Introducing Corpora in Translation Studies

Maeve Olohan

• Publisher: Routledge
• Year: 2004
• ISBN:
  9781134492206
• Edition: 1
• Topic: Filología
• Subtopic: Lingüística

The use of corpora in translation studies, both as a tool for translators and as a way of analyzing the process of translation, is growing. This book provides a much-needed assessment of how the analysis of corpus data can make a contribution to the study of translation.
Introducing Corpora in Translation Studies:

• traces the development of corpus methods within translation studies
• defines the types of corpora used for translation research, discussing their design and application and presenting tools for extracting and analyzing data
• examines research potential and methodological limitatis
• considers some uses of corpora by translators and in translator training
• features research questions, case studies and discussion points to provide a practical guide to using corpora in translation studies.

Offering a comprehensive account of the use of corpora by today's translators and researchers, Introducing Corpora in Translation Studies is the definitive guide to a fast-developing area of study.

1: Introducing translation studies research

1.0 Introduction

This chapter begins with the notion of the translator as a linguistic and cultural mediator who is often not visible to the reader of translations. This point serves to introduce translation studies as a thriving academic discipline that studies the activity of translation and the role of the translator, but that can do so from a range of perspectives and using a variety of methods. Translation studies has often been described in terms of competing paradigms or approaches - linguistics vs. cultural studies, prescriptive vs. descriptive, empirical vs. theoretical, and so on - and a short summary is given here of one recent attempt to reconcile two diverging perspectives, and subsequent reactions to it. It is argued that an understanding of the nature and status of these kinds of oppositions is necessary in order to situate research using corpora in the wider context of translation studies. However, it is also stressed that it is more fruitful to view the use of corpora as a research methodology, with its own strengths and limitations, than to see it as a paradigm occupying one or other pole. Against that backdrop, the comparative model of translation research is introduced as the one that underlies corpus analysis in translation studies.

1.1 Where is the translator?

A recent essay by Ian McEwan, prizewinning English novelist, discusses the role of the relationship between literature and science in the quest to understand the human condition. His essay, published in a British newspaper’s weekend supplement, starts as follows:

Greatness in literature is more intelligible and amenable to most of us than greatness in science. All of us have an idea, our own, or one that has been imposed upon us, of what is meant by a great novelist. Whether it is in a spirit of awe and delight, duty or scepticism, we grasp at first hand, when we read Anna Karenina, or Madame Bovary what people mean when they speak of greatness. We have the privilege of unmediated contact. From the first sentence, we come into a presence, and we can see for ourselves the quality of a particular mind; in a matter of minutes we may read the fruits of a long-forgotten afternoon, an afternoon’s work done in isolation, 150 years ago. And what was once an unfolding personal secret is now ours. Imaginary people appear before us, their historical and domestic circumstances are very particular, their characters equally so. We witness and judge the skill with which they are conjured.

(McEwan2001: 1)

McEwan contrasts this greatness in literature with greatness in science, which is ‘harder to grasp’; scientific work is ‘objectifying, therefore distancing’ and ‘scientific ideas happily float free of their creators’ (ibid.). What is interesting from the point of view of translation studies is that only the tiniest minority of McEwan’s English-speaking readership have had ‘unmediated contact’ with the writers used here as examples; admittedly a small number of the newspaper readers may have read Flaubert in French, but significantly fewer will have experienced Tolstoy in Russian. In the context of access through translation, has the greatness of this literature really been grasped ‘at first hand’? McEwan says that ‘from the first sentence, we come into a presence and we can see for ourselves the quality of a particular mind’. But whose presence is this - author or translator? Whose mind are we seeing ‘for ourselves’ - author or translator? McEwan asserts that ‘we may read the fruits of a long-forgotten afternoon … 150 years ago’, but taking the writing of Madame Bovary in 1857 as an example, when we read it in English translation, are we not also reading the fruits of a period of mental exertion by one of numerous translators who have translated it into English since the first British translation, produced by Eleanor Marx-Aveling, appeared in 1886?¹ Are we experiencing the skill with which the circumstances and characters of imaginary people are conjured by the source author, or by the translator?

