<DOCINFO AUTHOR ""TITLE "Claims, Changes and Challenges in Translation Studies: Selected contributions from the EST Congress,

Copenhagen 2001"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 Associate editor

Gideon Toury 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 UNISA University of Joensuu
ISIT Paris 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
 TM The paper used in this publication meets the minimum requirements
8

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 2004041064
isbn 90 272 1656 8 (Eur.) / 1 58811 509 7 (US) (Hb; alk. paper)

© 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
<TARGET
< R/ RE E F"toc"
F DOCINFO AUTHOR ""TITLE "Table of contents"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150" VOFFSET "4">

"intro">
"che">
"tou">
"pal">
"sto">
"hel">
"kor">
"sch1">
"zab">
"pok">
"nob">
"mal">

Table of contents

Introduction vii
Hypotheses about translation universals 1
Andrew Chesterman
Probabilistic explanations in Translation Studies: Universals — or a
challenge to the very concept? 15
Gideon Toury
A thousand and one translations: Revisiting retranslation 27
Outi Paloposki and Kaisa Koskinen
Creating “presence” in translation 39
Radegundis Stolze
Ready-made language and translation 51
Pál Heltai
Les attributs indirects en français et en danois: Différences typologiques
et problèmes de traduction 73
Hanne Korzen
Kontrastive Linguistik und sprachenpaarbezogene
Translationswissenschaft 83
Michael Schreiber
Translating non-segmental features of textual communication:
The case of metaphor within a binary-branch analysis 99
Patrick Zabalbeascoa
Challenging the myth of native speaker competence in translation
theory: The results of a questionnaire 113
Nike K. Pokorn
Expectativas y evaluación en la traducción de folletos turísticos 125
Marie-Louise Nobs
Censorship or error: Mary Howitt and a problem in descriptive TS 141
Kirsten Malmkjær
< /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 157
Thorsten Schröter
The figure of the factory translator: University and professional domains
in the translation profession 169
John Milton
Migrating from translation to technical communication and usability 181
Hanna Risku
From raw data to knowledge representation: Methodologies for
user-interactive acquisition and processing of multilingual terminology 197
Barbara Dragsted and Benjamin Kjeldsen
The translator as a creative genius: Robert Schumann 209
João Azenha Junior
Übersetzung zwischen Nationalismus und Internationalismus 219
Luc van Doorslaer
Non-verbal phenomena in simultaneous interpreting:
Causes and functions 227
Barbara Ahrens
Simultaneous interpreting A-B vs. B-A from the interpreters’ standpoint 239
Magdalena Bartlomiejczyk
Interpreters’ notes: On the choice of form and language 251
Helle V. Dam
Expressing a well-founded fear: Interpreting in convention refugee
hearings 263
Sabine Fenton
Cross-cultural dynamics in community interpreting. Troubleshooting 271
Mette Rudvin
The child in the middle: Agency and diplomacy in language brokering
events 285
Nigel Hall
The editorial process through the looking glass 297
Daniel Gile and Gyde Hansen

Name index 307

Concept index 313
<TARGET "intro" DOCINFO AUTHOR ""TITLE "Introduction"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 process-
ing. 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 prescrip-
tive 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 engi-
neering, 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 transla-
tion and retranslation of foreign literature into Finnish and find some counter-
evidence. 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 ready-
made 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 ix

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, transla-
tion difficulties can be explained by morpho-syntactic differences and the compres-
sion 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 “sprachen-
paarbezogene Translationswissenschaft”, literally “language-pair-specific transla-
tion science”, which is positioned at the interface between contrastive linguistics
and general Translation Studies. He stresses that in spite of the increasing impor-
tance 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 transla-
tion-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 questionnaire-
based pilot study dealing with expectations and assessment of German translations
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 choice-
based 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 transla-
tion 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, standard-
setting, etc.).
Translators do not always remain translators. They sometimes migrate into
different activities. Professional migration from translation to technical communi-
cation 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 xi

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 intelli-
gence. 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 institu-
tions 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 transla-
tions 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
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 profession-
als 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
</TARGET "intro">

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 non-
English 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 enrol-
ment 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.
<LINK "che-n*">

<TARGET "che" DOCINFO AUTHOR "Andrew Chesterman"TITLE "Hypotheses about translation universals"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 rela-
tion), and (b) the relation between translations and comparable non-transla-
tions 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 descrip-
tive research in its historical context and then consider some of the termino-
logical, 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 non-
edible 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 some-
thing 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 subsequent-
ly 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 3

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 hypothe-
ses 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)
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 conse-
quence 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 5

– 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 equiva-
lent 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 transla-
tion, particularly the translations of his own works (Kundera 1993: 123f.). He
complains of the way translators violate metaphors, seek to enrich simple vocabu-
lary, 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
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 7

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.
8 Andrew Chesterman

Potential S-universals
– Lengthening: translations tend to be longer than their source texts (cf. Ber-
man’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 9

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 descrip-
tive 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 transla-
tions 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
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 transla-
tion 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 unfortunate-
ly 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 explici-
tation hypothesis (Blum-Kulka and others) or the simplification hypothesis
(Laviosa), or the retranslation hypothesis. Others speak of laws: cf. Toury’s pro-
posed 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. standard-
ization, 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 non-
translations. 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 11

such as communicating in a non-native language or under special channel restric-

tions, or any form of communication that involves relaying messages, such as
reporting discourse, even journalism. It may be problematic, eventually, to differen-
tiate 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 non-
nativeness; 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.
<DEST "che-n*">

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’horlo-
gerie 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.
</TARGET "che">

Hypotheses about translation universals 13

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 transla-
tion”. 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 Intercul-
tural 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.
<LINK "tou-n*">

<TARGET "tou" DOCINFO AUTHOR "Gideon Toury"TITLE "Probabilistic explanations in Translation Studies"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 as-

pects of the quest for universals of translation: (1) the level where transla-
tion 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 probabi-
listic and conditioned rather than deterministic, and that they should be
found and formulated by means of a combination of “top-down” and “bot-
tom-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 transla-
tor 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 17

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 realiza-
tions 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).
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 counter-
evidence 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 so-
called 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 19

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 occur-
rence 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 “transla-
tion 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 observ-
ables. 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 transla-
tion 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,
20 Gideon Toury

addition/loss of information, generalization/concretization, etc. Consequently,

it is possible to formulate the following rule: If a shift occurs, it necessari-
ly 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. Popovič 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 non-
realizability 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, socio-
cultural (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 21

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)
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, socio-
cultural 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 determin-
ing 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) expecta-
tion. 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 23

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 deter-
mined, 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 transla-
tion 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 ever-
growing 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 proce-
dures 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
<DEST "tou-n*">

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
</TARGET "tou">

Probabilistic explanations in Translation Studies 25

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 responsibili-
ty). 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 Manipula-
tion 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 Transla-
tion Perspectives III: Selected Papers, 1985–86, M. Gaddis Rose (ed), 69–83. Binghamton: State
University of New York at Binghamton.
Popovič, 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. Popovič (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 introduc-
tion”. 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 Traduc-
ción, M.Á. Vega i R. Martín-Gaitero (eds), 69–80. Madrid: Editorial Complutense.
Weintraub, Ruth. 1988. “A paradox of confirmation”. Erkenntnis 29. 169–180.
<TARGET "pal" DOCINFO AUTHOR "Outi Paloposki and Kaisa Koskinen"TITLE "A thousand and one translations"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 hy-
pothesis 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 pro-
files 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 trans-
lations 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 compre-
hensible 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, retransla-
tions 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 accom-
modate 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-retrans-
lation 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 29

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 consid-
ered 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.
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 — Suoma-
lainen has simply omitted the whole passage, whether out of concern for under-
standability 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 31

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 transla-
tion. Idiosyncratic constraints — the translator’s own preferences, or even difficul-
ties 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 manu-
script 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 transla-
tors, 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
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 availabili-
ty, 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 rope-
maker 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 introduc-
ing 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 undoubt-
edly 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 33

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 contem-
porary 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 domesti-
cation, 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
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 motivat-
ed 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 well-
known 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 35

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 transla-
tion 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 transla-
tion 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).
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 hypothe-
sis: 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 accom-
plishment 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 Kaisa Koskinen
Research Institute for the Languages School of Modern Languages and
of Finland Translation Studies
Sörnäisten rantatie 25 University of Tampere
00500 Helsinki 33014 Tampere
Finland Finland
e-mail: outi.paloposki@kotus.fi e-mail: kaisa.a.koskinen@uta.fi
 A thousand and one translations 37

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 chronological-
ly 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.
</TARGET "pal">

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, Şebnem. 2001. “Is one case always enough?” Perspectives 9 (3): 167–176.
Susam-Sarajeva, Şebnem. 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.
<TARGET "sto" DOCINFO AUTHOR "Radegundis Stolze"TITLE "Creating “presence” in translation"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

Creating “presence” in translation

Radegundis Stolze
Darmstadt

This article aims to show how translation theory may be based on herme-
neutic language philosophy. It discusses the problem of translation — a
dynamic task from the translators’ perspective. Translation means present-
ing a message from a given text as understood by the translator, to other
readers. The theoretical view switches from the analysis of source- and tar-
get-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 cogni-
tive activity in understanding written texts, which also requires critical re-
flection of one’s own knowledge base. Subsequent text production in trans-
lation is seen as a problem-solving process coordinating various dimensions.
Aspects of holistic text understanding based on linguistics and the applica-
tion 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 orienta-

tion 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 transla-

tion 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 commu-
nication 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. Tempo-
rary 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 41

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 motiva-
tion 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
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 motiva-
tion, 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 compre-
hending, 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 43

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 ethno-
centric 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 dia-
cultures within one cultural system. There are special groups of scientists, house-
wives, 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 dia-
cultures. 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.
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 non-
specialists 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 authenti-
cally 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 “position-
ing” 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 view-
points 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 45

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 predica-
tive 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 Specialist texts

Culture People, country Area of research and work (science,

humanities)
Discourse field Social background & field Discipline & domain, level of commu-
of communication nication
Concepts Cultural associations, key Terminological conceptualization
words, author’s ideology (definition/deduction vs. convention/
interpretation)
Predicative mode Idioms, mood, focusing, Speech acts, passive voice
quotations
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 trans-
lators 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 47

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 Text types, macrostructure,

perspective, tense, genre
Coherence Semantic web, titles,
compatibility, structural axes,
intertextuality
Style Expressive means, rhyme,
metaphors, alliteration
Shape Rhythm, prosody, text form,
pictures, verse order
addressees’ expectation
Equivalence of terminology,
specification of concepts, word
compounding
Functional style, standard
formulae, controlled language,
phraseology
Illustrations, layout, script fonts

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.
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 49

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”.
</TARGET "sto">

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 Herme-
neutik. 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 process-
es.” 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.
<TARGET "hel" DOCINFO AUTHOR "Pál Heltai"TITLE "Ready-made language and translation"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

