. In the case of the more relational adjuncts, translation difficulties can be explained by morpho-syntactic differences and the compression of content in the French relational adjuncts. Korzen suggests that these constructions can be translated into Danish with adverbials or relative clauses. Linguistics comparisons per se do not tell the whole story. Schreiber stresses the value of language-pair-specific translation analysis within what he calls “sprachenpaarbezogene Translationswissenschaft”, literally “language-pair-specific translation science”, which is positioned at the interface between contrastive linguistics and general Translation Studies. He stresses that in spite of the increasing importance of extra-linguistic analyses of translation, which have marginalized the role of linguistics as such in TS, in actual translation, language-specific issues still represent the bulk of practical problems that translators must solve. He suggests that translation-solution-oriented language-specific analysis can help. Zabalbeascoa constructs a tree-diagram with successive binary branching at each level to represent all possibilities not only for translating metaphors (the focus of his paper), but for all other translation problems/solutions as well (jokes, insults, rhymes, etc.). This abstract structure (illustrated with concrete examples in the paper) can be adapted to descriptive and to prescriptive analyses, and allows identification of regularities, including translation strategies, which may help identify underlying norms. It is a binary structure, which could perhaps be used in a computerized model designed to automate the quantitative analysis on larger corpora. Zabalbeascoa deliberately adopts a speculative approach to his topic. Also linked to the issue of languages is Pokorn’s study on directionality (see also, below, a reference to Bartlomiecjczyk’s contribution on directionality in interpreting). Seven fragments from English translations of Slovene texts, translated by Slovene native speakers, English native speakers, and mixed pairs of translators working together, were presented to native speakers of English, who were asked by questionnaire to judge whether the translators were native speakers of English, and how many translators were involved in the translation. A substantial number of respondents did not identify correctly native vs. non-native English translators, and many respondents did not judge correctly how many translators were in the team. These findings challenge some received wisdom on the desirability of translation into one’s native language only. The reception of translations is also addressed by Nobs, who refers to general expectations of a certain type of translated text. She reports on a questionnairebased pilot study dealing with expectations and assessment of German translations
ix
x
Introduction
of a Spanish tourist leaflet. In the expectations questionnaire, priority was clearly given by respondents to content-related quality parameters. Translations in which deviations from the relevant target-language norms had been corrected were given only marginally higher assessments than translations which had not been similarly corrected. According to Nobs, some evidence suggests that form and content have a comparable weight in actual assessments. The issue of expectations vs. actual assessment has practical relevance to the professional world, and has recently been the focus of some attention in interpreting research as well. The question of shifts in translation is taken up by Malmkjær, who argues that through pattern analysis, it may be possible to make a distinction between choicebased shifts and actual errors. Shifts would be documented by semantic patterning, and errors by formal patterning. In an analysis of Mary Howitt’s translation of Hans Christian Andersen, she shows consistent censorship-like patterns in some translation solutions, which suggest choices, and the lack thereof in other surprising deviations, which can in almost every case be related to graphical or phonological error-inducing similarities. Beyond local shifts, the issue of translation strategies is of interest to the community. In an analysis of one British and six American film comedies and of their dubbing and subtitling into five target languages, Schröter identifies strategies for producing humour in both the source and target texts. The number of strategies expressing humour in the target text turns out to be roughly one third lower than in the source text. The study does not yet tell the reader whether dubbing or subtitling is superior in dealing with language-based screen humour or whether it is easier to translate English wordplays, which are an important ingredient in many comedies, more into some languages than into others. How relevant are scholarly analyses of translation issues by translation scholars to actual translation work? Not very relevant, according to the two surveys on the question reported in Milton’s contribution. The first was carried out in Brazil in 1999 and showed the existence of two distinct worlds of translation: the academic one, concerned with academic issues, and the professional one, concerned with professional issues, often related to technical texts. A follow-up survey carried out in 2001 confirmed the findings. Milton asks whether the university should attempt to adapt to the needs of professional translators, in particular by providing training in newly developing translation fields, or play other roles (arbitration, standardsetting, etc.). Translators do not always remain translators. They sometimes migrate into different activities. Professional migration from translation to technical communication is the topic taken up by Risku, who reports on an investigation of six cases of such migration, using semi-structured, individual interviews. Her paper presents information on the reason for the career shift, on the perception of the advantages of the new career, and on similarities and differences between the two careers. In
Introduction
particular, technical communication was seen as more autonomous and active, and as involving more project management and self management. The findings suggest that training and professional experience as translators may provide a good foundation for further training and migration into technical communication, but that they do not equip a person adequately for working in technical communication. From terminology, another neighbouring field, Dragsted and Kjeldsen refer to the practical issue of bringing together and optimizing cooperation between human translators, translation software developers and the police in an effort to collect terminology relevant to a project aimed at the exchange of law-enforcement intelligence. The needs of the three groups are analysed and shown to be different and sometimes conflicting. A particular event-based knowledge model was developed, and workshops in terminology for police officers and seminars with individual users were organized. Besides the large terminology data base produced, the project resulted in a methodology for terminology exchange across countries and institutions and in a specific knowledge model. The connection between translation and literature is obvious, but Azenha shows how translation can also be related to music. Robert Schumann is famous as a musician, but his interest in foreign languages, literature and translation is not well known. Azenha describes this interest, which started when Schumann was a young pupil at school and lasted all his life. He lists some of Schumann’s translations from Greek and Latin, and highlights his interest in poetry. Music could be considered a type of translation of his literary experience. Schumann’s translations are difficult to assess, because only a few fragments of his original manuscripts are available, but Azenha emphasises the important part played by translating in the artist’s development, as it enabled him to increase his expressive resources and enrich his work with input from other cultures. Moving from the individual to the group, in terms of linguistic and national communities, Van Doorslaer highlights the role of translation in the context of tensions between linguistic communities, both international and intra-national. Bilingual and multilingual countries such as Finland, Canada and the Netherlands are cases in point. In European Union working groups, where all European languages cannot be used, certain working languages have to be selected for translation and interpreting, and this choice can also be analysed in political, as well as in pragmatic terms. Thus, translation is central in identity-forming mechanisms, often reflects political and social power relationships, and may be a promising field of investigation for translation scholars who adopt this vantage point. On the interpreting side, three papers are devoted to issues in conference interpreting, and three to community interpreting. Conference interpreting is considered from the technical viewpoint, with three problem areas in mind. It is trivial to say that the most obvious difference between translation and interpreting is the written form of the former, and the oral form of the latter. What
xi
xii
Introduction
is less trivial is the exploration of the phonological dimension of interpreting, in both descriptive and functional terms, and very few studies are devoted to this topic. Following in the footsteps of Daró, Williams and Shlesinger, Ahrens looks at the interpreter’s intonation, confirms the previous finding that the interpreter’s pitch movements are not those that would occur naturally in spontaneous speech, and suggests that rather than mimicking the speaker’s, they may well be used as structuring aids in target-speech production. Speech production issues are particularly important in interpreting, where there is no time to examine and correct one’s target utterance. Directionality is therefore as central an issue in interpreting as in translation (see also the discussion of Pokorn’s contribution above). Bartlomiecjczyk reports on a questionnaire survey on the perception of directionality by interpreting students and professionals. The latter were found to prefer to work into their A language, while the preference was not as clear-cut among students. Bias is possible, due to the fact that the professionals were AIIC interpreters, which presupposes a directionality norm. Further bias may have come from the large proportion of students from Poland, where the prevalent norm may well be the opposite, as in the former USSR. In consecutive interpreting, the production of notes during the listening phase, while the speaker is producing his/her utterance, is of particular importance, because during the second phase, interpreters rely on them heavily for speech reformulation. Dam analyses a corpus of consecutive notes and looks at the interpreters’ preferences for symbols vs. language and for source language vs. target language. From this exploratory study, she develops interesting hypotheses, in particular the idea that interpreters prefer noting in the target language when they have no particular difficulty with the source speech, and that they shift to the source language when the going gets tough, presumably in order to save processing capacity. Moving on to public service interpreting (in the wide sense of the word), we find a concern with sociological issues: Fenton reports on interpreters working in convention refugee hearings in New Zealand. Her survey questionnaire showed that nearly all respondents had come to New Zealand as migrants themselves, that they were active in their respective communities, that the majority were university graduates but had little training in interpreting, and that most of them felt valued by their employer. They were often placed in an uncomfortable position because of pressure by appellants, or because of incompetence and sometimes unfair practices on the interviewers’ part, and because they felt that, at times, they were party to false claims by asylum seekers. Their role perception was clearly that of professional, impartial language mediators. Rudvin also looks at pressures, arising both from power asymmetries and from differences in language and/or culture. She presents a number of illustrations of the kinds of problem faced by public service interpreters. For instance, metaphors and indirect speech may be used far more often in so-called traditional cultures than in
Introduction xiii
Western cultures, and may cause comprehension difficulties in a hospital setting, or a misperception of testimony as inaccurate. In one case, reluctance to indicate a rapist’s action explicitly for cultural reasons resulted in the testimony sounding unconvincing and in the rapist’s acquittal. Like Fenton, Rudvin recalls that interpreters often have ethnic affiliations as well, and these may stand in the way of acceptability. Hall’s paper focuses on children’s perception of their role when they act as ‘language brokers’ in conversations between adults. An interview between a nonEnglish Urdu-speaking mother of a young boy who wanted to enrol her child in a school and an English-speaking representative of the school was simulated, with children acting as interpreters. The main issue explored in the study was the children’s role perception in the face of contradictory social pressures: the mother said something about her son’s naughtiness, which was negative in the context of seeking his enrolment in school. The evidence shows that children reacted on multiple levels and actually shaped and controlled the event to a considerable extent. In the process of bringing this volume into being, the detailed comments received from referees on the contributions were remarkably similar, considering the wide range of topics covered and the authors’ very different backgrounds. The editors felt that the data and the message conveyed by their convergence were worth analysing and presenting to the EST community, which Gile and Hansen do in the last contribution in this volume. They discuss the fact that most of the referees requested that the authors provide more complete reporting of data and a more precisely formulated rationale. The referees’ most frequent comments are presented and analysed. The Editors
References Chesterman, Andrew, Natividad Gallardo San Salvador and Yves Gambier (eds). 2000. Translation in Context. Selected papers from the EST Congress, Granada 1998. Amsterdam/Philadelphia: John Benjamins. Snell-Hornby, Mary, Zuzana Jettmarová and Klaus Kaindl (eds). 1995. Translation as Intercultural Communication, Selected papers from the EST Congress. Amsterdam/Philadelphia: John Benjamins.
Hypotheses about translation universals* Andrew Chesterman University of Helsinki
Contemporary proposals about universals are descriptive hypotheses about (a) the relation between translations and source texts (the equivalence relation), and (b) the relation between translations and comparable non-translations in the target language (the relation of textual fit). The article analyses the main trends in the thinking about translation universals, pointing out connections between the prescriptive tradition, work in literary translation criticism, and current corpus-based research on universals. In order to characterize some of the main similarities and differences between these different approaches to universals, I use a classification of different kinds of hypotheses. After introducing these central notions, I place modern descriptive research in its historical context and then consider some of the terminological, conceptual and methodological problems of this kind of research.
Different kinds of hypotheses Most research involves, at some point, the generation and/or testing of hypotheses. This is a normal part of the scientific process of investigation and observation, of constructing and testing theories. I shall use the notion of a hypothesis here in order to analyse and compare different kinds of claims about translation universals. A hypothesis is a tentative statement. It is not a statement of fact, but a claim that something might be true or worth considering. Some scholars prefer not to talk about hypotheses but prefer “claims”, or “arguments”; what I want to say below does not depend on the term “hypothesis” itself. We can distinguish various kinds of hypotheses. Some are more typical than others in particular fields of research, but all are relevant in the interdiscipline of translation studies. (For further discussion, see Chesterman 2000.) The main characteristics of hypotheses are of course well known in the philosophy of science (see e.g. Cohen and Nagel 1934, Popper 1963, Hempel 1966). The basic hypotheses are explanatory ones. These are statements that suggest explanations for a given phenomenon; they suggest probable causes, reasons,
2
Andrew Chesterman
influences. Alongside these there are descriptive hypotheses: these make claims about what something is, what characteristics or structure it has. A descriptive hypothesis is a claim about how to describe something, and hence how to relate it to, and distinguish it from, other things. More specifically, it is a claim about particular features that (you think) are common to all instances of the thing you are describing. Formulating descriptive hypotheses is thus part of the categorization of phenomena. Both explanatory and descriptive hypotheses can be tested via predictive hypotheses. For instance, if you think you have a good idea of the causal conditions that lead to a given phenomenon, you can test this idea by predicting that whenever these conditions exist, the phenomenon will follow, and then check how good your predictions are. Predictive hypotheses are usually more specific than explanatory ones, so that they can be better tested. Descriptive hypotheses also incorporate predictions: if all instances of a phenomenon are claimed to have feature F, this implies the prediction that the next instance you find will also have the feature F. Explanatory and descriptive hypotheses combine to form a theory of the explanandum, the phenomenon they claim to describe and explain. Suppose you are developing a theory about bananas and people’s use of them, for instance. One relevant descriptive hypothesis might be that bananas have nonedible skins: this is one characteristic of all bananas, you claim, which distinguishes them from other kinds of fruit that have edible skins. Formulated as a prediction: all future bananas found or studied will have this feature. If you then find a kind of banana with a skin that can be eaten, you have to reject or refine your hypothesis. Formulating a descriptive hypothesis is thus a way of trying to see what the distinguishing features of something are. Testing a descriptive hypothesis is a way of checking how general the description is: does it really apply to all cases? If you then move on to wonder why banana-skins are like this, you need to generate explanatory hypotheses. (Because they preserve the fruit better than edible skins would? Would this explanation also apply to other such fruit?…) Finally, there are interpretive hypotheses. These are claims about what something means, how it can best be understood or interpreted. Interpretive hypotheses are typical of hermeneutic research, although the actual term “hypothesis” is not so current in this sense. Niiniluoto (1983: 174), however, argues that the basic method of hermeneutic research is indeed the presentation and testing of “interpretive hypotheses”. (See also Føllesdal 1979.) In my own work, I have used the term in order to underline a basic similarity between positivist and hermeneutic research approaches: both seek to make justified claims (hypotheses) that can be subsequently evaluated and tested in different ways. A key word in interpretive hypotheses is the word as (in English). Translating, for instance, has been seen as a great many things: as rebuilding, recoding, changing clothes, performing, travelling, eating the dead… Each new way of seeing highlights
Hypotheses about translation universals
different aspects of it, and allows different insights. You test interpretive hypotheses in practice (not directly, against empirical evidence): do they indeed bring new insights, new understanding? Do they stimulate us to generate other new hypotheses of various kinds? Some scholars prefer not to use the term “hypothesis” in this meaning, because the process of hypothesis-testing works differently for these claims. Yet interpretive hypotheses also have a predictive sense: they implicitly predict that their adoption will bring new understanding etc. If it turns out that they do not seem fruitful, they eventually fade away. With these conceptual tools in mind, let us now take a historical look at what scholars and translators have thought about translation universals. Any such thinking must have to do with descriptive hypotheses: the underlying goal is to discover something about what all translations have in common, something that distinguishes them from texts that are not translations. This would contribute to the general categorization of texts. If we find evidence of such features, we can then look for explanatory hypotheses which would plausibly account for them. We must start, however, with a preliminary interpretive hypothesis, about what we mean by a translation universal. In simple terms, we can define a translation universal as a feature that is found (or at least claimed) to characterize all translations: i.e. a feature that distinguishes them from texts that are not translations. More strictly: to qualify as a universal, a feature must remain constant when other parameters vary. In other words, a universal feature is one that is found in translations regardless of language pairs, different text-types, different kinds of translators, different historical periods, and so on.
“Universal” prescriptions The first evidence of universalist thinking about translation goes back to the prescriptive claims of early writers about what all translations should be like. In many of these writings we find a kind of ideal picture of a Platonic translation, and this image is described in terms that are assumed to apply indeed to all translations. Classic examples abound in the early literature: Dolet’s and Tytler’s translation principles, for instance. Here they are, paraphrased: Translations should not be word-for-word renderings of the original. Translations should avoid unusual words and expressions. Translations should be elegant, not clumsy. (Dolet 1540; three of his five general principles) Translations should give a complete transcript of the ideas of the original. Translations should be in the same style as their source texts. Translations should be as natural as original texts. (Tytler 1797)
3
4
Andrew Chesterman
Hidden beneath these prescriptive statements about the universal characteristics of good translations there actually lie predictive hypotheses. What the authors of these statements are implying is this: if a translation has such-and-such a feature, people (readers, I myself) will react by thinking that it is a good translation. In other words, I say that this is the right way to translate. (For a more detailed discussion of this point, see Chesterman 1999.) The weaknesses of this approach are obvious. These undifferentiated claims assume that all translations are of the same kind, and that the same quality criteria always apply. They are usually based on either Bible or literary translation: both these are rather special cases, so generalizations cannot validly be drawn from them. The claims are thus based on overgeneralizations. One way to get round this fallacy is to narrow the scope of the claims so that they apply not to all translations, universally, but only some of them, a subset. Recall the two prescriptive principles of St. Jerome (395), one for each of two different kinds of translation: – –
Translations of sacred texts must be literal, word-for-word (because even the word order of the original is a holy mystery and the translator cannot risk heresy). Translations of other kinds of texts should be done sense-for-sense, more freely (because a literal translation would often sound absurd).
These too are implicitly predictive hypotheses, for the same reason as the previous cases mentioned.
“Universal” criticisms Running alongside the Platonic translation ideal we also have another stream of thought seeking to establish generalizations about translations. It is a direct consequence of the prescriptive approach. Because the ideal translation must necessarily remain an ideal, all translations fail in some way. All translations are less than ideal; they are secondary, deficient, they always lose or change something, they betray the original, they are not faithful, or not beautiful, or both. This view of the typical features of translations is presented by some critics of literary translations, and often by the general public who read translations of tourist brochures, menus and the like. The literary critics complain because translations are too free, and the general public complain if the translations they read are not free enough — i.e. if they sound unnatural. One of the most recent examples of the first kind of criticism is represented by much of the work of Antoine Berman. Berman, it must be said, is not writing about all translation; but he does intend his comments to pertain to all literary translation, to all of this subset. Here is his list of what he calls the “deforming tendencies” of literary translation (Berman 1985; see also Munday 2001: 149–151).
Hypotheses about translation universals
– – – – – – – – – – – –
Rationalization (making more coherent) Clarification (explicitation) Expansion Ennoblement (more elegant style) Qualitative impoverishment (flatter style) Quantitative impoverishent (loss of lexical variation) Destruction of rhythms Destruction of underlying networks of signification Destruction of linguistic patternings (more homogeneous) Destruction of vernacular networks or their exoticization (dialect loss or highlighting) Destruction of expressions and idioms (should not be replaced by TL equivalent idioms) Effacement of the superimposition of languages (multilingual source texts)
A similar line of argument is to be found in Milan Kundera’s ideas about translation, particularly the translations of his own works (Kundera 1993: 123f.). He complains of the way translators violate metaphors, seek to enrich simple vocabulary, reduce repetition, spoil sentence rhythms by altering punctuation, even change the typography. It is interesting to note how some of these putative deficiencies recur in the descriptive work we shall come to below. Examples of the second kind of reaction (the reaction to unnaturalness) can be frequently found in letters to newspapers by people complaining about translations of administrative documents, and in the amusement of tourists when faced with what has come to be known as Tourist English (see the Internet for examples!). This pejorative approach also suffers from overgeneralization. In particular, it suffers from a restricted view of what constitutes an acceptable translation in the first place. This view is so narrow that a great many translations are automatically criticized, although they might be perfectly acceptable according to other criteria than those selected by the critic in question, e.g. relating to strict formal equivalence or flawless target language. With respect to the alleged weaknesses of much literary translation, one can point out that most readers of literary translations may well prefer a freer, more natural version anyway. More importantly, these literary critics overlook the fact that a given formal feature (repetition, say) may have quite different effects on readers in different cultures, where there may be quite different rhetorical and stylistic norms. Berman’s list of what he sees as typical features of literary translation is first of all a list of descriptive hypotheses: the claim is that these features can be found in literary translations (or at least in most of them, or in typical examples); that the features characterize literary translation in general. But Berman is actually doing more than merely describing. By calling these features “deforming tendencies” (tendances déformantes) he is also making a value
5
6
Andrew Chesterman
judgement, and implying a prescriptive statement: literary translations, he implies, should not have these features. His list thus represents a mirror image of the earlier prescriptive claims. The claims made by Berman and Kundera (et al.) are also covert predictive hypotheses. Their authors imply: if a translation has these features, which I think are nasty features, people will react by not liking the translation. At least, discerning critics will react in this way. We can also comment on an implicit interpretive hypothesis here. By choosing to call these features “deforming tendencies”, Berman reveals his underlying metaphor, the image by means of which he interprets his observations. The features are interpreted as “deformations”, i.e. as something twisted out of its true, natural shape, as a person’s foot may be deformed at birth. Only the pure, undistorted shape seems to be acceptable, i.e. the shape given by the original work. Any change of shape is a disfigurement (rather than, for instance, a new creation, a new form, valuable as such). Thus does Berman’s choice of image force us to see his argument and his data in a certain light, rather than in some other light. To counter-balance these widespread criticisms we can of course also mention the self-evident fact that all translations have the positive characteristics of enabling people to read texts that are otherwise not accessible to them, etc. The point of this brief section has merely been to illustrate what has been one longstanding and influential tradition of universalist thinking about translation, and to show the links this tradition has with some of the hypotheses currently being investigated in descriptive work.
Two linguistic relations Before we turn to contemporary descriptive research, let us pause here to consider the nature of the linguistic relations upon which all statements about universals seem to be based. What we have is three sets of texts. First, we have translations. These are related to two other kinds of texts, which we can call reference corpora: to source texts on the one hand, and to non-translations in the target language on the other. The first relation is usually known as the relation of equivalence, interpreted in various ways. The second relation has been given many labels: we have referred to it above as the quality of naturalness; another term is the relation of acceptability. We could call it the relation of target text family fit (following the suggestion of an anonymous referee), or textual fit for short. This relation concerns the degree to which the linguistic profile of a translation matches the linguistic profile of the relevant family of texts in the target language; this textual family is made up of independently produced texts (not translated from source texts) of the same kind, with the same kind of subject matter and with the same kind of function. Some scholars refer to these texts as comparable texts; others call them parallel texts; still
Hypotheses about translation universals
others call them original texts. All these terms are problematic, as they all also have other uses. We could perhaps use the term non-translations to describe this group. As we can see from the prescriptive and pejorative traditions discussed above, claims about the universal features of translations are actually claims about the two relations of equivalence and textual fit. The pejorative claims, more specifically, concern differences between translations and the two reference corpora: either differences in equivalence (translations are not faithful) or differences in textual fit (translations are not beautiful, not like natural texts). In descriptive research, we deal with potentially universal features of the same two relations: between translations and their source texts, and between translations and non-translations in the target language. In order to facilitate the discussion, I shall call the first kind of universals S-universals, because they concern the way translators process the source text; the second kind are then T-universals, because they concern the way translators process the target language. (I originally thought this latter type could be called P-universals, from parallel texts; however, the ambiguous usage of the term “parallel” by different scholars is misleading. Hence this change of label, and hence also my preference for the term “non-translations”.)
Descriptive universals Descriptive research using electronic corpora is the most recent approach to thinking about possible translation universals. I call this approach “descriptive” because it does not incorporate ideas about what translations should or should not be, but about what translations (typically) are. One of the origins of such work has been Frawley’s notion (1984) of translations as constituting a third code in their own right, distinct from the source-language and target-language codes. Another origin has been hypotheses like that of Blum-Kulka (1986) on explicitation, and yet another has been Toury’s (1995) proposals about translation laws. We should also mention the background of work in linguistics on language universals, and in sociolinguistics on language variation. Mona Baker’s work (starting with a seminal article in 1993) has helped to make this kind of corpus-based research into one of the main paradigms of contemporary translation studies. Below are some examples of possible S-universals and T-universals. Note that these claims are hypotheses only; some have been corroborated more than others, and some tests have produced contrary evidence, so in most cases the jury is still out. Some of these hypotheses are also more abstract than others. They are all descriptive hypotheses: they purport to describe universal features of translations. However, as we saw above, descriptive hypotheses also have predictive implications, and such hypotheses can also be tested predictively, against new translations.
7
8
Andrew Chesterman
Potential S-universals – – – – – – – – –
Lengthening: translations tend to be longer than their source texts (cf. Berman’s expansion; also Vinay and Darbelnet 1958: 185; et al.) The law of interference (Toury 1995) The law of standardization (Toury 1995) Dialect normalization (Englund Dimitrova 1997) Reduction of complex narrative voices (Taivalkoski 2002) The explicitation hypothesis (Blum-Kulka 1986, Klaudy 1996, Øverås 1998) (e.g. there is more explicit cohesion in translations) Sanitization (Kenny 1998) (more conventional collocations) The retranslation hypothesis (later translations tend to be closer to the source text; see Palimpsestes 4, 1990) Reduction of repetition (Baker 1993)
Potential T-universals – – – –
Simplification (Laviosa-Braithwaite 1996: less lexical variety, lower lexical density, more use of high-frequency items) Conventionalization (Baker 1993) Untypical lexical patterning (and less stable) (Mauranen 2000) Under-representation of TL-specific items (Tirkkonen-Condit 2000, 2002)
The link between S-universals and T-universals can be quite complex. Take the hypothesis about interference, for instance. This claims that some aspects of the source-text form will inevitably be carried over into the translation. How do we arrive at this hypothesis? First of all, we notice certain differences between a set of translations and a set of non-translations: the translations do not, in some way, seem entirely natural. We wonder why this unnaturalness is there. Then we notice certain formal similarities between the translations and their source texts. We assume that these similarities are evidence of interference. The existence of this interference would offer an explanation for the unnatural features that first caught our attention: the source text has exerted an influence on the translator, a pull away from completely natural target language. If we have evidence to suggest that this pull also affects other translations, and still others, we end up thinking that interference may be a universal: an S-universal, because it manifests a relation (a similarity relation) between translations and their source texts. This example illustrates the kinds of relations we are concerned with. In the case of S-universals, we are looking at both similarities (> interference) and differences (> shifts, strategies, changes). In the case of T-universals, the focus is more on differences; similarities here would merely indicate naturalness, not universal indicators of translations as a distinct class of text.
Hypotheses about translation universals
Scope, conditions Genuine universals are the subject of unrestricted hypotheses: these claims aim to be valid for all translations of all kinds, in all times and places, universally. Such hypotheses thus have an unrestricted scope, or range of application. Other descriptive hypotheses start off with a more restricted scope in the first place, and concern only a subset of translations (recall Berman’s list for literary translation, above). These are not therefore hypotheses about genuinely universal features, but they can still bring new knowledge when they are tested. They may even turn out to have a wider scope than first thought. Translation research has made use of various types of limiting conditions in order to narrow the scope of claims. There are claims about features of translations between particular language pairs, in a given direction, perhaps concerning particular language items (e.g. the classic Vinay and Darbelnet 1958). Or claims about translations in a particular period in a particular culture (e.g. Toury 1995: 113f. on early 20th century Hebrew norms). Or claims about a particular type of translation (characterized e.g. by a given text type or skopos type: subtitling, technical, poetry, comic strips, gist translation…). Or claims pertaining to translations done by a particular translator (e.g. Baker 2000), or by translators of a particular kind (trainees; men/women; into L1 or L2; …). Or claims pertaining to particular conditions of the publishing or editorial process, in-house stylistics conventions and the like (e.g. Milton 2001). Each of these scope limitations thus defines a particular subset of translations or translators. We might find that given features are typical (or not typical) of some subset; or that given features seem to be typical (or not typical) of more than one subset. We can thus make progress either by starting with a very general claim and testing it out on specific data, or by starting with a more limited claim and testing to see whether it can also apply to a wider set of data. Additionally, of course, we can generate entirely new hypotheses, e.g. as a result of a case study that yields interesting results.
Problems with descriptive hypotheses One problem is that of representativeness. Since we can never study all translations, not even all translations of a certain type, we must take a sample. The more representative the sample, the more confidence we can have that our results and claims are valid more generally. Yet our data may still be unrepresentative in some way that we have not realized. It is not a priori obvious, for instance, what we should count as corpus-valid translations in the first place: there is not only the tricky borderline with adaptations etc., but also the issue of including or excluding
9
10
Andrew Chesterman
non-professional translations or non-native translations, and even defining what a professional translation is (see Halverson 1998). Should we even include “bad” translations? They too are translations, of a kind… Related to this is the problem of universality. Claims may be made that a given feature is universal, but sometimes the data may only warrant a restricted claim, if the data are not representative of all translations. Many “universal” claims have been made that actually seem to pertain only to literary or to Bible translation. More fundamentally, though: whatever translations we study, there is always the risk that our results will be culture-bound (Tymoczko 1998). Concepts of translation itself are culture-bound, for a start; even prototype concepts may be, too. We can perhaps never totally escape the limits of our own culture-boundness, even if this might be extended e.g. to a general “Western culture”. As regards conceptualization and terminology in this area there is unfortunately still a great deal to be clarified. I made one proposal above, about distinguishing between S-universals and T-universals. Baker’s original use (1993) of the term “universal” seems to have to refer to T-universals, since her point of comparison is non-translations; however, several of the examples of previous research that she mentions are based on evidence from a comparison with source texts, and hence concern S-universals (such as the reduction of repetition). Some scholars prefer to refer to these claims as hypotheses, such as the explicitation hypothesis (Blum-Kulka and others) or the simplification hypothesis (Laviosa), or the retranslation hypothesis. Others speak of laws: cf. Toury’s proposed laws of interference and standardization. Chevalier (1995) writes about “figures of translation”, comparable to rhetorical figures. Still other scholars prefer to look for core patterns, or simply widespread regularities. When it comes to the hypotheses themselves we find a plethora of abstract terms that appear at first sight to mean more or less the same thing (e.g. standardization, simplification, levelling, normalization, conventionalization). Different scholars often operationalize these abstract notions in different ways. Sometimes they are used to refer to a feature of difference between translations and their source texts, and sometimes to a feature of difference between translations and nontranslations. The resulting confusion leads to much reinventing of the wheel, and makes it hard to compare different results and claims. Furthermore, some of the terms used appear to be ambiguous between a process reading (from source text to translation) and a product reading (e.g. terms ending in -tion in English). A final major problem has to do with causality: universals, if they exist, presumably have both causes and effects. Here, we can currently do little more than speculate as rationally as possible. The immediate causes of whatever universals there may be must be sought in human cognition — to be precise, in the kind of cognitive processing that produces translations. Constraints on cognitive processing in translation may also be present in other kinds of constrained communication,
Hypotheses about translation universals
such as communicating in a non-native language or under special channel restrictions, or any form of communication that involves relaying messages, such as reporting discourse, even journalism. It may be problematic, eventually, to differentiate factors that are pertinent to translation in particular from those that are pertinent to constrained communication in general. Other kinds of explanations may be sought e.g. in the nature of translation as a communicative act, and in translators’ awareness of their socio-cultural role as mediators of messages for new readers (see e.g. Klaudy 1996). Translators tend to want to reduce entropy, to increase orderliness. They tend to want to write clearly, insofar as the skopos allows, because they can easily see their role metaphorically as shedding light on an original text that is obscure — usually unreadable in fact — to their target readers: hence the need for a translation. Their conception of their role may give a prominent position to the future readers of their texts; this may even have been emphasized in their training. Additionally, economic and technical factors may be relevant, exerting their own pressure on the translator’s work process. All such potential non-cognitive factors must eventually take their effect via the translator’s cognition, though — consciously or unconsciously. Research into the effects caused by potential universals is still in its infancy. Effects on readers, on translator trainers, and on translators themselves would all be worth studying. It may be that the more we know about T-universals, for instance, the more scholars or trainers will see them as undesirable features that should be avoided — at least in translations whose skopos includes optimum naturalness. On the other hand, as the sheer quantity of translations grows and target-language norms become blurred, it may be that readers will become more tolerant of apparent nonnativeness; different cultures might differ considerably in this respect. One effect of knowledge about S-universals on source-text writers might be a greater concern for the clarity of the source text, in order to facilitate the translator’s task and lessen the need for explicitation. This in turn could lead to greater fidelity to the original.
Patterns The search for universals is one way of seeking generalizations. This does not mean that unique particulars are overlooked: all translations are unique in some way. All three approaches mentioned above have been interested in general patterns rather than what makes a given translation unique. But even in unique particulars we may find patterns that remind us of other patterns we have come across elsewhere: no translation is absolutely unlike every other translation. What ultimately matters is perhaps not the universals, which we can never finally confirm anyway, but new knowledge of the patterns, and patterns of patterns, which helps us to make sense of what we are looking at.
11
12
Andrew Chesterman
Author’s Address: MonAKO Unioninkatu 40 00014 University of Helsinki Finland Email: andrew.chesterman@helsinki.fi
Note * This article is based on three conference presentations, during which my ideas on the topic have developed. The first was read at the Third EST Congress in Copenhagen in August-September 2001, as part of the session on universals; another was read at the CoLLaTE Symposium on Contrastive Analysis and Linguistic Theory at Ghent in September 2001; and a third at the Conference on Translation Universals at Savonlinna in October 2001. I am grateful for all the critical comments and feedback I have received at these meetings and from referees later.
References Baker, Mona. 1993. “Corpus linguistics and translation studies: Implications and applications”. In Text and Technology: In Honour of John Sinclair, M. Baker et al. (eds), 233–250. Amsterdam: Benjamins. Baker, Mona. 2000. “Towards a methodology for investigating the style of a literary translator”. Target 12 (2): 241–266. Berman, Antoine. 1985. Traduction et la Lettre ou l’Auberge du Lointain. Paris: Seuil. Blum-Kulka, Shoshana. 1986/2000. “Shifts of cohesion and coherence in translation”. In Interlingual and Intercultural Communication: Discourse and Cognition in Translation and Second Language Acquisition Studies, J. House and S. Blum-Kulka (eds), 17–35. Tübingen: Narr. Chesterman, Andrew. 1999. “The empirical status of prescriptivism”. Folia Translatologica 6, 9–19. Chesterman, Andrew. 2000. “A causal model for translation studies”. In Intercultural Faultlines. Research Models in Translation Studies I. Textual and Cognitive Aspects, M. Olohan (ed), 15–27. Manchester: St. Jerome Publishing. Chevalier, Jean-Claude. 1995. “D’une figure de traduction : le changement de ‘sujet’”. In L’horlogerie de Saint Jérôme, J-C. Chevalier and M-F. Delport, 27–44. Paris: L’Harmattan. Cohen, Morris R. and Ernest Nagel. 1934. An Introduction to Logic and Scientific Method. London: Routledge and Kegan Paul. Dolet, Étienne. 1540. La Manière de Bien Traduire d’une Langue en Aultre. Paris: Marnef. Englund Dimitrova, Birgitta. 1997. “Translation of dialect in fictional prose — Vilhelm Moberg in Russian and English as a case in point”. In Norm, Variation and Change in Language. Proceeedings of the centenary meeting of the Nyfilologiska sällskapet, Nedre Manilla 22–23 March, 1996, 49–65. Stockholm: Almqvist and Wiksell. Føllesdal, Dagfinn. 1979. “Hermeneutics and the hypothetico-deductive method”. Dialectica 33: 319–336.
Hypotheses about translation universals
Frawley, William. 1984. “Prolegomenon to a theory of translation”. In Translation. Literary, Linguistic and Philosophical Perspectives, W. Frawley (ed.), 159–175. Newark: University of Delaware Press. Halverson, Sandra. 1998. “Translation studies and representative corpora: Establishing links between translation corpora, theoretical/descriptive categories and a conception of the object of study”. Meta 43 (4): 494–514. Hempel, Carl G. 1966. Philosophy of Natural Science. Englewood Cliffs, N. J.: Prentice-Hall. Jerome, E. H. [395] 1997. “De optime genere interpretandi”. Translated by P. Carroll as “On the best kind of translator”, in Western Translation Theory from Herodotus to Nietzsche, D. Robinson (ed.), 22–30. Manchester: St. Jerome Publishing. Kenny, Dorothy. 1998. “Creatures of habit? What translators usually do with words”. Meta 43 (4): 515–523. Klaudy, Kinga. 1996. “Back-translation as a tool for detecting explicitation strategies in translation”. In Translation Studies in Hungary, K. Klaudy et al. (eds), 99–114. Budapest: Scholastica. Kundera, Milan. 1993. Les Testaments Trahis. Paris: Gallimard. Laviosa-Braithwaite, Sara. 1996. The English Comparable Corpus (ECC): A Resource and a Methodology for the Empirical Study of Translation. Unpublished PhD thesis, UMIST, Manchester. Mauranen, Anna. 2000. “Strange strings in translated language. A study on corpora”. In Intercultural Faultlines. Research Models in Translation Studies I. Textual and Cognitive Aspects, M. Olohan (ed.), 119–141. Manchester: St. Jerome Publishing. Milton, John. 2001. “The figure of the factory translator”. Paper presented at the Third EST Congress, Copenhagen, August 30–September 1, 2001. Munday, Jeremy. 2001. Introducing Translation Studies. Theories and Applications. London: Routledge. Niiniluoto, Ilkka. 1983. Tieteellinen Päättely ja Selittäminen. Helsinki: Otava. Popper, Karl. 1963. Conjectures and Refutations: The Growth of Scientific Knowledge. London: Routledge and Kegan Paul. Taivalkoski, Kristiina. 2002. “Traduire la mixité formelle : l’exemple des premières (re)traductions de Fielding en France”. Faits de Langues 19: 85–97. Tirkkonen-Condit, Sonja. 2000. “In search of translation universals: non-equivalence or ‘unique’ items in a corpus test”. Paper presented at the UMIST/UCL Research Models in Translation Studies Conference, Manchester, 28–30 April 2000. Tirkkonen-Condit, Sonja. 2002. “Translationese – a myth or an empirical fact? A study into the linguistic identifiability of translated language”. Target 14 (2): 207–220. Toury, Gideon. 1995. Descriptive Translation Studies and beyond. Amsterdam: Benjamins. Tymoczko, Maria. 1998. “Computerized corpora and the future of translation studies”. Meta 43 (4): 653–659. Tytler, Alexander Fraser. [1797] 1978 [reprint]. Essay on the Principles of Translation. Amsterdam: Benjamins. Vinay, Jean-Paul and Jean Darbelnet. 1958. Stylistique Comparée du Français et de l’Anglais. Paris: Didier. Øverås, Linn. 1998. “In search of the third code: an investigation of norms in literary translation”. Meta 43 (4): 571–588.
13
Probabilistic explanations in Translation Studies Universals — or a challenge to the very concept?* Gideon Toury Tel Aviv University, ISRAEL […] the search for universals in every field of human activity constantly punctuated and qualified with the apparently unique […] constitutes a kind of universal. (Newman 1985–86: 73)
The following article presents a brief overview of three interconnected aspects of the quest for universals of translation: (1) the level where translation universals could be located, (2) the possible format of those universals, when formulated in language, and (3) the way to expose them, in view of this level and that format. It is claimed that universals shouldn’t be sought on either too low or too high a level, that their format should be probabilistic and conditioned rather than deterministic, and that they should be found and formulated by means of a combination of “top-down” and “bottom-up” research procedures. Finally, the question is tackled of whether the probabilistic formulations themselves (i.e. beyond the mere principle of being probable) would constitute universals, or whether they represent a challenge to the very concept.
1.
Searching for regularities in translation
The quest for universals is all but common practice among translation scholars. This is so even if Translation Studies as a scholarly discipline — i.e., a “conceptual framework of scientific research” — is taken to cover not every single occupation with translating and translation(s), from whatever point of view and for whatever purpose, but, more narrowly, only those activities “whose focus is translation”, as suggested by Daniel Gile in his contribution to the “Shared Ground” debate currently ongoing in Target (Gile 2001: 149, 151; emphases added). In fact, it is
16
Gideon Toury
almost the other way around: there are, always have been, and will no doubt always be many scholars — among them very able ones — who would value differences over similarities any time. Quite a few show suspicion of, if not complete hostility towards, the very idea of conducting a deliberate search for recurring phenomena (or, better still, patterns); especially when these so-called regularities threaten to transcend the issue under scrutiny at a particular point in time, let alone cases where such regularities are claimed to represent universals, i.e., to be there because it is translation we are talking about. Scholars adopting this stance claim a sweeping wish to put their finger on that which is unique instead: a particular textual-linguistic phenomenon (metaphors [of a specific kind], names, puns, allusions, whatever), a single text (selected at random or based on a particular criterion), a closed body of texts (put together on the basis of a [more or less] well-defined principle), a single text-type, an individual translator or “school” of translators, a specific historical period, a pair of languages/ cultures; in brief, anything that can be presented as a “single” entity, simple or complex, in terms of the level it purports to. Nor is there much point in waging a campaign against these “separatists”. It is not even difficult to sympathize with them. Haven’t we all been there ourselves, at one time or another? And haven’t we enjoyed ourselves immensely? At the same time, one cannot but wonder how they reckon one might determine what would justifiably count as “unique” unless one has some idea of what one’s immediate object of study shares with other possible objects. True enough, translation scholars as individuals have to start somewhere, and the first case one happens upon indeed tends to seem fraught with revelations.1 However, as one continues to plough through the study of translation practices or texts, certain phenomena will manifest themselves as more predictable than others, more often than not in line with what the “veterans” already know. A sense of déjà vu will thus arise which will gradually evolve into some notion of regularity. Widening one’s scope of interest, in terms of either the issues (or phenomena) addressed, or the number of instances of performance/texts considered, especially if controlled by some definable principle(s), is thus bound to contribute towards modifying one’s (evidently erroneous) first impression, until it is finally reversed. From this point on one is likely to experience very few surprises: almost everything, certainly everything of consequence, will have become highly predictable. No sense, then, in going to fight a battle which we, “non-separatists” — quantitatively, a minority among translation scholars — are unlikely to win anyway. Instead, I would suggest again (see e.g. Toury 1995: 266) a division of labor between approaches and individuals according to their different foci and interests. And what I would offer in lieu of excessive attempts to establish a “shared ground” for all translation scholars, which is somewhat artificial and not very promising anyway, is a large enough dose of tolerance to each other’s whims. For whims we all have!
Probabilistic explanations in Translation Studies
In this vein, I wish to proceed from the (whimsical? simply naïve?) assumption that we all acknowledge the existence of “regularities” in translation and share at least some basic readiness to give the notion of “universals” a fair shot. I will therefore say no more about possible justifications for the very pursuit of universals in the field of translation. On the other hand, I won’t submit to detailed scrutiny and analysis any specific candidate either which may have been suggested as a translation universal. After all, no disagreement concerning the qualification of any individual proposition as a universal can be perceived as invalidating the very concept of “universals” nor can it render the pursuit thereof unfounded, let alone illegitimate. Finally, I will attempt no classification of any kind: personally, I tend to regard pigeonholing as rather simplistic and — at any rate — I don’t think the time is ripe. What I will do is present a brief overview of three interconnected aspects of the quest for universals of translation behavior and some of the problems this quest may involve; namely, – – –
the level where such universals might be located in the first place (Section 2), their possible format, should they be found and in need of formulation in language (Section 3), and the way to go about establishing what the possible universals are (Section 4).
On the basis of these preparatory phases I will then venture to give a first tentative answer to the question in the subtitle (Section 5). Part of what I’ll be saying I’m already convinced of; another part is still tentative. I’ll be the first to admit that there is still a lot to be done before these issues can be resolved! Let us begin with the question of level, then.
2.
A matter of level
2.1 Not too low, not too high First of all, I feel quite safe in claiming that universals shouldn’t even be sought at too low a level. Above all, regularities pertaining to individual instances of behavior, even groups thereof that emerge as typical of a particular culture at a particular moment in time, should not be over-generalized — unless good reasons for such a generalization have been found and spelt out. To be sure, even if those phenomena are later found to have realized higher-level, more abstract regularities, the realizations themselves — i.e. that which lent itself to direct observation to begin with — would still count as norm-governed, at best, if not as mere idiosyncrasies (which tend to be structured too, but on the level of the individual alone).
17
18
Gideon Toury
On the other hand, there seem to be levels that are too high for the very hunt for universals, especially if the uncovered regularities are to be non-trivial, that is, to make a contribution to our understanding of translation and whatever it may involve. Thus — to me, at least — propositions such as “translation involves explicitation”, whose omnipresence in translation so many have tried to tackle in recent years (for a detailed overview see Klaudy 1998), are already suspicious, in that respect, whether their (inherently vague) formulation is given a “weaker” or a “stronger” interpretation. Thus, if “translation involves explicitation” is read as a claim for exclusiveness — that is, if the implication is taken to be that it is only instances of explicitation that will be exhibited in translation, to the exclusion of non-explicitation, let alone implicitation — then the statement is obviously false. In fact, it shouldn’t even be taken to imply that, in any individual case, more instances of explicitation than implicitation will necessarily be found. In other words, even cases where the complete opposite will have been shown might still be regarded as translations. By contrast, if “translation involves explicitation” is understood to simply state the fact that there are cases of explicitation in translation processes, resulting in an increase of explicitness in translated texts (alongside cases of non-explicitation and implicitation, that is), then my reaction would be to wonder whether we have gained any real knowledge, or understanding, by giving the proposition a verbal formulation. In fact, there may even be some risk involved here: if taken at face value, the implication may easily be that the two complementary phenomena — “explicitation” and “implicitation” — should be assigned the exact same status, which is somewhat counter-intuitive, especially when one probes the regularity of the one in relation to the other. As will soon become clear, it is precisely the gradation of their status that needs to be tackled, not any absolute figures indicating overall “distribution”. Obviously enough, the less indeterminate a statement, the easier it is to disprove it; and not only speculatively, on the level of initial possibilities alone (which is what I have been doing so far), but on empirical grounds as well; that is, in the face of factual evidence. Clearly, no more than a single instance of counterevidence is required to shake the universality of any such statement, and exceptions are not too difficult to find. Actually, the fact that instances which could serve as counter-evidence are quite easy to fabricate should be reason enough to reject such statements’ claim to unconditioned universality. After all, everything that can be fabricated at will can also be produced in authentic, socially relevant cases of translation, given the appropriate circumstances. Now, there is one important feature that all statements of this format share; namely, their predicates being realization of one underlying feature, that of a socalled shift from some “zero-point”, be it ever so imaginary, a feature which used to be presented as characterizing the (hypothetical construct of) “optimal translation”.
Probabilistic explanations in Translation Studies
Explicitation, implicitation, simplification, complexification, addition, deletion, foreignization, domestication, etc. etc., are all shifts, in this crude sense.2 The ensuing question seems obvious enough: What would be the status of an underlying (and hence higher-level) generalization like “translation involves shifts”? My claim would be that we have stepped into the domain of analytic statements, if not tautologies. Thus, unlike lower-level generalizations like “translation involves explicitation”, the proposition “translation involves shifts” is no doubt true. However, its truth is “by definition”, so to speak. This is to say, the regular occurrence of shifts in translation(s) is not something we happened to uncover in our empirical studies, much less so by working in a “bottom-up” manner. Rather, it forms an integral part of the very concept of translation; no less than, e.g., “translation involves memory” (where the predicate represents a cognitive variable), or “translation is greatly influenced by norms” (where the predicate is socio-cultural in nature). Like them, it may have any number of different surface realizations in actual performance, but research projects tend to limit themselves to the level of observables. On occasion, they may tackle the way the underlying categories are realized, but the categories as such are hardly approached: They normally serve as little more than accepted points of anchoring. It is not that there is anything wrong with tautologies as such. In fact, many may be quite instructive, even helpful; in teaching contexts, for example. The point is only that, in using them, nothing new is added to our understanding of translation as such, and this quest for understanding, after all, lies at the heart of the very search for universals. To sum up, even though, on the face of it, statements like the ones cited in the last-but-one paragraph have often been offered as a universal (see a concise presentation in Laviosa-Braithwaite 1998), I would suggest that we continue our search and look for more meaningful, more informative universals. It is not that the suggestions made thus far are wrong in any way; it is only that they seem insufficient for the goals we wish to pursue. 2.2 The in-between domain If indeed translation universals should be neither too concrete nor too abstract, there is only one option left short of giving up the whole endeavor; namely, to conduct our search somewhere in-between the specific and the general. Does such an in-between domain have any existence at all? — This is what I’ll try to establish next. I will start with an old quotation which would add a historical dimension to the discussion: By virtue of their definition as shifts from one focal point, […], all shifts fall into dichotomous pairs […]. For instance, explicitation/implicitation,
19
20
Gideon Toury
addition/loss of information, generalization/concretization, etc. Consequently, it is possible to formulate the following rule: If a shift occurs, it necessarily occurs in one direction — or in its complete opposite. The selection of one of the two options, which in themselves are given, to the extent that it is ordered, is governed by translational norms. (Toury 1987: 6)
This is a 1987 translation of something I wrote in Hebrew more than ten years earlier, in my doctoral dissertation (book version: Toury 1977: 32) — a reservoir of half-baked ideas which may be worth returning to, from time to time. In those days, my main concern was to adapt the notion of “norms” to the specific requirements of TS and apply it to one particular case, and I failed to pursue my own observations in any other direction. Only recently did I realize that they had something to offer in terms of a possible interim area where meaningful, non-trivial claims for universality could maybe be made.3 Obviously, once the (quasi-)novelty of the overall claim that “translation selections are governed by norms” underlying my earlier attempt had worn off, the proposition itself became analytical again (see Section 1 above). Pretty much the same happened to ideas such as “translation involves shifts”, which had been presented as a near-revelation in the late nineteen-sixties (e.g. Popovicˇ 1970), or “translation entails manipulation of the source text” of the mid-eighties (e.g. Hermans 1985). It is the idea that translational shifts come in pairs, so that any selection is between modes of behavior which would yield shifts of a diametrically opposite nature, which should have been pursued, but wasn’t, along with the nonrealizability of the problematic “zero-point”. Several years later I returned to this hitherto overlooked issue; namely, when I began wondering where the study of norms was going to lead TS (e.g. Toury 1995: Part Four), or, more generally, “What lies beyond Descriptive Translation Studies” at large, as the title of one of my more recent papers reads (Toury 1997). I soon realized that one important goal might be the advancement of the theoretical branch of our TS beyond the initial stage where it can only account for potentialities and not actual behavior, not even the stronger vs. weaker tendencies it may show. Such an advancement, I claimed, could only be achieved gradually, through conscious attempts to link together different modes of translation behavior and a vast array of heterogeneous factors, which would be found to have a capacity to bring about the selection (or avoidance) of a particular kind of behavior and the ensuing regularities of the existence/non-existence of certain features in translated products (and/or in the relationships between those products and their sources); be the determining factors cognitive, cross-linguistic, textual, text-typological, sociocultural (including all kinds of ideology), or whatever, and be their influence on the translator’s behavior translation-specific or not. These linkings of modes of translational behavior and particular determining factors is what I have normally referred to as laws in the last few years (e.g. Toury 1995: Part Four).
Probabilistic explanations in Translation Studies
3.
The format of an explanation in TS
Needless to say, such a mode of reasoning hardly bears deterministic statements, tying modes of behavior with “determining factors” in a 1-to-1 relationship. After all, there doesn’t seem to be any single factor which cannot be enhanced, mitigated, perhaps even neutralized by the presence of another; the more so as, in actual reality, there will always be more than just two variables interacting, and hence influencing both each other and the selected behavior as a whole. Thus, a translator is male or female, older or younger, more or less experienced, more or less tired, under greater or lesser time-pressure, translating into a strong(er) or weak(er) language of his/hers, well- or less-well paid, and so on and so forth, all at once, not one at a time. Unfortunately, at this point, we don’t have so much as an exhaustive list of possible determining factors. Even less can we say with any amount of certainty what factors have a stronger or weaker influence on translational behavior (in themselves, so to speak), let alone how different factors interact. I suggest that, rather than being deterministic, and hence relatively easy to refute, the appropriate format of any candidate for universality should be both conditional and probabilistic.4 The aim should be to turn translation theory into a system of interconnected, even interdependent, laws of such a format, but it is certainly premature to say what the system might look like. By contrast, the format of individual theorems can, and should be addressed, even if only tentatively. Thus, the basic format of an explanatory statement that is both probabilistic and conditional could be of the following kind: If A and/or B, and/or … Y, then there is great likelihood that Z (or: small likelihood that minus-Z)
where the letters A-to-Y represent an array of possible variables and Z — the behavior in question and/or its result and/or the shift caused by its execution. (Minus-Z represents the opposite, complementary option.) A variation, which is probably easier to use, could be: The presence of A and/or B, and/or … Y, enhances the likelihood that Z (or: reduces the likelihood that minus-Z)
For example, The coincidence of lack of experience (variable A) and fatigue (variable B) increases the likelihood that translation procedures will be applied to small and/or low-level entities (behavior Z) (or: reduces the likelihood that they will be applied to long and/or high-level entities)
21
22
Gideon Toury
(Note that I am making no claim to validity here. The last passage was intended as an example of the format only!) To be sure, even this presentation is not elaborate enough for our purposes, as it reflects too linear a way of reasoning: the implication is that variables are taken up one by one and ordered consecutively, like discreet beads on a thread, as if each one of them were operating with complete independence from all other variables and hence it can only be put next to the one preceding it, to be followed by another variable, and so on and so forth. A better formulation would certainly be something like the following: If A and B, then the likelihood that Z is enhanced by the presence of C; in extreme cases, C can offset either A or B (or both)
The beginning of a concrete example might be: If a translator is both inexperienced (variable A) and tired (variable B), the likelihood that translation procedures will be applied to small and/or low-level entities (behavior Z) is enhanced if the translator’s culture is tolerant towards the results (variable C). In fact, when such tolerance exists, there is no real need for A and/or B to be present in order for it to be activated.
This kind of formulation — were it found to be valid — reflects the assignment of a rather high regulative power to cultural tolerance/intolerance of [the observable results of] translation activities as possibly different from non-translational ones, which seems to concur with the findings of many empirical studies: indeed, sociocultural variables often take precedence over factors of other kinds.5 This possible methodological gain is a clear outcome of the very search for generalizations that are probabilistic and conditional, and would hardly have emerged, had a different kind of method been chosen. One advantage of formulating statements of this kind on a level which lies in between the particular and the general is that they are likely to allow systematic accounts, including elaborate explanations, of low-level regularities, and accounts which are all of the same kind, at that. They also offer the possibility of predicting what the findings would be in a particular case, provided the appropriate determining factors will have been identified and brought to bear on the study; at least when it comes to “backward” predictions, those made with respect to translation acts that have come to their end (or translated texts which already exist) but haven’t yet been submitted to study. Finally, should a certain behavior (or its observable result) emerge as different from the one predicted, it would be possible to account for the apparent deviation without throwing the baby out with the bath-water; that is, without needing to discard the methodological basis for the (frustrated) expectation. I presume that, more often than not, the account will involve adding to the list variables which weren’t taken care of before, or refining the distinctions between
Probabilistic explanations in Translation Studies
different realizations of variables that were; either way, a ramification rather than a complete change. To be sure, processes of ramification can easily go on and on incessantly, which is a major way of elaborating the underlying theory.
4. How would an explanation be established? Supposing we agreed to try and advance translation theory in this way, how would we arrive at the required probabilistic-conditional statements? Above all, how would we be able to unearth the variables which influence translation behavior and its observable results, and how would their relative relevance and power be determined, alongside the ways they act upon each other? It is too early to venture any detailed answer here, and, in any case, reaching the final destination, even if there is one, seems to me less intriguing than the process of moving along the road which leads there. As to the road itself, it seems clear that some combination of “top-down” and “bottom-up” movements will have to be involved. Thus, existing, mostly speculative hypotheses formulated within translation theory will be examined against instances of real-world behavior. On the other hand, and at the same time, a series of empirical studies, both observational and experimental,6 will be conducted, moving gradually, and in as controlled a way as possible, from individual phenomena to culture-specific norms to the exposure of more and more general regularities on higher and higher levels, always together with the factors that are deemed responsible for their occurrence/non-occurrence. As I see it, descriptive-explanatory studies therefore stand to play an evergrowing role in our discipline. Not only in and for themselves, as a means of satisfying our curiosity with respect to one or another low-level issue, but also as a way of putting high-level hypotheses to the test as well as coming up with new ones, which would then have to be submitted to confirmation again, and again, in the kind of helical movement of refinement which is so typical of the progression of an empirical science. Luckily enough, there will always be new kinds of evidence to process (had anybody considered TAPs before the 1980s?), as well as new procedures to apply to them as well as to the old data (could anybody have thought of Corpus Studies before the invention of the computer?). As some philosophers have put it, “[…] confirmation is a tripartite relation, holding between the hypothesis, the test procedure, and its outcome” (Weintraub 1988: 176–177).
5.
Universals — or a challenge to the concept?
There remains one last question: Can we claim that — with the idea of probabilistic explanations of regularities encountered in translation(s) — we have hit upon our
23
24
Gideon Toury
coveted universals, or must we conclude that, if all regularities in the field of translation are indeed conditional, and only more or less probable, the whole quest is pointless? In other words (to paraphrase the subtitle of the present paper), the probabilistic formulations themselves (i.e., beyond the undeniable universality of the very principle that translation regularities can only present probabilities), would they constitute universals — or does this nature of theirs present a challenge to the very notion? Any attempt to answer this $64,000-question would require a lot more conceptual work on the notion of “universal” and the demands we would like to set for it in its application to translation. For me, the issue is not one that could be resolved by saying either “yes” or “no”. Rather, it is a matter of the gains vs. the losses involved in the way the notion is used. And I wouldn’t be at all surprised if we were to end up not only with different universals, but even with universals of different kinds, on a variety of different interim levels. When this is achieved, we will also be in a better position to decide whether the term “universal” should be retained or perhaps replaced with another term, less demanding and better suited for the phenomena we are dealing with. Perhaps one of the good old notions such as (the descriptive) “high-level regularity”, even (the explanatory) “law”, could be charged with a new meaning. What we should always remember, though, is that the terminological issue is only a minor one, and what we should first solve are the conceptual problems we still seem to have.
Notes * A more developed (and slightly updated) version, focusing on somewhat different issues, was presented in the Conference “Translation Universals — Do They Exist?” held in Savonlinna in October 2001, and will be included in the Proceedings. 1. Think of your students, or of yourselves as students! At times, almost everything one encounters may appear to be truly revelatory. (Good teachers certainly capitalize on this when introducing novices to TS, to the extent that they know how to evoke in their students a crucial sense of “aha”.) 2. The notion of “shift” itself will be left unspecified in order to avoid the controversial issue of how, or in respect to what, a shift may be identified and/or measured, as any involvement in the controversy is bound to take us way off course. 3. The quotation is from Section 2.8 of the book version, entitled “Rules, Universals of Translation Behavior and Translational Norms”; and it was “Shifts” which were tentatively presented as universal. Since the early eighties, I have refrained from using the word “universals” and used “laws” instead. I have only now returned to “universals”, because this was the term used by the organizers of the session. However, I won’t insist on this label. (See concluding paragraph.) 4. Here I have accepted — tentatively, at least — Andrew Chesterman’s argument that these are two different characteristics that do not imply each other. 5. This inference finds further corroboration in the emergence of the translation editor in individual cultures, an institution which seems to become necessary as soon as a (target) culture
Probabilistic explanations in Translation Studies
loses its tolerance of the results of the translation process as such. Thus, one of the editor’s major tasks is to eliminate instances of interference, especially of the “negative” type occurring on lower linguistic levels. This task definition reflects not only a growing intolerance (i.e. a negation of the cultural variable C), but also an understanding of the fact that, as long as variables A, B, and others like them, are still present, interference is going to persist. To be sure, in most extreme cases, an editor may go over a translator’s output as if it were a regular text in the target language and never even check it against the source text (which would still be regarded as the translator’s responsibility). See now Mossop 2001. 6. Here I follow Gile’s useful distinctions (1998: 70–71).
References Gile, Daniel. 1998. “Observational studies and experimental studies in the investigation of conference interpreting”. Target 10(1): 69–93. Gile, Daniel. 2001. “Being constructive about shared ground”. Target 13(1): 149–153. Hermans, Theo. 1985. “Introduction: Translation Studies and a new paradigm”. In The Manipulation of Literature: Studies in Literary Translation, T. Hermans (ed), 7–15. London and Sydney: Croom Helm. Klaudy, Kinga. 1998. “Explicitation”. In Routledge Encyclopedia of Translation Studies, M. Baker (ed), 80–84. London and New York: Routledge. Laviosa-Braithwaite, Sara. 1998. “Universals of translation”. In Routledge Encyclopedia of Translation Studies, M. Baker (ed), 288–291. London and New York: Routledge. Mossop, Brian. 2001. Revising and editing for translators. Manchester, UK & Northampton, MA: St Jerome. Newman, Aryeh. 1985–86. “Translation universals — Perspectives and explorations”. In Translation Perspectives III: Selected Papers, 1985–86, M. Gaddis Rose (ed), 69–83. Binghamton: State University of New York at Binghamton. Popovicˇ, Anton. 1970. “The concept ‘shift of expression’ in translation analysis”. In The Nature of Translation: Essays on the Theory and Practice of Literary Translation, J. S Holmes, F. de Haan and A. Popovicˇ (eds), 78–87. The Hague — Paris: Mouton & Bratislava: Publishing House of the Slovak Academy of Sciences. Toury, Gideon. 1977. Translational Norms and Literary Translation into Hebrew, 1930–1945. Tel Aviv: The Porter Institute for Poetics and Semiotics, Tel Aviv University. [Hebrew] Toury, Gideon. 1987. “Integrating the cultural dimension into Translation Studies: An introduction”. In Translation across Cultures, G. Toury (ed), 1–8. New Delhi: Bahri. Toury, Gideon. 1995. Descriptive Translation Studies and beyond. Amsterdam and Philadelphia: John Benjamins. Toury, Gideon. 1997. “What lies beyond Descriptive Translation Studies, or: Where do we go from where we assumedly are?”. In La Palabra Vertida; Investigaciones en torno a la Traducción, M.Á. Vega i R. Martín-Gaitero (eds), 69–80. Madrid: Editorial Complutense. Weintraub, Ruth. 1988. “A paradox of confirmation”. Erkenntnis 29. 169–180.
25
A thousand and one translations Revisiting retranslation Outi Paloposki and Kaisa Koskinen Research Institute for the Languages of Finland / University of Tampere
The Retranslation hypothesis — the claim that first translations are more domesticating than retranslations — is often referred to in Translation Studies literature but only in passing, without looking at the issue in great detail. On the other hand, many of the studies that explicitly deal with first translations and retranslations and would thus be ideal for testing the hypothesis do not mention it. Thus, there seems to be no substantial body of evidence either in support of or against the retranslation hypothesis. This article aims at taking a closer look at the hypothesis and comparing its claims with data from Finnish first translations and retranslations. It is suggested that many different factors, not just the order of appearance, affect the profiles of these translations.
Introduction In his preface to the special edition of Palimpsestes, dealing with retranslation, Paul Bensimon (1990: ix) claims that there are essential differences between first translations and retranslations. First translations, according to Bensimon, are often ‘naturalizations of the foreign works’ (naturalisation de l’oeuvre étrangère).1 They are ‘introductions’, seeking to integrate one culture into another, to ensure positive reception of the work in the target culture. Later translations of the same originals do not need to address the issue of introducing the text: they can, instead, maintain the cultural distance. In the same issue of Palimpsestes, Antoine Berman (1990: 1–7) outlines his ideas of retranslation as a way of or space for accomplishment. First translations date; hence the need for new translations. The position of these two scholars, Bensimon and Berman, is briefly what constitutes the basis for ‘the retranslation hypothesis’ (RH), as we understand it here. It is formulated most explicitly in an evaluation by Yves Gambier (1994: 414): “[…] une première traduction a toujours tendance à être plutôt assimilatrice, à réduire l’altérité au nom d’impératifs culturels, éditoriaux
28
Outi Paloposki and Kaisa Koskinen
[…] La retraduction dans ces conditions consisterait en un retour au texte-source.” (“[…] a first translation always tends to be more assimilating, tends to reduce the otherness in the name of cultural or editorial requirements […] The retranslation, in this perspective, would mark a return to the source-text”, emphasis in the text). It is conceivable and even to be expected that many translations indeed follow this schema. It is, after all, based on intuitive observations if not thorough empirical groundwork. The reasons behind it, too, seem plausible. If a text seems extremely foreign or alien in the receptor culture or to the translator him/herself, the result often is a domesticated version. The translator may share his/her audience’s unfamiliarity, or at least acknowledge it, wanting to produce a text that is comprehensible to the readers. Later translators can benefit from increased familiarity with the source culture. It seems that it is these kinds of cases that the retranslation hypothesis is based on. Because of their obvious comparative and contrastive possibilities, retranslations are an attractive topic for research.2 For example, Lefevere (1992), Du-Nour (1995), Kujamäki (1998; 2001), Tymoczko (1999), and Oittinen (1997; 2000) all deal with retranslation. RH, however, has not been taken up explicitly or tested extensively in any of these studies. Instead, it surfaces elsewhere, often as an aside. Taking up the issue, Andreas Poltermann (1992: 18) attributes the phenomenon of domesticating first translations to genre expectations within a target system, and Theo Hermans (1999: 139–140) expands on this explanation. RH can also be used as a paradigmatic example of a hypothesis: Andrew Chesterman (2000) splits RH into subcategories to illustrate how different models in translation studies accommodate hypotheses. Thus, while RH clearly provides a wealth of issues to work on, it quite as clearly needs to be examined in an empirical setting. While there seem to be several things tangled up in the discussion, at least two major issues can be discerned. The first concerns the reason for retranslations. Do first translations really ‘date’, always?3 Or is it that domesticating first translations date, creating a need for foreignizing retranslations (to test this, one would actually have to look at cases of non-retranslation as well)? Gambier (1994: 413–414) proposes a number of starting points for a further study of retranslation: do translations ‘date’ because our knowledge of the source languages (or cultures, one would like to add) has increased, or is retranslation a question of reinterpretation? The first alternative would seem more in line with RH, whereas the second option turns our gaze towards other solutions. Anthony Pym (1998: 82–83), too, presents some explanatory hypotheses about the reasons for retranslation, ranging from different pedagogical functions of texts to rivalry in the possession of the knowledge contained in the document to be translated. The second issue deals with the profiles of first and retranslations, and is, of course, linked to the first issue of reasons behind retranslation: do first translations tend to be more domesticating, and retranslations more foreignizing? According to
A thousand and one translations
RH, they do: assumedly, there is something inherently causative in the fact that a translation is a retranslation. A different perspective may be needed to distinguish other variables that bear on the issue of retranslation. In order to pay close attention to the context where retranslations appear, we will take up case studies of specific translations. Our aim is both to look at the actual profiles of some first and retranslations, and to try and uncover the reasons why particular texts are considered either worthy or in need of retranslation, looking also at the causes that determine how they are retranslated at particular times and in particular contexts.
Finnish 19th-century translations: Balancing between comprehensibility and foreignness (OP) To start with the intuitive attraction of RH, let us first examine to what extent the early Finnish data support it, before going to contrary examples. An earlier study on translations into Finnish during 1809–1850 (Paloposki 2002) seems to confirm the claims in RH, at least partially. As this era saw the beginning of literary translation into Finnish, all translations of fiction were first translations. Also, they were often more or less domesticated. This kind of a ‘domesticating phase’ is not uncommon in early literary development: Gideon Toury (1995: 132–133) has observed that translations tend to be unmarked in this situation. Unmarkedness is often associated with domestication: translations are indistinguishable from same-language original works. It would only be natural to expect subsequent translations to mark a return towards the source texts. However, we must bear in mind that the ‘adaptation stage’ is but a phase in the long history of translations, and there are first translations into a language during other times as well, probably with very different profiles. Thus, RH may apply during an initial stage in the development of a literature, but not to all individual first translations: domesticating first translations may be a feature of a phase in a literature, not of translation in general. Let us now turn to some first translations and retranslations of a later stage, the second half of the 19th century and early 20th century: different versions of Oliver Goldsmith’s The Vicar of Wakefield and of the Thousand and One Nights, both of which exist in several subsequent translations into Finnish. The Vicar of Wakefield, first published in 1766, was one of the most popular books among the ‘ordinary people’ in England (Rose 2001: 120). It was also the sole representative of the English 18th century novel in Finland for decades.4 In the following, the first two Finnish translations (of 1859 and 1905) are studied from the point of view of the linguistic profiles of the texts: syntax and lexicon (Bensimon 1990: ix–x). The focus is on how ‘closeness’ works on different levels of the text, and on the different causes behind textual closeness.
29
30
Outi Paloposki and Kaisa Koskinen
The first translation appeared in 1859. The translator, Gustaf Erik Eurén, was a prolific writer, grammarian, teacher, newspaperman and translator — nothing unusual at the time. Written Finnish was still very heterogeneous, but Eurén’s command of the language was generally highly esteemed: his grammar of the Finnish language was praised and recommended in the newspaper Suometar twice, in 1852 and again in 1860. Eurén’s translation was reviewed favourably in several newspapers, although critical attention was paid to the large number of personal pronouns and to a Swedish influence in the lexicon.5 The second translator, Samuli Suomalainen, was an experienced translator. His translation activity is set in a specific context: much of his work was targeted at younger audiences. He also adapted for young people Seitsemän Veljestä (Seven Brothers), an original work in Finnish by Aleksis Kivi, at a time when adaptation of original works was not yet customary (Kiuru 2002: 331). Thus, Suomalainen’s ideas of textual reworking tended towards adaptations more than his predecessor’s. Both the syntax and the lexicon in Eurén’s translation are closer to the original than in Suomalainen’s. This literalness is even more striking, considering the fact that between Goldsmith’s original and Eurén’s translation there was the anonymous Swedish translation of 1788. Even the morpho-syntax sometimes penetrates through the different languages, as in the opening sentence: Goldsmith 1766: I was ever of opinion, that the honest man who … Anon (Swedish): Jag har altid warit af den tankan, at en hederlig man… I have always been of the thought Eurén 1859:
minä olen aina ollut siitä ajatuksesta, että … I have always been it (pron + CASE elat.) thought (CASE elat.)
The elative case is a direct equivalent of the English and Swedish prepositions of and af. There are no articles in the Finnish language, and Eurén’s version thus replaces the definite article with a pronoun. The second Finnish translation departs from this morpho-syntax and tends towards a more fluent and idiomatic version, where partitive is used: Suomalainen 1905: Aina minä olen ollut sitä mieltä, että kunnon mies… that opinion (+ CASE partit.) There is also lexical (and stylistic) closeness/freedom in the texts. Whereas Eurén’s translation replaces unfamiliar concepts or words not existing in Finnish by borrowings — quatuor for quatre, dus äss for deuce ace, for example — Suomalainen has simply omitted the whole passage, whether out of concern for understandability or because of his own stylistic preferences, it is hard to know. The literality of the first translation may be a sign of a changed attitude towards translation after the initial adaptive stage, with more space for introducing foreign
A thousand and one translations
texts as such. This is indeed what RH would predict, but not in the case of first translations but of retranslations. This fact strengthens the point (alternative new hypothesis) that domestication may be a feature of a first phase in literature, not of first translations as such. Another reason for literality in this specific translation may be the translator’s idiosyncratic view of translation, or his background as a grammarian. Likewise, lexical and syntactic closeness to the source text can be a feature of a first translation. Idiosyncratic constraints — the translator’s own preferences, or even difficulties in interpreting the text — may have a role to play. The second case, The Thousand and One Nights, challenges the very idea of one single original that could be returned to, more or less closely. The Thousand and One Nights is, in several senses, a multi-layered collection — a sea, not a book, in the words of the most recent Finnish translator, Jaakko Hämeen-Anttila (2001: 7). Chronologically, it consists of different layers of Indian, Persian and Egyptian elements from the 10th to the 14th century (Gerhardt 1963: 9).6 Also, the manuscript tradition is varied. There exist several manuscripts of the stories, including a fragmentary page from a 9th-century version. The Frenchman Antoine Galland used a Syriac manuscript from the 14th-century and complemented his work with the help of an oral informant, to write down his translation of the 1001 nights, which was the one that made the stories famous in the West. Several translations into different languages were made out of Galland’s translation, and an additional number of translations appeared after the manuscript work of the 19th century. Jorge Luis Borges (2000) discusses some of these translations and observes that whereas the Englishman Edward Lane in the early 19th century translated against Galland, another Briton, Richard Burton, translated against Lane half a century later. This was no return to the original but an attempt at being different (Borges 2000: 34). They were all regarded as ‘the most faithful’ of translators in their time, but their faithfulness did not consist of an increased measure of fidelity to the source text, but of something that their contemporaries regarded as the most important facet of fidelity: in Lane’s case, it was the scientific rigour, in Burton’s, an inclusion and expansion of the erotic. The first Finnish translations appeared in newspapers starting from 1831, and the first book appeared in 1855. After this, there were volumes in 1856–57, 1874–78 (for children), and 1878–1880. At this time, a number of different translations in Swedish and German, which were still the most widely used transfer languages for books originally written in third languages, were available for the Finnish translators, but this wealth of sources is hardly ever mentioned. The minutes of the Society for Finnish Fiction (Suomalaisen Kauniskirjallisuuden Yhdyskunta) for 3.3.1849 mention the decision to translate the Thousand and One Nights as soon as possible and to acquire a suitable text, preferably in German. It was also noted that August Ahlqvist had promised to undertake the task. Ahlqvist’s translation came out, first
31
32
Outi Paloposki and Kaisa Koskinen
in the newspaper Suometar in 1853 and then as a book in 1855 (not in the Society’s series, though: censorship intervened, and the society was never given permission to function). Different originals were used, different stories selected: partly due to availability, partly, obviously, to audience considerations. In the prefaces and introductions (where they existed) it was emphasized that these tales were internationally known, they were part of universal folklore heritage, like the Finnish folktales; they were beautiful, enticing, educational.7 The very first published tale, that of the ropemaker Hassan, from Galland’s selection, appeared at regular intervals. Its first appearance was in the newspaper Oulun Wiikko-Sanomat in 1831, then in Ahlqvist’s translation in the 1850s, and again in the children’s version of 1874–78. The rope-maker’s tale exhibits relatively little variation throughout the decades. Lexical variation is greater in some of the later translations as some of the concepts have become more familiar, such as ‘turban’ (lakki > turbaani), but otherwise there is no evident dissimilarity between the texts. Both the 1831 version and the 1874 version give glosses (e.g., ‘our Prophet’ — ‘Mohammed’ in 1831; ‘Allah’ — ‘the God of the Musulmans’ in 1874). The 1855 version does not have any explanations or commentary, not even an introduction. It seems that the first attempts at introducing something so ‘foreign’ were exotic enough (with oriental names, customs and realia); there was no need in the span of a few decades to foreignize in later versions. Domestication mainly consisted of the selection of certain originals, and of glosses, which could also be interpreted as signs of foreignization (attempts at increased accuracy). The 20th century saw new translations; the study of these will undoubtedly change the picture again. Measuring domestication and closeness is problematic, since they may work on different levels of the text simultaneously. Similarly, source-text orientedness is a complex issue where different motivations and constraints on the translator may coincide, thus producing a text where it is difficult if not impossible to judge whether an individual lexical item, a calque for example, is an instance of deliberate strategic choice or simply of a lack of suitable alternatives. Apart from exploring the causes behind the different profiles of translations, the analysis of early Finnish translations helps to understand why RH has arisen: there certainly are domesticating first translations in Finnish texts as well. It is however argued, with the help of the Finnish data, that a domesticating first stage can be more easily associated with certain phases in the development of a literature rather than with individual translations.
A thousand and one translations
Alice’s Adventures in Finland: From domestication to foreignization and back (KK) Riitta Oittinen’s study (1997) of the three different Finnish versions of Lewis Carroll’s Alice’s Adventures in Wonderland (orig. 1865) illustrates the assumption of a literature going through different phases, and it is also a good example of the attractiveness of the retranslation hypothesis. Oittinen does not discuss the theoretical problem of retranslations, but her findings are nicely in line with RH: the first translation by Anni Swan in 1906 is domesticating (Oittinen describes it as ‘cannibalistic’), the second by Kirsi Kunnas and Eeva-Liisa Manner in 1972 is free (‘carnivalistic’), and the (then) newest one by Alice Martin in 1995 is foreignizing (‘postmodern’). Oittinen’s labels are debatable, but her interpretation is easy to agree with. It is also understandable. Keeping in mind the relatively short history of Finnish literature, and the prevalent domesticating tendencies during the early phases, and also taking into account the scarcity of overtly domesticating contemporary translations, this is a likely outcome of many contemporary comparisons between a recent (re)translation and a noticeably older version. This is also Oittinen’s point. She sets the translations against their own time, and their particularities are largely explained by the temporal difference (1997: 119–139 et passim.; see also Bensimon 1990: ix). The argument thus runs as follows: as opposed to the era of the first translation, the Finland of the 1990s is sufficiently postmodern and globalized, and the readers sufficiently familiar with the British way of life, to accept a foreignizing translation. The emblem of this new attitude was the fact — also reflected in the title of Oittinen’s study Liisa, Liisa ja Alice — that in the 1995 version Alice remained Alice also in Finnish, whereas previous translators had renamed her Liisa. However, developments that took place after the publication of Oittinen’s book complicate the picture. In 2000, a new translation of Alice by a different publisher (Otava) and a different translator (Tuomas Nevanlinna) was published. In it, Alice was again finnisized as Liisa. Similarly, comparing the solutions that Oittinen takes up in Alice Martin’s version for their foreignizing tendencies, one finds that Nevanlinna often opts for more domesticating strategies. In other words, within a time frame of five years we have two retranslations, one of them clearly far more foreignizing than the other. It is unlikely that times would have changed that much (cf. Gambier 1994: 413). Interestingly, in spite of their different degrees of domestication, in a review of Nevanlinna’s translation (Oittinen 2001) these two recent translations are singled out as the two most ‘exact’ and faithful rerenderings of the original classic. There are two conclusions one can draw: either it appears that while postmodern translation theories advocate emancipatory and radical translation strategies, the actual translations of our postmodern era follow a directly opposite tendency, or it may be that the perceived ‘fidelity’ of the translation depends on
33
34
Outi Paloposki and Kaisa Koskinen
how closely it resembles our own interpretation of the text, and contemporary translations thus appear to be more faithful than previous ones that address different audiences and meet different expectations. Or both. To complicate the picture further, the existence of any retranslation cannot be straightforwardly attributed to assumed datedness of the previous versions: Kunnas’s translation from 1972 was reprinted by Gummerus in 1996; and Anni Swan’s translation from 1906 was reprinted in 2001 by WSOY (the same publisher also published Martin’s translation in 1995). Instead of a temporal succession from a domesticating first translation to a more “accomplished” foreignizing version we thus now have four coexisting interpretations of Alice, with varying degrees of domestication. The retranslation hypothesis does not sufficiently explain the developments. The reasons behind the profiles need to be sought elsewhere as well. It may well be that the publishers’ decisions and translators’ strategies were not primarily motivated by concerns related to the source text or previous translations. Perhaps the most formative influence for the latest translation are its illustrations by Helen Oxenbury (see also Oittinen 2001). They direct the text to a child audience, and children’s literature is also the genre the book is related to in the blurb. This orientation may have had a bearing on both the selection of the translator (Nevanlinna is wellknown for his Roald Dahl translations) and the translator’s choice of strategies. Alice Martin’s translation, on the other hand, is oriented towards adult readers (Oittinen 1997: 133). In the blurb, it is described as a ‘classic’ (with no reference to children’s literature), and its appeal to adults is emphasised. This orientation has no doubt affected the strategies. Commenting on her work, Martin has also brought up the aim to retain quotability in her translation, and the norm of harmony between translation and illustration (Martin 2001). A further explanation of her deliberate choice of a “faithful” attitude is that apart from her all other Finnish translators of Alice are well-known writers and/or poets who are typically expected to take more liberties (similar to RH, this is a statement that would need extensive empirical testing). Martin, on the other hand, wanted to prove how far one can get with the “boring old” translators’ virtues of careful reading, sympathetic understanding and fidelity (Martin in Oittinen 1997: 137) — an indication that postmodern translation theories and practices may have different agendas. On a more personal level, it may not be far-fetched to attribute her choice of using the English name ‘Alice’ instead of the ‘Liisa’ of all other Finnish translations to her own name. To sum up, the retranslations are affected by a multitude of factors, relating to publishers, intended readers, accompanying illustrations and — not least — the translators themselves. These are not adequately covered by the retranslation hypothesis.
A thousand and one translations
Pentti Saarikoski’s version of the Gospel according to Saint Matthew: The political and the personal (KK) Some cases of retranslation follow the schema of RH, some fit in uneasily, but for some RH is completely dysfunctional. The (re)translation of the Gospel according to Saint Matthew (Evankeliumi Matteuksen mukaan) by Pentti Saarikoski in 1969 belongs to the latter group. While his translation was (among other things) produced as a response to the (then) existing official version authorized in 1938 (another retranslation!), his explicitly stated aim was not to reduce the amount of domestication but rather to increase it. In an interview he stressed that he wanted to produce a text that would be easily understandable for everyone, written in colloquial and contemporary language instead of the scholastic and alienating style of the official translation. To use Schleiermacherian phraseology, he wanted to bring the text to the reader, not the other way round. However, in this particular case the issue of domestication versus foreignization is not a central theme. There are several other causes that give more plausible answers to the questions why and how this retranslation came into being. First, Saarikoski’s translation needs to be set in its context in the cultural climate of the 1960s. The translation has obvious affinities with its contemporaries, such as Pier Paolo Pasolini’s movie version of the same gospel and Clarence Jordan’s “Cotton Patch” Bible. Saarikoski produces a profane and political translation of the sacred text. He portrays the gospel as a communist manifesto, and the Marxist interpretation follows the trend of the generation. In analyzing this particular retranslation, the main issue is not its relation to the previous translation (in any case, Saarikoski’s version was made to supplement, not to surpass its predecessor). It is far more revealing to study the translation against its own historical context. A second, and even more important cause, however, is the translator Pentti Saarikoski himself. A poet and a translator, a youth idol and a cultural radical, and an enfant terrible of his time, Saarikoski was a public figure. In his translations the translator is never invisible. Neither did he differentiate between his two roles as a writer and a translator, but the two constantly fed on each other. Evankeliumi Matteuksen mukaan thus needs to be understood as a part of a greater whole (that of his whole oeuvre). There are also obvious parallels to his private life: the translation project feeds on both the devoted Christian faith of his youth (that he later abandoned) and the open political convictions of his adult years. The text is, in fact, the Gospel according to Pentti Saarikoski. This case, in other words, leads us back to the translator. The important question to ask seems to be “who is the translator?” (see Berman 1995; Chesterman 2000).
