Translation Universals: Do They Exist? (Benjamins Translation Library, 48)
By Anna Mauranen, Pekka Kujamaki
* Publisher: John Benjamins Publishing Co
* Number Of Pages: 221
* Publication Date: 2004-04
* ISBN-10 / ASIN: 1588114686
* ISBN-13 / EAN: 9781588114686
Table of contents
Introduction 1
I. Conceptualising universals
Probabilistic explanations in translation studies:Welcome as they
are, would they qualify as universals? 15
Gideon Toury
Beyond the particular 33
Andrew Chesterman
When is a universal not a universal? Some limits of current
corpus-based methodologies for the investigation
of translation universals 51
Silvia Bernardini and Federico Zanettin
II. Large-scale tendencies in translated language
Corpora, universals and interference 65
AnnaMauranen
Untypical frequencies in translated language: A corpus-based study
on a literary corpus of translated and non-translated Finnish 83
Sari Eskola
Untypical patterns in translations: Issues on corpus methodology
and synonymity 101
Jarmo Harri Jantunen
III. Testing the basics
Translation-specific lexicogrammar? Characteristic lexical and
collocational patterning in Swedish texts translated from English 129
Per-Ola Nilsson
Explicitation: A universal of translated text? 143
Vilma Pápai
Explicitation of clausal relations: A corpus-based analysis of clause
connectives in translated and non-translated Finnish children’s literature 165
Tiina Puurtinen
Unique items – over- or under-represented in translated language? 177
Sonja Tirkkonen-Condit
IV. Universals in the translation class
What happens to “unique items” in learners’ translations?
“Theories” and “concepts” as a challenge for novices’ views
on “good translation” 187
Pekka Kujam鋕i
The fate of “The Families of Medellín”: Tampering with a potential
translation universal in the translation class 205
Riitta J滗skel鋓nen
Author index 215
Subject index 219
Introduction
The search for universals of translation has experienced a surge of research interest
since the mid-nineties, in particular since the advent of electronic corpora
as research tools in translation studies. The seminal paper was Mona
Baker’s (1993) article where she suggested that large electronic corpora might
be the ideal tool for investigating the linguistic nature of translations: either
in contrast to their source texts or in contrast to untranslated target language
texts. Baker saw in electronic corpora a useful testbed for a series of hypotheses
on universal features of translation that had been put forward by other
scholars on the basis of small-scale, manually conducted contrastive studies
only. Included in her list were features such as a tendency towards explicitation
(Blum-Kulka 1986; Toury 1991a), disambiguation and simplification (Blum-
Kulka & Levenston 1983; Vanderauwera 1985), growing grammatical conventionality
and a tendency to overrepresent typical features of the target language
(Toury 1980; Vanderauwera 1985; Shlesinger 1991) as well as the feature of
cleaning away repetitions from translations (Shlesinger 1991; Toury 1991b).
Since this article, the idea of linguistic translation universals has found a place
at the centre of discussion in translation studies.
The idea of translation studies searching for general laws and regularities
is not new; the best-known advocate for general laws of translation has been
Gideon Toury (1980, 1995), who proposed this as a fundamental task of
descriptive translation studies. Similarly, more recently Andrew Chesterman
(e.g. 1998, 2000) has wished to see translation studies as a rigorously scientific
pursuit, seeking generalisations like any other science. A clearly linguistic
flavour to the issue has been added by those who have suggested that translated
language is a kind of ‘hybrid language’ (see e.g. Trosborg 1996 and 1997;
Sch鋐fner & Adab 2001), or a ‘third code’ (Frawley 1984).
The issue remains highly controversial: while some scholars (e.g. Laviosa-
Braithwaite 1996) claim that they have found clear support for hypotheses concerning
general linguistic properties of translated language such as simplification,
others (e.g. Tymoczko 1998; Paloposki 2002) maintain that the very
idea of making claims about universals in translation is inconceivable since we
have no way of capturing translations from all times and all languages. Others,
again, are proposing new subtypes of universals (Chesterman 2001), questioning
or further developing already established concepts, (e.g. Toury 2001,
Klaudy 2001) or wondering if the term was felicitous after all (Baker 2001).
The discussion is very much alive, and to fuel it further, we are now rapidly
accumulating evidence from actual data which demands interpretation.
In linguistics, universals have been discussed for quite a while, and it
has become clear that a fruitful study of language universals needs to take
into account several different kinds, including important tendencies shared
by many languages, not only ‘absolute’ universals, or, as Greenberg et al.
(1966) put it in their classic ‘Memorandum concerning language universals’:
“Language universals are by their very nature summary statements about
characteristics or tendencies shared by all human speakers.” Such an extended
view – which includes tendencies – also seems to suit translation studies.