Ready-made language and translation

Pál Heltai
Institute of Foreign Languages, Szent István University, Gödöllő

Translation studies remains an interdisciplinary field, making good use of

new insights in other linguistic disciplines. The present paper gives an over-
view 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 ready-
made 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 lan-
guage effectively and to evaluate ready-made language from the socio-
linguistic point of view. The paper also reports on a pilot experiment dem-
onstrating 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 transla-
tion. 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 proto-
typical 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 convention-
alized 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 53

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, routin-
ized language, prefabricated language, as well as multi-word units, phraseological
54 Pál Heltai

units, ready-made units, routine formulae, memorized formulae, memorised sequences,

prefabricated elements and prefabs are sometimes used synonymously. Unfortunate-
ly, 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 memo-
rized 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 expres-
sions recalled as wholes from memory may be put at several hundred thousand.
Nattinger and DeCarrico (1992) claim that “we are creatures of habit”, and “routin-
ized 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 55

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.
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 communi-
cation. 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 ready-

made 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 com-
plaints 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 57

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 metaphori-
cal, idiomatic and emotionally loaded words: familiarity breeds contempt. Hack-
neyed expressions abound in colloquial language, newspapers and political lan-
guage, 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 differenc-
es compared to stylistically neutral ready-made units. The difference is not structur-
al, 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 expecta-
tion 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
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, scientif-
ic, legal and bureaucratic language, on the other hand, call for a high degree of
routinization, although, of course, different types of technical texts (and conse-
quently 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 59

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, especial-
ly 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 socio-
linguistic 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 transla-
tion shifts, especially explicitation (discussed in Klaudy 1999), while direct interfer-
ence is manifested first of all in collocations and other ready-made units. Odd,
foreign-sounding collocations, however, seldom disrupt meaning, since speakers
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 register-
specific 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 61

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. Transla-
tor 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 Gödöllő 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
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 transla-
tions 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 transposi-
tions, 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 particu-
lar 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 tényezők (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 auto-
matic 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 63

(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 experi-
enced 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 conclu-
sions, 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 ready-
made 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
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 ready-
made 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 ready-
made language. Inappropriate use or non-use of ready-made phrases might con-
tribute to the impression that the translated text is “jargon” or “translationese”.
– Translationese can be tolerated to some extent depending on the circumstanc-
es, 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 ready-
made 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 Gödöllő, Páter K. u. 1.
heltai@gtk.gau.hu

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-to-
say./ 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-conclusion-
is-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/ the-
whole-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 65

the-shelf/ ready-made/ because they-say-what-you-want-to-say./ /I-don’t-think-I’m-going-

out-on-a-limb-here./ However, /it-is-appropriate-to-say-at-this-point/ that-much-work-
remains-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 environ-
ments, 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.
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 tényezők a következők: a klíma, a talaj milyensége, a domborzati
viszonyok és a biotikus tényezők. Minden fafajta alkalmazkodik ezekhez a tényezőkhöz
egyfajta integrációs módon: kifejleszti a specifikus szubpopulációit, amelyek alkalmazkodtak
az adott környezethez. Mint már az előzőekben tárgyaltuk, ilyen szempontból a
legjelentősebb környezeti tényező a hőmérséklet csökkenése, amit okozhat a nagy tengerszint
feletti magasság, illetve a sarkokhoz való közeledés. A szubpopulációk úgy alkalmazkodnak,
hogy módosítják azt az optimum hőmérsékleti tartományt, amelyben a fotoszintézis lényeges
folyamatai zajlanak.

Elaborated translation
A fákra ható környezeti tényezők a következők: éghajlat, talaj, domborzat és a biotikus
tényezők (flóra és fauna). Minden fafaj egyfajta komplex módon alkalmazkodik ezekhez a
tényezőkhöz, ami jelen esetben azt jelenti, hogy a fajokból specifikus alfajok fejlődnek ki,
amelyek képesek alkalmazkodni a sajátságos környezetükhöz. Mint ahogy azt már tárgyaltuk,
a legfontosabb tényező a hőmérséklet-csökkenés, ami vagy a növekvő tengerszint feletti
magassággal, vagy a sarkokhoz való közeledéssel párosul. Minden egyes alfaj alkalmazkodik
ehhez a környezeti tényezőhöz azzal, hogy módosul az optimum hőmérséklete, amelyen
minden lényeges fotoszintetikus folyamat zajlik.

Back-translation of instant translation

The environmental factors affecting trees are the following: climate, the quality of soil, topo-
graphic 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 67

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 tényezők a következők: a klíma, a talaj milyensége, a domborzati
viszonyok és a biotikus tényezők. Minden fafajta alkalmazkodik ezekhez a tényezőkhöz
egyfajta integrációs módon: kifejleszti a specifikus szubpopulációit, amelyek alkalmazkodtak
az adott környezethez. Mint már az előzőekben tárgyaltuk, ilyen szempontból a
legjelentősebb környezeti tényező a hőmérséklet csökkenése, amit okozhat a nagy tengerszint
feletti magasság, illetve a sarkokhoz való közeledés. A szubpopulációk úgy alkalmazkodnak,
hogy módosítják azt az optimum hőmérsékleti tartományt, amelyben a fotoszintézis lényeges
folyamatai zajlanak.
Számos fafajta, amely kedvezőtlen élőhelyen is jól megél, a kedvező élőhelyen ha nincs
verseny az élőhelyért, akkor nagyon jól fejlődik. Az ilyen fák nem tolerálják az élőhelyért való
versengést, de jól tűrik a szélsőséges környezeti viszonyokat. Például a feketefenyő (Picea
mariana) megtalálható lápokban és hegytetőkön is az Egyesült Államok északkeleti részén, de
nem tudnak más fákkal, például a? — fenyővel (Picea rubens) versenyezni jobb élőhelyeken.
Ezért a Picea rubens Amerika északkeleti részén, a New Hampshire-i White Mountains-ban
a hegyek lábánál él, a fekete fenyő viszont a hegytetőn, illetve hibridizációval keletkezett
szubpopulációi előfordulnak még közepes magasságokban is.
A fajok közötti versengés (néhány esetben beszélhetünk törzsek közötti versengésről) a
legintenzívebb, mert az egyedek ugyanazokért a környezeti forrásokért versengenek. Mivel a
fák képtelenek megkeresni a szükséges erőforrásaikat, ezért a lényeges harc a helyért és a
felhasználható erőforrásokért folyik, míg a föld alatti versengést a vízért, a helyért, a
tápanyagokért és a szimbiontákért (mikroorganizmusokért, például mikorrhizákért és
nitrogénkötő baktériumokért) való küzdelem jelenti.

Elaborated translation
A fákra ható környezeti tényezők a következők: éghajlat, talaj, domborzat és a biotikus
tényezők (flóra és fauna). Minden fafaj egyfajta komplex módon alkalmazkodik ezekhez a
tényezőkhöz, ami jelen esetben azt jelenti, hogy a fajokból specifikus alfajok fejlődnek ki,
amelyek képesek alkalmazkodni a sajátságos környezetükhöz. Mint ahogy azt már tárgyaltuk,
a legfontosabb tényező a hőmérséklet-csökkenés, ami vagy a növekvő tengerszint feletti
magassággal, vagy a sarkokhoz való közeledéssel párosul. Minden egyes alfaj alkalmazkodik
ehhez a környezeti tényezőhöz azzal, hogy módosul az optimum hőmérséklete, amelyen
minden lényeges fotoszintetikus folyamat zajlik.
Sok fafaj, amely kedvezőtlen élőhelyen is megél, valójában kedvezőbb körülmények
között jobban fejlődne, feltéve, hogy nincs konkurencia. Ezek a fák kevéssé képesek tolerálni
a konkurenciát, de jól viselik a szélsőséges körülményeket. Például a fekete luc (Picea
mariana) megtalálható Északkelet-Amerikában mocsaras területeken és hegytetőkön
egyaránt, viszont ez a faj más élőhelyeken nem képes versenyezni más fafajokkal, például a
vörös luccal (Picea rubens). Következésképpen a new hampshire-i White Hegységben, az
68 Pál Heltai

Egyesült Államok északkeleti részén a vörös luc a hegyek lábánál található, miközben a
lejtőkön alfaj populációk kialakulása figyelhető meg (hibridizáció).
A versengés a fajok között, illetve egyes esetekben nemzetségek között a legjelentősebb,
mert az egyedek ugyanazokért a környezeti erőforrásokért versengenek. Mivel a fák nem
képesek az aktív mozgásra, hogy így megkeressék a szükséges erőforrásaikat, ezért nagyon
jelentős lehet a versengés a rendelkezésre álló termőhelyért és a szükséges erőforrásokért. 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-kötők).

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 tényezők a következők

soils a talaj milyensége
topography domborzati viszonyok földrajzi viszonyok
biota biotikus tényezők élő szervezetek
major factor környezeti tényező
elevation tengerszint feletti tengerszint feletti
magasság magasság
temperature hőmérsékleti tartomány hőmérsékleti
tartomány
threshold for élőhelyért folytatott versenykényszer más fajokkal
competition verseny folytatott
verseny
extremes szélsőséges környezeti szélsőséges környezeti szélsőséges
viszonyok viszonyok körülmények
move helyváltoztató mozgás
symbionts szimbionta hasznos szervezetek
társszervezetek
nitrogen fixers nitrogénkötõ
baktériumok
 Ready-made language and translation 69

Least proficient students

Student X Student Y Student Z

SL expression Addition in TT

factors are
soils
topography földrajzi viszonyok
biota
major factor
elevation
temperature
threshold for
competition
extremes szélsőséges viszonyok
move
symbionts
nitrogen fixers nitrogénkötő
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.
70 Pál Heltai

Heltai, P. 1997. “Minimal translation”. In A 2. Transferre Necesse Est konferencia előadásai

(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), Gödöllő:
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 hipotézisről”. (On the Explicitation Hypothesis) Fordítás-
tudomány 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. Descrip-
tion, 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, Győr — 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 English-
German 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.
</TARGET "hel">

Ready-made language and translation 71

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.
<TARGET "kor" DOCINFO AUTHOR "Hanne Korzen"TITLE "Les attributs indirects en français et en danois"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 rela-
tional 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 prédica-
tion 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 “descrip-
tif”. 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 circonstan-
ciel. 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” circonstan-
ciel, moins il paraîtra acceptable.2 Dans ce qui suit, nous verrons que cette diffé-
rence 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 75

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 fonction-
ner 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
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 construc-
tion est bien plus fréquente en français qu’en danois. Quand l’attribut est descrip-
tif, 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 co-
verbe 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. Hers-
lund 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 77

(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
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 “desententiali-
sation”, 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 “déverbal-
isation”, cf. (17)):
(16) 0. verbe fini dans une principale
1. verbe fini dans une subordonnée
2. verbe non fini
3. nominalisation
(17) 0. verbe prototypique
1. perte d’autonomie illocutoire, modale, temporelle et aspectuelle
2. perte de désinences verbales et de sujet explicite
3. 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 79