35
36
Outi Paloposki and Kaisa Koskinen
Conclusions The retranslation hypothesis, thus, arises out of particular historical observations. In a particular case it may, for example, be explicable by the current phase in literature. It can also be explained on the basis that it is normally not possible to be more foreignizing (or accurate in relation to the source text) in the translation than what will be understood in the target culture. Consequently, when information about this specific foreign culture or text increases, new possibilities are opened up for translation as well. However, as we have shown, RH only covers part of the ground of all retranslations: while there are numerous (re)translations that fit in the RH schema, there also exist several counter-examples where the schema is turned the other way round, and also cases where the whole issue of domestication/ assimilation versus foreignization/source-text orientation is irrelevant. It is possible to desire that a retranslation be more target-oriented, and it is also possible to use a foreignizing strategy in a first translation. In the latter case there are at least three possible outcomes: the source culture may be familiar to the readers through means other than previous translations of the same text, foreignization therefore not constituting an undesired alternative; foreign elements may be explained in a preface or in footnotes; or much of the contents of the translation may be left — deliberately or not deliberately — unclear to the audience. There are examples of all three alternatives in Finnish translations. Consequently, we do not find sufficient support for the retranslation hypothesis: there are no inherent qualities in the process of retranslating that would dictate a move from domesticating strategies towards more foreignizing strategies. The hypothesis may have arisen from the particular historical observation point of the researcher: what is considered faithful or accurate at any given time. It also includes a disquieting tinge of hubris and a questionable tendency to reduce historical development into straightforward evolution or linear progress: the ideas of accomplishment and interpretative evolution indicate a certain superiority towards predecessors and seem to suggest that contemporary reinterpretations are somehow more adequate or more mature than those produced during previous eras. Authors’ Addresses: Outi Paloposki Research Institute for the Languages of Finland Sörnäisten rantatie 25 00500 Helsinki Finland e-mail: outi.paloposki@kotus.fi
Kaisa Koskinen School of Modern Languages and Translation Studies University of Tampere 33014 Tampere Finland e-mail: kaisa.a.koskinen@uta.fi
A thousand and one translations
Notes 1. Translations of quotations are ours. 2. A recent survey showed that approximately 10% of the articles in ten international translation journals in the year 2000 dealt with retranslation (Susam-Sarajeva 2001). 3. It is worth noting here that RH presupposes that retranslation always takes place chronologically after the first translation, whereas some current usage of the term does not necessarily imply this. The multiplicity of translations, not necessarily their chronological order, is apparent in e.g. Pym (1998: 82): “[…] retranslations sharing virtually the same cultural location or generation […]” (emphasis added), and in Susam-Sarajeva’s (2002) study on Turkish retranslations of Roland Barthes’ texts. 4. Excluding books by Swift and Defoe, adapted for children (Kovala 1992: 68). 5. Finnish is a pro-drop language, where rich verb inflection allows for an empty category in the subject position (null subject). Pronominal specification was criticized as Swedish influence. 6. Note, however, that very little is known of what may or may not have been in the story collection before the fifteenth century (Irwin 1994: 4). 7. Nevertheless, the newspaper Kanawa was forced to cut short the story of Aladdin which it had been publishing as a continuing story in 1846, because of public dislike.
References Bensimon, Paul. 1990. “Présentation”. Palimpsestes XIII (4): ix–xiii. Berman, Antoine. 1990. “La retraduction comme espace de la traduction”. Palimpsestes XIII (4): 1–7. Berman, Antoine. 1995. Pour une critique des traductions: John Donne. Paris: Gallimard. Borges, Jorge Luis. 2000. “The translators of the Thousand and One Nights”. In The Translation Studies Reader, L. Venuti (ed), 34–48. London & New York: Routledge. Chesterman, Andrew. 2000. “A causal model for Translation Studies”. In Intercultural Faultlines, M. Olohan (ed), 15–27. Manchester: St. Jerome. Du-Nour, Miriam. 1995. “Retranslation of children’s books as evidence of changes of norms”. Target 7 (2): 327–346. Gambier, Yves. 1994. “La Retraduction, retour et détour”. Meta 39 (3): 413–417. Gerhardt, Mia. 1963. The Art of Story-Telling. Leiden: E. J. Brill. Hermans, Theo. 1999. Translation in Systems. Manchester: St. Jerome. Hämeen-Anttila, Jaakko. 2001. “Esipuhe kirjaan jota ei ole”. In Kuka murhasi Kyttyräselän? Tarinoita Tuhannesta ja yhdestä yöstä (Trans. Jaakko Hämeen-Anttila), 7–13. Helsinki: Basam Books. Irwin, Robert. 1994. The Arabian Nights: A Companion. London: Allen Lane. Kiuru, Silva. 2002. “Seitsemää veljestä Suomen nuorisolle”. In Äidinkielen merkitykset, I. Herlin, J. Kalliokoski, L. Kotilainen and T. Onikki-Rantajääskö (eds), 331–352. Helsinki: SKS. Kovala, Urpo. 1992. Väliin lankeaa varjo. Jyväskylä: University of Jyväskylä. Kujamäki, Pekka. 1998. Deutsche Stimmen der Sieben Brüder. Frankfurt am Main: Peter Lang. Kujamäki, Pekka. 2001. “Finnish comet in German skies. Translation, retranslation and norms”. Target 13 (1): 45–70. Lefevere, André. 1992. Translation, Rewriting and the Manipulation of Literary Fame. London and New York: Routledge.
37
38
Outi Paloposki and Kaisa Koskinen
Martin, Alice. 2001. “A translator’s view of translation norms”. In http://www.eng.helsinki.fi/hes/ Translation/a translator.htm. Oittinen, Riitta. 1997. Liisa, Liisa ja Alice. Tampere: Tampere University Press. Oittinen, Riitta. 2000. Translating for Children. New York: Garland. Oittinen, Riitta. 2001. “Liisasta Aliceksi ja taas Liisaksi” (book review). Helsingin Sanomat B2, 10.3.2001. Paloposki, Outi. 2002. Variation in Translation. Literary Translation into Finnish 1809–1850. Unpublished PhD thesis. University of Helsinki, Finland. Poltermann, Andreas. 1992. “Normen des Literarischen Übersetzens im System der Literatur”. In Geschichte, System, Literarische Übersetzung, H. Kittel (ed), 5–31. Berlin: Erich Schmidt Verlag. Pym, Anthony. 1998. Method in Translation History. Manchester: St. Jerome. Rose, Jonathan. 2001. The Intellectual Life of the British Working Classes. New Haven & London: Yale University Press. Susam-Sarajeva, S¸ebnem. 2001. “Is one case always enough?” Perspectives 9 (3): 167–176. Susam-Sarajeva, S¸ebnem. 2002. Translation and Travelling Theory. The Role of Translation in the Migration of Literary Theories Across Culture and Power Differentials. Unpublished PhD thesis. University College London. Toury, Gideon. 1995. Descriptive Translation Studies and beyond. Amsterdam: John Benjamins. Tymoczko, Maria. 1999. Translation in a Postcolonial Context. Manchester: St. Jerome.
Sources Carroll, Lewis. 1995 [1865]. Alicen seikkailut ihmemaassa. Trans. Alice Martin. Helsinki: WSOY. Carroll, Lewis. 2000. Liisa Ihmemaassa. Trans. Tuomas Nevanlinna. Helsinki: Otava. Goldsmith, Oliver. 1986 [1766]. The Vicar of Wakefield. Harmondsworth: Penguin. Goldsmith, Oliver. 1788. Land-prästens i Wakefield Lefwerne. Stockholm. (2nd edition). Goldsmith, Oliver. 1859. Maapapin Wakefjeldistä elämä. Trans. Gustaf Erik Eurén. Turku: Wilen. Goldsmith, Oliver. 1905. Wakefieldin kappalainen. Trans. Samuli Suomalainen. Helsinki: Otava. Ihmeellinen Onni. Itämainen satu. 1855. Trans. August Ahlqvist. G. W. Edlund. (Without title: the story of the rope-maker Hassan, 1001 Nights) 1831. Oulun Wiikko-Sanomat, issues 26–28, 30, 31, 33–38. Itämaan satuja. I–III. 1856–1857. Trans. G. E. Eurén. Turku: Wilén. Saarikoski, Pentti. 1969. Evankeliumi Matteuksen mukaan. Helsinki: Otava. Tuhannen ja yksi yötä. Kokoelma itämaalaisia satuja. 1878–1880. Trans. G. E. Eurén et al (anon.). Turku: Wilén. Tuhat yksi yötä. Kokous itämaan satuja. I–VIII. 1874–1879. Trans. N. F. Spolander. Porvoo: G. L. Söderström.
Creating “presence” in translation Radegundis Stolze Darmstadt
This article aims to show how translation theory may be based on hermeneutic language philosophy. It discusses the problem of translation — a dynamic task from the translators’ perspective. Translation means presenting a message from a given text as understood by the translator, to other readers. The theoretical view switches from the analysis of source- and target-text structures and cultures to the translators’ approach as experts who motivate their own decision processes. The modern hermeneutic approach sheds light on the translator’s cognitive activity in understanding written texts, which also requires critical reflection of one’s own knowledge base. Subsequent text production in translation is seen as a problem-solving process coordinating various dimensions. Aspects of holistic text understanding based on linguistics and the application of assessment categories for writing and revision will be described.
The cognitive act of understanding The translator’s competence is at stake. Translators are challenged to find orientation in their worlds of languages, cultures and sciences. The individual translator as a human being is faced — for each and every new task — with a special text and must understand it, in order to write a new text as a translation, which will then be understood by others and serve their needs. Comprehending a written text creates a mental representation which will then be presented in the translation. Thus any translation is linked to the original — not by structural equivalence — but by its mental representation in the translator’s mind. There is no “transfer” relationship between a source and a target text. Both texts together are identical, an external form of the cognitive representation of the message. We should change the traditional horizontal concept of translation as “navigating on a sea of words between languages and cultures” (Bassnett 2000: 106) to a different perspective. The relation is between the translator and the message, and not between texts in various cultures.
40
Radegundis Stolze
The act of presenting a perceived message from another language in a translation is rather similar to acting on stage, where an idea — one’s own or one received from the playwright — is “made present” for the audience. What I translate is the same as what I have in my mind, what I have understood beforehand — be it right or wrong. This problem has not been solved by traditional translation theory with its assumption of two separate texts: a source text and a target text. Designing models for the interlingual coordination of semiotic structures or intercultural transfer does not solve the primary problem of understanding the text to be translated, and connecting it with successful target language text production. The question should be raised whether understanding a “text” concerns oral or written communication. This question is important, as there is a significant difference between interpreting as an oral reproduction of heard speech, and translation as a presentation of written texts in intercultural communication. In speech acts with a shared presence of speaker and audience, speakers may well negotiate their text understanding with others. According to Sperber & Wilson’s relevance theory (1986), communication depends on the “principle of relevance”, in that hearers adapt their interpretations of an utterance through an optimization of means and resources. Relevance theory applies to “ostensive-inferential communication making manifest to an audience one’s informative intention” (Sperber/ Wilson 1986: 54). The underlying cognitive strategy of all human behavior is to select the most plausible assumption from the momentary cognitive environment. It is based on the general principle of spending as little cognitive processing effort as possible on supplying contextual information. According to Sperber & Wilson (1986: 15), the crucial mental faculty which enables humans to communicate with one another is the ability do draw inferences from people’s behavior. Gaining new information requires “that the outcome of an act of communication has to modify some previously held assumptions in order to be found rewarding” (Gutt 2000: 28). However, this functions only in direct oral communication, when the speaker and the audience are both present and share the same context of the utterance. Temporary misunderstandings are gradually eliminated within the prevailing situation. A written text, on the contrary, is a means of carrying a message fixed in language to other times, places and cultures and to unknown readers. Written texts are severed from their original situation, and there is no overt behavior as a context. Understanding written texts requires the reconstruction of the original situation. This occurs when the textual input creates a cognitive scene in both the reader’s and the translator’s mind (Fillmore 1977: 63). Even if texts materially “transfer” information between cultures, this information is “revived” in the relationship between the reader and the text, as it induces a cognitive representation. Texts allow us to look beyond the text structures and to view a distant, external world. The identification of a specific subject in a visual scene is one of the most important comprehension tasks in written texts. Non-professional readers and
Creating “presence” in translation
translators run the risk of inferring inappropriate text meanings because there is no “cognitive environment that is mutually shared between communicator and audience” (Gutt 2000: 27, n.7) which can invoke the principle of relevance. Therefore, translators’ orientation in the world and their textual approach as readers are central issues. The cognitive tasks required to understand a written text are totally different from those required to understand an oral text. Written texts have lost the direct stimulus character of utterances; they are now open to various interpretations, and readers, and in particular translators, have to reflect on their textual understanding. Comprehension cannot be taken for granted. There is no partner to negotiate the sense, no matter how available the text remains for repeated reading. Initial inference of supposedly relevant information may later prove inadequate and thus be transformed during the whole reading process.
An attitude of knowledge-based receptivity The translator’s approach may be based on hermeneutics. Hermeneutics as a language philosophy is not a feature of texts, as might be supposed from Barthes’ (1970: 10) definition of an “hermeneutic code” amongst five “semic codes” to be used as analytical tools for reading the “open text”. Barthes treats the text as an object, and the “hermeneutic code” is defined as anything “unconventional in a text, whether in its texture, denotative image or connotative sense”. The problem with this definition, however, is obvious: what appears as “unconventional” or “hard to understand” depends on the reader’s competence, and is not an objective text characteristic. What perhaps is difficult for me to understand might be totally clear for another reader who had traveled the country and already seen certain things, or knows his/her author, or has access to the relevant domain of science. What I understand depends on what I already know. The main problem in translation, therefore, is the translator’s knowledge base, which will be activated hermeneutically by the textual input. The whole dispute between “foreignizing and domesticating translators” (Venuti 1995) tends to neglect the primary problem of how a translator, endowed with more or less limited world knowledge, can grasp the text’s message. This is the hermeneutic issue. Hermeneutics is a philosophically oriented theory of interpretation (Gadamer 1960); for a more detailed survey of some aspects thereof see (Stolze 2003: 113). It can be summarized as follows: 1. As a starting point, hermeneutics questions the chance of any individual to understand otherness. Orientation of human beings in the world and motivation of their activity is pivotal. 2. Any exact description of rules with compelling logic does not lead to a singular expression of human behavior. Different convictions and deviating actions are
41
42
Radegundis Stolze
always possible. This is the main difference between humanities and science. 3. Hermeneutics distinguishes — from a personal worldview — between objects/ facts with their analysis/cognition, and human activity with its internal motivation, i.e., between objectivity and subjectivity, analysis and evidence, method and experience, rationale and intuition, inference and feeling. 4. People’s understanding and their actions are historically situated and need interpersonal relationships. Instead of the permanent objective clarity of perception, there is only a temporary inter-subjective control of opinions through discussion within a group. 5. Understanding is possible on the basis of something shared, e.g., common experience and knowledge. The individual never absolutely and exclusively possesses the uniting aspect; he or she only has a share in it. The cultural experience of history as a critical appropriation of a people’s tradition is also the basis of cultural understanding. There is no autonomous individual existence, since we all live in the traditions present in our languages, and handed down in texts. 6. Gadamer sees understanding as a dialogue with the text. Reading involves a learning process, and this finally leads to the famous “melting of horizons” (1960: 289) between the understanding person and the text. When comprehending, I receive a share of the text’s meaning as formed by the author. The receptive acceptance of the strange — until the moment of understanding — discloses the truth, when in understanding a foreign text I enter into its sphere of discourse. 7. Truth thus appears as evidence, a disclosure of sense felt like an experience. Therefore, hermeneutics considers an attitude of openness, of unprejudiced receptivity, as decisive for the person trying to understand, so that the message may disclose itself in the process of reading. This is significantly different from cognitive “text processing”, which is conceived as a strategy directed towards creating mental representations by adapting information (Rickheit 1995: 15). 8. Any understanding experienced as a learning process influences and enriches the reader’s mind, which then opens new horizons for further understanding. Within the “hermeneutic circle” (Gadamer 1960: 250) it is true that, as the relationship between one’s given knowledge base and the novel input from understanding develops, the cognitive and even the translating capacity will increase constantly. 9. The hermeneutic philosophy of language first developed a set of rules for text analysis which critically consolidate one’s understanding. For example, there is an intrinsic connection between the whole entity and its individual elements, or the original text constitution and its later effect, or the respective genre and its analogy to other genres, etc. This was later developed by modern text linguistics, and we may apply it to reflect on and explain our own understanding.
Creating “presence” in translation
10. Texts in their faithful interpretation are dynamic, evolving their potential meaning at every new reading. You can experience this when you re-read older versions of your own papers or translations: they appear partly defective or strange, because you have learnt something in the interim.
Translation as an assignment yet to be fulfilled Hermeneutics discussed the human approach to the world, and it can provide a valid basis for translation which is seen as a process-oriented human activity. More precisely: translation is seen here as an assignment not yet fulfilled, and thus not suitable for pressing into a model. Translation is an open process towards an optimal solution, responsive to orientation, motivation and revision. This includes a change of viewpoint — from a relationship between texts to the translator’s perspective. The issue is not a description of any “hermeneutic truth” found in a text, but a translator’s question of: How can I understand? Understanding the truth of a text is never fixed or completed, it constitutes a dynamic process, as we are constantly learning. On the other hand, understanding is not impossible or necessarily ethnocentric in a “power struggle of the meaning”, as long as we really open ourselves to the foreign message. Understanding a text is not the same as integrating a text/ translation into another cultural polysystem. The core problem in translation appears to be the fact that translators are faced with an unfamiliar text — maybe from a foreign culture or from a specific domain of specialist communication (e.g. legal or medical texts) — and they have to produce a text for target readers who, again, are mostly not members of their own social group, but often come from the same discourse field as the original text. There is a double bond: to the original message and to the target readers’ expectations. Translators do not stand “between cultures” (Bassnett 2000: 113), but they are rooted in one culture, and by having access to the other, they cognitively reach out into both. The two culture systems establish contact within the translator’s mind: in other words, her cognition as an expert reaches out into two different cultures and into various discourse fields of scientific and academic knowledge. The concept of cultures determining communities that share a common language, history, social life, etc. remains abstract however, as there are various diacultures within one cultural system. There are special groups of scientists, housewives, church people, political parties, translation teachers, staff in the European institutions, medical doctors, business men, etc., all rooted in their own “native” culture. People can adopt several roles, thus having cognitive access to various diacultures. The dia-cultural way of speaking is mirrored in their texts. A message circulating among people has its motivational background, it answers certain social questions, and that is what translators have to discover.
43
44
Radegundis Stolze
Within the multiple cognitive outreach into various dia-cultures, translators perceive the content of texts. Their special expert qualification is their access to various specializations, not only to the one an author has in his own field of work. In a normal professional situation, for instance in the language industry, translators are constantly faced with different types of texts. They are the only nonspecialists in the whole communication process of translating, while authors and even target readers often are members of a similar group. The core problem is then: how to understand that message, how to enter that strange sphere of discourse, since the task of creating a presence for the message in a translation means authentically continuing that discourse between author and readers in another language.
The translator as an expert reader The difficulty in understanding a text to be translated is not the given text; rather it is the translator with his/her adequate or inadequate knowledge base in the hermeneutic circle. Language is a means for many different purposes, and the words themselves are neutral; they only get their real meaning when seen against that background. The necessary translational reading means firstly an expert “positioning” of the text in its culture, discourse field and conceptual world, and this will activate the translator’s knowledge base regarding cultural characteristics and features of specialist communication. Any text to be understood and translated presents itself to the reader/translator in an individual and holistic way. The text as an individual message consists not only of semiotic structures requiring analysis; it also contains some singular aspects that can only be perceived intuitively. Language is always a combination of both rule application and intuitive creation by human beings. In the process of understanding I will also have to consider the text as a whole, since any text element only receives its specific meaning within the entirety of the text. The sense of a text is not determined by the sum of the meanings of single words and phrases, but rather by holistic comprehension. A responsible translator will have to transcend his or her own subjective viewpoints and try to grasp as many aspects of the text as possible. Initially one will only have an individual understanding of the text and will need further research, before being able to present the message responsibly and comprehensively. Later on, the translator will have acquired more experience and will be able to infer the relevant knowledge right from the beginning. She has a double perspective on texts: viewing their cultural constellations and analyzing the text structure to find confirmation for her initial understanding. The holistic approach to texts in literature and in specialist communication is the same, but the required knowledge base is different. Instead of proceeding to an initial “text analysis” of language and grammar structures, translators will ask
Creating “presence” in translation
themselves when positioning the text: What are the knowledge elements at my disposal? Am I able to understand the message? From which country, which social group does it come? What do I know about that culture? Generally, in language and literature, we should take into account the social position of a group and the rooting of people, as a culture is not a homogeneous entity. Every field of communication has its specific concepts and they recur in the texts. Determination of key words is decisive for grasping the holistic sense of the text and the author’s ideology. We should be interested in cultural associations and metaphors representing stereotypes. If we consider specialist communication: which area of research and work — science or humanities — determines the discourse field of the text? Which level of communication is applicable — does the text address experts or lay people? We should be able to recognize the characteristics of specialist texts, know that there are various disciplines with several domains, and that communication here functions against a specific background. There is a difference in terminological conceptualization between exact definitions with methodical deduction in the sciences, and academic convention and interpretation of terms in the humanities (Stolze 2003: 201). With respect to the intrinsic relationship between the text as a whole and its constitution, the holistic view is being complemented by an analysis of the predicative mode which shows in the particular style of the text characterizing its author. Discourse markers may structure the message, and the speaker’s perspective is illuminating. Mood and focusing in the sentence structure tell us something about what was important for the author and what was not. In specialist texts we observe verbal speech acts indicating e.g. directives or legal obligations. Specialist texts are informative, anonymous, and linguistically economical. The aspects to be considered in reading for translation are shown in the table: Translator’s reading Literature Culture Discourse field Concepts
Predicative mode
Specialist texts
People, country
Area of research and work (science, humanities) Social background & field Discipline & domain, level of commuof communication nication Cultural associations, key Terminological conceptualization words, author’s ideology (definition/deduction vs. convention/ interpretation)
Idioms, mood, focusing, quotations
Speech acts, passive voice
45
46
Radegundis Stolze
Translation as a holistic text production Translators are responsible for presenting the message understood from a text as comprehensively as possible for target culture readers, so that these may interpret it and react to it according to their own interests. It is never the translator’s task — except for a special commission — to provide an ethnocentric interpretation of the original for them, or to change and adapt it to make it more adequate for the readers (Venuti 1995: 93). As with the reading of an original text, it is the readers who decide whether the translated message is acceptable or not. Competent translators will question themselves whether the understanding gained may not still be a subjective, ethnocentric interpretation requiring deeper research for a more authentic comprehension of the message. Any foreign idea, once grasped, can certainly be expressed somehow in another language, but a translation is not a subjective pleasure for linguists, nor a narrow ethnocentric interpretation bearing the risk of being misunderstood. Also, the responsible translator will have to motivate the decision on how to formulate the message which she has understood. Translation aims at “good text production”; this has long been seen as a kind of “reverse text reception” (Antos), including comparable restrictions for the acting person: poor knowledge, limited proficiency in language and genres, problems of activating and focusing one’s knowledge, etc. (Antos 1982: 6). On the other hand, there are some differences. Comprehension competence is organized more globally, as shown, and it does not focus on grammar structures but creates a cognitive scene, whilst the productive competence really focuses on the text level and formulation aspects (Antos 1982: 119). Many researchers view text production as a “problem solving process”, in accordance with the studies of Hayes & Flower (1980) who analyzed the process of professional text production. They isolated phases of planning, e.g. tentatively translating one’s ideas into text structures, and reviewing the solutions according to communicative goals. Draft writing and reviewing are repeated in a cyclical fashion several times, until a final text is produced that corresponds best to the initial writing goal. In translating, this overall goal is already given in the task of presenting the original message, but without regard for the linguistic structures of the source text. The problem solving process of writing is oriented holistically towards the whole of a text following global production schemes. The specific problem in translational text production, then, is the coordination of the various rhetorical features that contribute to the intended sense of the text as a whole. Such features are developed gradually during constant reviewing and reformulating of the first draft, and they involve concrete aspects such as text organization, understandability and adequacy for addressees, effect, style, emotion,
Creating “presence” in translation
institutional background of the communication, aesthetics, thematic progression, coherence, etc. All these global features can be described by rhetorical and stylistic forms on the text level. Text genres and formulaic writing with fixed blocks and phraseological elements are in certain recurrent situations an application of some writing problems which have now been socially accepted and solved. Understanding of the text leads to an initial draft of a translation regarding the global vision of the message. In order to make it more adequate for possible readers, the first draft will then be reviewed in several stages, according to the quality assessment categories presented below. The individual coordination hierarchy of the various aspects has to be determined anew for every case of translation; no general rule for all possible situations is yet available. The revision of the drafts may be based on linguistic and rhetorical aspects, as they have been traditionally analyzed for text production. Rhetorical categories
General language
LSP
Text function
Discourse markers, speaker’s perspective, tense, genre Semantic web, titles, compatibility, structural axes, intertextuality Expressive means, rhyme, metaphors, alliteration
Text types, macrostructure, addressees’ expectation Equivalence of terminology, specification of concepts, word compounding Functional style, standard formulae, controlled language, phraseology
Rhythm, prosody, text form, pictures, verse order
Illustrations, layout, script fonts
Coherence
Style
Shape
In the revision of an initial translation draft, all the categories mentioned play an important role, and they are interrelated and never equally valid. The translator will have to decide in every single case which aspect is dominant. Multifaceted texts include all these aspects at varying levels of the holistic integrity.
How to reach authenticity in translation The translation task is to create a “presence” for the message in empathy with it, so that it appears like an original text about one’s own opinion. The translator is a co-author for the text. This includes, on the one hand, a certain identification with the message and, on the other hand, coherent text production oriented towards the target readers’ level of comprehension, as required for authentic translation. If the content of a text is not acceptable for a translator, he or she can only refuse the assignment.
47
48
Radegundis Stolze
Translation is seen here as an intuitive formulative impulse regarding the message — instead of an interlingual transfer. The cognitive representation of the message first understood leads, via this impulse, to the appearance of target language words and phrases as the frames for the cognitive scene, in which this message is “present”. Such a cognitive event is an “autopoietic process” of semiosis, not completely controllable methodologically. The message simply changes its linguistic form in the process of finding words for it. The semantic meaning is more important here, than the grammar aspect. When I know what to say, I will find the right words for it. “Poiesis” means transformation of something, contrary to creation out of nothing (Antos 1982: 100). And autopoiesis is a characteristic feature of living systems, such as human beings when they act as translators. Intuition is a core aspect of translation. In this quest for authenticity — in order to create “presence” for the message — the translators focus on the intended discourse field for the translation. The text function will be important, as it determines the general set-up of the whole text that is written. In translation for specific purposes the text function normally remains the same, since it is a continuation of specialist communication across language barriers. Text types are cognitive formulation models, and it is useful to know about target parallel texts. Competent translators command certain macrostructures and apply them, as they correspond to the addressees’ reading expectations. There is a difference when you write for lay persons or for specialists; intelligibility is a relative concept. The concrete formulations may be revised according to said assessment categories. In order to present a message in a responsible and authentic way, the text produced will also have to be coherent. According to hermeneutics, texts permit recipients to look beyond the language structures into the surrounding world, so that an overall cognitive scene will be created. Coherence in the target text may be realized by means of a semantic web. Having detected word fields of cultural concepts on the holistic text level, one might search for corresponding lexemes in the other language. Visual imagination helps in that process. Tenuous translations, due to language interference, often prevent the creation of a cognitive scene in the reader, and thus do not create presence for the message. Of course the text does not consist solely of words. The style of the whole text is as much an integrating element of sense as is the content and cannot be detached from the form. In literature, we will try to find concise formulations, and this is often not achieved at the first attempt. But there are many possibilities to better express an idea once grasped. A sensitive translator who detects them will also find a solution. Also, there are indefinite formulations open to interpretation in texts; this is a well-known topic of literary and reception studies. Translators might try to keep this suspense in the text, rather than fix it into a very clear proposition. In
Creating “presence” in translation
specialist texts, functional stylistics is decisive in order to create presence, since scientific contents need their adequate representation in texts. Finally the shape of the text as an extra-linguistic feature is relevant as well. In publicity texts, the cultural embedding of illustrations might be interesting, as it influences their understanding. Quite often in such translations, it is not only the text that must be altered; different pictures may even be used in order to present the message truly as such. Emotion is expressed in the sentence rhythm, and the verse order in poems often hinders literal translation. In specialist texts, the internal relationship to illustrations and the layout prerequisites are often a special writing problem. Technical translations may be dominated by the space available, and even script fonts have culturally different effects. Translation means writing a text, and we have to coordinate the various aspects — text function, coherence, style, and shape — in our quest to create presence for the message.
Conclusion Literature and general language communication is situated in a society with its cultural facets and linguistic creativity. It exploits the full potential of the language; the need for linguistic creativity and visual freedom remains valid. The translator needs courage and confidence in his or her own idiomatic proficiency. Specialist communication on the contrary is situated in a certain area of work and research within a domain and discipline, including a particular scientific conceptualization. This makes texts more difficult to translate, as we have to do research and learn about domain specificities. Such texts make use of a selection of the linguistic potential only, and are responsive to routine. This hermeneutic approach is innovative insofar as it focuses on presenting a “message” rather than a “source text”. The practical task for the translator then shifts from analyzing equivalence on the linguistic level to text production in a social situation. This means that instead of describing translation processes in a contrastive way, it would be more interesting to prescribe what target texts should look like, so that text production can follow a model. Only if there is responsibility in the professional translator, will the reader accept the translation as an adequate text. Translation should help to enter a literary world created by a novel, or to continue specialist communication initiated by the original text. Why should readers ask for a translation, when it does not, for them, represent the original? Seen in this way as a responsible task, translation is no longer a secondary work of slaves that have to follow their authors in all directions. It is, rather, a very creative operation of text production, never absolutely completed: it remains a “work in progress”.
49
50
Radegundis Stolze
References Antos, Gerd. 1982. Grundlagen einer Theorie des Formulierens. Textherstellung in geschriebener und gesprochener Sprache. Tübingen: Niemeyer. Barthes, Roland. 1970. S/Z. Paris: Editions du Seuil. Bassnett, Susan. 2000. “Authenticity, travel and translation”. In Translationswissenschaft. Festschrift für Mary Snell-Hornby zum 60. Geburtstag, M. Kadric, K. Kaindl and F. Pöchhacker (eds.), 105–114. Tübingen: Stauffenburg. Fillmore, Charles J. 1977. “Scenes-and-frames-semantics.” In Linguistic Structures Processing, A. Zampolli (ed.), 55–81. Amsterdam: New Holland. Gadamer, Hans-Georg. 1960. Wahrheit und Methode. Grundzüge einer philosophischen Hermeneutik. 5th ed. 1986. Tübingen: J. C. B. Mohr (Paul Siebeck). Gutt, Ernst-August. 2000. Translation and Relevance. Cognition and Context. 2nd ed. Manchester: St. Jerome. Hayes, John Richard and Flower, Linda S. 1980. “Identifying the organization of writing processes.” In Cognitive Processes in Writing. L. W. Gregg and E. R. Steinberg (eds.), 3–30. Hillsdale: Erlbaum. Sperber, Dan and Wilson, Deirdre. 1986. Relevance. Communication and cognition. Oxford: Blackwell. Rickheit, Gert. 1995. “Verstehen und Verständlichkeit von Sprache”. In Sprache. Verstehen und Verständlichkeit. Kongressbeiträge zur 25. Jahrestagung der Gesellschaft für Angewandte Linguistik GAL e.V., B. Spillner (ed.), 15–30. Frankfurt am Main: Peter Lang. Stolze, Radegundis. 2003. Hermeneutik und Translation. Tübingen: Narr. Venuti, Lawrence. 1995. The Translator’s Invisibility: A History of Translation. London/N. Y.: Routledge.
Ready-made language and translation Pál Heltai Institute of Foreign Languages, Szent István University, Gödöllo˝
Translation studies remains an interdisciplinary field, making good use of new insights in other linguistic disciplines. The present paper gives an overview of the problems of ready-made language, drawing on developments in second language acquisition, lexicology and sociolinguistics, and evaluating the translation of non-literary texts from all these different points of view. The paper argues that routinization is an important feature of non-literary texts, while overstepping the limits of routinization may lead to jargon and translationese respectively. It also maintains that, since the use of readymade language enables translation to be more efficient in terms of the effort invested, the training of non-literary translators should equip translators with a phraseological competence enabling them to use ready-made language effectively and to evaluate ready-made language from the sociolinguistic point of view. The paper also reports on a pilot experiment demonstrating some of the effects of such a phraseological competence.
1.
Interdisciplinary character of Translation Studies
In recent decades, Translation Studies (TS) has developed into an independent discipline, yet it remains an interdisciplinary field, constantly borrowing ideas and methods from other disciplines. Linguistics remains an important source and inspiration, although attempts to directly transplant the results of linguistics and other disciplines into TS are received with well-deserved mistrust (e.g. Gutt’s approach based on relevance theory; cf. Tirkkonen-Condit 1992). Snell-Hornby (1988/1995: 33/32) classifies translation on the basis of prototype theory, and points out that different types of translation are influenced to different degrees by linguistics and various other disciplines. This is worth remembering, since many authors in TS develop theories on the basis of literary translation alone, and then treat them as general theories of translation. Nida, for example, drew inspiration from Bible translation (Nida 1964, Nida and Taber 1969), Toury (1980, 1995) talks about the position that the TL text occupies in the social and literary
52
Pál Heltai
systems of the target culture (not very typical of technical translation), and Gutt (1991/2000) altogether excludes technical translation from the concept of translation. As a result, the application of such theories to technical translation often runs into difficulties. (The term technical translation is used throughout this paper in a broad sense corresponding to Snell-Hornby’s special language translation, including translations of business, medical, legal, etc. texts as well as the translation of texts about science and technology.) Literary and technical translation show such significant differences that one wonders whether these differences are of degree or of kind, and indeed, if a unified theory of translation is possible at all. If so, it must be based on prototype theory, recognizing that there are prototypical and non-prototypical literary and technical translations, and that there is a gradual transition between the two. Yet the two opposite poles are so wide apart that talking about translation in general is always a risky business, and to avoid confusion, one should always state the kind of translation one is talking about. This paper, drawing on language acquisition research, lexicology and sociolinguistics, is aimed at examining certain aspects of L2-L1 technical translation. It is not designed to go into a detailed discussion of differences within the broad category of technical translation: it applies to prototypical technical texts as against prototypical literary texts.
2.
Ready-made language
2.1 Ready-made language in language acquisition research In the early 1980s several studies suggested that language acquisition may be based, at least in part, on memorizing multi-word units or whole sentences, which are stored and recalled as unanalyzed chunks (Wong-Fillmore 1976, Peters 1983, Pawley and Syder 1983). These chunks are subsequently analyzed and the rules of syntax are derived from them. However, as syntactic rules emerge, the conventionalized association between the chunk and its function in context is retained and continues to be available. In the course of language use we rely more often on memorized sequences and reproduce them as wholes rather than generate sentences from scratch. Automatic recall of memorized sequences gives speakers more time to plan the message, providing them with “islands of reliability” when, due to limiting factors, they cannot concentrate on both content and form. The theory raises the possibility that language use is not as creative as supposed by transformational generative grammar, and fluent language use is not based on rules at all but on memory. Nattinger and DeCarrico (1992) claim that in speech production we usually start with phraseological units:
Ready-made language and translation
The process of speaking is compositional: We start with the information we wish to express or evoke, and we haul out of our phrasal lexicon some patterns that can provide the major elements of this expression. Then the problem is to stitch these phrases together into something roughly grammatical, to fill in the blanks with the particulars of the case at hand, to modify the phrases if need be, and if all else fails, to generate phrases from scratch to smooth over the transitions and fill in any remaining conceptual holes.
This model of speech production is yet to be elaborated, but some of the principles have already been applied in the practice of foreign language teaching, with the “lexical approach” advocated by Lewis (1993, 1997) placing great emphasis on collocations and other ready-made units. 2.2 Ready-made language in linguistics While the study of ready-made language is a relatively new issue in language acquisition research, it is not quite so in other branches of linguistics: various aspects of it have been touched upon or even studied in depth by theoretical linguistics, lexicology and stylistics. In theoretical linguistics the problem of “idioms” surfaced relatively early as an “anomaly” in the transformational generative model (Chafe 1968). Indeed, Bolinger (1976) claimed that most language is remembering rather than generating. Later on Sinclair (1991) suggested the “idiom principle”, and corpus linguistics has done extensive research on collocations and idioms. Within traditional lexicology, the study of ready-made language was at least partly covered by the discipline of “phraseology”. Neither is the phenomenon of ready-made language a new issue in stylistics, where ready-made units are usually frowned upon, and have traditionally been called names like stereotypic expressions, jargon, hackneyed phrases, clichés and so on. Ready-made language has drawn criticism from writers, journalists and members of the public worried about the standards of language use (with variable support from linguists). Orwell (1946/1978) talks about “dying metaphors” (toe the line, ride roughshod over, play into the hands of, no axe to grind, fish in troubled waters) and “verbal false limbs” (render inoperative, make contact with, take effect). 2.3 Definition and classification A number of terms have been used in theoretical linguistics, lexicology, language acquisition research and stylistics to describe the phenomenon I have referred to as ready-made language. The terms formulaic language, ready-made language, routinized language, prefabricated language, as well as multi-word units, phraseological
53
54
Pál Heltai
units, ready-made units, routine formulae, memorized formulae, memorised sequences, prefabricated elements and prefabs are sometimes used synonymously. Unfortunately, it is not only a terminological problem: in a review article Weinert (1995) points out that the underlying concept itself is not satisfactorily defined. Some of the confusion stems from the different perspectives adopted by lexicology and language acquisition research. Lexicological research seems to favour the term multi-word units or items, focusing on idioms and collocations as semantic units. According to Moon (1997: 43) “A multi-word item is a vocabulary item which consists of a sequence of two or more words (a word being an orthographic unit)”. The definition, however, cannot accommodate the sentence-length memorized sequences that are not equivalent to vocabulary items, and cannot be used as constituents of sentences. The meaning of the term ready-made language in language acquisition research is different from that in traditional lexicological and lexicographical approaches. The main differences are as follows. – – –
–
Ready-made language is defined using psycholinguistic criteria (ready-made units are stored and recalled as wholes). Idiomaticity is not a central concern. Whole sentences and all sorts of memorized material (memorized linguistic material above the sentence level such as quotations, slogans, poems, songs, etc.) are also taken into consideration. The enormous number of memorized sequences and their central role in language acquisition and production is emphasized.
The criterial feature of ready-made units is that they are stored and recalled as unanalyzed wholes. According to Pawley and Syder (1983), the number of expressions recalled as wholes from memory may be put at several hundred thousand. Nattinger and DeCarrico (1992) claim that “we are creatures of habit”, and “routinized formulas and other sorts of pre-fabricated language chunks, which are products of … ritualization, seem to play a large part in both acquiring and performing language”. According to Ellis (1996), the process of language acquisition is the learning of sequences: learning the sequences of phonemes in words and the sequences of words in collocations and sentences. Classifications of ready-made language in language acquisition research (Pawley and Syder 1983, Nattinger and DeCarrico 1992) focus on full sentences or sentence-length units reproduced as wholes (What do you mean by that?) and phrases with particular pragmatic functions (I can assure you …). Quotations, proverbs, sayings, slogans, situation-bound utterances and clichés are usually whole sentences and are reproduced invariably in the same form: To be or not to be; I wandered lonely as a cloud; Charity begins at home; Just do it; Do you want to talk about it? I know what you feel; A good time was had by all etc.
Ready-made language and translation
Smaller units than the sentence are what Pawley and Syder (1983) term “lexicalized sentence stems”, i.e. sentence frames that allow some variability, having open slots filled in according to the topic of discourse, but essentially available as wholly assembled units. One example given by Pawley and Syder is I’m sorry to keep you waiting, where the object and the tense are variable. Lexical phrases (Nattinger and DeCarrico 1992) have pragmatic and/or discourse functions, and are usually phrases, not sentences (By the way, … In conclusion … Further research is necessary to … I’m awfully sorry… All in all… If I were you I would …) It is this function that sets them apart from idiomatic expressions and restricted collocations, which have no such function: spill the beans (idiom) or bitter frost (restricted collocation); i.e. they do not have any identifiable pragmatic function in themselves. Idioms and restricted collocations seem to be classed as ready-made units in language acquisition research, although the focus seems to be on longer (clause or sentence-length) memorized sequences, while traditional lexicological research is focussed on idioms and collocations. It is by no means an easy task to reconcile the two approaches. Moon (1997: 47) suggests that prefabricated elements are a subcategory within the category of multi-word units, but from a psycholinguistic perspective it seems more logical to regard ready-made language as the generic term. Howarth (1996: 12) divides word combinations into functional expressions and composite units, the former corresponding to the discourse and pragmatic units emphasized in language acquisition research, the latter to multi-word lexical units (idioms and collocations) studied by lexicology. In this paper, since I wish to emphasize the psycholinguistic aspect, I shall use the term ready-made language and ready-made units to refer to both functional expressions and composite units, and reserve the term multi-word unit for composite units (idioms, collocations, binomials and other vocabulary items). It should be noted, however, that the status of collocations as ready-made units is as yet unclear. According to Howarth (1996), some classes of restricted collocations are probably stored as wholes, while others are not. The proverb Waste not, want not or the idiom to bite the dust are obviously semantic units and according to psycholinguistic evidence belong to ready-made language, but it is not clear whether all of the following more or less restricted collocations behave in the same way: hit hard, an earthquake hit (+ Object), hit the ground/road, hit the jackpot etc. It is possible that some subclasses of collocations behave as units, while other subclasses (less restricted, or “weak” collocations) do not, although we might hypothesize that even with the latter the activation of one constituent has a “priming effect” on the activation of the other one, and thus access to hit hard takes less time than access to the two constituents separately.
55
56
Pál Heltai
2.4 Lack of psycholinguistic research on ready-made language While researchers seem to agree that ready-made units are stored and recalled as units, psycholinguistic research on how this happens is more or less lacking. According to Nattinger and DeCarrico (1992), language is stored redundantly, i.e. hit and hard may be stored separately and together, too. Aitchison (1987) claims that compound and derived words are stored in the mental lexicon as wholes but it is also possible to generate them from stems and affixes in the course of communication. However, she does not spell out the details of this process. According to Howarth (1996), this finding can be extended to other multi-word units, such as restricted collocations. Models of the mental lexicon are usually concerned with lexical items and connections between lexical items. Research on the bilingual lexicon also focuses on words, especially the problem of whether the words of the two languages are stored in one lexicon or two (e.g. Schreuder and Weltens, 1993, de Bot et al. 1995, Singleton 1999). Studies of lexical access based on Levelt’s 1989 model are also aimed at exploring access to words. However, as Howarth (1996: 48) remarks, except for idiomatic expressions so far little attention has been directed at other categories of ready-made language, and psycholinguistic research has not yet tackled the problem of access to the wide range of memorized sequences that constitute ready-made language.
3.
Sociolinguistic aspects of ready-made language
Second language acquisition (SLA) research has a descriptive approach to readymade language and emphasizes its positive role in communication. Traditional phraseology also noted the image-power and expressivity of idioms and other phraseological units. On the other hand, stylistics has often deplored the use of ready-made language, as shown above. We must therefore examine these complaints to see if they are justified, and if so, to what extent, what the causes are, and what kinds of ready-made phrases are usually criticized. This issue has exceptional relevance for translation, since translations are par excellence targets for criticism on such grounds. The phrases disapproved of in stylistics correspond by and large to Orwell’s categories, “verbal false limbs” and “dead metaphors”. We shall look at these categories separately.
Ready-made language and translation
3.1 Jargon Orwell’s “verbal false limbs” seem to correspond to ready-made units often used in special languages and often regarded as important constituents of jargon. The term jargon has several senses: it may refer to the specialized terminology of a science or occupation, especially if used outside its natural provenance, and in another sense it is used to refer to a certain style, typical of bureaucratic and also of technical language, which is unnecessarily complicated and contains many ready-made units. The irritation caused by ready-made phrases may be due to vague meaning, excessive use, pretentiousness and use outside the natural context. 3.2 Hackneyed expressions “Dying metaphors”, “hackneyed” or “trite” expressions may be irritating because they are too frequent and there is a contradiction between frequency and claim to novelty. Too frequent use seems to wear down the expressive power of metaphorical, idiomatic and emotionally loaded words: familiarity breeds contempt. Hackneyed expressions abound in colloquial language, newspapers and political language, where emotions play an important role. In technical and scientific registers, apart from a few overworked phrases, hackneyed expressions are not very frequent. 3.3 Ready-made language, jargon and hackneyed phrases Ready-made units that are felt to be irritating do not show any structural differences compared to stylistically neutral ready-made units. The difference is not structural, but sociolinguistic: jargon, hackneyed or trite expressions and similar terms simply mean that for various reasons and in given contexts such phrases are regarded by many as difficult to comprehend or irritating. While such evaluations may not always be well-founded, the existence of such opinions is a sociolinguistic fact, and the translator does well to take them into consideration. 3.4 Expected degree of routinization Texts composed in different registers and different situations are expected to show different degrees of routinization. Fairy tales contain a number of ready-made phrases that listeners expect to hear and indeed enjoy hearing when their expectation is fulfilled. Rituals cannot do without ready-made language: it is inadvisable to change the usual expressions in a marriage ceremony or a funeral service. In works of literature, however, ready-made language is usually not appreciated. Apparently, in a literary context the expected degree of routinization is low. In everyday conversation and newspaper language ready-made phrases may pass unnoticed, but
57
58
Pál Heltai
romantic situations are rather intolerant of ready-made language, since in such situations you are expected to be original and poetic, and use novel, never-heard-of expressions to express your unprecedented and unique feelings. Technical, scientific, legal and bureaucratic language, on the other hand, call for a high degree of routinization, although, of course, different types of technical texts (and consequently technical translations) contain different amounts of ready-made language; indeed, different sections of a technical text may contain different proportions of ready-made language.
4. Ready-made language in technical texts and technical translation 4.1 Degree of routinization in technical texts Technical texts are often repetitive: they are about the same topic, describe the same procedures, and report the results in the same way. As a result, language forms will also tend to be repetitive, and prefabricated, routine phrases will abound. The most common ready-made units in technical and scientific registers are collocations and certain discourse-organising lexical phrases. Sentence-length memorized units, focused on in language acquisition research, are less frequent, but definitely exist here, too. A nice example of ready-made units used in technical texts is provided by Ellis (1996), who summarizes a paper concerned with ready-made language using ready-made units, and marking the ready-made units he uses (see Appendix 1). From the sociolinguistic point of view technical texts will become problematic if the degree of routinization does not match the expected degree of routinization. Divergence in both directions will cause problems. If routinization is excessive, the text will be perceived as pretentious and jargon-like, if it is inadequate, it will be perceived as not matching the register. 4.2 Ready-made language in translation: routine and creativity Translation theories in the past have often emphasized the creative or recreative aspect of translation. Kussmaul (1995:39–53) devotes a chapter to “Creativity in Translation”. Klaudy (1994: 19) claims that translation, as a decision-making process, is a creative activity. The shift away from grammatical and lexical contrasts and linguistic theories of translation and increased attention paid to pragmatic, discourse and sociolinguistic factors seem to have reinforced the emphasis on creativity. While this emphasis on creativity may be justified in the case of literary translation, it is less valid for technical translation, in which, just as in monolingual technical communication, routinized, automatic language processes, involving the use of ready-made language, acquire a significant role. Evidence for the routine
Ready-made language and translation
nature of technical translation is provided by the growing number of translation tools aimed at taking the drudgery out of technical translation. The routine nature of technical translation is due to two main factors: technical translators have limited time, and the register itself is characterized by the use of ready-made units, especially collocations. Howarth (1996: 131) notes that “it is partly by means of collocations restricted in a technical sense that one recognizes register”. The time factor also calls for the use of ready-made units, which frees processing capacity (cf. Gile 1995: 166–7). Indeed, in emergency situations translating becomes similar to sight translation or interpretation. On the basis of the above, I suggest that the process of technical translation is not as creative as is sometimes claimed by literary-based translation theories. Translators, like monolingual communicators under conditions of stress (e.g. football commentators), will inevitably and necessarily rely on automatic strategies and use ready-made units, helping them to cope with time constraints (and meet deadlines) and, at the same time, to achieve the expected degree of routinization in the target text. The use of ready-made language, however, always involves the risk of sociolinguistic error: if the translator oversteps the limits or falls short of the expected (and acceptable) degree of routinization, the target text may be perceived by readers as jargon or translationese. 4.3 Translationese Jargon and translationese seem to be overlapping terms. One component of jargon is excessive routinization, which can make it more difficult to comprehend a text, although it does not necessarily disrupt communication. Under certain conditions (e.g. time pressure), when there is no chance of getting a translation that conveys the message of the text more adequately, jargon can, to some extent, be tolerated. In this case the reader’s processing effort will increase, that is, the reader undertakes the role of post-editor in order to receive timely communication as against no communication. In relevance theoretic terms, the greater processing effort is justified by the contextual effect the reader can derive from such non-perfect translations. Translationese shows features similar to those of jargon, in that it is difficult to comprehend, but the difficulty may come from direct and indirect source language interference. Indirect interference concerns primarily distribution: the number of routinized expressions, nominalizations and impersonal constructions may be higher (Heltai 2000). Indirect pragmatic influence may manifest itself in what has been called quasi-correctness (Klaudy 1987), i.e. the omission of optional translation shifts, especially explicitation (discussed in Klaudy 1999), while direct interference is manifested first of all in collocations and other ready-made units. Odd, foreign-sounding collocations, however, seldom disrupt meaning, since speakers
59
60
Pál Heltai
have the ability to interpret novel collocations. According to Baker (1992: 50) there is no such thing as an impossible collocation, and within a given text almost any novel (or erroneous) collocation can be interpreted. The collocations created by the translator under the influence of the source text (ST) will be interpreted by readers in the same way as novel collocations in their native language: as unusual, novel collocations. The processing effort will increase, of course. The same may apply to some syntactic features: according to Schmid (1999), with some constructions in English a structure-preserving translation does not lead to an ungrammatical or unacceptable German sentence. In such cases the thematic structure of the target text (TT) may suffer, but can be recovered through some extra processing effort by the reader. Translators working under time pressure feel a great temptation to work in this way (Heltai 1999). A typical feature of translationese is the high incidence of ready-made units, which might be unusual in the given target language register (Heltai 2000), or just the opposite: instead of a ready-made unit, anticipated by the reader of a specialized domain, the translator uses a “creative” word combination. This can be due to the translator not recognizing ready-made units in the ST, or not knowing the registerspecific ready-made unit in the target language. Such a unit may also be missing, and a literal translation may lend a foreign flavour to the translation. Surprisingly, then, translationese may be the result of the translator’s “creativity”, i.e. ignorance of the ready-made phrases expected in the given register. Because of its prescriptive and pejorative connotations, the term translationese is usually avoided in present-day Translation Studies. Translated language, however, is the subject of much descriptive research. The object of such research is not to determine how translations are worse than non-translated texts, but to find out how they are different. This is a perfectly natural scientific attitude, but one must not forget the sociolinguistic fact that the users of translations are not descriptive linguists and have negative attitudes to the language variety they call translationese. On the other hand, the difference between translated and original may not be as easy to spot as people sometimes think: non-translated texts have been identified as translations (Pápai 2001). This could also be the result of the fact that many “original” texts are now composed using background material written in another language, and so the difference between non-translated and translated language is disappearing. However, it is important to remember that the texts used in these comparisons could very well represent different quality levels: a poorly written “original” text might be identified as a poor translation, while a well-written translation may sound better than a poorly-written non-translated text. Again, if judges in an experiment know that some texts are translations, they will tend to blame poor quality on translation.
Ready-made language and translation
5.
Translation evaluation and translator training
The quality of translation is subject to negotiation between the client and the translator, and translationese may be tolerated under certain conditions (Heltai 1997, 1999), but in translator training some prescriptivism is unavoidable. Translator training is based on the belief that untrained translators will usually provide lower quality translation than trained translators, so there is a difference between good and bad. With respect to ready-made language this means that we must develop a phraseological competence (cf. Howarth 1996) in trainee translators that includes knowledge of ready-made phrases used in various registers in both languages, an ability to match them and an ability to evaluate them from the sociolinguistic point of view. This means that translators should be aware of the fact that while all technical registers presuppose a certain degree of routinization, they are expected to hit the target accurately: a higher or lower degree of routinization might give grounds for complaint. The traditional complaints of prescriptive stylistics should not be taken at face value, but a trained translator must know when and how much ready-made language and which ready-made units they can use. Without such a phraseological competence translation under time pressure would be impossible — and in real life there are hardly any translation jobs without time pressure. In translation quality assessment, the efficient and judicious use of ready-made language, i.e. correspondence to the expected degree of routinization could also be used as an important criterion.
6. A pilot test In investigating the acceptability and characteristic features of translation under time pressure (Heltai 1999) a pilot experiment was carried out to explore the differences between translation of technical texts under time pressure and under non-stressful conditions, and to find the characteristics of minimally acceptable translation (“minimal translation”, Heltai 1997). As a spin-off, this pilot test also provided some information on the role of ready-made language. The study conducted by the author involved 25 trainee translators (18 female and 7 male students aged 22–23). The subjects were in their fourth year of training of a five-year translator training course at the Gödöllo˝ University of Agriculture, Hungary. The students were asked to translate the section Adaptations of the entry for forest from Encyclopaedia Britannica without a dictionary (see Appendix 2). The text can be classed as a technical text, although not too close to the prototype, since it was not specifically chosen to study ready-made language. The instructions required the students to “translate as much of the text as they could possibly do, not minding too much about quality” within one hour. It was also announced that this was a competition, the winner being the one who translated the most, provided the
61
62
Pál Heltai
translation was acceptable. After a lapse of two weeks the same students were asked to translate the same text again as a home assignment. The present author undertook a subjective quality analysis, comparing the translations of different students and the two versions written by the same student. The analysis was focused on exploring deficiencies that would not affect the comprehensibility of the translation (problems in word order, unusual collocations and other minor lexical problems). I must stress again that the test was not specifically designed to explore the role of ready-made language in translation, but in the course of analysing the translations certain characteristic tendencies emerged. Further experiments aimed explicitly at exploring the role of ready-made language in translation are needed to test the preliminary findings described below. In the following, the translation written under time pressure will be called instant translation, while the second one, written with no time constraint and with access to translation tools will be referred to as elaborated translation. 6.1 Results of the pilot test Analysis of the translations showed that the more proficient translators were able to use automatic shifts, while less proficient translators proved less skilled in this respect. (More proficient and less proficient refers to the students’ previous record during the course.) The most important automatic shift, apart from grammatical transpositions, seemed to be addition: apparently, proficient translators will automatically add words for fluency, comprehensibility and greater explicitness, that is, they seem to use optional explicitation (Klaudy 1999) more or less automatically. Of course, it is difficult to decide on the basis of the product whether a particular shift was automatic or non-automatic. (No think-aloud protocols were taken.) In the present case automaticity is supposed on the basis of common-sense reasoning: if a student, under conditions of time pressure, translates biota as biotikus tényezo˝k (biotic factors), adding an often-used collocate to the word, it is probably automatic. Again, if the overwhelming majority of students translate the phrase the environmental factors affecting trees with the correct Hungarian word order (with the premodifying phrase preposed), it is probably the result of automatic grammatical transposition. In many cases addition involved the use of ready-made units, mostly the use of a register-specific collocation in the translation corresponding to a single lexical item in the ST. (See Appendix 5.) An interesting finding was that some additions disappeared from the elaborated versions (see Appendices 3 and 4), which suggests that when the translator has enough time to ponder and decide, the ST form may exert more influence. In analysing translation under time pressure and without time pressure Hansen
Ready-made language and translation
(2002: 17) found that some students achieved better results under time pressure. Surprisingly, students who had a high level of competence in both languages tended to translate well under time pressure, while they made a number of errors under conditions of unrestricted time. This finding lends support to the hypothesis that, depending on the personal profile of the translator, a slow and analytic translation process may be subject to a higher degree of source language interference, while a faster process (when the translator “gets into the swing”), induces the translator to think holistically, which may yield a more fluent translation. “Getting into the swing” also means finding the correct register, and the ready-made units belonging to that register. Another finding was that the translations made by the same translator in two different conditions tended to be similar, and the quality difference between the two translations, especially with the more proficient translators, was not as great as one could expect. Individual students were often consistent in their choice of lexical items and grammatical constructions, although the time lapse between the two assignments was two weeks, during which the students had no access to the SL text. Since they had to translate under time pressure in the first assignment, it is unlikely that they could have remembered how they translated the text for the first time. The apparent consistency of the translations might have several causes. One cause could be memory, but in this case it is unlikely. Another cause could be related to the nature of technical translation: technical texts allow more scope for literal translation, using only obligatory shifts, than other types of translation (cf. Cardoso de Camargo 2001). (This may also account for the fact that with experienced translators the difference between instant and elaborated translation is often slight: a translation based on automatic processes using ready-made language may be near to the final version.) It is also possible that individual technical translators have their own styles and preferences that are unaffected by text or situation. Let me emphasize again that this was a pilot test with highly tentative conclusions, based on a subjective analysis, and no sweeping generalisations should be based on them. Work on a test specifically designed to explore the role of readymade language is in progress.
7.
Summary
In the above I have looked at various psycholinguistic and sociolinguistic aspects of ready-made language in translation. The main points are summarised below. – –
Different text types exhibit different degrees of routinization. Technical registers are characterised by a high degree of routinization. In translating technical texts, automatic strategies, involving the correct use of ready-made phrases, are of paramount importance. In technical translation
63
64
Pál Heltai
–
–
– –
–
–
creative language use is limited. Translation under time pressure is impossible without the use of ready-made phrases. Efficient and economical translation depends on the use of readymade phrases. Use of ready-made phrases in technical translation is generally evaluated positively if the translator uses the register-specific ready-made units and achieves the expected degree of routinization. Deviation from the expected degree of routinization may have negative sociolinguistic consequences. One component of translationese may be excessive or inadequate use of readymade language. Inappropriate use or non-use of ready-made phrases might contribute to the impression that the translated text is “jargon” or “translationese”. Translationese can be tolerated to some extent depending on the circumstances, but in general the translator is responsible for observing target language norms both for effective communication and for sociolinguistic reasons. The translator’s phraseological competence should include knowledge of readymade phrases in different registers in both languages, an ability to match them and an ability to evaluate them from the sociolinguistic point of view through awareness of the expected degree of routinization in the given register.
Author’s Address: Pál Heltai Institute of Foreign Languages, Szent István University H-2103 Gödöllo˝, Páter K. u. 1.
[email protected] Appendix 1: Ellis (1996: 118) demonstrating the use of ready-made language in his own writing There are, however, obvious differences in emphasis between a sequencing model and that of UG. One important area of conflict concerns the units of language. The present account is much more Markovian in that it stresses the importance of chunked sequences of language in fluent language use. As-far-as the idiom principle-is- concerned/ I-have-little-more-tosay./ I-would-refer-the-reader-to-the-sections-on/ “Sequence in Discourse Learning” and “SLA of Phrases, Collocations and Idioms.” /Taking-these-points-together/ the-conclusionis-obvious/: in-a-nutshell/ it-is-important-to-note-that/ a-large-part-of- communication /makes-use-of-/ fixed-expressions./ As-far-as-I-can-see/ for-many-of-these-at-least/ thewhole-is-more-than-the-sum-of-its-parts./ The meaning of an idiomatic expression cannot be deduced by examining the meanings of the constituent lexemes. /On-the-other-hand/ there-are-lots-of phrases that/ although they can be analyzed using normal syntactic principles/ nonetheless/ are not created or interpreted that way./ Rather, /they are picked-off-
Ready-made language and translation
the-shelf/ ready-made/ because they-say-what-you-want-to-say./ /I-don’t-think-I’m-goingout-on-a-limb-here./ However, /it-is-appropriate-to-say-at-this-point/ that-much-workremains-to-be-done.
Appendix 2: The source text used in the pilot test From the entry for forest, Encyclopaedia Britannica (Macropaedia), Fifteenth Edition, 1996. Adaptations The environmental factors affecting trees are climate, soils, topography, and biota. Each species of tree adapts to these factors in an integrated way, that is, by evolving specific subpopulations adapted to the constraints of their particular environments. As discussed above, the major factor is the decrease in temperature with increasing elevation or extremes in latitude. Each subpopulation adapts to this by modifying the optimum temperature at which the all-important process of photosynthesis takes place. Many tree species that survive in unfavourable habitats actually grow better in more favourable habitats if competition is eliminated. Such trees have a low threshold for competition but are very tolerant of extremes. For example, the black spruce (Picea mariana) is found in bogs and mountaintops in the northeastern United States but cannot compete well with other trees, such as red spruce (Picea rubens), on better sites. Consequently, in the White Mountains of New Hampshire in the northeastern United States, red spruce is found at the base of the mountains and black spruce at the top, with some development of subspecies populations (hybridization) at intermediate elevations. Competition within a species (and in some cases genus) is often most intense because the individuals compete for the same environmental resources. Since trees are unable to move in search of resources, competition for available space and resources can be important. Competition above ground centres on light, space and symbionts (largely pollinators) while that below ground is over water, space, nutrients and symbionts (microorganisms such as mycorrhizae and nitrogen-fixers). The ability of a tree to coexist with other members of the species in a given habitat may depend on the diversification of the space and resources they require. In extreme environments, such as are found on mountains and the subarctic, survival depends on the physical factors of the environment, whereas in more moderate habitats biotic factors become increasingly important. Flexibility and efficiency of resource use then become more important in determining survival and reproduction.
65
66
Pál Heltai
Appendix 3: Comparison of a paragraph from the instant and the elaborated translation by one of the best translators Bold type: additions Instant translation A fákra ható környezeti tényezo˝k a következo˝k: a klíma, a talaj milyensége, a domborzati viszonyok és a biotikus tényezo˝k. Minden fafajta alkalmazkodik ezekhez a tényezo˝khöz egyfajta integrációs módon: kifejleszti a specifikus szubpopulációit, amelyek alkalmazkodtak az adott környezethez. Mint már az elo˝zo˝ekben tárgyaltuk, ilyen szempontból a legjelento˝sebb környezeti tényezo˝ a ho˝mérséklet csökkenése, amit okozhat a nagy tengerszint feletti magasság, illetve a sarkokhoz való közeledés. A szubpopulációk úgy alkalmazkodnak, hogy módosítják azt az optimum ho˝mérsékleti tartományt, amelyben a fotoszintézis lényeges folyamatai zajlanak. Elaborated translation A fákra ható környezeti tényezo˝k a következo˝k: éghajlat, talaj, domborzat és a biotikus tényezo˝k (flóra és fauna). Minden fafaj egyfajta komplex módon alkalmazkodik ezekhez a tényezo˝khöz, ami jelen esetben azt jelenti, hogy a fajokból specifikus alfajok fejlo˝dnek ki, amelyek képesek alkalmazkodni a sajátságos környezetükhöz. Mint ahogy azt már tárgyaltuk, a legfontosabb tényezo˝ a ho˝mérséklet-csökkenés, ami vagy a növekvo˝ tengerszint feletti magassággal, vagy a sarkokhoz való közeledéssel párosul. Minden egyes alfaj alkalmazkodik ehhez a környezeti tényezo˝höz azzal, hogy módosul az optimum ho˝mérséklete, amelyen minden lényeges fotoszintetikus folyamat zajlik. Back-translation of instant translation The environmental factors affecting trees are the following: climate, the quality of soil, topographic conditions and biotic factors. Each tree species adapts to these factors in a kind of integrational way (?): it evolves its specific subpopulations, which are adapted to the particular environment. As we discussed in the foregoing, from this point of view the most siugnificant environmental factor is the decrease in temperature, which may be caused by high elevation above sea level, or nearness to the poles. Subpopulations adapt by modifying the optimum temperature range in which the essential processes of photosynthesis take place. Back-translation of eleborated translation The environmental factors affecting trees are the following: climate, soil, topography and biotic factors. Each tree species adapts to these factors in a kind of complex way, which in this case means that specific subpopulations evolve from the species, which are able to adapt to their particular environment. As we discussed in the foregoing, the most important factor is the decrease in temperature, which is associated with either increasing elevation above sea level, or nearness to the poles. Each subpopulation adapts to this environmental factor by a modification of the optimum temperature, at which all the essential photosynthetic processes take place.
Ready-made language and translation
Appendix 4: Comparison of the instant and the elaborated translation by one of the best translators Underlined parts: completely identical formulations. Instant translation A fákra ható környezeti tényezo˝k a következo˝k: a klíma, a talaj milyensége, a domborzati viszonyok és a biotikus tényezo˝k. Minden fafajta alkalmazkodik ezekhez a tényezo˝khöz egyfajta integrációs módon: kifejleszti a specifikus szubpopulációit, amelyek alkalmazkodtak az adott környezethez. Mint már az elo˝zo˝ekben tárgyaltuk, ilyen szempontból a legjelento˝sebb környezeti tényezo˝ a ho˝mérséklet csökkenése, amit okozhat a nagy tengerszint feletti magasság, illetve a sarkokhoz való közeledés. A szubpopulációk úgy alkalmazkodnak, hogy módosítják azt az optimum ho˝mérsékleti tartományt, amelyben a fotoszintézis lényeges folyamatai zajlanak. Számos fafajta, amely kedvezo˝tlen élo˝helyen is jól megél, a kedvezo˝ élo˝helyen ha nincs verseny az élo˝helyért, akkor nagyon jól fejlo˝dik. Az ilyen fák nem tolerálják az élo˝helyért való versengést, de jól tu˝rik a szélso˝séges környezeti viszonyokat. Például a feketefenyo˝ (Picea mariana) megtalálható lápokban és hegyteto˝kön is az Egyesült Államok északkeleti részén, de nem tudnak más fákkal, például a? — fenyo˝vel (Picea rubens) versenyezni jobb élo˝helyeken. Ezért a Picea rubens Amerika északkeleti részén, a New Hampshire-i White Mountains-ban a hegyek lábánál él, a fekete fenyo˝ viszont a hegyteto˝n, illetve hibridizációval keletkezett szubpopulációi elo˝fordulnak még közepes magasságokban is. A fajok közötti versengés (néhány esetben beszélhetünk törzsek közötti versengésro˝l) a legintenzívebb, mert az egyedek ugyanazokért a környezeti forrásokért versengenek. Mivel a fák képtelenek megkeresni a szükséges ero˝forrásaikat, ezért a lényeges harc a helyért és a felhasználható ero˝forrásokért folyik, míg a föld alatti versengést a vízért, a helyért, a tápanyagokért és a szimbiontákért (mikroorganizmusokért, például mikorrhizákért és nitrogénköto˝ baktériumokért) való küzdelem jelenti. Elaborated translation A fákra ható környezeti tényezo˝k a következo˝k: éghajlat, talaj, domborzat és a biotikus tényezo˝k (flóra és fauna). Minden fafaj egyfajta komplex módon alkalmazkodik ezekhez a tényezo˝khöz, ami jelen esetben azt jelenti, hogy a fajokból specifikus alfajok fejlo˝dnek ki, amelyek képesek alkalmazkodni a sajátságos környezetükhöz. Mint ahogy azt már tárgyaltuk, a legfontosabb tényezo˝ a ho˝mérséklet-csökkenés, ami vagy a növekvo˝ tengerszint feletti magassággal, vagy a sarkokhoz való közeledéssel párosul. Minden egyes alfaj alkalmazkodik ehhez a környezeti tényezo˝höz azzal, hogy módosul az optimum ho˝mérséklete, amelyen minden lényeges fotoszintetikus folyamat zajlik. Sok fafaj, amely kedvezo˝tlen élo˝helyen is megél, valójában kedvezo˝bb körülmények között jobban fejlo˝dne, feltéve, hogy nincs konkurencia. Ezek a fák kevéssé képesek tolerálni a konkurenciát, de jól viselik a szélso˝séges körülményeket. Például a fekete luc (Picea mariana) megtalálható Északkelet-Amerikában mocsaras területeken és hegyteto˝kön egyaránt, viszont ez a faj más élo˝helyeken nem képes versenyezni más fafajokkal, például a vörös luccal (Picea rubens). Következésképpen a new hampshire-i White Hegységben, az
67
68
Pál Heltai
Egyesült Államok északkeleti részén a vörös luc a hegyek lábánál található, miközben a lejto˝kön alfaj populációk kialakulása figyelheto˝ meg (hibridizáció). A versengés a fajok között, illetve egyes esetekben nemzetségek között a legjelento˝sebb, mert az egyedek ugyanazokért a környezeti ero˝forrásokért versengenek. Mivel a fák nem képesek az aktív mozgásra, hogy így megkeressék a szükséges ero˝forrásaikat, ezért nagyon jelento˝s lehet a versengés a rendelkezésre álló termo˝helyért és a szükséges ero˝forrásokért. A föld feletti verseny a fényért, a térért és a szimbionta szervezetekért (leginkább megporzók) folyik, míg a földfelszín alatt a vízért, a rendelkezésre álló térért, tápanyagokért és a szimbiontákért folyik (olyan mikroorganizmusokért, mint a mikorrhizák és más nitrogén-köto˝k).
Appendix 5: The translation of selected items by the three most proficient and the three least proficient students Most proficient students Student A
Student B
Student C
SL expression Addition in TT factors are
tényezo˝k a következo˝k
soils
a talaj milyensége
topography
domborzati viszonyok
földrajzi viszonyok
biota
biotikus tényezo˝k
élo˝ szervezetek
major factor
környezeti tényezo˝
elevation
tengerszint feletti magasság
temperature
ho˝mérsékleti tartomány
threshold for competition
élo˝helyért folytatott verseny
versenykényszer
extremes
szélso˝séges környezeti viszonyok
szélso˝séges környezeti viszonyok helyváltoztató mozgás szimbionta társszervezetek
move symbionts nitrogen fixers nitrogénkötõ baktériumok
tengerszint feletti magasság ho˝mérsékleti tartomány más fajokkal folytatott verseny szélso˝séges körülmények hasznos szervezetek
Ready-made language and translation
Least proficient students Student X
Student Y
Student Z
SL expression Addition in TT factors are soils topography biota major factor elevation temperature threshold for competition extremes
földrajzi viszonyok
szélso˝séges viszonyok
move symbionts nitrogen fixers nitrogénköto˝ baktériumok
References Aitchison, J. 1987. Words in the Mind. Oxford: Blackwell. Baker, M. 1992. In Other Words. London: Routledge. Bolinger, D. 1976. “Meaning and Memory”. Forum Linguisticum 1 (1): 1–14. Bot, K. de, Cox, A., Ralaton, S., Shaufeli A. and Weltens, B. 1995. “Lexical processing in bilinguals”. Second Language Research 11 (1): 1–19. Cardoso de Camargo, D. 2001. “Corpus-based translation research on legal, technical and journalistic texts.” Paper presented at the EST Congress in Copenhagen, 30 August to 1 September, 2001. Chafe, W. 1968. “Idiomaticity as an anomaly in the Chomskyan paradigm.” Foundations of Language 4: 109–26. Ellis, N. 1996. “Sequencing in SLA: Phonological memory, chunking, and points of order”. Studies in Second Language Acquisition 18 (1): 91–126. Gile, D. 1995. Basic Concepts and Models for Interpreter and Translator Training. Amsterdam and Philadelphia: John Benjamins. Gutt, E.-A. 1991. Translation and relevance: Cognition and context. Oxford: Blackwell. Gutt, E.-A. 2000. Translation and Relevance. St. Jerome Publishing: Manchester and Boston. Hansen, G. 2002. “Selbstaufmerksamkeit im Übersetzungsprozess”. Copenhagen Studies in Language 27. Copenhagen: Samfundslitteratur. 9–27. Hatim, B. and Mason, I. 1990. Discourse and the Translator. London: Longman.
69
70
Pál Heltai
Heltai, P. 1997. “Minimal translation”. In A 2. Transferre Necesse Est konferencia elo˝adásai (Proceedings of the 2nd Transferre Necesse Est Conference), K. Klaudy and J. Kohn (eds), Budapest: Scholastica. Heltai, P. 1999. “Evaluation of technical translations”. Traduction — Transition. Actes du XVe congrés mondial de la Fédération Internationale des traducteurs (FIT), Mons (B), 6–10 aout 1999, Volume 1, 169–174. Heltai, P. 2000. “Translating official texts”. In Szaknyelvoktatás és szakfordítás. Tanulmányok a Szent István Egyetem Alkalmazott Nyelvészeti Tanszékének kutatásaiból, J. Dróth (ed), Gödöllo˝: Szent István Egyetem. Howarth, P. A. 1996. Phraseology in English Academic Writing. Tübingen: Max Niemeyer. Klaudy, K. 1987. “Fordítás és aktuális tagolás”. (Translation and topic-comment structure) Nyelvtudományi értekezések 123. Budapest: Akadémiai Kiadó. Klaudy, K. 1994. A fordítás elmélete és gyakorlata. (Theory and Practice of Translation). Budapest: Scholastica. Klaudy, K. “1999. Az explicitációs hipotézisro˝l”. (On the Explicitation Hypothesis) Fordítástudomány 1 (2): 5–21. Kussmaul, P. 1995. Training the Translator. Amsterdam and Philadelphia: John Benjamins. Levelt, W. 1989. Speaking: From Intention to Articulation. Cambridge, MA.: Bradford Books. Lewis, M. 1993. The Lexical Approach. London: LTP. Lewis, M. 1997. “Pedagogical implications of the Lexical Approach”. In Second Language Vocabulary Acquisition, J. Coady and T. Huckin (eds), Cambridge: Cambridge University Press. Moon, R. 1997. “Vocabulary connections: multi-word items in English.” In Vocabulary. Description, Acquisition and Pedagogy, N. Schmitt, and M. McCarthy (eds), Cambridge: Cambridge University Press. Nattinger, J. and DeCarrico, J. 1992. Lexical Phrases and Language Teaching. Oxford: Oxford University Press. Nida, E. A. 1964. Toward a Science of Translating. Leiden: Brill. Nida, E. A. and Taber, C. R. 1969. The Theory and Practice of Translation. Leiden: Brill. Orwell, G. 1946/1978. “Politics and the English language.” Reprinted in Language Awareness, P. Escholz, A. Rosa, and V. Clark (eds), New York: St. Martin’s Press. Pápai, V. 2001. Universals of Translated Texts. Az explicitációs hipotézis vizsgálata angol-magyar és magyar-magyar párhuzamos korpuszok egybevetésével. (Investigating the explicitation hypothesis using English-Hungarian and Hungarian-Hungarian corpora). Unpublished dissertation, Gyo˝r — Pécs. Pawley, A. and Syder, F. H. 1983. “Two puzzles for linguistic theory: nativelike selection and nativelike fluency.” In Language and communication, J. C. Richards and R. W. Schmidt (eds), 191–221. London: Longman. Peters, A. 1983. The units of Language Acquisition. Cambridge: Cambridge University Press. Raupach, M. 1984. “Formulae in second language speech production.” In Second Language Production, H. Dechert, D. Möhle, and M. Raupach (eds), Tübingen: Günter Narr. Schmid, M. S. 1999. Translating the Elusive: Marked Word Order and Subjectivity in EnglishGerman Translation. Amsterdam and Philadelphia: John Benjamins. Schreuder, R. and Weltens, B. 1993. The Bilingual Lexicon. Amsterdam: John Benjamins. Sinclair, J. 1991. Corpus, Concordance, collocation. Oxford: Oxford University Press. Singleton, D. 1999. Exploring the Second Language Mental Lexicon. Cambridge: Cambridge University Press. Snell-Hornby, M. 1988/1995. Integrated Translation Studies. Amsterdam and Philadelphia: John Benjamins.
Ready-made language and translation
Tirkkonen-Condit, S. 1992. “A theoretical account of translation — without translation theory?” Target 4 (2): 237–245. Toury, G. 1980. In Search of a Theory of Translation. Tel-Aviv: The Porter Institute. Toury, G. 1995. Descriptive Translation Studies — And beyond. Amsterdam and Philadelphia: John Benjamins. Weinert, R. 1995. “The role of formulaic language in second language acquisition.” Applied Linguistics, 16 (2): 180–205. Wong-Fillmore, L. 1976. The Second Time Around: Cognitive and Social Strategies in Second Language Acquisition. Doctoral dissertation, Stanford University.
71
Les attributs indirects en français et en danois Différences typologiques et problèmes de traduction Hanne Korzen Copenhagen Business School, Denmark
This article examines the problems related to the translation of free adjuncts from French into Danish. Although free adjuncts exist in both languages, Danish accepts them to a considerably lesser extent than French. Thus the translator, when confronted with a free adjunct in the source language, first has to decide whether he can maintain the construction in Danish. If this is not the case, he has to choose between different kinds of transpositions. In order to establish rules capable of guiding the translator, the author gives a systematic description of the free adjuncts in the two languages and proposes to divide them into two types: The descriptive type and the relational type. It appears that only the former type is usual in Danish, whereas both types are frequent in French. The relational type normally has to be rephrased, and different kinds of transpositions are discussed.
0. Introduction Le but du présent travail est de discuter les problèmes que peut poser la traduction vers le danois des attributs indirects du sujet en français tels que (1) et (2): (1) a. Dupont est rentré ivre. b. Ivre-mort, Dupont descendait le Boulevard des Italiens. (2) a. Ivre-mort, Dupont était incapable de conduire. b. Travailleur méthodique, Dupont augmentait sa clientèle. c. Enfant, Dupont vivait avec sa grand-mère. Extrêmement fréquent en français, l’attribut libre existe, certes, en danois, mais d’une façon bien plus rare. Il y a, en effet, une différence statistique énorme, dans l’emploi de cette forme syntaxique, entre les deux langues ou, d’une façon plus
74
Hanne Korzen
générale, entre les langues romanes et les langues scandinaves (cf. Nielsen 1990, Eriksson 1997, I. Korzen 1998 et 1999, H. Korzen 2000a. et 2000b.).
1.
Description générale
La construction en question relève de la prédication seconde dans la mesure où l’attribut indirect joue le rôle de prédicat, formant un nexus avec le sujet de la phrase primaire. On peut parler d’une prédication “greffée” sur la prédication primaire (cf. Herslund 2000: 89): (3) Dupont est rentré/descendait le Boulevard des Italiens/était incapable de conduire + Dupont était ivre Dupont augmentait sa clientèle + Dupont était un travailleur méthodique Dupont vivait avec sa grand-mère + Dupont était enfant1 Si les constructions sous (1) et (2) contiennent toutes deux prédications: la prédication primaire et la prédication seconde, la relation qui s’établit entre ces deux prédications semble varier. Ainsi, dans les constructions (1), les attributs indirects (du moins dans la lecture par défaut) ne font que fournir une description du référent du sujet, valable pour l’intervalle de temps où se déroule le procès exprimé par la prédication primaire. Dans ce cas, je dirai que l’attribut indirect est “descriptif”. Pour ce qui est des constructions (2), il s’établit, en outre, une relation logique ou temporelle entre les deux contenus propositionnels (‘comme il était ivre-mort/un travailleur méthodique, …’, ‘quand il était enfant, …’). Dans ce cas, je dirai que l’attribut indirect est “circonstanciel” (cf. Combettes 1998:47). La distinction entre les deux types d’attributs indirects n’est, toutefois, pas absolue. D’une façon générale, un attribut indirect peut être plus ou moins descriptif ou plus ou moins circonstanciel. L’attribut indirect étant à lui seul, sous-déterminé quant au sens, les différentes nuances de sens attribuées à la construction sont le résultat d’une interprétation faite à partir des contenus sémantiques de l’attribut indirect et du contexte.
2.
Les différents types de constructions et leurs traductions
J’émettrai, d’emblée, l’hypothèse que les attributs indirects en danois sont, avant tout, descriptifs, et que plus un attribut indirect s’approche du “pôle” circonstanciel, moins il paraîtra acceptable.2 Dans ce qui suit, nous verrons que cette différence entre le français et le danois ne constitue pas un phénomème isolé, mais qu’elle est liée à une différence d’ordre typologique beaucoup plus générale. Regardons maintenant les différents types de constructions susceptibles de fonctionner comme attribut indirect.
Les attributs indirects en français et en danois
2.1 Syntagmes substantivaux et adjectivaux Incapable de donner une description du référent du sujet limitée au moment où se déroule le procès, un syntagme substantival fonctionnant comme attribut indirect sera nécessairement circonstanciel (cf. (2b et c)). Un tel syntagme ne peut jamais fonctionner comme attribut indirect en danois. Le traducteur est donc obligé de recourir à une transposition. Le plus souvent, il optera pour une subordonnée adverbiale appropriée. Ainsi, une construction comme (2b) se traduira comme (2b¢): (2) b¢. Da F. gik systematisk til værks, fik han sin kundekreds til at vokse. ‘Comme F. était un travailleur méthodique, il …’3 Une autre solution consiste à rendre le contenu de l’attribut indirect par une relative, qui peut, dans les deux langues, avoir une nuance causale: (2) b≤. F., som gik systematisk til værks, fik sin kundekreds til at vokse. ‘F., qui était un travailleur méthodique, …’ En se servant d’une adverbiale, le traducteur explicite la relation logique qui existe entre l’attribut indirect et la prédication primaire, relation qui est implicite en français. En se servant d’une relative, il atténue, au contraire, cette relation logique, car un attribut indirect antéposé acquiert plus facilement une nuance causale qu’une relative. La construction à attribut indirect occupant ainsi une position intermédiaire entre l’adverbiale et la relative pour ce qui est de l’explicité, le traducteur est condamné soit à la “sur-traduction”, soit à la “sous-traduction” (cf. Lundquist 2001). Pour ce qui est du syntagme adjectival fonctionnant comme attribut indirect, il peut être soit descriptif (cf. (1)), soit circonstanciel (cf. (2a.)).4 Un attribut descriptif se traduit sans problèmes par une construction équivalente en danois: (1¢) a. Dupont kom fuld hjem. b. Døddrukken, gik Dupont ned ad Boulevard des Italiens. Quand l’attribut est circonstanciel, le traducteur doit avoir recours aux mêmes types de transpositions que dans le cas de (2b) (déjà illustrées par (2b¢ et 2b≤). 2.2 Les syntagmes participiaux Les syntagmes participiaux dont le noyeau est un participe passé se rencontrent très fréquemment comme attribut indirect dans les deux langues. Le verbe qui entre dans un tel syntagme est souvent un verbe transitif employé au passif (exprimant soit le procès (cf. (4), soit l’état qui en résulte (cf. (5)). Mais il peut aussi s’agir d’un verbe inaccusatif, le participe exprimant alors le résultat d’une action préalable (cf. (6) et (7)). Dans des cas comme (4) et (5), où l’attribut indirect fournit une
75
76
Hanne Korzen
description du référent du sujet au moment où se déroule l’action dénotée par la prédication primaire, on peut, sans aucun problème, garder la même construction en danois: (4)¢ Mon neveu est arrivé par le train, accompagné de sa femme. (4¢) Min nevø kom med toget ledsaget af sin kone. (5)¢ Muni d’un litre de gin, il s’est enfin mis en route. (5¢) Udstyret med en liter gin begav han sig endelig på vej. Dans des cas comme (6) et (7), où le participe passé exprime une action préalable, la meilleure traduction sera ou bien celle qui fait de l’attribut et de la prédication primaire deux phrases coordonnées (cf. (6¢)) ou bien celle qui fait de l’attribut indirect une temporelle (cf. (7¢)): (6)¢ Né en Pennsylvanie en 1958, il y expose ses premiers dessins à l’âge de 19 ans. (6¢) Han blev født i Pennsylvania i 1958 og udstillede 19 år gammel sine første tegninger der. ‘Il naquit en P. en 1958 et y exposa …’ (7)¢ Arrivé au parapet de la rue Beethoven, il s’arrêta (Green, Eriksson 1993: 134) (7¢) Da han var ankommet til rækværket ved rue B., standsede han. ‘Quand il fut arrivé au parapet (…), il …’ Comme nous le verrons dans 3. ci-dessous, il est, cependant, possible de traduire des exemples comme (6) et (7) littéralement. Mais on peut se demander s’il s’agit là d’une bonne traduction. Pour ce qui est des participes présents comme attributs indirects, la construction est bien plus fréquente en français qu’en danois. Quand l’attribut est descriptif, il se traduit directement: (8)¢ Tremblant de tous ses membres, elle attendait le verdict du jury. (8¢) Skælvende over hele kroppen, ventede hun på nævningenes dom. Cependant, dans la plupart des cas, une transposition s’impose, car le participe présent comme attribut indirect est typiquement circonstanciel. Cela vaut pour les construction comme (9), où le syntagme participial peut être analysé comme un coverbe qui forme avec le verbe principal un prédicat complexe (cf. Herslund 2000):5 (9)¢ Et prenant Selma par le bras, elle l’entraîne d’autorité (Mourad, cit. Herslund op. cit.: 91) Une telle construction se traduit normalement par une phrase coordonnée à la prédication primaire:
Les attributs indirects en français et en danois
(9¢) Og hun tager Selma i armen og trækker hende uden videre med ‘et elle prend S. par le bras et …’ Et, d’une façon générale, cela vaut pour toutes les constructions contenant une négation ou un objet direct ou pour celles dont le participe est à la forme composée. Ainsi, l’attribut indirect de (10) ne saurait se traduire que par une causale: (10)¢ N’ayant pas compris la question du professeur, il a été incapable de répondre. (10¢) Da han ikke havde forstået lærerens spørgsmål, kunne han ikke svare. ‘Comme il n’avait pas compris …, …’ Dans le style littéraire d’autrefois, on rencontrait, en danois, des constructions rappelant les constructions françaises (9) et (10). Ainsi, les participes présents contenant un objet direct n’étaient pas rares: (11) Hæren drog videre, ødelæggende alt på sin vej (Mikkelsen 1911: 412) ‘L’armée continua sa marche, détruisant tout sur son passage’ Et on rencontrait même des participes composés: (12) … havende begrebet, saá han i et nu … (Bang, cit. Hansen 1967) ‘… ayant compris, il réalisa…’ Mais (11) est tombé en désuétude, et (12) est carrément exclu de nos jours. Ce genre de constructions étaient calquées sur le latin. Aujourd’hui le danois subit une autre influence, ainsi que nous le verrons tout à l’heure.