Moreover, distinctions between universals which can be traced back to general
cognitive capacities in humans, and those which relate linguistic structures
and the functional uses of languages (see, Comrie 2003) provide food for
thought for the study of translations and characteristics of translated language
as well. We may want to differentiate our search for that which is most
general first of all in cognitive translation processes, secondly, the social
and historical determinants of translation, and finally, the typical linguistic
features of translations. However, the greatest part of empirical investigation
into translation universals has so far focused on linguistic characteristics –
while theoretical discussion has concerned the plausibility, kinds and possible
determinants of universal tendencies. There is a need to clarify the issues and
also to bring together these angles, to the extent that it is possible.
Clearly, the quest for translation universals is meaningful only if the
data and methods we employ are adequate for the purpose. The value of
universals in deepening our understanding of translation lies in developing
theory and accumulating evidence from all the three main domains that
are relevant to universals: cognitive, social, and linguistic. There is therefore
no reason to subscribe to any methodological monism, even though the
impetus for systematic linguistic research of translation universals originated
in corpus studies. There are good reasons to expect corpusmethods to make an
important contribution to the field in that they allow comparisons of linguistic
features on a large scale; this goes both for the more traditional approach
of comparing translations with their source texts (parallel corpora) and the
more recent discovery of the potential in comparing translations to similar
texts written originally in the target language (comparable corpora). One of
the main methodological principles in an ambitious domain like this is to keep
in mind the diversity of languages, and not draw excessively hasty conclusions
on the basis of comparing typologically very close languages only, or a very
small range of languages.
The present volume is a selection of articles from an international conferencewith
the same topic as the book, “TranslationUniversals – Do They Exist?”
held in Savonlinna, October 2001, on questions relating to translation universals.
Despite the uniformfocus on the topic, it comprises a number of different
approaches from theoretical discussion of the issues to empirical studies testing
some of the main hypotheses put forth so far. The research field is still
very new, as empirical work only seriously began in the late nineties. Several
papers discuss the established hypotheses on universals in the light of recent
work in different languages, and some move on to test new hypotheses that
have emerged out of the research carried out in the last two or three years.
One of the central issues is the role of interference in relation to translation
universals, and a number of suggestions are made as to its position, based on
various empirical approaches. Most studies report work based on large translational
corpora, which have begun to appear in many languages now, with
applications to translator education also included. The papers cover a number
of source and target languages, which makes a welcome change in the heavily
English-dominated field.
The volume is divided into four main sections, according to the main foci
of the papers. Those in the first section, Conceptualising Universals, address
issues concerning the notion of universals and universality, and the extent to
which this is appropriate or fruitful as an avenue for translation studies to take.
The first two articles, by Gideon Toury and Andrew Chesterman, discuss the
concept of universals, reflecting upon the possibility, and indeed desirability,
of discovering them in translations. Both stress the demanding nature of the
enterprise, and the methodological difficulties involved. Nevertheless, both
also see the search for universals as an important step forward for translation
studies, particularly as regards the character and credibility of translation
studies as a ‘science’. Moreover, both welcome corpus-based work as a major
road towards progress in the field, while neither is actively personally involved
in corpus-based studies. Gideon Toury’s opening article discusses the roles of
different levels of abstraction in discovering regularity, and posits probabilistic
statements at the highest level of generality.He then raises the questionwhether
probabilistic propositions, or conditioned regularities, are the best we can hope
for in descriptive translation studies, and if this is so, are these the universals we
have been looking for. The value of the concept of universals for Toury lies not
in the possible existence of such laws, but in their explanatory power, which, at
least for the time being, shows great promise. Toury prefers the term ‘laws’
to ‘universals’, but concedes to talk about universals in the present context,
without too much concern.
Andrew Chesterman continues the thread of thought stemming from the
quest for generalities which characterises all science. He considers the different
ways in which translation studies have sought the general, distinguishing what
he calls the prescriptive route, the pejorative route, and the descriptive route.
The contributions and problems of each are discussed, with the main focus
on the currently prevailing descriptive study of translation, where further
distinctions are made, such as the very useful one between universals which
relate to the process from the source to the target text (what he calls Suniversals),
and those which compare translations to other target language
texts (T-universals). He raises many other fundamental questions, relating to
the nature of evidence, the concept of tendency, and the problem of testing
very high-level hypotheses, questioning for each the current conceptualisation
and terminology, which, not surprisingly, tend to vary widely and often suffer
from vagueness. Finally, Chesterman invites us to go beyond descriptions,
to explanations, and consider questions of causes as well as effects. He calls
for wider testing of hypotheses, standardisation and operationalisation of
concepts, and generation of new hypotheses.
The final paper in this section, by Silvia Bernardini and Federico Zanettin,
assesses the appropriateness of corpus-based approaches to the search for universals.
Terminologically, they share Toury’s preference for ‘law’ over ‘universal’,
although for different reasons (its better fit into the framework of Firthian
linguistics). They address the issue of corpus design in view of the claims that
have been made for their ability to offer a testbed for translational hypotheses
at the highest levels of generalisation. The discussion is filtered through
an illustrative case, the compilation of an English-Italian translational corpus,
which is of the parallel corpus type, and bidirectional. The organising concept
is Toury’s “preliminary norms”, that is, the translation policies which largely
determine things like the selection of texts for translation. A survey of texts
that are available in translation quickly reveals that a considerable asymmetry
prevails between languages as regards the proportions of genres. Sheer overall
numbers show that for a given language pair, more gets translated in one direction
than the other. In addition, translations in one direction are likely to be
differently biased for prestige, date of original, and other social determinants.