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 circons-
tanciel 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 fré-
quents 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.
80 Hanne Korzen

Author’s Address:
Hanne Korzen
Dalgas Have 15
DK-2000 Frederiksberg
e-mail: hk.fra@cbs.dk

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.
</TARGET "kor">

Les attributs indirects en français et en danois 81

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 franco-
danoise 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.
<TARGET "sch1" DOCINFO AUTHOR "Michael Schreiber"TITLE "Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 im-
portant 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 transla-
tion practice when it goes beyond the comparison of languages (as is done in
contrastive analysis). While the object of contrastive analysis is the compari-
son 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 Translationswissen-
schaft (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 Forschungs-

schwerpunkte 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 Lösungs-
möglichkeiten 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 Hinter-
grund 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 Wortstellungs-
varianten in den verglichenen Sprachen (ggf. in einzelnen Varietäten oder Textsor-
ten) üblich und welche weniger üblich sind. Zweitens hat auch die sog. pragmati-
sche Wende in der Linguistik zu einer Ausweitung der KL geführt (Järventausta
1996: 98f.). Neben die anfangs dominierende, systemorientierte, kontrastive Gram-
matik ist eine kontrastive Pragmatik getreten, in der u.a. die Realisierung bestimmter
 Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft 85

Sprechakte (z.B. Anweisungen) in verschiedenen Sprachen untersucht wird (Snell-

Hornby 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 bescheide-
nen 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 Übersetzungs-
probleme 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
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 Zuord-
nungsvarianten (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 über-
setzerische 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 Arznei-
mittelgesetzes und im Einklang mit der EU-Gesetzgebung ist eine deutsch-
sprachige Packungsbeilage anzufertigen. (Nord 2001: 121)
 Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft 87

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 des-
kriptiv vorgehen, indem sie empirisch beschreibt, wie Übersetzer und Dolmetscher
mit sprachenpaarbedingten Problemen umgehen (wie wir unter 4. anhand kon-
kreter 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-Rhema-
Gliederung, 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 gespro-
chenen Französischen und Deutschen erwähnt. Gülich untersucht hier anhand
sprechsprachlicher Korpora Vorkommen von Konstruktionen mit redundanter
Besetzung einer syntaktischen Position wie
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 Personalprono-
mina 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 Unter-
schieden zwischen beiden Sprachen wenig hilfreich. Aus dolmetschwissenschaftli-
cher Sicht wird das Phänomen der Segmentierung in einem Aufsatz von Sabine
Bastian behandelt. Hier geht es um die Wiedergabe der französischen Segmentie-
rung sowie der sog. mise en relief beim Dolmetschen ins Deutsche. Bastian führt u.a.
aus, daß die hervorhebende Funktion der pronominalen Redundanz im Französi-
schen 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 Rhematisie-
rung) 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 Deut-
schen weniger gebräuchlich sind, kann es in Übersetzungen angebracht sein, eine
entsprechende Hervorhebung durch lexikalische Mittel auszudrücken, wie Gisela
 Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft 89

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 mani-
festent… 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 Spra-
chenpaar 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 vor-
angestellte 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
90 Michael Schreiber

Beibehaltung der Informationsstruktur nicht mit der Beibehaltung der Wort-

stellung 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 Untersuchungsgegen-
stand. Als solcher spielt sie innerhalb der Linguistik insbesondere in der Sprachty-
pologie 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 auf-
grund der Satzklammer und der unterschiedlichen Serialisierung im Haupt- und
Nebensatz am adäquatesten als “Mischtyp” zwischen Verbzweit- und Verbletzt-
sprache 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 Formalisierungs-
probleme 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 Serialisierungs-
unterschieden zwischen Ausgangs- und Zielsprache umgehen, und darzulegen, wie
die entsprechenden Strategien didaktisch vermittelt werden können.
Im folgenden möchte ich einige der wichtigsten wortstellungsbedingten Dol-
metschstrategien 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 Simultandolmet-
schen vgl. besonders die entsprechenden Publikationen der Triester Dolmet-
schwissenschaftler, 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 Simultan-
dolmetschen 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 91

(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 Dolmet-
schers 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 empiri-
schen Untersuchung zum deutsch-französischen Dolmetschen durch eine Heidel-
berger 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 Französi-
sche häufig angewandt, da das Französische aufgrund der möglichen Mehrfachbe-
setzung des Vorfeldes zahlreiche Möglichkeiten der Anteposition bietet:
(8) Er wurde im Alter von nicht einmal dreißig Jahren dank seiner unerschüt-
terlichen Überzeugung, daß keine noch so gut gemeinte Hilfe wirksam sein
kann, wenn ein Kind ohne Zuhause, ohne stabile soziale Rahmenbedingun-
gen aufwachsen muß, zum Begründer der SOS-Kinderdorf-Idee…
À l’âge de moins de tre- trente ans et grâce à sa conviction inébranlable que-
qu’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 verdolmet-
schende 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 Ver-
meidung 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):
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 organisier-
te Tagung setzt eine lange Tradition… fort….
Professor Sieber and his colleagues have organised this meeting most extra-
ordinarily 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 Sprachver-
gleich 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 Ziel-
sprachen 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 gesproche-
nen 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 Über-
setzungen anzugeben. In einigen Fällen funktioniert dies recht gut, denn es findet
sich eine vergleichbare Partikel im Französischen, wie in folgenden Übersetzungs-
vorschlägen von Blumenthal:
 Kontrastive Linguistik und sprachenpaarbezogene Translationswissenschaft 93

(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 deut-
schen 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 Proble-
matik der Null-Entsprechung behandelt René Métrich in einem Aufsatz zur “Nicht-
Übersetzung” von Modalpartikeln und anderen Gesprächswörtern im Sprachen-
paar 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)
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 Über-
setzung 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 irgend-
einer 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 zurück-
schleudern, 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 95

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 lösungsorien-
tiert 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 (Prunč 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 sprachenpaarbe-
dingten Probleme des Übersetzens und Dolmetschens möchte ich dem französi-
schen Sprach- und Übersetzungswissenschaftler Maurice Pergnier überlassen:
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 Überset-
zungswissenschaft”. 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 corrispon-
denti 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 langua-
ges”. 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 Über-
setzerische Kompetenz, Andreas F. Kelletat (ed), 91–104. Frankfurt: Lang.
Jörg, Udo. 1997. “Bridging the gap: Verb anticipation in German-English simultaneous inter-
preting”. 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 97

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’ens-
eignement, 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.
Prunč, 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 Snell-
Hornby 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: Hochschul-
Verlag.
Wandruszka, Mario. 1969. Sprachen — vergleichbar und unvergleichlich. München: Piper.
Weydt, Harald. 1969. Abtönungspartikel. Bad Homburg: Gehlen.
</TARGET "sch1">

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.
<LINK "zab-n*">

<TARGET "zab" DOCINFO AUTHOR "Patrick Zabalbeascoa"TITLE "Translating non-segmental features of textual communication"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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, follo-
wing 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 analy-
sed in its own terms as a way of gaining insight into how problems are sol-
ved. 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 analo-
gous 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 over-
another. 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 unconven-
tionally 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 classifica-
tion; 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 classifica-
tions. 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 101

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-translation-
problem, 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
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 parame-
ter. 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 meta-
phor and segment 0 belong to different paradigms, thus challenging Toury’s unit-
based 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-meta-
phor 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 relations-
hip 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 inter-
linguistic 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 bloodthir-
sty 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 borrow-
ing. 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 metaphori-
cal 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 b. S-set for P

[1] campo NO [1] NO

[2] sembrado NO [2] NO

n binary splits
[3] mar NO

[4] inacabables [5] todos [n] NO=[n+1]

aquellos

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 conven-
tions 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] NO

B B [2] NO

n binary splits
verbal non-verbal B

L2 not L2 S0 S0+X [n] NO=[n+1]

S0 S0+X S0 S0+X 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(meta-
phor) 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 trans-
lation. 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 illustra-
ted in the Figures. Indeed, Figure 1a illustrates the following analysis of example 1.
 Translating non-segmental features of textual communication 109

[1] Same metaphor, in Spanish; for example 1a (literal translation).

[2] Same — rural — type of metaphor; for instance 1b.
[3] Other — not rural — types of metaphor, including 1c.
[4] Other — not metaphorical — rhetorical, e.g. hyperbole, as in 1d.
[5] 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. hiato
‘hiatus’
b. cesura
‘caesura’
c. paréntesis
‘parenthesis’
d. freno
‘brake’
e. alto en el camino
‘stop on the way’
f. interrupción (tan) breve como un hiato
‘interruption (as) brief as a hiatus’
g. interrupción
‘interruption’
h. 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 inconsisten-
cy 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] Same irony; for example 3a.
[2] Same type of (metaphorical antithetical) irony; for instance 3b.
[3] Non-metaphorical antithetical irony; including 3c.
[4] Non-antithetical irony; for example 3d and 3e.
[5] 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”/com-
pound; 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.
<DEST "zab-n*">

</TARGET "zab">

Translating non-segmental features of textual communication 111

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 Non-
verbal Communication and Translation, F. Poyatos (ed.) 327–342. Amsterdam: John Benjamins.
<LINK "pok-n*">

<TARGET "pok" DOCINFO AUTHOR "Nike K. Pokorn"TITLE "Challenging the myth of native speaker competence in translation theory"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 com-
munity. 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 trans-
lation 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. Kocijančič 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; Skutnabb-
Kangas 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 115

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 assassina-
ted, at his home in Milford, New Jersey. Jože Paternost, the second Slovene trans-
lator, 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, Massachu-
setts. 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 educa-
tion, 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 Zmajić (1878–1944) was Croat, born in Croatia
but lived in Austria where she also died) and a native speaker of English M. Peters-
Roberts, 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
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 introduc-
tion 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 117

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 Native Non-native Cannot No
the translator(s) speaker speaker tell answer

Adamic Slovene 20 26 0 0
Zmajić & Peters-Roberts Croat and English 35 9 0 2
Paternost Slovene 14 32 0 0
Jereb & MacKinnon Slovene and 31 14 0 1
English
Leeming English 34 8 1 3
Klancar English 9 35 1 1
Klancar & Noyes English and 36 7 0 3
English

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

70
percent of responses

60

50

40

30

20

10

0
Zmajic & Jereb & Klancar &
Adamic Paternost Leeming Klancar
Peters-Roberts MacKinnon Noyes (Engl. &
(Slovene) (Slovene) (English) (English)
(Croat&Engl.) (Slov. & Engl.) Engl.)
native 43,48 30,43 79,07 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

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
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 compe-
tence 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 One More Cannot No
the translator(s) translator than one tell answer

Adamic Slovene 12 13 19 2
Zmajić & Peters-Roberts Croat & English 28 1 14 3
Paternost Slovene 15 4 22 5
Jereb & MacKinnon Slovene & English 15 6 23 2
Leeming English 22 1 17 6
Klancar English 19 4 15 8
Klancar & Noyes English & English 18 6 18 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 119