3.
Considérations typologiques
Comme je l’ai déjà fait remarquer, la très grande différence dans l’emploi des attributs libres, en français et en danois, est liée à une différence d’ordre typologique beaucoup plus générale entre les langues romanes et les langues scandinaves. Pour expliquer cette différence, je me suis appuyée sur le modèle proposé par I. Korzen (1998, 1999 et 2000), qui a fait une étude contrastive entre l’italien et le danois. Ce modèle rend compte des manières différentes de combiner deux ou plusieurs contenus propositionnels. Ainsi, celui qui désire combiner les contenus de (13): (13) Le vent s’est levé + Nous avons pu partir peut choisir la coordination ou la juxtapposition: (14) Le vent s’est levé, (et) nous avons pu partir. Dans ce cas, où les deux informations sont présentées au même niveau, et où le soin
77
78
Hanne Korzen
de démêler la relation qu’elles entretiennent entre elles est laissé à l’allocutaire, nous avons une structure linéaire, une simple “liste”. Mais le locuteur peut aussi opter pour une présentation plus explicite, subordonnant l’un des contenus à l’autre de façon à créer une structure hiérarchique: (15) a. Comme le vent s’était levé, nous avons pu partir. b. Le vent s’étant levé, nous avons pu partir. etc. Par la coordination ou la juxtaposition, les deux contenus propositionnels seront représentés par deux principales, chacune avec un verbe fini. Par la subordination, le verbe subordonné perdra un ou plusieurs des traits qui caractérisent un verbe fini se trouvant dans une principale. Pour illustrer cette réduction, I. Korzen (1998: 68) propose deux échelles, une qui indique les différentes constructions de moins en moins “phrastiques” (Korzen parle, ici, avec Lehmann 1988, de “desententialisation”, cf. (16)), et une qui représente les traits auxquels le verbe doit renoncer à mesure que l’on descend l’échelle en question (Korzen parle ici de “déverbalisation”, cf. (17)): (16) 0. 1. 2. 3.
verbe fini dans une principale verbe fini dans une subordonnée verbe non fini nominalisation
(17) 0. 1. 2. 3.
verbe prototypique perte d’autonomie illocutoire, modale, temporelle et aspectuelle perte de désinences verbales et de sujet explicite réduction des actants en constituants “secondaires”
Les attributs indirects se trouvent tous au niveau 2. de cette échelle, étant donné que ce sont ou bien des syntagmes verbaux non finis (comme p. ex. (9): prenant S. par le bras, …) ou bien des syntagmes participiaux réduits, le verbe être ayant été omis (cf. p. ex. (2a.), qui peut être considéré comme une réduction de: Étant ivre-mort, Dupont était incapable de conduire). D’une façon générale, les langues latines ont une forte tendance à descendre plus bas sur l’échelle indiquée que les langues scandinaves, et notamment dans la langue écrite, elles préfèrent une structure fortement hiérarchisée. Nous avons pu constater, dans ce qui précède, que l’attribut indirect en français (qui représente le niveau 2.), se traduit le plus souvent en danois ou bien par une proposition subordonnée (qui représente le niveau 1.) ou bien par une proposition principale coordonnée (qui représente le niveau 0.). Cette différence entre les langues romanes et les langues scandinaves s’explique naturellement par des différences morpho-syntaxiques importantes. Les langues romanes possèdent des constructions qui sont inexistantes dans les langues
Les attributs indirects en français et en danois
scandinaves, telles que la construction absolue (cf. (15b.)) et le gérondif (Il chante en se rasant). L’existance de ce type de constructions, qui permettent de comprimer le message de manière à placer plusieurs contenus propositionnels dans une seule et même phrase a, sans aucun doute, contribué à renforcer la tendance générale à la “compression”, favorisant par là même l’emploi étendu de l’attribut indirect.6 Mais la différence mentionnée entre les deux groupes de langues s’explique, comme le souligne I. Korzen, également par une tradition rhétorique tout à fait différente, la tradition latine, qui favorise à la fois la clarté et la concision, ayant fortement empreint tous les territoires de langues romanes et exercé une influence énorme sur leurs traditions scolaires. Il est intéressant de remarquer que l’emploi de l’attribut indirect (tant circonstanciel que descriptif) est quasiment aussi étendu en anglais qu’en français (cf. Brøndum 2001 et Vestergaard 1997). Quoiqu’il y ait des différences subtiles entre les deux langues (cf. Brøndum op. cit.), on retrouve en anglais les mêmes types de syntagmes dans cette fonction, à savoir des syntagmes substantivaux (A graduate of Groton and Stanford, the handsome young Bostonian had been a natural for The Bachelors when he moved to San Francisco as Bank of America trainee in 1971 (Maupin, cit. Brøndum op. cit.: 62)), des syntagmes adjectivaux (Cold, he borrowed Hardy’s sweater and put it on top of his (Ondaatje, cit. ib.: 32)), des participes passés (Emboldened, Arafat decided to hang with his original proposal (Time, cit. ib.: 36)), des participes présents (Israel’s army flubbed its response, at first claiming none of its soldiers could have been the one that killed the boy (Time, cit. ib.: 38)), et même des constructions absolues (… the young soldier, his eyes not leaving their focus, put out his palm and snapped his fingers (Ondaatje, cit. ib.: 40)). Cette différence entre l’anglais et les autres langues germaniques7 est, comme l’explique Vestergaard (op. cit.), due à la grande influence que le français a exercée sur l’anglais. Mais maintenant c’est l’anglais qui, à son tour, influence les autres langues, dont le danois. Comme je l’ai déjà fait remarquer, il est possible de traduire des constructions comme (6) et (7) directement: (6≤) Født i Pennsylvania udstiller han 19 år gammel sine første tegninger der. (7≤) Ankommet til rækværket ved rue B. standsede han. De telles constructions semblent quelque peu “traduites”. Force est cependant de constater que ce type d’attributs circonstanciels deviennent de plus en plus fréquents en danois, notamment dans le langage journalistique, phénomène qui est probablement dû à l’influence de l’anglais. Après avoir subi l’influence du latin, le danois subirait donc, maintenant, l’influence de l’anglais. Il faut pourtant souligner que l’attribut indirect circonstanciel est toujours un phénomène relativement rare en danois sauf dans quelques expressions plus ou moins consacrées telles que (6≤) et (7≤). Nous sommes (encore?) très loin de nous rapprocher de la structure fortement hiérarchisée des langues latines.
79
80
Hanne Korzen
Author’s Address: Hanne Korzen Dalgas Have 15 DK-2000 Frederiksberg e-mail:
[email protected] Notes 1. Pour une discussion des positions que peut occuper l’attribut indirect dans la chaîne linéaire (intégrée ou détachée), on est renvoyé à H. Korzen (2000a. et 2000b.). 2. Cette particularité de l’attribut indirect en danois ressort des descriptions qu’en donnent les grammairiens danois (cf. Hansen 1967 et Mikkelsen 1911). Cependant, l’hypothèse est, en outre, corroborée par l’étude d’un corpus danois (cf. H. Korzen 2000a.). 3. Pour ce qui est de l’attribut libre de (2c.), elle se traduira ou bien par une temporelle ou bien par une construction introduite par l’opérateur som: Da han var barn/som barn … ‘Quand il était enfant, …’. 4. Un adjectif dénotant une propriété essentielle, c’est-à-dire une propriété non-transitoire et constitutive chez un individu, provoque presque toujours la lecture circonstancielle: Petite, cette main devait appartenir à un enfant (‘Comme la main était petite …’). Par contre, un adjectif dénotant une propriété accidentelle, c’est-à-dire un état passager, valable pour la seule situation décrite par le verbe en question, peut, comme le montre (1) et (2.a), prendre les deux valeurs. 5. Pour une analyse détaillée de la syntaxe de cette construction et des nuances de sens qui peuvent s’y rattacher, on est renvoyé à Herslund. 6. Le caractère compact d’un texte latin par rapport à un texte scandinave ressort également de la fréquence élevée de nominalisations, qui, très souvent, ne peuvent pas se traduire littéralement en danois, mais doivent se rendre par des propositions subordonnées: L’achat de Citroën par Peugeot a choqué beaucoup de gens — Det har chokeret mange at Peugeot har købt Citroën ‘Cela a choqué beaucoup de gens que P. a acheté C.’. 7. Selon Vestergaard (op. cit.), l’attribut indirect serait encore moins courant en allemand que dans les langues scandinaves.
References Brøndum, Mette Qvist. 2001. Et sammenlignende studie af frie prædikativer på fransk og engelsk. Maîtrise: Université d’Aarhus, juin 2001. Combettes, Bernard. 1998. Les constructions détachées en français. Collection l’essentiel français. Ophrys. Eriksson, Olof. 1993. La Phrase Française. Essai d’un inventaire de ses constituants syntaxiques. Göteborg: Acta Universitatis Gothoburgensis. Eriksson, Olof. 1997. Språk i kontrast. En jämförande studie av svensk och fransk meningsstruktur. Göteborg: Akademiförlaget. Hansen, Aage. 1967. Moderne dansk, I–III. Copenhague: Grafisk Forlag.
Les attributs indirects en français et en danois
Herslund, Michael. 2000. “Le participe présent comme co-verbe”. In: Langue française 127, édité par Cardiot Pierre & Naoyo Furukawa: La prédication seconde, 86–94. Korzen, Hanne. 2000a. En kontrastiv analyse af frie prædikativer på dansk og fransk. Copenhagen Working Papers in LSP 6–1999/2000. École des hautes études commerciales de Copenhague. Korzen, Hanne. 2000b. “Les attributs indirects en français et en danois: problèmes de traduction”. A paraître dans Voprosy Filologii, Journal of Philology 2000/3. Moscou. Korzen, Iørn. 1998. “On grammaticalisation of rhetorical sattelites. A comparative study on Italian and Danish”. In: Korzen, Iørn & Michael Herslund (eds): Clause Combining and Text Structure. Copenhagen Studies in Language 22, 65–86. Korzen, Iørn. 1999. “Tekststruktur og anafortypologi”. In: Skytte, Gunver, Iørn Korzen, Paola Polito & Erling Strudsholm (eds.): Tekststrukturering på italiensk og dansk. Resultater af en komparativ undersøgelse. Copenhague: Museum Tusculanum Press, 331–417. Korzen, Iørn. 2000. “Reference og andre sproglige relationer”. In: Skytte, Gunver & Iørn Korzen: Italiensk-dansk sprogbrug i komparativt perspektiv. Reference, konnexion og diskursmarkering Vol. II. Copenhague: Samfundslitteratur. Lundquist, Lita. 2001. Conférence donnée à Bordeaux au cours du Rencontre linguistique francodanoise du 30 au 31 mai 2001. Mikkelsen, Kristian. 1911. Dansk Ordföjningslære. Copenhague. Réédition 1975: Hans Reitzels Forlag. Nielsen, Anni Odgaard. 1990. En kritik af Karen Mathiasens oversættelse af “l’Amour en plus” med henblik på ledfunktionen frit prædikat. Maîtrise. École des hautes études commerciales de Copenhague. Vestergaard, Torben. 1997. “‘Free adjuncts’ in English”. In: Bache, Carl & Alex Klinge (eds.): Sounds, Structures and Senses. Essays Presented to Niels Davidsen-Nielsen on the Occasion of his Sixtieth Birthday. Odense University Press, 251–65.
81
Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft Michael Schreiber Heidelberg
The objective of my paper is twofold: 1. I would like to stress that linguistic translation studies still forms an important subdiscipline of Translation Studies. Although the new “claims, changes, and challenges” of our discipline seem to be so fascinating that we sometimes overlook the linguistic problems of translation, these problems continue to be omnipresent in the practice of translation and interpreting. 2. I want to show that linguistic translation studies is only useful for translation practice when it goes beyond the comparison of languages (as is done in contrastive analysis). While the object of contrastive analysis is the comparison of linguistic systems and norms, linguistic translation studies is always oriented towards the solution of translation problems. The arguments of my paper are illustrated by examples from three domains: – information structure (topic-comment) – word order (with regard to simultaneous interpreting) – modalisation. The following languages are considered: German, French, Italian, Spanish, and English.
1.
Einleitung
Mit diesem methodisch ausgerichteten Beitrag verfolge ich zwei Ziele: 1. Ich möchte darauf hinweisen, daß die sprachenpaarbezogene Translationswissenschaft (STW), d.h. diejenige Teildisziplin der Translationswissenschaft (TW), die sich mit sprachenpaarbedingten Problemen des Übersetzens und Dolmetschens beschäftigt, trotz veränderter Rahmenbedingungen einen wichtigen Platz innerhalb der TW einnimmt. Zwar haben sich in der TW im Zuge der Emanzipation dieser
84
Michael Schreiber
Disziplin und im Zusammenhang mit technischen Innovationen neue Forschungsschwerpunkte herausgebildet (z.B. kulturelle Aspekte der Translation, mentale Prozesse beim Übersetzen und Dolmetschen, computergestützte Übersetzung), so daß die rein sprachlichen Probleme des Übersetzens und Dolmetschens in den Hintergrund getreten sind (Teich 1999: 507), doch zeigt sich in der translatorischen Praxis, daß sprachenpaarbedingte Probleme nach wie vor einen großen Teil der tatsächlich auftretenden Übersetzungs- und Dolmetschprobleme ausmachen. 2. Ich möchte zeigen, daß die STW nur dann verwertbare Ergebnisse für die translatorische Praxis liefert, wenn sie über einen reinen Sprachvergleich, wie er in der kontrastiven Linguistik im engeren Sinne (KL) durchgeführt wird, hinausgeht. Während es bei der KL um einen synchronischen Sprachvergleich auf der Ebene des Sprachsystems und der Sprachnorm geht, d.h. um das Aufzeigen von strukturellen Gemeinsamkeiten und Unterschieden zweier oder mehrerer Sprachen, arbeitet die STW lösungsorientiert: Sie zeigt mögliche Translationsverfahren auf (also Lösungsmöglichkeiten für sprachenpaarbedingte Übersetzungs- und Dolmetschprobleme). Es geht also im folgenden nicht um eine generelle Diskussion des Verhältnisses von Linguistik und TW aus wissenschaftstheoretischer Sicht (hierzu vgl. Albrecht 1987 und Steiner 1999), sondern lediglich um das Verhältnis von KL und STW, d.h. um den Bereich, in dem sich beide Disziplinen berühren.
2.
Kontrastive Linguistik
Die KL ist diejenige Teildisziplin der Sprachwissenschaft, die dem synchronischen Sprachvergleich gewidmet ist. Die KL ist daher primär eine Wissenschaft der langue (Koller 1992: 223). Diese oft gemachte Feststellung ist allerdings vor dem Hintergrund der jüngeren Enwicklung der modernen Sprachwissenschaft in zweierlei Hinsicht zu relativieren. Erstens kann innerhalb der langue zwischen Sprachsystem und Sprachnorm (im Sinne Coserius) unterschieden werden: Die KL muß sich nicht auf einen Systemvergleich beschränken, denn auch die Norm (d.h. das in einer Sprache Übliche, aber nicht notwendig Funktionelle) kann in den Vergleich einbezogen werden (Coseriu 1987: 72). So muß sich eine kontrastive Untersuchung zur Wortstellung nicht damit begnügen, die grundsätzlichen Möglichkeiten der Wortstellung der betreffenden Sprachsysteme zu vergleichen, sondern es kann auch (z.B. durch quantitative Analysen) untersucht werden, welche Wortstellungsvarianten in den verglichenen Sprachen (ggf. in einzelnen Varietäten oder Textsorten) üblich und welche weniger üblich sind. Zweitens hat auch die sog. pragmatische Wende in der Linguistik zu einer Ausweitung der KL geführt (Järventausta 1996: 98f.). Neben die anfangs dominierende, systemorientierte, kontrastive Grammatik ist eine kontrastive Pragmatik getreten, in der u.a. die Realisierung bestimmter
Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft
Sprechakte (z.B. Anweisungen) in verschiedenen Sprachen untersucht wird (SnellHornby 1998: 69). Die praktische Zielsetzung der KL bezog sich zunächst fast ausschließlich auf den Fremdsprachenunterricht. Man ging dabei von der sog. Interferenzhypothese aus: “Those elements [of the foreign language] that are similar to his [the student’s] native tongue will be simple for him, and those elements that are different will be difficult” (Lado 1957: 2). Spätere Untersuchungen haben allerdings gezeigt, daß diese Hypothese in ihrer strengen Form nicht haltbar ist, da Ähnlichkeiten oft größere Lernschwierigkeiten verursachen als scharfe Kontraste (Kielhöfer 1980:121ff.). Im heutigen Fremdsprachenunterricht spielt die KL (aufgrund ihrer eher bescheidenen praktischen Erfolge, aber auch aufgrund der kommunikativen Orientierung der modernen Fremdsprachendidaktik) keine zentrale Rolle mehr. Zudem hat sich in jüngerer Zeit neben der angewandten (didaktisch orientierten) KL eine theoretische KL (ohne direkten Anwendungsbezug) etabliert (Fisiak 1990: 5f.). Es darf aber nicht vergessen werden, daß es eine Variante der KL gab, die (auch) auf das Übersetzen ausgerichtet war: die stylistique comparée. Malblanc nennt als Ziel der stylistique comparée ”la comparaison des caractères propres de deux ou plusieurs langues” (1968: 15) und als Zielgruppe seines Buches “ceux pour qui la traduction est une tâche scolaire ou professionnelle” (1968: 288). Die stylistique comparée wollte also offenkundig beides sein: KL und STW — was nicht unproblematisch ist (Schreiber 1999: 100f.). Ähnlich verhält es sich mit dem auf der stylistique comparée aufbauenden multilingualen Übersetzungsvergleich von Mario Wandruszka (vgl. z.B. Wandruszka 1969), bei dem der Übersetzungsvergleich zwar primär im Dienst des Sprachvergleichs steht, bei dem aber auch Übersetzungsprobleme als solche diskutiert werden.
3.
Sprachenpaarbezogene Translationswissenschaft
Die STW ist diejenige Teildisziplin der TW, die sich mit sprachenpaarbedingten Problemen des Übersetzens und Dolmetschens beschäftigt (vgl. Koller 1992: 125f.). Es handelt sich also nicht um eine Teildisziplin der Linguistik, wie dies in Termini wie Translationslinguistik (Jäger 1975) zum Ausdruck kommt (vgl. jetzt auch den Titel des Sammelbandes Argomenti per una linguistica della traduzione von Korzen/ Marello 2000 sowie die Ausführungen hierzu von Lita Lundquist 2000, 111f.). Es handelt sich aber m.E. auch nicht um eine eigenständige Richtung oder Schule der TW (wie etwa die funktionalistische TW), da die STW nur ein Teilgebiet der modernen TW abdeckt. Koller weist darauf hin, daß sich TW nicht auf die langue, sondern auf die parole beziehe (1992: 223). Diese Aussage läßt sich dahingehend präzisieren, daß der Translationsprozeß selbstverständlich zur Ebene der parole gehört, da nur Texte
85
86
Michael Schreiber
übersetzt werden, daß die Aufgabe der STW aber gerade im Grenzbereich zwischen langue und parole angesiedelt ist, wie aus Kollers eigener Funktionsbestimmung der “linguistischen” (= sprachenpaarbezogenen) Übersetzungswissenschaft hervorgeht: Die linguistische Übersetzungswissenschaft beschreibt die potentiellen Zuordnungsvarianten (Äquivalente) und gibt die Faktoren und Kriterien an, die die Wahl von aktuellen Entsprechungen bestimmen. (Koller 1992: 125)
Die “potentiellen Zuordnungsvarianten” gehören dabei zur langue, die “aktuellen Entsprechungen” dagegen zur parole. Einzelne Arbeiten zur STW können innerhalb dieses Grenzbereichs näher am langue-Pol oder näher am parole-Pol angesiedelt sein. Dies möchte ich anhand dreier neuerer Lehrbücher illustrieren, die sich mit dem Übersetzen ins Deutsche beschäftigen: Das Buch Übersetzen Englisch-Deutsch von Karin Königs beschreibt sprachenpaarspezifische Probleme der englisch-deutschen Übersetzung anhand zahlreicher, meist kurzer Beispiele. Ziel des Buches ist es, die jeweils in Frage kommenden Übersetzungsvarianten möglichst vollständig zu inventarisieren und damit ein Angebot potentieller Entsprechungen zusammenzustellen. Bewußt nicht thematisiert werden mögliche Kriterien, nach denen eine Entscheidung zwischen verschiedenen Varianten zu treffen ist. (Königs 2000: IX)
Die Textebene wird also bewußt ausgeblendet. Einen Schritt weiter in Richtung auf die parole geht dagegen Käthe Henschelmanns Arbeitsbuch Problem-bewußtes Übersetzen, das der französisch-deutschen Übersetzung gewidmet ist. Hier werden nicht nur sprachenpaarbedingte, sondern auch kulturell bedingte und textsortenbedingte Übersetzungsprobleme behandelt. Vor allem aber werden nicht nur potentielle Entsprechungen in Form von Übersetzungsverfahren angegeben, sondern übersetzerische Entscheidungen am Beispiel von Übersetzungen ganzer Texte diskutiert: Zur Förderung des problem-bewußten Übersetzens genügt es nicht, für jede Texteinheit Übersetzungslösungen vorzustellen. Zur weiteren Reflexion wird vielmehr die Begründung des textglobalen Vorgehens sowie der einzelnen Übersetzungsentscheidungen angeboten. (Henschelmann 1999: 13)
Noch stärker text- und kontextbezogen ist Christiane Nords Einführungskurs mit dem Titel Lernziel: Professionelles Übersetzen Spanisch-Deutsch. Diese Einführung, in der 15 spanische Ausgangstexte unterschiedlicher Textsorten behandelt werden, ist funktionalistisch ausgerichtet. Daher werden jeweils möglichst realistische Übersetzungsaufträge angegeben, wie: Das Schnupfenmittel EGARONE soll demnächst auf dem deutschen Markt vertrieben werden. Entsprechend den Bestimmungen des deutschen Arzneimittelgesetzes und im Einklang mit der EU-Gesetzgebung ist eine deutschsprachige Packungsbeilage anzufertigen. (Nord 2001: 121)
Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft
Aufgrund der funktionalistischen Ausrichtung spielen pragmatische und kulturelle (“konventionsbedingte”) Übersetzungsprobleme in Nord 2001 eine große Rolle, aber auch sprachenpaarspezifische Übersetzungsprobleme und die entsprechenden Übersetzungsverfahren werden anhand der ausgewählten Texte diskutiert. Alle drei Lehrbücher können trotz ihrer Unterschiede zur STW gezählt werden, da sie über einen reinen Sprachvergleich hinausgehen und lösungsorientiert (im Hinblick auf Übersetzungsverfahren) sind. Diese Lösungsorientiertheit ist m.E. das wichtigste Kriterium zur Abgrenzung der STW von der KL. Lösungsorientiertheit ist allerdings nicht mit Präskriptivität zu verwechseln: Die STW kann auch deskriptiv vorgehen, indem sie empirisch beschreibt, wie Übersetzer und Dolmetscher mit sprachenpaarbedingten Problemen umgehen (wie wir unter 4. anhand konkreter Beispiele sehen werden).
4. Beispieldiskussion Im folgenden sollen die oben gemachten theoretischen Erläuterungen anhand konkreter Translationsprobleme exemplifiziert werden. Ausgewählt habe ich hierfür einige Beispiele aus drei Bereichen herangezogen, für die sowohl kontrastive als auch translationswissenschaftliche Voruntersuchungen vorliegen: Thema-RhemaGliederung, Serialisierung und Modalisierung. Die Beispiele beziehen sich jeweils auf die Translation aus romanischen Sprachen oder dem Englischen ins Deutsche (und umgekehrt). 4.1 Thema-Rhema-Gliederung (Informationsstruktur) Im Bereich der Thema-Rhema-Gliederung kann die Unterscheidung zwischen KL und STW folgendermaßen skizziert werden: Während die KL aufzeigt, welche Konstruktionen in beiden Sprachen zur Thema-Rhema-Gliederung eingesetzt werden können, ist es Aufgabe der STW darzulegen, wie trotz unterschiedlicher morphologischer und syntaktischer Verfahren der Thema-Rhema-Gliederung in einer Übersetzung die Informationsstruktur bewahrt werden kann. Es geht also nicht um die Nachbildung von Strukturen, sondern um semantisch-pragmatische “Äquivalenz”. (Auf eine Diskussion des Äquivalenzbegriffs möchte ich hier verzichten. Meine Position hierzu habe ich dargelegt in Schreiber 1993: 55ff.). Als Beispiel für eine kontrastiv-linguistische Untersuchung aus diesem Bereich sei ein Aufsatz von Elisabeth Gülich zur Segmentierung (Dislokation) im gesprochenen Französischen und Deutschen erwähnt. Gülich untersucht hier anhand sprechsprachlicher Korpora Vorkommen von Konstruktionen mit redundanter Besetzung einer syntaktischen Position wie
87
88
Michael Schreiber
(1) Vous me parliez du latin tout à l’heure, qu’est-ce que vous pensez du latin à l’école? — Ben moi je trouve que c’est une bonne chose. (Orléans-Korpus, zit. nach Gülich 1982: 50) (2) Nee, das war ne totale Schnapsidee… Nur weil sie die fünfzig Mark für die nächste Miete nicht zahlen will, sondern die gleich abziehn will, ne, die… die Kaution. (Kallmeyer-Korpus, zit. nach Gülich 1982: 56) Diskutiert werden die möglichen Funktionen solcher Segmentierungen, z.B. Thematisierungs- und Überbrückungssignal am Beginn einer Äußerung (wie in Beispiel (1)) oder Disambiguierung eines mehrdeutigen Pronomens (wie bei der “Rechtsversetzung” in Beispiel (2)). Translationsprobleme werden bei Gülich nicht explizit behandelt. Die strukturellen Unterschiede im Bereich der Personalpronomina werden z.B. folgendermaßen kommentiert: Comme il n’y a pas deux séries de pronoms […], le type ‘moi je’, extrêmement fréquent en français, n’existe pas en allemand; on peut seulement répéter le même pronom, […] mais ce procédé ne semble pas être très fréquent. (Gülich 1982: 45)
Für einen Übersetzer oder Dolmetscher, der an möglichen Translationsverfahren interessiert ist, sind solche allgemeinen Aussagen zu den strukturellen Unterschieden zwischen beiden Sprachen wenig hilfreich. Aus dolmetschwissenschaftlicher Sicht wird das Phänomen der Segmentierung in einem Aufsatz von Sabine Bastian behandelt. Hier geht es um die Wiedergabe der französischen Segmentierung sowie der sog. mise en relief beim Dolmetschen ins Deutsche. Bastian führt u.a. aus, daß die hervorhebende Funktion der pronominalen Redundanz im Französischen durch stärkere Akzentuierung (Insistenzakzent) des Personalpronomens im Deutschen wiedergegeben könne (vgl. auch Schreiber 1999: 190): (3) Il ne peut pas de toute façon, malgré sa valeur qui est reconnue, quoi, être, lui, dans une université maintenant. Er kann jedenfalls jetzt trotz seiner anerkannten Fähigkeiten nicht mehr an einer Universität arbeiten. (Bastian 1988: 120) In Beispiel 3 (mit Rechtsversetzung des Pronomens lui) kann dieses Verfahren allerdings nur bei einem ausreichenden Abstand (décalage) zwischen Ausgangstext und Verdolmetschung funktionieren. Ein anderes Verfahren der Thema-Rhema-Gliederung (speziell der Rhematisierung) im Französischen ist die mise en relief mit c’est … qui/que, die auch im geschriebenen Französisch häufig zu finden ist. Da solche “Spaltsätze” im Deutschen weniger gebräuchlich sind, kann es in Übersetzungen angebracht sein, eine entsprechende Hervorhebung durch lexikalische Mittel auszudrücken, wie Gisela
Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft
Thome in einer übersetzungswissenschaftlichen Studie an Beispielen wie dem folgenden gezeigt hat (hier aus der Übersetzung eines linguistischen Fachtextes): (4) Le diachroniste… ne peut oublier que c’est dans ce domaine que se manifestent… les déséquilibres. Der Diachroniker… kann nicht vergessen, daß sich gerade auf diesem Gebiet die Gleichgewichtsstörungen zeigen. (André Martinet, zit. nach Thome 1976: 408) Unterschiede im Bereich der Thema-Rhema-Gliederung bestehen auch im Sprachenpaar Italienisch-Deutsch. Aus sprachvergleichender Sicht weist Peter Koch z.B. auf Unterschiede der Informationsstruktur von Präsentativkonstruktionen wie den folgenden hin: (5) C’è qualcuno! Da ist jemand! (Koch 1994: 48) Zu diesem Beispiel führt Koch aus: “Im Italienischen trägt das nachgestellte Subjekt den stärkeren Ton und wird damit rhematisiert, im Deutschen trägt das vorangestellte Verb den Hauptton und ist damit rhematisiert” (Koch 1994: 48). Diese Beobachtung ist aus sprachvergleichender Perspektive sicherlich interessant, für den Übersetzer jedoch nicht ausreichend. Im Rahmen der STW müßte man sich daher fragen, wie eine dem Italienischen entsprechende Rhematisierung in diesem Fall auch im Deutschen erreicht werden könnte. In dem zitierten Beispiel könnte man sich im Deutschen etwa mit lexikalischen Mitteln behelfen (falls die Rhematisierung im Kontext eine wichtige Rolle spielt), wie in dem Satz “Da ist ein Mensch”, der die gleiche Akzentstruktur aufweist wie der italienische Ausgangssatz, allerdings mit dem Nachteil, daß er weniger idiomatisch ist. Hier müßten also mehrere potentielle Invarianten (Informationsstruktur, Idiomatizität) gegeneinander abgewogen werden (vgl. Schreiber 2002). Für das Sprachenpaar Deutsch-Englisch hat Mary Snell-Hornby darauf hingewiesen, daß es bei der Übersetzung ins Englische angebracht sein kann, die Satzgliedstellung zu ändern, um die Thema-Rhema-Gliederung zu erhalten, z.B. wenn im Deutschen ein umfangreiches rhematisches Satzglied am Satzbeginn steht, wie im folgenden Beispiel: (6) Die Schweiz […] ist eines der führenden Reiseländer der Erde. […] Ein dichtes, gut ausgebautes Netz von Bahnlinien und Straßen überzieht das Land. […] The country is served by a well-developed network of railways and roads. (Snell-Hornby 1998: 67) Für den Übersetzer ist in diesem Zusammenhang die Erkenntnis wichtig, daß die
89
90
Michael Schreiber
Beibehaltung der Informationsstruktur nicht mit der Beibehaltung der Wortstellung oder anderer Verfahren der Thema-Rhema-Gliederung verwechselt werden darf. Kontrastive Studien ohne Lösungsorientierung (wie die von Gülich 1982 oder Koch 1994) könnten ansonsten den Schluß nahelegen, die Informationsstruktur eines Textes sei grundsätzlich unübersetzbar. 4.2 Serialisierung (Wortstellung) Die Wortstellung (präzisier: Serialisierung, da nicht nur “Wörter” betroffen sind) kann einerseits als Teilbereich der Thema-Rhema-Gliederung betrachtet werden (wie unter 4.1 geschehen), andererseits als eigenständiger Untersuchungsgegenstand. Als solcher spielt sie innerhalb der Linguistik insbesondere in der Sprachtypologie eine Rolle, wo nach dem Kriterium der Satzgliedstellung u.a. SVO-Sprachen und SOV-Sprachen unterschieden werden. Daß diese Grobklassifikation nicht unproblematisch ist, zeigt sich bei der Einordnung des Deutschen, das sich aufgrund der Satzklammer und der unterschiedlichen Serialisierung im Haupt- und Nebensatz am adäquatesten als “Mischtyp” zwischen Verbzweit- und Verbletztsprache beschreiben läßt (Lang 1996: 12f.). Doch für Übersetzer und Dolmetscher sind solche Klassifikationsversuche keine große Hilfe. Ähnliches gilt für die Behandlung der Serialisierung innerhalb der generativen Grammatik: Hier geht es zwar z.T. um recht subtile Phänomene, oft auch aus kontrastiver Sicht (vgl. z.B. Abraham 1995), aber im Zentrum der Betrachtung stehen oft eher Formalisierungsprobleme als die eigentlichen Sprachdaten. Die spezifischen Anforderungen der STW zeigen sich im Bereich der Serialisierung besonders deutlich im Hinblick auf das Simultandolmetschen. Aufgabe der STW ist es aufzuzeigen, mit Hilfe welcher Strategien Dolmetscher mit den Serialisierungsunterschieden zwischen Ausgangs- und Zielsprache umgehen, und darzulegen, wie die entsprechenden Strategien didaktisch vermittelt werden können. Im folgenden möchte ich einige der wichtigsten wortstellungsbedingten Dolmetschstrategien nennen, die beim Dolmetschen aus dem Deutschen ins Englische und ins Französische auftreten (zu strategischen Prozessen beim Dolmetschen allgemein vgl. Kalina 1998; zu sprachspezifischen Strategien beim Simultandolmetschen vgl. besonders die entsprechenden Publikationen der Triester Dolmetschwissenschaftler, z.B. Snelling 1992, Fusco 1995 und Riccardi 1996). Zunächst sei die Antizipation genannt, d.h. die Vorwegnahme eines im Ausgangstext noch nicht verbalisierten Elements durch den Dolmetscher (zur Antizipation beim Simultandolmetschen vgl. Gile 1995: 176ff. und Zanetti 1999; zur Verbantizipation beim Dolmetschen aus dem Deutschen vgl. Jörg 1997). Wie bereits Wolfram Wilss am Beispiel des Sprachenpaars Deutsch-Englisch gezeigt hat, finden sich gelungene Antizipationen vor allem bei Routineformeln, z.B. Gruß- und Dankesformeln, wie im folgenden Beispiel (Redebeitrag aus dem Europäischen Parlament):
Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft
(7) Namens meiner Fraktion darf ich den beiden Herren On behalf of my political group Berichterstattern für die Arbeit, die sie aufgewendet haben, sehr I should like to thank the two Rapporteurs very herzlich danken. cordially for their work (Wilss 1978: 348) Die Antizipation des Verbs ist hier möglich aufgrund der Vertrautheit des Dolmetschers mit dem Ablauf einer Debatte im Europäischen Parlament (Wilss 1978: 348). Mit Gile 1995: 178 könnte man hier von einer extralinguistic anticipation sprechen. Ob eine Antizipation möglich ist, hängt aber auch von strukturellen Faktoren ab (vgl. den Begriff der linguistic anticipation bei Gile 1995: 176ff.). Laut einer empirischen Untersuchung zum deutsch-französischen Dolmetschen durch eine Heidelberger Diplomandin, Anja Krogh, ist es z.B. leichter, lexikalische Klammern zu antizipieren (z.B. bei Phraseologismen) als grammatikalische Klammern (Krogh 2000: 159ff.). Eine mit der Antizipation verwandte Strategie ist die Anteposition, d.h. die Vorziehung von frei beweglichen Konstituenten (speziell Umstandsbestimmungen) vor das Subjekt, wodurch ebenfalls eine Klammerstruktur im Zieltext vermieden wird. Wie Krogh gezeigt hat, wird diese Strategie beim Dolmetschen ins Französische häufig angewandt, da das Französische aufgrund der möglichen Mehrfachbesetzung des Vorfeldes zahlreiche Möglichkeiten der Anteposition bietet: (8) Er wurde im Alter von nicht einmal dreißig Jahren dank seiner unerschütterlichen Überzeugung, daß keine noch so gut gemeinte Hilfe wirksam sein kann, wenn ein Kind ohne Zuhause, ohne stabile soziale Rahmenbedingungen aufwachsen muß, zum Begründer der SOS-Kinderdorf-Idee… À l’âge de moins de tre- trente ans et grâce à sa conviction inébranlable quequ’aucune aide ne peut être superflue s- si un enfant vit dans la misère il est devenu le fondateur des villages d’enfants SOS… (Festansprache über Hermann Gmeiner, zit. nach Krogh 2000: 135) Wenn keine Antizipation oder Anteposition möglich ist und der zu verdolmetschende Satz nicht zu lang und unübersichtlich ist, greifen routinierte Dolmetscher z.T. auf die Strategie des (vergrößerten) décalage zurück, d.h. sie warten mit der Verdolmetschung, bis der zweite Teil der Satzklammer verbalisiert ist. Zur Vermeidung von Sprechpausen können verschiedene Hilfsstrategien eingesetzt werden; im folgenden deutsch-französischen Beispiel verwendet die Dolmetscherin ein reduziertes Sprechtempo sowie explikative Einfügungen (vgl. Setton 1999: 138ff. zum Englischen):
91
92
Michael Schreiber
(9) Nur dann werden wir als Europäer einen tatkräftigen Beitrag zur […] C’est la Bewältigung der neuen Herausforderungen leisten und gleichzeitig seule possibilité de relever le défi unsere amerikanischen Partner spürbar entlasten können. […] des nouvelles exigences. C’est aussi le seul moyen de venir en aide à nos partenaires américains. (Rede von Rudolf Scharping, zit. nach Krogh 2000: 116) Bei langen, unübersichtlichen Konstruktionen empfiehlt sich des weiteren die Technik der Satzaufspaltung, d.h. der Segmentierung in mehrere kurze Abschnitte — eine Technik, die zudem bereits von weniger routinierten Dolmetschern beherrscht wird (Ilg 1978: 88ff.; Kalina 1998: 193f.). Beim Dolmetschen aus dem Deutschen kommt diese Technik u.a. bei erweiterten Attributen zum Tragen, wie Nicole Kropat, eine weitere Heidelberger Diplomandin, anhand des Sprachenpaars Deutsch-Englisch gezeigt hat (zu erweiterten Attributen beim Simultandolmetschen aus dem Deutschen vgl. auch Dalitz 1983): (10) Diese von Professor Sieber und seinen Mitarbeitern so vorzüglich organisierte Tagung setzt eine lange Tradition… fort…. Professor Sieber and his colleagues have organised this meeting most extraordinarily well. It continues the long tradition of… (Rede von Th. Berchem, zit. nach Kropat 1994: 97) Aus den wenigen zitierten Beispielen dürfte deutlich geworden sein, wie sehr gerade die sprachenpaarbezogene Dolmetschwissenschaft über einen reinen Sprachvergleich hinausgehen muß. Allerdings gibt es in diesem Bereich immer noch einen besonderen Forschungsbedarf. So ist noch keineswegs befriedigend untersucht, wie stark die Möglichkeiten bestimmter Dolmetschstrategien in verschiedenen Zielsprachen variieren. 4.3 Modalisierung (Abtönung) Bei der Modalisierung oder Abtönung geht es um den Ausdruck der Stellung des Sprechers zum Gesagten. Hierzu werden im Deutschen, vor allem in der gesprochenen Sprache, häufig Abtönungspartikeln wie da, ja, denn oder eigentlich verwendet, für die es in vielen Sprachen keine direkten Entsprechungen gibt. Peter Blumenthal hat in seinem Buch zum deutsch-französischen Sprachvergleich versucht, für typische Verwendungsweisen deutscher Abtönungspartikeln französische Übersetzungen anzugeben. In einigen Fällen funktioniert dies recht gut, denn es findet sich eine vergleichbare Partikel im Französischen, wie in folgenden Übersetzungsvorschlägen von Blumenthal:
Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft
(11) Sei doch nicht so griesgrämig! Ne sois donc pas si morose! (12) Und ich erst! Et moi alors! (Blumenthal 1997: 94f.) In anderen Fällen ist es zwar möglich, die jeweiligen Redebedeutungen der deutschen Partikeln im Französischen explizit wiederzugeben, jedoch z.T. nur relativ umständlich: (13) Haben Sie auch nichts vergessen? Vous n’avez rien oublié, n’est-ce pas? (14) Wo habe ich bloß meine Brille? Où ai-je bien pu mettre mes lunettes? (15) Ich krieg dich schon! N’aie pas peur, je t’aurai. (Blumenthal 1997: 94f.) Da eine Häufung von Formulierungen wie in den Beispielen 13–15 in einem französischen Zieltext unidiomatisch wirken würde, bleiben in der Praxis viele Abtönungspartikeln ohne materielle Entsprechung (vgl. bereits Weydt 1969: 72). Blumenthal geht es im Rahmen seines sprachvergleichenden Ansatzes jedoch weniger um Übersetzungsverfahren als um Entsprechungen auf langue-Ebene, weswegen er die “Null-Entsprechung” zwar erwähnt, aber nicht näher thematisiert. Für Übersetzer wäre es aber gerade von Nutzen, nicht nur potentielle explizite Entsprechungen zu kennen, sondern auch über die Verwendungsbedingungen der einzelnen Entsprechungstypen (einschließlich der “Null-Entsprechung”) näheres zu wissen. Aus übersetzungswissenschaftlicher Sicht beschreibt Cornelia Feyrer die Null-Entsprechung daher als “legitime Variante” direkter Entsprechungen, wobei sie zwischen “echten” und “vermeintlichen” Null-Entsprechungen differenziert: “Wo keine direkte Entsprechung zu finden ist, kann sich sehr wohl auf Textebene oder auf pragmatisch-situativer Ebene eine finden” (Feyrer 1998: 279). Die Problematik der Null-Entsprechung behandelt René Métrich in einem Aufsatz zur “NichtÜbersetzung” von Modalpartikeln und anderen Gesprächswörtern im Sprachenpaar Deutsch-Französisch. Métrichs empirische Untersuchung von Übersetzungen beider Richtungen kommt zu dem Ergebnis, daß l’équivalence ‘zéro’ est presque systématiquement plus fréquente lorsque l’allemand est langue cible que lorsqu’il est langue source. Cela confirme l’idée d’une pression de la langue allemande en faveur de l’emploi de mdc [= mots de la communication] et fait apparaître que loin de négliger la traduction de ces mots les traducteurs français ont au contraire globalement plutôt tendance à les surtraduire. (Métrich 1997: 151)
93
94
Michael Schreiber
Untersuchungen zu anderen Sprachenpaaren kommen zu ähnlichen Ergebnissen. Auch für das Spanische, wo z.B. Partikeln wie pero, pues und ya und andere Ausdrucksformen (z.B. bestimmte Tempora und Modi) als Teiläquivalente deutscher Abtönungspartikeln fungieren können (Prüfer 1995: 256f.), gilt nach Ansicht von Christiane Beerbom: Würde man beim Übersetzen eines Textes jede einzelne deutsche MP [= Modalpartikel] im Spanischen wiedergeben, so würde dies einen Verstoß gegen die Norm des Spanischen darstellen und hätte zur Folge, daß der spanische Text völlig überladen und unidiomatisch wirken würde. […] In vielen Fällen ist also die Nullentsprechung die einzig adäquate Lösung. (Beerbom 1992: 461)
Als Beispiel für einen Verwendungstyp, der im Spanischen in der Regel ohne Entsprechung bleibt, nennt Beerbom den Gebrauch der Partikel schon in direktiven, formal deklarativischen Äußerungen, wie im folgenden Beispiel aus der Übersetzung eines Jugendbuchs: (16) [Momo kann mit der Puppe, die der graue Herr ihr geschenkt hat, wenig anfangen] “Mit einer so fabelhaften Puppe kann man nicht spielen wie mit irgendeiner anderen, das ist doch klar. Dazu ist sie auch nicht da. Man muß ihr schon etwas bieten, wenn man sich mit ihr nicht langweilen will.” “… Hay que ofrecerle algo, si uno no quiere aburrirse con ella.” (Michael Ende: Momo, zit. nach Beerbom 1992: 235) Für das Sprachenpaar Deutsch-Italienisch hat Renata Buzzo Margari gezeigt, daß zwar auch im Italienischen Abtönungen ausgedrückt werden können (z.B. durch Partikeln wie ma oder morphologische Mittel wie das Futur), daß dies aber wiederum seltener geschieht als im Deutschen, weswegen bei vielen Übersetzungen ins Deutsche aus Gründen der Idiomatizität Abtönungspartikeln hinzugefügt werden, wie im folgenden Beispiel aus einer literarischen Übersetzung: (17) Invece ci trovò un pesciolino lungo un mignolo, lo afferrò con rabbia e stava per ributtarlo in mare quando udì una vocina sottile che diceva: Ahi, non mi stringere così forte! Aber nur ein Fischchen lag darin, kaum so groß wie ein kleiner Finger. Der Fischer packte es voller Wut und wollte es eben ins Meer zurückschleudern, als ein feines Stimmchen sagte: Au, au, drück mich doch nicht so fest! (Gianni Rodari: Favole al telefono, zit. nach Buzzo Margari 1997: 155) Die bisher zitierten Beispiele gehörten der gesprochenen Sprache an. Wie Karin Königs für die Übersetzung aus dem Englischen gezeigt hat, können jedoch in bestimmten Fällen auch in schriftsprachlichen Textsorten Abtönungspartikeln im
Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft
Deutschen eingefügt werden, wie die Partikel etwa im folgenden Beispiel aus einem linguistischen Fachtext: (18) The colloquial preterite is a genuine preterite, not a perfect with the auxiliary slurred or omitted. Das colloquial preterite ist ein echtes Präteritum, nicht etwa ein Perfekt, bei dem das Hilfsverb verschluckt oder weggelassen worden ist. (Königs 2000: 569) Solche Beobachtungen können “Mut zur Hinzufügung” für Übersetzungen ins Deutsche und “Mut zur Lücke” für Übersetzungen aus dem Deutschen machen. Für den Übersetzer (oder Dolmetscher) sind die Verwendungsbedingungen der “Null-Entsprechung” daher genauso wichtig wie die potentiellen Möglichkeiten der expliziten Entsprechung deutscher Abtönungspartikeln. Die Entscheidung, in welchem konkreten Einzelfall die Null-Entsprechung die adäquate Lösung ist und in welchem Fall eine explizite Entsprechung vorzuziehen ist, kann die STW dem Übersetzer natürlich nicht abnehmen — aber sie kann ihm Entscheidungshilfen anbieten, besonders wenn sie über eine statische Auflistung potentieller Äquivalente hinausgeht und die Verwendungsbedingungen der einzelnen Lösungsmöglichkeiten stärker in den Blick nimmt.
5.
Schlußbemerkungen
Abschließend möchte ich folgendes hervorheben: 1. Meine Ausführungen sind keineswegs als globale Kritik an der KL zu verstehen. Da die KL im engeren Sinne (als reiner Sprachvergleich) jedoch andere Ziele verfolgt als die STW, ist sie für Übersetzer und Dolmetscher weniger nützlich als eine speziell auf Translationsprobleme zugeschnittene STW. 2. Die Unterscheidung zwischen KL und STW entspricht nicht einer scharfen Trennung: Viele Arbeiten bewegen sich im Grenzbereich beider Disziplinen. Für Übersetzer und Dolmetscher sind sie umso relevanter, je stärker sie lösungsorientiert im oben diskutierten Sinne ausgerichtet sind. 3. Problematisch an der STW, wie sie bisher betrieben wurde, ist allerdings, daß sie sich auf funktionskonstante Übersetzungen beschränkt (Pruncˇ 2001: 103). Ein Desideratum besteht noch in der deskriptiven Untersuchung interlingualer Bearbeitungen, z.B. Zusammenfassungen (Schreiber 1993: 301ff.). Das Schlußwort meines Plädoyers gegen die Vernachlässigung der sprachenpaarbedingten Probleme des Übersetzens und Dolmetschens möchte ich dem französischen Sprach- und Übersetzungswissenschaftler Maurice Pergnier überlassen:
95
96
Michael Schreiber
Quand j’entends dire que la traduction n’a rien à voir avec la linguistique, je suis amené à comprendre qu’elle n’a rien à voir avec le langage. Et j’en viens à me demander si, quand deux langues sont en présence, les humains se mettent à communiquer par télépathie. (Pergnier 1998: XV)
References Abraham, Werner. 1995. Deutsche Syntax im Sprachenvergleich. Tübingen: Narr. Albrecht, Jörn. 1987. “Wissenschaftstheoretischer Status und praktischer Nutzen der Übersetzungswissenschaft”. In Übersetzen im Fremdsprachenunterricht, Rolf Ehnert und Walter Schleyer (eds), 9–23. Regensburg: DAAD. Bastian, Sabine. 1988. “‘Mise en relief ’ und ‘phrase segmentée’ im gesprochenen Französisch und ihre Wiedergabe beim Dolmetschen”. In Leipziger romanistische Beiträge, Klaus Bochmann et al. (eds), 115–125. Leipzig: Universität. Beerbom, Christiane. 1992. Modalpartikeln als Übersetzungsproblem. Frankfurt: Lang. Blumenthal, Peter. 1997. Sprachvergleich Deutsch-Französisch. Tübingen: Niemeyer. Buzzo Margari, Renata. 1997. “Considerazioni sulle particelle modali tedesche e sulle corrispondenti espressioni italiane”. In Italiano e tedesco: un confronto, Sandra Bosco Coletsos (ed), 139–171. Alessandria: Edizioni dell’Orso. Coseriu, Eugenio. 1987. “Über Leistung und Grenzen der kontrastiven Grammatik.” In idem. Formen und Funktionen, 67–84. Tübingen: Niemeyer. Dalitz, Günter. 1983. “Deutsche erweiterte Attribute beim Simultandolmetschen”. Fremdsprachen 27: 157–162. Feyrer, Cornelia. 1998. Modalität im Kontrast. Frankfurt: Lang. Fisiak, Jacek. 1990. “On the present status of some metatheoretical and theoretical issues in Contrastive Linguistics”. In Further Insights into Contrastive Analysis, Jacek Fisiak (ed), 3–22. Amsterdam und Philadelphia: Benjamins. Fusco, Maria Antonietta. 1995. “On teaching conference interpretation between cognate languages”. The Interpreters’ Newsletter 6: 93–109. Gile, Daniel. 1995. Basic Concepts and Models for Interpreter and Translator Training. Amsterdam und Philadelphia: Benjamins. Gülich, Elisabeth. 1982. “La ‘phrase segmentée’ en français et en allemand: une technique particulière à la communication orale”. In Didactique des langues étrangères: français — allemand, 33–66. Lyon: Presses Universitaires de Lyon. Henschelmann, Käthe. 1999. Problembewußtes Übersetzen: Französisch-Deutsch. Tübingen: Narr. Ilg, Gérard. 1978. “L’apprentissage de l’interprétation simultanée. De l’allemand vers le français”. Parallèles 1: 69–99. Jäger, Gert. 1975. Translation und Translationslinguistik. Halle: Niemeyer. Järventausta, Marja. 1996. “Wieviel kontrastive Grammatik braucht der Übersetzer?”. In Übersetzerische Kompetenz, Andreas F. Kelletat (ed), 91–104. Frankfurt: Lang. Jörg, Udo. 1997. “Bridging the gap: Verb anticipation in German-English simultaneous interpreting”. In Translation as Intercultural Communication. Selected Papers from the EST Congress, Prague 1995, Mary Snell-Hornby, Zuzana Jettmarová und Klaus Kaindl (eds), 217–228. Amsterdam und Philadelphia: Benjamins. Kalina, Sylvia. 1998. Strategische Prozesse beim Dolmetschen. Tübingen: Narr. Kielhöfer, Bernd. 1980. Fehlerlinguistik des Fremdsprachenerwerbs. Kronberg: Scriptor.
Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft
Koch, Peter. 1994. “Valenz und Informationsstruktur im Sprachvergleich Italienisch-Deutsch”. Italienisch 32: 38–58. Königs, Karin. 2000. Übersetzen Englisch-Deutsch. München: Oldenbourg. Koller, Werner. 1992. Einführung in die Übersetzungswissenschaft. Heidelberg: Quelle & Meyer. Korzen, Iørn und Marello, Carla (eds). 2000. Argomenti per una linguistica della traduzione. Alessandria: Edizioni dell’Orso. Krogh, Anja. 2000. Warten auf das Verb. Empirische Untersuchung über Verbklammern als Problem beim Simultandolmetschen am Beispiel des Sprachenpaars Deutsch-Französisch. Universität Heidelberg: Institut für Übersetzen und Dolmetschen (Diplomarbeit). Kropat, Nicole. 1994. Die Satzaufspaltung als Strategie beim Simultandolmetschen am Beispiel des Sprachenpaars Deutsch-Englisch. Universität Heidelberg: Institut für Übersetzen und Dolmetschen (Diplomarbeit). Lado, Robert. 1957. Linguistics across Cultures. Ann Arbor: University of Michigan Press. Lang, Ewald. 1996. “Das Deutsche im typologischen Spektrum”. In Deutsch — typologisch, Ewald Lang und Gisela Zifonun (eds), 7–15. Berlin und New York: De Gruyter. Lundquist, Lita. 2000. “Translating associative anaphors”. In Argomenti per una linguistica della traduzione. Iørn Korzen und Carla Marello (eds), 111–129. Alessandria: Edizioni dell’Orso. Malblanc, Alfred. 1968. Stylistique comparée du français et de l’allemand. Paris: Didier. Métrich, René. 1997. “De la non traduction des ‘mots de la communication’ de l’allemand en français”. Zeitschrift für französische Sprache und Literatur 107: 143–172. Nord, Christiane. 2001. Lernziel: Professionelles Übersetzen Spanisch-Deutsch. Wilhelmsfeld: Egert. Pergnier, Maurice. 1998. “Préface”. In Enseignement de la traduction et traduction dans l’enseignement, Jean Delisle und Hannelore Lee-Jahnke (eds), IX-XVI. Ottawa: Presses de l’Université d’Ottawa. Prüfer, Irene. 1995. La traducción de las partículas modales del alemán al español y al inglés. Frankfurt: Lang. Pruncˇ, Erich. 2001. Einführung in die Translationswissenschaft. Bd. 1: Orientierungsrahmen. Universität Graz: Institut für Translationswissenschaft. Riccardi, Alessandra. 1996. “Language-specific strategies in simultaneous interpreting”. In Teaching Translation and Interpreting 3, Cay Dollerup und Vibeke Appel (eds), 213–222. Amsterdam und Philadelphia: Benjamins. Schreiber, Michael. 1993. Übersetzung und Bearbeitung. Tübingen: Narr. Schreiber, Michael. 1999. Textgrammatik — Gesprochene Sprache — Sprachvergleich. Frankfurt: Lang. Schreiber, Michael. 2002. “Thema-Rhema-Gliederung im Italienischen und Deutschen”. In Translation zwischen Theorie und Praxis. Lew Zybatow (ed), 421–436. Frankfurt: Lang. Setton, Robin. 1999. Simultaneous Interpreting. Amsterdam und Philadelphia: Benjamins. Snell-Hornby, Mary. 1998. “Kontrastive Linguistik”. In Handbuch Translation, Mary SnellHornby et al. (eds), 66–70. Tübingen: Stauffenburg. Snelling, David. 1992. Strategies for Simultaneous Interpreting. From Romance Languages into English. Udine: Campanotto. Steiner, Erich. 1999. “Linguistik und Translationswissenschaft — (getrennte) Disziplinen?”. In Modelle der Translation, Alberto Gil et al. (eds), 477–506. Frankfurt: Lang. Teich, Elke. 1999. “Contrastive Linguistics and Translation Studies Revisited”, In Modelle der Translation, Alberto Gil et al. (eds), 507–521. Frankfurt: Lang. Thome, Gisela. 1976. “‘La mise en relief ’ und ihre Wiedergabe im Deutschen”. In Proceedings of the Fourth International Congress of Applied Linguistics. Bd. 2, 399–411. Stuttgart: HochschulVerlag. Wandruszka, Mario. 1969. Sprachen — vergleichbar und unvergleichlich. München: Piper. Weydt, Harald. 1969. Abtönungspartikel. Bad Homburg: Gehlen.
97
98
Michael Schreiber
Wilss, Wolfram. 1978. “Syntactic anticipation in German-English simultaneous interpreting”. In Language Interpretation and Communication, David Gerver und H. Wallace Sinaiko (eds), 343–352. New York und London: Plenum. Zanetti, Roberta. 1999. “Relevance of anticipation and possible strategies in the simultaneous interpretation from English into Italian”. The Interpreters’ Newsletter 9: 79–98.
Translating non-segmental features of textual communication The case of metaphor within a binary-branch analysis* Patrick Zabalbeascoa Universitat Pompeu Fabra
This paper presents a tree diagram for analysing metaphor translation, following a speculative approach (Toury 1995: 15) within translation theory. Other translational problems may also benefit from this proposal, especially non-segmental features of texts. Each translation problem should be analysed in its own terms as a way of gaining insight into how problems are solved. The structure of the tree is a type-within-type binary branch analysis of all of the solutions that actually exist and other possible solutions. Thus, we may understand a translation’s rationale more fully by having a structured framework on which to pin real solutions along with what might have been. The concept of metaphor I use combines Richards 1965, Lakoff and Johnson 1980, and Goatly 1997. For the translation of metaphor, Toury 1995 and Newmark 1988 are useful starting points, but it is important to consider metaphors as features of texts and not always as segments. Examples are mostly animal metaphors plus a few from a short passage of The Bonfire of the Vanities.
1.
Introduction
This paper presents a proposal for analysing metaphor translation, or more precisely, metaphor as a problem that may be solved in a number of ways in translation. The various points of departure for the proposal are challenged to some extent so that new claims can be made about metaphor translation, theoretical constructs of metaphor translation, and ultimately, theoretical models for analogous situations. Following Toury’s (1995: 14–19) suggestion that fruitful proposals can come out of a speculative approach, I will not be looking at the issue from an entirely descriptive or entirely prescriptive point of view. The solutions proposed
100 Patrick Zabalbeascoa
and the relationships between types are not born out of the study of a massive corpus of texts nor do they express a preference for one way of translating overanother. In speculating, one looks into many possible solutions, and this, in turn, contributes to a greater understanding of actual solutions found in descriptive studies and of the ideal solutions suggested by prescriptive models. The hope is that other translational problems will benefit from this proposal for metaphors, especially non-segmental features of texts. Metaphor is a feature of both literary and nonliterary language. Newmark (1988: 84–96) tells us that original metaphors are “born” and “live” in literature and “die” once they move into everyday speech. But for Lakoff and Johnson (1980: Chapter 1) metaphor is a trait of natural language. Following the latter, we might say that metaphor is picked up in literature, experimented with and taken to new heights of creativity and expressive power. Furthermore, metaphorical expressions can also be entirely non-verbal or supported by accompanying non-verbal signs (Hammond and Hughes 1978; Zabalbeascoa 1997). For many scholars (e.g. Richards 1965 and Leech 1969) metaphor involves two “entities”, material or otherwise, and one or more common characteristics which the two entities (are seen to) share. One entity is what is ultimately and unconventionally being referred to and the other is the conventional meaning or referent of an expression that points to the former in order to create the metaphorical effect. Here I will adopt Goatly’s (1997: 9) terminology, Topic (the actual unconventional referent); Vehicle (the conventional referent), and Ground (the similarities and/or analogies involved). A metaphor requires all three components, not just the expression (the Vehicle-term) that merely signals its presence. Metaphors are more than purely ornamental literary artifacts. They can fulfil various functions and textual metafunctions (term borrowed from Halliday and Hasan 1976). For instance, metaphors can help in constructing cohesion and coherence (textual function), provide greater explanatory power (ideational function), and create a certain tenor (interpersonal function). Indeed, they have been grouped according to many typologies (e.g. Goatly 1997, Gozzi Jr. 1999, Grady 1997, Lakoff and Johnson 1980, Newmark 1988). Typologies may vary depending on which one of the three components of the triangular relationship is used as the main parameter for classification; e.g. semantic field of Tenor, Vehicle or Ground, intertextual relationships, discourse membership, function, lexical make-up, structure and metaphorical networking. A translational model should be adaptable to such a range of classifications. The diversity of typologies may be seen as a hindrance, or may simply respond to the need to highlight different kinds of relationships among metaphors, depending on the occasion. Because this may in fact be part of the dynamics of translators’ behaviour it would seem contradictory to try to impose a definitive classification. There is probably a difference between categorizing metaphors (i) for the purpose of understanding or explaining what a metaphor is and how it works, and (ii) for
Translating non-segmental features of textual communication
establishing relationships between a source text (ST) and its target text (TT). In his vision of Translation Studies, Holmes (1988: 74) regarded a theoretical account of metaphor translation as a partial theory, within the set of partial theories restricted to different problems (other partial theories being restricted by area, mode, rank, time and text-type). Although a model accounting for the translation of metaphor-as-problem does not make up a theory of translation it may be a partial contribution that will have to be complemented with models accounting for other problems (e.g. the translation of humour) and other partial theories. In picking up Holmes’ claim I will have to challenge Toury’s (1995: 81–84) analysis of the metaphor as a unit, although there is also much to be gained from his study of metaphor translation as a theoretical problem, especially linking it to binary and speculative approaches. Indeed, the present study is aimed at evaluating the usefulness and possible characteristics of binary branching.
2.
The challenges behind Toury’s claims about metaphor translation
From Toury (1995: 82–83) I quote the following ‘replacing + replaced segments’: (i) metaphor into ‘same’ metaphor; (ii) metaphor into ‘different’ metaphor; (iii) metaphor into non-metaphor; (iv) metaphor into 0 (i.e. complete omission, leaving no trace in the target text); (v) non-metaphor into metaphor; (vi) 0 into metaphor (i.e. addition, pure and simple, with no linguistic motivation in the source text). Each one of Toury’s first four possibilities shows a corresponding segment in the translation for metaphor-segments of the source text, including the possibility of 0 (no segment). His last two options, however, regard metaphor exclusively as a segment of the TT when it is related to a non-metaphorical segment of the ST. Options where metaphor is not the starting point are discarded in this study so as to be consistent with a metaphor-as-problem approach, unlike one which involves studying all the problems for which metaphor is the TT solution or part of it. Here, I am dealing exclusively with finding out three things: (a) what possibilities there are for translating source-text metaphors; (b) what kinds of relationships and classifications might be established for the solutions to the metaphor-translationproblem, along with the (methodological or theoretical) advantages of certain typologies over others; (c) ultimately, what conditions will favour a given type of solution in the everyday practice of translation. Slightly adapting the fifth item on Toury’s list I would contend that it would be a perfectly legitimate research project in its own right to find out how effective and frequent metaphor is as a solution in translated texts and what kinds of problems it solves other than the translation of metaphors. However, a project of this sort would not set out to discover how metaphors are dealt with in translation. For all options involving metaphor as part of the TT solution (for ST metaphors and ST non-metaphors alike), the research
101
102 Patrick Zabalbeascoa
aim could be to discover the norms that regulate the presence of metaphors in translation, either universally or restricted by language pairs or some other parameter. The particular case of 0 into metaphor could be looked into from the wider context of “addition” as a translational strategy, i.e. 0 into metaphor as an instance of all TT segments that appear to have no ST counterpart. But if we consider that metaphors do not come out of the blue it might be more productive to consider looking for some kind of ST item or feature on which to base the analysis. Possibly the biggest problem posed by a metaphor-as-unit approach is that of finding other units in the text that can be “added” to the metaphor-units to give us the total amount of the text. Moreover, authors like White (1996: 58) point out that even counting the metaphors of a text is no straightforward matter. You cannot compare metaphors to linguistic units (phoneme, morpheme, lexeme, clause, sentence) because they are of a different order. The same would be the case for textual units (e.g. title, introduction, conclusion). Likewise, I contend that metaphor and segment 0 belong to different paradigms, thus challenging Toury’s unitbased reasoning that leads him to propose metaphor into 0 and 0 into metaphor as possible solution-types with the unstated entailment that 0 is either a non-metaphor or a third category. Metaphor into 0 and 0 into metaphor are, of course, possible but they tell us nothing about the metaphorical potential of 0 since segment 0 has not yet been proven to be a non-metaphor in all cases. Metaphorical value might be assigned to a pause, a silent moment, an ellipsis, a blank space, a reluctance or inability to respond, etc. However, in terms of units, it does make perfect sense to speak of 0 to mean the absence of a unit. To delete a metaphor in the translating process is not necessarily equated with deleting the segment where it appears, but with destroying by whatever means available the triangular relationship between Topic, Vehicle and Ground. In some films, for example, there is an apparent linguistic void or ellipsis, which is actually filled by a non-verbal representation or image, a metaphor which may even be of linguistic origin. The example I have in mind is a scene in Pulp Fiction where a female character says “Don’t be [square]” to her male escort to get him to change his old habits and try something new, only she does not actually utter the last word but instead draws a square in the air, reinforced by a special effect that makes a dotted line appear on the screen. Another example of the problems of apparent “nothingness”, this time for rhyme not metaphor, can be seen in the film Lion King, or heard, rather, when the character called Timon sings (referring to his buddy Pumbaa, who has problems retaining gas) “He became downhearted / every time that he [farted]” the last word is omitted but clearly implied by a number of contextual markers, with the additional remark “Not in front of the kids” to Pumbaa who is about to say the taboo word. It does not seem very convenient to go into a long discussion as to whether these two examples actually constitute instances of “nothingness” or are an implied presence of some sort. The category of segment
Translating non-segmental features of textual communication 103
0 should not enter the discussion at an early abstract stage of binary branching although it might appear more meaningfully later on (i.e. beyond type-within-type, as shown in Figure 2). In a binary-branching approach, a given segment 0 solution might be an instance of metaphor or non-metaphor, or, rhyme or non-rhyme, (or, as I shall argue below, whatever the source-text problem may be). To retain the same ST metaphor in the TT may not inevitably entail interlinguistic rendering (see discussion on the L2/not L2 dichotomy in Section 3). By “same metaphor” I am not referring to mathematical or objective sameness but to subjective sameness, i.e. what is the same according to the translator or researcher. In fact, it seems counterproductive to take for granted that all solutions that are not L2 renderings, such as L1 borrowings, or the use of L3 (presence of items from any number of different languages) and non-verbal textual elements cannot express the same metaphor. Likewise, literal translation (interlinguistic lexical equivalence) does not entail that the same metaphor will be retained in all cases, since it can only guarantee that one part of the metaphorical relationship is kept (the Vehicle), and even that guarantee is questionable. For example, literal translation does not guarantee that the same metaphor is rendered if two cultures use a given word or expression with different metaphorical (idiomatic) meanings even though they may be equivalents for their non-metaphorical meanings. Let us look at the following animal metaphors to illustrate this point. There is a clear difference in the metaphorical image of a dragon depending on whether we are using it within a Christian tradition or a Chinese one. In English, a person might be referred to as a fox or as being foxy, but their Grounds are rather different from the idiomatic metaphors zorro and zorra, the Spanish literal equivalents for fox and foxy. Zorro and zorra convey cunning and whorish qualities, respectively, whereas fox would more frequently refer to being sly and foxy to sexy (especially US) or crafty. The Spanish zorro metaphor is not usually as negative as fox; this might be explained by such cultural differences as foxhunting practices and fables. However, Spanish zorra is much more negative than foxy and cannot be used to refer to craftiness. This might be explained by certain linguistic phenomena, such as the existence in Spanish of many masculine/feminine pairs where the masculine morpheme has a positive connotation and the feminine one has a demeaning connotation,1 or as the case may be, a negative metaphorical extension. In English the (sexy) foxy metaphor refers mostly to women, and there is the word vixen for the female fox, which also happens to be a literal equivalent of zorra, but rarely its metaphorical equivalent. Another example from English and Spanish is taken from metaphorical idioms or similes that are used to describe a person as being busy and hardworking; in such a case, one would expect bee and hormig(uit)a (ant or little ant), respectively. From the translator’s point of view, it is important to note in this particular example that the choice of bee is motivated not only by the Ground of the metaphor, but also because of other factors, like alliteration, as a popular cohesive
104 Patrick Zabalbeascoa
device, sound symbolism associated with bees buzzing, and possibly a certain bias towards monosyllables for such idioms. When these additional factors are not present there is often a greater coincidence between Spanish and English. A further example can be found in the passage from The Bonfire of the Vanities, Section 4. Claw as it is used in the English text is meant to conjure up an image of a bloodthirsty predator or vulture, whereas the Spanish version uses a translation of claw (zarpas) which overlaps somewhat with paw; this, added to the fact that zarpas is a noun, makes the most plausible interpretation of the translated metaphor an image of a clumsy bear fumbling with the newspaper while trying to open it. Zarpas is quite a common Spanish metaphor for clumsiness in handling things. Toury also considers the possibility that a metaphor may not constitute a whole unit, and be a part of a unit, either [M+X] or [X+M]. However, if I do not consider each metaphor as a unit or segment, I can build on Toury’s insight to make the following proposal. Let us consider a metaphor M and whatever it might be coupled to (X) not as units but as features. This gives us the means to consider that a textual metaphor will either have no accompanying feature, or will be coupled to some other feature(s) or component(s), regardless of where they appear in the text. So, for metaphor as problem, or solution, we might wish to discriminate (in a binary division) between a “bare” metaphor (M=M0) and a “coupled” metaphor (M=M0+X), but both within the broader category of metaphor. The conclusion that I have reached at this point is that in establishing binary divisions we must be very careful not to mix up different planes of reference. If metaphor does not entail verbal expression, if metaphor does not entail the impossibility of ellipsis or empty-unit, if metaphor does not entail a given language to the exclusion of all others, then we will have to discard binary divisions such as metaphor versus nothing, metaphor versus picture, and metaphor versus borrowing. First, we need to know what the solutions mean in metaphorical terms if our goal is to understand what happens to source-text metaphors in translation. Second, we cannot impose as the most useful typology of metaphors one that will always be the same and will always involve, for instance, a consideration as to whether some of the solutions are verbal and others are non-verbal.
3.
Binary-branch mapping and type-within-type categorization
Considering metaphor as a linguistic property, a text function, a literary/rhetorical device, or a means of conceptualization, I outline criteria for mapping all possible solutions for the translation of a given metaphor. First of all, there is the subjectivity of identifying metaphors (or treating certain expressions as being metaphorical). This subjectivity affects both the translator and the scholar as analyst or critic. Moreover, the scholar will have the additional task of distinguishing between those metaphorical
Translating non-segmental features of textual communication 105
interpretations which are shared with the translator and those which are not. Second, the aim is to understand how metaphors are translated on metaphorical grounds, first and foremost, i.e. what does a given solution entail in terms of metaphoricalness, metaphor type and any metaphor-related factors? To propose a model of analysis with this in mind forces us to discard metaphor into 0 in an abstract tree-diagram, not because it cannot happen but because first we need to know what 0 means on a metaphorical plane. If it means an omission of a segment, a physical nothing, then it seems rash to discard the possibility that segment 0 might turn out to be meaningful (i.e. something), metaphorically or otherwise. Moreover, in the event that binary-branching proved its usefulness beyond the case of metaphor analysis, the formal and functional implications of a 0 branch (option) would become more and more unpredictable. For Figures 1 and 2, the empty segment could a priori appear anywhere, i.e. a possible value for S0. Third, the present proposal is an account of the set of possible target-text solution-types (S-set) for a given feature or item of the source text which is perceived as posing a translation problem (P). Solutions are distributed according to no other criterion than how they relate to the ST feature as defined in the problem. So, when P is metaphor, we must look at the solutions in terms of metaphor membership and type, or, on a more general plane, categories of which metaphor could be regarded as a member (e.g. figures of speech, see Section 4, solution-type [4] in examples 1 and 2). In our tree analysis, the first binary split involves opting for either [1] “the same” metaphor or “no”, i.e. something else, and gradually widens its scope to cover ever more general categories. a. S-set for field metaphor [1] campo
[1]
NO
[2] sembrado
S-set for P
b.
NO [2]
NO
NO n binary splits
[3] mar
NO
[4] inacabables
[5] todos aquellos
[n]
NO=[n+1]
Figure 1.Binary tree diagrams for type-within-type set of solutions in translation.
This is illustrated in Figure 1b where [1]+[2]+(…)+[n+1] covers the set of all theoretically possible solutions and each split or node is determined by membership or non-membership to a certain type, which, in turn, is a subcategory of the type which defines the next split as we go down the tree. The actual number and definition of types will depend on the typology chosen, the only limitation being
106 Patrick Zabalbeascoa
that as the numbers rise the categories they represent get broader. Let us look at the case of P(metaphor) and a five-node tree for an S-set like the one in Figure 1a, which illustrates example 1a–e. The different forms of “the same metaphor” are all in [1]; all metaphors “of the same type” as [1] are included in [1] and [2]; [1], [2] and [3] cover all metaphors; [1], [2], [3] and [4] account for all rhetorical devices, if I define [4] as “all rhetorical devices except metaphor”; [1], [2], [3], [4] and [5] include all types of expressions. There are instances of all five solution-types in examples 1 and 2 in Section 4. A cline defined by [1] and [n] at opposite ends might be useful as a measure of “equivalence” since each category has a weaker condition for type-membership as one moves towards [n]. We must not forget, however, that a greater or lesser degree of equivalence according to one typology or tree analysis says nothing about similarity according to other criteria, which would require a separate tree (see simple vs. complex analyses below). It now becomes apparent that binarism presents each category or type in opposition to all others, e.g. category [3] is the complementary of [1], [2], [4], and [5] put together. This type of analysis probably has a number of limitations but it might be productive for certain case studies or theoretical models, so it is worth looking into. The following dichotomies can be used as binary splits that may occur for any of the solution-types, i.e. they would branch off from [1], [2]… [n+1], see B in Figure 2. Verbal/non-verbal. Not all solutions will necessarily be purely verbal. In this respect “non-verbal” (Figure 2) does not only imply “purely non-verbal” but a broader category of “not only verbal”. L2/not L2. L2, or TTL, is often defined as the main language of the TT, but because texts can include various languages as well as a range of dialects and sociolects, not to mention nonlinguistic code systems, L2 would probably be more accurately defined as the signs that are meaningful to a given prototypical TT readership. Likewise for L1 in relation to ST. If L2 is the language of the translation, then “not L2” includes both borrowings from L1 and any third-language solutions. If L2 is further restricted to cover only a certain variety (e.g. official or standard) then “not L2” will also have to include all other varieties. L2 could split again to express the literal translation/non-literal translation dichotomy. Bare/not bare. Not-bare, compound solutions would include some instances of what Newmark (1988: 32) calls couplets (or triplets), the combined effort of two (or three) translation strategies. So, a compound metaphor solution is M=[M0+X] and +X stands for anything before, within and/or after M0 and is understood to be part of the rendering of a source-text problem. Here are three examples of what X might be for M0(rose): (a) some sort of clarifying premodifier, a beautiful rose; (b) a clause or phrase, a rose standing alone; (c) a footnote; (d) some additional trait, such as rhyme, e.g. rose rhyming with grows, especially if rhyme is used to signal the presence of the metaphor. This could apply to other textual elements. Solution S,
Translating non-segmental features of textual communication 107
which may appear bare (S=S0) or compound (S=S0+X), could be an idiom, a rhetorical question, a noun clause, a line of verse, a proper noun, a culture-specific word, etc. Segmental/non-segmental. Non-segmental solutions include certain types of compensation that involve the use of intonation patterns or typographical conventions or otherwise adding features onto already-existing segments (e.g. irony, understatement, intertextuality). Something/nothing. Here it is essential to clarify whether we are using these terms as units (thus nothing means empty segment) or features (whereby nothing entails the absence of any features). S-set for P [1] B
B
verbal L2 S0
non-verbal not L2
S0+X S0
S0
S0+X
S0+X
NO [2] B
NO n binary splits
[n]
NO=[n+1]
B
B
Figure 2.An example of binary branching B beyond type-within-type.
This model of metaphor-translation analysis could be adapted to many other problems. The proposal is that each feature or element be analysed on a single plane of reference, the one that characterises it as a translation problem. If I wished to analyse many instances of P(joke), then each solution could be analysed as belonging to one of the following categories: [1] is the same joke; [1] and [2] are the same type of joke; [1], [2], and [3] are any joke; [4] something that is not a joke but belongs to the same category according to some more general parameter (e.g. phaticism or attention-getter); [5] all solutions that have nothing to do with jokes or any broader category that jokes might belong to. The same could be done for translations of examples, insults, rhymes, acrostics, parodies, dialectal traits, etc. as well as more traditional linguistic, literary and semiotic units, e.g. noun, sonnet, cultural element. The resulting number of categories/types is meant to be variable (n), depending mostly on the number of subtypes within the same type of joke. In each case, it is up to the researcher, the critic or the translator to decide on the most convenient number of categories and their definitions. Simple analyses would involve a single binary branch structure. Complex analyses would involve applying more than one tree to a given source-text element, when it
108 Patrick Zabalbeascoa
is interpreted as posing two or more significant problems. For example, I might choose to analyse dragon as a P(metaphor), P(insult), P(idiom), P(lexical unit), etc., or, as we have seen in the previous section, claw can either be analysed as P(metaphor) or P(verb). Of course, for certain analyses, a subtype of metaphor could be a grammatical category such as verb (and vice versa). This is illustrated in example 3 in Section 4 where metaphor is a subtype of irony for that particular analysis.
4. Three examples of analysis This section provides three examples taken from the following passage of The Bonfire of the Vanities (Tom Wolfe 1987: 169). Fallow took advantage of this hiatus to make his way across the room toward his cubicle. Out in the middle of the field of computer terminals, he stopped and (…) picked up a copy of the second edition, (…) Just think of the fine sense of gutter syntax that inspired them to create a headline that was all verbs and objects, with the subject missing, the better to make you claw your way inside these smeary black pages to find out what children of evil were fiendish enough to complete the sentence!
Two problems are analysed as metaphors and one as an instance of irony. All of the solutions provided in Spanish are plausible, depending on the aims of the translation. Solutions from Murillo’s (1988: 169) translation are marked TT. I provide a ranking order of preference, from most to least preferable, merely as an aid for readers who are not proficient in Spanish. It is based largely on a criterion of naturalness and idiom; there is no particular preference for retaining the same metaphor or for literal translation. Example 1 analyses field as metaphor, with the newspaper office as Topic, field as Vehicle, and large open space as Ground. Rank: c, d, a, e, b. (1) a.
campo ‘field’ b. sembrado (TT) ‘sown field’ c. mar ‘sea’ d. inacabables ‘neverending’ e. todos aquellos ‘all those [terminals]’
The numbers in square brackets below correspond to binary branching as illustrated in the Figures. Indeed, Figure 1a illustrates the following analysis of example 1.
Translating non-segmental features of textual communication 109
[1] [2] [3] [4] [5]
Same metaphor, in Spanish; for example 1a (literal translation). Same — rural — type of metaphor; for instance 1b. Other — not rural — types of metaphor, including 1c. Other — not metaphorical — rhetorical, e.g. hyperbole, as in 1d. Non-rhetorical solutions; for example, a literal description, as in 1e.
Example 2 is the hiatus metaphor. The Topic is coffee break, hiatus is the Vehicle, the brevity of the pause is the Ground. Rank: c, h, a, e, f, b, g, d. (2) a. b. c. d. e. f. g. h.
hiato ‘hiatus’ cesura ‘caesura’ paréntesis ‘parenthesis’ freno ‘brake’ alto en el camino ‘stop on the way’ interrupción (tan) breve como un hiato ‘interruption (as) brief as a hiatus’ interrupción ‘interruption’ descanso (TT) ‘rest’
[1] Same metaphor; for example 2a (by means of literal translation into L2 Spanish). [2] Same — metalinguistic — type of metaphor; for instance 2b and 2c. [3] Other metaphors (i.e. not metalinguistic), including 2d and 2e. [4] Other rhetorical devices (i.e. not metaphorical), for example a simile, as in 2f. [5] Non-rhetorical solutions, e.g. literal language, as in examples 2g and 2h. Murillo solves hiatus at [5] but field at [2]. This may be due to a certain inconsistency in his solutions. It may, however, be due to my particular analysis in defining metaphor as problem, hiatus and field as metaphors, and common semantic field of the Vehicle (metalinguistic and rural) as the requirement for type-membership. Example 3 is the ironic expression fine sense of gutter syntax. Rank: b, a, d, g, c, f, e. (3) a.
magnífica sintaxis del arroyo ‘magnificent gutter syntax’ b. magnífica sintaxis de cloaca ‘magnificent syntax of the sewer’
110
Patrick Zabalbeascoa
c.
cuánta sensibilidad sintáctica por el morbo más cruel ‘all that syntactic sensitivity at the service of a most cruel morbid pleasure’ d. fina sensibilidad sintáctica ‘sophisticated feel for syntax’ e. Qué magnífica capacidad de síntesis (TT) ‘What a magnificent ability to synthesize’ f. horrible estilo ‘horrible style’ g. sintaxis inmoral ‘immoral syntax’ [1] [2] [3] [4] [5]
Same irony; for example 3a. Same type of (metaphorical antithetical) irony; for instance 3b. Non-metaphorical antithetical irony; including 3c. Non-antithetical irony; for example 3d and 3e. No irony; for example 3f and 3g.
5.