McEwan is ultimately interested in the universality of literature, ‘illuminating human nature at precisely the point at which it is most parochial and specific’ (ibid.). There is ‘an unspoken agreement, a kind of contract, between writer and reader’, by which ‘it is assumed that however strange these people are, we will understand them readily enough to be able to appreciate their strangeness’ (ibid.). However, while we will recognize, understand or appreciate those traits of human nature in the characters we encounter in literature, and while we have the ability to place ourselves in their position - in the terminology of cognitive science, to construct alternative cognitive models of the world - we gain access to the strangeness through the familiarity of language produced for us by the translator as linguistic and cultural mediator, unacknowledged by McEwan in this piece.

This lack of consideration or relative invisibility of the translator of literary and other works is not unusual, and has been highlighted by Lawrence Venuti in particular (e.g. Venuti 1995), who attributes it to a combination of attempts by translators to produce fluent, transparent texts and the nature of acceptability judgements by readers of translation, who wish translations to appear as though they were originals (ibid.: 1). Both of these factors can be seen in Peter Fawcett’s (2000) study of literature reviews published in British broadsheet newspapers since 1992. Only a small number of published reviews are of translations and even fewer (a total of eleven for the period under investigation) make any mention of this fact, or consider the role of the translator. A similar trend is observed by Isabelle Vanderschelden (2000), in the same special issue of The Translator, in her study of reviews of foreign literature in French translation. She remarks that, while foreign literature and non-fiction appear to be published, promoted and read in France, the literary reviews of translations do not promote the visibility of translation, since most reviews comment on the translation as though it were a French original (ibid.: 282). A further example is offered by Jeremy Munday (2001: 157-9), who takes a collection of short stories by Garcia Marquez (Doce cuentos peregrinos) that has been translated into English by Edith Grossman {Strange Pilgrims) and examines reviews of that English-language translation in the UK and the USA. He notes that most reviewers do not acknowledge the fact that they are reviewing a translation, but instead comment on it as though the author had written in English. One review, for example, comments on a ‘a characteristic Marquez sentence’ (ibid.: 158). Munday points out that the sentence referred to in this way is actually only part of the original source-text sentence and that circumstantial adjuncts have been reordered; the target-text sentence thus bears little syntactic resemblance to the source-language one.

Of the eleven reviews examined by Fawcett that do, in fact, talk about the translation, most are seen to make disparaging comments, largely without convincing justification, about the nature and/or quality of the translation or translator and thus ‘constitute an exercise in institutionalized irresponsibility’ (Fawcett 2000: 305). Specific features of the reviews, as identified by Fawcett, include a preference for transparent translation and a dislike of source-oriented modes of translation, both corroborating Venuti’s (1995) views on fluency mentioned above. Vanderschelden seems to confirm some of Fawcett’s findings, noting that, where translations are reviewed as translations, the evaluation criteria ‘amount to no more than a set of presuppositions and subjective assumptions which vary from one reviewer to the next’ (Vanderschelden 2000: 288). Thus, when the presence of the translator is recognized, it is often as a result of perceived shortcomings in their work.

Translation studies, as a discipline that studies the phenomenon of translation from diverse angles and perspectives, is, of course, interested in the mediating role of the translator, and the translation process, the translation product, causes and effects of the translation activity and so on. However, as we will see in the next section, the questions of what translation studies focuses on and how it can go about studying its object of investigation are hotly debated issues.

1.2 Theoretical approaches, methodological issues and conflicting viewpoints

In the course of its development, translation studies has distanced itself from earlier, prescriptive approaches, and much research on translation now has a distinctly descriptive focus, concerned with describing what translations and translators actually do, what translations are like, etc. Within this framework, the issue of research models and methodology is important and has attracted increasing attention within translation studies in recent years. This may be seen as part of a general trend in the humanities (Simeoni 2000: 337; Baker 2001: 8) but is also very much a direct result of how translation studies has developed as an academic discipline. The increase in quantity and, arguably, quality of research being carried out is certainly linked to the surge in academic establishments offering translation studies qualifications, including research degrees. This is coupled with the emergence of more and more means of dissemination of research findings, including specialist publishers, journals, specialist publication series, conferences, workshops and seminars, all increasingly international in nature.²