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

70

60
percent of responses

50

40

30

20

10

0
Zmajic& Peters- Jereb & Klancar &
Adamic Paternost Leeming Klancar
Roberts MacKinnon Noyes
(Slovene) (Slovene) (English) (English)
(Croat&Eng.) (Slov.&Eng.) (Eng.&Eng.)
one translator 27,27 65,12 36,58 34,09 55 50 42,86
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

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
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%
Zmajic & Peters-
Roberts (Croat &
English)
Klancar & Noyes 21%
(English & English)
Paternost (Slovene)
45%
6%
Jereb & MacKinnon
(Slovene & English)
Klancar (English)
17%
0%
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 trans-
lation classes, instead of dwelling on the demand that all translations be done only
into one’s mother tongue.
<DEST "pok-n*">

Challenging the myth of native speaker competence in translation theory 121

Author’s Address:
Nike K. Pokorn
University of Ljubljana
Department of Translation Studies
Askerceva 2
1000 Ljubljana
Slovenia
e-mail: Nike.kocijancic@guest.arnes.si

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.”
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 Zmajić, 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 123

“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 Wednes-
day. 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 Zmajić, 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.
</TARGET "pok">

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.
Kocijančič 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.
<TARGET "nob" DOCINFO AUTHOR "Marie-Louise Nobs"TITLE "Expectativas y evaluación en la traducción de folletos turísticos"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 para-
meters, 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 expectati-
vas 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 requerimien-
to 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 esencial-
mente 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 transpa-
rente 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. Veracidad de la información transmitida
b. Claridad y comprensibilidad del TM
c. Información útil y accesible para un usuario real del TM
d. Corrección gramatical
e. Estilo funcional acorde al tipo del TM
f. Presentación visual y gráfica atractiva
g. 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 fundamen-
talmente 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 momen-
to 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 involunta-
rio. 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 germanoha-
blantes 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 encuesta-
dos, 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 131

razones prácticas optamos por la autoadministración de los cuestionarios, es decir

que los mismos sujetos fueron los que los rellenaran, sin ayuda de ningún entrevis-
tador. Esta opción imponía, evidentemente, una redacción clara e inequívoca de las
preguntas, con el fin de evitar ambigüedades que pudieran dar pie a una cumpli-
mentación 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 Tradma-
nipul 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, centrán-

donos 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, signifi-
cando 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-
132 Marie-Louise Nobs

Información que corresponde

con la realidad
4,75
Información útil para un turista
4,58
Texto claro y fácilmente
4,37 comprensible
3,54 Ausencia de errores
gramaticales
3,37 Presentación visual y gráfica
atractiva
2,95
Estilo adecuado que no provoca
hilaridad

0 1 2 3 4 5
Gráfico 1.Valoración de parámetros
Tabla 1.Valoración de parámetros
1 2 3 4 5 ns/nc media

Información que corresponde con la realidad 0 0 1 4 19 0 4,75

Presentación visual y gráfica atractiva 0 2 11 11 0 0 3,37
Ausencia de errores gramaticales 0 3 9 8 4 0 3,54
Información útil para un turista 0 1 0 7 16 0 4,58
Texto claro y fácilmente comprensible 0 0 3 9 12 0 4,37
Estilo adecuado que no provoca hilaridad 2 5 10 4 2 1 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 133

Información no fehaciente
4,82 Comprensión dificultosa
4,62 Cambios respecto al contenido
del TO
3,91
Folleto redactado en estilo poco
3,54 habitual
Presentación visual/gráfica poco
3,5 atractiva
3,16 Folleto contiene errores
gramaticales
2,75 Se nota que es una traducción
2,33 Folleto produce hilaridad
2,04 Cambios respecto a la forma del
TO

0 2 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
1 2 3 4 5 ns/nc media

Folleto produce hilaridad 5 9 8 1 1 0 2,33

Folleto redactado en estilo poco habitual 0 2 9 11 2 0 3,54
Folleto contiene errores gramaticales 0 1 12 9 2 0 3,16
Se nota que es una traducción 4 5 8 7 0 0 2,75
Presentación visual/gráfica poco atractiva 1 2 5 11 4 1 3,5
Comprensión dificultosa 0 0 0 9 15 0 4,62
Información no fehaciente 0 0 0 4 20 0 4,82
Cambios respecto al contenido del TO 2 1 6 7 6 2 3,91
Cambios respecto a la forma del TO 5 10 5 1 1 2 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 previa-
mente. 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
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 paráme-
tros 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).

4,75 4,83
5 4,62
4,37
veracidad de la información
4 3,54 transmitida
3,16
claridad y comprensibilidad
3 del TM
corrección gramatical
2

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
Valor medio Valor medio
pregunta 1 pregunta 8

veracidad de la información transmitida 4,75 4,83

claridad y comprensibilidad del TM 4,37 4,62
corrección gramatical 3,54 3,16
presentación visual y gráfica atractiva 3,37 3,5
estilo usual que no provoca hilaridad 2,95 2,33
 Expectativas y evaluación en la traducción de folletos turísticos 135

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

1 2 3 4 5 ns/nc media

Trpubl 0 0 8 3 0 1 3,27
Trmanipul 0 1 5 4 2 0 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.
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

1 2 3 4 5 ns/nc media

Trpubl 0 0 6 4 1 1 3,54
Trmanipul 0 1 3 3 5 0 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 137

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 encabeza-
miento 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 expectati-
vas 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 conteni-
do 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.
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 esencial-
mente 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 consulta-
dos 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).
</TARGET "nob">

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 Interpreta-
ción 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.
<LINK "mal-n*">

<TARGET "mal" DOCINFO AUTHOR "Kirsten Malmkjær"TITLE "Censorship or error"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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-Text-

oriented 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) distinc-
tion between semantic patterning in the case of the former and formal (gra-
phemic/phonemic) patterning in the case of the latter. Since it is necessary to
pre-establish the distinction between error and choice in translation on non-
linguistic 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 argu-
ment 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 transla-
tional 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 unintentio-
nal, 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 motiva-
ted 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 motiva-
ted 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 Ander-
sen’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 awarenes-
ses 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 145

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, produ-
ced 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 differen-
ces 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
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 Howitt 1846

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 “Little Ida’s Flowers”
Tommelise [Thumbelina] “Tommelise”
Den uartige Dreng “The Naughty Boy”
Reisekammeraten [The Travelling
Companion]
1837
Den lille Havfrue [The Little Mermaid]
Keiseren’s nye Klæder [The Emperor’s New
Clothes]
1838
‘Gaaseurten’ ‘The Daisy’
‘Den standhaftige Tinsoldat’ ‘The Constant Tin Soldier’
‘De vilde Svaner’ [The Wild Swans]
1839
‘Paradisets Have’ ‘The Garden of Paradise’
‘Den flyvende Koffert’ (“Svovlstikkerne”) ‘A Night in the Kitchen’
‘Storkene’ ‘The Storks’
1842
‘Ole Lukøie’ ‘Ole Lukoie — The Story-Teller at Night’
‘Rosenalfen’ ‘The Rose-elf ’
‘Svinedrengen’ [The Swine-boy]
‘Boghveden’ [The Buckwheat]

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 repulsi-
ve, 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 interrelations-
hips 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 skar hans Hoved af og begravede det med Kroppen i den bløde Jord
gloss cut his head off and buried it with the body in the soft earth
tt ‘and then buried him in the bloody earth’ (58)
st saa tog hun det blege Hoved med de lukkede Øine, kyssede den
gloss then took she the pale head with the closed eyes kissed the
st kolde Mund og rystede Jorden af hans deilige Haar
gloss cold mouth and shook the earth off his lovely hair
tt ‘so she cut away a beautiful lock of his hair, and laid it near her heart’ (60)
st tog hun Hovedet med sig hjem
gloss took she the head with her home
tt ‘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 Gud, hvor den stakkels Tommelise blev forskrækket,
gloss God how the poor Thumbelina became frightened
tt ‘Poor Tommelise! How frightened she was’ (38)
st Tak Du Gud for ham!
gloss Thank you God for him
tt ‘Be thankful that thou canst get such a one!’ (49)
st at de aldrig vilde døe, og at Paradisets Have skulde evig blomstre
gloss that they never would die and that Paradise’s garden should ever bloom
tt 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 De skjønneste Piger, svævende og slanke, klædte i bølgende Flor,
gloss The most beautiful girls, floating and slim, dressed in waving gauze,
st saa man saae de deilige Lemmer, svævede i Dandse,
gloss so one saw the lovely limbs floated in dances
tt ‘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 og altsom de blev større, vilde de mindre taale det; Moderen

gloss and as they grew bigger would they less bear it the mother
st maatte tilsidst love dem, at de nok skulde faa Hævn
gloss must finally promise them that they part should get revenge
tt ‘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 “Nu skulle vi hævnes!” sagde de.
gloss “now shall we be revenged said they
st “Ja vist!” sagde Storkemoderen. “Hvad jeg have udtænkt,
gloss “yes certainly said the stork mother “what I have planned
st det er just det rigtige!”
gloss that is just the right (thing)
tt “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 Strudsen
gloss the ostrich
tt ‘the simoom’ (9)
st Bajonetten nede imellem Brostenene
gloss the bayonet down between the cobble stones
tt 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 han vil bare have at de skulle være rolige
gloss he will only have that they shall be still
tt ‘he only wants to amuse them’ (1)
st see bare, hvor rolig Jer Fader staaer
gloss see only how still your father stands
tt ‘Look how funnily your father stands’ (121)
(In Swedish, rolig means, roughly, amusing; in Danish it means, roughly, still/quiet/
peaceful).
st over Isen
gloss across the ice
tt ‘over the iron’ (70)
(In German, Eisen means ‘iron’; in Danish, isen means ‘the ice’)
st gjorde Cuur til
gloss courted
tt ‘cured’ (18)
st det kan man saa rart
gloss that can one so nicely
tt ‘people can do that so seldom’ (93)
Other errors look as if they are caused by graphemic similarities that create rela-
tionships 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 151

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 Næste Nat listede hun sig igjen ned til den
gloss next night tip-toed she reflexive again down to it
tt ‘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 “inconsis-
tencies” of this type in detail.
st jeg har kjørt i en Bøtte
gloss I have driven in a pot
tt ‘I have gone in a boat’ (16)
st jeg skal ud paa Marken
gloss I shall out on the field
tt ‘I shall go out into the market’ (17)
st Den bløde Jord
gloss the soft earth
tt ‘the bloody earth’ (58)
st Græsset er blødt
gloss the grass is soft
tt ‘the grass is in flower’ (23)
It is possible that the above is an influence from German, where blühen means ‘to
be in bloom’.
152 Kirsten Malmkjær

st Her er en slem Træk!