Conclusions
I hope this type-within-type binary-branch model of analysis can be used to show up translational norms that might otherwise be difficult to account for or even identify. The model involves trying out different “trees” to find the most suitable one for each analysis, not only of individual problems but also large numbers of instances of the same problem, in search of elusive regularities. A single segment of text might require more than one “tree” analysis for a full account of what is going on or what is at stake. Nonprescriptive tree diagrams can easily be adapted to prescriptive flowchart diagrams, whereby solution-type [2] is allowed only if none can be found for [1], for instance. A binary-tree analysis can be more meaningful if it has a justifiable structure. The present proposal has found its justification in a type-within-type structure that allows for a variable number of subtypes, according to the typology used. The complexity of the translator’s job is shown by the fact that a given segment might present several phenomena or problems that require separate tree analyses. This could also be a limitation of the model. For example, translations of dragon might have to be analysed separately as metaphor, insult, joke, allusion, etc., each time according to a different typology. The basic binary branch, based on type-within-type, can be further completed by incorporating other binary divisions such as: verbal/non-verbal; “bare”/compound; and L2/not-L2 splits. Finally, I could not find sufficient grounds to assign a definitive location for empty segment, and thus I am forced to regard it as having the potential to belong anywhere.
Translating non-segmental features of textual communication
Notes * This study was carried out as part of Research Project PB98–1062-C04–01, financed by the Spanish Ministry for Education and Science. 1. This negative bias is changing due to progress towards reducing sexist language in Spain. For example, jueza used to refer to a judge’s wife, now it is used mostly to refer to a female judge.
References Allers, Roger and Minkoff, Rob. 1994. The Lion King. Film produced by Walt Disney. Goatly, Andrew. 1997. The Language of Metaphors. London: Routledge. Gozzi, Raymond Jr. 1999. The Power of Metaphor in the Age of Electronic Media. New Jersey: Hampton Press. Grady, Joseph. 1997. “A typology of motivation for conceptual metaphor in five steps”. In Metaphor in Cognitive Linguistics. R. W. Gibbs, Jr. and G. J. Steen (eds.) Amsterdam: John Benjamins. Halliday, M. A. K. and Hasan, Ruqaiya. 1976. Cohesion in English. London: Longman. Hammond, Paul and Hughes, Patrick. 1978. Upon the Pun. Dual Meaning in Words and Pictures. London: W. H. Allen. Holmes, James. 1988. “The name and nature of Translation Studies”. In Translated! Papers on Literary Translation and Translation Studies. J. S. Holmes (ed.) 67–80. Amsterdam: Rodopi. Lakoff, George and Mark Johnson. 1980. Metaphors We Live By. Chicago: University of Chicago. Leech, Geoffrey. 1969. A Linguistic Guide to English Poetry. Harlow: Longman. Newmark, Peter. 1988. Approaches to Translation. New York: Prentice-Hall. Richards, I. A. 1965. The Philosophy of Rhetoric. Oxford: Oxford University Press. Tarantino, Quentin. 1994. Pulp Fiction. Film produced by Lawrence Bender. Toury, Gideon. 1995. Descriptive Translation Studies and beyond. Amsterdam: John Benjamins. White, Roger. 1996. The Structure of Metaphor. Oxford: Blackwell. Wolfe, Tom. 1987. The Bonfire of the Vanities. New York: Bantam Books, Farrar. Spanish version by E. Murillo. 1988. La hoguera de la vanidades. Barcelona: Anagrama. Zabalbeascoa, Patrick. 1997. “Dubbing and the non-verbal dimension of translation”. In Nonverbal Communication and Translation, F. Poyatos (ed.) 327–342. Amsterdam: John Benjamins.
111
Challenging the myth of native speaker competence in translation theory The results of a questionnaire* Nike K. Pokorn University of Ljubljana, Slovenia
In linguistics and translation theory it is generally assumed that every native speaker is able to rapidly detect any non-member of his/her linguistic community. A questionnaire was designed to test this assumption and to find out to what extent native speakers of English could identify a native English translator vs. a non-native translator into English and a single translator vs. a team of translators in a set of translations from Slovene into English. It emerged that native speakers of English cannot always discriminate between native vs. non-native translators, or between single translators and teams of translators. This leads to the conclusion that the definition of the term “native speaker” is still open and far from being final, and also that translation theory should be cautious when referring to the innate capacities of an ideal native speaker.
Many translation theoreticians argue that translations should be done only into the translator’s mother tongue. They are convinced that if translators are working into a foreign language their translations inevitably sound strange to native speakers of the TL, i.e. to the intended public of the text (cf. Kocijancˇicˇ Pokorn 2000). For example, Peter Newmark in his Approaches to Translation expresses this opinion very clearly: The translator is in the best position to appreciate the “total” difference between one language and another. He himself usually knows that he cannot write more than a few complex sentences in a foreign language without writing something unnatural and non-native, any more than he can speak one. He will be “caught” every time, not by his grammar, which is probably suspiciously “better” than an educated native’s, not by his vocabulary, which may well be wider, but by his unacceptable or improbable collocations. (Newmark 1981: 180)
114
Nike K. Pokorn
The assumption that the translator “will be ‘caught’ every time” by native speakers of the TL stems from an equally widely accepted hypothesis in linguistics that every native speaker is able to rapidly detect any non-member of his/her linguistic community. For example Alan Davies in his book The Native Speaker in Applied Linguistics, after admitting that there is no consensus among linguists on the definition of the term “native speaker”, claims that the detection of non-members of one’s native linguistic community is one of the basic and essential characteristics of every native speaker: Let me say what I expect of the native speaker. I expect the native speaker to have internalised rules of use, the appropriate use of language, to know when to use what and how to speak to others. I expect control of strategies and of pragmatics, an automatic feeling for the connotations of words, for folk etymologies, for what is appropriate to various domains, for the import of a range of speech acts, in general for appropriate membership behaviour in him/herself and of implicit — and very rapid — detection of others as being or not being members. (Davies 1991: 94, my emphasis)
Since the claim that native speakers have an infallible ability to distinguish native speakers from non-native speakers used to pass unchallenged in applied linguistics, the majority of translation theorists accepted it and consequently demanded that texts be translated only into the translator’s mother tongue. However, since many of the traditionally unquestioned generalisations concerning the innate abilities of the native speaker have recently proved to be not as binding and absolute as they seemed at first sight (cf. Medgyes 1994; Paikeday 1985; Rampton 1990; SkutnabbKangas and Phillipson 1989), the assumption of the unsuitability of translations done by non-native speakers needed further justification. Thus to answer the question whether every native speaker is in fact always able to immediately detect a non-native translator, and if so, which elements of a text are crucial for such identification, a questionnaire was designed. Included in the questionnaire were seven fragments taken from different published English translations of two Slovene short stories and a novel — all written by Ivan Cankar (1876–1918), one of the most praised classical authors in Slovenia (see appendix). Two texts were translated by Slovene native speakers. The first Slovene translator was Louis Adamic (1898–1951), who was born to Slovene parents in Slovenia but emigrated to the US in 1913 at the age of 15. Although he claimed in some of his interviews that he had completely lost his Slovene and embraced English as his new language, according to the findings of contemporary linguistics (e.g. Davies 1991: 65, 91–92; Krashen 1981: 76; Pinker 1994: 293; Crystal 1994: 368), he was too old when he moved to a new linguistic environment to be able to became a native speaker of English. In the US, Adamic joined the army and became an American citizen in 1917. He worked as a journalist and later on as a
Challenging the myth of native speaker competence in translation theory
free-lance writer, creating all of his works, among them several novels, in English. When he received a Gugenheim reward for his work in 1932, he travelled back to Yugoslavia and harshly criticised the Serbian regime. He had good connections with his homeland after the Second World War and even became a member of the Yugoslav Academy of Science and Arts. He died in 1951, most probably assassinated, at his home in Milford, New Jersey. Jože Paternost, the second Slovene translator, was born in Slovenia in 1931 and moved to the US in 1945. Paternost was also too old (14 years) when he moved to the US to be able to attain the knowledge of English that could be comparable to that of an English native speaker. He graduated in Russian at the University of Ohio in 1955 and received a PhD in Slavonic languages at the University of Indiana in Bloomington. In 1977 he became full professor of Slavonic languages at the State University of Pennsylvania. Two texts were translated by English native speakers. One passage is taken from the translation by Henry Leeming, who was born in 1920 in Manchester. Henry Leeming holds a doctorate in Slavonic languages and taught for 30 years (from 1955 to 1985) at the Slavonic Department of London University. He also lectured at the universities of Cambridge, Oxford and Canberra. He is a corresponding member of Slovene Academy of Arts and Sciences. The second English native speaker, Anthony J. Klancar, was born to Slovene parents in 1908 in Cleveland, USA. He graduated in English and worked as a journalist all his life. In 1977 he died in Cambridge, Massachusetts. Since he was a representative of a second generation of Slovene immigrants, i.e. born in the US, where he had spent all his life and also received his formal education, for the purposes of this study, he is considered to be a native speaker of English in accordance with the findings of modern linguistics (e.g. Davies 1991: 64). Three texts were translated by pairs of translators: one pair consisted of a native speaker of Slovene, Elza Jereb (1935–), who graduated in English and French and became a teacher at the Department of Romance Languages at the University of Ljubljana, and an English native speaker, Alisdair MacKinnon (1934–). MacKinnon was born in South Wales and graduated in English at Cambridge. He worked as a lektor for English language at the University of Ljubljana for three years. The second text was translated by a pair consisting of a person who was familiar with Slovene but was not a native speaker (Agata Zmajic´ (1878–1944) was Croat, born in Croatia but lived in Austria where she also died) and a native speaker of English M. PetersRoberts, who helped her with her translations. The third pair consisted of Anthony J. Klancar, as mentioned above a second generation Slovene immigrant in the US, and a native speaker of American English George R. Noyes, an American professor of Slavonic languages. The questionnaire was intended for English native speakers only. All subjects were born in an English-speaking community, where they also live and work and they indicated that their mother tongue was English. To ensure homogeneous socio-economic background and competence in English, all subjects were students
115
116
Nike K. Pokorn
or faculty at different universities. Included in the group of highly educated native speakers were 5 university undergraduates, 8 graduates, 15 masters of arts, and 18 professors, all working in the field of humanities. There were 46 subjects in total, who varied in age but were all over 20. Since some of the passages were translated by British and others by American translators, 23 subjects were from the UK (11 from the University of Durham, England, 12 from Heriot-Watt University, Scotland), 8 from the US (the University of Kansas) and 15 from Canada: 8 from Vanier College in Quebec and 7 from the University of Alberta. The selected passages in the questionnaire were preceded by a short introduction explaining that the fragments were taken from different translations of two short stories and a novel by Ivan Cankar; and that the original texts were written in Slovene, i.e. in a language spoken by approximately 2 million speakers in Central Europe. The names of the translators were not given, since a foreign sounding name could influence their answers; however, the dates indicating the year when the translations were done were given in brackets at the end of each passage. Subjects were asked to read the passages and indicate whether the translator was a native speaker of American English, English English, some other English (Scottish, Canadian, Australian, etc.) or some other language, not English. By giving them those options, an attempt was made to inform the subjects that the translators might be members of different English communities. The English native speakers were also asked to briefly describe what influenced their decision. Each translated passage was followed by a question about how many translators were, according to their judgement, involved in the translation. The following three options were given: one, more than one, I could not tell how many. At the end of the questionnaire they were asked to define their “ideal” translation, i.e. they could provide their own definition or decide for one of two options: that the translation should be easy to read and fluent in the TL or that it should be as close to the original as possible, even if the structure of sentences in the TL sounds awkward. Finally the subjects were asked to specify the translated passage they liked best. The results showed that even competent native speakers cannot always distinguish between a native and a non-native translator when faced with the translated text only. In responding to the passages translated by one translator an average of 56% of subjects correctly indicated whether the translator of a particular passage was a native or a non-native speaker of English; 56% and 70% respectively of subjects thought that the two Slovene translators were not English. The results for the English translators were not so close: in one case 79% of subjects thought correctly that the translator’s native tongue was English, while in the second case only 20% of subjects granted the translator the status of a native speaker of English. The second “native English” translator seems to present a particular problem: he was born in a Slovene immigrant family in Cleveland, but later became a
Challenging the myth of native speaker competence in translation theory
Table 1.Answers given by native speakers indicating the native language of the translators. The third column gives the number of answers indicating that the translator of a particular passage was an English native speaker, the fourth that the passage was translated by a non-native speaker of English, the fifth column indicates when the subjects could not decide whether the translator was a native or a non-native speaker of English and the sixth column indicates how many interviewees did not answer the question. Names of the translator(s) Mother tongue of the translator(s) Adamic Zmajic´ & Peters-Roberts Paternost Jereb & MacKinnon Leeming Klancar Klancar & Noyes
Native speaker
Slovene Croat and English Slovene Slovene and English English English English and English
Non-native Cannot No speaker tell answer
20 35 14 31
26 9 32 14
0 0 0 0
0 2 0 1
34 9 36
8 35 7
1 1 0
3 1 3
Table 2.Answers given by the interviewed native speakers indicating the native language of the translators in percentages. The language in the brackets indicates the mother tongue of the translators 90 80
percent of responses
70 60 50 40 30 20 10 0 Leeming (English)
Paternost (Slovene)
native
43,48
30,43
20
79,54
68,89
83,72
non-native
56,52
69,56
18,60
77,78
20,45
31,11
16,28
cannot tell
0
0
2,32
2,22
0
0
0
79,07
Klancar (English)
Zmajic & Jereb & Klancar & Peters-Roberts MacKinnon Noyes (Engl. & (Croat&Engl.) (Slov. & Engl.) Engl.)
Adamic (Slovene)
journalist, finished his studies of English and spent his whole life in the US. According to criteria accepted in linguistics, he should be classified as an English native speaker, however, the response shows that his English would not be accepted as a native variety by other members of the community. Further problems arose
117
118
Nike K. Pokorn
when the analysis of his translations showed that he had problems with Slovene as well, since he often misunderstood Slovene passages. It seems that this suggests that the number of years of use of a particular language and even the fact that one is born in a particular linguistic community do not correlate with linguistic competence and performance in translation. Since he was most probably aware of the fact that his English did not sound right to others, the second translator decided to rewrite his translations with an American linguist. A passage from that revised translation got a much better response in the questionnaire: 84% of subjects thought it was done by a native speaker of English. In fact, when assessing the passages translated by pairs of translators, an average of 77% of subjects thought that the passages were translated by native speakers of English (79%, 69%, and 84%). When indicating the number of translators involved in the translation, the native speakers interviewed were again often in doubt. Table 3.Answers given by the interviewed native speakers indicating the number of translators involved in the translation. The third column gives the number of answers indicating that the passage was translated by one translator, the fourth that the passage was translated by more than one translator, the fifth column indicates when the subjects could not decide on the number of translators involved in the translation of a particular passage, and the sixth column indicates how many interviewees did not answer the question. Names of the translator(s) Mother tongue of the translator(s) Adamic Zmajic´ & Peters-Roberts Paternost Jereb & MacKinnon Leeming Klancar Klancar & Noyes
Slovene Croat & English Slovene Slovene & English English English English & English
One More Cannot translator than one tell 12 28 15 15 22 19 18
13 1 4 6 1 4 6
19 14 22 23 17 15 18
No answer 2 3 5 2 6 8 4
An average of 44% of subjects cannot specify the number of translators involved in the translation of a particular passage. Of 46 subjects interviewed, an average of only 10% correctly indicated that a translated passage was done by a pair of translators, and 42% of them were right in saying that a passage was translated by one translator only. In accordance with Venuti’s findings that English-language translation is dominated by the principle of transparency, i.e. by the desire to create an illusory effect of “naturalness” (Venuti 1995: 1–42), the vast majority of subjects (86%) think that an ideal translation should be fluent and easy to read in the TL.
Challenging the myth of native speaker competence in translation theory
Table 4.Answers given by the interviewed native speakers indicating the number of translators involved in the translation in percentages. The language in brackets indicates the mother tongue of the translator(s) 100 90 80
percent of responses
70 60 50 40 30 20 10 0 Adamic (Slovene)
Paternost (Slovene)
Jereb & MacKinnon (Slov.&Eng.)
Leeming (English)
Klancar (English)
Klancar & Noyes (Eng.&Eng.) 42,86
65,12
36,58
34,09
55
50
more than one
29,54
9
9,76
13,64
2,5
10,53
14,28
cannot tell
43,18
32,56
53,66
52,27
42,5
39,47
42,86
one translator
27,27
Zmajic& PetersRoberts (Croat&Eng.)
Table 5.The “ideal” translation strategy according to the interviewed native speakers expressed in percentage Other 7% SL-oriented translation 7%
TL-oriented translation 86%
However, it is surprising that translations done by pairs of translators are most appreciated: 83% of subjects chose one of the three translations done by pairs of
119
120 Nike K. Pokorn
Table 6.The “best” translated passage according to the interviewed native speakers expressed in percentages. The language in brackets indicates the mother tongue of the translator(s). Adamic (Slovene) 3%
Klancar & Noyes (English & English) 45%
Klancar (English) 0%
Zmajic & PetersRoberts (Croat & English) 21% Paternost (Slovene) 6% Jereb & MacKinnon (Slovene & English) 17% Leeming (English) 8%
translators as their personal best, which shows that a collaboration, apparently, does not influence TL fluency, i.e. does not have an adverse effect on the quality of the TL in the translation. To sum up, the results show that competent native speakers of English cannot always recognise the foreign and disturbing elements in translations by non-native speakers. This finding is even more marked when assessing passages translated by pairs of translators where there seemed to be no elements that the readers found ‘foreign and disturbing’. The questionnaire has thus shown that the assumption that every native speaker is able to rapidly detect any non-member of his/her linguistic community, when confronted only with a written document, has no solid foundation. Native speakers also cannot tell if the work was done by one or more translators. Since native speakers of the TL of the translated text do not recognise the “foreign and disturbing elements” in some translations by non-native speakers and the majority of them do not recognise those elements in translations which are the result of a collaboration between a native and a non-native translator, this leads to a conclusion that the definition of the term “native speaker” in linguistics and translation theory is still open and far from being final, and that translation theory should therefore be cautious when referring to the innate capacities of the ideal native speaker. Translations into a non-mother tongue, especially translations done by pairs of translators, need to be examined more closely by translation theorists. And last but not least, translating out of one’s mother tongue should also be openly discussed by those scholars involved in developing teaching techniques for translation classes, instead of dwelling on the demand that all translations be done only into one’s mother tongue.
Challenging the myth of native speaker competence in translation theory
Author’s Address: Nike K. Pokorn University of Ljubljana Department of Translation Studies Askerceva 2 1000 Ljubljana Slovenia e-mail:
[email protected] Note * I would like to thank those who graciously contributed the time and intuitions to this study: in particular Prof. Andrew Louth from the University of Durham, Dr. Donald McLaughlin from Heriot-Watt University, Prof. Tom Priestly from the University of Alberta, Prof. Marc Greenberg from the University of Kansas and Prof. Tom Lozar from Vanier College.
Appendix: The translated passages included in the questionnaire I. One day I craved black coffee. I don’t know how it came to my mind; I simply wanted some black coffee. Perhaps because I knew that there was not even a slice of bread in the house, and that much less coffee. Sometimes a person is merciless, cruel. Mother looked at me with her meek, surprised eyes but would not speak. After I informed her that I wanted some black coffee, I returned to the attic to continue my love story, to write how Milan and Breda loved each other, how noble, divine, happy and joyful they were… “Hand in hand, both young and athrob with life, bathed in morning dew-drops, swaying —” Then I heard light steps on the stairs. It was mother, ascending carefully, carrying a cup of steaming coffee. Now I recall how beautiful she was at that moment. A single ray of sun shone directly onto her eyes through a crack on the wall. A divine light o’ heaven, love and goodness were there in her face. Her lips held a smile as those of a child bringing one a gift. But — “Oh, leave me alone!” I said harshly. “Don’t bother me now! I don’t want any coffee!” (Louis Adamic, 1926) II. One day I thought I should like to have some black coffee. I don’t know why such an idea came into my head, but I just thought I would ask for some. Perhaps it was that I knew perfectly well there was not even bread in the house, much less coffee. Lack of imagination can make a man cruel and wicked. When I asked for a cup of coffee, my mother looked at me with big, shy eyes and made no reply. Peevishly and grumpily, I left the room, without another word, and went up to my garret, where I wrote of “Milan and Breda who loved each other”; people of rank, happy, and serene, who went “hand in hand, young and gay, through the morning dew and rising sun.”
121
122 Nike K. Pokorn
And now I heard soft steps on the stairs. It was my mother, walking slowly and carefully and in her hands was a cup of coffee. I remember now that she had never seemed to look so lovely. The slanting rays of the setting sun fell through the door of my garret on to her eyes which were big and clear, and as if filled with a heavenly light; love and kindness were reflected in them. Her lips smiled like those of a child who wants to surprise you with a pretty gift. But I turned away, and in a cold, cruel voice said, “Leave me in peace; I do not want your coffee.” (Agata Zmajic´, M. Peters-Roberts, 1933) III. And then, once I craved black coffee. I don’t know how it came to my mind; I merely wished to have it. Perhaps it was because I knew there was no bread in the house, not to speak of coffee. A man is in his very thoughtfulness malicious and cruel. Mother looked at me with big, timid eyes and did not answer. Ill-humored and peevish, without a word I returned under the roof in order to write how Milan and Breda loved each other and how they were so happy and gay. “Hand in hand, together, young, illuminated by the morning sun, washed in the dew…” I heard quiet steps on the stairs. It was Mother; she was treading slowly and cautiously. In her hand she was carrying a cup of coffee! I recall now that she was never so beautiful as in that moment. Thru the door shone a sloping beam of the noon sun, straight into my mother’s eyes; they were bigger, purer, all heavenly light was reflected from them, all heavenly nobleness and love. She smiled as if to a child to whom she was bringing a joyful gift. Yet I looked around and said in a withering voice: “Leave me in peace! … I don’t want it now!” (Joze Paternost, 1957) IV. Once I felt a strong craving for a cup of coffee. I do not know how this came to my mind; but I wanted it. Perhaps simply because I knew that we had not even bread at home, let alone coffee. Out of pure inadvertence man may be evil and pitiless. My mother looked at me wide-eyed and timid and gave no answer. Sour and full of ill humour, without as much as a word I went up to my loft to write about the love of Milan and Breda, and how noble, fortunate, happy and gay they both were. “Hand in hand, the two young people, in the full glow of the morning sun, bathed in dew…” At that moment I heard a quiet step on the stairs. It was my mother; she was treading slowly and carefully; in her hand she carried a cup of coffee. I recall now hat she was never as beautiful as at that moment. Through the door came a shaft of midday sun, right into my mother’s eyes. They were larger and purer, all the light of heaven shone out of them, all heaven’s love and tenderness. On her lips there was a smile like that of a child bringing a happy gift. But I turned and said nastily: “Leave me alone! … I don’t want it now!” (Elza Jereb, Alisdair MacKinnon, 1971) V. Her mother got up and kissed her. As her cheeks began to smart with her mother’s tears a sudden feeling of tenderness came over her. She raised her hand to touch her mother’s face, which was all hot and damp. She saw that face close up before her and it was as if she was seeing it for the first time, broad, flushed and furrowed with tears. The eyes were terrified, swollen, dulled with grief, the lips quivered. Sister Cecilia led her mother to the door.
Challenging the myth of native speaker competence in translation theory
“Is she never going to come away from here again?” her mother asked the sister with a strange look in her eyes, like a child begging for a present. “God’s will be done!” Sister Cecilia said quietly. The door closed and the footsteps died away along the corridor. (Henry Leeming, 1968) VI. The children were in the habit of conversing before they went to sleep. They sat for awhile on a broad, flat stove and told each other what happened to occur to them. Evening dusk peeped into the room through dim windows, with its eyes full of dreams; the silent shadows writhed upward from all corners and carried off their extremely wonderful fairly tales. They related whatever entered their minds, but their thoughts were only of beautiful stories spun out of the sun and its warmth, out of love and hope woven of dreams. All futurity was just one long, splendid holiday; between their Christmas and Easter came no Ash Wednesday. There somewhere behind the variegated curtains silently overflowed all life, twinkling and flashing from light to light. Their words were half-understood whispers; no story had either a beginning, nor distinct images, no fairy tale an end. Sometime all four children spoke at the same time, and nobody disturbed another; they all gazed fascinated at that wondrously beautiful celestial light, and there every word rang true, there every story had its own pure, living and lucid visage; every tale its splendid end. (Anthony Klancar, 1933) VII. The children were in the habit of talking together before they went to sleep. They sat for awhile on a broad, flat stove and told one another whatever happened to occur to them. The evening dusk peered into the room through dim windows, with its eyes full of dreams; the silent shadows writhed upward from every corner and carried away with them their marvellous fairy tales. The children related whatever entered their minds, but their thoughts were only of beautiful stories spun from the sun and its warmth, from love and hope woven of dreams. All their future was just one long, glorious holiday; between their Christmas and Easter came no Ash Wednesday. Somewhere behind variegated curtains all life silently overflowed, twinkling and flashing from light to light. Their words were half-understood whispers; no story had either a beginning or distinct images; no fairy tale had an end. Sometimes all four children spoke at once, yet no one of them disturbed another; they all gazed fascinated at that wondrously beautiful celestial light, and in that setting every word rang true, every tale had its splendid end. (Anthony Klancar, George R. Noyes, 1934/35)
References Cankar, Ivan. 1926. “A cup of coffee”. Trans. by Louis Adamic. Juvenile — Mladinski list 5: 82–83. Cankar, Ivan. 1933. “Children and old people”. Trans. by Anthony J. Klancar. New Era — Nova doba 9, 36: 6. Cankar, Ivan. 1933. “A cup of coffee”. Trans. by Agata Zmajic´, M. Peters-Roberts. Review of Reviews 84,1: 52–53. Cankar, Ivan. 1934/35. “Children and old people”. Trans. by Anthony J. Klancar in George R. Noyes. The Slavonic Review 13: 494–496. Cankar, Ivan. 1957. “A cup of coffee”. Trans. by Jože Paternost. Ameriška domovina (8. February): 6.
123
124 Nike K. Pokorn
Cankar, Ivan. 1971. “A cup of coffee”. My Life and Other Sketches. Trans. by Elza Jereb, Alasdair MacKinnon. Ljubljana: Državna založba Slovenije. Cankar, Ivan. 1976. The Ward of Our Lady of Mercy. Trans. by Henry Leeming in 1968. Ljubljana: Državna založba Slovenije. Crystal, David. 1994. (1987.) The Cambridge Encyclopedia of Language. Cambridge, New York, Melbourne: Cambridge University Press. Davies, Alan. 1991. The Native Speaker in Applied Linguistics. Edinburgh: Edinburgh University Press. Kocijancˇicˇ Pokorn, Nike. 2000. “Translation into a non-mother tongue in translation theory: Deconstruction of the traditional”. Translation in Context: Selected Contributions from the EST Congress, Granada 1998, Andrew Chesterman & Navidad Gallardo & Yves Gambier (eds), 61–72. Amsterdam, Philadelphia: John Benjamins. Krashen, Stephen D. 1981. Second Language Acquisition and Second Language Learning. Oxford, New York, Toronto, Sydney, Frankfurt: Pergamon Press. Medgyes, Peter. 1994. The Non-Native Teacher. Hong Kong: MacMillan Publishers. Newmark, Peter. 1981. Approaches to Translation. Oxford, New York, Toronto, Sydney, Frankfurt: Pergamon Press. Paikeday, Thomas. M. 1985. The Native Speaker is Dead!: An informal discussion of a linguistic myth with Noam Chomsky and other linguists, philosophers, psychologists, and lexicographers. Toronto, New York: Paikeday Publishing Inc. Pinker, Steven. 1994. The Language Instinct: The New Science of Language and Mind. London, New York, Ringwood, Toronto, Aukland: Penguin Group. Rampton, M. B. H. 1990. “Displacing the ‘native speaker’: Expertise, affiliation, and inheritance”. ELT Journal 44, 2: 97–101. Skutnabb-Kangas, Tove & Phillipson, Robert. 1989. “‘Mother Tongue’: The theoretical and sociopolitical construction of a concept”. Status and Function of Languages and Language Varieties, Ulrich Ammon (ed), 450–477. Berlin, New York: Walter Gruyter. Venuti, Lawrence. 1995. The Translator’s Invisibility: A History of Translation. London, New York: Routledge.
Expectativas y evaluación en la traducción de folletos turísticos Marie-Louise Nobs Universidad de Granada
This paper describes the design and carrying out of a two-part pilot study on translation quality. The first part of the study is descriptive in nature and deals with the expectations of real users of translated tourist leaflets; the second part is experimental and is designed to allow a comparison of the assessment made by these users of a published translation, on the one hand, and a second translated version of the same source text, manipulated by the researcher, on the other. The questionnaires drawn up to collect the data cover seven basic quality parameters, including accuracy of information, clarity and readability of the target text, attractive visual presentation, and professional behaviour of the translator. Preliminary results obtained in the pilot study are presented and discussed and modifications are proposed for the final version of the questionnaires. The conclusions describe the tendencies observed during this pilot stage.
1.
Introducción
El objeto de este artículo lo constituye un estudio piloto sobre evaluación de calidad en traducción realizado con usuarios reales de folletos turísticos traducidos. Una vez aclarado el enfoque de la investigación nos centraremos en algunos aspectos relacionados con su diseño y realización. Asimismo presentaremos y discutiremos algunos de los resultados obtenidos y propondremos determinados cambios en los cuestionarios para la investigación futura hacia la que se orienta la prueba piloto. En las conclusiones destacaremos las tendencias observadas en el marco del estudio realizado y sus posibles implicaciones.
2.
Planteamiento de la investigación
Plantearse una investigación que enfoca las expectativas y la evaluación de usuarios reales de folletos turísticos traducidos, tiene su origen en una serie de preguntas que
126 Marie-Louise Nobs
como traductora y docente de la traducción nos venimos haciendo desde hace algunos años. Son preguntas que giran en su mayoría alrededor de una hipótesis no probada, según la cual un texto meta (TM) sólo sería percibido por parte de los usuarios reales del mismo como un texto aceptable si tiene en cuenta las expectativas de los destinatarios, no sólo en relación con la corrección lingüística formal, sino también en relación con el contenido, su presentación visual y la interrelación entre ambos. Siendo el objetivo general de nuestra investigación la comprobación de la hipótesis de partida a la que acabamos de aludir, su consecución pasaba por elaborar un diseño de estudio que permita dar respuestas a las siguientes preguntas: 1. ¿Cuáles son las expectativas que un grupo de usuarios reales tiene en relación con un determinado tipo de texto, en concreto con un folleto turístico? ¿Muestran estas expectativas entre los sujetos encuestados una clara tendencia o es imposible llegar a observar rasgos comunes? 2. ¿Cuál es la evaluación de la calidad que usuarios germanohablantes hacen de un folleto turístico determinado en su versión traducida del español al alemán? ¿Muestra esta evaluación entre los sujetos encuestados una clara tendencia o es imposible llegar a observar rasgos comunes? 3. ¿Cómo ponderan los sujetos encuestados cada uno de los parámetros de calidad barajados? 4. ¿En qué medida influyen las características de los sujetos encuestados en los resultados, especialmente en relación con las siguientes variables: pertenencia a una determinada cultura y lengua, pertenencia a una determinada franja de edad, nivel cultural y grado de conocimiento del tema del folleto en cuestión? 5. ¿Refleja la evaluación de los sujetos consultados las expectativas que éstos habían formulado con anterioridad? 6. ¿Existe correlación entre la evaluación de la calidad global que realizan usuarios reales de un folleto turístico traducido y el grado de profesionalidad que éstos atribuyen a la persona que lo ha traducido? Partimos de un enfoque funcional de la traducción que nos lleva a plantear el por qué, el para qué, el para quién y además el dónde se traduce. Se trata de un enfoque que parte de la base de que al traducir el traductor actúa con un determinado objetivo y que traducir es, por tanto, un acto de comunicación transcultural intencionado que se desarrolla en un marco de actuación concreto y a requerimiento de terceros (cf. Holz-Mänttäri, 1984 y 1993) en el que cada actor y cada factor implicado se puede estudiar. Consideramos que comprender un texto es esencialmente un proceso subjetivo y selectivo (cf. Hönig 1989: 126 y Hatim & Mason, 1995: 123), lo que nos prohíbe negar o disimular el carácter subjetivo del mismo. Exceptuando muy pocos estudios (cf. Jänis & Priiki, 1993), la mayoría de las investigaciones que enfocan, como el nuestro, las reacciones de determinadas audiencias respecto a una determinada traslación se efectuaron por investigadores
Expectativas y evaluación en la traducción de folletos turísticos 127
de la interpretación. Para nuestros fines son especialmente interesantes los estudios sobre evaluación realizados por Gile (1990) y Collados (1998). Gile introduce por primera vez el parámetro de la calidad global de una traslación mientras que Collados destaca la importancia de parámetros no verbales para evaluar una determinada traslación y advierte que no conviene equiparar conceptos como ‘calidad’ de una traducción y ‘éxito’ de una traducción. En palabras de Collados: “Fracaso [comunicativo] no equivale necesariamente a deficiente calidad, ni éxito a constatada calidad” (Collados 1998: 25).
3.
Diseño de la investigación
Somos conscientes de que la aplicación del método experimental en el campo de la traducción plantea dificultades reales. Ello se debe, entre otros, a la dificultad de controlar las variables externas, la falta de instrumentos de medida específicos y el hecho de que tanto en las Ciencias Sociales como en Traductología el investigador forma parte de lo observado. El hecho de proseguir con el diseño experimental a pesar de las dificultades se debe a nuestro convencimiento de que la aplicación de una metodología transparente que posibilita la replicabilidad puede contribuir a aclarar cómo determinados parámetros de calidad cooperan en el éxito o fracaso de una traducción. 3.1 Parámetros de calidad Sobre la base de la bibliografía consultada establecimos siete parámetros mediante los cuales los usuarios reales de un folleto turístico traducido tenían que evaluar la calidad concreta de los textos traducidos que en el marco del experimento les presentamos. Concretamente se trata de los parámetros siguientes: a. b. c. d. e. f. g.
Veracidad de la información transmitida Claridad y comprensibilidad del TM Información útil y accesible para un usuario real del TM Corrección gramatical Estilo funcional acorde al tipo del TM Presentación visual y gráfica atractiva Comportamiento profesional
3.2 Tipo de diseño del estudio piloto El estudio concreto que nos ocupa es un estudio exploratorio. Se trata de un estudio seccional en el que los sujetos son consultados una sola vez. Los datos se recaban
128 Marie-Louise Nobs
mediante encuesta por administración de cuestionarios a grupos determinados de usuarios reales de una traducción de un folleto turístico sobre Sierra Nevada (Granada/España). El diseño propuesto corresponde a un cuasi-experimento, teniendo en cuenta la falta de aleatoriedad a la hora de seleccionar la muestra. El estudio consta de una primera parte descriptiva y una segunda parte experimental. La parte descriptiva se refiere a las expectativas de los usuarios reales de folletos turísticos y la parte experimental se centra en la evaluación que éstos realizan respecto a dos textos traducidos concretos, una traducción publicada (Tradpubl) y una traducción manipulada (Tradmanipul), ambas elaboradas sobre la base del mismo TO. 3.3 Objetivos del estudio piloto Teniendo en cuenta que todo estudio piloto se orienta por definición en los objetivos de una futura investigación, recordaremos que la consecución de éstos pasa por encontrar respuestas a las seis preguntas que más arriba planteamos. En cuanto a los objetivos específicos del pretest (cf. Diekmann 1998: 415ss.) es evidente que el objetivo principal del mismo es de tipo metodológico y se centra fundamentalmente en comprobar la coherencia y la validez de los cuestionarios como herramienta de observación y de medida de datos. En el marco de esta aportación optaremos, no obstante, por resaltar y comentar también algunos de los resultados obtenidos, al considerar que, aún no siendo representativos y concluyentes, pueden señalar ciertas tendencias. 3.4 Sujetos del estudio piloto Los sujetos consultados para el pretest fueron 24, procedentes de cuatro Escuelas de Español para Extranjeros, todas ellas sitas en Granada capital y con acreditada experiencia. En los cuatro centros docentes estaban matriculados en aquel momento 76 alumnos germanohablantes de los cuales 24 devolvieron el cuestionario cumplimentado (31,6%). Existen varias razones por las que escogimos como sujetos para el estudio piloto a estudiantes de español como lengua extranjera siendo la razón fundamental la de formar parte del grupo de turistas que Latiesa (2000: 115) denomina “viajeros de alto contacto”, caracterizados por ser más exigentes con la calidad global de la oferta turística que otro tipo de visitantes. 3.5 Material del estudio piloto 3.5.1 Los textos La selección del texto1 base fue el resultado de un compromiso entre las exigencias planteadas por el método científico y las condiciones reales dictadas por las exigencias
Expectativas y evaluación en la traducción de folletos turísticos 129
de la situación investigadora concreta. El texto utilizado en nuestro experimento debería ajustarse a las siguientes características, que en el caso óptimo tendrían que cumplirse todas a la vez: a.
Ser un texto que pertenece al género de texto de los folletos turísticos editados por organismos oficiales y distribuidos de manera gratuita entre sus usuarios. b. Ser un texto traducido al alemán sobre la base de un texto original redactado en español. c. Ser un texto auténtico, es decir, un texto que efectivamente se ha publicado y que todavía está en circulación. d. Ser un texto corto (máximo 350 palabras). e. Ser un texto autónomo (un texto que funciona sin información añadida). f. Ser un texto que no sea neutro en relación con uno o varios de los parámetros de calidad barajados. El texto seleccionado fue finalmente un texto sobre Sierra Nevada editado en el año 1999 por el Patronato Provincial de Turismo de Granada. Forma parte de un folleto en color de unas dimensiones reducidas de 10 por 21 centímetros. Optamos por este texto, por considerar que cumple con todas las características buscadas. 3.5.1.1 La traducción publicada. El texto base de nuestro experimento es una traducción de un TO en español. Ambos textos, el TO y la traducción alemana publicada (Tradpubl), tienen una presentación visual idéntica, manteniendo exactamente el mismo diseño gráfico, reproduciendo la misma disposición espacial de los elementos verbales y no verbales. En cuanto a su aparición externa, se podría decir que el original y la traducción no se distinguen, o dicho de otro modo, que el TM es un calco del TO. Parece que la idea de calcar o, tal como dice Larose (1998: 172), de “clonar” el TO ha sido aplicada también al tratamiento del material verbal por parte del traductor. Para nuestros fines el texto traducido en un folleto bajo el título “Granada monumental” era especialmente aprovechable debido a que se trata de un texto que en relación con varios de los parámetros de calidad barajados, puede ser percibido como un texto no neutro, o dicho parafraseando a Hönig (1995: 26–27), como un texto que no corresponde, ni a las convenciones de género vigentes en la cultura meta, ni a las expectativas de los usuarios, convirtiéndose de este modo en un texto que involuntariamente llama la atención sobre aspectos no deseados. Nuestra primera percepción fue contrastada antes de seleccionar el texto con la de otros usuarios germanohablantes, para descartar la arbitrariedad de la misma. Por ello reunimos en mayo de 1999 un grupo de estudiantes universitarios de intercambio a los que presentamos el texto en cuestión y a los que pedimos que intentaran verbalizar el efecto que les causaba.2 En el grupo de discusión hubo unanimidad al afirmar que la Tradpubl provocaba en el lector una pequeña sonrisa debido a que el estilo utilizado
130 Marie-Louise Nobs
no correspondía a las convenciones textuales vigentes en la cultura meta, sin que ello disminuyera considerablemente la comprensión del contenido del texto. 3.5.1.2 La traducción manipulada. La idea que nos ha guiado a la hora de manipular la Tradpubl era esencialmente la de producir una traducción a la que no se le notara a primera vista que se trata de una traducción, es decir, un TM sin incorrecciones de tipo gramatical, escrito de modo comprensible en un estilo neutro, usando el nivel de lengua apropiado, de modo que no llamara la atención de modo involuntario. Para ello procedimos a la manipulación de los siguientes parámetros de calidad: corrección gramatical, claridad y comprensibilidad y estilo funcional acorde al tipo del TM. Los demás parámetros, es decir los relacionados con la adecuación pragmática, el tratamiento del material visual y la veracidad de la información transmitida no los manipulamos, dejándolos tal como en la Tradpubl. 3.5.2 Los cuestionarios Para garantizar una recogida sistematizada de la información nos basamos en la elaboración de los cuestionarios en las recomendaciones hechas por distintos autores procedentes de las Ciencias Sociales, especialmente en Sierra Bravo (1994), Diekmann (1998) y Schnell et al. (1999). 3.5.2.1 La estructura de los cuestionarios. En la encuesta piloto, utilizamos dos cuestionarios. El cuestionario 1 enfoca las expectativas que usuarios germanohablantes tienen respecto a un folleto turístico sobre Granada y el cuestionario 2 se centra en la evaluación que estos mismos usuarios realizan de un TM concreto. El cuestionario 1 comienza con preguntas referidas a los parámetros e indicadores de calidad que más arriba presentamos y concluye con preguntas mediante las cuales pretendemos conocer los datos sociodemográficos y socioculturales de los encuestados, lo que permitirá evaluar los datos según la atribución de los sujetos a las siguientes variables: lengua materna, nacionalidad, lugar habitual de residencia, sexo, edad, actividad profesional, formación, tipo de viaje, conocimiento de Granada, conocimiento de Sierra Nevada, afición al esquí, afición al senderismo y conocimiento de la lengua española. El cuestionario 2 enfoca la evaluación que los mismos sujetos hacen del texto turístico concreto, cuya lectura es necesaria para poder contestar a las preguntas que contiene. Las preguntas retoman los mismos parámetros e indicadores de calidad que utilizamos en el cuestionario 1. Con ello pretendemos crear la posibilidad de comparar los datos de los dos cuestionarios, o, más concretamente, establecer una correlación entre expectativas y evaluación. 3.5.2.2 La administración de los cuestionarios. Los cuestionarios fueron distribuidos durante el mes de julio de 1999 en los cuatro centros de enseñanza citados. Por
Expectativas y evaluación en la traducción de folletos turísticos
razones prácticas optamos por la autoadministración de los cuestionarios, es decir que los mismos sujetos fueron los que los rellenaran, sin ayuda de ningún entrevistador. Esta opción imponía, evidentemente, una redacción clara e inequívoca de las preguntas, con el fin de evitar ambigüedades que pudieran dar pie a una cumplimentación no completa. El diseño de la parte experimental del estudio preveía que la evaluación de la calidad se llevase a cabo sobre la base de dos textos traducidos, siendo uno la traducción publicada del folleto (Tradpubl) y el otro una traducción manipulada por nosotros (Tradmanipul). Este planteamiento implicaba que la mitad de los sujetos consultados hiciesen su evaluación sobre la Tradpubl y la otra mitad sobre la Tradmanipul. Los sujetos desconocían este planteamiento. Los respectivos textos les fueron suministrados con el cuestionario de modo que la Tradpubl y la Tradmanipul fueron evaluadas cada una por 12 sujetos, repartidos de modo proporcional entre los 24 estudiantes de los cuatro centros docentes. Por motivos eminentemente prácticos no preveíamos en el marco del pretest ningún grupo de control que hubiera tenido acceso a ambas traducciones, a la Tradpubl y la Tradmanipul, por lo que la comparación de las dos traducciones por los mismos sujetos quedó excluida.
4. Resultados del estudio piloto A continuación procedemos a una presentación selectiva de los resultados, centrándonos exclusivamente en dos aspectos relacionados con las expectativas y dos con la evaluación de los sujetos.3 4.1 Las expectativas de los usuarios reales (cuestionario 1) 4.1.1 La valoración de los parámetros barajados En la primera pregunta del cuestionario se les pide a los usuarios que determinen la importancia que atribuyen a seis parámetros en relación con un folleto turístico puesto a su disposición de forma gratuita por la oficina de turismo de Granada. La tabla 1 refleja los resultados frecuenciales al respecto y el gráfico 1 los presenta en el orden de preferencia que los encuestados establecieron, indicando los valores medios de cada uno. Para la evaluación disponían de una escala del 1 al 5, significando 1: ninguna importancia y 5: máxima importancia. 4.1.2 La incidencia de los parámetros en la merma de la calidad Mediante la pregunta 8 del cuestionario pretendíamos conocer en qué medida los usuarios reales de un folleto turístico traducido atribuyen la disminución de la calidad a parámetros concretos. En la tabla 2 presentamos los resultados frecuencia-
131
132
Marie-Louise Nobs
Información que corresponde con la realidad
4,75
Información útil para un turista
4,58 Texto claro y fácilmente comprensible
4,37
Ausencia de errores gramaticales
3,54 3,37
Presentación visual y gráfica atractiva
2,95
0
1
2
3
Estilo adecuado que no provoca hilaridad
4
5
Gráfico 1.Valoración de parámetros Tabla 1.Valoración de parámetros
Información que corresponde con la realidad Presentación visual y gráfica atractiva Ausencia de errores gramaticales Información útil para un turista Texto claro y fácilmente comprensible Estilo adecuado que no provoca hilaridad
1
2
3
4
5
0 0 0 0 0 2
0 2 3 1 0 5
1 11 9 0 3 10
4 11 8 7 9 4
19 0 4 16 12 2
ns/nc media 0 0 0 0 0 1
4,75 3,37 3,54 4,58 4,37 2,95
les correspondientes a la pregunta 8. Mediante el gráfico 2 visualizaremos el peso que los encuestados atribuyeron a cada parámetro. Para determinar la incidencia de los parámetros en la merma de la calidad de un folleto turístico traducido los sujetos disponían de una escala del 1 al 5, significando 1 = ninguna merma de calidad y 5 = máxima merma de calidad. 4.1.3 Discusión En relación con la valoración de los parámetros (cf. Tabla 1 y Gráfico 1) llama la atención que los tres parámetros que obtienen una valoración superior a 4 se refieren a aspectos relacionados con el contenido del folleto. La valoración de los otros tres parámetros, todos relacionados con aspectos relativos a la forma del folleto, es claramente inferior, siendo la ‘corrección gramatical’ el parámetro formal más valorado (3,54) seguido por la ‘presentación visual y gráfica atractiva’ (3,37) y el ‘estilo adecuado que no provoca hilaridad’ (2,95). El hecho de que la adecuación estilística sea el único parámetro que haya obtenido una valoración por debajo de
Expectativas y evaluación en la traducción de folletos turísticos
Información no fehaciente
4,82
Comprensión dificultosa
4,62
Cambios respecto al contenido del TO Folleto redactado en estilo poco habitual
3,91 3,54
Presentación visual/gráfica poco atractiva Folleto contiene errores gramaticales Se nota que es una traducción
3,5 3,16 2,75 2,33
Folleto produce hilaridad
2,04
0
2
Cambios respecto a la forma del TO
4
6
Gráfico 2.Incidencia de los parámetros en la merma de la calidad Tabla 2.La incidencia de los parámetros en la merma de la calidad
Folleto produce hilaridad Folleto redactado en estilo poco habitual Folleto contiene errores gramaticales Se nota que es una traducción Presentación visual/gráfica poco atractiva Comprensión dificultosa Información no fehaciente Cambios respecto al contenido del TO Cambios respecto a la forma del TO
1
2
3
4
5
ns/nc
media
5 0 0 4 1 0 0 2 5
9 2 1 5 2 0 0 1 10
8 9 12 8 5 0 0 6 5
1 11 9 7 11 9 4 7 1
1 2 2 0 4 15 20 6 1
0 0 0 0 1 0 0 2 2
2,33 3,54 3,16 2,75 3,5 4,62 4,82 3,91 2,04
3 podría indicar que los usuarios de folletos turísticos traducidos estan dispuestos a aceptar que éstos tengan un estilo poco usual. Esta misma idea se ve confirmada en la pregunta 6, en la que se les pregunta a los sujetos, si les molesta si a un folleto turístico se le nota que se trata de una traducción, a la que un 62,5% contestaron con un no, frente a un 37,5% que contestó de modo afirmativo. En relación con la incidencia de los parámetros en la merma de la calidad (cf. Tabla 2 y Gráfico 2) podemos observar que los encuestados confirman, a grandes rasgos, el orden de preferencia de los parámetros que habían manifestado previamente. Parece confirmarse que los parámetros relacionados con el contenido son considerados más importantes que los relacionados con la forma, considerando que lo que más afecta a la disminución de la calidad de un folleto turístico es la inclusión de
133
134 Marie-Louise Nobs
información que no corresponde con la realidad y el hecho de que su comprensión sea dificultosa. La pregunta sobre la incidencia de los parámetros en la merma de la calidad, al retomar los parámetros barajados en la pregunta 1, sirve, entre otros, de control a las respuestas recabadas en relación con la valoración de los parámetros. De hecho una comparación de los resultados al respecto pone de relieve que los usuarios establecen una correlación importante entre la valoración de los parámetros y la valoración de la incidencia de los parámetros en la merma de la calidad. Sirva como ilustración de ello la comparación de los siguientes valores (recabados en las preguntas 1 y 8). 6 5
4,83
4,75 4,37
4
4,62
veracidad de la información transmitida
3,54 3,16
claridad y comprensibilidad del TM
3
corrección gramatical 2 1 0
pregunta 1 / pregunta 8
Gráfico 3.Comparación entre la valoración de los parámetros y su incidencia en la merma de la calidad
Tabla 3.Comparación entre la valoración de parámetros y su incidencia en la merma de la calidad
veracidad de la información transmitida claridad y comprensibilidad del TM corrección gramatical presentación visual y gráfica atractiva estilo usual que no provoca hilaridad
Valor medio pregunta 1
Valor medio pregunta 8
4,75 4,37 3,54 3,37 2,95
4,83 4,62 3,16 3,5 2,33
Expectativas y evaluación en la traducción de folletos turísticos
A pesar de que los valores comparados difieren ligeramente se puede observar que el orden jerárquico en la valoración de los parámetros y el orden a la hora de determinar la incidencia en la merma de calidad es el mismo, con la excepción de los parámetros de la ‘presentación visual y gráfica atractiva’ y de la ‘corrección gramatical’, cuyo orden se invierte (cf. Tabla 3). En relación con la pregunta sobre si los usuarios se sienten molestos cuando a un texto turístico se le nota que se trata de una traducción, se puede observar que sólo el 37,5% opina que sí, frente a un 62,5% que manifestó no sentirse molesto por esta circunstancia. Ni el hecho de que un folleto turístico traducido contenga expresiones que producen cierta hilaridad ni que esté redactado en un estilo poco usual es considerado por los usuarios como un factor que implica una importante merma de calidad. Ello podría indicar que los usuarios de folletos turísticos traducidos asumen que un texto traducido no ha de satisfacer las mismas exigencias en cuanto a la forma que un texto original mientras que consideran que las exigencias en relación con la transmisión del contenido no han de variar. Este tratamiento desigual de aspectos relacionados con el contenido y con la forma del texto traducido se refleja también en el hecho de que los usuarios admiten que un texto traducido puede contener cambios respecto a la forma del TO sin que la calidad del folleto se vea apenas afectada, mientras que opinan que la calidad del TM disminuye considerablemente si el TM contiene cambios respecto al contenido del TO. 4.2 La evaluación de los usuarios reales (cuestionario 2) 4.2.1 La evaluación global de la calidad En la primera pregunta relacionada con la evaluación del texto previamente leído, les pedimos a los sujetos que evalúen de modo global la calidad del TM en una escala del 1 al 5, significando 1= pésimo y 5= excelente. Tabla 4 refleja los resultados obtenidos distinguiendo entre la traducción publicada (Tradpubl) y la traducción manipulada (Tradmanipul), presentando los valores absolutos y la media resultante. Tabla 4.La calidad global de la traducción
Trpubl Trmanipul
1
2
3
4
5
ns/nc
media
0 0
0 1
8 5
3 4
0 2
1 0
3,27 3,58
Podemos observar que ambas traducciones obtienen casi la misma valoración, siendo la de la Tradmanipul ligeramente más alta que la de la Tradpubl.
135
136 Marie-Louise Nobs
4.2.2 La evaluación del grado de profesionalidad del traductor Mediante la pregunta 11B pretendíamos conocer la evaluación del grado de profesionalidad que los sujetos atribuyen a la persona que ha traducido el texto leído. En la Tabla 5 agrupamos los resultados obtenidos, presentando los valores recabados y las medias calculadas para la Tradpubl y la Tradmanipul en una escala del 1 al 5, significando 1 = nada profesional y 5 = extremadamente profesional. Tabla 5.El grado de profesionalidad del traductor
Trpubl Trmanipul
1
2
3
4
5
ns/nc
media
0 0
0 1
6 3
4 3
1 5
1 0
3,54 4,
4.2.3 Discusión Respecto a la evaluación global de la calidad (cf. Tabla 4) llama la atención que ambas traducciones obtengan casi la misma valoración, siendo la de la Tradmanipul ligeramente más alta que la de la Tradpubl. Más allá de las medias matemáticas se pueden hacer algunas observaciones que nos parecen dignas de ser comentadas: ningún sujeto considera que la calidad de la Tradpubl es excelente, mientras que dos encuestados opinan que la Tradmanipul lo es. Es cierto que todos los sujetos sólo opinan sobre la calidad de uno de los dos textos traducidos, por lo que los resultados no pueden reflejar un valor comparativo. Es probable que los resultados hubieran sido distintos si el diseño experimental hubiera posibilitado llevar a cabo una comparación entre las dos traducciones. Es llamativo que los valores obtenidos en relación con la evaluación de la profesionalidad de la persona que ha traducido el texto concreto (cf. Tabla 5) son más altos que los valores referidos a la calidad concreta que los sujetos atribuyen al TM (cf. Tabla 4). En relación con el grado de profesionalidad la valoración media de la Tradpubl se sitúa en 3,54 y la de la Tradmanipul en 4, lo que equivale a una traducción muy profesional. Asimismo podemos constatar que en el caso de la evaluación del grado de profesionalidad del traductor, las diferencias entre la valoración de la Tradpubl y la Tradmanipul son mayores que las diferencias observadas en relación con la evaluación global de la calidad del TM.
Expectativas y evaluación en la traducción de folletos turísticos
5.
Conclusiones
5.1 Conclusiones de tipo metodológico A grandes rasgos hemos podido comprobar la coherencia y la validez de los cuestionarios y del diseño de la investigación en la que se basan. No obstante, la realización de la prueba piloto nos induce a reconsiderar para el diseño de la futura investigación la inclusión de un grupo de control que tendrá acceso a ambas traducciones utilizadas, es decir a la Tradpubl y a la Tradmanipul. 5.1.1 Modificaciones de las herramientas de observación La utilidad más palpable del pretest se reflejará en la adopción de una serie de modificaciones concretas en los dos cuestionarios de la futura investigación para posibilitar una recogida más operativa de los datos y para excluir al máximo valoraciones personales. Con los cambios que proponemos pretendemos conseguir fundamentalmente los siguientes objetivos: a.
evitar contestaciones mediatizadas: entre otros, no nombrando en el encabezamiento de modo explícito el tema de la investigación, suprimiendo los términos ‘calidad’ y ‘traducción’ para posibilitar que las primeras preguntas relacionadas con las expectativas se contesten de modo espontáneo. b. garantizar la comparabilidad entre las preguntas relacionadas con las expectativas y la evaluación concreta: entre otros, incluyendo en el cuestionario 2 una pregunta respecto al parámetro ‘información fehaciente’ y otra para averiguar en qué medida aprueban los sujetos cambios en el TM con respecto al contenido o la forma del TO. c. permitir una valoración más matizada: entre otros, cambiando la escala fija de cinco valores por una escala decimal para evaluar la calidad global de la traducción presentada y el grado de profesionalidad atribuido al traductor del texto. d. permitir la formación de subgrupos significativos entre los sujetos consultados: entre otros, dividiendo las franjas de edad de los sujetos de modo que reflejen con más precisión las diferencias socioculturales que una determinada edad suele llevar consigo. e. averiguar cuales son los principales objetivos que los usuarios atribuyen a un folleto turístico: introduciendo una pregunta específica al principio del cuestionario 1. f. evitar confusión a la hora de atribuir un valor númerico a una determinada evaluación: invirtiendo los valores de la escala del 1 al 5, en el sentido de que, tal como es usual en la mayoría de los países germanohablantes a la hora de calificar un trabajo, servicio o examen el 1 equivalga a excelente y el 5 a muy deficiente.
137
138
Marie-Louise Nobs
5.2 Conclusiones relacionadas con los resultados obtenidos No es fácil que en un estudio piloto de las características del que nos ocupa se puedan observar tendencias claras. Salvaguardando las limitaciones que impone la pequeña muestra de nuestro estudio piloto y teniendo plena conciencia de que todas las tendencias observadas requerirán ser comprobadas posteriormente nos limitaremos a nombrar las siguientes: 1. Los datos recabados indican que los sujetos sólo aprecian una ligera diferencia entre la calidad de la Tradpubl y de la Tradmanipul, a favor de esta última. En principio, ello podría ser la consecuencia del pequeño volumen de la muestra o bien de haber modificado sólo tres de los siete parámetros barajados, lo que podría manifestarse en que los parámetros no manipulados tengan tanta influencia que la calidad de la Tradmanipul no sea percibida como esencialmente mejor que la de la Tradpubl. 2. Los datos indican que, a la hora de nombrar las expectativas respecto a un determinado tipo de folleto turístico, la gran mayoría de los usuarios consultados concede mayor importancia a los parámetros relacionados con el contenido que a los relacionados con la forma del TM. Sin embargo, existen indicios de que a la hora de realizar una evaluación concreta, la forma y con ella todos los elementos no verbales del texto adquieren casi la misma importancia. En caso de confirmarse esta última tendencia obligaría a revisar el papel que tradicionalmente se viene atribuyendo a un traductor profesional. De esta forma, al disponer de una base empírica, sería posible rediseñar un perfil de experto en traducción que incluyese de modo adecuado el tratamiento de los elementos no verbales y tuviese en cuenta que la compleja relación existente entre elementos verbales y no verbales de un texto varía de cultura a cultura. Su estudio podría resultar interesante para futuras investigaciones y de utilidad para traductólogos, formadores de traductores y profesionales de la traducción.
Notes 1. Nos basamos en la definición de ‘texto’ de Schmitt (1998: 147): “Ein Text ist ein thematisch und/oder funktional orientierter, kohärenter Komplex aus verbalen und/oder nonverbalen Zeichen, der eine für den Adressaten erkennbare kommunikative Funktion erfüllt und eine inhaltlich funktional abgeschlossene Einheit bildet.” 2. En las Ciencias Sociales se habla de “grupos de discusión” como técnica cualitativa de obtención de datos frente a la técnica cuantitativa de la “encuesta” (cf. Bericat, 1998). 3. Una presentación exhaustiva de los resultados se encuentra en Nobs (2001).
Expectativas y evaluación en la traducción de folletos turísticos 139
References Bericat, Eduardo. 1998. La integración de los métodos cuantitativo y cualitativo en la investigación social. Barcelona: Ariel. Brunette, Louise. 2000. “Towards a terminology for translation quality assessment. A comparison of TQA practices”. The Translator 2: 169–82. Collados Aís, Angela. 1998. La evaluación de la calidad en Interpretación simultánea. La importancia de la comunicación no verbal. Granada: Comares Interlingua. Diekmann, Andreas. 1998. Empirische Sozialforschung. Grundlagen, Methoden, Anwendungen. Reinbek bei Hamburg: Rowohlt. Gile, Daniel. 1990. “L’évaluation de la qualité de l’interprétation par les délégués: une étude de cas”. In The Interpreter’s Newsletter 3: 66–71. Hatim, Basil y Mason, Ian. 1995. Teoría de la traducción. Una aproximación al discurso. Trad. Peña, Salvador. Barcelona: Ariel. Holz-Mänttäri, Justa. 1984. Translatorisches Handeln. Theorie und Methode [Annales Academiae Scientiarum Fennicae B226]. Helsinki: Suomalainen Tiedeakatemia. Holz-Mänttäri, Justa. 1993. “Textdesign — verantwortlich und gehirngerecht”. In Traducere Navem. Festschrift für Katharina Reiss zum 70. Geburtstag, Holz-Mänttäri, Justa & Nord, Christiane (eds), 301–20.Tampere: University. Hönig, Hans G. 1989. “Die übersetzerrelevante Textanalyse”. In Übersetzungswissenschaft und Fremdsprachenunterricht. Neue Beiträge zu einem alten Thema, Königs, Frank G.(ed), 121–45. München: Goethe-Institut, Ref.42. Arbeitsstelle für wissenschaftliche Didaktik. Hönig, Hans G. 1995. Konstruktives Übersetzen. Tübingen: Stauffenburg. Hönig, Hans G. 1998. “Positions, power and practice: Functionalist approaches and translation quality assessment”. In Translation and Quality, 6–34. Clevedon: Multilingual Matters. Jänis, Marja y Priiki, Timo. 1993. “User satisfaction with translated tourist brochures: The response of tourists from the Sovjet Union to Russian translations of Finnish tourist brochures”. In Translation and the (Re)production of Culture. Selected papers of the CETRA Research Seminars in Translation Studies 1989–1991, 49–56. Leuven: University. Larose, Robert. 1998. “Méthodologie de l’évaluation des traductions”. Meta XLIII, 2: 163–86. Latiesa, Margarita. 2000. Granada y el turismo. Granada: Editorial Universidad de Granada. Nobs, Marie-Louise. 2001. Expectativas y evaluación en la traducción de folletos turísticos: estudio piloto. Trabajo de investigación presentado en el Departamento de Traducción e Interpretación de la Universidad de Granada. Granada: sin publicar. Schmitt, Peter A. 1998. “Defekte im Ausgangstext”. In Handbuch Translation, Snell-Hornby, Mary et al.(eds), 147–51. Tübingen: Stauffenburg.
Censorship or error Mary Howitt and a problem in descriptive TS Kirsten Malmkjær Centre for Research in Translation, Middlesex University, London, UK
This article argues that the otherwise praiseworthy liberalism of Target-Textoriented Descriptive Translation Studies makes it difficult to justify any distinction we might like to draw for e.g. pedagogical or criticism-related reasons between choice and error in translation. It proposes that such a distinction may be drawn on the basis of a (linguistically oriented) distinction between semantic patterning in the case of the former and formal (graphemic/phonemic) patterning in the case of the latter. Since it is necessary to pre-establish the distinction between error and choice in translation on nonlinguistic grounds if we are to argue for informative concomitance between it and the two types of linguistic patterning, the linguistic distinction is illustrated with reference to a set of texts in the case of which a strong argument for distinguishing choice from error can be made on (non-linguistic) historico-sociological grounds.
Introduction In Descriptive Translation Studies after about 1980, translational equivalence is usually considered to be realised by the relationships between texts (or text parts) which we think of as related to one another as Target Text and Source Text. Observation of these relationships proceeds from Target Text to Source Text, and explanations for characteristics of the equivalence relationships are generally provided in terms of translational and other norms that govern choices of translational equivalents within the receiving Target culture (Toury 1980/1981; 1995: 61). Reading translations and their source texts from this point of view, it is constantly surprising and delightful to discover the range of items that can serve as equivalents of certain other items (and also to discover how often items that we might have expected to function as equivalents of certain other items do not, in fact, function thus). Still, from time to time, wonderment gives way to unease. Some selections may
142 Kirsten Malmkjær
seem in some way not quite right: not, perhaps, either exciting, innovative, unusual or unfortunate, but simply wrong; and the TT-oriented philosophy makes it difficult to justify such judgements on theoretical grounds, because it is always possible, in theory, that the translator was obeying some norm or principle, however eccentric, or had some reason, however peculiar, guiding their choice. It is bothersome for intra-theoretical reasons if there is data which cannot be accounted for within an otherwise satisfying theory. And it is obviously also important to be able to distinguish intentional, motivated choice from unintentional, caused error in the applied branches of our discipline, such as e.g. translation criticism and translation pedagogy. An ability to justify their criticisms is vital for educators and critics; and for trainees and more experienced translators alike, it is extremely helpful to have confidence in the distinction between error and motivated choice, so that they can defend instances of the latter against perceptions of them as examples of the former. To err is incompetent whereas the worst that can be said of motivated choices is that they are objectionable if made without due consideration of ethical issues. In this article, I seek to illustrate the difference between errors and motivated choice through a case study of translations made by Mary Howitt of a selection of Hans Christian Andersen’s stories. This set of translations is helpful to my purpose because there is good reason to suppose that some of those of Howitt’s choices of translational equivalents which appear surprising on semantic grounds are motivated by considerations of suitability for her perceived readership, whereas others remain inexplicable on these grounds, but display characteristics which suggest influences from other languages together with an incomplete comprehension of Danish, the language of the Source Texts.
Error, motivated choices and manipulation I suppose that a translational-equivalent choice appears aberrant either when the semantic field from which the Target Text item is drawn seems to the analyst different from that to which the Source Text item belongs, or when a semantic field that can encompass both terms has to be conceived of at a very abstract level (which may ultimately be the same thing since at a sufficiently abstract level, everything resembles everything else). For example, Mary Howitt (1846: 4–5) gives ‘doormat’ as a translational equivalent for a term that means ‘spittoon’, and we would probably not think to classify doormats and spittoons as items falling into one semantic field normally. Again, a number of translators of Hans Christian Andersen’s story known in English as “The Steadfast Tin Soldier” translate a term that means ‘borne’ as ‘made’ (see Malmkjær 1998), and we would need to posit the fairly abstract semantic field of “coming into being” to encompass both of these.
Censorship or error 143
Studies such as those reported in e.g. Aijmer et al. (1996), Johansson and Oksefjell (1998) and Meta (1998: 43/4), show that translators tend not to stray very often beyond the semantic fields, fairly narrowly conceived, of their Source Text items in their Target Texts: so rarely, in fact, that these aberrant choices are statistically virtually non-existent. This, however, does not mean that such choices are theoretically and pedagogically insignificant — far from it. In many cases, they are exactly the kinds of choice that make a text or text part significant, or which contribute particularly significantly to the individuality of a text. When translators stray in their translations beyond the semantic fields of their Source Texts in selecting translational equivalents, it is sometimes possible to ask them for their reasons, which they may give; or the reasons can be surmised, especially if we have some historical, cultural and even personal knowledge about and relevant to the translators and their times. In this respect, motivated choices differ from errors, where by ‘error’, I mean the kind of choice that the translator believes is correct but which is, in Pym’s (1992) terminology, simply wrong in a ‘binary’ (yes-no) way. Pym contrasts binary errors with choices which may be considered more or less suitable than other choices for a variety of reasons but about which there can be legitimate argument. Simple mistakes are less of a worry, because they will be spotted and corrected given good (self-)editing, and I shall have no more to say about them. Usually, motivated choices are made by translators with finely tuned awarenesses of translational and other norms and in the interest of the success of the translation as such, or of its Skopos (Vermeer 1986), or of its projected audience. Sometimes, such choices are motivated by a set of well-articulated principles and beliefs and in such cases, we are justified, I think, in speaking of manipulation on the part of the translator. This sense of manipulation is also worth having a theoretical grip on: any carefully considered choice a translator makes may be motivated by considerations that not everyone would agree with, but because translational manipulation, as defined here, is informed by particularly well articulated beliefs and principles, it can have effects on and through translations that are more far-reaching and better tried, tested and planned for than other motivated choice types. Between the two extremes of error and manipulation are many cases of translator “interference”, which might be explained by attributing various “single” motivations to the translator. For example, reluctance to cause her readership distress may have moved Howitt (1846: 38) not to include a translation of the following indirect thought extract: dersom han nu ikke kunde komme løs, maatte han jo sulte ihjel, (‘if he could not get loose, he must starve to death’ (my translation)) in her translation of the story usually known in English as “Thumbelina”. It occurs at a point where Thumbelina has been carried off from a leaf on which she had been floating along a stream, pulled by a butterfly tied fast to the leaf; it is likely that the
144 Kirsten Malmkjær
butterfly will not be able to escape. A wish to help her readership understand the concept of a faldhat (a form of protective headgear) is likely to have moved Howitt to translate han havde en Faldhat paa (he was wearing a Faldhat) as ‘he had one of those pads round his head, which children used to wear formerly’ (Howitt 1846: 71).
Manipulation, motivation and textual patterning As mentioned in the previous section, I understand manipulation in translation to be instantiated in translational choices that are motivated by sets of principles or beliefs which a translator could articulate fairly precisely if asked. Because of the very close relationship (not to be discussed further here) between principles and beliefs on the one hand and linguistic expression on the other, it is almost inevitable that translational choices made with reference to sets of beliefs and principles will result in identifiable patterns in the relationships between Source and Target Texts, given a certain volume of these. In contrast, although large numbers of translational choices within a translation or set of translations might be guided by considerations such as those I attributed to Mary Howitt in the previous section (wishing to prevent reader distress and incomprehension) or by e.g. a wish to produce an aesthetically pleasing TT, it is unlikely that those choices would produce perceptible patterns in the relationships between the texts, because it is not likely that one particular set of strategies for choosing Target Text equivalents for a specific, identifiable set of Source-Text features would achieve the desired outcome; if it did, translation of e.g. literature would be easier than it is reputed to be (see e.g. Bush 1997; 1998).
The data The data I have chosen in order to try to illustrate the distinction that I wish to draw between (binary) error and manipulative choices are Mary Howitt’s translations of nine of Hans Christian Andersen’s stories plus one of the stories within the longer story, “The Flying Trunk”. Mary Howitt (1799–1888) had already translated Andersen’s three first novels (details at the end of this article) before she embarked on the stories, and she used the fact of his authorship of those novels to advertise the stories, her title page proudly presenting them as: Wonderful Stories for Children. By Hans Christian Anderson [sic], Author of ‘The Improvisatore’ etc. But whereas the novels were translated from German translations, Howitt claimed to have translated the stories from Danish. Bredsdorff (1954: 489–491) berates Howitt for her imperfect knowledge of Danish, and lists around forty errors in the collection. He also points to omissions
Censorship or error
and additions, which he ascribes to Howitt’s puritan morals (1954: 492–494), and he considers Howitt’s translation to be so poor that (1954: 494, my translation): ‘if we only had Mary Howitt’s translation of the fairy tales, Hans Christian Andersen would be unreadable in English today!’ My own search through Howitt’s collection confirms and extends Bredsdorff’s findings: there is no doubt that Howitt errs prolifically, though as Hjørnager-Pedersen (1993) points out, not all the errors are equally serious in the sense of altering the sense of the Source Text, and few would be noticed by readers of the Target Texts alone. Furthermore, it is possible to consider the omissions, additions and substitutions in Howitt’s collection (whether we approve of them or not) to be evidence of excellent censorial abilities, exercised in order to produce a set of stories in a form suitable for the Victorian middle-class children whom she perceived, or so her title suggests, as her main audience. For these reasons, Howitt’s translations of the tales (1846) are useful data for the kind of study I want to carry out. In addition, the personal and socio-historical background of the translator (and author) is well documented, and Howitt’s translations of Andersen’s early stories are the first in English, so they cannot have been influenced by earlier translations of these exact texts into English. It is a complete corpus of one translator’s treatment of one genre by one author, produced over a fairly brief stretch of time, and it is small enough to analyse in some depth in its entirety. Furthermore, the socio-cultural contexts that surrounded Andersen and Howitt are sufficiently distant from the present for us to take a relatively objective view of them, yet not so distant as to be incomprehensible to all but the specialist. The two contexts differ sufficiently from one another for the translator to have felt a need to engage in significant manipulation, but the differences remain contained within shared parameters of understanding of concepts such as e.g. religion, genre, society, family, and education; the two contexts are not so different that intra-parametric comparison must give way to inter-parametric contrast, obscuring the clean line I would like to draw between the notions of caused error and intentional manipulation.
Data analysis Howitt’s title suggests that she intended to produce a set of texts that would be suitable for a (Victorian) child audience, so it is reasonable to assume that some of her translational choices were motivated by beliefs about what kind of language and what kinds of story were suitable for such an audience. Her collection was generally well received by the press (see Bredsdorff 1954: 438–447), and one reviewer expresses the kind of effect that Howitt probably sought to achieve especially eloquently, praising Andersen’s “graceful wit, and delicate humour, and sweet, naïve kindness” and describing him as “one of those friends whom Heaven has sent
145
146 Kirsten Malmkjær
us — those sweet, Christian messengers of peace and goodwill’ (Titmarsh 1847 quoted in Bredsdorff 1954: 446). Clearly there were a number of Andersen stories available to Howitt to translate, had she wanted to, which are incompatible with this description of their author; and before looking at the translations as translations, it is interesting to look at Howitt’s selection of stories to translate. Below is a list of stories published by Andersen by 1842, the latest date for any story included in Howitt’s selection, side by side with a list of Howitt’s corresponding titles: Andersen 1835 Fyrtøiet [The Tinderbox] Lille Claus og store Claus [Little Claus and Big Claus] Prindsessen paa Ærten [The Princess on the Pea] Den lille Idas Blomster Tommelise [Thumbelina] Den uartige Dreng Reisekammeraten [The Travelling Companion] 1837 Den lille Havfrue [The Little Mermaid] Keiseren’s nye Klæder [The Emperor’s New Clothes] 1838 ‘Gaaseurten’ ‘Den standhaftige Tinsoldat’ ‘De vilde Svaner’ [The Wild Swans] 1839 ‘Paradisets Have’ ‘Den flyvende Koffert’ (“Svovlstikkerne”) ‘Storkene’ 1842 ‘Ole Lukøie’ ‘Rosenalfen’ ‘Svinedrengen’ [The Swine-boy] ‘Boghveden’ [The Buckwheat]
Howitt 1846
“Little Ida’s Flowers” “Tommelise” “The Naughty Boy”
‘The Daisy’ ‘The Constant Tin Soldier’
‘The Garden of Paradise’ ‘A Night in the Kitchen’ ‘The Storks’ ‘Ole Lukoie — The Story-Teller at Night’ ‘The Rose-elf ’
Howitt’s collection does not include any of Andersen’s indulgences in farce, bawdy, revelry, or story telling for its own sake or for mere entertainment (as Howitt might
Censorship or error 147
have interpreted e.g. “The Tinderbox”, “Little Claus and Big Claus” or “The Princess on the Pea”). Instead, she selects two types of tale, the sweetly entertaining and the morally edifying. But even these, especially examples of the morally edifying kind, contain a good deal of what Howitt might well consider inappropriate for her collection of Wonderful Stories for Children, and it is possible to group her obvious manipulations of this material into five categories on the basis of TT-ST relationship patterning and to identify probable motivations for the relevant translational choices.
Patterns of manipulation Type 1: Avoidance of the repulsive by omission (major strategy) or substitution (minor strategy) The first type of manipulation takes the form of substitution or omission of any ST word or expression denoting objects, concepts or events that might appear repulsive, in particular, anything touching on death and putrefaction, which is usually omitted, though also including the replacement of a spittoon with a doormat. The list below includes three of around thirteen textual instances of this strategy, though it is difficult to quantify data of this kind accurately because of the interrelationships between examples from the same story or sections of stories. For instance, the examples below are mutually dependent in that once the Source Text murderer’s action of severing his victim’s head from its body has been omitted, it is not open to the young lady who disinters her dead lover in the Target Text to do more than take home a lock of his hair. Numbers in brackets refer to pages of the Howitt collection.
Examples st gloss tt
skar hans Hoved af og begravede det med Kroppen i den bløde Jord cut his head off and buried it with the body in the soft earth ‘and then buried him in the bloody earth’ (58)
st gloss st gloss tt
saa tog hun det blege Hoved med de lukkede Øine, kyssede den then took she the pale head with the closed eyes kissed the kolde Mund og rystede Jorden af hans deilige Haar cold mouth and shook the earth off his lovely hair ‘so she cut away a beautiful lock of his hair, and laid it near her heart’ (60)
st gloss tt
tog hun Hovedet med sig hjem took she the head with her home ‘she went home’ (60)
148 Kirsten Malmkjær
Type 2: Avoidance of blasphemy by omission and substitution Patterning in this category is especially clear because it regularly involves omission of vocabulary that refers to the inhabitants of the metaphysical spheres evoked in the texts, or replacement of it by vocabulary with more secular connotations. I reproduce three of nine examples I have identified:
Examples st gloss tt
Gud, hvor den stakkels Tommelise blev forskrækket, God how the poor Thumbelina became frightened ‘Poor Tommelise! How frightened she was’ (38)
st gloss tt
Tak Du Gud for ham! Thank you God for him ‘Be thankful that thou canst get such a one!’ (49)
st gloss tt
at de aldrig vilde døe, og at Paradisets Have skulde evig blomstre that they never would die and that Paradise’s garden should ever bloom that they who were purified by trial should never die, and that the Garden of Paradise for them should bloom for ever! (86)
Type 3: Avoidance of associations with sexuality by omission and substitution Again, patterning in this category is clear because it regularly involves omission of words and phrases denoting body parts, acts of intimacy, or actions that might provoke sexually charged thinking. I reproduce one of eleven examples, many of considerable length: st gloss st gloss tt
De skjønneste Piger, svævende og slanke, klædte i bølgende Flor, The most beautiful girls, floating and slim, dressed in waving gauze, saa man saae de deilige Lemmer, svævede i Dandse, so one saw the lovely limbs floated in dances ‘The most beautiful maidens floated in the dance’ (86)
Type 4: Avoidance of improper sentiments/emotions (other than sexuality, e.g. pride; vengeance) by substitution and omission I reproduce extracts from the story of the storks, a large section of which concerns, in the original, revenge, instigated by the young storks but supported and indeed executed by their mother, in which the storks take on some boys who have teased them; and from Howitt’s translation, in which the young storks meekly comply with their mother’s advice to let revenge give way to forgiveness:
Censorship or error 149
st gloss st gloss tt
og altsom de blev større, vilde de mindre taale det; Moderen and as they grew bigger would they less bear it the mother maatte tilsidst love dem, at de nok skulde faa Hævn must finally promise them that they part should get revenge ‘and the more they were determined on revenge, the less they said of it to their mother. Their mother, they thought, would at last grant their wishes’ (125–126)
st gloss st gloss st gloss tt
“Nu skulle vi hævnes!” sagde de. “now shall we be revenged said they “Ja vist!” sagde Storkemoderen. “Hvad jeg have udtænkt, “yes certainly said the stork mother “what I have planned det er just det rigtige!” that is just the right (thing) “Now let’s have revenge,” said they. “Leave off talking of revenge,” said the mother. “Listen to me, which is a great deal better. (126)
Discussion I think it is clear that 1. These additions, omissions and substitutions can be explained with reference to Howitt’s over-riding desire to produce a set of stories corresponding to what she considers the norm for literature for children. 2. There are no formal patterns involved, only semantic patterns (mainly in lexis and phrasing) determined by the motivation referred to in (1).
Error identification The unifying characteristic of motivated choices, including manipulation, is the fact that they can be explained with reference to semantic notions such as intention and aboutness; granted that the explanation may be incorrect or correct, at least the examples invite or even invoke an explanation in terms of translator motivation. In contrast, one error-identifying factor is apparent inexplicability. Of course it is always possible to discover later that there is an explanation for a translational choice, and that what we thought was an error was actually a motivated choice, but in the present data, I can think of no plausible motivation for the two translations in the following section:
150 Kirsten Malmkjær
Type 1: Inexplicability st gloss tt
Strudsen the ostrich ‘the simoom’ (9)
st gloss tt
Bajonetten nede imellem Brostenene the bayonet down between the cobble stones His bayonet down among the stones of a sink (114)
Other error types are indeed explicable, but never with reference to intention and only with reference to formal features of the expressions affected. The first type of this kind is caused by relationships between ST and TT terms which we usually think of as False Friends:
Type 2: False friends The polyglot Howitt finds false friends for her Danish, not only in English, but also in Swedish and German, and translates the false friends: st gloss tt
han vil bare have at de skulle være rolige he will only have that they shall be still ‘he only wants to amuse them’ (1)
st gloss tt
see bare, hvor rolig Jer Fader staaer see only how still your father stands ‘Look how funnily your father stands’ (121)
(In Swedish, rolig means, roughly, amusing; in Danish it means, roughly, still/quiet/ peaceful). st gloss tt
over Isen across the ice ‘over the iron’ (70)
(In German, Eisen means ‘iron’; in Danish, isen means ‘the ice’) st gloss tt
gjorde Cuur til courted ‘cured’ (18)
st gloss tt
det kan man saa rart that can one so nicely ‘people can do that so seldom’ (93)
Other errors look as if they are caused by graphemic similarities that create relationships that are less powerful than the relationships between false friends are; some in the list below suggest that Howitt may have worked from an unclear source
Censorship or error
such as a handwritten copy, for example. In some cases, the Danish term looks like an English term (marken [the field] — market); in others, the Danish term looks like another Danish term with the meaning Howitt has selected (talerknerne [the plates] — talerne — ‘the talkers’; torvekurven [the shopping basket] — tørvekurven — ‘the peat basket’;1 Kaalbøtter [somersaults] which becomes something to do with nøtter [Swedish for ‘nuts’] and ends up as ‘gingerbread nuts’); in yet others, the Danish term resembles an English term for which Howitt has selected another term that relates to the latter in meaning (træk [draft] — ‘treck’ — ‘travel’). The Taage [fog] — ‘ceiling’ example below is a combination of the last two categories, via Tage [roofs]’:
Type 3: Misleading graphemics/phonemics/morphemics st gloss tt
Næste Nat listede hun sig igjen ned til den next night tip-toed she reflexive again down to it ‘Next night she listened again’ (45)
This is one of the errors that Bredsdorff lists in evidence of Howitt’s poor Danish skills. There are, however, two cases in which she translates lister/listede respectively appropriately as ‘steals softly’ (p. 1) and ‘stole’ (p. 48) and another instance in which ‘listened’ (p. 104) appropriately translates lyttede. Hjørnager-Pedersen (1993) draws attention to the financial stress the Howitts lived under at the time when Mary was translating the stories and the consequent need for speed, and this may have caused the lapse of concentration on page 45. However, as it is not my intention here to judge Howitt’s Danish skills, I will not discuss further “inconsistencies” of this type in detail. st gloss tt
jeg har kjørt i en Bøtte I have driven in a pot ‘I have gone in a boat’ (16)
st gloss tt
jeg skal ud paa Marken I shall out on the field ‘I shall go out into the market’ (17)
st gloss tt
Den bløde Jord the soft earth ‘the bloody earth’ (58)
st gloss tt
Græsset er blødt the grass is soft ‘the grass is in flower’ (23)
It is possible that the above is an influence from German, where blühen means ‘to be in bloom’.