The dilemma that follows is that comparability of the texts conflicts with the
objective of reflecting the prevailing preliminary norms, although an ambi
tious corpus would wish to incorporate both criteria. The suggested solution is
a broad-based methodological approach to translational corpus compilation,
which gives due recognition to the social contexts that translations reside in.
The second section, Large-Scale Tendencies in Translated Language, is
methodologically fairly uniform in that each paper reports a corpus-based
study, and addresses questions of capturing universals with this approach.
Moreover, they all make use of the same corpus, the Corpus of Translated
Finnish, a comparable corpus which consists of 10 million words altogether,
consisting of both translations into Finnish in several genres, and comparable
texts originally written in Finnish. The texts are contemporary, and include
translations from a number of different source languages. The corpus, which
is one of the largest comparable corpora in existence, was compiled at the
Savonlinna School of Translation Studies in the last five years of the 1990s. The
first of the papers, by AnnaMauranen,who was the initiator and director of the
Savonlinna project, gives an account of the structure and origins of the corpus.
Her paper sets out by considering the problem of interference in translation,
which has been used rather carelessly and given diverse interpretations, and
then moves on to explaining and trying out a procedure for comparing
different corpora in search for evidence on the role of interference and transfer.
A corpus comparison on an overall basis is problematic; the present solution is
based on lexis and rank order, and it obviously needs other types of evidence
to support or refute the findings.Nevertheless, the method yields results which
suggest that translations are more similar to one another than to originals
in the target language, but that translations from particular source languages
and cultures differ from each other in their distance from the target language
texts. This suggests that interference is a fundamental property of translations,
but that not all linguistic features specific to translations are reducible to
interference – other sources are required to explain the rest of the distance
between translations and non-translations on the one hand, and the proximity
of translations to one another.
The topic of interference is followed on by Sari Eskola, who advocates a new
reading to the concept of interference as a neutral, non-pejorative term. Her
theoretical interest is also in clarifying the concepts of ‘norm’ and ‘universal’
with respect to regularities, and she suggests the common term for observed
regularities should be Toury’s ‘law’, with a distinction being made between
local and global laws, the latter representing universals. She has investigated the
syntax of texts translated into Finnish in comparison with originally Finnish
texts, with Russian and English as source languages, both typologically very
distant from Finnish. Her particular focus is on non-finite constructions,
which, on the face of it, could be assumed to be typical of translations in that
they offer convenient ways of overcoming syntactic differences in the source
and target languages. Her findings indicate that translations, compared with
original TL texts, overrepresented those SL features which had straightforward
translation equivalents in the TL, but, conversely, underrepresented features
which were specific to the TL. This supports Tirkkonen-Condit’s (2000, this
volume) hypothesis on the relative underrepresentation of unique items, and
alsoMauranen’s (2000) findings on word combinations. Since the latter studies
were based on lexis, Eskola’s syntactic results provide an important support.
Eskola’s finding that the differences between translations from Russian and
Finnish originals are greater than between translations from English vis-a-vis
Finnish originals are in line with Mauranen’s lexical results (this volume).
Jarmo Harri Jantunen takes up the methodological issues involved in the
quest for universals with the help of comparable target-language corpora.
His study is also based on a subsection of the Corpus of Translated Finnish
(CTF). His particular focus is on lexical patterning, more specifically nearsynonymous
frequent intensifiers, but the main objective of the paper is to
present a quantitative methodological solution for investigating the influence
of the SL on translations. The three-phase method of comparisons is enabled
by the compilation principles of the CTF, and Jantunen takes pains to explore
the suitability of various statistical measures for discovering meaningful
regularities in the data in a reliable way. His findings are interestingly complex
in that the very small selection of near-synonyms showed different patterning
both in terms of collocations and colligations, and the main conclusion is that
it is imperative to continue fine-tuned research into specific cases to be able to
appreciate the extension of SL influence and other determinants of difference
and similarity in translated and untranslated language.
The third section, Testing the basics, is devoted to papers in which some
basic assumptions on the specificity of translated language are tested with
different parallel and comparable corpora. The section is opened by Per-Ola
Nilsson, who reports on a methodologically rigorous corpus-driven study
of translation-specific lexicogrammar in texts translated from English into
Swedish. The quantitative comparison of original and translated Swedish reveals
that in the translated text corpus, the grammatical word av as well as many
collocational patterns and frameworks including av were significantly overrepresented.
Nilsson uses the fiction part of the English-Swedish Parallel Corpus
(ESPC), which with its aligned subcorpus enables himto move on to the search
for causes for this overrepresentation. The analysis shows a strong structural
correspondence between English sources and Swedish translations: the transfer
of several frequent SL patterns give rise to these frequency differences between
translated and non-translated Swedish................................................................ |
楼主: 帆帆