gloss here is a bad draft
tt ‘It is bad travelling to-night,’ (67)
st (en) Taage
gloss (a) fog
tt ‘ceiling’ (80)
st sagde alle Talerknerne
gloss said all the plates
tt said all the talkers (93; also on p. 94)
st Torvekurven
gloss the (market)square basket
tt the coal-box (92; also twice on p. 95)
st Adolph og Jonas skjød med Flitsbuer over Graven
gloss Adolph and Jonas shot with bows over the grave
tt ‘Jonas and Adolph stood with their cross-bows above the grave’ (110)
st Nøddeknækkeren slog Kaalbøtter,
gloss the nutcrackers did (literally hit) somersaults
tt ‘The nutcrackers knocked about the gingerbread nuts’ (113)
st at han ikke kunde standse.
gloss that he not could stop
tt ‘That he could no longer stand upright’ (116)
st kukkelure
gloss mope about
tt ‘cough’ (123)
st nu flyve vi høire om
gloss now fly we (to the) right around
tt ‘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 sommerti-
der (‘summer-times’), that is, it includes misleading graphemics:
st Sommerfuglene
gloss the butterflies
tt ‘the summer birds’ (6)
Note that this term is appropriately translated on four subsequent occasions.
 Censorship or error 153

st da vendte hans Tanker tilbage.

gloss then turned his thoughts back [the phrasal verb vende tilbage means
‘return’, so the sense is that his thoughts returned]
tt ‘His thoughts turned to the past’ (88)
st “Sommetider, om (Natten,
gloss “sometimes at (the) night
tt ‘In the summer season at night’ (99)
st “Nei lad være med det!”
gloss “no let be with that [meaning ‘do not do that’]
tt ‘No, let them be’ (124)

Type 5: Homonym selection

Some of the examples below seem to be selections of the wrong sense of a homo-
nym, though other factors, such as misunderstanding of syntax or phrasal verb
status are also at play.
st det skal være saadan en deilig Leilighed!”
gloss it shall be such a lovely flat
tt ‘it will be such a nice opportunity (10)
st Svalen
gloss the swallow
tt the breeze (15; three occurrences)
st “Nu tør man nok ikke mere sige sin Mening!”
gloss “now dare one particle not more say one’s meaning
tt ‘“Now, one cannot tell what he means”’ (18)
st Muldvarpen stødte til den
gloss the mole knocked (in)to it
tt ‘the mole stood beside it’ (43)
st der ville vi nyde Forfriskninger
gloss there will we enjoy refreshments
tt ‘it will be refreshing to us’ (83)
st det syntes ogsaa, at det hørte med til Blomsterne
gloss 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’]
tt which had heard the flowers (105)
st Ansigtet satte de lige ud;
gloss the face set they straight ahead
tt ‘their faces were all the same’ (111)
<DEST "mal-n*">

154 Kirsten Malmkjær

st de vilde være med,

gloss they would be with [the phrasal verb være med means ‘participate’]
tt ‘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 Træerne fortalte om Røvere og Hexe og Blomsterne om de nydelige
gloss the trees told of robbers and witches and the flowers of the pretty
st smaa Alfer og hvad Sommerfuglene havde fortalt dem.
gloss small elves and what the butterflies had told them
tt ‘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 da kunde Blomsten ikke, som Aftenen forud, folde sine Blade
gloss then could the flower not like the evening before fold its petals
st sammen og sove, den hang syg og sørgende ned mod Jorden
gloss together and sleep it hung ill and grieving down towards the earth
tt 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.
</TARGET "mal">

Censorship or error 155

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 transla-
tors?”. 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.
<TARGET "sch2" DOCINFO AUTHOR "Thorsten Schröter"TITLE "Of holy goats and the NYPD"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 difficul-

ties 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” transla-
tions, 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 Swedish Swedish Danish Norwegian French
audio subtitles subtitles subtitles subtitles subtitles subtitles
(DVD) (TV) (DVD) (DVD) (DVD) (DVD)

Analyze… x x x x x x
Four Wed. x x
MiB x x x x x x x
Mickey… x x x x x x
Road Trip x x x x
Nutty P. I x x x x x x
Nutty P. II 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 Norwe-
gian 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 161

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 inappro-
priate 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 near-
homophones, 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 appro-
priate 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.
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 be-
deutet: 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 manipula-
tion 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 interpre-
tation 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 Nor- Danish French
original dubbing subtitles subtitles subtitles wegian subtitles subitles
audio (TV) (DVD) subtitles

N-Y-P-D 2 1/2 1/2* 0 1/2 1/2 1 0/1

big butt 1 0 0* 0 0 0/1 0 0
back
half the man1 1 1* 0/1 1 1 1 0
there-square1 1 1* 0 0* 0 1 1
memory- 1/2 1 1* 0 0* 0 0/1 0/1
thing
White/Black 1 0/1 0/1* 0/1 0/1* 0/1 0/1 0/1
flashy mem- 1 1 0/1 0 0/1 0 0/1 0/1
ory
TOTAL 8/9 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 No. of I. II. III. IV. Average
strategies strategies min TL min TL max TL max TL of I. to
in SL in TL max SL min SL max SL min SL IV.
min/max min/max

German dubbing 83/93 46/59 0.49 0.55 0.63 0.71 0.60

German subtitles 69/79 39/54 0.49 0.57 0.68 0.78 0.63
Swedish (TV) 17/18 7/9 0.39 0.41 0.50 0.53 0.46
Swedish (DVD) 74/84 43/57 0.51 0.58 0.68 0.77 0.64
Norwegian 74/84 43/58 0.51 0.58 0.69 0.78 0.64
Danish 74/84 42/59 0.50 0.57 0.70 0.80 0.64
French 43/51 27/39 0.53 0.63 0.76 0.91 0.71

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 lan-
guage-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 correspon-
ding 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 transla-
tion 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 German Swedish Swedish Norwegian Danish French
dubbing subtitles subtitles subtitles
(TV) (DVD)

68 71 50 64 71 74 73
 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 langua-
ge, 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
</TARGET "sch2">

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 Shakes-
peare’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.
<TARGET "mil" DOCINFO AUTHOR "John Milton"TITLE "The figure of the factory translator"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 171

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, Trad-
Prt List trad-prt@egroups.com (now trad-prt@yahoogroups.com), 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 undergra-
duate 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 non-
research 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 Appen-
dix, 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.
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, undergra-
duate 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 173

teachers teaching translation and translation theory courses, publishing in speciali-

zed 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 know-
ledge 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 profes-
sion. 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 tradinfo@egroups.com (now tradinfo@yahoogroups.com.br), a list which
provides information about translation events in Brazil; two general lists for
translators from and into Portuguese, Trad-Prt List trad-prt@egroups.com (now
trad-prt@yahoogroups.com.br, tradutores tradutores@egroups.com (now traduto-
res@yahoogroups.com.br); Abrapt abrapt@egroups.com (now abrapt@yahoo-
groups.com.br), a list for members of the Brazilian Association for Translation
Research; litterati litterati@egroups.com (now litterati@yahoogroups.com.br), a
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 engpos@yahoo-
groups.com (now engpos@yahoogroups.com.br). The questionnaire was subse-
quently 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 175

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 scholars-
hip 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 commit-
ments, 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
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 American-
Brazilian 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 profes-
sional life of translators?

Author’s Address:
John Milton
University of São Paulo, Brazil
DLM, FFLCH, USP, 05508–900
jmilton@usp.br

Appendix

Table 1a.Reasons for choosing course — Rio Pretoa

1st year 2nd year 3rd year 4th year Total

Course 4 11% 3 7% 6 17% 11 45% 24 17%

qualities
Language 19 53% 27 61% 16 46% 6 25% 68 53%
skills
Free public – 1 2% 6 17% 4 16% 11 8%
univ.
Not to teach – 3 7% 1 3% 2 8% 6 4%
Near home 2 5% – 2 6% 1 4% 5 4%
To be 10 28% 6 14% 3 9% – 19 14%
translator
Others 1 3% 2 4% 3 9% 6 4%
Total 36 42 37 24 139
a
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 2nd year 3rd year 4th year Total
7%
Course – – 5 17% 8 10% 3 16 8%
qualities
More job 4 9% – – 3 7% – – 7 4%
opportunities
Language skills 28 62% 13 45% 52 63% 27 66% 120 62%
Part of job – – 5 17% 5 6% – – 7 4%
To be 6 13% 1 3% 8 10% 5 12% 20 10%
translator
Others 7 15% 5 17% 7 8% 6 15% 23 12%
Total 45 29 83 41 193

Table 3.Profession of USP translation students

Profession No. %

Translator 16 25%
None 9 14%
Teacher 29 45%
Others 10 16%

Total 64

Table 4.Students who would have taken undergraduate translation course: Letras and
non-Letras
Letras Non-Letras

Yes %Letras No % Letras Yes % non-Letras No % non-Letras

Students 22 73% 8 27% 8 32% 17 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 Transla-
tor. 6:1, 49–65.
</TARGET "mil">

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.
<LINK "ris-n*">

<TARGET "ris" DOCINFO AUTHOR "Hanna Risku"TITLE "Migrating from translation to technical communication and usability"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 transla-
tors 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 inclu-
des 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 qualita-
tive) 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 conside-
ration:
– 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 co-
operative 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 transla-
tion 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 183

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-technical-
communicators 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. What originally motivated you to shift to TC?
2. Did you notice any similarities to or differences from translation work?
3. Do TC projects resemble translation projects or do they differ greatly?
4. Which tasks did you find especially easy or difficult?
5. Do you think your translating background gave you any advantages or disad-
vantages over technical communicators from other backgrounds?
6. Did your professional role and status change or do you consider both profes-
sions 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
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 male-
dominated 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 origina-
ting 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 185

The presentation of the study is structured into the four sub-competencies of

translation presented in Risku (1998):
1. Target analysis
2. Source analysis
3. Development of an overall strategy
4. 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
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 well-
defined 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, back-
ground 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 communi-
cators 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 transla-
tion 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 unders-
tand 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 unstruc-
tured 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 expe-
rience 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 191

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 self-
confidence 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 (Human-
Computer-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 depart-
ments, 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
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 well-
researched 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 organisatio-
nal 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 characteris-
tics that are equally important for both TC and translation. These essential compo-
nents 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 informa-
tion society. In this situation, translators migrating to technical communication are
in a unique, advantageous position to fill such vacancies now and in the future.
<DEST "ris-n*">

194 Hanna Risku

Author’s Address:
Hanna Risku
Danube University Krems, Austria
Hackenberggasse 45 C5
A-1190 Vienna
hanna.risku@donau-uni.ac.at

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 vermit-
teln. 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 Translationswis-
senschaft. 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.
</TARGET "ris">

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. Amster-
dam: Benjamins.
Suchman, Lucy. 1987. Plans and Situated Actions. Cambridge: Cambridge University Press.
Suojanen, Tytti. 2000. Technical Communication Research: Dissemination, Reception, Utiliza-
tion. 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.
<TARGET "dra" DOCINFO AUTHOR "Barbara Dragsted and Benjamin Kjeldsen"TITLE "From raw data to knowledge representation"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 pro-
viding 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 soft-
ware 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 descri-
bes a partly new approach to leveraging the potential of machine translation
and intelligence processing by involving human translators in the develop-
ment 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 expecta-
tions 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 endeavou-
red 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 transla-
tion 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 termino-
logy 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 re-
structuring 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 near-
synonyms in the target language and hence requires clearly distinguished termino-
logy 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 infor-
mation 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 informa-

tion, 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 develo-
pers and police translators. The NLP software developers were, as already mentio-
ned, 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 develo-
pers. 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 termino-
logy exchange methodology ‘the electronic basket procedure’, because the terminolo-
gy 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 collec-
tion, (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 Rules &
Regulations 2
1
Person Authorities

Perpetrator Victim Other Police Courts Other

Activities Activities Activities

3 Event
Time/
Place Procedures Procedures Procedures

Objects Objects Objects

4
MO Docu- Docu- Docu-
ments ments ments
5 Physical
Objects Evidence Evidence Evidence

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 Docu-
ments 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 multi-
tude 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 informa-
tion relevant to the user groups as well as to the linguistic software system develo-
pers. 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 identifi-
cation 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 terminologi-
cal 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 institu-
tions 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, conse-
quently, that of the project in general.