151
152
Kirsten Malmkjær
st gloss tt
Her er en slem Træk! here is a bad draft ‘It is bad travelling to-night,’ (67)
st gloss tt
(en) Taage (a) fog ‘ceiling’ (80)
st gloss tt
sagde alle Talerknerne said all the plates said all the talkers (93; also on p. 94)
st gloss tt
Torvekurven the (market)square basket the coal-box (92; also twice on p. 95)
st gloss tt
Adolph og Jonas skjød med Flitsbuer over Graven Adolph and Jonas shot with bows over the grave ‘Jonas and Adolph stood with their cross-bows above the grave’ (110)
st gloss tt
Nøddeknækkeren slog Kaalbøtter, the nutcrackers did (literally hit) somersaults ‘The nutcrackers knocked about the gingerbread nuts’ (113)
st gloss tt
at han ikke kunde standse. that he not could stop ‘That he could no longer stand upright’ (116)
st gloss tt
kukkelure mope about ‘cough’ (123)
st gloss tt
nu flyve vi høire om now fly we (to the) right around ‘Now we fly round, higher than ever’ (124)
Type 4: Mis-calques Mistaken calques tend to affect compounds and phrasal verbs, though ‘summer season’ below derives from a misreading of sommetider (‘sometimes’) as sommertider (‘summer-times’), that is, it includes misleading graphemics: st gloss tt
Sommerfuglene the butterflies ‘the summer birds’ (6)
Note that this term is appropriately translated on four subsequent occasions.
Censorship or error
tt
da vendte hans Tanker tilbage. then turned his thoughts back [the phrasal verb vende tilbage means ‘return’, so the sense is that his thoughts returned] ‘His thoughts turned to the past’ (88)
st gloss tt
“Sommetider, om (Natten, “sometimes at (the) night ‘In the summer season at night’ (99)
st gloss tt
“Nei lad være med det!” “no let be with that [meaning ‘do not do that’] ‘No, let them be’ (124)
st gloss
Type 5: Homonym selection Some of the examples below seem to be selections of the wrong sense of a homonym, though other factors, such as misunderstanding of syntax or phrasal verb status are also at play. st gloss tt
det skal være saadan en deilig Leilighed!” it shall be such a lovely flat ‘it will be such a nice opportunity (10)
st gloss tt
Svalen the swallow the breeze (15; three occurrences)
st gloss tt
“Nu tør man nok ikke mere sige sin Mening!” “now dare one particle not more say one’s meaning ‘“Now, one cannot tell what he means”’ (18)
st gloss tt
Muldvarpen stødte til den the mole knocked (in)to it ‘the mole stood beside it’ (43)
st gloss tt
der ville vi nyde Forfriskninger there will we enjoy refreshments ‘it will be refreshing to us’ (83)
st gloss tt
det syntes ogsaa, at det hørte med til Blomsterne it thought also that it belonged with to the flowers [the phrasal verb, høre med til means ‘belong with/be one of ’; høre alone means ‘hear’] which had heard the flowers (105)
st gloss tt
Ansigtet satte de lige ud; the face set they straight ahead ‘their faces were all the same’ (111)
153
154
Kirsten Malmkjær
st gloss tt
de vilde være med, they would be with [the phrasal verb være med means ‘participate’] ‘they wanted to be with the rest of the things’ (113)
Type 6: Grammar and syntax Many of the errors in the Howitt corpus involve misreading — often multiple misreadings — of more or less complex syntactic relationships, often aided and abetted by misunderstandings of individual terms as in e.g. the second example below, where Howitt misunderstands kunde to mean ‘knew’, possibly confused by German kannte, possibly by Danish kendte. These examples are taken from a list of twenty one. st gloss st gloss tt
Træerne fortalte om Røvere og Hexe og Blomsterne om de nydelige the trees told of robbers and witches and the flowers of the pretty smaa Alfer og hvad Sommerfuglene havde fortalt dem. small elves and what the butterflies had told them ‘the trees talking about robbers, and witches, and flowers, and the pretty little fairies, and all that the summer birds had told them of. (5–6)
st gloss st gloss tt
da kunde Blomsten ikke, som Aftenen forud, folde sine Blade then could the flower not like the evening before fold its petals sammen og sove, den hang syg og sørgende ned mod Jorden together and sleep it hung ill and grieving down towards the earth The little flower knew this not; before the evening was ended, it had folded its petals together and slept upon the earth, overcome with sickness and sorrow. (27)
Discussion Some errors appear inexplicable. The majority seem to be caused by formal characteristics of the language items involved, whether graphemic, morphemic or syntactic. They are never functionally or thematically motivated, because of course they are not motivated, but, rather, incidental results of what the translator happens not to know about the languages involved.
Note 1. I am grateful to Henrik Gottlieb, Viggo Hjørnager-Pedersen and other Danish-speaking members of the audience for the spoken version of this paper for pointing out to me the source of this error.
Censorship or error
References Aijmer, Karin, Altenberg, Bengt and Johansson, Mats (eds). 1996. Languages in Contrast: Papers from a Symposium on Text-Based Cross-Linguistic Studies, Lund 4–5 March 1994. Lund: Lund University Press. Bredsdorff, Elias. 1954. H. C. Andersen og England. Copenhagen: Rosenkilde og Baggers Forlag. Bush, Peter. 1997. “The translator as reader and writer”. Donaire 8: 13–18. Bush, Peter. 1998. “Literary translation: Practices”. In Routledge Encyclopedia of Translation Studies, Mona Baker (ed.), 127–130. London and New York: Routledge. Hjørnager-Pederson, Viggo. 1993. “Mary Howitt’s translations of Hans Christian Andersen”. In Proceedings of the Fifth Nordic Conference on English Studies, Julian Meldon D’Arcy (ed.), 286–296. Reykjavik: University of Iceland Publishing Co. Johansson, Stig and Oksefjell, Signe (eds). 1998. Corpora and Cross-linguistic Research: Theory, Method and Case Studies. Amsterdam and Atlanta GA: Rodopi. Malmkjær, Kirsten. 1998. “Love thy neighbour: Will parallel corpora endear linguists to translators?”. Meta 43 (4): 534–541. Pym, Anthony. 1992. “Translation error analysis and the interface with language teaching”. In Teaching Translation and Interpreting: Training, Talent and Experience, Cay Dollerup and Anne Loddegaard (eds), 279–288. Amsterdam and Philadelphia: John Benjamins. Titmarsh, Michael Angelo. 1847. “A grumble about the Christmas-Books”. Fraser’s Magazine January. Toury, Gideon. 1980/1981. “Translated literature: System, norms, performance: Toward a TT-oriented approach to literary translation”. In In Search of a Theory of Translation, 35–50. Tel Aviv: Porter Institute. Reprinted in Poetics Today 2(4): 9–27. Toury, Gideon. 1995. Descriptive Translation Studies and beyond. Amsterdam: John Benjamins. Vermeer, Hans J. 1986. Voraussetzungen für eine Translationstheorie: Einige Kapitel Kultur- und Sprachtheorie. Heidelberg: Selbstverlag.
Sources Andersen: H. C. Andersens Eventyr: Kritisk udgivet efter de originale Eventyrhæfter med Varianter ved Erik Dal og Kommentar ved Erling Nielsen. I:1835–42, 1963; II:1843–55, 1964. Copenhagen: Hans Reitzels Forlag. Improvisatoren 1835 Copenhagen: C. A. Reitzel. O. T. 1836 Copenhagen: C. A. Reitzel Kun en Spillemand 1837 Copenhagen: C. A. Reitzel Howitt: Wonderful Stories for Children. By Hans Christian Anderson [sic], Author of ‘The Improvisatore’ etc. 1846 Mary Howitt, London: Chapman and Hall. The Improvisatore, or, Life in Italy 1845 London: Richard Bentley, based on German. Only a Fiddler; and O. T., or Life in Denmark. By the Author of the Improvisatore, or, Life in Italy 1845 London: Richard Bentley.
155
Of holy goats and the NYPD A study of language-based screen humour in translation Thorsten Schröter Karlstad University, Sweden
1.
Language-play on screen and its translation
Humour, especially when based on language-play, is known to pose great difficulties for the translator. This is true for all kinds of media, but the screen translator faces additional problems that are due to the special nature of the texts s/he has to produce. While both dubbing and subtitling can usually be said to render the spoken aspects of screen productions more or less adequately, especially in view of their medium-specific limitations, it is interesting to see how the challenge of reputedly ‘untranslatable’ play with linguistic elements, which forms an essential part of many humorous films and shows, has been addressed. The results of an empirical study described below will suggest a partial answer to this question. The issue of screen translation of language-play has already been addressed a few times, e.g. by Zabalbeascoa (1994, 1996), who explores it within the framework of dubbing, and by Pisek (1997), who touches on both dubbing and subtitling, but quantitative comparisons are hard to find: Gottlieb (1997) is an exception in that he counts the instances of wordplay in both the SL and the TL version of the subtitled TV show analyzed by him. And Whitman-Linsen (1992: 301–314) is one of few researchers who deal with the translation of film humour into more than one language simultaneously. In the present paper, both dubbed and subtitled versions, as well as several target languages, will be included, and the objective is to make quantitative comparisons between the different versions. Comic effect through language-play can be achieved by means of a large array of different strategies, ranging from simple alliteration and rhyme to puns and manipulations of the grammar. Even whole series of gags building onto each other, as well as combinations of several identifiable strategies helping to form one joke, are common. These strategies can be categorized, which makes it possible to abandon purely subjective judgements about the degree of ‘funniness’ in the source- and target-language versions of a certain passage in favour of more
158
Thorsten Schröter
objective comparisons based on the type and number of the contributing elements. From this procedure, conclusions can be drawn concerning the performance of dubbers and subtitlers when it comes to dealing with language-based screen humour. As indicated, this study deals only with the quantitative side of that performance. While the use and transfer of some strategies will be exemplified and discussed in this paper, the goal has not been to offer a finished and comprehensive model for the detailed analysis of language-play, but rather to gain a first impression of how such language-play has been treated in screen translation in terms of numbers. (For comprehensive discussions of possible categorizations of wordplay, which to a large extent coincides with what has been labelled language-play here, the reader may be referred to Delabastita (1993) and Heibert (1993).) On the basis of the data presented below, qualitative judgements can only be made to the extent that quantity may be an indicator of quality.
2.
Material
For the study described here, seven American and British film comedies have been examined in a search for instances of language-based humour. These films are Analyze This (1999), Four Weddings and a Funeral (1994), Men in Black (1997), Mickey Blue Eyes (1999), Road Trip (2000), The Nutty Professor I (1996), and The Nutty Professor II (2000). When language-play could be identified in the original, it was first analyzed in terms of contributing strategies and then compared to different “official” translations, i.e. translations that have been broadcast and/or been added to the interactive menu of a DVD. As pointed out above, the focus was on the number of strategies in the translations as compared to the original. For all films, the German dubbed Table 1.Translations included in the study Film
German German audio subtitles (DVD)
Analyze… Four Wed. MiB Mickey… Road Trip Nutty P. I Nutty P. II
x x x x x x x
Swedish subtitles (TV)
x x x x x
x x
Swedish subtitles (DVD)
Danish Norwegian French subtitles subtitles subtitles (DVD) (DVD) (DVD)
x
x
x
x
x x x x x
x x x x x
x x x x x
x x x x
Language-based screen humour in translation 159
version and at least one Swedish subtitle version (TV or DVD) were among the translations covered. In the case of Men in Black, both of the Swedish subtitle versions formed part of the corpus. For six of the seven films, Danish and Norwegian DVD captions could be included in the analysis. A German subtitle version (DVD) was available for five of the films and French DVD subtitles for four. See Table 1 for the exact distribution.
3.
Strategies
In the films studied, language-play depending on the following broad strategies could be identified: –
– – – – – – – –
multiple interpretations of words, expressions, clauses etc, supported by – homonymy (complete or “near-”) – homophony (complete or “near-”) – polysemy, extended meanings (e.g. literal vs. figurative) rhyme alliteration nonce-combinations of words and morphemes creative use of established words deviant pronunciation reinterpretation of abbreviation exploitation of spelling-pronunciation discrepancy play with grammar
The list is not exhaustive, nor are the different items on it mutually exclusive. It is not always easy to distinguish what has been referred to here as language-based humour, i.e. the creative play with form, meaning and grammar, from other types of humour that also rely on language rather than on actions or pictures. Many things people say or write can be amusing without qualifying as language-play, and a grey area certainly exists. Conversely, people may employ some form of language-play not to cause roaring laughter, but in order to appear witty or simply to attract attention. This use is often exemplified in the headings of newspaper and magazine articles, but it can be found in films as well. In the study, all instances of language-play were included in the analysis, regardless of their potential for the generation of laughter. But let us look at what must be considered clear examples of language-play. These will help to illustrate the notion of strategy and show that determining the type and number of strategies is not always straightforward.
160 Thorsten Schröter
3.1 Example 1: Rhyme The first example is from the movie Four Weddings and a Funeral: (1) At the last moment, Charles (played by Hugh Grant) arrives at a church wedding. As he is the groom’s best man, he is very stressed. A friend of his, Fiona (Kristin Scott Thomas), is already seated in a bench. Fiona: There’s a greatness to your lateness. Charles: Thanks. It’s not achieved without suffering. German dubbing: Dein Zuspätkommen hat eine gewisse Grösse. [ª‘There’s a greatness to your lateness.’] Danke. Dafür hab’ ich auch schwer gelitten. [ª‘Thank you. I’ve been suffering a lot for it.’] Swedish TV subtitles: -Du är storslaget sen. [ª‘You are magnificently late.’] -Inte utan att lida pin. [ª‘Not without suffering.’] We disregard the irony in Fiona’s comment and in Charles’s answer. Though it stems from the choice of words (how could it be otherwise?), the irony does not depend on any wordplay or similar manipulation of the language. We can see, then, that only one identifiable strategy that would fall under language-play in the sense used here has been employed to make Fiona’s statement funnier, or at least slightly witty, namely rhyme. In Charles’s reply, no such strategy has been used. While both translations render the content of the brief exchange adequately, neither attempts to reproduce the rhyme of the source language (SL) utterance or to replace it with another strategy. This is of course the easiest way to deal with language-based humour in screen translation, and by no means uncommon, as will be seen. 3.2 Example 2: Use of inappropriate near-homophone A second example from the same film, Four Weddings and a Funeral, shows more ambitious translations: (2) At another church wedding: the newly-examined priest (Rowan Atkinson) leads a wedding ceremony for the first time. He is very nervous. Priest: …who lives and reigns with you and the Holy Goat — ggghh — Ghost. German dubbing: …der mit dir lebt und regiert, und mit dem Heiligen Geiz — ggghh —
Language-based screen humour in translation
Geist. [ª‘… who lives and reigns with you and with the Holy Miserliness — ggghh — Ghost.’] Swedish TV subtitles: – som lever och härskar med dig och den helige Spöke…Ande. [ª‘who lives and reigns with you and the holy Spectre… Ghost.’] The strategy employed in the SL slip of the tongue might be labelled use of inappropriate near-homophone. More specifically, the near-homophone has one phoneme less than the intended word. In the German dubbed version, the strategy chosen is almost identical to the original one: Geiz (‘meanness, miserliness’) instead of Geist is also inappropriate, if not as outrageous as Goat, and the difference in pronunciation between the wrong and the right word is rather small and limited to the last phonemes. No phoneme has fallen away as in the original joke, but the last two have been transposed (/gaist/ Æ /gaits/). A different approach has been taken in the Swedish subtitles: instead of nearhomophones, words with overlapping meaning have been used to reproduce the comic effect. Possibly inspired by the alternative meaning of ghost, the translator lets the priest say Spöke (‘ghost’, in the sense of ‘spectre, spook’) and then the appropriate Ande (‘ghost’, in the sense of ‘spirit’), which sounds completely different. Thus, one strategy has been replaced by another. That could have worked well in principle since slips of the tongue do not have to be phonetically related to the intended utterance, but in the case under discussion a slight problem arises in the form of the definite article and the adjective. While both remain unchanged in English, no matter what noun they precede, the forms of the German and Swedish equivalents depend on the gender of the following noun. In the German dubbed version, both the slip of the tongue, Geiz, and the intended Geist have the same gender, and the grammar remains unaffected. In the Swedish text, on the other hand, the article and the adjective only go together, grammatically, with Ande, not with Spöke — which makes the replacement of the former by the latter a rather improbable slip of the tongue. But all this has more to do with the quite different question of success in translation, and the subtitle discussed here has to be counted as an instance of one strategy being replaced by one other strategy. 3.3 Example 3: Two in one — reinterpretation of abbreviation
and exploitation of spelling-pronunciation discrepancy The third example is from Men in Black. All seven translations are given as they will serve to illustrate that it can be difficult at times to determine the exact number of strategies employed. The asterisks are added to indicate a new subtitle.
161
162 Thorsten Schröter
(3) Plain-clothes police officer James Edwards (Will Smith) manages to catch a suspect whom he has been pursuing for a while. He holds his badge under the suspect’s nose. Edwards: Do you see this? Huh? N-Y-P-D! Means: I will (k)Nock Your Punk ass Down! German dubbing: Siehst du das hier? Hah? N-Y-P-D! [pronounced as in English] Das bedeutet: Nervenden Junkies Piss ich aufs Dach! [ª‘Do you see this? Huh? N-Y-P-D! That means: I piss irritating junkies on the roof (= head).’] German subtitles: *Siehst du das hier? *N-Y-P-D! *Das bedeutet: Nervenden Junkies piss ich auf ’s Dach! Swedish TV subtitles: *Ser du den här? New York-polisen! *-Det betyder att jag tar dig. [ª‘Do you see this? New York police! That means you’ve been caught / I will arrest you.’] Swedish DVD subtitles: *Ser du den här? N. Y. P. D! Det betyder: *-Nu Jävlar Passar Du dig! [ª‘… Now you (better) behave, damn it!’] Norwegian: *Ser du dette? N-Y-P-D! *-“Nå Ypper Purken Med Deg”. [ª‘Do you see this? N-Y-P-D! Now the cop will pick a fight with you / will show you who’s in charge.’] Danish: *Kan du se den her? N-Y-P-D. *-”Nu Ydmyger Panseren Dig!” [ª‘Can you see this? N-Y-P-D. Now the cop (will) humiliate(s) you!’] French: *Tu vois ça? *Police! *Je vais te faire ta fête! [ª‘Do you see this? Police! Now you’re in for it!’] This is an example of a wordplay where two strategies involving language manipulation have been combined: reinterpretation of an abbreviation and exploitation of a spelling-pronunciation discrepancy. In the German translations, a noticeable effort has been made to find an equivalent for the joke. This has not succeeded completely, though. For while the
Language-based screen humour in translation 163
original interpretation of the N in NYPD as a misspelt but correctly pronounced Knock renders the punch-line extra funny, there is hardly a connection between the letter Y and the word Junkies, which has become common in German but is still pronounced and written the same way as in English. (The only link, not exploited in the joke, is that in words of German origin, J is pronounced like the Y at the beginning of English words). If this was not simply a blunder, perhaps the dubbing team and the subtitler hoped that the German audience would not be able to react to this inconsistency in the very short space of time before the dialogue and the action continue. Depending on the intention, the line can be a case of one strategy instead of two (involving a new, only partially successful, reinterpretation of an abbreviation) or a ‘2:2’ (with a hard-to-classify second strategy that seems to have missed its point). The question does not need to be settled, however, as both possibilities can be taken into account simultaneously (see below). The joke is completely lost in the Swedish TV subtitles, which only reflect a ‘hard-core’ interpretation of what officer Edwards says. They thus form an instance of two strategies having been replaced by none. The Swedish and the Norwegian DVD subtitles are ambiguous in about the same way as the German translations. It is obvious that the strategy reinterpretation of an abbreviation has been transferred in both cases, but it is doubtful how far the interpretation of the Y as Jävlar and the use of five capital letters in Nå Ypper Purken Med Deg can be seen as purposeful strategies or just imperfections (accepted or unnoticed). In the Danish version one and only one strategy has been used, but for the French translation, the number of strategies can yet again be discussed. This time the question is whether there is simply no playful use of the language worth counting — or whether the double alliteration in the established sequence te faire ta fête should be considered a strategy. It might also be pointed out that some of the target language (TL) solutions, especially the Norwegian one, sound rather forced. Here and in other cases where some sort of reproduction of the original joke had been attempted, the result turned out to be less than perfect under closer inspection. This reflects the fact that translating language-play is a difficult business.
4. Results for one film Table 2 shows the quantitative results for one film, Men in Black, chosen because it is the only film with seven different translations included in the study. As can be seen from the totals, the number of strategies used in the different translations lies in all cases below, and sometimes well below, the number of strategies to be found in the English dialogue. This is true not only for Men in Black, but for all the films investigated in the study. It will also be noticed that there are
164 Thorsten Schröter
Table 2. Men in Black: number of identifiable strategies in English original wordplay and its translations (minimum/maximum, depending on analysis) Wordplay
English German German Swedish Swedish NorDanish French original dubbing subtitles subtitles subtitles wegian subtitles subitles audio (TV) (DVD) subtitles
N-Y-P-D 2 big butt 1 back half the man1 there-square1 memory- 1/2 thing White/Black 1 flashy mem- 1 ory TOTAL 8/9
1/2 0
1/2* 0*
0 0
1/2 0
1/2 0/1
1 0
0/1 0
1 1 1
1* 1* 1*
0/1 0 0
1 0* 0*
1 0 0
1 1 0/1
0 1 0/1
0/1 1
0/1* 0/1
0/1 0
0/1* 0/1
0/1 0
0/1 0/1
0/1 0/1
5/7
4/7
0/2
2/5
2/5
3/6
1/5
* = identical or very close to the other version in the same language
rather many instances where the number of strategies employed has been left undetermined. As the purpose of the study was to get a general idea of how language-play is dealt with in quantitative terms, no strict criteria for deciding on an exact number of strategies per wordplay were deemed necessary at this point. (This issue will be dealt with in extended future studies.) As both the minimum and the maximum have been taken into account, the results remain valid nonetheless. Only the arithmetic involved becomes somewhat more complicated.
5.
Results for all films
In Table 3, the number of identifiable strategies (minimum/maximum) is given for each target version. Of course, only those films have been included in the count where the language/version in question could be investigated. That explains the different figures in the SL column. The figures in columns I to IV result from the four possible ways of relating the minimum and maximum numbers of strategies in the SL and TL versions with each other. It will suffice to consider the combined values given in the column to the right instead. They represent the approximate proportion of SL strategies that have been transferred or substituted in the different languages and modes. If the figures are converted to percentages, this proportion lies between 60 and 71 percent in all but one case, i.e. around two thirds of the total number. The Swedish TV subtitles
Language-based screen humour in translation 165
Table 3. Number of strategies in translations related to number of strategies in English SL dialogues Target Language No. of strategies in SL min/max
No. of I. strategies min TL max SL in TL min/max
II. min TL min SL
III. max TL max SL
IV. max TL min SL
Average of I. to IV.
German dubbing German subtitles Swedish (TV) Swedish (DVD) Norwegian Danish French
46/59 39/54 7/9 43/57 43/58 42/59 27/39
0.55 0.57 0.41 0.58 0.58 0.57 0.63
0.63 0.68 0.50 0.68 0.69 0.70 0.76
0.71 0.78 0.53 0.77 0.78 0.80 0.91
0.60 0.63 0.46 0.64 0.64 0.64 0.71
83/93 69/79 17/18 74/84 74/84 74/84 43/51
0.49 0.49 0.39 0.51 0.51 0.50 0.53
are the exception, with less than half as many strategies as in the SL dialogues. On the other hand, only two films were analyzed in this category, in one of which, Men in Black, the TV subtitler had hardly made an effort at all to transfer the SL language-play to the captions (see Table 2). However, the results for the other languages and modes should also be viewed with some reservation. The number of films studied is of course limited, as is the amount of language-play detected. With such a corpus, any one of a variety of factors can have influenced the results to a significant extent: The perceptiveness, priorities, imagination and problem-solving abilities of the individual translators are probably reflected in the figures. Their performance might also be affected by the general quality of the films and the importance attributed to their translation. When it comes to the closely related Scandinavian languages, translators may have had a chance to copy from each another (cf. Gottlieb (1994: 117f.) on the spread of pivot subtitling in Scandinavia); and in many cases the German subtitles seem to build on the dubbing rather than on the SL text. Obviously, the results also depend on how the individual wordplays have been analyzed. With all this in mind, however, the figures can still serve as an indicator, if not a true reflection, of the general picture. It is of course to be expected that a certain percentage of strategies in language-play will fall away in translation. Jokes and amusing comments based on the manipulation of linguistic elements are among the greatest challenges for translators. At the same time, such jokes cannot simply be ignored when they contribute substantially to the character of a text, as is the case in comedies. Given these competing forces, a result suggesting the loss of about one third of the original strategies is not surprising. The outcome is roughly the same if one considers the proportion of SL language-play that has been replaced by some kind of language play in the TL
166 Thorsten Schröter
version, rather than the more complicated and unwieldy strategies. The figures in Table 4 are in most cases somewhat higher than the corresponding ones in Table 3, for the Danish subtitles even as much as ten percentage points (74 percent as opposed to 64 for the strategies), while they are identical for the Swedish DVD subtitles (64 percent in both cases). The results are thus more or less in line with each other. However, if one dares to draw conclusions from the small differences that are there after all, one might surmise that in general, it has been more important for the translators, or perhaps just more feasible (or both), to have some wordplay in the translation where there is one in the English text than it was to mirror the number of contributing strategies. Both hypotheses would be plausible enough: firstly, a wordplay based on just one strategy can be as surprising and amusing as one based on two strategies or more — and might thus serve the same purpose. Secondly, the more strategies contribute to a wordplay, the more elaborate it is and the less likely that a translator would come up with an equally elaborate equivalent. Larger and more detailed studies would have to confirm these suggestions, though. What is striking about the results of the present study is that they compare rather well with those of other investigations where the quantification of translated wordplay has been one of the aims: not only is the average percentage of SL language-play translated into TL language-play higher for the films analyzed here than the approximately 50% found in the study by Gottlieb (1997) referred to earlier, it is also higher than for written translations of literary classics. Delabastita (1993: 268f.), summing up his analysis of different translations of Hamlet, found that only about half of the puns in that play had been rendered by a pun in at least one of the TL versions and that many more of the original puns had been replaced by a TL pun in only a minority of the translations. In a study of French translations of three other Shakespeare plays, Offord (1997: 254f.) found in all cases reductions in the number of puns by more than 50%, and by more than 90% for the translation of one of the plays. Heibert (1993: 210ff.), finally, looking at translations of another classic, James Joyce’s Ulysses, also found considerably fewer instances of wordplay in the TL versions than in the original, and while he could show big differences between the translations in this respect, not even the best reaches quite the same percentage of wordplay > wordplay solutions as the highest-scoring
Table 4. Instances of SL language-play translated with the help of some kind of corresponding language-play (in percent) German dubbing
German subtitles
Swedish subtitles (TV)
Swedish subtitles (DVD)
Norwegian Danish
French
68
71
50
64
71
73
74
Language-based screen humour in translation 167
translations investigated in this study. This somewhat surprising finding confirms that screen translation does not need to have an inferiority complex vis-à-vis literary translation any more. Something that is not confirmed by the present study is the superiority of either dubbing or subtitling in dealing with language-based screen humour. Even less does it show that any language or group of languages would serve better as the target for translations of English wordplay than others. There are sizeable differences within one and the same language (Swedish TV vs. Swedish DVD subtitles), and the French subtitles, the only translations into a non-Germanic language included in the study, are among the better ones when it comes to preserving language-play and strategies. This is not to say that translators from English into other Germanic languages do not often have an advantage in this respect, but the study was not designed to prove this.
6. Conclusion and outlook Based on the analysis of one British and six American film comedies, the study that has been described here reveals that a substantial proportion of the language-play evident in the English dialogues has no equivalent in their translations. The SL dialogues have been compared with seven different combinations of target language, mode and medium, and the results show that both the instances of language-play and the number of individual strategies that help to create these instances have been reduced, on average, by roughly one third in the dubbing as well as in the subtitles. Several conflicting factors are probably responsible for keeping the reduction at about this level. On the one hand, language-play is an important ingredient in many comedies and for this reason alone it ought to be preserved to as large an extent as possible in translation. On the other hand, language-play is often very difficult to transfer, and the effort required to find good equivalents might sometimes be felt to be out of proportion to the expected benefits. The more complex or language-specific a wordplay is, the more likely it is to be simplified, replaced or omitted in the translation. The many cases where only one common strategy has been employed are generally less problematic, especially if the crucial aspects of the source and the target language are similar. Finally, the differences between dubbing and subtitling probably also play a role — and might, taken together, both contribute to and limit the loss of language-play in screen translation. For example: a dubbing team could easily be tempted to ignore a wordplay altogether because the audience would not notice the omission, while a subtitler, at least if working from English, will be quite conscious of the target audience’s proficiency in that language and do his or her best to avoid any glaring mismatches between spoken and written text. Yet the results could also be
168 Thorsten Schröter
the exact opposite: with the freedom that dubbing entails, an entirely new wordplay could be built up that fits the target language and serves the target audience better than a clumsy transfer of the original one. This is also advocated by e.g. Chiaro (1992: 86, 95). And the subtitler might prefer not to distract the viewers from the original joke, which most can understand anyway. Instead of attempting a difficult transfer, the result of which would be liable to criticism, the subtitler might then consider an unpretentious translation of the hard-core meaning of the troublesome sequence to be the best solution. Further studies, featuring a larger source material and refined analyses, will have to take up these issues and suggest more detailed and well-founded answers. The figures presented in this paper can, at the very least, serve dubbers, subtitlers and other interested parties as a first pointer as to where the standards seem to lie in the translation of language-play on screen.
References Chiaro, Delia. 1992. The Language of Jokes: Analysing Verbal Play. London & New York: Routledge. Delabastita, Dirk. 1993. There’s a Double Tongue: An Investigation into the Translation of Shakespeare’s Wordplay. Amsterdam & Atlanta: Rodopi. Delabastita, Dirk (ed.). 1997. Traductio: Essays on Punning and Translation. Manchester: St. Jerome & Namur: Presses Universitaires de Namur. Gottlieb, Henrik. 1994. “Subtitling: Diagonal translation”. Perspectives: Studies in Translatology 2 (1): 101–121. Gottlieb, Henrik. 1997. “You got the picture? On the polysemiotics of subtitling wordplay”. In: Delabastita, Dirk (ed.), 207–232. Heibert, Frank. 1993. Das Wortspiel als Stilmittel und seine Übersetzung: am Beispiel von sieben Übersetzungen des »Ulysses« von James Joyce. Tübingen: Gunter Narr. Offord, Malcolm. 1997. “Mapping Shakespeare’s puns in French translations”. In: Delabastita, Dirk (ed.), 233–260. Pisek, Gerhard. 1997. “Wordplay and the dubber/subtitler”. AAA — Arbeiten aus Anglistik und Amerikanistik 22 (1): 37–51. Zabalbeascoa, Patrick. 1994. “Factors in dubbing television comedy”. Perspectives: Studies in Translatology 2 (1): 89–99. Zabalbeascoa, Patrick. 1996. “Translating jokes for dubbed television situation comedies”. In: Delabastita, Dirk (ed.) Wordplay and Translation. Special issue of The Translator: Studies in Intercultural Communication 2 (2): 235–257. Whitman-Linsen, Candace. 1992. Through the Dubbing Glass: The Synchronization of American Motion Pictures into German, French and Spanish. Frankfurt am Main: Peter Lang.
The figure of the factory translator University and professional domains in the translation profession John Milton University of São Paulo, Brazil
The aim of this paper, based on two surveys which were carried out in Brazil in 1999 and 2001, is to examine the relationship between translators and academia in Brazil. It shows that Translation Studies as an academic area exists as an almost separate domain from that of professional translation in Brazil, and that there is minimal contact between these areas. University translation courses have traditionally been based on Letras courses and many have little contact with the profession. Indeed, many professional translators, especially men, come into the translation profession through other disciplines such as engineering, law and informatics. Although the results of my surveys examine the Brazilian situation, I believe they bear some relation to the situation which exists between professional translators and academia in other countries.
Introduction The aim of this paper, based on two surveys which were carried out in Brazil in 1999 and 2001, is to examine the relationship between translators and academia in Brazil. Although its results primarily examine the Brazilian situation, I believe they bear some relation to the situation which exists between professional translators and academia in other countries. My hypothesis is that Translation Studies as an academic area exists as an almost separate domain from that of professional translation in Brazil, and that there is minimal contact between these areas, a situation which is found, though maybe to a lesser extent, in other countries. Despite this somewhat bleak statement, it is possible to find an increasing amount of interest in the work of the professional translator in a number of areas, for example, in Interpreting Studies (see Shlesinger 1995), in Protocol Studies (see
170 John Milton
Kussmaul 1995), and the publication of such works as Translators through History (Delisle & Woodsworth 1995). Specific case studies can also be found. Johann Hermans and José Lambert (Hermans & Lambert 1998) examine the position of the professional translator in companies in Belgium. In the companies studied, assistant managers and secretaries who have a working knowledge of the foreign language are often called on to carry out translations. And although in-house translators may earn relatively well, they are seldom integrated into the company and usually feel low category employees who are carrying out a non-vital mechanical task. Another important article, by Kaisa Koskinen, analyzes translation in the European Commission in Brussels (Koskinen 2000). Final documents are usually hybrid, in that they contain translations of sections from different languages, which may have been worked on by a large number of different translators. All documents have to be translated into all EU languages, but in the case of “small” languages such as Finnish, the documents will already have been read in English or French. Frustration at the prospect of having no audience, and the considerable anonymity of the translator, who will have absolutely no contact with those who draft original documents, leads to a high turnover, despite excellent pay.
The 1999 survey In the first survey I examined three translation courses in Brazil: the undergraduate course of UNESP (São Paulo State University) São José de Rio Preto, in the hinterland of the state of São Paulo; the undergraduate course of the Universidade Ibero Americano in the city of São Paulo; and the Universidade de São Paulo postgraduate diploma course (for complete results and analysis see Milton 2001). The following questions were given to the UNESP and Ibero university students: Why did you choose this undergraduate translation course? Which professional career do you intend to follow? Are you working / Have you worked with translation / interpretation? Which types? Please give details. In which areas of translation would you like to work in the future? In which areas will this course most help your skills as a translator? How will this course help you in the translation market? What are the other positive effects of the course?
The USP questions were slightly different: Present Occupation. Academic background. Why are you taking this “Especialização” course in translation?
The figure of the factory translator
Are you working with translation / interpretation? Which types of translation? Please detail. Do you intend to work with translation in the future? In which areas of translation? In which areas will this course most help your skills as a translator? How will this course help you in the translation market? How will the course help you professionally? What are the other positive effects of the course? What is your opinion of undergraduate courses in translation? Would you have taken an undergraduate course in translation? Why?
I also asked professional translators who were members of an Internet group, TradPrt List
[email protected] (now
[email protected]), their opinions on the connection between the university and the profession and whether they had taken or would take a university translation course before attempting to enter the translation profession. This informal survey consisted of the following questions: Academic background. What type of translation and/or interpretation do you do? Which courses in the area have you taken? What type of university course would you recommend to someone beginning to work in the area? How do you see the relationship between practicing translators and undergraduate courses in translation? Other comments.
The main results of the surveys were as follows: –
–
–
Most students taking university translation courses do so in order to improve their language skills (RP 53%; Ibero 62%; see Appendix, Tables 1 & 2), rather more than those who aimed to work in the translation profession (RP 14%; Ibero 10%). Thus university translation courses are seen as excellent ways of improving language skills. Indeed, at a number of faculdades, private nonresearch colleges offering degrees, former low prestige Letras courses have been closed and reopened as translation courses in order to attract more students. The new syllabus is often very similar to the old. The USP Diploma course attracts a large number of teachers (45%; see Appendix, Table 3) who are seeking an alternative profession, and who are also interested in improving language skills, but far fewer practicing translators (25%). The fact that it is held in the afternoon prevents in-house translators from attending. None of the courses specializes in the newer areas of translation practice such as subtitling or online translation. Much course work is carried out in groups with minimal use of computers.
171
172 John Milton
–
–
–
–
Respondents (n = 30) in the informal survey given to translators were divided on the question as to whether university undergraduate translation courses should actually exist, particularly as many students graduate with a limited knowledge of the foreign language. Surely, say opponents of such courses, diploma courses directed towards students with good language skills are much more practical. Supporters of undergraduate translation courses admit the limited skills of many students but argue that they have a basis which may be built upon by future courses and potential employers. Moreover, undergraduate translation courses give Translation Studies a considerable institutional importance and weight. Many translators enter the translation profession through other professions, particularly in technical areas. For example, many translators of technical works are qualified engineers, etc. This is truer of men than women, who seem to believe more in such courses, and who have more professional and academic qualifications from translation and Letras university courses. In general, those working translators who took Letras or translation courses were happy with their choices and would recommend such a course; those who came in to translation from other areas were satisfied with their choice. The same could be found in the USP Diploma course, where those who had come from Letras courses were pleased to have done so (73%; Appendix, see Table 4), and those from more applied courses were also satisfied with their original choices (68%; Appendix, see Table 4). The university is generally seen by translators to be very distant from the real translation world and fails to train potential translators for the market. All university translation courses in Brazil originated from Letras courses, and few courses have practical contact with other areas. Many students obtain degrees in Translation/Interpretation with a limited knowledge of the foreign language and translation skills. Thus employers of translators often totally ignore degrees in the area when employing translators.
I can thus propose the existence of two almost separate domains in the translation world: that of professional translators, many of whom work in technical areas, who have had little contact with the university; and that of the academic translation world, of which the main constituents are translation undergraduate, diploma and postgraduate courses, the production of M. A. and Ph. D. theses, and publications and conferences. I suggest that there is only a very limited area where the two domains overlap, and that they remain virtually separate, with different professional activities such as workshops and congresses, different publications, and most importantly, a different focus, that of the professional area being the production of translations, and that of the university area the study of translations. I believe that the university translation domain can be considered a separate self-sustaining domain, with university
The figure of the factory translator
teachers teaching translation and translation theory courses, publishing in specialized journals and attending university translation conferences, while they may have little or, in many cases, no, contact with the profession itself, and many will never actually carry out translations. There is a dearth of published material analyzing translation courses in Brazil. One of the few critiques of existing translation courses appears in Luzia Araújo’s thesis, De Big Bangs a Buracos Negros no Universo da Tradução no Brasil: Um Estudo sobre o Papel da Terminologia na Prática Tradutória e na Formação de Tradutores [From Big Bangs to Black Holes in the Universe of Translation in Brazil: A Study on the Role of Terminology and Translation Practice and in the Training of Translators] (Araújo 2001), in which she compares undergraduate translation courses in Brazil with their counterparts in the UK. Courses in Brazil fail to take this very important aspect of translation practice seriously: little time is devoted to it; it is often placed too early in courses, thereby preventing students from taking advantage of knowledge of terminological practices in the final parts of their courses; few teachers of terminology have experience as practising translators — most are teachers of Linguistics; and students received very limited training in the use of appropriate software. Araújo concludes her thesis by suggesting a programme for a course on Terminology into translation courses in Brazilian universities: an initial course in “Applied Terminology”, and a more advanced course in “Managing Terminology for Translators”. My study concentrates on the situation specifically in Brazil, where, as I have emphasized, there is a distinct lack of contact between university and the profession. I suggest that this overlap may be greater in countries such as Germany and Belgium, where the well-developed system of Hochschulen trains a large number of professional translators.
The 2001 Survey In order to extend this hypothesis and see whether my hypothesis on the two almost separate domains would be substantiated, I asked a number of questions which extended some of the points made above in another questionnaire, which was sent in Portuguese to a number of Portuguese-English translation internet groups: tradinfo
[email protected] (now
[email protected]), a list which provides information about translation events in Brazil; two general lists for translators from and into Portuguese, Trad-Prt List
[email protected] (now
[email protected], tradutores
[email protected] (now
[email protected]); Abrapt
[email protected] (now
[email protected]), a list for members of the Brazilian Association for Translation Research; litterati
[email protected] (now
[email protected]), a
173
174 John Milton
group for translators of literature in Brazil; and a group for postgraduate students in the English Department of the Universidade de São Paulo, engpos
[email protected] (now
[email protected]). The questionnaire was subsequently sent to a list of ex-students from the Catholic University of São Paulo (PUC) Interpretation Course. Hopefully, this broad survey can be followed by more specific qualitative and quantitative surveys. M/F Age: Approximate number of hours spent on translations/interpretation per week? Do you work as a freelance? For an agency In house translator? Sworn translator? Approximate monthly income as a translator/interpreter? (You are not obliged to give an answer) In which areas of translation/interpretation do you work? Do you have other professions? Which? When filling out forms do you put your profession as a translator? If not which profession do you put? What was your training as a translator/interpreter? What is your academic background? How do you keep up-to-date as a translator/interpreter? What is your present contact with the university? Which congresses have you taken part in the last five years? Would you think about taking an MA? Ph D? Why? Which publications on translation do you read? Have you published in academic journals? Which professional associations do you belong to? How do you see the connection between the university and the translator/ interpreter professions?
The questionnaire was returned by 34 translators, 13 male (average age 45.53) and 23 female (average age 43.13) (overall average age 44). The number of hours worked on translation varied enormously, between 65 and 4, with a large number putting in a 40-hour-week. Likewise, salaries varied considerably, but a number of translators earned between three and five thousand reais per month (US$1 = R$3.5), a fairly good salary in Brazil. The answers to the questionnaire allow us to establish that there are very different types of translators. Firstly, a group of translators working in technical and/or business translation can be identified. The background of these translators may be a degree in a technical area such as Engineering or Business Studies, or, as
The figure of the factory translator
in a couple of cases, the translators may have no degree at all, or, as in one case, may be high school dropouts. This group of translators had served their apprenticeship “on the job”, working for agencies, under the supervision of an experienced translator, or they may even be self-taught. In several cases, they had spent a number of years living abroad. They take occasional specific courses but in general have minimal contact with the university, do not read academic publications, have no intention of taking any further university studies, and do not attend academic conferences. They improve their skills “on the job” and through contact with other translators. Some members of this group were members of professional associations such as the American Translators Association (ATA), the Brazilian Translators’ Union (SINTRA), the Brazilian Association of Translators (ABRATES), and the São Paulo Association of Sworn Translators (ATIPIESP). On other occasions, contact with the rest of the profession is made merely through Internet groups. The translation activities of this group may be connected to other activities such as localization (adapting manuals etc. to local conditions or specifications) or editing previously translated material. Members of this group may call themselves language consultants, “consultores”, or technical specialists, “técnicos”. Such translators, of whom nearly all work as freelance translators, and who, as mentioned, are often unqualified in any area related to “Translation” or “Letras”, have a high sense of individualism and are strongly anti-corporative. They work a 40-hour-week or longer and usually earn from R$3.000,00 (approx. ð1.000,00) upwards. This is one reason for a lack of interest in taking an M. A. or Ph. D, as the time needed to take such a course would severely reduce income, and any scholarship they might receive would be much less than their current income. Due to lack of contact with the university, this group often left the question on their opinion of the contact between university and the profession, unanswered. On other occasions, the lack of contact between the university and the translation profession and market was stressed. The second major group was that of teachers and postgraduate students who do a certain amount of translating as a secondary profession, often working only a few hours a week, from 4 to 10. This group was mostly female, and many of its members had taken university Letras or Translation courses or a Postgraduate Specialization Diploma, and comments about these courses were generally positive. Many of the members of this group translate for publishing companies, and, as the members of this group spend fewer hours translating due to academic commitments, the amounts earned through translation are much lower. In fact, a number of respondents had reduced the hours they had spent translating when starting to teach or take a postgraduate course. University life will also usually involve reading and publishing in academic journals and participation in academic conferences. This group emphasized the positive role of the university in providing a space for the circulation of ideas and the link between practice and theory, and the fact that
175
176 John Milton
a theoretical background that may help practising translators. They also mentioned the fact that contact between the university and the profession has improved to a certain extent in recent years. Yet there were some dissenting voices, who criticized the lack of contact with the translation industry and market; the inflexibility of the university system, which makes it difficult to offer specific modular courses; the impossibility of accepting non-degree holders in a specialized diploma course; and the lack of computer resources. The suggestion was also made that the university could play a much more active role in the profession: it would have the authority to establish some kind of professional qualification, perhaps similar to that of the Institute of Linguists in the UK, which might bring a little order to this very confused area.
Conclusions The responses from the survey carried out helped to substantiate my initial hypothesis: that there are two very distinct domains in the translation profession in Brazil — full-time translators who have often entered the profession through a technical profession such as Engineering or Computer Studies, and who have little or no university contact and often a rather negative idea of the contribution of the university to Translation Studies; and teachers, often of Translation or Translation Studies, whose main profession is teaching but who will carry out a small amount of translation, often for publishing houses. This group is highly involved in the university world of translation, publishing and reading translation journals and books and attending and organizing congresses. The breadth of the translation “community” was summed up by one respondent: “The translation “community” is too broad and diversified for its relations with the university to be homogeneous”. The survey does raise a number of questions as to the position of the university with regards to translation: should it make a concerted attempt to provide courses in the developing areas of translation such as media translation, interpretation, and computerized translation tools, or could this function be best performed by other entities: translators’ associations or more flexible schools, such as the AmericanBrazilian bi-national centre Associação Alumni, which runs a highly successful translation and interpretation course in São Paulo? Likewise, to what extent should the university attempt to approach, work with and learn from the translating profession? Might it also play an arbitrating role in the profession, helping to provide some kind of professional qualification and standard? Should this be a worry of the university? After all, except for a relatively small number of creative writing courses in Anglo-American universities, the university has little influence on creative writers. At the moment, university systems in different countries view translation practice in very different ways. As mentioned, in countries such as
The figure of the factory translator 177
Germany and Belgium, it is seen as the responsibility of higher technical schools, which have less status than universities. In Finland it is fully integrated into the university system. Newer universities in the UK are also attempting to provide practical training for translators and interpreters. In Brazil and Spain, university translation courses are becoming a much more attractive option than Letras courses. But many of these courses are little more than general language courses with little very specific translator or interpreter training. Here in Brazil the great majority of translation training takes place outside the university. Should this worry those of us working in Translation Studies at the university? Or should we attempt to come out of the ivory towers and involve the Brazilian university in the professional life of translators? Author’s Address: John Milton University of São Paulo, Brazil DLM, FFLCH, USP, 05508–900
[email protected] Appendix Table 1a.Reasons for choosing course — Rio Pretoa 1st year Course qualities Language skills Free public univ. Not to teach Near home To be translator Others Total a
2nd year
3rd year
4th year
Total
4
11%
3
7%
6
17%
11
45%
24
17%
19
53%
27
61%
16
46%
6
25%
68
53%
1
2%
6
17%
4
16%
11
8%
3 – 6
7%
1 2 3
3% 6% 9%
2 1 –
8% 4%
6 5 19
4% 4% 14%
3 37
9%
6 139
4%
– – 2 10 1 36
5% 28% 3%
2 42
14% 4%
24
Here, as in a number of the other questions, respondents often gave more than one choice.
178 John Milton
Table 2.Reasons for choosing course — Ibero 1st year Course qualities More job opportunities Language skills Part of job To be translator Others Total
2nd year
3rd year
–
–
5
17%
8
10%
4th year 7% 3
Total
4
9%
–
–
3
7%
–
–
7
4%
28 – 6
62% – 13%
13 5 1
45% 17% 3%
52 5 8
63% 6% 10%
27 – 5
66% – 12%
120 7 20
62% 4% 10%
7 45
15%
5 29
17%
7 83
8%
6 41
15%
23 193
12%
16
8%
Table 3.Profession of USP translation students Profession
No.
%
Translator None Teacher Others
16 9 29 10
25% 14% 45% 16%
Total
64
Table 4.Students who would have taken undergraduate translation course: Letras and non-Letras Letras
Students
Yes
%Letras No
22
73%
8
Non-Letras % Letras
Yes
% non-Letras No
27%
8
32%
17
% non-Letras 68%
References Araújo, Luzia. 2001. “De Big Bangs a Buracos Negros no Universo da Tradução no Brasil: Um Estudo sobre o Papel da terminologia na Prática Tradutória e na Formação de Tradutores”, Ph. D. thesis. Instituto de Estudos da Linguagem, Unicamp. Delisle, Jean, and Judith Woodsworth. 1995. Translators through History. Amsterdam: John Benjamins and UNESCO. Hermans, Johann, and José Lambert. 1998. “From translation markets to language management: The implications of translation services”. Target 10:1, 113–132. Koskinen, Kaisa. 2000. “Institutional illusions: Translating in the EU Commission”. The Translator. 6:1, 49–65.
The figure of the factory translator 179
Kussmaul, Paul. 1995. Training the Translator. Amsterdam: John Benjamins. Milton, John, and Kirsten Malmkjaer and Veronica Smith. 2000. “Translation and mass culture”. In Translation in Context, Selected Contributions from the EST Congress, Granada 1998, Andrew Chesterman, Natividad Gallardo San Salvador and Yves Gambier (eds), 243–259. Amsterdam: John Benjamins. Milton, John. 2001. “To become a translator or to improve language skills? Why do Brazilian students choose university translation courses?” TradTerm 6.1., 61–81. Pym, Anthony. 1998. Method in Translation History. Manchester: St. Jerome. Shlesinger, Miriam. 1995. “Stranger in paradigms”. In Target 7:1,1–28.
Migrating from translation to technical communication and usability* Hanna Risku Danube University Krems, Austria
The main aim of this paper is to illustrate the personal experiences of translators who have migrated to the field of technical communication and/or usability. The results are based on theoretical analysis and a qualitative interview study and are then discussed in the context of translator training and career development. On a theoretical level, translation competence would appear to provide an excellent foundation for professional development into the fields of international technical communication and usability. Professional translators are capable of analysing both text and situation, carrying out research on requirements and conventions, managing intercultural communication projects and writing reader and user-oriented texts. Thus, their work includes several of the components required in International Technical Writing and International Usability Testing. However, there are also important differences between translation and technical communication. This paper compares the competencies required and draws upon both theoretical and empirical (non-representative qualitative) research.
Introduction This article examines the similarities and differences in the competencies required for technical translation and technical communication (TC). The theoretical “distant-observer-viewpoint” is compared with a “first-person-viewpoint”, namely the personal real-life experiences of six former translators who became technical communicators. The goals of this study are to determine why TC is attractive to translators and then to ascertain why so many translators succeed in migrating to this field. The results are discussed in the context of translator training and career development and are thus relevant for translators and trainers alike.
182 Hanna Risku
Approximately one third of the students on the postgraduate TC course at the Danube University Krems in Austria have a translating background. The course caters for technical communicators who want either to keep abreast of current developments or to deepen their knowledge of TC. It also offers translators an opportunity to acquire the additional knowledge and skills required to migrate from translation to a career in one of the TC fields: technical writing, illustration, usability engineering, information management, information design, web page design, TC consulting or project management. We are often asked what translators migrating to TC should take into consideration: – – – – –
Are there any similarities to translation work? Do TC projects resemble translation projects? Will it be disadvantageous to have a translating rather than, say, an engineering background? How might my role and my status change? What additional skills might I still have to learn?
Until now, we have replied using a conceptual comparison of the competencies needed, drawing from the literature available (see Risku 1998, Schäffner and Adab 2000, Göpferich 1998, Risku 2000a, b). Empirical data on translators who have migrated to TC is limited to statistics on the percentage of translation graduates working in TC (e.g. for data on Austria see Moisl 1996, Wußler 1999: 33f.). There are few personal accounts available (Farthofer and Haussteiner 1997, Eronen 1994). However, we generally agreed with the former translator John D. Graham who claims “good technical translators are automatically good technical communicators” (2000, personal communication). Both translation and TC can be described as cooperative text design, i.e. the effective use of source texts and other knowledge sources to produce documents suitable for a specific target group, situation and environment (Risku 1998: 263). Based on this premise, translators are well suited to careers in TC and, in particular, in multilingual TC and international information design. However, conversations with TC recruitment managers led us to the revised conclusion that technical translation alone was often not considered a suitable background. Indeed, other areas of study and work experience were deemed more appropriate or valued more highly. For example, many Cognitive Science and Human-Computer Interaction graduates opt for a career in TC, although they usually have neither theoretical nor practical backgrounds in source text analysis, terminology, writing or text production. These deliberations prompted us to study the realities of migrating from translation to TC. Where do former translators position themselves between the poles of similarity and difference? Employers and users of technical documentation have no insight into the tasks and competencies involved. Experience-based insider knowledge
Migrating from translation to technical communication and usability
can only really be obtained from those who have made this career change. This describes the view of “translators-turned-technical-communicators” instead of documenting the evaluative, prescriptive views of non-translators and non-technicalcommunicators with potentially only superficial knowledge of these professions.
Methodology In summer 2001, six qualitative, semi-structured, individual interviews of about two hours each were carried out in a neutral, non-business environment and a complete transcript of each was produced. Excerpts from these transcripts are given in italics and were translated from the original German by the author. At the beginning of each interview, the practical and personal goals of the study were explained and interviewees were guaranteed anonymity. Each was asked to describe their educational and professional background and experience. The interviewer assumed a listening role, only prompting the interviewee if necessary. This ensured that the information given was highly personalised, not simply a premeditated reaction to the interviewer’s questions with “expected” answers. Nevertheless, the fact remains that humans always post-rationalise their own actions, regardless of how recent the memory might be (Suchman 1987). If the interviewee did not volunteer certain information during the interview, the following questions were asked: 1. 2. 3. 4. 5.
What originally motivated you to shift to TC? Did you notice any similarities to or differences from translation work? Do TC projects resemble translation projects or do they differ greatly? Which tasks did you find especially easy or difficult? Do you think your translating background gave you any advantages or disadvantages over technical communicators from other backgrounds? 6. Did your professional role and status change or do you consider both professions to be equal? 7. Can you name any special skills and competencies you either already had or needed to learn? Finally, the subject was asked to identify any competencies she had either clearly mastered or had to learn for TC from the following list: – – – – – –
Target group analysis Task analysis Media analysis Consulting clients on the use of different media Text analysis Research: electronic sources, libraries, personal contacts and interviews
183
184 Hanna Risku
– – – – – – – – – – – – – – – –
Teamwork Use of source material Argumentation strategies for client negotiations Development of communication/text concepts Producing an initial product draft Writing Editing Terminology Management of a variety of languages and cultures Use and development of style guides Incorporation of text, images, etc. Ordering and managing translation projects Usability testing Project/time management Financial calculation Obtaining feedback from clients and readers/users
All six interviewees (identified below as A to F) were university graduates of translation studies programmes in Austria and had initially worked as translators before migrating to TC. One interviewee had also completed postgraduate studies in TC. They were all in full-time employment at the time of the interviews and not studying formally. All the interviewees were female, which was probably not coincidental given the high percentage of female translators. This added another perspective, providing data on the experience of moving from a female- to a maledominated working environment. The former translators interviewed all appear to have felt the need to move on from translation. An interview often offers participants a welcome opportunity to talk about the negative aspects of a former occupation, giving their description of translation a rather negative touch. Drawing a clear dividing line between other professions and focussing on one’s own knowledge community often helps establish a new professional identity. The qualitative, non-representative interview method was chosen because neither observation, representative surveys with statistical analysis, nor interviews with employers or colleagues would have given the desired subjective insight. On the contrary, people without experience of translating or TC are rarely aware of what these professions really entail. Thus, expert interviews were held, with the participants treated as subjects (expert), not study objects (implying the interviewer has more knowledge of the topic than the interviewee). To determine any identifiable structures or patterns, the transcript material was first analysed by source (data-driven) to determine any categories of analysis originating from the interviews and thus from the participants themselves. The material was subsequently categorised according to the seven questions mentioned above.
Migrating from translation to technical communication and usability
The presentation of the study is structured into the four sub-competencies of translation presented in Risku (1998): 1. 2. 3. 4.
Target analysis Source analysis Development of an overall strategy Project management and self-management
Results We will begin with the original motivations to migrate to technical communication.
Motivating factors Only one interviewee had studied TC prior to migration. Attending a conference on TC had been her initial motivation to do so. She was fascinated by people who knew so much about HTML and localisation and wanted to learn about them too. None of the others had had any comprehensive training; only an introductory presentation by a visiting lecturer and a lecture series at an American university. Factors motivating the career shift can be divided into three categories: career-, technology- and translation-related. 1. Career-related factors – Well-defined job profile – A job without secretarial work (assuming that only translators are abused for secretarial tasks) – Well-paid job in an international organisation – Career development opportunities – On-the-job learning 2. Technology-related factors – Interest in technology – Prior research experience for a terminological masters thesis – Experience in translating software manuals – Experience of communication with technology experts – Challenge of being female in a male-dominated environment (to make a competent impression) 3. Translation-related factors – Application of language and text competencies acquired for translation – Avoiding the thankless job of translating – Another profession and competence to balance uneven translation workloads – Advantage on the sparse translation market
185
186 Hanna Risku
The results suggest that the major migration factors are a positive experience in technical translation, career development opportunities and sparse translation markets. Succeeding in a man’s world was also a motivating factor. Language and text competencies and a willingness to learn on-the-job made the career shift possible.
Competence 1 — Target Analysis: Situation and Target Group Analysis Conceptual analysis Technical communicators define relevant target groups through task analysis. The make-up of these groups is highly relevant since products are used for different purposes in different environments (e.g. first-time/expert users, technophobes or technophiles). Technical communicators interview end-users and observe them in their working environments. Since their main concern is to enable the use of products, any instructions must follow user rather than developer logic. Target group and situation analyses forms an essential part of TC courses. Conceptual comparison of competencies shows professional translators are already well equipped for target situation and target group analysis. They are aware of the situative aspects of information design and already analyse source texts, target groups and target situations. This requires research using online-databases, technical publications and parallel literature and increasing use is made of the Internet instead of libraries. However, the Internet does not provide all the answers and cannot replace contact with subject matter and language experts. Therefore, translators must be competent in information research and also have the necessary social and communicative skills to obtain information from their contacts (see also Vienne, 2000). Translation clients and product developers often do not understand that translators and technical communicators serve the interests of their target readers/ users. An incomprehensible text is a disservice to source text authors, translation clients, developers and producers alike. Interviewees’ statements The participants partly confirmed our views. Translators and technical communicators must enjoy and be competent in research (A, E, F) and carry out similar target group analysis (E). They do not need to be Subject Matter Experts (SMEs), e.g. developers, programmers or financial experts (A, B). Indeed, being an SME or understanding software code would be counter-productive to the role of advocate of the reader/user (E, F). Unlike SMEs (F), translators see things from a target- or user-oriented point of view (A, E, F). Communicating with new SMEs was not a problem (E, F), because translation was perfect training for that (E). Indeed, people with other backgrounds lack that experience (E). Translators are adept at listening carefully (A, B), picking
Migrating from translation to technical communication and usability 187
up words and phrases (A) and preparing for an interview to avoid looking totally stupid (B). Interview techniques were lacking in translation training (E). However, several interviewees talked in emotional, unambiguous terms of their lack of competence in target analysis (I felt totally out of my depth, like a complete moron, A). Target analysis and maintaining the user perspective when talking to SMEs are the most important and difficult tasks (A, E). One interviewee felt unprepared to systematically assess user/reader knowledge level (B). When they did manage to assess the target group, they didn’t know how to carry out task analysis (i.e. research the environment, actions and needs of the user/reader) and thus lacked the necessary criteria to select the content and structure of the target text (E, F). As translators, they learned to concentrate on linguistic and terminological research, whereas in TC they had to determine, select and structure argumentation and content (A, B, D, F). Attending user-training courses might be beneficial to understand the needs and questions of users/readers (B). The lack of preparatory research in translation and translation training conflicted with our assumptions. In translation, they always had the source text as a starting point and began translating immediately (B, C, D). Any research generally started later and accompanied the translation process, whereas in TC, preparatory research was required (B, D). Furthermore, in translation, only text was analysed, whereas in TC, the most suitable medium for the target situation also has to be determined (E). Translators can usually select people who know about the subject and are willing to provide information as contacts (government departments, companies, friends), whereas contacts allocated in TC often don’t know enough about the subject/product or really provide answers (D).
Competence 2 — Source Analysis: Situation and Source Material Analysis Conceptual analysis Both translators and technical communicators work with source material written in a specific language. In TC, this is the language of the subject matter and marketing experts. Source material includes product specifications, construction drawings or internal training documents. Technical communicators also work with previous versions or similar products or use print manuals and/or instructions as source material for web content and CDs. However, unlike translation, there is no welldefined source text. They interview SMEs to discuss content issues. Discussion with relevant experts is considered part of their job. According to Fraser (2000), turning to clients and contacts for assistance is not always an option for translators. Questions are often frowned upon (“you don’t have to understand it, just translate it!” is a common response). Thus, translators are not automatically accepted as communication partners searching for source-related information.
188 Hanna Risku
Social and communicative skills are prerequisites for success in both fields since they enable the correct formulation of needs and requirements. These skills must rest on a sound theoretical base and require appropriate argumentation, background knowledge, intuition, creativity and experience. It is this combination of skills and competencies that allows successful problem resolution. This is confirmed by Dörner (1989), who studied problem solvers and troubleshooters in action. Interviewees’ statements The participants largely supported these views. Translators and technical communicators both know how to analyse source material and quickly acquaint themselves with new subject matter (B, C, E, F). Understanding is no problem since you get or can find enough information! (B). Both are initial test users/readers (B, F), a motivating aspect in translation training (B). You learn to sit down, understand new material and recognise the central theme (C). Employers also recognise you have to learn on-the-job and offer training (E, C). However, co-operation partners do not recognise the necessity of source analysis. No-one is responsible for helping me understand source texts (D). Learning the basics of a variety of fields during translation studies was considered positive (E, F), whereas the need to study at least two foreign languages was questioned (C). Knowledge of foreign language(s) and the mother tongue is advantageous to SMEs and native speakers (B, D, E). Language and communication competence are prerequisites for both careers (D). Two talked emotionally about their love of language (B, D). The ability to recognise and serve other people’s communication needs is advantageous, e.g. to journalists who have individual styles or SMEs who just report on the latest product functionality (D). The quality of source texts and material was criticised as pseudo-information presented in a pseudo-systematic form; it is incomplete and unreadable (D). The presumption that the availability of a source text would be seen as a difference between translation and TC was confirmed. The source text was a good and sufficient starting point in translation (D, F). However, some unexpected differences were also uncovered. Source analysis for translation was based purely on the source text and language, whereas in TC the thing itself was studied (A, B). However, it is unlikely they would have expressed themselves in the same way about some issues during their translating careers (e.g. translators don’t need to understand what they translate, whereas technical communicators need to understand the source material (B)). In translation, the use of two terms for the same concept in the source text is considered relevant and equivalent terms are searched for in the target language, whereas in TC inconsistent terminology is avoided (A). The need to underline their new professional identity (see “Methodology”) appears to play a role here in simplifying the task of the translator. While the linguistic focus of translator training was fascinating as students, it is disadvantageous in TC where it is more important to use and understand different
Migrating from translation to technical communication and usability 189
reference materials than to master and find vocabulary (B). Source analysis in translation is based mainly on written material, whereas in TC there is a greater need for personal interviews (E). In some areas, the knowledge of the SMEs can be advantageous, e.g. machine construction (C). It would be beneficial to graduate and work in another field before completing a post-graduate translation programme (You should study something sensible first) (C).
Competence 3 — Developing an Overall Strategy Conceptual analysis Technical communicators first develop a general picture of the target situation and text (see above). These target hypotheses are continuously refined during the text design and production processes to develop an overall strategy. Target hypotheses and overall strategies also play a decisive role in professional translation. Translators do not simply reproduce or follow source texts slavishly and without criticism, they develop a strategy for the complete translation task, including a situation-specific approach to the use of the source text (Risku and Freihoff 2000). The selection of appropriate media and tools forms part of the text production phase where technical communicators specify usage of different elements, e.g. text, graphics, photos, animation, video sequences, etc. Technical communicators need to know the consequences of combining certain media types and be aware of cognitive aspects of text comprehension and reader/user behaviour to make informed decisions. According to Resch (2000: 184), non-sequential and unstructured reading of text is characteristic of today’s typical readers. Their behaviour is similar to that of individuals who channel-hop or impatiently surf the web. This emphasises the importance of usability considerations (see below). Technical communicators establish a general concept for text design and layout based on specific requirements (e.g. corporate style guides and design principles). Technical translators must also meet format and layout requirements. Increasingly, clients expect translations to be delivered ready for publication (e.g. in print form, on the web or on CD). The ability to provide these gives them a commercial advantage but cannot (yet?) be considered a core translation competency. In contrast to technical communicators, technical translators rarely influence the choice of media and tools. However, they must often question the suitability of proposed media for a specific target situation, e.g. the availability of appropriate hardware among the target group. Thus, technical communicators and translators play different roles in strategy development. Technical communicators participate in the selection of media and content, whereas translators are seen primarily as language experts responsible for the linguistic transformation of source material.
190 Hanna Risku
Interviewees’ statements The design process and use of graphics and tools was described as new and uncharted territory (A, B, D, F). Combining and handling text and images was a new, fun competence (C, F), as were new tools (C). The focus on language in translation was seen as an obstacle because the use of images and other media tends to be ignored (B). Tools are a career booster and can be used to consult clients on selection and appropriate use (E). They are also advantageous for translation (C). A translator will definitely benefit from working as a technical communicator! (C). Writing skills are based on personal experience and talent rather than a specific background (A, B). Selecting and keeping to a common layout is equally important in both fields (C). One interviewee enjoyed and felt competent in writing texts in her own words, also considering this an important skill for translation (B).1 Others had not learned to formulate content without a source text and only knew how to reformulate other people’s existing texts (A, C, D). Despite their importance, writing skills are totally lacking in translation studies (D). However, translation experience teaches you to structure information (D). Whilst adapting the same information for different product formats is enjoyable (B), one participant lacked that competence altogether (E). Learning to take notes, use symbols, train graphical recognition (A) and communicate content rather than reformulate linguistic material (D) in interpreter training is good preparation for TC. The technical communicator’s role is different and more active (D). They are free to make design decisions (B) and being involved in the complete development process increases job satisfaction and identification (A, E). Developers are grateful for screen design suggestions and it feels good to argue for the selection of content from a user perspective (E). The translator’s role is more passive and isolated (E, C). Starting out as a technical communicator is difficult because you have no experience of selecting content (A), developing overall concepts or plans (A, C), structuring complete texts (B, D) or presenting information in small didactic bites (A, B, C). You are trained to think differently in translation studies (A). They had not even used translation tools (C, F): Translation management and terminology tools should be as integral a part of teaching translation as interpreting booths are for training interpreters (F). In contrast to our view of translation, some participants described translation as primarily linguistic and TC as primarily communicative activities. Whilst the principal goal of translation work is high linguistic quality, in TC it is logical structure and content (A). Even in technical translation, the goal is to translate the source text into the target language and not necessarily to enable users to handle the product (B). Thus, translations don’t have the same relationship to practical situations as technical documents. A translator can write well-formulated sentences without necessarily considering if the content is correct (B).
Migrating from translation to technical communication and usability
Competence 4 — Project Management and Self-Management Conceptual analysis Both technical communicators and translators require project management skills. They need to plan, organise, meet deadlines and calculate the resources required. Feedback from co-workers and readers/users documents success and failure factors, but is difficult to collate. However, it is needed to ensure continuous improvement in future projects and both technical communicators and translators strive to raise awareness in their co-workers. They need frequent further education and the selfconfidence to handle conflicts, criticism and ignorance. Information, documentation and terminology management skills are further prerequisites for the success of any projects, since they result in time and cost savings in Internet and database research or in interviewing information sources (e.g. SMEs and users). It is often difficult to predict the desired effect of graphics and icons. This can only be guaranteed by observing and interviewing target users, and HCI (HumanComputer-Interaction) usability tests play an important role. International usability testing will provide fascinating job opportunities for translators in the future, since they are well equipped for usability testing with readers/users from different cultural and language backgrounds. With high development costs, producers of localised computer software cannot afford to leave acceptance to chance. Usability testing will become an integral part of the TC product life cycle and should have equal status with target group and situation analysis. Technical communicators co-operate with product developers, marketing experts, terminologists, illustrators, usability engineers, customer service departments, web designers, clients and, last but not least, translators. Therefore, those with a translating background are uniquely aware of translators’ needs and capable of producing documents suitable for translation and localisation. They know translators need contacts and reference material and understand the importance of consistent terminology and phraseology in source texts. Interviewees’ statements Translators and technical communicators are mediators (A, D, E), advocates (A, B), consultants and all-round geniuses (B), although the latter is not always seen as positive by partners (B). Both were for different reasons also described as “Don Quixotes” (A) and outsiders (A, B): Translators only transform texts written by others, whereas technical communicators are not SMEs (A, D). In neither group does one own the intellectual property or receive much recognition; rather, one disappears or becomes invisible compared to members of other professions, e.g. journalists (B). Low visibility and lack of general awareness of the competencies
191
192 Hanna Risku
involved lead to both being exploited for secretarial tasks (A, E). Thus, a migration meant to exclude secretarial work was not necessarily justified. Competencies such as writing and structuring original text and increased contact with other company employees, led to the TC role being rated higher (F). Whilst two interviewees described technical communicators as translators or interpreters between developers and users (C, F), others did not accept this comparison, since technical communication is not just translating (D). Again, whilst such expressions suggest a feeling of career advancement through TC, their opinion might well have been different while working as translators. Conversely, some felt that translators (and in particular interpreters) were more likely to be considered as experts since few co-operation partners feel capable of carrying out these tasks, while everyone thinks they can write user documentation (C, D). Translators and technical communicators need good contacts and gain respect by using the terminology of partners and employers (A, B, E) and presenting wellresearched target analysis (A). Personality plays a key role in both areas (B, D). You must be communicative (B, D), independent and self-confident (E, D), yet able to handle criticism (C). The participants did not gain self-confidence through translation training. Indeed, they asked themselves: What do I really know? Am I really able to do anything? (A, C). A year abroad helped increase self-confidence (B, C, D) either due to the different teaching approach (suddenly we heard “You are good! You can do it!”) (B, C) or the work experience (D). You must continually learn and update your knowledge through further education (A, B, E). Both have to work hard to obtain feedback (E, F). For TC, more knowledge of usability and testing is needed (C), because methods for assessing text are required (at least translators have the source text for comparison) (C, also D). Terminology management is necessary for both career paths, despite the fact that translators document linguistic terminology while technical communicators document concepts (A). Again, we find the surprisingly common contrast between translation as a linguistic and TC as a communicative activity. As translators, the interviewees tended to avoid management and organisational tasks although they consider them necessary for both professions (B, E). Translators have the advantage of clear deadlines whereas technical communicators must adapt to changing product development cycles (E, D).
Consequences and discussion: A challenge through further education The study demonstrates that translation training and professional experience provided participants with a good foundation to further educate themselves into the
Migrating from translation to technical communication and usability 193
role of technical communicator. Their background did not automatically make them technical communicators, but provided core competencies such as – – – – –
carrying out terminological research, communicating with SMEs, analysing source material, getting acquainted with new subject matter, and mastering both a foreign language and one’s mother tongue
that made them ideally suited to this profession. Some missing competencies were clearly TC-specific and would be best taught in postgraduate study programmes, such as – – – – – –
information design, combining text and images, selecting appropriate media, common TC tools, usability testing, and enhanced project management.
The most interesting consequence for translation scholars is that a conceptual comparison reveals a lack of specific knowledge, skills, experiences and characteristics that are equally important for both TC and translation. These essential components of translation curriculae are needed to provide translation training that enables students to feel competent and find their place in today’s job market. To summarise, the interviewees emphasised the importance of the following competencies: – – – – – – –
systematic target analysis, formulation of criteria for the selection of appropriate content for the target group, ability to analyse and produce content and not just linguistic material, ability to abstract from the source text and concentrate on the needs of the target group, writing in your own words, familiarity with different text types, trust in your own ability.
All these aspects could easily be integrated into translation training. Companies are coming to realise that unprofessional TC puts them at a disadvantage. They must invest in the appropriate training to equip their technical communicators with both a theoretical background and practical skills. If willing to make this investment, they can maintain their competitive edge in today’s information society. In this situation, translators migrating to technical communication are in a unique, advantageous position to fill such vacancies now and in the future.
194 Hanna Risku
Author’s Address: Hanna Risku Danube University Krems, Austria Hackenberggasse 45 C5 A-1190 Vienna
[email protected] Notes * We would like to thank the participants for taking part in the interviews and giving us an insight into their work. We also thank the government of Lower Austria for its support. Finally, we would like to thank both the anonymous reviewers and the participants for reading this paper and for their valuable comments and feedback. 1. As one of the participants noted when reading the results of the study, you can never use “your own words” as a technical communicator either. Technical text production is always highly constrained by the communicative conventions of the particular situation.
References Carliner, Saul. 1999. Overview of the Technical Communication Industry: The Work of Technical Communicators. URL: http://saulcarliner.home.att.net/workoftcs.htm. Dörner, Dietrich. 1989. Die Logik des Mißlingens. Strategisches Denken in komplexen Situationen. Hamburg: Rowohlt. Eronen, Mari. 1994. “Tekninen kirjoittaminen”. Kääntäjä 7/1994. Göpferich, Susanne. 1998. Interkulturelles Technical Writing: Fachliches adressatengerecht vermitteln. Ein Lehr- und Arbeitsbuch [Forum für Fachsprachen-Forschung 40]. Tübingen: Narr. Farthofer, Andrea and Haussteiner, Ingrid. 1997. “Von Schnörkeln, Grenzschnichten und Blickskodes — Herausforderungen für Technische Redakteure in der EDV-Branche”. In Berufsbilder für Übersetzer und Dolmetscher. Perspektiven nach dem Studium, I. Kurz and A. Moisl (eds), 80–87. Wien: WUV. Fraser, Janet. 2000. “The broader view: How freelance translators define translation competence”. In Developing Translation Competence, C. Schäffner and B. Adab (eds), 51–62. Amsterdam: Benjamins. Kurz, Ingrid/Moisl, Angela. (eds) 1997. Berufsbilder für Übersetzer und Dolmetscher. Perspektiven nach dem Studium. Wien: WUV. Moisl, Angela. 1996. Die Beschäftigungssituation der Absolventinnen und Absolventen des ITAT Wien. Die Jahrgänge 1988 bis 1991. Zwischenbericht 1996 [unpublished]. Resch, Renate. 2000. “Werbetexte — multisemiotisch, intertextuell und zapper-gerecht. Neue Herausforderungen für ÜbersetzerInnen und Translationswissenschaft”. In Translationswissenschaft. Festschrift für Mary Snell-Hornby zum 60. Geburtstag, M. Kadric, K. Kaindl and F. Pöchhacker (eds), 183–193. Tübingen: Stauffenburg. Risku, Hanna. 1998. Translatorische Kompetenz. Kognitive Grundlagen des Übersetzens als Expertentätigkeit. Tübingen: Stauffenburg.
Migrating from translation to technical communication and usability 195
Risku, Hanna. 2000a. “Postgraduate program in technical communication at the Danube University Krems”. TC Forum 4: 18–20. Risku, Hanna. 2000b. “Berufsbegleitender Master-Studiengang”. Technische Kommunikation 6: 46–47. Risku, Hanna and Freihoff, Roland. 2000. “Kooperative Textgestaltung im translatorischen Handlungsrahmen”. In Translation in Context. Selected contributions from the EST Congress, Granada 1998, A. Chesterman, N. G. San Salvador and Y. Gambier (eds), 49–59. Amsterdam: Benjamins. Schäffner, Christina and Adab, Beverly. (eds) 2000. Developing Translation Competence. Amsterdam: Benjamins. Suchman, Lucy. 1987. Plans and Situated Actions. Cambridge: Cambridge University Press. Suojanen, Tytti. 2000. Technical Communication Research: Dissemination, Reception, Utilization. URL: http://tutkielmat.uta.fi/pdf/lisuri00001.pdf. Vienne, Jean. 2000. “Which competences should we teach to future translators, and how?” In Developing Translation Competence, C. Schäffner and B. Adab (eds), 91–100. Amsterdam: Benjamins. Wußler, Annette. 1999. Anforderungsprofile für Übersetzer und Dolmetscher. TextconText Beiheft 6 (tct 13 = 3): 1–166.
From raw data to knowledge representation Methodologies for user-interactive acquisition and processing of multilingual terminology Barbara Dragsted and Benjamin Kjeldsen Copenhagen Business School, Department of English
The Copenhagen Business School (CBS), Faculty of Modern Languages, recently participated in SENSUS, an EU Language Engineering project providing new technology aimed at analysing multilingual data and breaking down language barriers between law enforcement agencies in Europe. The natural language processing technologies developed and enhanced by the project rely heavily on the fields of machine translation and terminology. The CBS project contribution focused on interpreting, communicating and operationalising the needs of human translators, machine translation software developers and the intelligence community, all of whom formed part of the project. This task was solved by triangulating the positions of the three groups around a common interface for collecting, structuring and validating multilingual police and intelligence-related terminology. This paper describes a partly new approach to leveraging the potential of machine translation and intelligence processing by involving human translators in the development process, and discusses various aspects and practical experience related to the interaction with police translators. Particular emphasis is put on the challenge of coping with the very different approaches, needs and expectations on the part of the different user groups.
1.
Introduction to the SENSUS project
SENSUS was a Language Engineering project operating under the European Commission Fifth Framework Programme. The project ran from 1998 to 2000 with the overall aim of breaking down language barriers by facilitating efficient and fast communication between law enforcement agencies in Europe. SENSUS endeavoured to combine existing language technology products with new research and
198 Barbara Dragsted and Benjamin Kjeldsen
development. Focus was on providing linguistic resources and technology for two central areas: multilingual communication and multilingual document analysis.1 With reference to multilingual communication, SENSUS supports the workflow of international communication by providing translation (i.e. traditional translation as well as machine translation) and content analysis. Different levels of sophistication for the translation tools were explored and developed by the project, from simple term replacement to full machine translation. Human translators can access a terminology and knowledge database designed by CBS as part of the project, and automated translation tools and other natural language processing (NLP) applications can draw on the content of this database. The multilingual communication component could be incorporated in an independent operational communication network allowing police officers to exchange information across borders, but the final SENSUS system does not include such a network, since this was not the goal of the project. As to multilingual analysis, the focus of SENSUS was to support the war on crime by providing intelligence and police analysts with intelligent, multilingual data analysis, data mining, visualisation and reporting tools in support of, and in accordance with, their specific intelligence analysis requirements. Unstructured and structured data elements are analysed by the project end application and made available to the analyst by various tools. In specific terms, the task of the project was to compose a system that would enable police and intelligence services to quickly examine documents obtained from various sources with a view to determining the content and relevance of the document in question. Technically, this requires a powerful scanning device and multilingual Optical Character Recognition (OCR) tools (in the case of paper documents) as well as storage and retrieval capacity. Once a document is in the system, it is indexed for rapid retrieval and possible cross-referencing to other documents containing similar types of events, names of persons, places, dates, cars etc. The goal was to make it possible for, say, a German border control officer to check whether a specific license plate, for instance, is mentioned in a Portuguese police report, and if so, what the police report is about. Through an intuitive graphical interface, the border officer can also check whether documents seized from a suspect vehicle contain any information useful to an ongoing investigation, such as names of persons, events etc. It is possible to check the document for specific terms (including slang expressions) that are indicative of criminal action or otherwise relevant to police work. The SENSUS project involves end users from police, security and intelligence services of more than 10 European countries. The SENSUS user group, which was co-ordinated by Europol, was not only the intended users of the end product, but also participated actively in the project already in the development stage by providing terminology for the project.
From raw data to knowledge representation 199
The NLP software developers in the project are leading companies in the areas of linguistics, retrieval, data mining, visualisation and software integration and workflow applications.
2.
CBS contribution to the SENSUS project
Both the user group and the technology group in the project were assisted by academic research institutions, of which the Copenhagen Business School (CBS) was one. Drawing on our experience and knowledge in the domains of linguistics, communication, terminology and translation and on experience from previous police communication projects (such as LinguaNet), the Business School formed the linguistic hub connecting the real-life requirements of the user group and the technical work of the developers of NLP technology for the project. More specifically, CBS were responsible for collecting and processing terminology for the project’s different users and uses. The participating police forces were asked to submit relevant wordlists and texts that they might have in electronic format or on paper. The ‘working basis’ for the terminology collection process was thus a large collection of wordlists and texts in a wide variety of electronic formats, as well as hard copies of domain-specific dictionaries and full texts. The material comprised very different information items structured in a number of different ways, and the language coverage varied greatly. Therefore, comprehensive restructuring of the input was required to make it conform with the agreed project formats. Once the terminology was compiled into a common format, actual terminology work could begin. The multi-faceted applicability of the system required linguistic resources of varying degrees of sophistication. Police translators need detailed encyclopaedic monolingual and bilingual dictionaries for correct translation of specific terms in all situations. A machine translation system has little room for synonyms and nearsynonyms in the target language and hence requires clearly distinguished terminology with one-to-one translation equivalence. The CBS SENSUS Team were responsible for catering for this variation in the users’ needs by identifying their requirements and setting up a method for channelling and sorting linguistic resources in the project.
3.
Different groups with different interests
It was clear from the outset that three different terminology user groups could be identified: NLP software developers, police officers, and police translators. It was also clear that these different groups had different kinds of needs that had to be met.
200 Barbara Dragsted and Benjamin Kjeldsen
The NLP software developers were interested in feeding as much terminology as possible into the machine translation component of the SENSUS system. NLP systems, including MT tools, are, of course, highly dependent on the volume of their vocabulary. Above all, the focus of the NLP software developers was therefore on quantity more than on quality. This group emphasised a need for vast amounts of terminological data in as many languages as possible, with a specified amount of linguistic information, including semantic categories (corresponding to those of already existing machine translation systems) (e.g. Hutchins and Somers 1992: 19f.), grammatical information such as part of speech, gender, countability etc., but little or no information beyond the level of the lexeme in terms of definitions, regional variance marking, terminologist and source identification etc. In the course of the project, the degree of linguistic information required was somewhat reduced. The police translators took more or less the opposite approach with a focus on quality above quantity. In order to provide this user group with an efficient tool, terminological data had to undergo comprehensive processing: again, language coverage was an important factor, but just as important was the amount of information included in each terminological record. Preferably, each record in the terminological database would include: subject area of the concept, any synonyms and abbreviations, linguistic information (such as part of speech, gender, number) and, very importantly, definitions of the concept in all languages, hyperlinks to related concepts, information on the origin and usage of the concept, and an example of context in which the term appeared. In this respect, a terminological database does not differ much from the ideal LSP (language for special purposes) dictionary (Dyrberg and Tournay 1999: 62). In practice, the CLS Framework data categories (ISO 12620) were adopted for structuring the project termbase in order to secure consistency of data storage as well as interchangeability.2 Another important piece of information to be included concerned the degree of equivalence with concepts in other languages, especially since a good deal of the terminological data concerned legal language with its inherent cultural, historical, and regional distinctiveness which often results in twisted conceptual scopes between languages (Bergenholtz and Tarp 1994: 109f.). In contrast, MT systems often have little room for such finely grained cultural variation or other conceptual distinctions. It is generally recognised, though, that qualitative terminology work can also be put to use in NLP applications from machine translation to content analysis and retrieval mechanisms (Slocum and Morgan 1995: 232). Finally, the police officers primarily wanted a tool that worked, i.e. a tool which would provide them with quick solutions to any language and data analysis problem they might have. The police officers would benefit from both the machine translation and data analysis component of the system, on which the technology providers were focusing, and from the terminology and knowledge database emphasised by the police translators. However, unlike the translators, who have a
From raw data to knowledge representation 201
natural understanding of the importance of comprehensive terminological information, the police officers were not used to doing terminological work and were therefore not immediately inclined to set aside the large amount of time required for thorough terminological processing. On the other hand, the knowledge possessed by this group was indispensable in the process of gathering and validating factual information in the database. Hence, it was necessary to direct attention towards an educational effort for this group to shed light on the practice and applicability of terminology work. As primary end users of the project, the police expert group and their needs also reflected the ultimate goal of the project, but it could be difficult for the individual member of the group to realise the potential of the final system and the complex processes involved in its development. One challenge was thus to visualise the link between terminology work and the efficiency of the final application for this group.