Author’s Address:
bd.eng@cbs.dk, bek.eng@cbs.dk
</TARGET "dra">

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 Termino-
logy 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 trans-
lation”. 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/
<TARGET "aze" DOCINFO AUTHOR "João Azenha Junior"TITLE "The translator as a creative genius"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 Por-
tuguese. 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 magni-
tude 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 Schu-
mann’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 211

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 above-
average 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, Schu-
mann 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 Jüngling-
lebens”, 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
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 birth-
day), 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 transla-
ted 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 particular-
ly 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 langua-
ges, 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 213

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 transla-
tions, 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 transla-
ted 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
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 ever-
present 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 philosophi-
cal 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 conside-
red 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 re-
reading of the past paves the way for a comprehension of the present, simultaneous-
ly 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 215

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 manus-
cripts 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)
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 [Schu-
mann] 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).
</TARGET "aze">

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 Schwer-
punkte, 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-For-
schungsstelle 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.
<TARGET "doo" DOCINFO AUTHOR "Luc van Doorslaer"TITLE "Übersetzung zwischen Nationalismus und Internationalismus"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 schwer-
lich 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 überge-
hen, 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 inter-
national 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 Problema-
tik 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 franzö-
sischsprachigen 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). Merk-
würdig 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 “inter-
national”, 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 Klassifizierungsver-
suchen 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 Qualitä-
ten. 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’ (ver-
weisend 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 grenzüber-
schreitend eine Fußball-Europameisterschaft, Euro 2000, haben organisieren
können…).
Dennoch, als allerneueste Tendenz zeichnen sich jetzt gelegentlich auch gegen-
sätzliche 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 Reregionalisierungs-
tendenz 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 Sprach-
brüdern 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 Übersetzungs-
realität 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-linguisti-
schen Kontext zu verstehen […]: zumal sie nicht nur strukturelle — phoneti-
sche, semantisch, syntaktische -, sondern auch funktionale, d.h. gesellschaftli-
che Aspekte aufweist.
Eine der wesentlichen gesellschaftlichen Funktionen literarischer Über-
setzungen 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 Internationali-
sierung ist öfter festzustellen, dass sich Identifikation, z.T. auch Bewusstseins-
bildung 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 Spra-
chenverhältnisse 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 durch-
geführten 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 be-
stimmten EU-Ausschüssen oder Arbeitsgruppen mit Blick auf die Effizienz nur
‘Arbeitssprachen’ benutzt werden. Die Kriterien für die Auswahl dieser Arbeits-
sprachen 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 Zusammen-
hang aufgeführt wird, ist die Übersetzung irischer Literatur ins Englische (Ty-
moczko 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 zuein-
ander? 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 Überset-
zungswissenschaft sprengen, andererseits könnten sie durch lehrreiche Querver-
bindungen mit den cultural studies vielleicht auch die Flexibilität und kommunika-
tive 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 approa-
ches, 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 Über-
setzers 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.
</TARGET "doo">

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.
<TARGET "ahr" DOCINFO AUTHOR "Barbara Ahrens"TITLE "Non-verbal phenomena in simultaneous interpreting"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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, typi-
cal patterns. The paper discusses the possible reasons for this fact and des-
cribes 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 trans-
parencies, 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 misinter-
pretations since they depend on the culture they represent (cf. Poyatos 1987: 103ff.).
This specific situation and the overlapping perception of the ST and the produc-
tion 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 stressed syllable written in capitals
·0.32Ò pause, duration: 0.32 seconds
·0.32:AÒ pause due to breathing, duration: 0.32 seconds
[xyz] context of the parts that are shown in Figures 1 to 4
azosia- false start
 Non-verbal phenomena in simultaneous interpreting 231

For legibility reasons, pauses have only been indicated in the passages correspon-
ding 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 telecom-
muniCAtions so WHY is there this ability to do] something WELL or efFECti-
vely 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 intona-
tion group (cf. Crystal 1969: 227). This is evident in Figure 1. In this example, the
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. Crutten-
den 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 informa-
tion 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 perfor-
mance, 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 simultane-
ously 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 investiga-
tion 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 simulta-
neous 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.
</TARGET "ahr">

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 Dol-
metschprozeß, Dolmetschmodelle und Dolmetschstrategien”. In Theorie und Praxis des
Übersetzens und Dolmetschens [Publikationen des Fachbereichs Angewandte Sprachwissen-
schaft 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 inter-
pretation: 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.
<TARGET "bar" DOCINFO AUTHOR "Magdalena Bartlomiejczyk"TITLE "Simultaneous interpreting A-B vs. B-A from the interpreters’ standpoint"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 stu-

dents’ 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 profes-
sional 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 percei-
ved, 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 particu-
lar 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 elegantly-
worded and an impeachably pronounced half-message or less. (1989: 157)
A small-scale empirical study by Tommola and Helevä (1998), in which they in-
vestigate 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 language-
specific, 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 Al-
Khanji (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 inter-
preting, approximately in the proportion 50:50.
Questionnaires for professional interpreters were distributed between Decem-
ber 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 mentio-
ned 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

0
n
ar

nt
ed
tio

in

lar
m

ce
pe

nd
na

bu
m

ac
ys

ta
ra

to

ca

r’s
rs
er
g

in

vo

te
de
liv
r’s

r’s

re
r’s
un
te

de

rp
te

te
re

re

xt

te
re
rp

rp

in
te

rp
te

te

ce

te
in

in

ur

in
so

B-A 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 inter-
pretation 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 differen-
tiate 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 respon-
dents 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 inter-
preting 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

0
n
ar

nt
ed
tio

in

lar
m

ce
pe

nd
na

bu
m

ac
ys

ta
ra

to

ca

r’s
rs
er
g

in

vo

te
de
liv
r’s

r’s

re
r’s
un
te

de

rp
te

te
re

re

xt

te
re
rp

rp

in
te

rp
te

te

ce

te
in

in

ur

in
so

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 conside-

rably in their opinions concerning retour interpreting in comparison with inter-
preting 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 systema-
tically 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 respon-
dents, 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
magdalenabartlomiejczyk@hotmail.com
248 Magdalena Bartlomiejczyk

Appendix: Questionnaire distributed to interpreting students

(All questionnaires were in English. The questionnaire distributed to professional inter-

preters 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 simultane-
ous 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)
</TARGET "bar">

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 inter-
pretation 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 interpreta-
tion”. 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.
<TARGET "dam" DOCINFO AUTHOR "Helle V. Dam"TITLE "The interpreters’ notes"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 dis-
cussion here are (1) interpreters’ choice of form for their notes (symbols vs.
language) and (2) their choice of language (source, target or other langua-
ges). Because of the pilot status of the study, it provides no rock-solid re-
sults, but actually raises more questions than it answers. The main contribu-
tion 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 litera-
ture, 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 small-
scale 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 observa-
tions alone, but it would be interesting to see whether a similar pattern of symbol-
word 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 task-
related, but reflect differences in the interpreters’ personal or professional back-
grounds, 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-generali-
sation, 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 reinfor-
cers, 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 symbol-

word 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 conver-
sion 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 summari-
sed, 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 107 (58%) 117 (63%) 163 (87%) 125 (77%) 102 (77%) 614 (72%)
SL 46 (25%) 22 (12%) 3 (2%) 3 (2%) 11 (8%) 85 (10%)
3rd L 9 (5%) 29 (16%) – 2 (1%) – 40 (5%)
? 23 (12%) 19 (10%) 22 (12%) 33 (20%) 20 (15%) 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 indivi-
dual 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 ex-
trapolated to other tasks performed by the same interpreters, because the high
degree of agreement between subjects found here indicates that the observed target-
language 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 so-
called 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 TL-
oriented 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 target-
language-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 consecu-
tive interpreting, the phase relevant to this study, namely the listening and note-
taking 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 note-
taking 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).
</TARGET "dam">

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. Frank-
furt: 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 Conferen-
cias. 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 sprachunab-
hängigen 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.
<TARGET "fen" DOCINFO AUTHOR "Sabine Fenton"TITLE "Expressing a well-founded fear"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 enhan-
ce the role of the legal interpreter with the advocacy functions that inter-
preters 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 asso-
ciation 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 ap-
plicant’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 well-

meaning 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 under-
standing 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 restrai-
ned 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 recommenda-
tions 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 pres-
cription 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,
</TARGET "fen">

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 inter-
preter 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.
<TARGET "rud" DOCINFO AUTHOR "Mette Rudvin"TITLE "Cross-cultural dynamics in community interpreting. Troubleshooting"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

Cross-cultural dynamics in community

interpreting. Troubleshooting

Mette Rudvin
University of Bologna

In this paper a number of issues related to interpreter-mediated communi-

cation for public services are examined. It is argued that the encounter bet-
ween host country and migrant is governed by power relations where hier-
archy 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, mismat-
ching 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 inter-
preter-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 imperia-
lism). 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 globalisa-
tion 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 developed-
developing 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 eloquent-
ly 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 hier-
archical 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 compara-
tive 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 differen-
ces 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 commu-
nication 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 “meaning-
creating” 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 much-
discussed 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 cross-
cultural 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, under-
standing “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 inter-
preter’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 interpreter-
mediated 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 meta-
phor 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 other-
wise), 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 streng-
thened 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 culture-
bound 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 accom-
plished best by staying within the politeness system, including the use of that
form of politeness expected by one’s culture in the specific context. Abroga-
tions 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 non-
adversarial 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 inap-
propriately 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 mis-
matching 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 deducti-
ve (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 cour-
troom 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 infor-
mation 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 communica-
tive encounter. Although a discussion of each of the following lies beyond the scope
of the present paper, some factors could be mentioned: gender, the culture-
boundedness 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 com-
pletely 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 inter-
preting 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 higher-
status 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 attach-
ment 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.criti-
callink.org/english/proceedings.htm.
</TARGET "rud">

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 Publica-
tions.
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.
<TARGET "hal" DOCINFO AUTHOR "Nigel Hall"TITLE "The child in the middle"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 pa-
pers 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 com-
plex experience; it challenges conventional power relationships within fami-
lies or schools, challenges chronological age expectations, and allows chil-
dren to act strategically as experts. This chapter considers the language bro-
kering behaviour of group of Asian heritage children taking part in a com-
plex simulation in which they mediated an interview between a school repre-
sentative 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 unrewar-
ded, 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 broke-
ring. 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 opportu-
nity 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 inter-
views). 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 Urdu-
speaking 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 participa-
tion 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 agree-
ment 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 What’s she saying?

C 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 She doesn’t want him to play a lot.

T Don’t worry it’s playing with things to help them to learn.
C They learn him something. They learn through play, then they learn something.
M 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 considera-
ble power is invested in school authorities. It starts with the compulsory nature of
schooling, is deeply embedded within the organisation of schooling, and is accom-
panied 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 practical-
ly 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 expecta-
tions 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 tran-
scriptions 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 in-
stitution 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 irresponsi-
ble 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, mecha-
nical 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 tri-
interactional 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 representa-
tive 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 inter-
vention 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 represen-
tative’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 partici-
pant 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 demon-
strates 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 pay-
ment; 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: n.hall@mmu.ac.uk

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.
</TARGET "hal">

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 Inter-
pretation 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 Chinese-
and 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. Cam-
bridge: 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.
<TARGET "gil" DOCINFO AUTHOR "Daniel Gile and Gyde Hansen"TITLE "The editorial process through the looking glass"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

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 institutio-
nal factors, sociological factors, and the availability of referees for peer-
reviewing. 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 repre-
sentation, 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 in-
built 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 in-
between, 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 weaknes-
ses 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, ambigu-
ous 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 com-
prehensively 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 unbalan-
ced 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 proposi-
tion, 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. Never-
theless, the following could be pointed out:
– The set of more than 42 papers reviewed for these proceedings can be conside-
red 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, McBur-
ney 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 corre-
sponding background have ventured into statistics, philosophy of science, lin-
guistics, 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

– 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.

5. 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.
</TARGET "gil">

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.
<TARGET "ni" DOCINFO AUTHOR ""TITLE "Name index"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

Name index

A Blum-Kulka 7, 8, 10, 12
Abraham 90, 96, 217, 218 Blumenthal 92, 93, 96
Adab 182, 194, 195 Bolinger 53, 69
Adamic 114, 117, 118, 121, 123 Borges 31, 37
Adegbija 282 Bredsdorff 144–6, 151, 155
Aguilar-Amat 223, 225 Brisset 220, 221, 225
Ahlqvist 31, 32, 38 Brøndum 79, 80
Ahrens vi, xii, 227–9, 236, 251, 261 Bühler 230, 235, 236
AIIC xii, 241, 242, 247, 256, 260, 261 Burton 31
Aijmer 143, 155 Bush 144, 155
Aitchison 56, 69 Buzzo Margari 94, 96
Al-Khanji 241, 249
Al-Salman 241, 249 C
Albrecht 84, 96, 296 Cankar 114, 116, 123, 124
Alexieva 230, 236, 251, 252, 256, 261 Cardoso de Camargo 63, 69
Andres 251, 260, 261 Carliner 194
Antos 46, 48, 50, 229, 236 Carroll 13, 33, 38
Araújo 173, 178 Chafe 53, 69
Chalmers 301, 306
B Chesterman v, vii, xiii, 1, 4, 12, 24, 28,
Babbie 304, 306 35, 37, 124, 179, 195, 225
Bailey 286, 295 Chevalier 10, 12
Baistow 282, 283 Chiaro 168
Baker 7–10, 12, 25, 60, 69, 155 Clifford 283, 286
Barsky 264, 265, 267–9 Cohen 1, 12
Barthes 37, 41, 50 Colin 264, 269
Bassnett 39, 43, 50 Collados 127, 139, 230, 236
Bastian 88, 96 Collados Aís 139, 230, 236
Beaud 304, 306 Combettes 74, 80
Beerbom 94, 96 Corbin 302, 304, 306
Bensimon 27, 29, 33, 37 Coseriu 84, 96
Bergenholtz 200, 207 Cronin 225, 249
Bericat 138, 139 Cruttenden 229, 231–3, 237
Berk-Seligson 264, 269, 276, 277, 283 Crystal 114, 124, 228, 229, 231, 234,
Berman 4–6, 8, 9, 12, 27, 35, 37, 216, 237
218
308 Name index

D G
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 Grave 152, 222, 226, 277
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 Hämeen-Anttila 31, 37
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
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
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
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 Kocijančič Pokorn 113, 124

Herbert 239, 249, 256, 261 Koller ii, 84–6, 97
Hermans 20, 25, 28, 37, 170, 178 Königs 86, 94, 95, 97, 139
Herslund 74, 76, 80, 81 Korzen v, viii, ix, 73, 74, 77–81, 85, 97
Hjørnager-Pedersen 145, 151, 154 Koskinen v, vii, viii, 27, 36, 170, 178
Holmes 25, 101, 111 Kovala 37
Holz-Mänttäri 126, 139 Krashen 114, 124
Hönig 126, 129, 139 Kreisig 217, 218
Howarth 55, 56, 59, 61, 70 Krockow 219, 226
Huber 227, 230, 232, 233, 237 Krogh 91, 92, 97
Hughes 100, 111 Kropat 92, 97
Hutchins ii, 200, 207 Kruse 216–8
Kuhn 301, 306
I Kujamäki 28, 37
Igarashi 301, 306 Kundera 5, 6, 13
Ilg 92, 96, 251, 252, 256, 261 Kunnas 33, 34
Irwin 37 Kussmaul 58, 70, 170, 179

J L
Jäger 85, 96 Lado 85, 97
Jänis 126, 139 Lakoff 99, 100, 111, 276, 277, 283
Jansen 212, 217, 218 Lambert ii, 170, 178, 225, 226, 237,
Järventausta 84, 96 251, 252, 261
Jereb 115, 117, 118, 122, 124 Lane 31, 37
Johnson 99, 100, 111 Lang 37, 50, 90, 91, 96, 97, 168, 236,
Jones 240, 249 237, 261
Jörg 90, 96 Larose 129, 139
Judd 298, 302–4, 306 Latiesa 128, 139
Laviosa 8, 10, 13, 19, 25
K Laviosa-Braithwaite 8, 13, 19, 25
Kalina 90, 92, 96, 228, 235, 237 Lederer 252, 256, 261
Kaur 286, 296 Leech 100, 111
Kenny 8, 13, 249 Leeming 115, 117, 118, 123, 124
Kidder 306 Lefevere 28, 37
Kielhöfer 85, 96 Lehmann 78
Kim 282, 283 Levelt 56, 70
Kintsch 235, 237 Lewis 33, 38, 53, 70
Kirchhoff 227, 228, 237, 251, 256, 261 Liebkind 282, 283
Kiuru 30, 37 Lieshout 229, 237
Kivi 30 Louth 121
Kjørup 301, 306 Lozar 121
Klancar 115, 117, 118, 123 Lundquist 75, 81, 85, 97
Klaudy 8, 11, 13, 18, 25, 58, 59, 62, 70
Knapp 285, 296 M
Koch 89, 90, 97 MacKinnon 115, 117, 118, 122, 124
310 Name index

Malblanc 85, 97 Niiniluoto 2, 13

Malmkjær ii, iii, iv, v, x, 141, 142, 155 Nobs v, ix, x, 125, 138, 139
Manner 19, 33, 298 Nooteboom 221
Marello 85, 97 Nord 86, 87, 97, 139
Martin 25, 33, 34, 38, 70, 217, 218, Noyes 115, 117, 118, 123
296, 306
Martinet 89 O
Mason 69, 126, 139 Offord 166, 168
Matyssek 251, 252, 261 Oittinen 28, 33, 34, 38
Maupin 79 Oksefjell 143, 155
Mauranen 8, 13 Ondaatje 79
McBurney 299, 304, 306 Orwell 53, 56, 57, 70
McLaughlin 121 Øverås 8, 13
McQuillan 286, 296 Oxenbury 34
Medgyes 114, 124
Mellman 274, 283 P
Meta 13, 37, 126, 129, 130, 139, 143, Paikeday 114, 124
155, 249 Paine 286, 296
Métrichs 93 Paloposki v, vii, viii, 27, 29, 36, 38
Mikkelsen 77, 80, 81 Pápai 60, 70
Millán-Varela 224, 226 Pasolini 35
Mills 286, 296 Paternost 115, 117, 118, 122, 123
Milton vi, x, 9, 13, 169, 170, 177, 179, Pawley 52, 54, 55, 70
306 Pergnier 95–7
Moeketski 272, 281, 282 Peters 52, 70, 115, 117, 118, 122, 123
Moisl 182, 194 Peters-Roberts 115, 117, 118, 122, 123
Moon 54, 55, 70 Phillipson 114, 124
Morgan 200, 207 Pinker 114, 124
Morin 225 Pisek 157, 168
Morris 12, 264, 269 Plastow 301, 306
Morrow 288, 296 Pöchhacker ii, 50, 194, 227, 230, 237
Mossop 25 Poltermann 28, 38
Mulisch 221 Popovič 20, 25
Munday 4, 13 Popper 1, 13
Murillo 108, 109, 111 Poyatos 111, 227, 228, 237
Priestly 121
N Priiki 126, 139
Nagel 1, 12 Prüfer 94, 97
Nattinger 52, 54–6, 70 Prunč 95, 97
Neubert 285, 296 Pym 28, 37, 38, 143, 155, 179
Nevanlinna 33, 34, 38
Newman 15, 25 R
Newmark 99, 100, 106, 111, 113, 124 Rahman 276, 281, 283
Nida 51, 70 Rampton 114, 124
Nielsen 74, 81, 155 Resch 189, 194
 Name index 311

Riccardi 90, 97 Stolze v, viii, 39, 41, 45, 50

Richards 70, 99, 100, 111, 288, 296 Strauss 302, 304, 306
Rickheit 42, 50 Suchman 183, 195
Risku vi, x, 181, 182, 185, 189, 194, 195 Suomalainen 30, 38, 139
Robbins 286 Susam-Sarajeva 37, 38
Rose ii, 25, 29, 38, 106, 146 Swan 33, 34
Rozan 251, 252, 256, 260, 261 Swift 37
Rudvin vi, xii, xiii, 271, 272, 282, 283 Syder 52, 54, 55, 70