4. Challenges Overall, the biggest concern in the terminology part of the SENSUS project was to unite the sometimes highly conflicting needs of the three different user groups. More specifically, we identified the following challenges: First of all, there was a conflict between the interests of NLP software developers and police translators. The NLP software developers were, as already mentioned, mostly interested in large quantities of data without elaborate information on individual terms, but with an approximated one-to-one equivalence, whereas the translators wanted a database with comprehensive terminological records including both linguistic and factual knowledge, and translational equivalence clarification. One of our challenges was therefore to consider the interests of both these groups at the same time. In short, we may call this challenge the quality–quantity conflict. Furthermore, there was a conflict between the needs of the police translators on the one hand and the motivation, and to some extent ability, of police officers to perform terminological work on the other. Another challenge was therefore to introduce police officers to the nature and benefits of terminological work, and then to teach them how to do it. This challenge could be termed the user involvement challenge.
5.
Approach
A composite approach was taken in response to these challenges. First of all, it was of course of great importance that all three user groups be involved in the process, so we arranged individual meetings with the different users to discover exactly what their needs and expectations were. The meetings allowed end users with little or no
202 Barbara Dragsted and Benjamin Kjeldsen
experience with terminology and computational translation aids to obtain a realistic background on which to base expectations of the final system, i.e. both the declared and the latent needs and expectations of the user groups were explored. The NLP software developers were also given the opportunity to voice their expectations and requirements. Thus, through dialogue, all users became familiar with the project potential, and mutual expectations were exchanged, which created a motivational and inspirational basis for further co-operation in the project.
The conflict between quality and quantity In order to consider the conflicting interests of both the NLP software developers and the police translators, it was decided to divide the terminology collection process into two parts: a quantity part which was to satisfy the NLP software developers’ requirement for vast amounts of terminological data, and a quality part which would meet the needs of the police translators for an elaborate terminological database and allow systematic maintenance and expansion of the database. The quantity part implied that all the raw word lists received from the different police officers and police translators were converted into a common format and supplied only with the type of information required by the NLP software developers. Furthermore, CBS extracted terms from texts received from users (using the corpus tool System Quirk) and, again, added the information required by NLP software developers. All the terminology processed in the quantity part also became part of the terminology database developed for manual consultation, even though it contained only minimal information. Moreover, work was instantiated to select a number of key terms for further processing in the quality part of the project. The quality part was highly dependent on user interaction in order to gather elaborate specialised information within the SENSUS domain. Our first initiative was therefore to involve the police translators and police officers in the terminology collection process by setting up a procedure for interactive terminology exchange aiming at validating and expanding terminological entries. We termed our terminology exchange methodology ‘the electronic basket procedure’, because the terminology would travel between CBS and the individual users in small batches or ‘baskets’ in electronic form. Technically, the format used for the baskets was Microsoft Excel, which everybody was expected to be familiar with. So for instance, we would send an Excel basket to the French police containing around 20 terms for monolingual expansion. The French police were thus tasked with adding definitions and other information to the terms. Upon completion of their task, they would return the basket to CBS, and we would then check the input and send it on to the German police, for instance, who would perform the same task and so on until sufficient information had been added in all SENSUS languages. After this process of filling linguistic and factual gaps in different languages, the next step in the basket
From raw data to knowledge representation 203
procedure was a validation process where the baskets were sent out once again for final multilingual validation by experts. After validation, the terms were ready to enter the final SENSUS terminology database. Apart from forming part of the database, the terminology collected through the basket procedure could also be used in the MT and data analysis component. The terminology collected through the basket procedure was very elaborate, and in this way satisfied the police translators’ need for quality terminology. However, the process was also extremely time consuming, and we soon found a need to narrow down the terminology exchange procedure to cover 200 highly relevant terms which were to undergo extensive terminological processing. While this number may seem relatively small, it provides a suitable testbed for the new methodology introduced, and the actual number of high-quality terms eventually exceeded this initial goal by far. The terms were to be identified by the users themselves, and at this point a new and very important aspect was added, namely the formulation of a knowledge model, cf. Figure 1, the purpose of which was to (1) secure more systematic collection of data, (2) discover any holes in the data collection, (3) structure the terminological data in a logical way, and (4) parallelise multilingual and multicultural data. As appears from Figure 1, the SENSUS Knowledge Model is structured around an event, e.g. a criminal act such as burglary. The term burglary would thus be assigned the knowledge model code event. The boxes surrounding the event relate
6 1
Rules & Regulations
2
Person
Perpetrator
3
Authorities
Victim
Other
Police
Courts
Other
Activities
Activities
Activities
Procedures
Procedures
Procedures
Objects
Objects
Objects
Documents
Documents
Documents
Evidence
Evidence
Evidence
Event
Time/ Place
4 MO
5
Physical Objects
Figure 1.SENSUS Knowledge Model
204 Barbara Dragsted and Benjamin Kjeldsen
to how and when an event took place (the left side of the model), and to how it is investigated or handled by the authorities (the right side of the model). So for instance the term burglar would fit in the box perpetrator, a crowbar could be an object involved in a criminal act. The term crime scene investigation would be put in the box Procedures under police activities, and Police report would fit in the Documents box. The boxes surrounding an event can then be further elaborated by identifying subtypes of for instance Documents related to police investigation. The overall characteristics and ideology behind the event-based knowledge model can be applied to other domains and applications as well. The resulting structure of terminology and knowledge may help for instance a large corporation organise and share its knowledge in one or more languages, which would secure consistency in use of language and reduce the margin for error in communication, both internally in the corporation and externally. For “Event”, substitute “Product”, “Project” or “Process”, for instance, when talking about a corporation. The other labels in the model must be substituted correspondingly to match the processes of the corporation in question. Once a universal model for that particular corporation is built, new projects, products or processes may be categorised using the same custom shell. In a non-commercial or governmental organisation, the base point may be a “Process”, such as passing a law in parliament, publishing a document, implementing a computer system etc. The resulting organisation of sub-processes and relevant information pertaining to the product, process or event will find a multitude of possible uses, ranging from pure terminology or knowledge structuring over document routing applications or knowledge sharing to NLP applications such as machine translation, search engine facilities or other types of artificial intelligence. The logical structure of the SENSUS Knowledge Model helps bridge the gap between the requirements and capabilities of the three user groups. The model caters for the semantic category modelling required by NLP developers, the terminological information desired by police translators, and the comprehensibility barrier seen by the police officers actually contributing their knowledge to the system. The intuitive composition of the knowledge model provided police officers and translators alike with a tool for input and organisation of knowledge. In many ways, the model is a pre-fabricated conceptual system ready to structure information relevant to the user groups as well as to the linguistic software system developers. By pre-fabricating a conceptual system for any domain, you run the risk of simplifying conceptual nuances, as the labels of individual terminological categories will be broadly defined. On the other hand, you have a universal, accessible tool for organising terminology and the conceptual knowledge underlying the terms. As demonstrated in the following, this feature proved to be the decisive factor in securing user-interaction and creating a flexible, reliable, and efficient term bank from which all users and developers in the project would benefit.
From raw data to knowledge representation 205
User involvement challenge In response to the challenge of involving police officers in terminological work, CBS arranged two terminology workshops. The aim of these workshops was to explain the principles and application of terminology to police officers, in order for them to become more motivated and better equipped to do terminological work. Moreover, the workshops provided both police officers and police translators with hands-on experience so that any methodological and technical problems in relation to the processing of electronic baskets could be resolved as they arose. One of the technical issues discussed at the workshops was the medium of the electronic baskets. Microsoft Excel was supposed to be a known or recognisable application among European police users, but was in fact not supported at very many user sites. Realising that the “familiarity” gain of a known application was lost anyway, CBS decided to introduce users to a fully fledged terminology application, viz. Trados MultiTerm, a terminology database application. This allowed users flexible input options, ranging from plain text to direct typing into MultiTerm. At the same time, users were able to track progress in the database and start consulting it right away. Apart from the two terminology workshops involving both police officers and translators, three seminars were held with individual users, both police translators and police officers. The purpose of the seminars was mainly to arrive at the most logical and workable structure of the knowledge model which would form the basis for the identification of the first 200 key terms, and subsequently to actually identify the 200 terms. As discussed, the knowledge model not only served the purpose of systematising term collection and making the tasks easier for those new to terminology work, but also helped route terminological data to the various NLP applications. The seminars with individual users were very effective and resulted in the formulation of the knowledge model presented earlier, as well as the selection of 200 central terms in German and English. The two terminology workshops were given quite a positive reception by police officers as well as police translators, and were very important steps towards the aim of collecting and validating the 200 key terms. Project software developers also supported the proliferation of the knowledge model as a valuable tool for structuring linguistic input for the various artificial intelligence applications in the project.
6. Output and conclusion The CBS contribution to the SENSUS project was to collect, unify and structure terminological data from different users for different purposes by means of user interaction. The concrete output of the CBS contribution was terminological data which are used partly for automatic data analysis and machine translation in the final
206 Barbara Dragsted and Benjamin Kjeldsen
SENSUS system (the quantity part) and partly as an independent look-up tool for translators and police officers (the quality part). Technically, both types of terminological data are combined in one joint Trados MultiTerm database. The database holds approx. 13,000 terminological records, each containing data in one or more of the nine languages dealt with in the project. In total, the records contain almost 50,000 terms with varying degrees of conceptual and linguistic information and validation. This means that the average language coverage is 3.8 languages per record. Out of the 13,000 articles, some 200 are central concepts identified on the basis of the SENSUS knowledge model tailored for police terminology. These 200 central concepts serve as the basis for expanding qualitative terminology work in European police co-operation projects. The terminology has been implemented in the various components of the final SENSUS system, which, at this point, operates in four languages: English, German, French and Spanish. Apart from this concrete output in the form of a terminological database, the terminology project has contributed methodologies and experience which can be drawn on in future similar projects: Firstly, a methodology for exchange of terminology across countries and institutions has been developed. Secondly, a knowledge model has been drawn up, which is tailored for police terminology and usable for identification and structuring of police terminology in the future. As shown earlier, the knowledge model is an implementation of a process-based or event-based knowledge modelling strategy developed by CBS, which can be tailored to other processes, projects or products. Thirdly, experience has been gained concerning the difficulties of joining and motivating the diverse groups of police officers, human translators, and NLP developers. The gap between human translators and machine translation developers is particularly interesting, as it is increasingly becoming a challenge in the daily work of both parties. The project succeeded in creating mutual understanding through constructive dialogue and educational efforts. We resolved the challenge of motivating all three identified groups to make a joint contribution from which all benefited. By demonstrating, through dialogue and in workshops, that it was possible to create a sum greater than its parts and by providing a shared medium for interaction, we were able to fully leverage the potential of the groups and, consequently, that of the project in general. Author’s Address:
[email protected],
[email protected] From raw data to knowledge representation 207
Notes 1. For more information on the SENSUS project, see the CBS/SENSUS homepage at www.cbs.dk/ departments/vista/sensus which includes a full list of participants, publications and an electronic version of this paper. 2. The CLS Framework (Concept-oriented with Links and Shared references) was developed by Brigham Young University and the Kent State University. ISO 12620 has now been superseded by ISO 16642 (the TMF project).
References Bergenholtz, H. and S. Tarp. 1994. Manual i Fagleksikografi. Udarbejdelse af fagordbøger — problemer og løsningsforslag. Herning: systime. Dyrberg, G. and J. Touray. 1999. “Ækvivalens og brugerdeterminerede oplysninger i fagsproglige oversættelsesordbøger”. In Nordiske Studier i Lexikografi 4:1, Slotte, P. et al. (eds.) 61–74, Helsingfors: Nordiske föreningen för lexikografi. Hutchins, W. J. and H. Somers. 1992. An Introduction to Machine Translation. London: Academic Press. Kugler, M., K. Ahmad and G. Thurmair (eds.). 1991. Translator’s Workbench. Tools and Terminology for Translation and Text Processing. Berlin: Springer. Ohse, Katrin et al. 1995. TRADOS MultiTerm ’95+! User’s Guide. Stuttgart: TRADOS GmbH Sager, J. C. 1990. A Practical Course in Terminology Processing. Amsterdam: John Benjamins Publishing Co. Slocum, J. and M.G. Morgan. 1995. “The role of dictionaries and machine-readable lexicons in translation”. In Automating the Lexicon. Research and Practice in a Multilingual Environment, Walker, D. E., A. Zampolli and N. Calzolari (eds.), 221–247. New York: Oxford University Press. Teubert, W. 1996. “Comparable or parallel corpora”. International Journal of Lexicography 9:3, 238–264.
Internet References CBS/SENSUS Project http://www.cbs.dk/departments/vista/sensus/ CLS Framework Data Categories http://www.ttt.org/clsframe/datcats.html LinguaNet Project http://www.cbs.dk/departments/fir/linguanet/ SENSUS Project http://www.sensus-int.de/ System Quirk Language Engineering Workbench http://www.computing.surrey.ac.uk/SystemQ/
The translator as a creative genius Robert Schumann João Azenha Junior University of São Paulo, Brazil Musik ist die höhere Potenz der Poesie.1
This article is part of a broader research project, which aims to create an annotated translation of Robert Schumann’s Gesammelte Schriften über Musik und Musiker (Collected Writings on Music and Musicians) into Portuguese. Such a translation will give special consideration to Schumann’s relations with the German Romantic Movement — Literature and Aesthetics — and with the foundations of musical critique. In this introductory article, I attempt to present a bird’s-eye view of Schumann’s interest in foreign languages — the core of his school instruction at the Gymnasium in Zwickau — and his activity as a translator of Greek and Latin poets. Schumann’s musical rendering of his literary experience as a writer, reader, translator and editor reveals the composer in two different aspects: as a spokesman, in music, for the exponents of German Classicism and for the poets of his generation, and as an eager reader, who, in his compositions, provided a very personal interpretation of the literary canon of his time.
The work of Robert Schumann as a theoretician, musical critic, poet and author of fiction, essays and aphorisms is virtually unknown to the Brazilian public. In Brazil, Robert Schumann is known mostly among music lovers and experts, and even then, their appreciation of his works is almost always limited to the so-called piano works from his first phase — composed before ca. 1840 — and to part of his instrumental works. Similarly, as far as his biography is concerned, special emphasis has been given in Brazil and in many other countries to his depression and to the shadowy and eccentric side of his personality, an image traditionally conveyed by motion pictures, biographical syntheses and books. Nevertheless, it is a well-known fact to all those interested in Robert Schumann’s life and work that until the age of twenty, the composer lived in doubt between a
210 João Azenha Junior
musical and a literary career. It is not until 1830, when he was a student of Law in Heidelberg, that he made the decision to embrace a career as a musician, explicitly declaring his decision in a letter to his mother in Zwickau. By that time Schumann had already written some of his piano works and intended to pursue a musical career as a pianist. Yet, an injury to the fourth finger of his right hand diverted him from his plans and made him devote himself to music as a composer and critic instead. The experience he had gained in all those early years of engagement with literature, eager reading2 and original writing3 were to indelibly mark his bonds with literature and the theoretical and aesthetic ideals of German Romanticism. Evidence of Schumann’s commitment to literature is seen in his activity as a translator. Schumann made his first translations in the early years of his education at the Gymnasium in Zwickau and was intensely devoted to translating until the 1830s. After this, he reduced the time dedicated to translation but, all through his life, he never completely stopped translating. This article is a brief introduction to Schumann’s devotion to translation, which according to Draheim (1996: 271; my translation), would surprise even those best acquainted with his biography with its magnitude and diversity, ranging from the tragedies of Sophocles to the sonnets of Petrarch, including the odes of Horace, and extending to neo-Latin literature with Sarbiewski, and English literature with Ossian.
Two preliminary facts should be taken into account regarding Schumann’s work as a translator: the example given by his father, August Schumann, and his education (above average, for the times) as a pupil of the Archdeacon Dr. Döhner and as a student at the Gymnasium in Zwickau. Schumann’s father, Friedrich August Gottlob Schumann, son of a pastor of humble origin, achieved a certain notoriety as a publisher and bookseller in the society of Zwickau, in Saxony. A voracious reader and a dedicated student of foreign cultures and languages, August Schumann earned himself some renown as a translator of Lord Byron and as a pioneer in introducing the works of foreign writers in Germany through a series of translations published in pocket-book format called Classics of All Peoples and of All Times. According to Schoppe (1987), apart from his appreciation of foreign languages and literatures, August Schumann’s efforts to introduce foreign literature through pocket translations bore a direct relationship to his determination to fight — through his work as a publisher and translator — against class divisions and the privileges of the nobility and for the dissemination of knowledge to the people. All this, at a time when the Napoleonic army was sweeping across Europe, kindling nationalistic feelings among the subjugated peoples. By the age of seven, when Schumann was admitted to the school of the Archdeacon Dr. Döhner, he was already well versed in elementary subjects. At that
The translator as a creative genius
time he began his Latin studies and, at the age of eight, French and Greek. Schoppe (1987) comments that at this phase in his life, Schumann outdid his classmates in the study of foreign languages (especially Latin) and “treads his own path and translates Greek texts for himself” (p. 9). At the time, Robert, the boy, had no intention of publishing these translations, but the experience of translating Greek and Latin poets left deep marks on the writer and composer-to-be: His literary expression was shaped according to the example given by the Greek and Latin poets. Robert is astoundingly gifted for languages and, besides, displays a very keen sense of metre. Anacreon’s chants, Bion’s, Theocritus’ and Mosche’s idylls — he translated them into German according to the metre of the original. He read Horace and the Greek tragic poets in the original (Schoppe 1987: 10; my translation)
In 1820, when he was nine, Schumann joined the Gymnasium in Zwickau, which had had an excellent reputation since the Renaissance, with a curriculum that set aside more than 20 hours of classroom instruction for Latin and Greek: Graduates were required to be capable of reading a relatively easy writer, like Livius, as fluently as they would read a book written in German […]. In Latin and Greek classes in the final year of school, only Latin was spoken; the pupils were required to take notes on the subjects discussed in Latin and elaborate on them at home. Only the most difficult passages of Greek writers would be translated into German; for less difficult passages, the Latin translation was deemed sufficient. They were frequently asked to produce versified German translations of the poets. All versified translations from Robert Schumann’s pen were preserved.4 [my translation]
According to Draheim (1996), the oldest piece of evidence of Schumann’s aboveaverage interest in classical languages is a note he wrote to his friend Emil Herzog at the age of twelve, in which he quotes Cicero.5 In 1827, when he was 16, Schumann was awarded a book as a prize: Tacitus’ Agricola. In that same year, another quotation from Horace’s Ars poetica appeared in his Diaries (“Tage des Jünglinglebens”, 1827), which points to the important role played by foreign languages and literatures in young Schumann’s everyday life: I frequently thought about Nanni [Petsch]: if I could only define what I feel — et amata relinquere pernix — says old Horace, with reference to youth: and yet, every time I think of Liddy [Hempel], Nanni [Petsch] appears right before my eyes with an expression of anger. It was the happiest time of my life, truly the happiest hours, last spring, when I loved and was loved.6 [my translation]
Robert Schumann’s interest in foreign languages and in translation, as well as the
211
212 João Azenha Junior
importance he attached to this period of his life, did not decline with time. More than 20 years later (about 1840), while gathering information for a curriculum vitae, he wrote: 1825. Translation (metrical) of Anacreon, later of Bion, Theocritus, Mosche, Homer, Sophocles, Tibullus, Horace, Sarbiewsky (a great deal). Great talent for the metre. 1825. Literary association. Also tried to write drama. (Coriolanus), later (1829) much of Ossian, Petrarch.7 [my translation]
Schoppe (1987) comments that, in parallel with all this reading, writing and translating, his inclination towards music also grew: at 12, Schumann composed dances, the “Psalm 150 for choir and orchestra”, some operatic pieces and many pieces for voice and piano. Schumann himself later said: Simultaneously, at that time (shortly before reaching my twenty-first birthday), I tried my hand at many poetry texts. I was already acquainted with the major poets of every country. When I was 18, I was delighted to read Jean Paul; at that time, I also listened to Franz Schubert for the first time. Goethe and Bach were unknown to me until then.8 [my translation]
In the Introduction to the 4th edition of his Writings, Jansen (1891) refers to this phase of Schumann’s life, when he was deeply involved with languages and literature: These simultaneous and diverse activities, performed earnestly and diligently, distracted the young man from the subjects taught at school. Yet, his intensive study of classical languages, outside the regular school curriculum, taught him what he would later learn at school. Long before his classmates were capable of reading Greek and Latin poets in the original, Schumann had already translated into German, according to the original metre, chants of Anacreon, idylls of Bion, Theocritus and Mosche. When he mastered the difficult forms of Horace and of the Greek tragic authors, he was already in possession of an aptitude, self-confidence and freedom of expression that would raise his poetry above the average and, most especially, his prose, awakened and winged by the reading of Jean Paul.9 [my translation]
Schumann’s devotion to the study of foreign languages and cultures, and particularly to Classical Antiquity, was not restricted to Greek and Latin. Draheim (1991, 1996) mentions translations done by Schumann from Italian (Petrarch) and from English (Ossian and Felicia Hemans, among others). This is an indication that Schumann had studied not only three (Latin, Greek and French), but five languages, with the addition of Italian and English. In order to achieve an overview of Schumann’s diversified production as a translator, I will now present a list of his translations from Greek and Latin. The translations made by Schumann from other languages are to be found throughout his writings, diaries and correspondence, so
The translator as a creative genius
that the lengthy and meticulous task of pinpointing and editing these writings is still to be carried out.
Translations from Greek a. Sophocles (496–406 B. C.): Antigone, verses 1 to 22 in the Prologue; Oedipus in Colonna, verses 1 to 291. [8 handwritten pages]. b. Bion (end of the 2nd century B. C.): 6 idylls and a proverb (full-text translations, though only the former in verse); Theocritus (first half of the 3rd century B. C.): 3 idylls (full-text translations, but not versified); Mosche (about 150 B. C.): 2 idylls (full translations). [33 handwritten pages]
Translations from Latin a. Horace (Quintus Horatius Flaccus, 65–8 B. C.): Selected odes of Horace translated according to the original metre by Robert Schumann. Comprising: II.18, II.19, II.20, III.3, III.9, III.12.10 Most of the odes are translated in full. In one of the odes [III.3], Schumann presents only the metrical scheme; some others [II.18, II.19 e II.20] are preceded by the metrical scheme. [11 handwritten pages]. It is quite remarkable that the composer-translator should analyse the Latin text to establish the metrical scheme on which his translation would be based. According to Draheim (1996), Schumann’s choice of 6 of the 88 odes of Horace — also part of the reading programme at the Gymnasium in Zwickau — was unusual. Apart from three well-known odes, which had already been translated several times (c.II, 19 e 20; c.III, 9), the other three (c.II, 20 e C. III, 7 e 11) are “less famous and less esteemed” (p. 277). No information is known about the sources for Schumann’s translations, nor any commentaries or previous translations he might have been acquainted with. Draheim suggests that Schumann’s selection of the odes from the 2nd book might be motivated by the epicurean satisfaction and the peacefulness of the philosopher-poet in ode 18 (c.II), the ravishing, youthful and dionysiac impetus in ode 19 (c. II), and the motif of eternal posthumous fame in ode 20 (p. 277; my translation).
b. Tibullus (Albius Tibullus, ca. 50–19 B. C.): Elegies of Tibullus translated according to the original metre by Robert Schumann.11 The opening page shows an epigraph by Ernst Schulze. Elegies 1 to 4 from the 1st book were fully translated. As to the fifth elegy, there is nothing but the title. [14 handwritten pages]. c. Maciej Kazimierz Sarbiewski/Matthias Kasimir Sabievius (1595–1640): Selected
213
214 João Azenha Junior
lyric poems by Horace and Sarbiewski translated according to the metre of the original text by Robert Schumann.12 This selection comprises odes and an epigram by Sarbiewski, all in full-text translations with the exception of Aufschwung, for which only the first 14 of the 22 stanzas and the beginning of the 15th are translated. [22 handwritten pages]. It is no surprise that the vast majority of Schumann’s work as a translator involves ancient works. Apart from the fact that classical languages — particularly Latin and Greek — constituted the core of his school instruction at the Gymnasium in Zwickau, the study and translation of ancient works of literature were an everpresent element in the humanistic education during the transition from the 18th to the 19th century in Germany. No detailed discussion of the motives of such praxis in view of the philosophical and literary background will be possible within the limits of the present article. Nevertheless, it is important to emphasize that such a praxis should not be considered as simple devotion to the past, but rather as a re-reading, as a means of bringing the past to the present. Time barriers are dissolved from this perspective: the rereading of the past paves the way for a comprehension of the present, simultaneously pointing towards the future. This is the very essence of Friedrich Schlegel’s concept of progression in poetry writing (dichten), a notion very dear to all translator-poets of his generation. Such a concept of progression, of motion, is not only adopted by Schumann during his education — as seen in his activities as a reader, writer, translator and critic; it is also introduced into his music, through his use of tempi, rhythm and dynamics, as the most suitable means of translating his experience in literature. From the 1830s onwards, Robert Schumann was intensely dedicated to music and to his work as a music critic. Nevertheless, his interest in literature, whether translated or not, remained an integral part of his creative activities. It is possible to suggest that from the very moment Schumann opted for music, a new translation process began in his life: the musical rendering of his literary experience as a writer, reader, translator and editor. The most evident and best-known influence of literature and translation on Schumann’s musical work is the extensive use of translated texts for his vocal compositions, such as the oratorio Das Paradies und die Peri, of 1843, based on a libretto by Thomas Moore, translated and adapted by the composer himself, and more than 60 Lieder composed for texts written by Burns, Álvaro de Almeida, Byron, Thomas Moore, Andersen, Shelley and Shakespeare, among others.13 The list of poets Schumann translated into music also includes Mörike, Eichendorff, Goethe, Schiller, Rückert and Heine. This reveals Schumann in two aspects: as a spokesman, in music, for the exponents of German Classicism and for the poets of his generation, and also as an eager reader, who, in his compositions, provided a very personal interpretation of the literary canon of his time.14
The translator as a creative genius
Nevertheless, Schumann’s commitment to literature and to translation is also present in his instrumental works. The so-called piano works of the first phase are examples of this: the piano cycles of literary inspiration like Papillons op. 2, based on the novel Flegeljahre by Jean Paul Richter, the Davidsbündlertänze op. 6, which give musical expression to artists congregated in the Davidsbund,15 and the Kreisleriana op. 16, inspired by the character of the Kapellmeister Kreisler, by E. T. A. Hoffmann, to cite just three examples. Much of his instrumental and symphonic work is also inspired by literature: the opera Genoveva (1848), based on texts by Tieck and Hebbel; the Scenes of Faust, based on Goethe; and the cycle inspired by Goethe’s Wilhelm Meister are only a few examples. Gieseler (1981) points out that the connection between Schumann’s musical work and literature cannot be measured with objectivity, as one could suppose at first. The transference from words to notes does not mean that literature as such will become ‘visible’ in music, but rather that a composer who has a vast literary experience and who worked ‘literarily’, created works of music from all those experiences that marked him as an integral person. (Gieseler 1981: 69; my translation)
Schumann’s accomplishment was therefore to re-write his literary experience as music, translating it into another medium. For this purpose, Schumann does not aim to reproduce the sensations elicited by the literary work — for example, emulating the murmur of rivers or the rustling of trees in the wind; from literature he imports motifs — for example, the split personality, the masquerade as a synthesis of reality and dream — and the formal elaboration. There, we can draw an analogy between the splitting of longer literary narratives into aphorisms and fragments, a characteristic of Romanticism, and the breakdown of the classical form of the sonata into miniature pieces, which allow the composer’s creative genius total freedom of expression. However, while Schumann’s involvement with reading and writing can be traced and assessed in his compositions, they do not provide access to his activity as a translator. Any study that aims to reveal the value and the particularities of his translations is hindered by the fact that only a few fragments of original manuscripts have been compiled, which makes any definitive generalization or judgement impossible. Draheim (1991) comments that such a study would have to overcome considerable difficulties for it would have to find the editions and the commentaries used by Schumann in his translations. Nevertheless, based on a small sample, Draheim says that Schumann’s translations not only […] display a high level of accuracy regarding the metre, to the point of revealing in some passages a certain virtuosic quality; they also often outdo, with the elegance of their formulation, translations printed at the beginning of the 19th century. (Draheim 1991: 42; my translation)
215
216 João Azenha Junior
Similarly, Draheim (1991) points out that: save for a few exceptions, they [the translations] are correct in content and metre, often elegant from the linguistic viewpoint and only very few passages take up that pompous style that, so frequently, renders virtually unbearable so many translations and ‘adaptations’ of ancient lyric. Quite aptly, he [Schumann] can try, at times, to reproduce in German, as far as possible, the exact position within the verse of certain words that were of crucial importance to Horace. By inserting expletive particles or by ingeniously altering mythological names, he creates for himself moments of freedom, which are also opened due to his intentional abandoning of rhyme (Draheim 1991:277–278; my translation)
In any case, a judgement of Schumann’s work as a translator would be premature at this point for lack of solid foundations. It is beyond any doubt, however, that translations played an important part in the formation of the artist, and especially in providing him with the opportunity to try his hand at verse, in a period of his life when he was still torn between literature and music. In this sense, the creative practice of translation enabled him — as well as other romantic German poets involved with translation — to expand his expressive resources, to explore and deepen the potentials of his mother-tongue, to enrich and diversify his work with forms and motifs taken from foreign literatures. But not only this: the practice of translation made it possible for Schumann to experience the contact with the Other. Such a contact is not only meant as an instrument for the construction of a universal literature (Weltliteratur) as defined by Goethe, but as a means of exalting the creative genius, as a means of “fertilising one’s Own through the mediation of the Foreign” (Berman 1984: 16; my translation). This openness of the spirit (der Geist) to the foreign (das Fremde), made possible by translation is a characteristic feature Schumann shares with poets, writers and translators of his time. Such openness reveals a Schumann in tune with his time, open to what foreign worlds have to say, open to universalisation. A Schumann who, in his engagement with literature, bears little resemblance to the alienated spirit depicted by some of his biographers.
Notes 1. “Music is poetry taken to its highest” [my translation]. 2. Cf. Kruse 1991. 3. Cf. Schoppe 1987. 4. Kurt Wagner: Robert Schumann als Schüler und Abiturient. Speech, held at the ceremony in honour of Schumann at the Zwickau Gymasium on March, 15, 1928. Also published as an addendum to Zwickau Gymnasium annual report of the schoolyear 1827/28. Zwickau, 1928. Cited in Draheim 1996: 272.
The translator as a creative genius 217
5. “Solem e mundo tollere videntur, qui amicitiam e / vita tollunt, qua a Diis immortalibus nihil amabi / lius nihil jucundius. Cic: Laelius: /Sansouci vivat!/ Zwickau / January, the 21th / 1823. When you read these few lines, / think of your faithful friend / and classmate: / [my translation of German original]: Rob.: Alex: Schumann. Discp: classis tertiae Lycei Zwickaviennsis”. In: Robert Schumann: Gesammelte Schriften über Musik und Musiker, hrsg. von Martin Kreisig, 2 Bde, Leipzig, 1914, Bd. 1, p. VIII. Kreisig explains that “Sansouci” was the name given by the boys at the Gymnasium to a hill outside Leipzig, where they would meet to play. 6. Robert Schumann: Tagebücher I, Band I (1827–1838), hrsg. von Georg Eismann, Leipzig, 1971, p. 20. [my translation] 7. Cited in Draheim 1996: 273, taken from the Manuscripts (Robert-Schumann-Haus Zwickau, Sign. 4871/VII B, 3-A3). In the note no. 9, Draheim explains that no translations of Anacreon and Homer were found. According to the same author, “Schumann’s keenness on the metre in Antiquity is not only visible in the metrical schemes that he included before his translations in most cases, but also in the excerpts taken from specialized books on the subject, published at the beginning of the 19th century”. [my translation] 8. Taken from the Archives of Robert Schumann House, in Zwickau, VII B, 1 A3. Cited in Schoppe 1987: 9. 9. Robert Schumann. Gesammelte Schriften über Musik und Musiker. Erster Band. Vierte Auflage, mit Nachträgen und Erläuterungen von F. Gustav Jansen. Vorbericht des Herausgebers. Leipzig, 1891, p.VIII und X. Cited in Draheim 1996: 274. 10. Taken from the opening page of the manuscript. 11. Idem. 12. Idem. 13. For a comprehensive list of the translated texts used by Schumann in the compilation of Lieder, cf. Abraham and Sams 1994. 14. For further information on Schumann’s readings and his relationship with the literary canon of the first half of the 19th century in Germany, cf. Kruse 1991. 15. Davidsbund was an association of young artists “created” by Schumann under the inspiration of the novel Serapion’s brothers, by E. T. A. Hoffmann, published in 1821. The association which, as Schumann himself would admit in the preface to the Gesammelte Schriften, existed “only in his mind” aimed to “introduce something new into the universe of fugues and sonatas, which had become academic, namely: poetry in music” (Gieseler 1981: 71). The association’s motto was David’s fight against the Philistines, that is, the young musicians’ fight against the feeble epigones of Classicism. The association was in fact a mixture of reality and fiction. The two most important Davidbündler were Florestan and Eusebius, pseudonyms of Schumann himself and a manifestation of his personality as inspired by Walt and Vult, characters of the novel Flegeljahre, by Jean Paul. Another member of the group was Meister Raro, also a pseudonym of Schumann himself, but — in contrast to Florestan and Eusebius — Meister Raro embodied some characteristic features of Clara’s father (Friedrich Wieck), who would become Schumann’s father-in-law. Gieseler (1981) mentions some other members of the Davidsbund who, in fact, were a cover of the identity of Schumann’s friends, writers and composers. Such unfolding of Schumann’s personality, however, — according to Gieseler — shouldn’t be interpreted as a sign of the illness that would take him to the asylum in 1854, and to death, in 1856. An explanation for all this can be found in the connection between his musical activities and literature: on one hand, the inspiration in Hoffmann’s novel; on the other, his pseudo-identities Florestan (the bold and wild) and Eusebius (the moderate and gentle).
218 João Azenha Junior
References Abraham, Gerald and Sams, Eric. 1994. Schumann. Stuttgart and Weimar: J. B. Metzler. Berman, Antoine. 1984. L’épreuve de l’étranger. Culture et traduction dans l’Allemagne romantique. Paris, Gallimard. Burger, Ernst. 1999. Robert Schumann. Eine Lebenschronik in Bildern und Dokumenten. Mainz: Schott Musik International. Draheim, Joachim. 1991. “Robert Schumann als Übersetzer”. In Robert Schumann und die Dichter. Ein Musiker als Leser. [Katalog zur Ausstellung des Heinrich-Heine-Instituts in Verbindung mit dem Robert-Schumann-Haus in Zwickau und der Robert-Schumann-Forschungsstelle e.V. in Düsseldorf], Bernhard R. Appel and Inge Hermstrüwer, (eds), 41–48. Düsseldorf: Droste. Draheim, Joachim. 1996. “Robert Schumann als Horaz-Übersetzer”. In Worte, Bilder, Töne. Studien zur Antike und Antikerezeption. Bernhard Kytzler zu Ehren, Richard Faber and Bernd Seidensticker (eds), 271–287. Würzburg: Königshausen und Neumann. Gieseler, Walter. 1981. “Schumanns frühe Klavierwerke im Spiegel der literarischen Romantik”. In Robert Schumann, Universalgeist der Romantik, J. Alf and J. Kruse (eds), 62–87. Düsseldorf: Droste. Kruse, Joseph A. 1991. Robert Schumann und die Dichter. Ein Musiker als Leser. [Katalog zur Ausstellung des Heinrich-Heine-Instituts in Verbindung mit dem Robert-Schumann-Haus in Zwickau und der Robert-Schumann-Forschungsstelle e.V. in Düsseldorf]. Düsseldorf: Droste. Kruse, Joseph A. 1991. “Robert Schumanns Lektüre. Zeitgenössischer Kanon, individuelle Schwerpunkte, kompositionsspezifische Auswahl und seine Urteile als Leser”. In Robert Schumann und die Dichter. Ein Musiker als Leser. [Katalog zur Ausstellung des Heinrich-Heine-Instituts in Verbindung mit dem Robert-Schumann-Haus in Zwickau und der Robert-Schumann-Forschungsstelle e.V. in Düsseldorf], Joseph A. Kruse. Düsseldorf: Droste. Meier, Barbara. 1995. Robert Schumann. Reinbeck bei Hamburg: Rowohlt. Nauhaus, Gerd. 1988. “Robert Schumann”. In Kindlers Neues Literatur Lexikon, Walter Jens (ed), Bd. 15, 40–42. München: Kindler. Schoppe, Martin. 1987. “Schumanns frühe Texte und Schriften”. In Schumanns Werke — Text und Interpretationen: 16 Studien, Akio Mayeda and Klaus Wolfgang Niemöller (eds). Mainz: Schott. Schumann, Robert. 1891. Gesammelte Schriften über Musik und Musiker. Vierte Auflage, mit Nachträgen und Erläuterungen von F. Gustav Jansen. Leipzig. Schumann, Robert. 1914. Gesammelte Schriften über Musik und Musiker, ed. Martin Kreisig. Leipzig. Schumann, Robert. 1971. Tagebücher I, Band I (1827–1838), ed. Georg Eismann. Leipzig.
Übersetzung zwischen Nationalismus und Internationalismus Luc van Doorslaer Lessius Antwerpen, CETRA Leuven
Translation is a phenomenon that by definition crosses borders and thinks internationally. Still, despite all the efforts to go beyond especially national restrictions, Translation Studies cannot deny the importance of the tensions between national, transnational, multinational and international thinking in our discipline. Important and interesting cases in this respect are territories or countries with an institutionalized multilingualism. It’s probably not a coincidence that Translation Studies flourish in countries like Finland, Canada or Spain/Catalonia. This article also describes certain aspects of the problematization of language and translation borders and classifications in the Dutch language territory (the Netherlands and Belgium). Translations have played an important role in creating an identity that searched for a balance between national and international tendencies. The current language situation in some EU committees and working groups is a revealing example of the use of translation in the politics of power and image. The topics of identity and (hidden) nationalism in translation still offer many challenges to Translation Studies as an interdiscipline. Die Wege der Nationen, ihr Mit- und Gegeneinander, auch die Konflikte und die Narben, die davon zeugen, haben die neuere Geschichte bestimmt, und aus der Zukunft werden sie schwerlich verschwinden. Wir mögen Europäer und womöglich sogar Weltbürger werden, aber es kann uns einzig als Deutschen, Dänen, Polen, Niederländern, Franzosen, Italienern gelingen. Die Völker bleiben das Fundament, und wenn wir sie übergehen, wird wenig Bestand haben, was wir erstreben. (Krockow 2000: 8)
In einer immer internationaler werdenden Welt mutet es immer ein bisschen altmodisch und sogar konservativ an, sich noch mit Nationen und Nationalem zu beschäftigen. Gerade wo es sich um Übersetzung handelt, ein Verfahren, das doch
220 Luc van Doorslaer
per definitionem grenzüberschreitend und zumeist international sein sollte? Trotzdem habe ich in letzter Zeit bei mehreren Übersetzungskongressen und -kolloquien festgestellt, dass Begriffe wie transnational, multinational oder international nicht immer differenziert und eindeutig benutzt werden. Versucht wird immer, das Nationale zu übersteigen, gleichzeitig aber führt dieser Versuch offensichtlich öfter zu Unsicherheiten und Unklarheiten, die sich in den meisten Fällen nicht auf die Präfixe trans-, multi- oder inter-, sondern vielmehr auf die wechselnden Inhalte des Begriffes ‘national’ beziehen. Wir interessieren uns alle für das Übersetzen, das Überqueren von Grenzen also. Auf jeden Fall schon mal Text- und Sprachgrenzen, gelegentlich aber auch Staats- bzw. Nationengrenzen (wo immer diese sich auch befinden). Es ist schon öfter darauf hingewiesen worden, dass die Grenzproblematik innerhalb der Übersetzungswissenschaft in so genannten ‘einheitlichen’ Sprachgebieten, d.h. wo Staats- und Sprachgrenzen zusammenfallen (z.B. Schweden oder Norwegen) viel weniger thematisiert wird. In Staaten dagegen, die sich tagtäglich mit der Problematik einer institutionalisierten Mehrsprachigkeit auseinander zu setzen haben, ist man für die Feinheiten der Themen Grenzüberquerung und Übersetzung wohl empfindlicher. Ist es Zufall, dass in letzter Zeit in den translation studies so viele Finnen aktiv sind? Finnland, ein Land mit einer breit angelegten institutionellen Zweisprachigkeit, obwohl es sich bei den schwedischsprachigen Finnen um eine kleine Minderheit handelt. Gleiches gilt für die wachsende Zahl der Spanier in unserer Disziplin, nicht wenige davon aus Katalonien. Schon deutlicher als im Falle Finnlands (obwohl auch da) war die spanisch-katalanische Auseinandersetzung (besser gesagt: die kastilisch-katalanische) in ihrer historischen Perspektive eine Frage, bei der Sprache als Mittel zur politischen und kulturellen Macht bzw. Machtausübung benutzt worden ist. Und Sprachen in einem Machtkampf führen auch oft zu interessanten Verhaltensmustern beim Übersetzen. Das geht auch aus einem weiteren geographischen Beispiel hervor: Kanada. Hier ist der sprachliche Machtkampf offensichtlich. Aus übersetzerischer Sicht ist aber interessant, dass die Lage hier besonders kompliziert ist. Es gibt nicht nur den bekannten Gegensatz zwischen Englisch- und Französischsprachigen, es besteht innerhalb des französischsprachigen Teils zudem eine weitere Empfindlichkeit, weil sich die Québécois auch gegen die ‘Sprachunterdrückung’ und Sprachdominanz von Frankreich wehren. So beschreibt Brisset den für Übersetzungsforscher recht merkwürdigen Fall einer Macbeth-Übersetzung “traduit en québécois” (Brisset 2000: 346). Merkwürdig aus dem Grund, weil zumeist die Ausgangssprache bei einer Übersetzung erwähnt wird, hier dagegen ist es die Zielsprache. In einem Versuch, einer gewissen Variante des Französischen einen höheren Status zu verleihen, heißt es nicht “traduit en français”, sondern “en québécois”. Hier wird eine Übersetzung ganz klar in einem politischen Machtkampf eingesetzt, um Grenzen und unterschiedliche Verhältnisse innerhalb eines Sprachgebiets (statt eines Staatsgebiets) zu betonen.
Übersetzung zwischen Nationalismus und Internationalismus 221
Die Betonung des nationalen Denkens ist bei diesem Übersetzungsbeispiel der internationalen, grenzüberschreitenden, in diesem Fall sogar der interkontinentalen Sprachsolidarität überlegen. Nationalist ideology rejects the notion of Quebec French being “international”. In this context, the word “international” has a negative connotation and reveals a desire to exclude; the “multicultural” and the “transcultural” are negative values, to be fought at all costs. Suddenly characterized as “international”, French has been defined as, and deliberately made into, a foreign language. Such an ideology emphasizes the illegitimacy of French, claiming that it is neither heard nor understood in Quebec. And proof of this assertion is to be found in the speech of the ordinary Québécois. (Brisset 2000: 357)
Unklarheiten bei der Grenzziehung und daher auch bei den Klassifizierungsversuchen gibt es auch reichlich im niederländischen Sprachgebiet: niederländisch, holländisch, flämisch, flandrisch, belgisch, nordniederländisch, südniederländisch, … all diese Begriffe sind mir begegnet, als ich die niederländischsprachige Literatur in deutscher Übersetzung am Anfang des 20. Jahrhunderts untersucht habe (s. van Doorslaer 2004). Auf den ersten Blick wurden diese Begriffe damals in deutschen Zeitschriften chaotisch und durcheinander benutzt, bei näherer Betrachtung aber waren doch einige relativ konsequent angewandte Kategorien zu unterscheiden. So wurden Übersetzungen aus der ‘flämischen Literatur’ fast immer wegen ihrer Darstellungskraft und ihrer Lebensfreude gelobt, wobei öfter auf die Tradition der Malerei und der Dorfgeschichte hingewiesen wurde. ‘Französische Literatur in Belgien’ (nie ‘wallonische Literatur’ genannt) hatte angeblich ganz andere Qualitäten. Sie war analytischer und durchdachter, in Kriegszeiten wurden dieselben Qualitäten dann aber auch dekadent oder volksfremd genannt. Innerhalb des niederländischen Sprachgebiets gab es damals, in der ersten Hälfte des vorigen Jahrhunderts, also eine klare Trennung zwischen den Begriffen ‘holländisch’ (verweisend auf das Königreich der Niederlande) und ‘flämisch’ (verweisend auf die nördliche Hälfte Belgiens). Obwohl es sich in den beiden Fällen um die niederländische Schriftsprache handelt, erlebten die deutschen Übersetzungen beider Gebiete eine ganz unterschiedliche Erfolgsperiode. Die holländische Literatur war vor allem um die vorige Jahrhundertwende herum populär, seit dem Ersten Weltkrieg aber war es dann jahrzehntelang vielmehr die flämische Literatur (u.a. durch das Phänomen Felix Timmermans). Und seit den achtziger Jahren des 20. Jahrhunderts tritt eine neue Akzentverschiebung in Richtung Königreich der Niederlande auf (durch Autoren wie Cees Nooteboom und Harry Mulisch). Voor de vertalingen van Nederlandstalige literatuur in Duitsland is 1914 een cruciaal jaar: tussen 1890 en 1914 werd er vooral literatuur uit Nederland — waaronder veel toneel — vertaald en na 1914 vooral literatuur uit Vlaanderen. De literatuur uit Vlaanderen zou in de 20ste eeuw in Duitsland dominant
222 Luc van Doorslaer
blijven, zeker wat het aantal drukken betreft. Pas aan het begin van de 21ste eeuw, die volgens historici in 1989 begon, zou de literatuur uit Nederland — grotendeels dankzij het Schwerpunkt op de boekenbeurs in Frankfurt am Main in 1993 — weer veel dominanter worden op de Duitse markt. (Grave 2000: 113) Für die Übersetzungen der niederländischsprachigen Literatur in Deutschland ist 1914 ein entscheidendes Jahr: Zwischen 1890 und 1914 wurde hauptsächlich Literatur aus den Niederlanden — darunter viele Theaterstücke — übersetzt und nach 1914 vor allem Literatur aus Flandern. Die Literatur aus Flandern dominierte sicherlich was die Anzahl der Ausgaben betrifft das 20. Jahrhundert in Deutschland. Erst am Anfang des 21. Jahrhunderts, das laut den Historikern 1989 begann, wurde die Literatur aus den Niederlanden — zum Großteil dank des Schwerpunkts auf der Frankfurter Buchmesse 1993 — auf dem deutschen Markt wieder viel dominanter. (Übersetzung LvD.)
Mittlerweile zeigt sich auch eine andere höchst interessante Entwicklung. Die nationalen Kategorien ‘holländisch’ und ‘flämisch’ sind seit dem zweiten Weltkrieg allmählich aus der Übersetzungspraxis verschwunden. Man liest jetzt fast nur noch “aus dem Niederländischen übersetzt”, auch wenn es sich um flämische Autoren handelt. Die nationalen Grenzen sind aufgegeben worden, und das nicht nur, weil zwischen den Niederlanden und Flandern jetzt eine offizielle ‘Niederländische Sprachunion’ besteht. Sicherlich spielen hier auch wirtschaftliche Gründe eine nicht unwichtige Rolle. Die Verlagswelt funktioniert ja zum Großteil großniederländisch. Man könnte hier von einer Tendenz der ‘Grenzverwischung’, von einer kleinen internen Globalisierung innerhalb des niederländischen Sprachraums sprechen: die ‘Niederen Lande’ zusammen, the Low Countries (es war auch wohl kein Zufall, dass diese beiden Länder zum ersten Mal erfolgreich international und grenzüberschreitend eine Fußball-Europameisterschaft, Euro 2000, haben organisieren können…). Dennoch, als allerneueste Tendenz zeichnen sich jetzt gelegentlich auch gegensätzliche Entwicklungen ab. Ich gebe ein Beispiel aus dem Übersetzungsbereich. Vor einigen Jahren haben der flämische Autor Tom Lanoye und der Theatermacher Luk Perceval im Rahmen eines riesigen Theaterprojekts mit dem Titel Ten oorlog die Königsdramen Shakespeares übersetzt bzw. bearbeitet. Das Ergebnis war ein sehr flexibles stilistisches Spiel mit einer Vielzahl von Sprachvarianten, Soziolekten, Regiolekten, Jargons usw. Diese komplexe Bearbeitung wurde von Rainer Kersten und Klaus Reichert wiederum ins Deutsche übersetzt und erschien 1999 unter dem Titel Schlachten! Auffallend ist, dass in dieser Veröffentlichung, soweit ich weiß, zum ersten Mal seit längerer Zeit, wieder stand: “aus dem Flämischen übersetzt”. Einerseits lässt sich diese Hinzufügung durch die spezifischen Qualitäten dieses
Übersetzung zwischen Nationalismus und Internationalismus 223
Werks erklären und zwar hauptsächlich aufgrund der sprachlichen Diversität in Flandern. Andererseits ließe sich auch fragen, ob eine solche Hinzufügung in einer Übersetzung als Beispiel einer Renationalisierungs- oder Reregionalisierungstendenz betrachtet werden könnte. Jetzt, wo Europa so allmählich weiß, dass die Flamen, die Niederländischsprachigen in Belgien, nicht eine Minderheit, sondern die Mehrheit bilden, fängt man damit an, die Unterschiede gegenüber den Sprachbrüdern im Norden wieder zu betonen, sich zu re-minorisieren (vgl. Aguilar-Amat & Santamaria 2000: 74 zu der Relativität der Begriffe ‘Minorität’ und ‘minorisiert’)? Diese Tendenz könnte man auch anhand verschiedener anderer gesellschaftlicher Beispiele in dem Zusammenspiel zwischen den Niederlanden und Flandern illustrieren, die hier aber zu weit führen würden. Wie paradox es auch sei, die gesellschaftliche Wirkung des Nationalbegriffs scheint auch in der Übersetzungsrealität größer, als manchmal angenommen wird. Die Funktion von Übersetzungen mit Bezug auf das delikate Gleichgewicht zwischen national und international ist gesellschaftlich zu deuten. Interessant wäre es, den Zusammenhang mit der ideologischen Katalysatorfunktion der literarischen Übersetzung zu erläutern, wie sie von Peter Zima beschrieben worden ist. [L]iterarische Übersetzung [ist] nicht unabhängig von ihrem sozio-linguistischen Kontext zu verstehen […]: zumal sie nicht nur strukturelle — phonetische, semantisch, syntaktische -, sondern auch funktionale, d.h. gesellschaftliche Aspekte aufweist. Eine der wesentlichen gesellschaftlichen Funktionen literarischer Übersetzungen ist die Konsolidierung bestimmter kultureller oder ideologischer Entwicklungen und Trends. Das Werk eines Autors, der nahezu unbekannt oder nur vom Hörensagen bekannt ist, wird übersetzt und wirkt oftmals als Katalysator in ästhetischen oder politischen Diskussionen. (Zima 1992: 231)
Als Ergänzung zu unserem modernen Drang zur Globalisierung und Internationalisierung ist öfter festzustellen, dass sich Identifikation, z.T. auch Bewusstseinsbildung auf einer beschränkteren, regionalen Ebene abspielt. Internationalisieren, entnationalisieren heisst nicht notwendigerweise auch entregionalisieren. Ein gewisses Bedürfnis an Stereotypen und Images ist offensichtlich sehr menschlich. Ich glaube, dass gerade die Übersetzungswissenschaft in diesem Bereich sehr viel interessantes Material zur Verfügung stellen könnte. Als go between ist sie die vermittelnde Disziplin par excellence, die die Spannung zwischen dem nationalen und dem internationalen Denken und Handeln thematisieren könnte. Die Sprachenverhältnisse innerhalb der Europäischen Union bilden hier wohl das beste Beispiel. Die Ausgangslage einer komplizierten und vielschichtigen (erwünschten) Identitätsbildung, auch im Bereich der Sprachenvielfalt, wird hier in eine konkrete Sprach- und Übersetzungspolitik umgesetzt. Aufgrund eines konsequent durchgeführten demokratischen Prinzips werden die Nationalsprachen beim Ausbau
224 Luc van Doorslaer
einer internationalen bzw. internationalisierten (europäischen) Identität eingesetzt. Gleichzeitig aber wird dieses Prinzip auch wieder in Frage gestellt, wenn in bestimmten EU-Ausschüssen oder Arbeitsgruppen mit Blick auf die Effizienz nur ‘Arbeitssprachen’ benutzt werden. Die Kriterien für die Auswahl dieser Arbeitssprachen sind aber gelegentlich undurchschaubar, weil sie eine Mischung von Pragmatik, Bedeutung, Geschichte und Image sind. Wer maximale Effizienz anstrebt, müsste sich auf nur eine Sprache beschränken. Das Englische wird international am meisten benutzt, das Deutsche ist die quantitativ wichtigste Muttersprache innerhalb der EU (und deswegen die demokratischste?), das Französische ist aus historischer Sicht die bedeutendste Sprache in diplomatischen Kreisen usw. Eine Fülle von zum Teil national bedingten Argumenten wird aufgeführt, um die privilegierte Stellung der eigenen Sprache zu verteidigen. Im Sog dieses Denkens versuchen auch die ‘kleineren’ EU-Sprachen, sich als die grösseren oder wichtigeren unter den kleineren darzustellen und so ihrerseits ebenfalls einige Privilegien einzufahren. Sprach- und Übersetzungsverhältnisse werden somit im macht- und imagepolitischen Spiel der Nationalstaaten eingesetzt. Gerade im letzten Jahrzehnt ist die Problematik der Identitätsbildung und -veränderung in Verbindung mit dem Kulturübergang beim Übersetzen allmählich auch Objekt der Übersetzungsforschung geworden (“The 1990s witness a series of historical studies that explore the identity-forming power of translation, the ways in which it creates representations of foreign texts that answer to the intelligibilities and interests of the translating culture” — Venuti 2000: 337). Denken wir dabei beispielsweise an den postkolonialen Ansatz in der Übersetzungswissenschaft, der zur Beschreibung von aus machtpolitischer Sicht höchst interessanten Fallbeispielen geführt hat. Übersetzung schafft ein bewusstes Umgehen mit der Geschichte und der Kultur, schafft somit eine Identität, wird gelegentlich aber auch bewusst dabei eingesetzt. Das traditionelle Beispiel, das u.a. von Tymoczko in diesem Zusammenhang aufgeführt wird, ist die Übersetzung irischer Literatur ins Englische (Tymoczko 2000). Dennoch ist das Verhältnis zwischen institutionellen Bedingungen und dem Einsatz von Übersetzungen in der Literatur, aber beispielsweise auch bei der internen und externen Kommunikation von lokalen, regionalen, überregionalen, nationalen und internationalen Behörden und sonstigen Institutionen von der Übersetzungswissenschaft noch relativ wenig untersucht worden. Wie verhalten sich beispielsweise nationale und internationale Beweggründe in der EU zueinander? Mit welcher Intensität werden die internen Machtverhältnisse durch den Einsatz (vielleicht vor allem Nicht-Einsatz?) von Übersetzungen beeinflusst? Inwieweit kann die bestehende Lage der offiziellen Sprachen in der EU unter den Druck der sprachlichen Entwicklungen innerhalb der Nationalstaaten geraten, wo Minderheitssprachen gelegentlich neue Rechte erwerben? So beschreibt zum Beispiel Millán-Varela 2000 den Zusammenhang der Übersetzungs- und Identitäts-
Übersetzung zwischen Nationalismus und Internationalismus 225
frage in Galizien, wobei sie auf eine bemerkenswerte Entwicklung und Veränderung in der Haltung der Galizier gegenüber den anderen Mittelmeersprachen hinweist. Solche Fragen könnten auf den ersten Blick die traditionellen Grenzen der Übersetzungswissenschaft sprengen, andererseits könnten sie durch lehrreiche Querverbindungen mit den cultural studies vielleicht auch die Flexibilität und kommunikative Ausrichtung unserer Disziplin betonen. Is all this also an issue for translation and translation studies? Is it relevant at all to examine in what kind of a society (linguistic community?) and in what kind of communication environment translation takes place? For theoreticians and trainers from the 1960’s, the answer would probably have been quite simple, and many translation trainers and translation training curricula are still satisfied with answers from the 1960’s. For all those who have supported the many “cultural turns” in the discipline and the shift into functional approaches, the consequences of the new environment are enormous. Rather than strict technical characteristics of the translation activities and processes, it is the overall cultural framework that keeps changing, as well as the position of translations […]. But even the questions and answers from the 1970’s need to be updated. (Lambert 2004)
Unsere Identität ist vielseitig und komplex, daher wird sie von Michael Cronin, unter Verweis auf Edgar Morin, als eine Poly-Identität beschrieben: eine Mischung aus Regionalem, Nationalem und Kontinentalem, mit Aspekten einer Klassen-, Geschlechts- sowie Rassenidentität (Cronin 2000: 18). Die Identität eines Übersetzers ist, aufgrund seiner Arbeit, genauso vielseitig und zweifelhaft. Die Identität einer Übersetzung ist polyvalent und zumindest zweideutig. Die Identität der Übersetzungswissenschaft ist grundsätzlich eine Poly-Identität, sie ist multi- und womöglich interdisziplinär. Ihre Vielseitigkeit sollte allerdings auch dazu führen, dass über die Querverbindungen etwa zur Identitäts- und Nationalproblematik das Wissen über das Phänomen Übersetzung selber ausgedehnt wird.
References Aguilar-Amat, Anna & Santamaria, Laura. 2000. “Terminology policies, diversity, and minoritised languages”. In Translation in Context. Selected Contributions from the EST Congress, Granada 1998, Andrew Chesterman, Natividad Gallardo San Salvador, Yves Gambier (eds.), 73–84. Amsterdam/Philadelphia: Benjamins. Brisset, Annie. 2000. “The search for a native language: Translation and cultural identity”. In The Translation Studies Reader, Lawrence Venuti (ed.), 343–375, London/New York: Routledge. Cronin, Michael. 2000. Across the Lines. Travel, Language, Translation. Cork: University Press. Doorslaer, Luc van. Geplant 2004. Übersetzungen der ‘kleinen’ flämischen in der ‘großen’ deutschen Literatur (vorläufiger Titel). Amsterdam: Rodopi.
226 Luc van Doorslaer
Grave, Jaap. 2000. “Paul Raché en Cyriel Buysse. Vroege recensies van Buysses werk in Duitsland.” Mededelingen van het Cyriel Buysse Genootschap, Nr. XVI, 113–140. Krockow, Christian Graf von. 2000. Über die Deutschen. München: List. Lambert, José. Geplant 2004. “Responsibilities in No Man’s Land: Translation as language and communication in the New Millennium”. Also on CETRA website: fuzzy.arts.kuleuven.ac.be/cetra Millan-Varela, Carmen. 2000. “Translation, normalisation and identity in Galicia(n).” Target, 12:2, 267–282. Tymoczko, Maria. 2000. “Translation and political engagement. Activism, social change and the role of translation in geopolitical shifts”. The Translator, Vol. 6, Nr. 1, 23–47. Venuti, Lawrence. 2000. “1990s”. In id., The Translation Studies Reader. London/New York: Routledge. 331–342. Zima, Peter V. 1992. Komparatistik. Einführung in die vergleichende Literaturwissenschaft. Tübingen: Francke.
Non-verbal phenomena in simultaneous interpreting Causes and functions Barbara Ahrens University of Mainz — Faculty of Applied Linguistics and Cultural Studies (FASK) Germersheim, Germany
In simultaneous interpreting, the source-language speaker can use gestures, facial expressions, proxemics, prosody and graphic material, whereas the interpreter, who is only perceived by her/his listeners through her/his voice in the headphones, has a more limited use of nonverbal elements such as prosody and suprasegmental elements, including intonation, stress, pauses and delivery speed. Ideally, according to Kirchhoff, the interpreter should sound like a speaker expressing him/herself spontaneously and naturally, but such output is difficult to find in real settings. In fact, as noted by Shlesinger and Williams, the interpreter’s situation with overlapping source-speech perception and target-speech production seems to generate unnatural, typical patterns. The paper discusses the possible reasons for this fact and describes the resulting nonverbal phenomena in the target speech on the basis of an authentic English-German corpus which was part of the author’s PhD project at the FASK in Germersheim.
Introduction In oral communication, the message of an utterance is transmitted via various communication channels: Due to their complementary and compensatory function, nonverbal elements such as facial expressions, gestures, proxemics, prosody as well as graphic material can help the speaker to convey the message he/she wants to get across to his/her communication partners (cf. Huber 1996: 259, Pöchhacker’s definition of text in simultaneous interpreting [1994: 95ff.], Poyatos 1987: 77ff.). Simultaneous interpreting (SI) is a very specific form of oral communication; the verbal and nonverbal elements of texts that are interpreted are equally important. The situation of nonverbal elements is particularly complex because those of the
228 Barbara Ahrens
source text (ST) and those of the target text (TT) are interwoven: The ST producer is completely free as to the means he/she chooses for encoding his/her message. The simultaneous interpreter, however, depends on the ST speaker’s message (cf. Kalina 1998: 108) and has no opportunity to influence the way it is presented. Since the simultaneous interpreter sits in a booth, the audience hears him/her only via headsets. Thus, during the production of the TT, his/her nonverbal communication — at least the part which is perceived by the audience — is limited to prosodic and suprasegmental elements such as intonation, stress, pauses and speed, whereas the audience’s visual — and, to some extent, also acoustic — input results directly from the ST speaker’s facial expressions, gestures, paralinguistic features etc. The interdependence between the nonverbal elements of the ST and those of the TT is even more evident if the ST speaker’s talk is accompanied by overhead transparencies, a beamer or a video presentation (cf. Ahrens 2002: 37ff.). The simultaneous interpreter is a recipient of the ST although he/she is not its genuine addressee (cf. Kalina 1998: 108). This results in complex relations between all communication partners: The ST speaker’s input has a direct impact on those persons among the audience who can understand the ST language. All persons listening to the ST who are not able to understand the ST language are addressees of the ST, but have to rely on the interpreter’s rendering of the ST message in the TT language. The ST speaker’s impact on them is direct with respect to kinesic and graphic elements, as well as some paralinguistic and non-linguistic information (e.g. voice quality, cf. Crystal 1969: 100ff.) and indirect with respect to linguistic input, for which they depend on the interpreter. It goes without saying that the features of nonverbal communication are susceptible to misunderstandings and misinterpretations since they depend on the culture they represent (cf. Poyatos 1987: 103ff.). This specific situation and the overlapping perception of the ST and the production of the TT seem to be the reason why the interpreter’s output is often marked by some salient features typical of SI (cf. Ahrens 1999 and 2002, Shlesinger 1994, Williams 1995), although in the literature, the ideal case of simultaneously interpreted text is described as sounding like spontaneous speech (cf. Kirchhoff 1976: 67).
Corpus and methodology For the PhD project presented here, it was possible to record the material to be analysed in an authentic setting: During a guest lecture for the students of the Faculty of Applied Linguistics and Cultural Studies (FASK) in Germersheim given in English (72 minutes) the performance of three parallel booths with professional interpreters all working from English into German were recorded. They had been hired to interpret for the students attending the lecture. The interpreters were all professional conference interpreters working as “freelancers” (with an average
Non-verbal phenomena in simultaneous interpreting 229
experience of 4,6 years). They had all been trained either at the Institute for Translating and Interpreting (IÜD) at Heidelberg University or the Faculty of Applied Linguistics and Cultural Studies (FASK) in Germersheim (University of Mainz). Four of them had English as their B language, two as their C language, i.e. they were all trained and used to working from English into German. They knew that they were going to be recorded but did not know the objective of the study. After their performance all interpreters filled in a retrospective questionnaire on the problems they had with the ST and how they dealt with them (problem-solving strategies, references to specific paragraphs of the text etc.). Since the lecture was given in one of the conference rooms at the FASK in Germersheim, the technical equipment for audio and video recordings of both the speaker and the booths was used. Dual track recordings were made of the ST and the three TTs. Thus it was possible to synchronize the two channels in order to draw conclusions about the ear-voice-span (EVS) and its influence on prosodic features of the TT. The ST speaker was a native British English speaker who was used to giving lectures in public. During his talk on an actual German-English translation job in the field of marketing communication, he spoke spontaneously without reading from his manuscript, which he used only for short quotations and for planning how to proceed in his speech. The speaker was aware that his lecture was going to be interpreted. The ST and the way it was presented can be described as rather informal, containing jokes and little polemic remarks, situational references, such as questions to the audience or comments about the microphone cable he was tripping over all the time, repetitions of what he had already said in order to think about the next thing he was going to say. As to paralinguistic and prosodic features, the ST presentation shows the typical characteristics of spontaneous speech: sudden variations in the speech rate (cf. Goldman-Eisler 1958:61), hesitation phenomena such as filled pauses (cf. Cruttenden 1997: 174) or drawled syllables (cf. Crystal 1969: 154) as well as many false starts due to changes in the planning of the respective utterance. This replanning affects not only sentences or words but also the order of arguments or topics (cf. Antos 1982: 183, Goldman-Eisler 1958: 66ff. and 1961: 171). Word-for-word transcripts of the ST as well as the TTs were made. The audio data was digitized, sampled and submitted to signal analysis in order to make pitch movement and other prosodic features visible. The analysis was carried out using PRAAT, a computer programme especially developed and designed for speech analysis by Paul Boersma and David Weenink at the Phonetic Sciences Department of the University of Amsterdam (cf. Lieshout 2000). Since the two channels of the dual track recording can be synchronized, it is possible to draw conclusions about pause patterns, the EVS etc. The autonomous effect of the TT can be studied without the disturbing interference of the ST channel (cf. Ahrens 1999:52) in order to see if the intonation “sui generis” (Shlesinger 1994: 226) in SI can also be confirmed
230 Barbara Ahrens
in longer samples and language pairs other than the ones investigated so far (cf. Alexieva 1988, Daró 1990 and 1994, Shlesinger 1994, Williams 1995).
Pitch movement and prominent structuring markers in the TT In her experiment, Shlesinger (1994) observed that “the low-rise nonfinal pitch movement” (Shlesinger 1994: 231) dominated in her TT samples of SI, even in positions where it would seem more likely to have a falling pitch movement (cf. Shlesinger 1994: 231). This feature was confirmed by the analysis presented here. Although this pitch movement seems to be a characteristic phenomenon of SI, there are hardly any comments on it in the literature on SI and the delivery of the TT. A large number of studies on quality in SI confirm that for many users of SI, the form of the TT is less important than its content (cf. Bühler 1986, Gile 1990, Kurz 1989 and 1993, for a comprehensive overview see Collados Aís 1998). One exception is Kurz/Pöchhacker’s study on expectations and requirements for SI on TV (cf. Kurz/ Pöchhacker 1995). Shlesinger’s study is the only one to consider the low-rise nonfinal intonational contour and its impact on the audience (cf. Shlesinger 1994: 231). Pöchhacker’s text description profile refers to the paraverbal dimension of text delivery (e.g. melody or rhythm) describing it with categories such as “noticeable” or “very noticeable”. But it does not explain the individual prosodic features for which the melody of a spoken text, for example, is perceived or evaluated as being “native” or “noticeable” (cf. Pöchhacker 1994: 112ff.). This situation gives rise to the hypothesis that the low-rise nonfinal intonational contour is regarded as the “normal” case in SI (by interpreters themselves as well as the audience) because nobody ever mentions it. Due to the fact that this characteristic pitch movement does not ideally mark the information structure of the ST — which is prosodically defined by the ST speaker’s pitch movement, pausing, phrase final lengthening and terminal junctures (cf. Huber 1996: 266) — it seems probable that simultaneous interpreters use other prosodic features to signal that there is a new unit of information, even if the preceding unit does not have a clear boundary. In the corpus, there are salient prosodic features that evidently serve this purpose. Two examples will be discussed here. In the examples presented below, the following methods of transcription have been applied: conSIdered ·0.32Ò ·0.32:AÒ [xyz] azosia-
stressed syllable written in capitals pause, duration: 0.32 seconds pause due to breathing, duration: 0.32 seconds context of the parts that are shown in Figures 1 to 4 false start
Non-verbal phenomena in simultaneous interpreting
For legibility reasons, pauses have only been indicated in the passages corresponding to the figures, not in the context given with them in the transcriptions. Figures 1 to 4 represent the pitch movement of the examples discussed here. The black line in the graphs represents pitch, the dashed line intensity. In examples (1) and (2) taken from the German TTs the English gloss is given in the respective lines below the German utterance. Example (1) In the ST paragraph in question, the speaker said: [… KOMpetenz um Computer und Telekommunikation aGAIN a HIGH tech or a COMplex process linked to competence is perhaps more EAsily mh underSTOOD COMpetence in everything to do with computers and telecommuniCAtions so WHY is there this ability to do] something WELL or efFECtively WHY is it conSIdered to BE ·0.32:AÒ an ADvertising [MESsage in the GERman CULture…]
Figure 1.Example (1) — ST pitch movement
Figure 1 clearly shows that the ST speaker finishes the intonation group (cf. Cruttenden 1997: 29) with the word “efFECtively” using a falling-rising-falling tone on its nucleus (cf. Crystal 1969: 207ff.). The immediately following word “WHY” is the onset syllable of a new intonation group, i.e. “the first prominent syllable (the ‘onset’) of any stretch of utterance definable as a tone unit [i.e. intonation group — B. A.] […]”(Crystal 1969: 143). One of the characteristics of the onset syllable is that a speaker returns to the pitch-level on which he/she normally starts a new intonation group (cf. Crystal 1969: 227). This is evident in Figure 1. In this example, the
231
232 Barbara Ahrens
boundary of the intonation group is not marked by a pause. Nevertheless, the following intonation group starting with “WHY” can be identified by other prosodic features such as declination in the previous intonation group (cf. Cruttenden 1997: 120) and the onset since these are also cues to the information structure underlying certain prosodic patterns (cf. Huber 1996: 266). In the TT paragraph corresponding to this ST paragraph, the interpreter says: [… bei der TElekommunikation ist das] ·0.64Ò wahrSCHEINlich in the telecommunication is this probably ·0.20:AÒ verSTÄNDlicher ·0.44:AÒ GUT WArum IST more.understandable good why is [KompeTENZ in DEUTSCHland eine WERbebotschaft …] competence in Germany an advertising.message
Figure 2.Example (1) — TT pitch movement
As can be seen in Figure 2, the interpreter finishes the intonation group with a rising pitch movement, although a native German-speaking audience would rather expect a falling tone due to the content of this phrase. One of the reasons for this rising pitch movement may be that the interpreter has a rather long EVS at this moment. The analysis shows that the interpreter has an EVS of 8,25 seconds, which is long especially when taking into account the speech rate of the ST speaker: During this period of 8.25 seconds, he utters 44 syllables, which means an average duration of 0.1675 seconds per syllable (plus three very short pauses with durations of 0.1, 0.38 and 0.4 seconds). After the rising pitch movement and a short pause for breathing (0.44 seconds, see Figure 2), the interpreter starts with a new intonation group “GUT” continuing with the most recent part of the ST and thus omitting the phrase “so WHY
Non-verbal phenomena in simultaneous interpreting 233
is there this ability to do something WELL or efFECtively”. By doing so, he reduces the EVS to 1.37 seconds (measured between “WHY” in the ST and “WArum” in the TT). The prominent structuring marker “GUT” could therefore be regarded as 1. a new onset indicating that a new intonation group establishing a new unit of information begins, i.e. it is used in a typical way as an opening marker (cf. Gülich 1970: 108ff.) and 2. a means of resuming the part of the ST that is omitted in order to establish or maintain cohesion. The first case — giving prominence to structuring markers in order to signal the onset of a new intonation group, although the former has not been finished with a tone clearly indicating the end of an intonation group — is a very frequent feature in the corpus (the study presented here is still in progress, and therefore only preliminary statistical data can be provided). Since the segmentation of the TT into intonation groups reflects the information structure of the ST (cf. Huber 1996: 266, Cruttenden 1997: 81ff.), the prominent structuring markers discussed here help to structure the progression of the ST message, maintaining the cohesion of the text. They are not necessarily induced by the ST, i.e. imitating the structuring signals used by the ST speaker. Example (2) The case of ST-induced prominence of structuring markers can be seen in the following example: [… working with] connoTAtion and asSOciation THIS is the adVERtisement for ROver […] The ST speaker uses the prominent demonstrative pronoun “THIS” as a structuring marker in order to proceed to the next unit of information, which is meant to be an illustration of what he just said. Therefore, a new intonation group beginning with the typical onset syllable is uttered. The former intonation group ending with “… and asSOciation” has a falling tone. Here again, the onset of the new intonation group which resets the declination of the former intonation group, as shown in Figure 3, identifies the intonational structure and its relation to the content of this ST paragraph. The interpreter imitates this structure in the TT: […] man ARbeitet mit Azosia- ASsoziatiONen und KONnotationen one works with azosia- associations and connotations ·0.32:AÒ ALso wie zum BEIspiel in der [Anzeige für ROver…] that.is as for example in the advertisement for Rover
234 Barbara Ahrens
Figure 3.Example (2) — ST pitch movement
Figure 4.Example (2) — TT pitch movement
The structuring marker “ALso” is prominent. It follows the intonation group ending with “… und KONnotationen”, whose tone can be characterized as fall-rise, although it is perceived as level tone (the fall is only approx. 25 Hertz, cf. Crystal 1969: 215ff.). This may be interpreted as a “continuance” (Crystal 1969: 145). Since it is followed by a pause and the new onset “ALso” bearing a rising tone, the boundary between these two intonation groups is very obvious. Nevertheless, the prominent structuring marker “ALso” plays a cohesive role, since it ties the unit of information starting with this marker to the former unit ending with “… und KONnotationen”.
Non-verbal phenomena in simultaneous interpreting 235
Conclusions The two examples, which have been taken from the performance of different booths, show prosodic features whose purpose is to structure the units of information of the TT. Example (1) is an example of the omitting and summarizing strategy that results in a TT having a segmentation of its own. Example (2) describes a case in which the interpreter imitates the explicit structure of the ST. In questionnaires filled in by the interpreters immediately after their performance, they all mentioned the fast speech rate as the main problem of the ST. It forced them to omit some information of the ST or to summarize it in order to keep up with the ST speaker. They also answered that for adapting themselves to the ST speaker’s pace they tried to reduce the EVS as far as possible. The omitting and summarizing strategy applied in the first example results in a TT segmentation and presentation that is not induced by the structure of the ST. Since interpreters tend to “overstress” TT elements in order to make the content and logical structure of the text clear to themselves and the audience (cf. Kalina 1998: 200), they probably stress TT elements that are important indicators of its thematic structure. Gülich (1970) found that structuring markers are used in monolingual communication in order to maintain the communicative flow (cf. Gülich 1970: 16). Their cohesive function is similar to that of Kintsch/van Dijk’s “Topic Change Markers” (cf. van Dijk/Kintsch 1983: 201ff.). Communicative flow and cohesion can be achieved not only verbally, but also prosodically (cf. Gülich 1970: 232ff.). Conference interpreters as communication experts have to be aware of these devices because they can help them to convey the ST message truthfully and completely. Since logical cohesion and sense consistency are the most important criteria for the evaluation of an interpreted text (cf. Bühler 1986, Kurz 1989 and 1993), it seemed probable that cohesive structuring markers would occur in the TTs. This hypothesis was examined in the interpreters’ performance. A salient feature in the corpus presented here — English interpreted simultaneously into German — is the evident prominence of structuring markers, especially when the intonation group preceding the marker has not been finished with a tone which clearly indicates the end of the intonation group. In the corpus, there are many examples in which the interpreters used a rising or level tone, although a low falling pitch movement would have been expected by a native German-speaking audience. Thus, the prominence of structuring markers seems to be a cohesive device for the TT, especially when ST elements are omitted in order to reduce the EVS (as in the first example discussed above). Based on the interpreters’ comments on the summarizing strategy, this might be considered as an indicator that confirms the hypothesis mentioned above. 135 structuring markers (single words and phrases, such as “let’s go on with…”) have been identified in the ST. Since the study presented here has not been
236 Barbara Ahrens
finalised yet, the following figures can only be taken as preliminary results. The quantitative analysis of two of the three booths shows that approximately 55% of the structuring markers used in the TT correspond to those of the ST whereas 45% do not have an equivalent in the ST. This seems to confirm the hypothesis that interpreters structure the TT also according to their needs as far as the segmentation and presentation of the TT are concerned. Prominence of structuring markers is a feature that is found in all three versions of the TT in the corpus, regardless of the interpreters’ personality or experience. The preliminary quantitative analysis of two of the three booths shows that approximately 40% of the structuring markers in the TT are stressed. This figure refers to prominent markers consisting of one or two words. Additionally, structuring phrases have to be taken into account since they usually constitute a separate intonation group with a prominent nucleus. Approximately 25% of all structuring markers in the TTs of two booths analysed so far are such structuring phrases (Gülich’s “Eröffnungssätze” [1970]). These preliminary figures seem to confirm that the stressing of structuring markers can be regarded as a universal prosodic device for delivering the TT. Simultaneous interpreters know that they have to rely on their voice and prosodic features when rendering the TT in order to convey the ST message clearly and unambiguously. Due to the summarizing strategy that the interpreters applied intentionally according to their answers given in the retrospective questionnaires, they were aware of the fact that they were not able to imitate the ST structure when omitting certain elements of the ST. This could be the reason for the use of prosodic structuring markers. Further investigation of this phenomenon would be helpful in order to shed more light on the specific intonation found in SI.
References Ahrens, Barbara. 1999. “‘Dolmetscher klingen alle gleich!’ — Ein Bericht über empirische Forschung im Bereich Intonation beim Simultandolmetschen”. In Modelle der Translation: Grundlagen für Methodik, Bewertung, Computermodellierung [Sabest 1], A. Gil, J. Haller, E. Steiner and H. Gerzymisch-Arbogast (eds), 43–58. Frankfurt: Peter Lang. Ahrens, Barbara. 2002. “The interdependence between verbal and nonverbal elements in simultaneous interpretation”. In Perspectives on Interpreting: Papers from the First Forlì Conference on Interpreting Studies, G. Garzone, P. Mead and M. Viezzi (eds), 37–46. Bologna: Clueb. Alexieva, Bistra. 1988. “Analysis of the simultaneous interpreter’s output”. In Translation, our Future: Proceedings of the XIth World Congress of FIT, P. Nekeman (ed), 484–488. Maastricht: Euroterm. Antos, Gerd. 1982. Grundlagen einer Theorie des Formulierens [Reihe Germanistische Linguistik 39]. Tübingen: Max Niemeyer. Bühler, Hildegund. 1986. “Linguistic (semantic) and extra-linguistic (pragmatic) criteria for the evaluation of conference interpretation and interpreters”. Multilingua 5 (4): 231–235. Collados Aís, Angela. 1998. La evaluación de la calidad en interpretación simultánea: La importancia de la comunicación no verbal [Interlingua 7]. Granada: Editorial Comares.
Non-verbal phenomena in simultaneous interpreting 237
Cruttenden, Alan. 1997 [1986]. Intonation, 2nd ed [Cambridge Textbooks in Linguistics]. Cambridge: Cambrigde University Press. Crystal, David. 1969. Prosodic Systems and Intonation in English [Cambridge Studies in Linguistics 1]. Cambridge: Cambrigde University Press. Daró, Valeria. 1990. “Voice frequency in languages and simultaneous interpretation”. The Interpreters’ Newsletter 3: 88–92. Daró, Valeria. 1994. “Non-linguistic factors influencing simultaneous interpretation”. In Bridging the Gap: Empirical Research in Simultaneous Interpretation [Benjamins Translation Library 28], S. Lambert and B. Moser-Mercer (eds), 249–271. Amsterdam and Philadelphia: John Benjamins. Dijk, Teun A. van and Kintsch, Walter. 1983. Strategies of Discourse Comprehension. New York: Academic Press. Gile, Daniel. 1990. “L’évaluation de la qualité de l’interprétation par les délégués: Une étude de cas”. The Interpreters’ Newsletter 3: 66–71. Goldman-Eisler, Frieda. 1958. “Speech analysis and mental processes”. Language and Speech 1: 59–75. Goldman-Eisler, Frieda. 1961. “The significance of changes in the rate of articulation”. Language and Speech 4: 171–174. Gülich, Elisabeth. 1970. Makrosyntax der Gliederungssignale im gesprochenen Französisch [Structura 2]. München: Wilhelm Fink. Huber, Dieter. 1996. “Prosodic transfer: Nonverbal language in intercultural communication”. In Scotland to Slovenia: European Identities and Transcultural Communication: Proceedings of the Fourth International Scottish Studies Symposium [Scottish Studies International 21], H. W. Drescher and S. Hagemann (eds), 259–277. Frankfurt: Peter Lang. Kalina, Sylvia. 1998. Strategische Prozesse beim Dolmetschen [Language in Performance 18]. Tübingen: Gunter Narr. Kirchhoff, Helene. 1976. “Das Simultandolmetschen: Interdependenz der Variablen im Dolmetschprozeß, Dolmetschmodelle und Dolmetschstrategien”. In Theorie und Praxis des Übersetzens und Dolmetschens [Publikationen des Fachbereichs Angewandte Sprachwissenschaft der Johannes Gutenberg-Universität Mainz in Germersheim, Reihe A, Band 6], H. W. Drescher and S. Scheffzek (eds), 59–71. Frankfurt: Peter Lang. Kurz, Ingrid. 1989. “Conference interpreting: User expectations”. In Coming of Age: Proceedings of the 30th Annual Conference of the American Translators Association, D. L. Hammond (ed), 143–148. Medford, NJ: Learned Information. Kurz, Ingrid. 1993. “Conference interpretation: Expectations of different user groups”. The Interpreters’ Newsletter 5: 13–21. Kurz, Ingrid and Pöchhacker, Franz. 1995. “Quality in TV Interpreting”. Translatio — Nouvelles de la FIT 14 (3–4): 350–358. Lieshout, Pascal van. 2000. “PRAAT workshop: A basic introduction”. http://www.fon.hum.uva.nl/praat. Pöchhacker, Franz. 1994. Simultandolmetschen als komplexes Handeln [Language in Performance 10]. Tübingen: Gunter Narr. Poyatos, Fernando. 1987. “Nonverbal communication in simultaneous and consecutive interpretation: A theoretical model and new perspectives”. TexTconTexT 2(2): 73–108. Shlesinger, Miriam. 1994. “Intonation in the production and perception of simultaneous interpretation”. In Bridging the Gap: Empirical Research in Simultaneous Interpretation [Benjamins Translation Library 28], S. Lambert and B. Moser-Mercer (eds), 225–236. Amsterdam and Philadelphia: John Benjamins. Williams, Sarah. 1995. “Observations on anomalous stress in interpreting”. The Translator 1 (1): 47–64.