S T
Saarikoski 35, 38 Taber 51, 70
Sams 217, 218 Taivalkoski 8, 13
Santamaria 223, 225 Tarp 200, 207
Schäffner 182, 194, 195 Teich 84, 97
Schmid 60, 70 Thiéry 251, 261
Schmitt 70, 138, 139 Thome 89, 97
Schnell 130 Tirkkonen-Condit ii, 8, 13, 51, 71
Schoppe 210–2, 216–8 Titmarsh 146, 155
Schreiber v, ix, 83, 85, 87–89, 95, 97 Tommola 241, 246, 249
Schreuder 56, 70 Tournay 200
Scollon 279, 280, 282, 283 Toury ii, v, vii–viii, 7–10, 13, 15–6, 20,
Seleskovitch 240, 249, 251, 252, 256, 25, 29, 38, 51, 71, 99, 101–2, 104,
261 111, 141, 155, 273
Setton 91, 97 Tse 286, 296
Sham 286, 295 Tymoczko 10, 13, 28, 38, 224, 226
Shannon 286, 296 Tytler 3, 13
Sherwood 286, 295, 296
Shipman 304, 306 V
Shlesinger ii, xii, 169, 179, 227–230, van Dijk 235
237 van Doorslaer vi, xi, 219, 221
Sierra Bravo 130 van Hoof 251, 260, 261
Sinclair 12, 53, 70 Vasquez 264, 269, 296
Singleton 56, 70 Venuti ii, 37, 41, 46, 50, 118, 124,
Skutnabb-Kangas 114, 124 224–6
Slocum 200, 207 Vermeer 143, 155
Smith 162, 179, 306 Vestergaard 79–81
Snell-Hornby ii, vii, xiii, 50–2, 70, 85, Vienne 186, 195
89, 96, 97, 139, 194 Vinay 8, 9, 13
Snelling 90, 97
Solberg 292, 296 W
Somers 200, 207 Wadensjö 264, 269, 273, 282, 283, 286,
Sperber 40, 50 296
St. Jerome 4, 12, 13, 37, 38, 50, 69, 168, Wagner 216, 292, 296
179, 249 Wandruszka 85, 97
Steiner 84, 97, 236 Weinert 54, 71
</TARGET "ni">

312 Name index

Weinstein-Shr 292, 296 Wolfe 108, 111

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
Wong-Fillmore 52, 71
Woodsworth ii, 170, 178
Wußler 182, 195
Z
Zabalbeascoa v, ix, 99, 100, 111, 157,
168
Zima 223, 226
Zmajić 115, 117, 118, 122, 123
<TARGET "ci" DOCINFO AUTHOR ""TITLE "Concept index"SUBJECT "Benjamins Translations Library, Volume 50"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">
Concept index
A causality 10
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 40, 41, 44, 46–7, 62, 142, 189, 204,
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 creative, creativity 49, 52, 58–60, 64,
calque 32, 152
career advancement/development 181,
185–6, 192
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,
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
100, 159, 176, 188, 209, 214–6
critical reading 301
cross-cultural 271, 273–5, 278, 280–1
314 Concept index

cultural brokering 286 explanatory hypotheses 2, 3, 28

cultural difference 271, 279 explicitation 5, 7, 8, 10–1, 18, 19, 59,
cultural referent 278 62

D F
dead metaphor 56 fair hearing 264
décalage 88, 91 faithful 4, 7, 31, 33–4, 36, 43
decision-making 58, 240, 298 false friends 150
deforming tendencies 4–6 feature 2–9, 10, 20, 46–7, 49, 59–61,
delivery, delivery speed 227, 230, 243, 99, 100, 102, 104–5, 107, 144, 150,
245, 248 228–230, 232–3, 235–6, 275, 276,
descriptive 1–3, 5–7, 20, 23–4, 56, 60, 278–9
99–100, 125, 128, 141, 273–5 fidelity 11, 31, 33–4, 255
developing world 272 film comedies 158, 167
dialogical 286 fluency 62, 120, 240
différence d’ordre typologique see foreign language 53, 113, 170, 172,
typological difference 188, 193, 221, 239, 246, 252
direction, directionality 9, 20, foreignization , foreignizing 19, 28,
239–247, 271, 278 32–6, 41
Disambiguierung 88 foreignness 29
discourse field 43–5, 48 formal elaboration 215
discourse modes 273, 276, 280 formal training 286
domesticated, domesticating , further education 191–2
domestication 19, 27–9, 31, 32–4,
36, 41 G
dubbing 157, 160, 162–8 generalization 4, 11, 17, 19–20, 22, 215
dying metaphors 53, 57 globalisation, Globalisierung 222–3,
271
E glosses 32
education 145, 183, 191–2, 266–8, grado de profesionalidad 126, 136–7
281–2, 294 grammatical markers 271, 276
Effort Models 259 group identity 276, 280–1
elaborated translation 62–3, 66–7
electronic basket 202 H
end users 198, 201 habitus 280–1
equivalence, equivalent 1, 5–7, 30, 39, hackneyed expressions 57
47, 49, 54, 87, 93, 103, 106, 141–4, hermeneutic, hermeneutics 2, 39,
161–2, 166–7, 188, 199–201, 266, 41–4, 48–9
273–4 hesitation 229
error identification 149 hierarchy 47, 271–2, 274, 276
ethical issues, ethics 142, 265–8, 273–4 holistic 39, 44–8
ethnic affiliation 279 homonym 153
expectations 125–6, 128–131, 137–8 homophone 160–1
expert interviews 184 hospital 275
explanandum 2 host institution 274

hypothesis, hypotheses 1–10, 23, 27–8,
31, 33–4, 36, 85, 166, 273
I 252
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 60, 113–120, 188, 193, 239, 240,
jargon 51, 53, 57–9, 64
K multilingual, multilingualism 198, 220
kinship 274–6
knowledge base 39, 41–2, 44
knowledge model 203–6
Korpora see corpus
Concept index 315
L
language acquisition 51–6, 58
language combination 239, 246, 248,
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,
242–7
motivated choice 142–3, 149
N
native language/speaker/tongue see
mother language
naturalizations 27
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 285
obligatory shifts 63
omission, omit 30, 59, 101, 105, 144–9,
167, 235
online translation 171
overgeneralization, over-
generalization 4, 5, 17, 255, 304
overstress 235
P public service interpreting 271–4, 282
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,
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
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

retranslation hypothesis 8, 10, 27–9,
31–7
role 264–5, 268
routine, routinization 49, 51, 54, 57–9,
61, 63–4
S modelling strategy 206
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
Concept index 317
cognitive strategy 40
Dolmetschstrategie 90, 92
domesticating strategy 33, 36
foreignizing strategy 36
interpreting strategy see
Dolmetschstrategie 264
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
</TARGET "ci">

318 Concept index

text analysis 42, 44, 182–3 unit 55, 101–2, 104, 107
text function 47–9, 104 composite unit 55
text linguistics 42 empty unit 104
text production 39, 40, 46–7, 49, 182, lexical unit 108
189, 194 linguistic unit 102
textual fit 1, 6, 7 multi-word unit 52–6
Thema-Rhema-Gliederung 87–90 note unit 253–5, 258
time pressure 21, 59, 60–4 orthographic unit 54
traditional cultures 275 phraseological unit 52, 56
training 185–7, 190, 274, 281, 286, 291, pragmatic unit 55
299 ready-made unit 53–9, 60–4
interpreter training 140, 229, semantic unit 54–5
239, 240–2, 246–7, 253–4, semiotic unit 107
260, 266, 272–4, 286 source-text unit 258
translator training 11, 51, 61, textual unit 102
142, 172–4 tone unit 231
user training 187 unit of information 230, 233–5
training in research 304 unit of language 64
translation competence 181 universality 10, 18, 20, 21, 24
translation criticism 1, 142 university translation 169, 171–3, 177
translation curriculae 193 unmarkedness 29
translation evaluation 61 usability 181–2, 184, 189, 191–3
translation tools 59, 62, 176, 190, 198 user groups 197, 199, 201–2, 204
translationese 51, 59–61, 64
translator motivation 149 V
transpositions 62, 73, 75 verbal false limbs 53, 56–7
truth-value 275 visual presentation see presentación
TT-oriented 142 visual
turn-taking 271, 278–9
type-within-type structure 110 W
typological difference 74, 77 wordplay 157–8, 160, 162, 164, 166–8
Wortstellung 84, 90
U writing skills 190, 304
Übersetzungsprobleme 85–7
In the series Benjamins Translation Library the following titles have been published thus
far or are scheduled for publication:

4 TOURY, Gideon: Descriptive Translation Studies – and beyond. 1995. viii, 312 pp.
5 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.
6 EDWARDS, Alicia B.: The Practice of Court Interpreting. 1995. xiii, 192 pp.
7 BEAUGRANDE, Robert de, Abdullah SHUNNAQ and Mohamed Helmy HELIEL (eds.): Language,
Discourse and Translation in the West and Middle East. 1994. xii, 256 pp.
8 GILE, Daniel: Basic Concepts and Models for Interpreter and Translator Training. 1995. xvi, 278 pp.
9 REY, Alain: Essays on Terminology. Translated by Juan C. Sager. With an introduction by Bruno de
Bessé. 1995. xiv, 223 pp.
10 KUSSMAUL, Paul: Training the Translator. 1995. x, 178 pp.
11 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.
12 BERGENHOLTZ, Henning and Sven TARP (eds.): Manual of Specialised Lexicography. The
preparation of specialised dictionaries. 1995. 256 pp.
13 DELISLE, Jean and Judith WOODSWORTH (eds.): Translators through History. 1995. xvi, 346 pp.
14 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.
15 WILSS, Wolfram: Knowledge and Skills in Translator Behavior. 1996. xiii, 259 pp.
16 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.
17 POYATOS, Fernando (ed.): Nonverbal Communication and Translation. New perspectives and
challenges in literature, interpretation and the media. 1997. xii, 361 pp.
18 SOMERS, Harold (ed.): Terminology, LSP and Translation. Studies in language engineering in honour
of Juan C. Sager. 1996. xii, 250 pp.
19 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.
20 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.
21 BUSH, Peter and Kirsten MALMKJÆR (eds.): Rimbaud's Rainbow. Literary translation in higher
education. 1998. x, 200 pp.
22 CHESTERMAN, Andrew: Memes of Translation. The spread of ideas in translation theory. 1997.
vii, 219 pp.
23 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.
24 ORERO, Pilar and Juan C. SAGER (eds.): The Translator's Dialogue. Giovanni Pontiero. 1997.
xiv, 252 pp.
25 POLLARD, David E. (ed.): Translation and Creation. Readings of Western Literature in Early Modern
China, 1840–1918. 1998. vi, 336 pp.
26 TROSBORG, Anna (ed.): Text Typology and Translation. 1997. xvi, 342 pp.
27 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.
28 SETTON, Robin: Simultaneous Interpretation. A cognitive-pragmatic analysis. 1999. xvi, 397 pp.
29 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