Simultaneous interpreting A-B vs. B-A from the interpreters’ standpoint Magdalena Bartlomiejczyk University of Silesia, Institute of English
The paper investigates, through a questionnaire survey, interpreting students’ and professional interpreters’ perception of how well they perform when interpreting simultaneously into their A language vs. into their B language. The research revealed considerable differences between the professional interpreters’ and trainee interpreters’ views on this issue. Professionals generally felt that they performed better when working into their mother tongue, while only half of the students felt the same and 26% thought they worked better into B. In conclusion, hypotheses as to what factors can be responsible for this discrepancy are put forward, and the author expresses her support for A-B interpreting training aimed at preparing interpreting students for the needs of the market.
1.
Introduction
The differences between Simultaneous Interpreting from the mother tongue into a foreign language (A-B, “retour” or “active” interpreting) and in the opposite direction (B-A, “passive” interpreting) have been a controversial subject for a long time. Numerous SI theoreticians, practitioners and trainers have been trying to answer the question as to which language combination is easier for the interpreter and provides better interpreting quality. Gile (1990a: 34 and 1990b: 233) notes that although many differences depending on the direction of SI are intuitively perceived, the subject has not been adequately dealt with by SI research yet. Although some articles dealing with language direction in SI have been published since, a lot still remains to be done in this field. From the very beginnings of interpreting research, most authors have been strongly convinced that the B-A combination is “superior”, often going as far as to consider it the only one acceptable except for the “ideal” A-A combination. Herbert, for example, states that “so far as possible and with few exceptions, he [the interpreter] should speak only in his mother tongue” (1953: 61), which applies to
240 Magdalena Bartlomiejczyk
both the consecutive and simultaneous mode. Seleskovitch allows consecutive interpreting into one’s B, but believes that “simultaneous interpretation can only be done properly into one’s native language” (1978: 100), for which the interpreter has a “total feel […] so that the expressions he uses will sound immediately familiar to his listeners” (1978: 74–75). Interpreting into one’s B, according to her, can only produce linguistically flawed speech even if the interpreter concerned is a very fluent speaker of his or her B language: “once he finds himself in the shoes of an interpreter, […] his native-like fluency disappears. His words no longer flow easily and naturally, and his pronunciation and vocabulary reflect the influence of his native language” (1978: 79). To provide a more recent example, Jones (1998: 134–136) also considers working into one’s B a more challenging task than working into one’s A, although by no means does he claim that the former should not take place at all. He believes that interpreters working into their B language should simply accept their limitations as to the powers of expression they possess in their target language as an inherent quality of retour, and be “modest in the style they adopt”. Views such as those cited above were mainly based on introspection and extensive observation, but sometimes the subject has been dealt with in a more objective and scientific way, e.g. by Kurz, who measured the brain activity (EEG) of an interpreter working (mentally, without uttering the target text) into and out of their mother tongue (1992: 173–186). The experiment has shown that more brain areas are involved when interpreting from A into B than the other way round, which may support the thesis that interpreting into a non-native language is more difficult and tiring for the interpreter. As a consequence of this B-A dominance, some interpreting courses offered by translation and interpreting schools do not provide at all for SI training into one’s B. It is also the general practice of many international organisations and in particular of EU institutions, to let their interpreters work only into their mother tongue, which does not always prove practicable in EU institutions, as some languages happen to be much less popular and difficult to “pick up” as B or C than are others, as is the case with Finnish and with languages of many new EU members, for example Polish, Czech or Hungarian. The opposite view (i.e. that interpreting into one’s B provides better quality) has also been claimed with much confidence, for example by Denissenko, on the basis of his experiences as a teacher of interpreting in Moscow. Denissenko believes that the comprehension part of the interpreting process is the most crucial and, therefore, the best effects can be produced when the interpreter is listening to a speech delivered in his or her mother tongue, as “the losses at input cannot be repaired” (1989: 157). Moreover, his second argument in favour of A-B interpreting is that the interpreter is faced with a smaller number of possible target language renderings of a particular source language phrase, which makes the decision making and control processes easier. Thirdly, Denissenko looks at the problem from the
Simultaneous interpreting A-B vs. B-A from the interpreters’ standpoint 241
perspective of the user’s expectations and states that: A full or near full message gotten across even if in a somewhat stiff, less idiomatic or slightly accented language serves the purpose much better than an elegantlyworded and an impeachably pronounced half-message or less. (1989: 157)
A small-scale empirical study by Tommola and Helevä (1998), in which they investigate the propositional accuracy of interpretations from English into Finnish and from Finnish into English produced by trainee interpreters having Finnish as their A, seems to confirm Denissenko’s views at least to some extent. Whereas no significant difference in propositional accuracy related to the language direction was observed in the case of simple texts, A-B interpretations of complex texts proved to be more accurate. These results, however, could also be explained as languagespecific, i.e. with the hypothesis that English requires fewer cognitive resources as a target language than does Finnish. To sum up, some comparisons between interpreting into A and into B have already been done, involving various research methods and leading to various conclusions. Apart from a small scale questionnaire survey by Al-Salman and AlKhanji (2003), which is restricted to the Arabic-English and English-Arabic context, nobody has systematically asked interpreters themselves what they think and feel about this question, and this is what the author of this article resolved to do. The aim of this study was to find an answer which would still be (inter-)subjective, but founded in the opinions of many different subjects either involved in interpreting professionally or being trained as interpreters. Questionnaires were used to find out what differences between the two directions of SI are actually intuitively perceived by professional interpreters and advanced interpreting students. The working languages of respondents were qualified as A, B and C in accordance with the AIIC criteria, but no differentiation concerning specific languages or language groups was made for the sake of this study.
2.
Method and subjects
One-page questionnaires (a copy of the questionnaire for interpreting students is provided as an Appendix) were distributed by the author to interpreting students at two institutions training interpreters: the Institute of English, University of Silesia, Poland, and the Institute for Translation and Interpreting, University of Vienna, Austria, between December 2000 and May 2001. Neither of the schools mentioned above systematically favours work into one’s A or into one’s B, and the interpreting students of both are required to be able to work in both directions when they finish the course. Students from Vienna attend separate SI classes devoted to their A-B, B-A and C-A or C-B combinations for as many terms as they
242 Magdalena Bartlomiejczyk
feel it necessary before taking the final diploma examination. Students of the University of Silesia, on the other hand, are required to attend one SI class weekly for 6 terms, and these classes cover both English-Polish and Polish-English interpreting, approximately in the proportion 50:50. Questionnaires for professional interpreters were distributed between December 2000 and August 2001 basically in two ways: either given personally by the author to teachers of interpreting at the two institutions mentioned above as well as to other professional interpreters with whom she is acquainted, or sent via the Internet to AIIC members, asking them for their responses. These methods enabled the author to gather answers from 93 respondents altogether (although at least four times more than that were asked to complete the questionnaire). The questionnaire was completed by 53 students who participated in SI classes and were trained in both B-A and B-A interpreting. All the students had practised SI for at least one term, but most of them were more advanced (3–4 terms or more). Their ages ranged from 21 to 28, most of them being between 22 and 24. The students from the University of Silesia (32 subjects) all had Polish as their mother tongue and English as their B language, some of them additionally having some other B or C language(s). The students from the Vienna Institute (21 subjects) displayed a wide variety of A, B and C languages; for all of them, German was either A or B. Other A and B languages were: English, French, Hungarian, Polish, Russian and Spanish. A further group of 40 respondents were professional conference interpreters with experience ranging from one to 42 years but in most cases reaching at least 15 years (30 interpreters), the majority of them being AIIC members (35 interpreters). Their language combinations comprised all the languages of the students in Vienna mentioned above and some more, including Arabic, Danish, Swedish and Portuguese. The most widely represented A and B languages were English, German and French.
3.
Results
3.1 Interpreting students Estimating the quality of their B-A and A-B interpretation, 26% of the student respondents thought themselves to be equally good (or equally bad) in both directions, and 74% made a distinction in their estimation depending on the direction: 48% thought they interpreted better from B into A, 26% from A into B. Therefore, only about half this group of respondents shared the widespread view referred to in the introduction of themselves being better at interpreting into their mother tongue. The total average grade for B-A interpreting was 4.9, whereas for
Simultaneous interpreting A-B vs. B-A from the interpreters’ standpoint 243
4
3
2
1
B-A
in te rp re te r’s vo ca bu lar y in te rp re te r’s ac ce nt
un de rs ta nd in g
in te rp re te r’s g
ra m m in ar te rp re te r’s in so to ur na ce tio te n xt de liv er ys pe ed
0
A-B
Figure 1.Reported frequency of specific problems encountered by students when interpreting into and from their mother tongues (based on answers to Questions 12 and 13 in the Questionnaire).
retour interpreting it was 6.8, with a 1.9 difference between them (for the grading scale used see the Appendix, Question 14). Among those who noticed a difference between the quality of their interpretation in both directions, most did not consider this difference to be radical: 40% of the total number of respondents gave themselves grades with just a one point difference between the two directions. Added to those who did not differentiate at all, this makes 66%. Incidentally, 66% is also the percentage of those students who think they will work in both directions in their future career as interpreters. 14% of the respondents awarded themselves grades varying by at least 3 points, but more than that, 26.4%, think they will work in only one direction (22.7% only B-A, which in practice means reducing their present B language to the status of C in their professional future, and 3.8% — 2 students — only A-B).
244 Magdalena Bartlomiejczyk
The answers to questions concerning the particular problems students experience when interpreting into their native and into their non-native language (Questions 12 and 13 in the Appendix) are graphically represented as Figure 1. On the vertical axis, the answers given have been ascribed the following numeric values: “never” is treated as 0, “rarely” – 1, “from time to time” – 2, “often” – 3, “usually” – 4, and “always” – 5. As might be expected, the greatest difference in favour of interpreting into one’s A was reported for interpreter’s accent (1.3 points), and the greatest difference in favour of retour interpreting for understanding the source text (almost 1 point). Among the specific problems mentioned in the questionnaire, interpretation into one’s mother tongue was considered to result in fewer problems in 3 aspects, and retour interpretation was judged to be easier in 3 aspects, too. Students generally refrained from mentioning any additional problems they had noticed or from providing any other remarks. 3.2 Professional interpreters Several considerable differences may be noticed between the responses given by students and by professionals. Firstly, professionals awarded themselves much better grades for their performance and reported problems considerably less frequently, which is not surprising and certainly corresponds to their possessing much higher skills than those of the interpreting students. Hardly anybody from this group used a grade worse than 5 for interpreting in either direction, and the average self-assessment was 2.0 for B-A combination and 3.6 for A-B combination. The average difference in assessment, however, being 1.6 points, did not radically differ from its counterpart in the case of the students. The fact that professional interpreters were convinced to a much greater degree that the quality of their B-A interpretation was better than that of their A-B interpretation constituted another crucial difference. As many as 29 of the respondents who agreed to give themselves grades (there were a few respondents in this group who answered all the questions except the last two), i.e. 82.9%, were of the opinion that they performed better when interpreting into their mother tongue. For 12 interpreters (34.3%) this difference was reflected in grades differing by only 1 point, and 6 interpreters (17.1%) differed by 3 points or more. Six interpreters (17.1%) believed themselves to be equally good in both directions. Nobody awarded herself or himself a better grade for retour interpreting than for interpreting into her or his A, although there was one interpreter, with English as B, who forcefully claimed, in an e-mail she sent in addition to the completed questionnaire, that interpreting into one’s B was both easier and conducive to higher quality (she did not, however, make this assertion visible in her self-assessment, giving herself 1 for both directions). The frequency of the problems encountered was generally lower than in the
Simultaneous interpreting A-B vs. B-A from the interpreters’ standpoint 245
4
3
2
1
un de rs ta nd in in te g rp re te r’s vo ca bu lar y in te rp re te r’s ac ce nt
in te rp re te r’s g
ra m in m te ar rp re te r’s in to so na ur tio ce n te xt de liv er ys pe ed
0
B-A
A-B
Figure 2.Reported frequency of specific problems encountered by professionals when interpreting into and from their mother tongue (based on answers to Questions 12 and 13 in the Questionnaire).
case of the students (see Figure 2). Professional interpreters reported that the problems they encountered most often resulted from the source text delivery speed (not a linguistic problem as such), and this was the most difficult aspect in the case of both directions (although these problems were considerably less frequent when interpreting into the mother tongue). The only aspect in which A-B interpreting was considered easier than B-A by this group was understanding the source text, although this difference in favour of retour was small (0.2 point) when compared to the same parameter in the case of the students. In all other aspects, the professionals considered B-A interpreting less problematic and the differences were more pronounced (0.9 for interpreter’s accent, 0.8 for interpreter’s grammar and source text delivery speed, 0.4 for interpreter’s intonation and 0.3 for interpreter’s vocabulary). Apart from a few cases where interpreters gave up interpreting language
246 Magdalena Bartlomiejczyk
combinations with C (C-A or C-B), 4 interpreters (10%) had given up A-B interpreting in the course of their professional careers. The reasons given for this were various, including not having enough opportunities to practise this particular combination and not being satisfied with one’s performance. Many of the respondents provided the author with some additional remarks and opinions relating to the differences between A-B and B-A interpreting. Except for the interpreter mentioned above, all of them were in favour of interpreting into one’s mother tongue, mentioning problems such as limited idiomatic use, getting tired more quickly and audiences being critical of the interpreter speaking with a non-native accent in connection with retour. Some expressed the opinion that the use of A-B interpreting depended heavily on the particular languages playing the role of B, with English being most readily accepted as a non-native target language by audiences and interpreters alike.
4. Conclusions On the whole, interpreting students and professional interpreters differed considerably in their opinions concerning retour interpreting in comparison with interpreting into one’s mother tongue. Students were in favour of A-B interpreting much more often and there were also more respondents in this group who judged the quality of their performance independent of the language combination used. This difference might be explained as a result of students’ having too high an opinion of their own mastery of their B language. They feel very good in the comprehension part of their work, but at the same time perhaps they are just not aware of the many errors which they do notice when interpreting into their mother tongue, and, as their B language improves, they become more critical of their retour interpreting, which can even lead some interpreters to give up the A-B combination they used to work in at the beginning of their career. The other explanation might be that there is a stage in interpreting training where students do perform better interpreting into a foreign language. This would account for Denissenko’s views, as he based them mainly on his observations of the students at the interpreting school at which he worked, as well as for Tommola and Helevä’s findings, since they tested the interpreting students at the end of SI course they took. Both these hypotheses, however, are suggested tentatively, as many more studies concerned with objective scientific research involving an external evaluation of students’ performance into and out of various A and B languages are definitely necessary to determine which one is valid and to what extent. When comparing B-A and A-B interpreting, we may also consider as one potential factor the specific language combination (Gile 1997: 209). Actually, it is often difficult to determine whether the influence of interpreting direction (into or
Simultaneous interpreting A-B vs. B-A from the interpreters’ standpoint 247
out of one’s native tongue) or the language specificity plays the greater role in interpreters’ performance. Some languages (e.g. English) seem to pose fewer difficulties and involve less production effort as target languages than others, and some require extra listening effort (e.g. Chinese and Japanese) or an extra strain on the memory effort (e.g. German) as source languages. Such intrinsic qualities may sway interpreters’ opinions and preferences regarding the language combinations they work in quite independently of the given languages being native or non-native for them. It is also possible that the opinion concerning the superiority of B-A language combinations widely held by the professional interpreters taking part in the survey is due, at least to some extent, to environmentally shared wisdom, as AIIC systematically favours work into one’s A and 87.5% of the respondents in the professionals’ group were AIIC members. Whatever our opinion of retour might be, A-B combinations are an absolute necessity for some languages which are hardly ever represented as C or B; this certainly includes Polish and the languages of many other new EU members. This is simply the demand of the market, and consequently we cannot and should not refrain from teaching A-B interpreting to students with A languages belonging to this group. As Gile states: It is a fact of life that many interpreters are required to work into their B language because of market conditions, chief among which is the lack of availability of native speakers for some rare languages. In spite of this, in many interpreting schools, no training into one’s B language is given. The result may be that when entering the marketplace, beginning interpreters are severely handicapped […]. (1997: 209–210)
As the character of interpreting into and out of one’s mother tongue differs, however, as confirmed by the subjective opinions of the majority of our respondents, and in particular by experienced and skilled professional colleagues, perhaps it would be a good idea to develop different teaching techniques for A-B and B-A interpreting, focusing on the development of particular skills and solving specific problems relating to interpreting in both directions. Author’s Address: Magdalena Bartlomiejczyk University of Silesia Institute of English ul. Zytnia 10 41–205 Sosnowiec, Poland
[email protected] 248 Magdalena Bartlomiejczyk
Appendix: Questionnaire distributed to interpreting students (All questionnaires were in English. The questionnaire distributed to professional interpreters differs very little from this one. Instead of the question about age, it contains one asking how long the respondent has worked as a simultaneous interpreter, and the one about future plans to work in various language combinations has been replaced by the question if the respondent has given up any language combinations he or she used to work in and, if yes, for what reasons. Some of the questions differ in phrasing, but not in their content, referring to “working as an interpreter” rather than “practising SI”.) 1.
Age:
2. Sex:
3.
A language(s):
4. B language(s):
5.
C language(s):
6.
Have you got any difficulties in determining which is your strongest working language? If yes, briefly explain why.
7.
At what age did you start to learn your B language(s)?
8.
At what age were you able to communicate in your B language(s)?
9.
In what language combinations do you practice simultaneous interpreting?
10. How long have you practised already for each of the combinations mentioned in 9? 11. Which combinations do you think you will be working in as a professional simultaneous interpreter? 12. What problems are you confronted with when interpreting from B into A? (please specify for each point: never, rarely, from time to time, often, usually, always) * your grammar *your vocabulary * your intonation * your accent * speed at which the source text is delivered * understanding the source text * others (please specify if you can think of any) 13. What problems are you confronted with when interpreting from A into B? (please specify for each point: never, rarely, from time to time, often, usually, always) * your grammar *your vocabulary * your intonation * your accent * speed at which the source text is delivered * understanding the source text * others (please specify if you can think of any) 14. On the whole, how would you judge the quality of your interpretation B-A? (using the scale from 1 to 10, 1 = excellent, 10 = unacceptable, please specify for each such language combination you work in) 15. On the whole, how would you judge the quality of your interpretation A-B? (using the same scale as in 14, please specify for each such language combination you work in)
Simultaneous interpreting A-B vs. B-A from the interpreters’ standpoint 249
References Al-Salman, Saleh and Al-Khanji, Rajai. 2002. “The native language factor in simultaneous interpretation in an Arabic/English context”. In: Meta 47(4), 607–625. Denissenko, Jurij. 1989. “Communicative and interpretative linguistics”. In The Theoretical and Practical Aspects of Teaching Conference Interpretation, Gran & Dodds (eds), 155–158. Udine: Campanotto Editore. Gile, Daniel. 1990a. “Scientific research vs. personal theories in the investigation of interpretation”. In Aspects of Applied and Experimental Research on Conference Interpretation, Gran & Taylor (eds), 28–41. Udine: Campanotto Editore. Gile, Daniel. 1990b. “Research proposals for interpreters”. In Aspects of Applied and Experimental Research on Conference Interpretation, Gran & Taylor (eds), 226–236. Udine: Campanotto Editore. Gile, Daniel. 1997. “Conference interpreting as a cognitive management problem”. In Cognitive Processes in Translation and Interpreting, Danks, Shreve, Fountain & McBeath (eds), 196–214. Thousand Oaks, London, New Delhi: Sage Publications. Herbert, Jean. 1952. The Interpreter’s Handbook: How to Become a Conference Interpreter. Geneva: Librairie de l’Université. Jones, Roderick. 1998. Conference Interpreting Explained. Manchester: St. Jerome. Kurz, Ingrid. 1992. Simultandolmetschen als Gegenstand der interdisziplinären Forschung. Vienna: WUV-Universitätsverlag. Seleskovitch, Danica. 1978. Interpreting for International Conferences: Problems of Language and Communication. Washington, D. C.: Pen and Booth. Tommola, Jorma and Helevä, Marketta. 1998. “Language direction and source text complexity. Effects on trainee performance in simultaneous interpreting”. In Unity in Diversity? Current Trends in Translation Studies, Bowker, Cronin, Kenny & Pearson (eds), 177–186. Manchester: St. Jerome.
Interpreters’ notes On the choice of form and language Helle V. Dam Aarhus School of Business (Denmark)
This paper reports on a small-scale pilot study on consecutive interpreters’ notes based on data drawn from a simulated conference with interpreting between Spanish and Danish. The issues selected for presentation and discussion here are (1) interpreters’ choice of form for their notes (symbols vs. language) and (2) their choice of language (source, target or other languages). Because of the pilot status of the study, it provides no rock-solid results, but actually raises more questions than it answers. The main contribution of the paper, therefore, does not lie in presenting new knowledge, but rather in raising questions and suggesting hypotheses that may serve as the point of departure for future studies on note-taking.
1.
Introduction
The issue of conference interpreters’ notes has generated a large volume of literature, from Rozan (1956) over van Hoof (1962), Seleskovitch (1975), Kirchhoff (1979), Ilg (e.g. 1980, 1982 and 1988), Thiéry (1981), Matyssek (1989) and Alexieva (1993), to Ahrens (2001) and Andres (2001) — just to mention some (for a detailed review of the literature and a comprehensive bibliography on note-taking, see Ilg & Lambert (1996)). So far, most of the writings have mainly focussed on didactic issues and aimed at giving recommendations about what interpreters’ notes should look like and how note-taking should be taught. As most of the recommendations are offered on the basis of personal experience and/or opinions only, with hardly any empirical studies to back them (see, however, Kirchhoff (1979) and Andres (2001)), it is not surprising that they often point in different directions. For example, opinions on the question of note form, i.e. whether symbols or natural language (words and abbreviations) are a better choice, differ significantly between authors. One line of thought is represented by e.g. Rozan (1956) — the author of the perhaps most widely used note-taking system among interpreters ever — who
252 Helle V. Dam
recommends taking down most notes in a natural language, limiting the use of symbols to some 10 or 20 (Rozan 1956: 27–28). By contrast, e.g. Matyssek (1989), whose note-taking system remains very influential in Germany, advocates an elaborate system of language-independent symbols — almost a new code language, or an “interpreter’s shorthand”, as it has also been referred to (Ilg & Lambert 1996: 72). Another example of differing opinions concerns the issue of language, the basic point of disagreement being whether the notes should be taken down in the source language or in the target language. For example Seleskovitch & Lederer (1989) firmly recommend using the target language, whereas others feel that the source language is a better choice (e.g. Ilg 1988; Alexieva 1993). I do not pretend to settle any of these issues in this paper, but I shall attempt to approach them in a different way. For one thing, I shall be concerned not so much with what interpreters’ notes ought to look like, but rather with what they actually look like. Furthermore, I shall be less preoccupied with answering questions than with raising them. Essentially, my paper reports on some of the preliminary results of an ongoing research project on conference interpreters’ notes. The long-term objective of the project is to study interpreters’ notes from different perspectives, including the notes per se, their relation to the source text, and their relation to the target text. In addition, the plan is to draw on different kinds of data, including notes produced in the context of interpreting between different language pairs, in different settings, using different materials, etc. At this point, however, only a smallscale pilot study has been conducted with the aim of generating the first questions and hypotheses, which are to be examined and tested at a later point. The pilot study has so far mainly focussed on describing interpreters’ notes per se, and the results I have chosen to present here relate to the two issues in note-taking outlined above, namely (1) interpreters’ choice of form for their notes (symbols vs. language) and (2) their choice of language (source, target or other languages).
2.
Data
The data used for the pilot study comprise five sets of interpreters’ notes produced in the context of an experimental study of consecutive conference interpreting. The original aim of the study was to address the topic of text condensation in consecutive interpreting (Dam 1995), which means that note-taking was not originally the object of study. The source text was a 7.5-minutes and 1081-words long speech in Spanish on the topic of drug policy. The speech was interpreted consecutively into Danish by five professional conference interpreters with Danish as their native language (A-language) and Spanish as a foreign language (B- or C-language). Apart from the fact that they were all professional interpreters working for the EU with the language combination Spanish-Danish, the subjects were not directly comparable.
Interpreters’ notes 253
Most importantly, their professional experience ranged from two to fourteen years, and not all of them had received their training at the same school.1 The five sets of notes obtained from the experiment comprised a total of 1442 note units (words, abbreviations or symbols), with each interpreter having rendered between 196 and 346 units (cf. Table 1 below).
3.
The choice of note form
The first aspect of notes studied was that of note form, i.e. the notes were analysed for the type of representation chosen for them by the subjects. Not surprisingly, three categories of note forms were found in the data, namely (1) words, (2) abbreviations, and (3) symbols. In this study, the category of words includes all notes represented as full words, i.e. words that had not been abbreviated, although words both with and without morphemes of inflection were registered as full words. In particular, nouns both with and without plural markers were indiscriminately analysed as full words, which means that no distinction was made between representations like for example ‘problems’ and ‘problem’ in the analyses. The category of abbreviations includes, of course, ‘real’ abbreviations, i.e. units in which only part of a word is represented (for example: ‘prob.’ or ‘prblm’ for ‘problem(s)’). But the category also counts acronyms (like ‘EU’ for ‘European Union’) and other short forms that cannot really be characterised either as real abbreviations or as acronyms, but rather as something in between. For example, in the note ‘L+G’ for ‘ladies and gentlemen’, the ‘L’ and the ‘G’ were categorised as abbreviations (but the plus (+) was analysed as a symbol, cf. below). The category of symbols is a broad one in this context. Basically, it comprises everything that is not language, including signs like pluses and colons, lines, arrows, drawings, etc. Table 1 shows the distribution of the subjects’ notes over the three categories (please note that the order of subject 3 and subject 4 has been inverted for illustrative reasons, cf. the accompanying text). Table 1.The distribution of the subjects’ notes over the tree categories of note forms Note form
Subject 1
Subject 2
Subject 4
Subject 3
Subject 5
Total
Words 118 (34%) 114 (33%) 54 (18%) 124 (49%) 89 (45%) 499 (35%) Abbreviations 67 (19%) 73 (21%) 109 (36%) 64 (25%) 44 (22%) 357 (25%) Symbols 160 (46%) 159 (46%) 139 (46%) 65 (26%) 63 (32%) 586 (41%) Total Units
345
346
302
253
196
1442
254 Helle V. Dam
The rightmost column of Table 1 shows the average distribution of the five subjects’ notes over the three categories of note forms. As can be seen, the subjects as a group show a preference for using symbols (41% of all note units), words are their second choice (35%), whereas abbreviations are used to a lesser extent (25%). However, this overall pattern is not consistently reflected in the individual subjects’ notes. Rather, the results obtained for the individual subjects exhibit what at first sight seems to be a high degree of variability. However, the subjects do in fact not perform in entirely different ways. Rather, they tend to divide into two major groups, each of them showing patterns that are similar within the group but different from those of the other group. On the one hand, the results obtained for subject 1 and subject 2 are remarkably similar in all respects (see Table 1). On the other hand, the results obtained for subject 3 and subject 5 are also quite similar, but at the same time they are rather different from the performances of subjects 1 and 2, at least as regards the use of words vs. symbols. Thus, subjects 1 and 2 show a preference for the use of symbols over the use of words (46% vs. 34% and 46% vs. 33%, respectively). By contrast, subjects 3 and 5 show a clear preference for the use of words over the use of symbols (49% vs. 26% and 45% vs. 32%, respectively). The performance of subject 4 differs somewhat in that (s)he uses more abbreviations than any of the other subjects (the proportion of abbreviations remains fairly stable around 20% in the other four subjects), but his/her results conform nicely with those of subjects 1 and 2 as regards their preference for symbols over words. Thus, subject 4 has exactly the same proportion of symbols in his/her notes (46%) as do subjects 1 and 2. To sum up, rather than exhibiting totally different patterns, the subjects of this study show a bipolar pattern of preferences in their choice of note forms: while most of them have approximately the same proportion of abbreviations, some of them show a preference for symbols (subjects 1, 2 and 4), whereas others prefer using words (subjects 3 and 5).2 It is clearly not possible to draw any conclusions on the basis of these observations alone, but it would be interesting to see whether a similar pattern of symbolword polarity can be found in future studies as well. If it can be confirmed that interpreters tend to be either relatively symbol-oriented or relatively word-oriented when taking notes, the next logical step would be to examine the reasons for and, in particular, the effects of their choice of strategy. As regards the reasons for the choice of a symbol- or a word-oriented notation, no evidence to support one or the other hypothesis can be found on the basis of the present study. All we can say is that, since all the subjects performed the same task, we have to assume that the different choices of strategy observed here are not taskrelated, but reflect differences in the interpreters’ personal or professional backgrounds, i.e. in their personality, experience, training, etc. Possible correlations between these factors and the choice of note-taking strategy need to be established
Interpreters’ notes 255
in future empirical studies. As regards the effects of the choice of either notation strategy, the present study does in fact allow us to make some observations that may serve as a basis for hypotheses for future studies. As we can see in Table 1, there is a correlation between the number of symbols used and the total number of units represented in the notes, as both the number of symbols and the total number of note units generally descend from the left to the right (the only small deviation is found in the results for subject 1 and subject 2, who exhibit insignificant differences between them). In other words, the pattern observed here can be expressed as follows: the fewer symbols, the fewer notes, or — vice versa — the more symbols, the more notes. The relation between the amount of words and total note units is less clear, but at a general level we can see that the group of subjects who have the highest proportions of words in their notes (subjects 3 and 5) has the lowest number of total note units (see Table 1). On the other hand, the group of subjects who have the lowest proportions of words (subjects 1, 2 and 4) has the highest number of total note units. At this general level, the relation observed here is therefore the opposite of the one found between symbols and total note units, and in simplified terms it can be expressed as follows: the more words, the fewer notes, or — vice versa — the fewer words, the more notes. The above observations make good sense. Thus, it seems intuitively plausible that symbols are more economical as notes than full words, provided that the symbols are both simple and fully mastered by the interpreter, and that the full words are relatively long.3 Consequently, on the basis of the results of the present study, we may formulate a hypothesis saying that the more notes the consecutive interpreter represents as symbols, the more notes (s)he can take down, and, conversely, the more notes (s)he writes in the form of full words, the fewer notes (s)he is able to produce. However, in all probability, this hypothesis will turn out to be an over-generalisation, both because no interpreter can be expected to have a nice and simple symbol available for every word or idea that (s)he may come across, and because some full words may in fact be very efficient notes. We can therefore only expect the hypothesis to hold up to a certain limit. Furthermore, it needs to be stressed that, if the hypothesis is confirmed, this does not necessarily mean that symbols are a better choice than full words, for the simple reason that a large amount of notes may not be desirable per se. The purpose of notes is invariably to serve as memory reinforcers, so as to enable the interpreter to produce a target text that reflects the meaning of the source text rather precisely. For that purpose, a small number of notes may very well be as efficient as — or perhaps even more efficient than — a large number of notes. At this point we do not know, and this question therefore needs to be resolved in future studies, essentially in studies on the relation between the number of notes and the degree of source-text fidelity of the ensuing target text. Only then
256 Helle V. Dam
will it be possible to start making recommendations about the optimum symbolword ratio on an informed basis.
4. The choice of language The second issue that will be addressed in this paper is the question of choice of language for the notes. As mentioned in Section 1, this has traditionally been a controversial issue, in particular as regards the choice between taking notes in the source language and writing in the target language. Thus, the advisability of using the target language has been stressed time and again (e.g. Herbert 1952; Rozan 1956; Seleskovitch 1975; Seleskovitch & Lederer 1989; Mikkelson 1983; AIIC 1994) because it is felt that this option forces the interpreter away from the surface form of the source language speech and therefore makes for better processing of the text, and that it facilitates production of the target language speech. But this position has been questioned by others (e.g. Kirchhoff 1979; Ilg 1988; Alexieva 1993; Gile 1995) on the grounds that writing notes in the target language requires language conversion during note-taking and therefore adds to the number of functions that the interpreter has to perform during the listening phase — a phase characterised i.a. by being paced by the speaker, unlike the production phase. In addition, some interpreting teachers report that their students perform markedly worse when they take notes in the target language (Alexieva 1993), while others contend that students perform poorly when writing in the source language (Seleskovitch 1975). Considering this disagreement, I found it could be interesting to look at whether there were general patterns in interpreters’ actual notes, and whether relevant hypotheses could be generated from the exercise. The analysis of language choice carried out in the present study was based only on the notes that the subjects had represented as words or abbreviations, whereas the notes represented as symbols were not taken into account because of their inherent non-language nature. Each of the words and abbreviations taken down by the subjects was categorised as representing either (1) the target language (TL, in this case Danish), (2) the source language (SL, in this case Spanish), or (3) a third language (3rd L, in this case mostly English). However, the language of some of the notes could not be identified because their form would be identical in Spanish (the source language), Danish (the target language) and/or English or other third languages. An example of a note of this kind could be ‘legal’, which is an existing word in Danish, Spanish and English — and probably in other languages as well. Notes for which the linguistic origin could not be determined were categorised separately, and in Table 2, where the results of the language analysis are summarised, they are represented in the column named ? (question mark). As we can see in Table 2, the results concerning the choice of language are quite
Interpreters’ notes 257
Table 2.The distribution of the subjects’ notes (words and abbreviations) over languages Language
Subject 1
Subject 2
Subject 3
Subject 4
Subject 5
Total
TL SL 3rd L ?
107 (58%) 46 (25%) 9 (5%) 23 (12%)
117 (63%) 22 (12%) 29 (16%) 19 (10%)
163 (87%) 3 (2%) – 22 (12%)
125 (77%) 3 (2%) 2 (1%) 33 (20%)
102 (77%) 11 (8%) – 20 (15%)
614 (72%) 85 (10%) 40 (5%) 117 (14%)
Total
185
187
188
163
133
856
consistent, insofar as the pattern of the overall results generally coincides with the patterns found in the individual subjects’ performances. Clearly, there is a strong preference for writing notes in the target language in all respects — both overall (72% of all notes are in the TL) and individually (the proportion of notes in the TL ranges from 58% to 87% in the performances of the individual subjects). Much less used is the source language (10% overall, and between 2% and 25% in the individual performances), and the use of a third language is even less frequent (5% overall, and between 0 and 16% individually). Only subject 2 deviates slightly from the overall pattern in that (s)he writes more notes in a third language than in the source language (16% in a 3rd language and 12% in the SL). If we assume that the subjects adopted a target-language strategy here because they find that this is the better option, these results apparently support the authors who are in favour of using the target language for note-taking. But it needs to be stressed that this small piece of evidence can in no way be generalised to other interpreters, other language pairs, other situations, etc. It cannot even be extrapolated to other tasks performed by the same interpreters, because the high degree of agreement between subjects found here indicates that the observed targetlanguage orientation may be exclusively task-related, which means that in other tasks the very same interpreters may well opt for a source-language strategy. In fact, the above results are not even sufficient to formulate a full-fledged hypothesis. For this to be possible, we need to ask either why the interpreters chose a TL-oriented strategy for their notes, or what effects it had. As to possible effects, so far no attempts have been made to chart them out in the context of this study. As to possible reasons, an additional analysis was conducted, in which it was explored in what cases the subjects had primarily used the target language for their notes, and when they had resorted to the source language. In this analysis of language shifts, the notes were not studied in isolation, but were matched with the source text. The analysis is still very preliminary, but it does give some indications of possible reasons for interpreters’ choice of source or target language for their notes.
258 Helle V. Dam
4.1 Analysis of language shifts In the analysis of language shifts, the source text and the notes were first divided into paragraphs, so that a series of shorter source-text units and their corresponding notes were established. Each paragraph of each subject’s notes was then analysed for deviations from the subject’s general distribution of target-language notes and source-language notes. For example, subject 1 represents 58% of all his/her notes in the TL and 25% in the SL (cf. Table 2), but for the first paragraph, only 23% of his/her notes were written in the TL, whereas 65% were taken in the SL. Subject 1’s notes for the first paragraph were, then, classified as representing a deviation, or shift, towards a more SL-oriented strategy. The next step was to identify the socalled consensus-based deviations, i.e. the paragraphs in which most of the subjects (three or more) had made a shift of the same type: either towards a more TLoriented strategy or towards a more SL-oriented strategy. Finally, the source-text paragraphs for which consensus-based deviations from the subjects’ usual SL-TL distribution of notes had been identified were given a rough description. From this analysis the following pattern of results emerged: 1. A general shift towards the source language was registered in the notes for the first two paragraphs of the source text, a tendency that was particularly strong in the notes for the very first paragraph. The same type of shift was found in the notes for two source-text passages that were dominated by numbers — the only two passages of the source text that contained numbers. 2. A general shift towards the target language was found in the notes for the paragraphs towards the end of the speech. On the whole, target-language notes were used increasingly over the speech, and in the last paragraph only 3 of a total of 69 note units were rendered in the source language. A shift towards target-language notes was also found in a mid-text paragraph presenting some very simple cause-effect relations, with neatly structured information given in particularly short sentences with explicit cohesive ties between them. I suggest that the paragraphs for which there was a shift towards the source language may be characterised as relatively difficult to interpret. Conversely, I would characterise the paragraphs for which there was a shift towards the target language as relatively easy to deal with for the consecutive interpreter. For one thing, at the beginning of a speech, the interpreter normally has to be particularly attentive, insofar as (s)he does not know the speaker, his/her speaking style, the exact topic or perspective chosen, how the argument will develop, etc. Then, as the speech develops, the interpreter becomes more familiar with the speaker’s style, gets a clearer idea of the development of the argument, etc., which should logically mean that — other things being equal — the task becomes easier as it progresses. Secondly, concerning the passages with many numbers, we may note that numbers are a well-known source of difficulty to interpreters (e.g. Gile 1995: 176; Dam 2000: 57).
Interpreters’ notes 259
Finally, the paragraph structured around cause-effect relations described above is not only easy to analyse and understand, it is also typical of the type of material for which note-taking is considered to be very straightforward, because it can readily be represented by means of well-rehearsed symbols such as arrows and equals signs, in combination with a few keywords and standard connectives. This was also the pattern typically found in the notes produced by the subjects in this study. In sum, it appears that when the source-text paragraphs were relatively difficult, the subjects generally shifted towards a more source-language-oriented strategy, and — vice versa — when the task was relatively easy, a shift towards a more targetlanguage-oriented strategy generally occurred. I suggest that this pattern can be explained within the framework of the Effort Models developed by Daniel Gile (e.g. Gile 1995). In the Effort Model of consecutive interpreting, the phase relevant to this study, namely the listening and notetaking phase, is described as consisting of four components: (1) listening and analysis, (2) note-taking, (3) short-term memory operations, and (4) coordination. In the effort models as a whole, the individual components are referred to as ‘efforts’ to stress the fact that each of them requires a certain amount of processing capacity, which is available only in limited supply. The components are inter-dependent in the sense that variations in the requirements for one effort may have implications for any of the other efforts. If, for example, a large amount of processing capacity is devoted to the note-taking component in the listening phase of consecutive, less processing capacity will be available for the other three components. In the context of this study, the components at stake seem to be (1) listening and analysis, and (2) note-taking, as it can be hypothesised that the different notetaking patterns observed here are a result of different requirements for listening and analysis. This hypothesis is based on two assumptions. On the one hand, it is assumed that difficult source-text passages imply higher processing capacity requirements than easy passages. In particular, requirements for the listening and analysis component should logically increase with an increase in source-text difficulty. On the other hand, it is assumed — as does Gile — that “writing notes in the target language […] requires more processing capacity than writing them in the source language because of the capacity required for conversion” (Gile 1995: 182). Clearly, if the interpreter writes notes in the source language, (s)he can choose to represent the input verbatim, whereas this is not an option when writing in the target language. When applying the above assumptions to this study, the following description of events seems plausible: when capacity requirements for listening and analysing were high — as they probably were for the source-text passages described above as relatively difficult — the subjects had less processing capacity left for note-taking. As a consequence, they turned towards the note-taking strategy that requires less capacity, i.e. the source-language strategy. And vice versa: when capacity requirements
260 Helle V. Dam
for listening and analysis were low — as they probably were for the passages described above as relatively easy — the subjects had more processing capacity available for the note-taking component. They were therefore better able to apply the more demanding note-taking strategy, i.e. target-language orientation. Based on the above assumptions and observations, we could therefore formulate a general hypothesis saying that — other things being equal — the more difficult the task, the more notes will be taken down in the source language, and — vice versa — the easier the task, the more notes will be rendered in the target language. This hypothesis may be summed up as in the following quotation: “TL where possible, SL where necessary” (AIIC 1994: 21). But whether it can be verified remains to be investigated in studies designed for that particular purpose.
5.
Concluding remarks
In this paper, several questions have been raised on the topic of interpreters’ notes, whereas practically no answers have been given. The main contribution of the paper therefore does not lie in presenting new knowledge, but rather in suggesting some concrete hypotheses that may serve as the point of departure for future studies on note-taking. Connecting to the keywords of the present volume — Claims, Changes and Challenges — I could therefore say that no Claims have been made in this paper, which in itself represents a Change with respect to most earlier publications on note-taking. Rather, questions have been asked and hypotheses have been suggested, the testing of which constitutes a Challenge for future research.
Notes 1. However, four of the five subjects had in fact been trained at the same school. The one who had been trained elsewhere is subject 4 (see Tables 1 and 2). 2. Note, however, that language-based notes (words and abbreviations) dominate over symbols in this study: both in the overall results, and in the results obtained for each of the subjects. For reasons of space, this observation will not be pursued in this paper. 3. It seems equally plausible that abbreviations would be more economical than full words, a point often made in the literature (e.g. Rozan 1956; van Hoof 1962; Mikkelson 1993). But as the relative number of abbreviations is not sufficiently differentiated between subjects here to allow studying their effects, the question of abbreviations is not pursued further in this context. Note, however, that with the exception of subject 4, the interpreters in this study show a clear preference for full words over abbreviations, whereas it is often stated that writing out words in full is an exception in note-taking for consecutive (e.g. Andres 2001: 246).
Interpreters’ notes 261
References Ahrens, Barbara. 2001. “Einige Überlegungen zur Didaktik der Notizentechnik”. In Dolmetschen. Beiträge aus Forschung, Lehre und Praxis, A. F. Kelletat (ed.), 227–241. Frankfurt: Peter Lang. AIIC. 1994. Interpreter Training Workshop, Poznan, 8–10 April, 1994. Genève: AIIC. Alexieva, Bistra. 1993. “On teaching note-taking in consecutive interpreting”. In Teaching Translation and Interpreting 2, C. Dollerup and A. Lindegaard (eds), 199–206. Amsterdam and Philadelphia: John Benjamins. Andres, Dörte. 2001. “Notation: gute Zeichen — schlechte Zeichen. Empirische Untersuchung zur (Un-)Möglichkeit von Notizen, dargestellt am Sprachenpaar Französisch-Deutsch”. In Dolmetschen. Beiträge aus Forschung, Lehre und Praxis, A. F. Kelletat (ed.), 243–265. Frankfurt: Peter Lang. Dam, Helle V. 1995. Tekstkondensering i Foredragstolkning. PhD dissertation. Faculty of Modern Languages, Aarhus School of Business, Denmark. Dam, Helle V. 2000. “On the option between form-based and meaning-based interpreting: The effect of source text difficulty on lexical target text form in simultaneous interpreting”. In Översättning och Tolkning, B. Englund Dimitrova (ed.), 51–81. Uppsala: Ekonomikum. Gile, Daniel. 1995. Basic Concepts and Models for Interpreter and Translator Training. Amsterdam and Philadelphia: John Benjamins. Herbert, Jean. 1952. Manual del Intérprete. Del Entrenamiento para Llegar a Intérprete de Conferencias. Genève: Georg. Ilg, Gérard. 1980. “L’interprétation consécutive. Les fondements”. Parallèles 3: 109–136. Ilg, Gérard. 1982. “L’interprétation consécutive. La pratique”. Parallèles 5: 91–109. Ilg, Gérard. 1988. “La prise de notes en interprétation consécutive. Une orientation générale”. Parallèles 9: 9–13. Ilg, Gérard and Lambert, Sylvie. 1996. “Teaching consecutive interpreting”. Interpreting 1 (1): 69–99. Kirchhoff, Hella. 1979. “Die Notationssprache als Hilfsmittel des Konferenzdolmetschers im Konsekutivvorgang”. In Sprachtheorie und Sprachenpraxis, W. Mair & E. Sallager (eds), 121–133. Tübingen: Günter Narr. Matyssek, Heinz. 1989. Handbuch der Notizentechnik für Dolmetscher. Ein Weg zur sprachunabhängigen Notation Vol. I–II. Heidelberg: Julius Groos. Mikkelson, Holly. 1993. “Consecutive interpretation”. The Reflector 6: 5–9. Rozan, Jean-François. 1956. La prise de notes en interprétation consécutive. Genève: Georg. Seleskovitch, Danica. 1975. Langage, Langue et Mémoire. Etude de la prise de notes en interprétation consécutive. Paris: Minard Lettres Modernes. Seleskovitch, Danica and Lederer, Marianne. 1989. Pédagogie Raisonnée de l’Interprétation. Paris: Didier Erudition. Thiéry, Christopher. 1981. “L’enseignement de la prise de notes en interprétation consécutive: Un faux problème?”. In L’Enseignement de la Traduction et de l’Interprétation. De la Théorie à la Pédagogie, J. Delisle (ed.), 99–112. Ottawa: Éditions de l’Université d’Ottawa. van Hoof, Henri. 1962. Théorie et Pratique de l’Interprétation. München: Max Hueber.
Expressing a well-founded fear Interpreting in convention refugee hearings Sabine Fenton The debate surrounding the rights and obligations of liaison interpreters juxtaposes two positions: interpreting and advocacy. In the face of increasing pressure from some academic quarters to enhance the role of the legal interpreter with the advocacy functions that interpreters often fulfil in social settings, this paper investigates the implications this might have for interpreters working in convention refugee hearings. It is argued that in these cases often the weakest members of the profession interpret for the weakest members of society and that an enlargement of the interpreter’s role could increase their clients’ ‘well-founded’ fear even further or could raise unrealistic expectations in the refugee applicant and exert an unbearable pressure on the interpreter. The author of this paper is the national president of a professional association of translators and interpreters and her interest in this investigation stems from the concerns of this body about the well-being of this group of associates of the New Zealand Society of Translators and Interpreters. Her findings are based on a corpus of questionnaires filled in by interpreters working for the New Zealand Refugee Status Appeals Authority.
Introduction The plight of convention refugees, also called asylum seekers or spontaneous refugees, touches every developed country today. In a world aflood with refugees, the question of who is granted political asylum of those fleeing the world’s civil conflicts becomes paramount. It is the task of the immigration officer to identify who is a true convention refugee on the basis of the claimants’ accounts of their life story. The interpretation of this account by an interpreter may bear heavily on a positive or negative outcome for the refugee. Success depends largely on the skills of the interpreter. The question of what these skills should include, however, and the role the interpreter should play, is far from clear. The question has in fact
264 Sabine Fenton
created a notable divide with some academic scholars on one side and the United Nations High Commission for Refugees, the League of Red Cross and Red Crescent Societies, and the judiciary on the other side of the debate, with the former favouring a liberal interpretation of the interpreter’s role and functions and making a case for interpreter latitude, while the latter insist on a more restricted role prescription. The context of asylum seeker interviews is a legal one. It is therefore fair to say that interpreters working in this environment are working under the conditions and obligations of legal and court interpreters as shown by the research of recent times (Berk-Seligson 1990; Gonzáles, Vasquez & Mikkelson 1991; Morris 1993; Edwards 1995; Colin and Morris 1996; Wadensjö 1998). This generally accepted view has, however, been challenged in recent times by a voice from one of the great immigrant receiving countries of the Western World, Canada. Barsky (1996) pleads for extending the role of the interpreter in convention refugee hearings. He maintains that interpreters should be “legally recognized as active intermediaries between the claimant and the adjudicating body”, and proposes interpreting strategies which would assist claimants by improving their chances of receiving a fair hearing. These range from “intervening with questions and clarifications that are pertinent to the case through “compensating for the claimant’s errors of judgement” to adding unsolicited supplementary information on the historical, political and social situation of the claimant’s country and finally “improving the narrative”. In short, Barsky advocates a degree of latitude, which would allow the interpreter to retell the claimant’s life story, on the assumption that the interpreter might simply be better at telling a good story. These are extreme demands of an interpreter by any measure and are grounded in Barsky’s grim view of incompetent immigration officers and unfair hearings. He sees the entire process as stacked against the claimant, and supports his argument by quoting the United Nations Handbook on Procedures and Criteria for Determining Refugee Status (1992): His (the claimant’s) application should therefore be examined within the framework of specially established procedures by qualified personnel having the necessary knowledge and experience, and an understanding of an applicant’s particular difficulties and needs. (cited in Barsky 1996: 47)
Here Barsky seems to see the interpreter as one of the “qualified personnel” examining the refugee’s case. On closer reading of the Handbook, however, one may seriously question whether the interpreter is in fact part of this “qualified personnel”. “A competent interpreter” is mentioned (p. 46) as one of the guarantees that the applicant is to be provided with, but the assessment of the applicant’s claim rests entirely with the examiner, “the competent official, e.g. immigration officer” with “the competent authorities of Contracting States” (p. 1). The Interpreter is included to make communication possible, but not to examine or judge a refugee’s case.
Expressing a well-founded fear 265
This paper attempts to demonstrate that such views as Barsky’s, as wellmeaning and compassionate as they may be, open the door to dangerous and unsafe practices for the interpreter, instil an exaggerated faith by the claimant in the powers of the interpreter, and if accepted by the interpreting community, would set the interpreting profession back by years. The arguments for this purpose are not based on any particular theory but rather on the voice and profile of the interpreters working for the New Zealand Refugee Status Appeals Authority, their work experiences and their situation within their own communities.
The Research Project and Method The research reported here was undertaken at the beginning of 2001. It investigated three main areas: 1. The personal profile of the interpreters working for the Refugee Status Appeals Authority. 2. The role perception these interpreters have of their professional activities. 3. The implications a widening of their role prescription would have for their professional practice as interpreters and for their private lives. It is hypothesised that interpreters do not see themselves as being part of the decision- making personnel in asylum seeker cases; that they have a solid understanding of interpreting ethics; and that an extension of their role would be understood by them as a violation of their professional ethics. It is also hypothesised that a more active role, which would go beyond interpreting, would also have a negative impact on their personal and professional lives. In summary, the following research represents an attempt to contribute to the literature on the role of the interpreter in legal settings in general and convention hearings in particular. The survey methodology used for data collection was a questionnaire focussing on the following main areas: the interpreter, the contractor, the client, and the job. A total of 110 survey packets were mailed out to the interpreters on the list of the Refugee Status Appeals Authority in Auckland, New Zealand’s largest city and Wellington, New Zealand’s capital. Each packet mailed to respondents contained two cover letters: one from the researcher, stating the purpose of the research project, and another from the Authority, encouraging the interpreter to take part. It also contained a questionnaire and a self-addressed franked return envelope. Full anonymity of the respondents was assured by the fact that the Authority mailed the packets out and the responses were directly mailed to the researcher. The overall response rate was 32%. The findings reported here are based on the analysis of the returned questionnaires.
266 Sabine Fenton
Analysis of the Questionnaires The first area of focus pertained to the identification of individual characteristics of the interpreters. 98% of them had come to New Zealand as migrants, a very small percentage (2%) as refugees and nearly all had been in New Zealand for more than 5 years (91%). All of them had studied English formally in a classroom setting, the majority of respondents (67%) for 10 years or longer. In terms of educational background as measured by the highest degree obtained, 60% of the respondents had received their bachelor degree or equivalent and 29% had postgraduate qualifications. With respect to respondents’ specific education and training in interpreting before they started to work for the Refugee Status Appeals Authority, only a little over half (57%) had attended a training course. The courses consisted of non-language-specific part-time courses attended in New Zealand and lasted between one week and one year. All respondents (100%) indicated that they played an active role in their communities, which varied in size from under 500 to over 2000 members. Activities ranged from interpreting and translating to teaching, to leading roles in their cultural organisations, and to voluntary community work, particularly being a contact person for newcomers, to helping them to settle in their new surroundings, and to being a minister of religion. In summary we can say that most respondents were well educated, had excellent linguistic skills but were generally lacking in interpreting education and training at the time they began working as interpreters. The roles they played in their communities made them easily identifiable. The second area of focus was that of the interpreters’ employer or contractor. An overwhelming majority felt valued by their employer (71%). This was reflected in the courteous manners of the staff, recognition of the importance of their job, praise and thanks expressed for good work, respect shown to them, and an appreciation of paid interpreter-training courses and professional development seminars. As much as this supportive attitude by the employer towards the respondents is to be welcomed, it carries a potential danger for the interpreting process. Feeling so positively about the employer and the importance of the role they are playing, interpreters may over-identify with the professional during the interpreting and thus lose their impartiality. Interestingly, however, results of the two following areas of investigation show that in fact the opposite was true. Having been treated with respect as professionals in their own right bolstered the interpreters’ pride in their own profession. Professional acceptance had the effect that they adhered closely to their professional ethics of which impartiality is, of course, a leading principle. The results of the third content area, that of the client, begin to show the pressures interpreters experience. 63% of respondents said that asylum seekers ask them for help during the interview, how they should answer this or that question. 57% said they were contacted after the interview at home. While all interpreters
Expressing a well-founded fear 267
stated that they politely but firmly denied such requests, they also said that they nevertheless felt under pressure. When it comes to being contacted at home by a claimant, this pressure intensifies to such an extent that interpreters often ask the manager of the interpreting service who arranges the appointments, “please don’t tell them my name”. The last area of the questionnaire concerned the job itself. Open-ended questions provided opportunities for unstructured, fuller answers. The enjoyment of working for the Appeals Authority came for the interpreters from using their skills (“energising and fulfilling”), working with a professional team on the one hand and helping people on the other, and the context of the assignments (“edge of seat stuff”). A great number of interpreters (68%), however, also expressed negative feelings concerning the job. Firstly there was unhappiness about the interviewers’ “hostile or wrong way of questioning”, the perceived ignorance of lawyers about a certain case, their “coldness”. “I have come across officers who have asked unclear and irrelevant questions, (some bordered on unfair) that have led to confusion on the part of the applicants, they panic, then they can’t answer the questions properly and correctly. I’ve seen some just freeze mentally and I feel sorry for them”. Other comments referred to the inability of the interviewers to judge the claimant’s educational background and adjust the tone and register of their questions accordingly. “The professionals should be aware that some of the appellants are not at all educated and that sometimes it is hard for a particular appellant to understand and to be able to answer a question, however simple it is to the professional. Getting uptight to the appellants does not really help”. And interpreters feel pity and sympathy for the appellants, who are, after all, their fellow countrymen and women and have escaped life-threatening situations and find themselves in a foreign land and foreign culture. “Some professionals and their staff should be more friendly towards the applicants. Remember they are going through a lot of stresses and horrors back at their own country. This country and the court are foreign to them”. This is close to the scenario Barsky describes and raises the expectation that these interpreters will want to intervene on behalf of the refugees. And indeed, some of the interpreters (28%) expressed the wish to be able to do just that but felt restrained by their code of ethics. But there is also resentment towards the asylum seekers when the interpreters (38%) feel they have been party to a false claim. “Their stories are so far-fetched, a lie or fabrication and I had helped the applicant to be successful”. And finally there is apprehension and wariness about their own welfare and safety. Interpreters (52%) report that they have been threatened during an interview. Well-known and easily identifiable within their communities, interpreters are being called by phone about the result of their interview. Appellants come to their house. “They bring papers for me to read about their case. They talk, they discuss, they ask my opinion.” Apprehension turns to fear for them (17%) when blame is laid on them by unsuccessful appellants
268 Sabine Fenton
who did not see their expectations realised. “If I don’t achieve what they want, they treat me with abuse, ring at night, threaten me with my life”. The last question of the survey asked the interpreters to describe how they saw their role. Surprisingly, the respondents unanimously understood their role to be strictly that of interpreting and were adamant about keeping it this way. They perceived themselves as linguistic “experts” adhering to their professional Code of Ethics. “I see my main role as the upholder of accurate interpreting, not biased towards either the Appeals Authority or the claimant”. They wanted to be “as invisible and unobtrusive as possible”. They proudly referred to their professional ethics and their role, “I take pride in my integrity and honesty”, and “I see myself as a professional aid to communication between two parties and a cultural bridge between two parties from two different cultures”.
Conclusion There were several reasons for engaging in this study that for the first time in New Zealand investigated a group of interpreters, namely those working for the Refugee Status Appeals Authority. One goal was to determine the profile of the interpreters in terms of their educational background and standing in their communities. Of further interest was their understanding of their professional role and the extent to which they would be prepared to compromise this role. In general each of these goals was achieved. We must now relate the final analysis to Barsky’s recommendations as set out at the beginning of this paper. His proposition was to extend the role of interpreters in convention refugee hearings to “active intermediaries” (Barsky 1996: 46) who follow strategies of intervention and advocacy, compensation and performance skills on behalf of their client. From the statements supplied by the interpreters in the questionnaires the following picture has emerged: 1. None of the respondents was prepared to violate their impartiality for either the professional or the client in the way Barsky proposed. The narrow role prescription was perceived and appreciated as a protection against a takeover by either side and reinforced in them their own professionalism. 2. All respondents were identifiable within their communities and access to them at home was therefore possible. If the onus of establishing an appellant’s credibility were extended to them as Barsky suggested, interpreters feared the pressure of an exaggerated faith in them. It would not only impact on their professional performance but also threaten their private lives. This is a clear rejection by the practitioners themselves of a role that goes beyond interpreting. It is perhaps grounded in the strong emphasis on ethics that part-time,
Expressing a well-founded fear 269
non-language-specific interpreting courses tend to have, and it rewards them with a sense of identity and security within their professional activity. The support and help that Barsky seeks for convention refugees must come from other sources, it cannot come from the interpreters. Involving them in anything other than interpreting would be as detrimental to the individual interpreter as to the profession as a whole.
References Barsky, Robert F. 1996. “The interpreter as intercultural agent in convention refugee hearings”. The Translator 2 (1): 45–63. Berk-Seligson, Susan. 1990. The Bilingual Courtroom: Court Interpreters in the Judicial Process. Chicago: University of Chicago Press. Colin, Joan and Morris, Ruth. 1996. Interpreters in the Legal Process. Winchester, UK: Waterside Press. Edwards, Alicia Betsy. 1995. The Practice of Court Interpreting. Amsterdam/Philadelphia: John Benjamins. González, Roseann D., Vásquez, Victoria and Mikkelson, Holly. 1991. Fundamentals of Court Interpretation. Durham, NC: Carolina Academic Press. Horvath-Lindberg, Judit and Miserez, Diana. 1991. Working with Refugees and Asylum Seekers — A Handbook. Geneva: League of Red Cross and Red Crescent Societies. Office of the United Nations High Commissioner for Refugees. 1992. Handbook on Procedures and Criteria for Determining Refugee Status. Geneva: UNHCR. Wadensjö, Cecilia. 1998. Interpreting as Interaction. London/New York: Longman.
Cross-cultural dynamics in community interpreting. Troubleshooting Mette Rudvin University of Bologna
In this paper a number of issues related to interpreter-mediated communication for public services are examined. It is argued that the encounter between host country and migrant is governed by power relations where hierarchy is constructed along a number of different axes: cultural, historical, institutional, political as well as communicative. The paper identifies some of the overt and covert power differentials which could become potential stumbling blocks in public service interpreting, both linguistic and language related features, such as: metaphor and truth value, connotation, mismatching of semantic fields, paraphrasing explicit language, social role and identity embedded in grammatical markers, indirectness as a politeness marker, turn-taking and direction.
More than other forms of language-mediated communication practices — notably (conference) interpreting and mainstream translation — community interpreting1 has had to confront the question of cultural difference between interlocutors. Of course, any interpreter-mediated situation is by definition cross-cultural (as is any interaction between two discourse communities even within the same nation-state) but community interpreting is more visibly and openly so than most other interpreter-mediated situations because the two participant groups tend to come from countries separated by large geographical distances which have had less contact, historically, linguistically and politically than, say, the European countries or the British-settled areas in North America and Australasia with evident historical and linguistic bonds (principally through settlement in the period of British imperialism). That is not to say that cross-cultural differences do not affect interaction between these groups, quite the contrary. These differences are mitigated by a common language, however, and by commonalities in certain institutional practices (e.g. legal, medical) that — for reasons of geographical, historical and political proximity — are to some degree similar. That similarity is increasing with globalisation and exchanges in the scientific community (especially in the health sector) and
272 Mette Rudvin
— within the EU — through continued ‘harmonization’ of legal and immigration practices. Another relevant issue that does not apply — or only to a lesser degree — between communities in the Western world is the clearly perceived if unstated power differential between the Western world and ‘traditional’ non-Western communities, especially in countries with recent immigration, such as Italy (the present author’s country) rather than countries which have by now a long-standing tradition of immigration and have learned not only to accept but also to appreciate the resources that new nationals bring to their countries. In those countries where the power differential is publicly expressed and blatantly part of the political discourse, especially after the events of September 11th (the “immigrants must adapt to our ways if they want to stay here” attitude), the immigrant group is expected to adapt to the host country’s norms, communication and institutional systems, as argued in Rudvin 2002 even more so than might be expected of a representative of another Western country. If the bulk of community interpreting in the West, especially in the three typical areas of community interpreting: legal, health, immigration, takes place between the host country and communities from the so-called developing world, that does not imply that community interpreting operates exclusively within a developeddeveloping country relationship. Although most recent community interpreting research has focussed on the situation in the Western world, much theoretical research and practical work (especially issues such as training and accreditation) is being done elsewhere. In South Africa, for example, Rosemary Moeketski eloquently addresses issues of power and inequality in the courtroom:2 When I first went to a trial court, I saw something else. I saw blatant inequality and disparity. I saw power juxtaposed with powerlessness; knowledge with ignorance; confidence with fear; arrogant control with humble submission; manoeuvring and manipulation with confusion and bewilderment. I saw character assassination at its worst. (Moeketski 1999: 24)
Although community interpreting in the Western world is over-represented by encounters between developed-developing (or North/South, West/East), I believe the ‘bottom line’ is that the encounter is essentially between the empowered, the master discourse community, and the subaltern, which is what Moeketski’s quote also illustrates. The role of hierarchy in the service-provider/client relationship is constructed not only in virtue of the ‘authority’ of the host country institution but, this paper argues, it is constructed historically, politically (in a First-Third World perspective), institutionally; indeed, one of the defining criteria, I would argue, of community interpreting versus other forms of interpreting is precisely its hierarchical nature and its relation to institutional power and power differentials. That is not to say, however, that the situation will remain unchanged in terms of which group constellation or set of power differentials is represented in public service
Cross-cultural dynamics in community interpreting 273
interpreting; for example, second-generation immigrants have a much stronger negotiating position than do their parents’ generation. Furthermore, demographic trends change over time, leading to changes in the power differentials between host and migrant communities. Public service interpreting has only just begun to emerge forcefully as an independent theoretical discipline. Much of the research that has been done so far has focussed either on practical aspects relating to institutions or terminology, accreditation, training programmes, or to codes of ethics. Insofar as it has addressed supralinguistic aspects, particularly the interpreter’s role, it has done so from a prescriptive standpoint, i.e. that the interpreter must at all costs be neutral and impartial and not take sides. Although there is a growing awareness of the need to study and incorporate the cross-cultural dynamics of the interpreting encounter (somewhat belated compared to the ‘cultural turn’ in translation/interpreting studies), few case-studies have as yet been conducted on concrete cases of comparative institutional practices or comparative communication systems outside the Western world, although significant work has been done within Europe (especially in legal interpreting) or between Europe/US and Spanish-speaking communities from Latin or South America. I believe that the discipline would profit greatly from such studies, and that their point of departure should be the immigration statistics of each country: that is, identify the main groups of migrants and languages, not just in absolute numbers but in terms of their distribution in the various services (legal, health, mental health, refugee administration etc.) and examine the differences between the institutional practices of host and migrant (e.g. cultural framework as it relates to religion, health, views on integration, crime and violence, sickness and health expectations from institutions and the host country, as well as communication models as they relate to institutional discourse modes). While forced to confront the mediation, management or bridging of cultural and linguistic difference, the interpreter has traditionally been expected to adhere to the traditional ideal of ‘translating everything sender says’ which so often results in a ‘tension’ between equivalence and effective provision of services. The present paper regards the community interpreter as an active “meaningcreating” agent in this type of cross-cultural interpreting encounter and thus is firmly rooted in the descriptive tradition, as presented in Toury’s seminal work. Cecilia Wadensjö’s approach is also descriptive: … I have started from the open question of how interpreting happens. I have not asked whether or not interpreting happens in certain ways, in which case I would have assumed a hypothesis to be verified or falsified. I have rather been asking, as part of my analyses, how control and responsibility for a communicative exchange is distributed between interpreters and those for whom they work.3
In that same spirit, this paper does not ‘prescribe’. More modestly, I hope to engage
274 Mette Rudvin
in a bit of troubleshooting, to identify some of the areas which I think represent potential significant stumbling blocks in public service interpreting, and in so doing also suggest possible areas of research for descriptive studies. In terms of the muchdiscussed code of ethics for community interpreters, I believe that it is crucial for this area of interpreting research to adopt a global approach in which the needs and realities of the service provider and the host institution are considered — as well as those of the client — rather than attempting to construct a code based on absolute values of ‘neutrality’ etc. divorced from the real world. If interpreters continue to be trained from the perspective of interpreting theory only and not also as active and acting agents in the community, the discrepancy between theory and practical needs will place an unsustainable burden on the interpreter’s shoulders and lead to psychological stress.4 It is important to highlight, in research, how differences in institutional systems and communication models impede the statutory rights of migrants and smooth interaction between migrants and the host community; it is therefore essential to educate service providers in order to make them aware of problems hindering migrants’ communication competence. If, in a narrow-minded equivalence-based approach of ‘translating what we hear’ we ignore the real-world needs of both client and service provider — and in doing so overestimate the ease of communication — we are doing a disservice to both parties.
The communication process and power asymmetries The following section will examine, then, some possible trouble spots in crosscultural dynamics.5 I would not go so far as to suggest solutions to such problems, but self-awareness, for all parties, is an important starting point. Indeed, as Lisa Mellman says when discussing counter-transference in court interpreters, understanding “is crucial for accurate interpreting. This awareness can then mitigate in the forensic setting against the natural vulnerabilities that everyone has” (Mellman 1995: 471) and in the context of the present discussion could, I believe, help avoid the most blatant communication blunders and foster a better relationship between client and service provider. Rather than putting the responsibility on the interpreter’s shoulders, the institutions themselves should address cross-cultural dynamics and train their operators to work with other language communities, drawing on the interpreters’ skills and know-how in their capacity as ‘informed party’. If, as discourse- and conversation analysis suggest, hierarchy and power differentials are crucial to discourse forms in Western society, especially discourse in institutionalised settings, it is even more so in many so-called ‘traditional societies’. Indeed even ordinary conversation which is reciprocal and therefore largely egalitarian (at least in the layman’s view), is often strictly governed by power dynamics (familial, social and professional status, age, gender, kinship). Below I have listed some of the cultural factors relating to overt and covert power relation-
Cross-cultural dynamics in community interpreting 275
ships that have emerged in the course of my own professional experience and research, as potential communication problems in inter-cultural interpretermediated encounters for public services.
Linguistic features Perhaps the most obvious examples of mismatches between languages at the supratextual level are features like metaphor, connotation, mismatching of semantic fields, paraphrasing explicit language, etc. Interpreting metaphor is not in fact just a question of finding another metaphor or paraphrasing: the perceived information value, communicative performance potential or contextual appropriacy of metaphor is highly culture-bound. Many so-called ‘traditional cultures’ use metaphor far more frequently than Western cultures and give more weight not just to the range of applicability but its validity as a descriptive tool and as a ‘truth-recounting’ tool. Symptoms by a client in a hospital room might well be described through images or metaphor; a patient may be used to describing her symptoms through metaphorical expressions or images, for example pain being ‘hot’ or ‘cold’ (the body itself being a social, cultural and historical production). Or in the courtroom a witness might use roundabout expressions or paraphrases during cross-examination, which could easily be perceived by the judge as ‘inaccurate’. Such paraphrasing can have disastrous consequences: at the Critical Link conference on community interpreting in Montreal in 2001, the participants were shown a staged video clip of a real rape case where the Farsi-speaking (Iranian) woman interpreter rendered the witnesses’ statement about sexually explicit acts through paraphrase (“he was disrespectful to her body”) rather than the more direct form which the prosecutor was soliciting to strengthen the case against the defendant. The witness came across as looking unconvincing and the defendant was acquitted. In explaining her conduct to the judge afterwards, the interpreter explained that she was greatly limited in her freedom of expression for cultural reasons of ‘shame’ and dishonour, that is she was unable to repeat sexually explicit language as it would “bring dishonour on her and her husband to say such words”. Her paraphrasing might have seemed as clear to her as the explicit language which the public prosecutor was trying to obtain and which to him had a higher truth-value in accordance with Western epistemology. For judge and jury metaphor did not carry adequate ‘truth-value’.5 Had both parties been more aware of the different uses of metaphor and appropriacy of using direct and explicit sexual language, the conflict might have been avoided. Perhaps the most obvious and frequent trap in cross-cultural communication at all levels of proximity (conceivably even within the same language) is when semantic fields do not overlap. One example that is pertinent to community interpreting (for reasons of identification, for example) is kinship terms. Terms such as ‘mother’ and ‘brother’ are particularly ambiguous. In Australian Aboriginal
276 Mette Rudvin
English (hereafter AAE) for example, ‘mother’ is also ‘mother’s sister’, an indication of the fluidity and continuity in aboriginal kinship systems.6 In Pakistan, as in many other countries, ‘brother’ can also mean ‘cousin’ (distinguished as ‘distant or proper brother’), indicating the strength of the extended family system and solidarity between kin. The use of the father’s name as surname in many Muslim cultures can also cause unnecessary misunderstandings, suspicion and confusion when a witness, defendant or patient is asked to identify him/herself or his/her kinsmen — or indeed identify another person in terms of their kinship relations. Kinship terms are also used for respect (e.g. ‘older sister’ for a woman one’s own age or slightly older and the ubiquitous ‘auntie’ or ‘uncle’ in both Hindi/Urdu and English, especially but not exclusively used by children). Kinship terms are a useful strategy to avoid using first names, which in many contexts is extremely inappropriate and/or impolite, as Tariq Rahman notes (1999). The use of titles (professional or otherwise), honorifics and other forms of address is also of crucial importance in many societies and governs conversation at all levels, in both formal and informal discourse. Social role and identity may be embedded in grammatical markers denoting humility (i.e. that the user is inferior to the addressee on a shared hierarchical scale) and/or rank, position, or professional role in a social relationship: for example the use of the 3rd person singular or 1st person plural about oneself (in Urdu or Farsi, for example). The 1st person plural ‘ham’ might be used to affirm professional status, for example.7 Group identity, hierarchy and role are consolidated and strengthened by the use of such social markers, which may be lost on the interlocutor. Untrained in both language and discourse modes of the host community, the client is unable to compensate for the lack of such strategies, and the interpreter is hard put to create, through compensation or other means, on the spur of the moment, a global strategy that corresponds in function to what the client believes s/he is communicating. Lack of communicative competence in the discourse modes of the host community, however competent the subject is in his own discourse community, can seriously jeopardize communication and the desired goal (justice, treatment, etc.).
Supralinguistic communication features Politeness, so crucial to interactive communication, is quintessentially culturebound and particularly prone to misunderstandings.8 Susan Berk-Seligson’s empirical study (1990) on discourse strategies in court interpreting is perhaps one of the best-known texts on how politeness rules and the interpreter’s mediation of these affect interpreting and sometimes the judicial outcome. In her stimulating paper on courtroom and therapeutic discourse Robin Lakoff notes how politeness can only be expressed if there is consensus for it in that particular discourse community (Lakoff 1989: 103): Discourse, she warns, is “intelligible to all partici-
Cross-cultural dynamics in community interpreting 277
pants” only when all agree “on the necessity for politeness, or the opposite; and agree on the form the politeness is to take. Otherwise one person’s meaningful contribution will be read by another as anomaly, craziness, or worse”. It is quite clear that in interpreter-mediated communication each party refers to his or her own cultural politeness model and the two are not necessarily “intelligible to all participants”. It is generally the client who must adapt to the host country’s communication norms and will be evaluated on the basis of his/her competence in managing those norms. If politeness competence is essential for communicative success (Lakoff 1989: 104), and in terms of each party’s success in obtaining the desired outcome (e.g. client acquittal or prosecutor charge), then one’s purpose in engaging in the discourse in the first place will be accomplished best by staying within the politeness system, including the use of that form of politeness expected by one’s culture in the specific context. Abrogations of the rules will typically be read as opting out of the system, ordinarily through incompetence, unsociability or desperation. (ibid.)
Clearly a member of a different speech and/or language community is at a grave disadvantage. Furthermore, despite the formality of the courtroom context, deliberate aggression or rudeness by the cross-examiner is often common (to ‘haul out the truth’). Defendants with different educational backgrounds from the lawyers or judge who have not therefore perfected the art of verbal confrontation (Lakoff mentions members from the working or under-class, but the same applies to members of other language groups) are disadvantaged when aggressiveness is used to provoke the witness into looking bad — and therefore guilty — in front of the jury through “systemic, interactional and non-reciprocal rudeness”, for example by means of aggressive tag questions (‘you don’t want to die, do you?’) (Lakoff 1989: 125). Unskilled or weaker parties, she notes, are denied their rights: “Their questions are not answered, their motions denied, their objections routinely overturned. The appearance of dispassion that pervades the language gives the impression of disinterestedness and non-confrontation” (Lakoff 1989: 113). In nonadversarial justice systems with no tradition of adversarial courtroom rhetoric, the defendant or witness might be at a great loss as to how to proceed; as Berk-Seligson demonstrated in a different context, when politeness is neglected or used inappropriately the jury’s or judge’s verdict may be unfavourably influenced. If in Western courtrooms direct communication is expected and required, this clashes strongly with those cultures which uphold indirectness as a politeness marker;9 questions are generally asked indirectly, for example the Pushtu and other languages “how have you come?” and never “why have you come?” as a sign of respect. In AAE information is requested in a ‘roundabout way’ and obtaining information is a two-way process (see footnote 6). Direct questions are considered to be rude, as is direct eye contact. In fact, AAE has no specific grammatical
278 Mette Rudvin
structures to formulate direct questions but tends to use rising intonation and tags (eh, inna, unna). Knowledge is gained by hinting without asking directly and this information-seeking norm governs language organization. Question-and-answer sessions, so essential to Western courtroom cross-examination, can become a communicative ordeal for clients used to mitigated questions and answers, and their discomfort can easily be misread by the interlocutor as untrustworthiness. Similarly, silence and hesitancy can be interpreted as shiftiness or non-cooperation while the person might just be thinking of an answer that will best please the interlocutor. Silence may indeed be a sign of respect to be observed by both speakers, to give the interlocutor time to think over what’s being said. One might argue that a professional interpreter should be able to identify and translate or somehow signal the importance of such linguistic markers or mismatching of semantic fields into the target language, but (apart from the fact that for practical reasons this is often not feasible) interpreters for clients from the large international language groups such as English/French/Spanish/Arabic will often not be from the client’s own country; apart from the fact that the vast variety of “englishes”, “frenches”, “spanishes” etc. have different semantic and grammatical features from the ‘standard models’ of these world languages, they clearly do not share their cultural referents. So not only do the communication models differ (directness, distance, politeness, semantic overlap, etc.), even if they are speaking the same language (indeed many interpreters find it difficult to even understand varieties of a particular language), their real-world referents do not necessarily overlap. Varieties of standard languages (esp. colonial languages) are sometimes described as ‘cross-cultural’, or ‘nativized’ — that is the language has adapted semantically, grammatically to its new terrain and culture (for example English in the Indian subcontinent or in East or Central Africa), and reflects that reality and world-vision rather than say the cultures of the UK or the US. So interpreters are in fact often interpreting without having access to their client’s cultural referents or language elements specific to that variety of English, for example. Turn-taking and direction in the conversation may vary greatly between discourse communities in terms of what is considered to be appropriate and polite. AAE speakers, for example, have very different conventions for group-speaking than their fellow countrymen of European origin. Whilst the Anglo-Saxon tradition requires individual speakers to alternate the direction of their speech, distributing turns evenly among interlocutors (at least in non-hierarchical conversation) and keeping constant eye-contact, ideally including all members of the group in a conversation. AAE talk, however, is not projected to any individual, nor is any response expected (hence eye contact is not important). Talk is simply broadcast generally, there being no mandated pattern of turns among speakers, and any response is up to an individual hearer. Further, participants in the conversation can freely tune in or out. (Foley 1998:252)
Cross-cultural dynamics in community interpreting 279
It is not only each individual communicative act which is culture-bound, however, but also the complex interplay between these acts. Ron Scollon notes that although turn-taking and topic introduction (which are closely linked, he argues) might coincide for different groups such as New Yorkers and Tokyoites, who are both said to “take over turns very rapidly”, “topic introduction in some cases is very deductive (early introduction) for New Yorkers and inductive (topic introduction follows extended face work) for Tokyoites” (Scollon 2001: 83). Scollon argues therefore that it is “the linkage of turn exchange to topic introduction which produces the experience of cultural difference, not the practices for turn exchange which are quite homologous” (ibid.). The ethno-linguist William Foley reports the case of the highly egalitarian Ilingot people in the Philippines and juxtaposes their special oratorial style, ‘crooked speech’ used in judicial negotiations with the Western adversarial courtroom rhetoric (Foley 1998: 255). In ‘crooked speech’ persuasion is sought and accepted, and conflict is resolved by finding terms for agreement that satisfies both parties. ‘Crooked speech’ is characterized by a number of structural features: altered stress patterns, phonological elaborations of words, metaphors, repetitions and paraphrases, and puns. People’s names are avoided, they are rather referred to through social position, and there is frequent recourse to hedging (for example by the reduplication of words) which is necessary to offset potentially insulting or threatening remarks (Foley ibid.). Ethnic affiliation or non-affiliation is an essential criterion to consider in interpreter screening, but far from straightforward. Although some clients prefer to use out-group interpreters, usually for fear of confidentiality breaches, many prefer the sense of security and familiarity — in what may often constitute a hostile host environment — that a member of the same ethnic group may give them. This may cause problems, and/or increase the stress level, for the interpreter in terms of client bonding (sometimes the client even seeking favours). This difficulty can be exacerbated by the fact that many service providers (especially in hospitals) expect the interpreter to function as emotional ‘care-giver’, confidante, or practical helper as well as ‘translator’. It can be extremely damaging to engage interpreters who are in social, political, ethnic or religious opposition to the client (for example cases where the interpreter is a representative of an oppressive regime from the home culture or where an interpreter of a majority religion has been chosen to interpret for a discriminated minority-religion client); loyalties in such cases are severely tried, and such pressure on both client and interpreter should be avoided. In certain cases, however, intergroup conflict and mutual suspicion can be so intense that the client would prefer an interpreter from outside his or her own ethnic group, fearful that information from the case may leak out into the community and damage his/her standing in that community. In other cases, an unknown interpreter from the same
280 Mette Rudvin
ethnic or language group may be suspected of siding with other sub-groups from the same community involved in the case. One client I interpreted for in the city court, for example, was far more suspicious of his own fellow nationals than he was of the host country representatives or system at large, trusting that the system would be ‘on his side’ against his co-national rival (another suspect). Unfortunately such blind trust placed in the host system is rarely rewarded. It is not uncommon for subjects to seek identification with, or even patronage from, the high-status group (in this case the host country), and to deliberately seek to distance themselves from and also ‘win one over’ their co-national(s). This is apparently not an infrequent phenomenon in the construction of group identity generally.10 It might be interesting to examine, statistically, connections between community size and identity and preference for in- or out-group interpreters. Naturally subject matter will also make a difference (a woman requesting an abortion, or some other delicate matter, might prefer an out-group interpreter). Naturally, many other factors impact heavily on the cross-cultural communicative encounter. Although a discussion of each of the following lies beyond the scope of the present paper, some factors could be mentioned: gender, the cultureboundedness of truth/objectivity, a subject’s primary loyalty to group or individual versus institution/nation, presenting and selecting information on the basis of seeking-top-lease the interlocutor (or on the basis of the appropriacy of divulging certain information), patronage, class hierarchies (internally in the host country, i.e. the judge/doctor/immigration officer are generally members of the middle classes), systemic institutional differences, colonial issues and the authority of the host country and its institutions.11 The result of power imbalances and cross-cultural differences such as these are potential obstacles to effective communication between service provider and client. Further complicating the situation is the possibility — or even likelihood — that each interlocutor, representing his/her own discourse community, may be completely unaware of the fact that these differences do in fact hinder communication, and see resulting communication problems or misunderstandings as a consequence of the interlocutor’s inadequacy or guilt, rather than recognising it as simply a lack of communicative competence in the host country’s discourse modes. In other words, as Scollon so convincingly argues,12 all social practices — even the most simple ones such as walking on the pavement or handing over an object — are constructed in the agent’s habitus. The interplay between these practices are constitutive of “group, class or categorical habitus” (Scollon 2001: 82). Problems of interaction among interlocutors within any given group arise when there is social or cultural incommensurability between these socially constructed acts, each practice situated within a network of social relations varying in strength and thus competing with eachother. In the context of immigrant-host community two complex communicative and social networks are interacting and the degree of
Cross-cultural dynamics in community interpreting 281
incommensurability is clearly much higher. As a rule, the newcomer is hard-put to compete in the host community’s communicative and social model and function competitively in that new habitus. S/he will not only have to identify and learn the interlocutors’ social practices, but must choose whether to adopt or jettison these practices. For this reason it is crucial that service providers and institutions are included in the educating and training process, and that the interpreting act is situated in the framework of the community as a whole, not solely in the interpreting profession.
Mandates and authority It has been suggested above that it is virtually always the client who must adapt to the host country’s communication system and institutional requirements, not only because the legal or medical institution demands it, but because of the power equilibrium between client and host country representative as ‘host-guest’ or ‘provider-seeker’ (e.g. seekers of residence and/or employment in general or of specific services). As Rahman notes, the norms and values of the less powerful culture tend to be replaced by those of the powerful culture in a meeting of the two (Rahman 1991: 193). Furthermore, the behavioural patterns as well as the needs of immigrants change, naturally, over time and when a higher degree of integration and language-sufficiency has been reached, the migrant might (but again might not) wish to make the host country’s norms their own and attempt to emulate the communication models of the host country. In some cases, the norms of the higherstatus group (the host society which enjoys higher economic and political status for the economic migrant) will be more attractive to the migrant, as it represents precisely privilege and status. This will depend of course on the degree of attachment to their own home country (for example, a member of a discriminated minority may seek to identify him/herself with the host country rather than with the oppressive climate of his/her own home country) and their own sense of group identity. The quote from Moeketski at the beginning of this paper clearly shows that the power differentials apply not only to a Western non-Western relationship but to any relationship between dominant group and minority group.13 We must not, however, lose sight of the fact that as well as being a representative of ethnic groups and institutions, each interlocutor is also an individual and each interpreting encounter is a unique event in which all parties strive for some, if not all, common goals — be it a diagnosis or an information exchange with a police officer/judge — and is motivated to resolve whatever the problem may be. I am not suggesting that cross-cultural communication is impossible but rather that both interpreter and service provider should be fully aware of what they are up against.
282 Mette Rudvin
Notes 1. Legal interpreting is frequently included under the umbrella term ‘community interpreting’ or ‘public service interpreting’ and the two are used interchangeably for the purposes of this paper. Its classification in this category has been the subject of some discussion, but is becoming increasingly accepted. The reason for its exclusion, I suspect, is perhaps that it enjoys higher status than interpreting for hospitals or immigration services. The communication dynamics and the power features (especially in terms of master discourse versus the subaltern) of interpreting in the health sector share much in common with police/court interpreting. 2. See e.g. Erasmus 1999 and Moeketski 1999. 3. Wadensjö, pre-conference seminar at the EST conference, Copenhagen, August-September 2001. 4. As Baistow reports in her findings from a survey in 2000 among community interpreters. 5. Briefly discussed in Rudvin 2002b. 6. Duranti (1997: 38–39) provides a very useful list of references to literature on metaphor. 7. The two following internet sources provide numerous examples: www.une.edu.au/langnet/ aboriginal.htm by D. Eades and www.ntu.edu.au/education/csle/student/goodwin/goodwin0.html by M. N. Goodwin. 8. This feature is similar to what happens when children and mothers speak of themselves in the 3rd person (‘mummy will put baby to bed now’) precisely to affirm and then consolidate that relationship in the formative stages of the relationship. Such affirmation serves to give both mother and child added security in terms of their own roles and the two roles in relationship to each other. 9. As discussed in Rudvin 2002b. See also Geertz 1982 on the culture-specificity of politeness rules. 10. See e.g. Adegbija 1989. 11. See e.g. Kim 1994 and Liebkind 1995. 12. Some of these issues are further discussed in Rudvin 2002. 13. Scollon 2001, esp. p. 25 and pp. 80–81. 14. It is my impression that this phenomenon is far more pronounced in legal and administrative settings (e.g. refugee applications) however, than in health settings where service providers are beginning to recognize the necessity of understanding their patients’ mode of expression, in all its facets, in order to be able to reach a correct diagnosis. The flourishing of cross-cultural awareness courses in many countries is a positive result of this attitude.
References Adegbija, E. 1989. “A comparative study of politeness phenomena in Nigerian English, Yoruba and Ogori”. Multilingua. Journal of Cross-Cultural and Interlanguage Communication 8/1: 37–56. Bahadir, Sebnem. 2001. “The empowerment of the (community) interpreter: The right to speak with a voice of one’s own”. Paper presented at the conference Critical Link 3: Interpreting in the Community: the Complexity of the Profession (Montréal: May 2001). Available on: www.criticallink.org/english/proceedings.htm.
Cross-cultural dynamics in community interpreting 283
Baistow, Karen. 2000. Dealing with Other People’s Tragedies: The Psychological and Emotional Impact of Community Interpreting. Vienna: Babalea. Barth, Fredrik. 1969. Ethnic Groups and Boundaries. Boston: Little, Brown. Berk-Seligson, Susan. 1990. The Bilingual Courtroom: Court Interpreters in the Judicial Process. London: The University of Chicago Press. Duranti, Alessandro. 1997. Linguistic Anthropology. Cambridge: Cambridge University Press. Erasmus, Mabel. 1999. Liaison Interpreting in the Community. Pretoria: Van Schaik. Foley, William. 1998. Anthropological Linguistics. An Introduction. Oxford: Blackwell. Geertz, Clifford. 1989. “‘From the native’s point of view’: On the nature of anthropological understanding”. In Culture Theory. Essays on Mind, Self, and Emotion, Richard A. Shweder and Robert LeVine (eds), 123–136. Cambridge: Cambridge University Press. Kleinman, Sherryl and Copp, Martha A. 1993. Emotions and Fieldwork. London: Sage Publications. Lakoff, R. T. 1989. “The limits of politeness: Therapeutic and courtroom discourse”. Multilingua. Journal of Cross-Cultural and Interlanguage Communication 8(2/3): 101–129. Kim, Uichol et al. 1994. Individualism and Collectivism. Theory, Methods, and Applications. London: Sage Publications. Eades, Diana. 1991. “Communicative strategies in Aboriginal English”. In Language in Australia, S. Romaine (ed), 84–94. Cambridge: Cambridge University Press. Liebkind, Karmela. 1995. “Ethnic identity — challenging the boundaries of social psychology”. In Social Psychology of Identity and the Self Concept, Glynis M. Breakwell (ed), 147–185. London: Surrey University Press. Malcolm, I.G. Malcolm and Rochecouste, Judith. 1991. “Australian Aboriginal English discourse”. In Englishes around the World. Caribbean, Africa, Asia, Australasia. Studies in Honour of Manfred Görlach, Edgar W. Schneider (ed), 268–278. Amsterdam and Philadelphia: John Benjamins. Mellman, Lisa A. 1995. “Countertransference in court interpreters”. Bulletin of Academic American Psychiatry Law 23/3:467–471. Moeketsi, Rosemary. 1999. Discourse in a Multilingual and Multicultural Courtroom: A Court Interpreter’s Guide. Pretoria: J L van Schaik. Rahman, Tariq. 1999. Language, Education and Culture, Karachi: Oxford University Press. Rudvin, Mette. 2002a. “Challenging the ‘Neutrality Maxim’”. An Interactional Approach to Community Interpreting”. In Perspectives on Interpreting, Giuliana Garzone and Maurizio Viezzi (eds), 217–233. Bologna: CLUEB. Rudvin, Mette. 2002b. “Cross-cultural aspects of community interpreting in Italy. In Community Interpreting and Translating: New Needs for New Realities, Carmen Valero Garcies and Mancho Barés (eds), 127–132. Madrid: University of Alcalà. Scollon, Ron. 2001. Mediated Discourse [The nexus of practice]. London and New York: Routledge. Wadensjö, Cecilia. 1998. Interpreting as Interaction. London and New York: Longman.
The child in the middle Agency and diplomacy in language brokering events Nigel Hall Manchester Metropolitan University
The term language brokering has now been used in a number of recent papers exploring the interpreting experience of children. The term brokering focuses attention on the whole cultural meaning of such an event, in which any interpretation is simply a part. Being a child language broker is a complex experience; it challenges conventional power relationships within families or schools, challenges chronological age expectations, and allows children to act strategically as experts. This chapter considers the language brokering behaviour of group of Asian heritage children taking part in a complex simulation in which they mediated an interview between a school representative and a non-English speaking mother. Built into the interview were dilemmas for these children, and this paper will explore the strategic nature of their responses, and the extent to which they were willing to exert their power as experts.
Introduction In 1987 Knapp-Potthoff and Knapp had a book chapter published which was titled, “The man (or woman) in the middle: discoursal aspects of non-professional interpretation.” This was a rare foray into the world of non-professional translation and interpretation, and the subject seems relatively untouched since. This is very surprising as their chapter provides a particularly interesting counterpoint to the substantial and increasing literature on professional translation and interpretation. Indeed, examining some of the professional literature may lead a reader to assume there is only professional translation and interpretation: From the point of view of the language mediator, namely the interpreter or translator, the languaging task is not normal. It is a professional venture. (Neubert, 1997: 2)
286 Nigel Hall
In fact, such claims are very far from the truth. The vast majority of translation and interpreting events are quite normal and are mediated by non-professionals as part of their everyday family and community experience. By and large it is the volunteers, friends, relatives and children who simply get on with the task of interpreting for those who need it. Such people rarely have formal training, are usually unpaid and unrewarded, and have varying levels of competence and experience. For all this inexperience, such activities are in many circumstances often critical to a family’s or community’s social, economic and administrative well being. If the role of adults as ‘natural translators’ and interpreters (Harris, 1992) has been largely ignored, then what about children who carry out these roles? They are truly the children in the middle. In this paper the children’s behaviour discussed is designated as language brokering. The term language brokering derives from the notion of cultural brokering developed by the anthropologists Wolf and Clifford (Robbins, 1996) to describe the activities of individuals who connect local with national worlds. Bailey (1969) distinguishes between a pure messenger, a type of intermediary most dependent on primary parties and their actions, and the ‘broker’, a more active and powerful role which has both overview and influence in the mediating situations. Paine (1971) suggests that a ‘broker’ takes more independent measures to initiate action or promote negotiation. The extent to which a young person takes this more powerful role may well depend upon particular indivduals or particular settings, but it is clear that in some circumstances this role is exactly that undertaken by children (Hall and Sham, 1998). The term language brokering has now been used in a number of recent papers which have sought to explore aspects of the translation and interpreting experience of children (Shannon, 1990; McQuillan and Tse, 1995; Tse, 1996; Hall and Sham, 1998). The purpose of this study is to explore what it means for a child to be ‘in the middle’ rather than an adult or a professional, by collecting data from events involving young children. The notion of ‘being in the middle’ represents an attempt to explore talk as activity rather than talk as text; as a dialogical process rather than monological one (Wadensjö, 1998). While it is important to know how successful they are in the translation role, and that while success is partly dependent upon textual knowledge, it is also important to understand how the child language brokers act contextually. To achieve the objective of studying the ‘child in the middle,’ it is necessary to investigate the child as a practitioner of language brokering. While this probably sounds obvious and straightforward, it is in fact highly unusual and complex. Research into childhood translation and interpreting is very limited, but in the few studies undertaken the primary form of data collection has been through retrospective interviewing of subjects (McQuillan and Tse,1995; Tse 1996; Hall, and Sham, 1998) or in one case, parents (Kaur and Mills, 1993), and the observation of events has been very rare (only Harris and Sherwood, 1978 and Shannon, 1990). The reliability and
The child in the middle 287
validity of accounts in interviews is inevitably untested in such studies, and as personal narratives they may be coloured by many factors. Accordingly, a study was developed to explore child language brokering practices.
The study The data To generate data from the actual practice of child language brokering and yet avoid getting caught in complex ethical, technical and procedural problems, it was decided to set up a simulation. It dealt with the problem of seeking permissions from the adult participants and it also took care of all the practical difficulties of locating events and recording them. It also provided the children with an opportunity to act in a situation that was going to treated as real and yet would not generate the negative consequences that might derive from a failure in a real event.
The children One year earlier a group of nine-year-old UK children from families of Pakistani origin took part in interviews about their language brokering experiences. The children had been selected after a very simple questionnaire asking if they ever had to translate for other family members. The children were in classes at a school in an area with a large population of families of Pakistani origin, and the main family language for all the children was Urdu. One year later, the children were invited to participate in the simulations. Most who were still in the school agreed to do so. This provided a group of seven children, five boys and two girls. All except two had been born in the UK, one of the others arriving as a baby and the other when aged seven. Six of the children spoke Urdu and English, while the other child also spoke some Punjabi.
The script The literature on translation and interpreting, the relatively few existing published studies on child language brokering, and previous work by the author, led to the development of a basic script for each simulation that contained a number of elements posing some challenges to the children. The word ‘challenges’ is not meant to imply anything that would put the children in some discomfort; it simply means that certain problems would occur in the interview that would need to be resolved by the children. The general constraints of the script were that it needed to create an interview event that would last no longer than around ten minutes (in practice the
288 Nigel Hall
average length of the interviews was just under ten minutes), and that with certain specific exceptions contained language and concepts in English and Urdu within the ability range of the child. The script was developed with an experienced Urdu/ English speaker (who later played the role of the mother in the simulated interviews). A trial was carried out with a child two years younger than the children in the research study, and some small modifications were made. The script was based around an interview between a non-English Urduspeaking mother of a young boy who wanted to enrol her child in a school and a representative of the school who spoke no Urdu. The script was conceived as having seven sections, although these were partly analytical categories. Two sections are considered in this paper. In Section 6, the mother is asked about her child’s behaviour. The mother’s script has her seek to enlist the child broker in lying to the interviewer. Section 7 has the mother asking what her child would do in school. When told, she reacts negatively leaving the child broker with the decision about answering accurately or modifying the response.
The setting The interviews took place in the deputy headteacher’s room, which had been rearranged slightly to facilitate seating and recording. The use of the deputy headteacher’s room was partly because it was offered by the school, but also because from the children’s perspective it was, however formalised, a perfectly appropriate place to conduct such a school-based interview. The chairs were set around one end of a long table, with the child sitting in the centre seat where he/she could see clearly both the other participants. The interview was recorded with both video and audio; only one child acknowledged the existence of the miniature video camera. The mother was in full Asian dress and spoke only Urdu during the event.
Ethical considerations and procedure It would have been possible to set up the simulation and yet allow the children to believe that it was a real event. However, children have always had a relatively unpowered position within research, often do not have their own views and needs fully considered by researchers, and are often unreflectively coerced into participation in research events over which they have no control and about which they have no say (Morrow and Richards, 1996). Our ethical position demanded that the children be genuine volunteers, who understood as far as we could make possible what was going to happen and why we wanted them to participate. Permission was secured from the parents and the school, and meetings took place with all the children beforehand to explain fully what we were doing and to seek their agreement to participate in the simulation. The notion of simulation was fully explained
The child in the middle 289
to the children who were, anyway, from their own lives and from school familiar with the concept through both socio-dramatic play and drama experiences. The children were also told that the simulations were going to be video-recorded. On the day before the interview all children were introduced to the participants, and the Urdu speaker acting as the mother spoke to all the children in Urdu. The only information not revealed to the children was that the mother spoke English.
The data and analysis As the data is being used to examine aspects of the child’s brokering as a social process, only the English text and the translation of the Urdu texts are used (but full transcripts for this article with the Urdu are available from the author). Participants are indicated by T for the teacher, M for the mother and C for the child. The translations of the Urdu speech are represented by italics; all other text represents speech in English. Episode 1: Discussion about the child’s behaviour Extract 1: Shahela T Is he a well-behaved child? C Does he behave? Like does he be like naughty? Is he well? M What can I tell her? He is a very naughty child, but what shall I say to her? Children should come to school. If he’s naughty or dirty will they send him home? C He’s naughty M Tell her he’s good! C She says, “If he’s naughty at school, will you keep him?” Extract 2: Syed T Does he get on well with other children? C Does he play properly? Does he play with other children properly? M Can I lie because actually my child is quite naughty? C What? M He’s naughty… should I tell her the truth or not? C Your child is very naughty? M He is very naughty. What should I do? C Is your child very naughty? T What’s she saying? C Tell her. M Should I tell her or not? C Tell her.
290 Nigel Hall
T C
What’s she saying? Her son’s a bit naughty.
Extract 3: Massood T Is he a well-behaved child? C Does he be good? He doesn’t be naughty? M Shall I lie or tell the truth? He’s a very naughty child, but what shall I tell her? T What’s she saying? C He’s a bit. He is sometimes a bit naughty, but sometimes good. T OK. C But he behaves most of the time. Episode 2: Discussion about children playing in school Extract 1 Shahela T Is there anything you would like to ask me? C Do you want to ask her anything? M What can I ask? What do they do in school the young children? C She wants to know what little children do at school. T Well, when the children are young we like them to play a lot. C Most of the time they play. M They play? They play? They come to school to play? T What is she saying? C She said they doesn’t come to school to playing. T Well don’t worry. It’s playing with things that help them learn. C They say work things that you play with that help them learn…like… with sand like with water. That’s like science. M To learn to read and write. That’s why we send them to school. C I know. Extract 2 Syed T Is there anything you would like to ask me? C Do you want to ask her anything? M Yeah. What do they do in school? C What? M What do they do in school? C What do you do in school? T When the children are young we like them to play a lot. C When the child is young they make them play. M Not playing! Do we send our children to learn to read and write or play? T What’s she saying? What’s she saying?
The child in the middle 291
C T C M
She doesn’t want him to play a lot. Don’t worry it’s playing with things to help them to learn. They learn him something. They learn through play, then they learn something. Oh, all right then.
Extract 3 Massood T …Is there anything you would like to ask me? C Do you need to ask her anything? M What do the children do in school? C What does…what do children do at school? T Well, when the children are young we like them to play a lot. C First they play a lot. M They play? Do we send them to play or to read? C He’s young now. He’ll go to the nursery, then play he’ll with everyone. Then they won’t do any work. They’ll play with everyone. M Yes, but he… C When he goes to a big class, then he’ll get into work. He’ll… T What’s she saying? C Eh…she said …uhm…do they work …uhm…and I said in the nursery you always play and when you go up to the juniors you have to work. T O. K. Is she happy with that idea? C Are you happy? M Who with? C About this idea? M I am …You should read… I don’t like playing…but … C Yeah.
Analysis of the children’s language brokering in the two situations Power is a central and very interesting issue in child language brokering events, especially those that take place within a school. On the whole, the power of children within schooling is quite limited. Institutionally, historically and legally considerable power is invested in school authorities. It starts with the compulsory nature of schooling, is deeply embedded within the organisation of schooling, and is accompanied by a wide range of sanctions if that power is not acknowledged by children. Typically, teachers are the ones who control classrooms: they adjudicate in disputes, administer justice, regulate work, hold knowledge, manage organisation, command behaviour, dispense favours, and administer rewards and punishments. Research into classroom language suggests that most teachers dominate oral dialogue, have difficulty in listening to children’s voices and manipulate children’s language performance by sanctioning particular forms of verbal behaviour. Within the home
292 Nigel Hall
children may have greater degrees of freedom, but they are still legally and practically subservient to parents, although the extent of this may vary from culture to culture and within cultures may vary between genders and ages. Child language brokering poses some difficulties for this distribution of power. Typically the adults, be they parents or teachers, become dependent upon the children, and the children become the experts. It is the adults who appear de-skilled and have to invest considerable trust in the children, even sometimes quite young children. This is not likely to be a comfortable relationship for either the adults or the children, and while the children will often insist that their parents trust them, evidence from some studies suggests that this trust is more imagined than real, while evidence from other studies suggests that even if such trust existed it might be misplaced (Weinstein-Shr, 1994; and Wagner, 1993). However, it is also the case that child participation in brokering events challenges notions of chronological age and childhood, two things that are frequently the subject of stereotypical expectations about what children can be expected to accomplish. Solberg (1990) uses the phrase ‘social age’ to represent the shifts that can occur when children take on and succeed with different kinds of problems that demand responses above and beyond those normally expected of children of the same chronological age. The dilemma for the child in both the situations represented in the transcriptions lies in whether having accepted the role of translating fairly what is being said by the other participants, s/he can take sides and in the process modify, maybe substantially, what is being said; can s/he act as an independent agent rather than simply as a translator/interpreter. Another way of putting this is, with whom does the child feel s/he has a social contract? There are a number of tensions inherent in the children’s role in the simulations: –
–
–
On the one hand, they are members of the community of that school and in many circumstances would be likely to show considerable support for that school; on the other hand, they are members of a family and ethnic community outside school, a grouping that has a strong sense of identity, reinforced through ethnicity, language, dress, etc. On the one hand, they have been enlisted as an agent by the school, an institution that has formal power invested in it, to represent it; on the other hand, natural justice demands that the views of the mother, who is in the position of supplicant, are represented fairly. On the one hand, they are children, and are well aware of their social, familial and institutional status, one which positions them as dependent and irresponsible rather than accountable, answerable, capable, competent, independent, reliable and trustworthy; on the other hand, they have been selected precisely because they have experiences and skills that are not possessed by the adult participants in the setting — they are the experts. They have their agency
The child in the middle 293
–
–
recognised and acknowledged by people who in other areas of life would be more likely to be distrustful of their actions. On the one hand, their role as translator could be seen as a monologic, mechanical role in which skills are involved but in which the translator is simply expected to perform a technical task; on the other hand, the event is a triinteractional one in which basic language skills are embedded within ability to comprehend and sustain complex dialogic processes. On the one hand, the task is simply about helping two individuals converse with each other: on the other hand, the event is a cultural exchange in which any meaning (hardly a simple concept at the best of times) is situated within partially overlapping and often contradictory networks of associations, histories and values.
Thus for any child self-positioning in the interview takes place amidst competing claims for allegiance, competing expectations relating to chronological and social age behaviour, the contrastive complexities of monologic and dialogic performance, and the contrastive contexts of the immediate setting and the broader cultural meanings. Any assumption of power, any use of initiative and any exercise of agency by the child is a consequence of decisions relating to all the above. How did the above three children respond in relation to these tensions and conflicts? Of the children involved in the episodes above, Shahela appears the most likely to avoid being agentive. While she does not offer precise translations, in a truncated form her responses express simply the point of the mother’s statement. It seems that far from offering allegiance to the mother, she carries out her task in a way which fulfils the school’s need for honest response from the mother to the questions asked. She seems less likely to exert power, but to simply perform in the task in a more neutral way, almost as a professional translator and interpreter might seek to do. The evidence available from the whole interview event offers another perspective. Shahela was, in the view of the Urdu-speaking researcher, by far the most fluent child in Urdu, and the task seemed almost too easy for her. Her responses came close to being simultaneous with the mother’s Urdu. Both adult participants and the author, who was observing on a monitor, got the impression that she was on a kind of automatic pilot and was not fully aware of what she was translating, until after she had said it. When she does realise what she is saying she moves into social repair strategies, in the first episode checking the reaction of the school representative to what she has said, and in the second spending some time on her own explanation. The other two children seemed much readier to explore the complexities of the social situating of their brokering. There was a clear attempt on the part of both to exert agency in modifying texts to maintain social equilibrium, so that the interview did not degenerate into argument or break down completely. Both attempted to qualify the negativity in the mother’s words or attitude. Syed seems to register the
294 Nigel Hall
mother’s comment about naughtiness with some disbelief and repeats his question to her three times, getting the same answer each time. After the teacher’s intervention to find out what has been going on, he is prompted to appeal to the mother for guidance about what to do. Her reply is emphatic but nevertheless Syed manages to answer with discretion. Massood again offers a striking diplomatic response, neither quite misrepresenting nor quite conveying completely the mother’s comments. The situation for the children was different in episode two. While the children could not have been expected to know anything about the mother’s child, when issues about schooling arise they are in possession of knowledge and experience. It should not be so surprising that Massood bypasses the school representative and answers the mother’s question himself. He responds to the mother’s annoyance with an explanation, and quite a good one, about the developmental notion of education, but at a prompt from the school representative tells her what he has said to the mother. Later, when there is a misunderstanding about being happy, he cuts off the mother’s complaints by simply offering as an answer to the school representative’s question, “Yeah.” Syed, perhaps as a result of having survived the problem in episode one, this time modifies the mother’s responses, softening them to “she doesn’t want him to play a lot,” completely ignoring in his interpretation the annoyance and annoyed tone of the mother’s statements. For these two boys resorting to a linguistic form of social lubrication not only keeps the interview moving, but avoids having to struggle with explaining to an adult that another adult disagrees with them. To some extent the above comments, focussing upon the behaviour of the children, may seem monologic in that they explore the text and strategies of only one of the participants in the triangle, but the children are actually significantly shaping the whole event, including the behaviour of the two adults. While in this simulation prior knowledge and a preconceived script allowed the adults to maintain an overall perspective on, and control of, the developing structure, the children significantly challenged this control by their behaviour. The children in these brokering episodes were offering clear demonstrations that for them brokering was multi-level interactive activity demanding awareness of local and global contexts as all the participants jointly constructed their text. For the children, moving from one language to another in a brokering event was a relatively minor part of the process; choosing what to say was a more complex process than simply translating what had been said. They were clearly able to use with agility, tact and discretion the social knowledge they possessed, and demonstrated high-order social, cognitive and linguistic skills that in all probability greatly exceeded those demanded by the monologic textual practices within schooling. To a large degree their decisions might be regarded as highly dubious by a professional translator; consistently they modified what was said by the other two participants, and in other
The child in the middle 295
circumstances this might cause significant problems. Only exploration of other types of brokering encounters will reveal whether children can shift strategies in response to different situations. But the problem of being a child in the middle is that the recognition of her/his status limitations ranked against the competing expectations of the other two participants makes ensuring harmony within the encounter a more secure experience than going for local text accuracy. For the child, both alternatives have costs and benefits, but in this case diplomatic decisions help the interview proceed, keep relationships intact, and avoid blame for participant dissonance.
Conclusion This chapter illuminates an area of interpreting behaviour that has been largely ignored by the research literature. It is not only professionals who translate and interpret; indeed, the contrary is true, and children, for many reasons, form a huge language brokering constituency. Harris and Sherwood (1978: 55) argued that the data for studying translation “should come primarily from natural translation”. While few would agree with this, it is important that research in this area should explore translation, interpreting and language brokering as social activity, not simply as professional activity. The behaviour of the children in this study demonstrates the complexities of the social context for child language brokers, the social and practical dilemmas they face when asked to broker between adults, and the highly subtle and strategic responses generated when negotiating a pathway through these complexities. Their language brokering worlds are worthy of critical enquiry, for unlike professionals, they do not have training, support mechanisms or payment; they have to work it out for themselves. Author’s Address: Nigel Hall Professor of Literacy Education Institute of Education, Manchester Metropolitan University, 799 Wilmslow Road, Didsbury, Manchester UK M20 2 RR Email:
[email protected] References Bailey, Frederick. 1969. Stratagems and Spoils: A Social Anthropology of Politics. Oxford: Blackwell. Hall, Nigel and Sham, Sylvia. 1988. “Language brokering by children”. Paper presented at the Annual Conference of the British Educational Research Association, Belfast, 27th–30th August 1998.
296 Nigel Hall
Harris, Brian. 1992. “Natural translation: a reply to Hans P. Krings”. Target 4: 97–103. Harris, Brian and Sherwood, Bianca. 1978. “Translating as an innate skill”. In Language Interpretation and Communication, D. Gerver and H. Sinaiko (eds), 155–170. New York: Plenum Press. Kaur, Sukhwant and Mills, Richard. 1993. “Children as interpreters” In Bilingualism in the Primary School: a Handbook for Teachers, E. Mills and J. Mills (eds), 114–125. London: Routledge. Knapp-Potthoff, Annelie and Knapp, Karlfried. 1987. “The man (or woman) in the middle: discoursal aspects of non-professional interpreting”. In Analysing Intercultural Communication, K. Knapp, W. Eninger and A. Knapp Potthoff (eds), 182–211. Berlin: Mouton de Gruyter. McQuillan, Jeff and Tse, Lucy. 1995. “Child language brokering in linguistic minority communities: effects on cultural interaction, cognition, and literacy”. Language and Education 9 (3): 195–215. Morrow, Virgina and Richards, Martin. 1996. “The ethics of social research with children: an overview”. Children and Society 10: 90–105. Neubert, Albrecht. 1997. “Postulates for a theory of Translation”. In Cognitive Processes in Translation and Interpreting, J. Danks, G. Shreve, S. Fountain and M. McBeath (eds), 1–24. Thousand Oaks, California: Sage Publications. Paine, Robert. 1971. “A theory of patronage and brokerage”. In Patrons and Brokers in the East Artic, R. Paine (ed.), Toronto: Toronto University Press. Robins, S. 1996. “Cultural brokers and bricoleurs of modern and traditional literacies: land struggles in Namaqualand’s coloured reserves”. In The Social Uses of Literacy, M. Prinsloo and M. Breier (eds), 123–140. Amsterdam: John Benjamins. Shannon, Sheila. 1990. “English in the Barrio: the quality of contact among immigrant children”. Hispanic Journal of Behavioural Sciences 12 (3): 256–276. Solberg, Anne. 1990. “Negotiating childhood: changing construction of age for Norwegian children”. In Constructing and Reconstructing Childhood: Contemporary Issues in the Sociology of Childhood, A. James and A. Prout (eds), 118–137. London: Falmer Press. Tse, Lucy. 1996. “Language brokering in linguistic minority communities: the case of Chineseand Vietnamese-American students”. The Bilingual Research Journal 20 (2/3): 485–498. Vasquez, Olga, Pease-Alvarez, Lucinda and Shannon, Sylvia. 1994. Pushing Boundaries: Language and Culture in a Mexicano Community. New York: Cambridge University Press. Wadensjö, Cecilia. 1998. Interpreting as Action. London: Longman Wagner, Daniel. 1993. Literacy, Culture and Development: Becoming Literate in Morocco. Cambridge: Cambridge University Press. Weinstein-Shr, Gail. 1994. “Literacy and second-language learners: A family agenda”. In Adult Biliteracy in the United States, D. Spener (ed.), Illinois: Centre for Applied Linguistics, and Delta Systems.
The editorial process through the looking glass Daniel Gile and Gyde Hansen Université Lumière Lyon 2 / Copenhagen Business School
Editors of scholarly publications are gatekeepers who decide whether to publish manuscripts or not, but their decisions are constrained by institutional factors, sociological factors, and the availability of referees for peerreviewing. The quality of the input of such referees depends not only on their thematic and methodological expertise, but also on their motivation, on their personal bias and on sociological factors. Although this suggests that editorial decisions are less objective and reliable than their idealized representation, the referees’ comments are often convergent to an encouraging degree. The main comments made by referees on 42 manuscripts submitted to these proceedings are listed. They share a large common denominator, which suggests that the fundamental norms they are based on are the same.
1.
Introduction
Conference proceedings are a well-established tradition, and are often awaited impatiently not only as a reflection of the meeting, but also as a publication vector for participants, and frequently as a justification to their university for their presence and participation in the event. However, publishers do not necessarily allow sufficient space for all papers to be published, and editors must choose. How they do this is not always clear to authors who submit their manuscripts. Basically, the process is very similar to the editorial process in other types of collective volumes and in journals in that it is based on the peer review system. This paper discusses the concept, as well as recurrent issues, in particular as they crystallise in the field of Translation Studies; it then discusses issues in refereeing, and briefly mentions follow-up possibilities after comments are received from the reviewers. It is based on the authors’ experience as editors in the field, and also includes, by way of illustration, a case study with an analysis of the most common types of weaknesses reported in papers submitted for publication in proceedings in the field of TS.
298 Daniel Gile and Gyde Hansen
2.
The editorial process
2.1 The editor’s role as a gatekeeper In the scientific community, where a scholar’s publications are his/her visiting cards, passports and stepping stones towards professional promotion, editors of journals and collective volumes have an important three-pronged institutional role to play. Firstly, they create and offer publication space to individual scholars. Secondly, they act as gatekeepers and guarantors of the scientific quality of texts they publish (see Judd et al. 1991: 17f.). Thirdly, through the peer-review system, they offer guidance and text-improvement opportunities. As gatekeepers who decide to publish or not to publish, editors seem to be in a position of power. And yet, their action is constrained by several factors, including difficulties in finding adequate referees for proper implementation of the peer review system, and “political” and diplomatic considerations that inevitably influence their decision-making process. 2.2 Constraints and policies Thus a system which is theoretically designed to provide reliable guidance and assessment to editors in their role as gatekeepers in the publication process has inbuilt limitations which can be difficult to overcome. To these limitations, one should add the expectations of the publisher in some cases, and of the researchers’ community in other cases (for instance, when editing the proceedings of a conference), not to mention the expectations and reactions of individual researchers, who may need to publish for institutional or personal reasons. Several types of policy are observed in the field. The easiest one to implement consists in accepting virtually all papers, and merely performing some cosmetic surgery on the wording, correcting errors in the lists of references, re-formatting papers to applicable standards. While this may not be the most efficient policy in terms of scholarly quality and promotion of solid scientific norms, it can be considered legitimate when the purpose is only to reflect the state of the art in a field, or to provide the written record of a scholarly meeting. At the other extreme, editors may decide to implement a strict quality policy, accepting only manuscripts which pass severe scrutiny. Such a strategy is possible for editors of prestigious journals in established disciplines where there is much research production and much competition. In TS, due to the small size of the community and the general state of research, it is doubtful whether any editorial board can afford to act in such a strict manner. Most often, the editorial policy with respect to quality is somewhere inbetween, and quality thresholds for acceptance vary as a function of supply and
The editorial process through the looking glass 299
demand: if many good papers are offered and fill the required number of pages, there is no need to accept mediocre submissions. When there is more publication space than material, acceptance standards fall. In such a situation, it is difficult to set standards in advance. For each issue of a journal or collective volume, submissions may have to be refereed first, and the decision to publish or not to publish is taken when the total picture is clear. 2.3 Referees in the peer review system As its name indicates, the peer review system is based on the work of scholars with a similar background to the authors’, whose role is to provide useful input, which takes the form of assessment for the benefit of the editor, as well as guidance to the authors. It is significant, but not often noticed, that in its very name, the system refers to “peers” rather than to “experts” or some other authority-carrying agent: the idea, which is corroborated by experience very regularly, is that even without a substantial difference in training and experience between referees and authors, the very fact that it is easier for anyone to detect another person’s weaknesses than one’s own, makes it possible for the former to help the latter (“… well-trained scientists can make mistakes and sometimes overlook important problems of design” (McBurney 1990: 16)). So much for theory. In practice, the system is a bit more complicated. Firstly, “peers” are not always peers. Sometimes, acknowledged authorities are asked to do the reviewing, and sometimes, referees are much less experienced and qualified than the authors themselves. In the ‘guidance’ part, i.e. specific comments on errors or weaknesses and specific suggestions for improvement, this is generally not too problematic. In fact, excellent criticism is often received from conscientious beginners, who may react with high sensitivity to all segments in a text which contain unclear, ambiguous writing or faulty logic, whereas more experienced readers already familiar with the field, who may also be busier, may focus on more specific points and overlook such segments. With respect to overall assessment of manuscripts by referees as opposed to specific comments, experienced, more highly qualified scholars are clearly in a better position than beginners to provide useful input. In established disciplines with a large community of scholars, referees may not be too difficult to enlist. In other cases, and in particular in the rather small, but diversified TS community, finding such reviewers may be a real problem. Firstly, in some sub-areas of a discipline, there may simply not be more than a very small number of scholars competent enough for refereeing purposes. The problem in international events (as opposed to national ones) is compounded in papers written in languages other than English, which has more or less become the international language of science, in the field of TS as in other fields. More generally, refereeing effectiveness can be discussed under several components.
300 Daniel Gile and Gyde Hansen
2.3.1 The language issue Though not a scientific or intellectual criterion, the language barrier is formidable, even in a multilingual society such as the TS community (or perhaps especially in this community, where authors may insist on their right to write in their language and count on their fellow-researchers’ comprehension). Not only are language skills essential when assessing and/or correcting the style of a text, but they are also important in terms of comprehension, especially in theory-oriented papers. How many TS scholars can review German texts on court interpreting, Spanish texts on research methodology, French texts on contrastive linguistics and translation? The problem was a very concrete one in this volume as well, when scholars well versed in certain fields turned out not to have the language in which the relevant papers were written. In several instances, the less than ideal choice was between refereeing a paper written in a language one knew well, but in a field not familiar enough, and refereeing a paper addressing a field one knew well, but written in a language in which one did not trust one’s comprehension ability fully. 2.3.2 ‘Thematic’ expertise The requirement for sufficient relevant knowledge on the specific theme in the field is obvious, but in practice, it is not always easy to meet, because editors do not always have access to referees with the required qualifications. When they lack ‘thematic’ knowledge, readers may miss errors in the manuscript, especially in the literature review and references, but also in the interpretation of quotations and in the use of concepts. On epistemological and methodological issues, they may also fail to see the relevance of certain points made by authors, or miss gross errors. When the referee’s lack of familiarity with a field leads to unjustified criticism, the refereeing process can become counter-productive. The most conscientious and careful referees faced with this dilemma tend to focus on the structure and rationale of the paper and to formulate questions rather than assertive criticism when in doubt, thus alerting the authors and editors to the existence of potential problems, without making any damaging claims. 2.3.3 Methodological expertise In a field as heterogeneous as TS, research paradigms are diverse and quite different from each other. Some scholars may adhere to the methods and norms of cognitive psychology, others to those of literary studies, of philosophical discourse, of sociology, of formal linguistics, etc. These are not necessarily mutually compatible, and if referees adhere to one paradigm and the author of a refereed text to another, this may lead to negative assessments predicated on an opposition between schools of thought, not on actual weaknesses in the paper within the paradigm chosen by the author. Challenging a paradigm is legitimate, and a potentially powerful driver of scientific progress, but criticising an author’s method using the criteria of a
The editorial process through the looking glass 301
paradigm other than the one s/he has chosen to follow can be counter-productive (see Kjørup 1996: 109, 113, 116 f.). 2.3.4 Motivation However ‘objective’ and impersonal the peer review system is supposed to be within the idealistic view of science, personal factors may play a significant role in the refereeing process. Critical reading is serious work, as it involves reading a text carefully, making sure one understands it, assessing it, reporting on the findings, and making suggestions for improvement. If there are serious design flaws, and if weaknesses are numerous, it may take much time and effort to list them comprehensively and to suggest alternative formulations, inferences or actions to help an author understand the problem and do the appropriate remedial work. In such cases, the willingness of referees to devote considerable work to the process can make much difference, and it is not rare to see one reviewer only send back one or a few general paragraphs as input to the editors, and another offer several pages of detailed comments on the same paper. 2.3.5 Personal bias and sociological factors One final component, which is not in line with the ideal of objectivity in science, but which is nevertheless endemic in human society, is personal bias. Scientific society is competitive, both in terms of theories and schools of thought and in terms of personal rivalry (see Plastow and Igarashi 1989). Actually, as analysed by Kuhn and many others (see Kuhn 1970, Chalmers 1990), such sociological forces may reasonably be considered useful in that they drive researchers into accumulating evidence to attack existing paradigms (or defend them, depending on which side they are on), thus contributing to scientific development. Nevertheless, in an activity having such a strong assessment component as refereeing, a positive or negative attitudinal bias, whether conscious or subconscious, can result in unbalanced judgement. Scholars may be eager to find fault with their opponents’ or rivals’ work, or may be reluctant to criticise their allies’ and friends’ writings. The problem is compounded by status issues: when reading a well-known, experienced scholar’s work, referees may be reluctant to criticise, not only out of respect for the author, but also out of fear of ill feelings or ‘retaliation’ if recognised. Indeed, in a community as small as that of Translation scholars, it is often easy to recognise an ‘anonymous’ author and an ‘anonymous’ referee through their distinct style, ideas and references.
302 Daniel Gile and Gyde Hansen
3.
Common problems in submitted papers: A case study
3.1 Introduction After the EST Congress in Copenhagen, more than 50 papers were submitted for inclusion in the proceedings. Following the tradition and conventions of scientific editing and publishing, they were processed through the peer-review process. All papers were reviewed by at least 2 referees, and most of them by three. In a few cases, only general assessments were received (see above), but specific comments were made on 42 of them. Referees mentioned both positive aspects, such as the presence of innovative components, interesting information, interesting ideas, etc., and problematic areas. In line with its focus as explained in the introduction, this paper analyses the latter, where constructive work could be done, but should not be misinterpreted as a pessimistic statement or diagnosis. The following types of perceived problems were reported by referees. Since this is a single case study, they are given not as a demonstration of any general proposition, but by way of an illustration of what we consider a not atypical pattern. Also note that for reasons that readers will understand easily, no attempt will be made to present the full set of data, with all the comments made by each referee on each paper. This information will remain confidential, and only a synopsis will be given here, in the hope that it will be sufficient to provide interesting and useful input. 3.2 Frequently mentioned weaknesses in papers for this volume –
–
–
–
The most numerous comments from referees report missing data: in order to gain better understanding of the author’s claims or conclusions, reviewers felt that they needed data which was not provided, on experimental conditions, on actual results, on the nature of the texts involved, etc. (also see Strauss/Corbin 1998: 268ff., Judd et al. 460ff.) This problem was reported in 18 out of the 42 papers (43%) for which detailed comments were received from referees. In a related but not identical category, in 11 papers (26%), reviewers reported problems with the bibliographical references, which were missing or incorrect, and/or cited secondary sources inappropriately (also see Gibaldi 1999). The second largest number of comments pertained to weaknesses in the rationale of papers. Reviewers reported that authors made unsubstantiated or excessive claims, speculated, mixed categories, over-generalised, and more generally made inferencing errors in 17 out of the 42 papers (40%) (see Judd et al. 1991: 27ff., 258f. on various types of validity and bias). This problem in the rationale is linked with problems in the design of empirical studies. Out of 20 papers reporting such studies, 12 (60% — the highest proportion) were reported as having a design that did not correspond to the
The editorial process through the looking glass 303
– –
– –
stated aim of the study, or as comparing groups that were not comparable, or using samples that were too small, or not representative, or as failing to control relevant variables (also see Judd et al. 1991: 75ff., 215ff.). The third largest set of comments considered that papers (13 out of 42, that is 31%) presented no innovation or added value. Other frequent comments had to do with specialised concepts, which were not defined at all or not defined clearly enough, or used in a way different from their established use in other disciplines or in paradigms from which they were taken (9 papers out of 42, that is 21%). Referees also commented on the excessive length of papers or segments thereof, in particular the introduction (8 papers out of 42, that is 19%). Still other comments reported clumsy or pompous style, repetitions in the structure of the paper, missing units in graphs, abbreviations and acronyms which were not spelled out, lack of indications as to the objectives of the papers or studies (see Judd et al. 1991: 472 about procedure and style).
3.3 Assessing the truthfulness of the referees’ comments Before analysing these data, a few points must be made: Firstly, no strong claim is made as to the degree of representativeness of the sample of papers, of referees and/or of comments. The numbers are given here as a rough indication. Nevertheless, the following could be pointed out: –
–
–
The set of more than 42 papers reviewed for these proceedings can be considered of reasonable size for a sample of this nature, and we (the authors of this paper and editors of this volume) did not feel it differed markedly in respect of problem areas from other sets of papers received and reviewed for other volumes in the field in recent years. An important point is that while there were a few disagreements (less than a handful) between the referees and/or editors on the general value of papers, no disagreement was expressed on any specific comment. Thus, while it is not claimed that referees are holders of the truth (they are as fallible as any scholar — see the discussion in Section 2.3), in this case, there is good reason to consider that their judgement on specific weaknesses has some reasonable inter-subjective value. Referees and editors in this case study came from different countries and have different backgrounds. These factors suggest that the picture which emerges is not likely to be a highly biased reflection of the general situation in TS. There has been a strong will on the part of the editors (there was at least one editor among the referees of each paper) to stress quality in the editorial process, in spite of the other factors mentioned in Section 2.1.
304 Daniel Gile and Gyde Hansen
3.4 Analysing the data Looking at the data, it appears that referees have reacted on the basis of a number of norms, which are made explicit in research textbooks (see for instance, McBurney 1983, Shipman 1988, Babbie 1992, Judd et al. 1991, Gibaldi 1999, Strauss/ Corbin 1998), in particular the requirements that science be explicit, in providing all the information required for readers to be able to understand and assess reports on scientific studies (“Great care is taken to specify the exact conditions under which observations are made so that other scientists can repeat the observations if they desire and can try to obtain the same results as the first scientists” (McBurney 1990: 7; see also Strauss/Corbin 1991: 266f. about replicability), that it be logical in its inferences, and that scientific endeavours contribute at least some innovation, rather than just review existing data and information (in the words of James Heckman, editor of the Journal of Bone and Joint Surgery who gave a plenary at a medical conference in Paris in the early nineties — undated manuscript), “the reviewer will ask three questions: Is it true? Is it new? Does it matter?” — also see Beaud 1988). A strikingly large proportion of papers received, more than one in three, did not observe these principles. This suggests a lack of specific awareness of the importance of these norms in research which, in turn, could be ascribed to the lack of appropriate training in research principles and their concrete implementation. A third interesting observation is the relatively high number of problems arising from trans-disciplinary ventures. As they explore new avenues with theories and concepts from related disciplines, scholars with a solid track record in TS seem to become susceptible to misperceptions. Beyond the case study of this volume of proceedings, problems have been observed when TS scholars without the corresponding background have ventured into statistics, philosophy of science, linguistics, and cognitive psychology. Such weaknesses could probably be avoided by co-operating with researchers from the relevant disciplines, and, of course, by appropriate training in such disciplines.
4. Referees’ comments and beyond The next and obvious questions are, firstly, what authors can/should do after they receive the referees’ comments. Most authors only view the peer review process as an assessment, and focus on its short-term effects in terms of acceptance or rejection of their manuscript and of requirements for rewriting. However, it could also be an excellent opportunity to improve research and writing skills. Comments can be classified into three levels of difficulty for authors:
The editorial process through the looking glass 305
–
–
–
5.
Some comments are easy to deal with. Following minor suggestions for grammatical correction and stylistic improvement, correcting factual errors, correcting errors in the bibliographical references, improving the graphic presentation of data are ‘mechanical’ and should cause no problems to authors. At a second level of difficulty, the referees’ suggestions challenge the authors’ ‘local’ assessments (on specific points): does a previous finding really mean what the paper says it means? Does another author really say what the paper reports him/her as saying? Is a quoted piece of work good enough to be used as a reference? Should another study not be cited to strengthen the case? Is the newly coined terminology appropriate, or should it be replaced by better terms? Should a particular assertion based on the findings not be toned down? At the highest level of difficulty are suggestions that challenge fundamental choices and inferences made by the author. In some cases, if the design of an empirical study is seriously flawed, if a theoretical discourse on a concept, model or theory is predicated on a serious misunderstanding of its nature, the paper cannot be salvaged. In other cases, it is possible to improve it without starting again from scratch. For instance, the data may be re-processed with a different method, and findings can be re-interpreted, which may lead to a new conclusion, which may not be in line with the ideas that authors set out to defend. Again, this may be associated with considerable psychological difficulty for them.
Conclusion
The peer-review process and the subsequent decisions and actions discussed here are not the only components of editorial work. Upstream, there are contacts with one publisher or several, definition of editorial policy, sometimes calls for papers with specific guidelines as to the length and format of texts. Downstream, when revised papers have been accepted, there is still much reading, checking, writing, copy-editing, etc. However, the most stimulating and constructive part certainly lies in dealing with the issues highlighted in this paper. The preparation of proceedings involves a chain of actions, decisions and reactions from authors, referees and editors. However clear and straightforward the principles may be, their implementation is therefore lengthy, complex, and sometimes a bit painful. However, if referees and editors have enough time and good environmental conditions for serious work on the manuscripts, and if authors accept their comments as potentially useful input, the editorial process offers excellent skill-honing opportunities for all parties in the process.
306 Daniel Gile and Gyde Hansen
References Babbie, Earl. 1992. The Practice of Social Research. (Sixth Edition). Belmont, California: Wadsworth. Beaud, Michel. 1988. L’art de la thèse. Paris: Editions la découverte. Chalmers, Alan. 1982. What is this Thing called Science? (Second Edition). Milton Keynes: Open University Press. Gibaldi, Joseph. 1999. MLA Handbook for Writers of Research Papers. (Fifth Edition). New York: The Modern Language Association of America. Judd, Charles M., Eliot R. Smith & Louise H. Kidder. 1991. Research Methods in Social Relations. (Sixth Edition). Orlando: Harcourt Brace Jovanovich College Publishers. Kjørup, Søren. 1996. Menneskevidenskaberne. Roskilde: Roskilde Universitetsforlag. Kuhn, Thomas. 1970. The Structure of Scientific Revolutions. (Second Edition, enlarged). Chicago: The University of Chicago Press. McBurney, Donald. 1990. Experimental Psychology. Belmont, California: Wadsworth. Plastow, Michael & Yoshihide Igarashi. 1989. The Mind of Science. Tokyo: Kyoritsushuppan. Shipman, Martin. 1988. The Limitations of Social Research. (Third Edition). London and New York: Longman. Strauss, Anselm & Juliet Corbin. 1998. Basics of Qualitative Research. (Second Edition). California: Sage.
Name index
A Abraham 90, 96, 217, 218 Adab 182, 194, 195 Adamic 114, 117, 118, 121, 123 Adegbija 282 Aguilar-Amat 223, 225 Ahlqvist 31, 32, 38 Ahrens vi, xii, 227–9, 236, 251, 261 AIIC xii, 241, 242, 247, 256, 260, 261 Aijmer 143, 155 Aitchison 56, 69 Al-Khanji 241, 249 Al-Salman 241, 249 Albrecht 84, 96, 296 Alexieva 230, 236, 251, 252, 256, 261 Andres 251, 260, 261 Antos 46, 48, 50, 229, 236 Araújo 173, 178 B Babbie 304, 306 Bailey 286, 295 Baistow 282, 283 Baker 7–10, 12, 25, 60, 69, 155 Barsky 264, 265, 267–9 Barthes 37, 41, 50 Bassnett 39, 43, 50 Bastian 88, 96 Beaud 304, 306 Beerbom 94, 96 Bensimon 27, 29, 33, 37 Bergenholtz 200, 207 Bericat 138, 139 Berk-Seligson 264, 269, 276, 277, 283 Berman 4–6, 8, 9, 12, 27, 35, 37, 216, 218
Blum-Kulka 7, 8, 10, 12 Blumenthal 92, 93, 96 Bolinger 53, 69 Borges 31, 37 Bredsdorff 144–6, 151, 155 Brisset 220, 221, 225 Brøndum 79, 80 Bühler 230, 235, 236 Burton 31 Bush 144, 155 Buzzo Margari 94, 96 C Cankar 114, 116, 123, 124 Cardoso de Camargo 63, 69 Carliner 194 Carroll 13, 33, 38 Chafe 53, 69 Chalmers 301, 306 Chesterman v, vii, xiii, 1, 4, 12, 24, 28, 35, 37, 124, 179, 195, 225 Chevalier 10, 12 Chiaro 168 Clifford 283, 286 Cohen 1, 12 Colin 264, 269 Collados 127, 139, 230, 236 Collados Aís 139, 230, 236 Combettes 74, 80 Corbin 302, 304, 306 Coseriu 84, 96 Cronin 225, 249 Cruttenden 229, 231–3, 237 Crystal 114, 124, 228, 229, 231, 234, 237
308 Name index
D Dahl 34 Dalitz 92, 96 Dam vi, xii, 251, 252, 258, 261 Darbelnet 8, 9, 13 Daró xii, 230, 237 Davies 114, 115, 124 de Bot 56 DeCarrico 52, 54–6, 70 Defoe 37 Delabastita 158, 166, 168 Delisle 97, 170, 178, 261 Denissenko 240, 241, 246, 249 Diekmann 128, 130, 139 Dolet 3, 12 Dollerup 97, 155, 261 Dörner 188, 194 Draheim 210–3, 215–8 Du-Nour 28, 37 Duranti 282, 283 Dyrberg 200, 207 E Eades 282, 283 Edwards 162, 163, 264, 269 Ellis 54, 58, 64, 69 Ende 94 Englund Dimitrova 8, 12, 261 Erasmus 282, 283 Eriksson 74, 76, 80 Eronen 182, 194 Eurén 30, 38 F Farthofer 182, 194 Feyrer 93, 96 Fillmore 40, 50, 52, 71 Fisiak 85, 96 Flower 46, 50, 151, 154 Foley 278, 279, 283 Føllesdal 2, 12 Fraser 13, 155, 187, 194 Frawley 7, 13 Freihoff 189, 195 Fusco 90, 96
G Gadamer 41, 42, 50 Galland 31, 32 Gambier ii, xiii, 27, 28, 33, 37, 124, 179, 195, 225 Geertz 282, 283 Gerhardt 31, 37 Gibaldi 302, 304, 306 Gieseler 215, 217, 218 Gile ii–iv, vi, xiii, 15–6, 25, 59, 69, 90–1, 96, 127, 139, 230, 237, 239, 246–7, 249, 256, 258–9, 261, 297 Goatly 99, 100, 111 Goldman-Eisler 229, 237 Goldsmith 29, 30, 38 Gonzáles 264 Goodwin. 282 Göpferich 182, 194 Gottlieb 154, 157, 165, 166, 168 Gozzi Jr. 100 Grady 100, 111 Graham 182 Grave 152, 222, 226, 277 Green 76 Greenberg 121 Gülich 87, 88, 90, 96, 233, 235–7 Gummerus 34 Gutt 40, 41, 50–52, 69 H Hall vi, xiii, 13, 111, 155, 285, 286, 295 Halliday 100, 111 Halverson 10, 13 Hämeen-Anttila 31, 37 Hammond 100, 111, 237 Hansen ii, iii, iv, vi, xiii, 62, 69, 77, 80, 297 Hasan 100, 111 Hatim 69, 126, 139 Haussteiner 182, 194 Hayes 46, 50 Heibert 158, 166, 168 Helevä 241, 246, 249 Heltai v, viii, 51, 59–61, 64, 70 Hempel 1, 13
Name index 309
Henschelmann 86, 96 Herbert 239, 249, 256, 261 Hermans 20, 25, 28, 37, 170, 178 Herslund 74, 76, 80, 81 Hjørnager-Pedersen 145, 151, 154 Holmes 25, 101, 111 Holz-Mänttäri 126, 139 Hönig 126, 129, 139 Howarth 55, 56, 59, 61, 70 Huber 227, 230, 232, 233, 237 Hughes 100, 111 Hutchins ii, 200, 207 I Igarashi 301, 306 Ilg 92, 96, 251, 252, 256, 261 Irwin 37 J Jäger 85, 96 Jänis 126, 139 Jansen 212, 217, 218 Järventausta 84, 96 Jereb 115, 117, 118, 122, 124 Johnson 99, 100, 111 Jones 240, 249 Jörg 90, 96 Judd 298, 302–4, 306 K Kalina 90, 92, 96, 228, 235, 237 Kaur 286, 296 Kenny 8, 13, 249 Kidder 306 Kielhöfer 85, 96 Kim 282, 283 Kintsch 235, 237 Kirchhoff 227, 228, 237, 251, 256, 261 Kiuru 30, 37 Kivi 30 Kjørup 301, 306 Klancar 115, 117, 118, 123 Klaudy 8, 11, 13, 18, 25, 58, 59, 62, 70 Knapp 285, 296 Koch 89, 90, 97
Kocijancˇicˇ Pokorn 113, 124 Koller ii, 84–6, 97 Königs 86, 94, 95, 97, 139 Korzen v, viii, ix, 73, 74, 77–81, 85, 97 Koskinen v, vii, viii, 27, 36, 170, 178 Kovala 37 Krashen 114, 124 Kreisig 217, 218 Krockow 219, 226 Krogh 91, 92, 97 Kropat 92, 97 Kruse 216–8 Kuhn 301, 306 Kujamäki 28, 37 Kundera 5, 6, 13 Kunnas 33, 34 Kussmaul 58, 70, 170, 179 L Lado 85, 97 Lakoff 99, 100, 111, 276, 277, 283 Lambert ii, 170, 178, 225, 226, 237, 251, 252, 261 Lane 31, 37 Lang 37, 50, 90, 91, 96, 97, 168, 236, 237, 261 Larose 129, 139 Latiesa 128, 139 Laviosa 8, 10, 13, 19, 25 Laviosa-Braithwaite 8, 13, 19, 25 Lederer 252, 256, 261 Leech 100, 111 Leeming 115, 117, 118, 123, 124 Lefevere 28, 37 Lehmann 78 Levelt 56, 70 Lewis 33, 38, 53, 70 Liebkind 282, 283 Lieshout 229, 237 Louth 121 Lozar 121 Lundquist 75, 81, 85, 97 M MacKinnon 115, 117, 118, 122, 124
310 Name index
Malblanc 85, 97 Malmkjær ii, iii, iv, v, x, 141, 142, 155 Manner 19, 33, 298 Marello 85, 97 Martin 25, 33, 34, 38, 70, 217, 218, 296, 306 Martinet 89 Mason 69, 126, 139 Matyssek 251, 252, 261 Maupin 79 Mauranen 8, 13 McBurney 299, 304, 306 McLaughlin 121 McQuillan 286, 296 Medgyes 114, 124 Mellman 274, 283 Meta 13, 37, 126, 129, 130, 139, 143, 155, 249 Métrichs 93 Mikkelsen 77, 80, 81 Millán-Varela 224, 226 Mills 286, 296 Milton vi, x, 9, 13, 169, 170, 177, 179, 306 Moeketski 272, 281, 282 Moisl 182, 194 Moon 54, 55, 70 Morgan 200, 207 Morin 225 Morris 12, 264, 269 Morrow 288, 296 Mossop 25 Mulisch 221 Munday 4, 13 Murillo 108, 109, 111 N Nagel 1, 12 Nattinger 52, 54–6, 70 Neubert 285, 296 Nevanlinna 33, 34, 38 Newman 15, 25 Newmark 99, 100, 106, 111, 113, 124 Nida 51, 70 Nielsen 74, 81, 155
Niiniluoto 2, 13 Nobs v, ix, x, 125, 138, 139 Nooteboom 221 Nord 86, 87, 97, 139 Noyes 115, 117, 118, 123 O Offord 166, 168 Oittinen 28, 33, 34, 38 Oksefjell 143, 155 Ondaatje 79 Orwell 53, 56, 57, 70 Øverås 8, 13 Oxenbury 34 P Paikeday 114, 124 Paine 286, 296 Paloposki v, vii, viii, 27, 29, 36, 38 Pápai 60, 70 Pasolini 35 Paternost 115, 117, 118, 122, 123 Pawley 52, 54, 55, 70 Pergnier 95–7 Peters 52, 70, 115, 117, 118, 122, 123 Peters-Roberts 115, 117, 118, 122, 123 Phillipson 114, 124 Pinker 114, 124 Pisek 157, 168 Plastow 301, 306 Pöchhacker ii, 50, 194, 227, 230, 237 Poltermann 28, 38 Popovicˇ 20, 25 Popper 1, 13 Poyatos 111, 227, 228, 237 Priestly 121 Priiki 126, 139 Prüfer 94, 97 Pruncˇ 95, 97 Pym 28, 37, 38, 143, 155, 179 R Rahman 276, 281, 283 Rampton 114, 124 Resch 189, 194
Name index
Riccardi 90, 97 Richards 70, 99, 100, 111, 288, 296 Rickheit 42, 50 Risku vi, x, 181, 182, 185, 189, 194, 195 Robbins 286 Rose ii, 25, 29, 38, 106, 146 Rozan 251, 252, 256, 260, 261 Rudvin vi, xii, xiii, 271, 272, 282, 283
Stolze v, viii, 39, 41, 45, 50 Strauss 302, 304, 306 Suchman 183, 195 Suomalainen 30, 38, 139 Susam-Sarajeva 37, 38 Swan 33, 34 Swift 37 Syder 52, 54, 55, 70
S Saarikoski 35, 38 Sams 217, 218 Santamaria 223, 225 Schäffner 182, 194, 195 Schmid 60, 70 Schmitt 70, 138, 139 Schnell 130 Schoppe 210–2, 216–8 Schreiber v, ix, 83, 85, 87–89, 95, 97 Schreuder 56, 70 Scollon 279, 280, 282, 283 Seleskovitch 240, 249, 251, 252, 256, 261 Setton 91, 97 Sham 286, 295 Shannon 286, 296 Sherwood 286, 295, 296 Shipman 304, 306 Shlesinger ii, xii, 169, 179, 227–230, 237 Sierra Bravo 130 Sinclair 12, 53, 70 Singleton 56, 70 Skutnabb-Kangas 114, 124 Slocum 200, 207 Smith 162, 179, 306 Snell-Hornby ii, vii, xiii, 50–2, 70, 85, 89, 96, 97, 139, 194 Snelling 90, 97 Solberg 292, 296 Somers 200, 207 Sperber 40, 50 St. Jerome 4, 12, 13, 37, 38, 50, 69, 168, 179, 249 Steiner 84, 97, 236
T Taber 51, 70 Taivalkoski 8, 13 Tarp 200, 207 Teich 84, 97 Thiéry 251, 261 Thome 89, 97 Tirkkonen-Condit ii, 8, 13, 51, 71 Titmarsh 146, 155 Tommola 241, 246, 249 Tournay 200 Toury ii, v, vii–viii, 7–10, 13, 15–6, 20, 25, 29, 38, 51, 71, 99, 101–2, 104, 111, 141, 155, 273 Tse 286, 296 Tymoczko 10, 13, 28, 38, 224, 226 Tytler 3, 13 V van Dijk 235 van Doorslaer vi, xi, 219, 221 van Hoof 251, 260, 261 Vasquez 264, 269, 296 Venuti ii, 37, 41, 46, 50, 118, 124, 224–6 Vermeer 143, 155 Vestergaard 79–81 Vienne 186, 195 Vinay 8, 9, 13 W Wadensjö 264, 269, 273, 282, 283, 286, 296 Wagner 216, 292, 296 Wandruszka 85, 97 Weinert 54, 71
311
312
Name index
Weinstein-Shr 292, 296 Weintraub 23, 25 Weltens 56, 69, 70 Weydt 93, 97 White 65, 67, 102, 111, 164 Whitman-Linsen 157, 168 Williams xii, 227, 228, 230, 237 Wilson 40, 50 Wilss ii, 90, 91, 98 Wolf 286
Wolfe 108, 111 Wong-Fillmore 52, 71 Woodsworth ii, 170, 178 Wußler 182, 195 Z Zabalbeascoa v, ix, 99, 100, 111, 157, 168 Zima 223, 226 Zmajic´ 115, 117, 118, 122, 123
Concept index
A A-B interpretation see directionality abbreviation 159, 161–3, 200, 251, 253–4, 256–7, 260 Abtönung 92 accent 241, 243–6, 248 acronym 253, 303 adaptation 9, 29, 30 addition 19, 20, 62, 66, 68, 101–2, 145, 149 adequacy 46 advocacy 263, 268 allegiance 293 anticipation, Antizipation 90–1 apprenticeship 175 Äquivalenz see equivalence artificial intelligence 204–5 asylum seeker 263–7 authentic, authenticity 18, 46–8, 227–8 authority, authorities 203–4, 263, 264–8, 291, 299 automatic 52, 58–9, 62–3, 114, 205, 293 avoidance 147–8 B B-A dominance see directionality basket, basket procedure 202–3 Bible translation 10, 51 binary error 143–4 borrowing 30, 51, 103–4, 106 brokering 285–7, 289, 291–6 C calque 32, 152 career advancement/development 181, 185–6, 192
causality 10 child, childhood 34, 144–5, 149, 282, 285–295 choice and error 141 closeness 29–32 co-author 47 code of ethics 267–8, 274 cognition, cognitive 10–3, 19–20, 39–44, 46, 48, 182, 189, 241, 294, 300, 304 coherence 47–9, 100 cohesion, cohesive, cohesive device 8, 100, 103–4, 233–5, 258 collocation 8, 53–6, 58–60, 62, 64, 113 communication, communicative competence/skills 186, 188, 274, 276–7, 280 community interpreting 271–2, 275, 282 competence, competencies 39, 41, 46, 51, 61, 63–4, 113, 115, 118, 181, 185–193, 231, 232, 264, 274, 276–7, 280, 286 comprehension, comprehensibility 29, 40, 41, 44, 46–7, 62, 142, 189, 204, 214, 240, 246, 300 conditional 21–4 content analysis 198, 200 corpus, corpora 1, 6–7, 9, 23, 53, 87, 100, 145, 165, 202 courtroom, court interpreters 264, 272, 274–9 creative, creativity 49, 52, 58–60, 64, 100, 159, 176, 188, 209, 214–6 critical reading 301 cross-cultural 271, 273–5, 278, 280–1
314 Concept index
cultural brokering 286 cultural difference 271, 279 cultural referent 278
explanatory hypotheses 2, 3, 28 explicitation 5, 7, 8, 10–1, 18, 19, 59, 62
D dead metaphor 56 décalage 88, 91 decision-making 58, 240, 298 deforming tendencies 4–6 delivery, delivery speed 227, 230, 243, 245, 248 descriptive 1–3, 5–7, 20, 23–4, 56, 60, 99–100, 125, 128, 141, 273–5 developing world 272 dialogical 286 différence d’ordre typologique see typological difference direction, directionality 9, 20, 239–247, 271, 278 Disambiguierung 88 discourse field 43–5, 48 discourse modes 273, 276, 280 domesticated, domesticating , domestication 19, 27–9, 31, 32–4, 36, 41 dubbing 157, 160, 162–8 dying metaphors 53, 57
F fair hearing 264 faithful 4, 7, 31, 33–4, 36, 43 false friends 150 feature 2–9, 10, 20, 46–7, 49, 59–61, 99, 100, 102, 104–5, 107, 144, 150, 228–230, 232–3, 235–6, 275, 276, 278–9 fidelity 11, 31, 33–4, 255 film comedies 158, 167 fluency 62, 120, 240 foreign language 53, 113, 170, 172, 188, 193, 221, 239, 246, 252 foreignization , foreignizing 19, 28, 32–6, 41 foreignness 29 formal elaboration 215 formal training 286 further education 191–2
E education 145, 183, 191–2, 266–8, 281–2, 294 Effort Models 259 elaborated translation 62–3, 66–7 electronic basket 202 end users 198, 201 equivalence, equivalent 1, 5–7, 30, 39, 47, 49, 54, 87, 93, 103, 106, 141–4, 161–2, 166–7, 188, 199–201, 266, 273–4 error identification 149 ethical issues, ethics 142, 265–8, 273–4 ethnic affiliation 279 expectations 125–6, 128–131, 137–8 expert interviews 184 explanandum 2
G generalization 4, 11, 17, 19–20, 22, 215 globalisation, Globalisierung 222–3, 271 glosses 32 grado de profesionalidad 126, 136–7 grammatical markers 271, 276 group identity 276, 280–1 H habitus 280–1 hackneyed expressions 57 hermeneutic, hermeneutics 2, 39, 41–4, 48–9 hesitation 229 hierarchy 47, 271–2, 274, 276 holistic 39, 44–8 homonym 153 homophone 160–1 hospital 275 host institution 274
Concept index
hypothesis, hypotheses 1–10, 23, 27–8, 31, 33–4, 36, 85, 166, 273 I Identity, Identität 184, 217, 219, 223–5, 269, 271, 276, 280–1, 292 idiom, idiomatic 5, 30, 45, 49, 53–7, 64, 89, 93–4, 103–4, 107–8, 241, 246 immigration 263–4, 266, 272–3, 280–2 implicitation 18–9 in-house translator 170–1 indirectness 271, 277 inequality 272 inexplicability 149, 150 instant translation 62, 66–7 institution, institutional 271–4, 280–1, 219–220 institutional multilingualism 292, 297–8 integration 27, 43, 48, 273, 281 interference, Interferenz 8, 10, 25, 48, 59, 63, 85, 143, 229 International Technical Writing 181 International Usability Testing 181, 191 Internationalisation 223 interpreter latitude 264 interpreter’s role 263–4, 273 interpreter 174, 177, 190, 192–3 interpreting ethics 265 interpreting metaphor 275 intonation 107, 227–9, 230–6, 243, 245, 248, 278 introspection 240 J jargon 51, 53, 57–9, 64 K kinship 274–6 knowledge base 39, 41–2, 44 knowledge model 203–6 Korpora see corpus
L language acquisition 51–6, 58 language combination 239, 246, 248, 252 language direction 239, 241 language specificity 247 language-play 157–160, 163–8 language-sufficiency 281 law, legal 43, 45, 52, 58, 169, 197, 200, 204, 263–5, 271–3, 281–2 law (in research) 8, 10, 20–2, 24, 85 legal interpreting 273, 282 lexical variation 5, 32 lexicon 29, 30, 53, 56 liaison interpreters 263 linguistische Übersetzungswissenschaft 86 literal, literality, literalness 4, 30–1, 49, 60, 63, 103, 106, 108–9, 159 literary translation 1, 4–6, 9, 29, 51, 58, 167 localization 175, 185, 191 Lösungsorientiertheit 87 M machine translation 197–200, 204–6 Machtkampf 220 Manipulation, manipulative 20, 142–5, 147, 149, 157, 160, 162, 165 Mehrsprachigkeit see multilingualism memorized sequences 52, 54–6 mental representation 39, 42 misreading 152, 154 Modalisierung 87, 92 monological 286 mother language, mother tongue 11, 60, 113–120, 188, 193, 239, 240, 242–7 motivated choice 142–3, 149 multilingual, multilingualism 198, 220 N native language/speaker/tongue see mother language naturalizations 27
315
316 Concept index
naturalness 6, 8, 11, 108, 118 negotiate, negotiation 40–1, 61, 184, 273, 279, 286 non-automatic 62 non-cooperation 278 non-explicitation 18 non-native speaker 114, 116–7, 120 non-professional translation 285 non-translations 1, 6–8, 10 nonverbal 227–8 norm 9, 11, 17, 19, 20, 23–4, 34, 64, 83–4, 94, 102, 110, 141–3, 149, 272, 277–8, 281, 297–8, 300, 304 normalization 10 notes 190, 251–260 O obligatory shifts 63 omission, omit 30, 59, 101, 105, 144–9, 167, 235 online translation 171 overgeneralization, overgeneralization 4, 5, 17, 255, 304 overstress 235 P P-universals 7 paralinguistic 228–9 paraverbal 230 partial theory 101 patterning 5, 8, 141, 144, 147–8 pauses 91, 102, 109, 227–9, 230–2, 234 peer review 297–9, 301, 304 phraseology 35, 47, 53, 56, 191 pitch 229–232, 234–5 planning 46, 229 poiesis 48 police 197–206, 281–2 politeness 271, 276–8 post-editor 59 postmodern 33–4 power 22–3, 56–7, 100, 219, 224, 240, 265, 271–4, 280–2, 285, 291–3, 298 pragmatic, pragmatics 54–5, 58–9, 114
prediction, predictive 2–4, 6–7, 16, 22, 105 prescriptive claims 3, 6 presence 39, 40, 44, 47–9 presentación visual 126–7, 129, 132–5 pressure 11, 263, 266–8, 279 priming effect 55 probabilistic 15, 21–4 problem-solving 39, 46, 165, 189, 229 process-oriented 43 processing capacity 59, 259, 260 processing effort 40, 59, 60 professional ethics 265, 266, 268 professional identity 184, 188 professional qualification 176 professional translation 10, 169, 189, 285 professional translator 49, 169, 171–3, 181, 186, 293–4 project management 182, 185, 191, 193 propositional accuracy 241 prosody 47, 227–230, 232, 235–7 prototype theory 51–2 proxemics 227 public service interpreting 271–4, 282 pure messenger 286 Q quality assessment 47, 61, 125 R receptivity 41–2 recurring phenomena 16 refereeing 297, 299–301 reference corpora 6, 7 refugee 263–9, 273, 282 register 58–60, 62–4, 267, 293 regularities 10, 15–8, 20, 22–4, 110 reinterpretation 28, 159, 161–3 relevance 23, 40–1, 51, 56, 59, 198, 300 representativeness 9, 303 research paradigms 300 retour 28, 239, 240, 243–7
Concept index
retranslation hypothesis 8, 10, 27–9, 31–7 role 264–5, 268 routine, routinization 49, 51, 54, 57–9, 61, 63–4 S S-universals 7, 8, 10, 11 sample 9, 303 Satzaufspaltung 92 scientific norms 298 scope 4, 9 screen translation 157–8, 160, 167 search engine 204 secretarial tasks/work 185, 192 Segmentierung 87–8, 92 self-management 185, 191 self-positioning 293 semantic field 100, 109, 142–3, 271, 275, 278 semantic web 47–8 Serialisierung 87, 90 shift, shifts 8, 18–9, 20–1, 24, 59, 62–3, 257–9 simultaneous interpreting 83, 227, 239, 248–9 skopos 9, 11, 143 social age 292–3 social practices 280–1 social role 271, 276 software developers 197, 199–202, 205 source analysis 185, 187–9 source material 168, 184, 187–9, 193 source-text orientation 36 special language translation 52 speculative approach 99 speech rate 229, 232, 235 spontaneous speech 228–9 sprachenpaarbedingte Probleme 84 strategy, strategies 10, 33–4, 42, 90–2, 102, 106, 119, 144, 147, 157–9, 160–7, 185, 189, 254, 260, 268, 276, 294–5 argumentation strategy 184 automatic strategy 59, 63
cognitive strategy 40 Dolmetschstrategie 90, 92 domesticating strategy 33, 36 foreignizing strategy 36 interpreting strategy see Dolmetschstrategie 264 modelling strategy 206 notation strategy 255 note-taking strategy 254, 259, 260 omitting and summarizing strategy 235 overall strategy 185, 189 problem-solving strategy 229 social repair strategy 293 source-language strategy 257–9 summarizing strategy 235–6 target-language strategy 257–9 translation strategy 33, 106, 119 stress 59, 61, 236, 274, 279 stress pattern 227–8, 230, 235–6, 279 structure hiérarchique 78 structure linéaire 78 structuring markers 230, 233, 235–6 stylistique comparée 85 substitution 147–8 subtitling 9, 157, 165, 167–8, 171 summarizing 235–6 suprasegmental 227–8 symbols 190, 251–6, 259, 260 T T-universals 7, 8, 10–1 target analysis 185–7, 192–3 tautologies 19 TC 181–193 technical communication/ communicator/communicators 58, 181–3, 185–9, 191–5 technical texts/translation/writing 52, 58–9, 61, 63–4, 181–2, 186, 190 tensions 219, 292–3 terminology 10, 24, 45, 47, 57, 173, 184–5, 187, 190–3, 197–9, 200–6, 273
317
318
Concept index
text analysis 42, 44, 182–3 text function 47–9, 104 text linguistics 42 text production 39, 40, 46–7, 49, 182, 189, 194 textual fit 1, 6, 7 Thema-Rhema-Gliederung 87–90 time pressure 21, 59, 60–4 traditional cultures 275 training 185–7, 190, 274, 281, 286, 291, 299 interpreter training 140, 229, 239, 240–2, 246–7, 253–4, 260, 266, 272–4, 286 translator training 11, 51, 61, 142, 172–4 user training 187 training in research 304 translation competence 181 translation criticism 1, 142 translation curriculae 193 translation evaluation 61 translation tools 59, 62, 176, 190, 198 translationese 51, 59–61, 64 translator motivation 149 transpositions 62, 73, 75 truth-value 275 TT-oriented 142 turn-taking 271, 278–9 type-within-type structure 110 typological difference 74, 77 U Übersetzungsprobleme 85–7
unit 55, 101–2, 104, 107 composite unit 55 empty unit 104 lexical unit 108 linguistic unit 102 multi-word unit 52–6 note unit 253–5, 258 orthographic unit 54 phraseological unit 52, 56 pragmatic unit 55 ready-made unit 53–9, 60–4 semantic unit 54–5 semiotic unit 107 source-text unit 258 textual unit 102 tone unit 231 unit of information 230, 233–5 unit of language 64 universality 10, 18, 20, 21, 24 university translation 169, 171–3, 177 unmarkedness 29 usability 181–2, 184, 189, 191–3 user groups 197, 199, 201–2, 204 V verbal false limbs 53, 56–7 visual presentation see presentación visual W wordplay 157–8, 160, 162, 164, 166–8 Wortstellung 84, 90 writing skills 190, 304
In the series Benjamins Translation Library the following titles have been published thus far or are scheduled for publication: 4 5
6 7 8 9 10 11 12 13 14 15 16 17 18 19
20 21 22 23
24 25 26 27
28 29
TOURY, Gideon: Descriptive Translation Studies – and beyond. 1995. viii, 312 pp. DOLLERUP, Cay and Annette LINDEGAARD (eds.): Teaching Translation and Interpreting 2. Insights, aims and visions. Papers from the Second Language International Conference Elsinore, 1993. 1994. viii, 358 pp. EDWARDS, Alicia B.: The Practice of Court Interpreting. 1995. xiii, 192 pp. BEAUGRANDE, Robert de, Abdullah SHUNNAQ and Mohamed Helmy HELIEL (eds.): Language, Discourse and Translation in the West and Middle East. 1994. xii, 256 pp. GILE, Daniel: Basic Concepts and Models for Interpreter and Translator Training. 1995. xvi, 278 pp. REY, Alain: Essays on Terminology. Translated by Juan C. Sager. With an introduction by Bruno de Bessé. 1995. xiv, 223 pp. KUSSMAUL, Paul: Training the Translator. 1995. x, 178 pp. VINAY, Jean-Paul and Jean DARBELNET: Comparative Stylistics of French and English. A methodology for translation. Translated and edited by Juan C. Sager, M.-J. Hamel. 1995. xx, 359 pp. BERGENHOLTZ, Henning and Sven TARP (eds.): Manual of Specialised Lexicography. The preparation of specialised dictionaries. 1995. 256 pp. DELISLE, Jean and Judith WOODSWORTH (eds.): Translators through History. 1995. xvi, 346 pp. MELBY, Alan K. and Terry WARNER: The Possibility of Language. A discussion of the nature of language, with implications for human and machine translation. 1995. xxvi, 276 pp. WILSS, Wolfram: Knowledge and Skills in Translator Behavior. 1996. xiii, 259 pp. DOLLERUP, Cay and Vibeke APPEL (eds.): Teaching Translation and Interpreting 3. New Horizons. Papers from the Third Language International Conference, Elsinore, Denmark, 1995. 1996. viii, 338 pp. POYATOS, Fernando (ed.): Nonverbal Communication and Translation. New perspectives and challenges in literature, interpretation and the media. 1997. xii, 361 pp. SOMERS, Harold (ed.): Terminology, LSP and Translation. Studies in language engineering in honour of Juan C. Sager. 1996. xii, 250 pp. CARR, Silvana E., Roda P. ROBERTS, Aideen DUFOUR and Dini STEYN (eds.): The Critical Link: Interpreters in the Community. Papers from the 1st international conference on interpreting in legal, health and social service settings, Geneva Park, Canada, 1–4 June 1995. 1997. viii, 322 pp. SNELL-HORNBY, Mary, Zuzana JETTMAROVÁ and Klaus KAINDL (eds.): Translation as Intercultural Communication. Selected papers from the EST Congress, Prague 1995. 1997. x, 354 pp. BUSH, Peter and Kirsten MALMKJÆR (eds.): Rimbaud's Rainbow. Literary translation in higher education. 1998. x, 200 pp. CHESTERMAN, Andrew: Memes of Translation. The spread of ideas in translation theory. 1997. vii, 219 pp. GAMBIER, Yves, Daniel GILE and Christopher J. TAYLOR (eds.): Conference Interpreting: Current Trends in Research. Proceedings of the International Conference on Interpreting: What do we know and how? 1997. iv, 246 pp. ORERO, Pilar and Juan C. SAGER (eds.): The Translator's Dialogue. Giovanni Pontiero. 1997. xiv, 252 pp. POLLARD, David E. (ed.): Translation and Creation. Readings of Western Literature in Early Modern China, 1840–1918. 1998. vi, 336 pp. TROSBORG, Anna (ed.): Text Typology and Translation. 1997. xvi, 342 pp. BEYLARD-OZEROFF, Ann, Jana KRÁLOVÁ and Barbara MOSER-MERCER (eds.): Translators' Strategies and Creativity. Selected Papers from the 9th International Conference on Translation and Interpreting, Prague, September 1995. In honor of Jiří Levý and Anton Popovič. 1998. xiv, 230 pp. SETTON, Robin: Simultaneous Interpretation. A cognitive-pragmatic analysis. 1999. xvi, 397 pp. WILSS, Wolfram: Translation and Interpreting in the 20th Century. Focus on German. 1999. xiii, 256 pp.
30 DOLLERUP, Cay: Tales and Translation. The Grimm Tales from Pan-Germanic narratives to shared international fairytales. 1999. xiv, 384 pp. 31 ROBERTS, Roda P., Silvana E. CARR, Diana ABRAHAM and Aideen DUFOUR (eds.): The Critical Link 2: Interpreters in the Community. Selected papers from the Second International Conference on Interpreting in legal, health and social service settings, Vancouver, BC, Canada, 19–23 May 1998. 2000. vii, 316 pp. 32 BEEBY, Allison, Doris ENSINGER and Marisa PRESAS (eds.): Investigating Translation. Selected papers from the 4th International Congress on Translation, Barcelona, 1998. 2000. xiv, 296 pp. 33 GILE, Daniel, Helle V. DAM, Friedel DUBSLAFF, Bodil Ringe MARTINSEN and Anne SCHJOLDAGER (eds.): Getting Started in Interpreting Research. Methodological reflections, personal accounts and advice for beginners. 2001. xiv, 255 pp. 34 GAMBIER, Yves and Henrik GOTTLIEB (eds.): (Multi) Media Translation. Concepts, practices, and research. 2001. xx, 300 pp. 35 SOMERS, Harold (ed.): Computers and Translation. A translator's guide. 2003. xvi, 351 pp. 36 SCHMID, Monika S.: Translating the Elusive. Marked word order and subjectivity in English-German translation. 1999. xii, 174 pp. 37 TIRKKONEN-CONDIT, Sonja and Riitta JÄÄSKELÄINEN (eds.): Tapping and Mapping the Processes of Translation and Interpreting. Outlooks on empirical research. 2000. x, 176 pp. 38 SCHÄFFNER, Christina and Beverly ADAB (eds.): Developing Translation Competence. 2000. xvi, 244 pp. 39 CHESTERMAN, Andrew, Natividad GALLARDO SAN SALVADOR and Yves GAMBIER (eds.): Translation in Context. Selected papers from the EST Congress, Granada 1998. 2000. x, 393 pp. 40 ENGLUND DIMITROVA, Birgitta and Kenneth HYLTENSTAM (eds.): Language Processing and Simultaneous Interpreting. Interdisciplinary perspectives. 2000. xvi, 164 pp. 41 NIDA, Eugene A.: Contexts in Translating. 2002. x, 127 pp. 42 HUNG, Eva (ed.): Teaching Translation and Interpreting 4. Building bridges. 2002. xii, 243 pp. 43 GARZONE, Giuliana and Maurizio VIEZZI (eds.): Interpreting in the 21st Century. Challenges and opportunities. 2002. x, 337 pp. 44 SINGERMAN, Robert: Jewish Translation History. A bibliography of bibliographies and studies. With an introductory essay by Gideon Toury. 2002. xxxvi, 420 pp. 45 ALVES, Fabio (ed.): Triangulating Translation. Perspectives in process oriented research. 2003. x, 165 pp. 46 BRUNETTE, Louise, Georges BASTIN, Isabelle HEMLIN and Heather CLARKE (eds.): The Critical Link 3. Interpreters in the Community. Selected papers from the Third International Conference on Interpreting in Legal, Health and Social Service Settings, Montréal, Quebec, Canada 22–26 May 2001. 2003. xii, 359 pp. 47 SAWYER, David B.: Fundamental Aspects of Interpreter Education. Curriculum and Assessment. 2004. xviii, 312 pp. 48 MAURANEN, Anna and Pekka KUJAMÄKI (eds.): Translation Universals. Do they exist? 2004. vi, 224 pp. 49 PYM, Anthony: The Moving Text. Localization, translation, and distribution. 2004. xviii, 223 pp. 50 HANSEN, Gyde, Kirsten MALMKJÆR and Daniel GILE (eds.): Claims, Changes and Challenges in Translation Studies. Selected contributions from the EST Congress, Copenhagen 2001. 2004. xii, 306 pp. + index. EST Subseries Volume 1 51 CHAN, Leo Tak-hung: Twentieth-Century Chinese Translation Theory. Modes, issues and debates. 2004. xvi, 277 pp. 52 HALE, Sandra Beatriz: The Discourse of Court Interpreting. Discourse practices of the law, the witness and the interpreter. xvi, 261 pp. + index. Expected Summer 2004 53 DIRIKER, Ebru: De-/Re-Contextualizing Conference Interpreting. Interpreters in the Ivory Tower? vii, 209 pp. + index. Expected Summer 2004 54 GONZÁLEZ DAVIES, Maria: Multiple Voices in the Translation Classroom. Activities, tasks and projects. xii, 259 pp + index. Expected Summer 2004 A complete list of titles in this series can be found on www.benjamins.com/jbp