Within much of this work there is a perceptible trend towards more widespread reference to a ‘scientific’ approach to the study of translation. Depending on who uses the term, this can simply mean ‘systematic’ or ‘rigorous’. It might mean ‘more objective’, or ‘less subjective’. For some people, perhaps particularly those who eschew linguistically oriented approaches to translation, it can mean ‘linguistic’; others associate it exclusively with classical approaches to experimental research, hypothesis testing, quantitative methods, etc. Each of these interpretations may be presented as positive or negative.³ When used in a positive sense, the term often reflects a belief that we need to test our unverified assumptions about what translation is and what it does, and why.⁴ Used in a negative sense, on the other hand, it often represents a view that sees this kind of research as positivistic, essentialist, deterministic or contrived.⁵ In this context, essentialism refers to a belief that ‘meanings are objective and stable, that the translator’s job is to find and transfer these and hence to remain as invisible as possible’ (Chesterman and Arrojo 2000: 151) and this is opposed to non-essentialism, which claims that ‘meanings … are inherently non-stable, that they have to be interpreted in each individual instance, and hence that the translator is inevitably visible’ (ibid.).

The debate initiated by Andrew Chesterman and Rosemary Arrojo (2000), in a recent issue of the journal Target, is illustrative of both divergence and convergence of perspectives on research in translation studies. Chesterman and Arrojo looked for points of concurrence between the postmodern cultural studies and textual theories approach to translation on the one hand, and the empirical, descriptive approach on the other, with Chesterman espousing the latter and Arrojo the former. They produced thirty theses to represent their shared ground. It is not possible to provide more than a summary here. The first fourteen theses are to do with defining what translation is; they include general statements on what translation research aims to do, i.e. ‘ to understand the phenomenon of translation, however this is defined and practised’ ( ibid.: 152). Other theses in this category talk about the importance of metaphors of translation and different ways of understanding translation, acknowledging the differences across cultures and languages of the conception of ‘translation’ and recognizing as a valid research aim an examination of the extent to which these conceptions overlap (ibid.: 152-3). A second set of theses, seven in total, refers to causes for translations being the way they are, emphasizing the need for research to provide explanation, recognizing the uniqueness of each translation but also the existence of similarities and patterns of translation behaviour. The remaining nine theses deal with the consequences of translations, advocating that translation scholars study the effects of translations, for particular periods and cultures, including the way in which translations can influence readers’ behaviour and ways of thinking.

This ‘shared ground’ position paper prompted a number of responses. Both Kirsten Malmkjær (2000) and Sandra Halverson (2000) question the philosophical underpinnings of the essentialist vs. non-essentialist dualism proposed by Chesterman and Arrojo as a possible label for the debate as witnessed in translation studies. Malmkjær focuses on the nature of meaning and argues for non-essential-ism and linguistic relativity, i.e. that it is possible ‘both to believe that meaning is always context bound and to understand how it can come to appear that some meanings are more stable than others’ (Malmkjær 2000: 342-3, author’s emphasis). The meaning of an utterance arises through language use and is a function of interaction between people in a communication situation. Halverson (2000) is equally dismissive of the essentialist vs. non-essentialist dichotomy, pointing out the fallaciousness of equating essentialism with empiricism. She observes that both Chesterman and Arrojo encourage empirical research, but that they disagree on the status of this research and its results. Like Malmkjær, Halverson returns to the notion of meaning as relative, taking a cognitive semantic view of meaning as being ‘grounded in our shared cognitive capacities and in the many common ways in which humans experience and interact with the world’ (ibid.: 360-1).

The emphasis on ‘shared ground’ and ‘reconciliation’ in Chesterman and Arrojo is not accepted as such by Anthony Pym (2000a), who sees this debate instead as one of linguistic empiricism vs. deconstruction (very provocatively described by Pym using statistical methods such as z-curves and P-values). He pits Chesterman’s ‘reasoned empiricism’ against Arrojo’s ‘screaming’; she is likened to ‘several hundred deconstructionist school marms’, and displays ‘timid purism’ (ibid.: 335-6). According to Pym, statements about what translation is are unimportant for the discipline but of relevance when they exclude other modes of cross-cultural communication, as do Chesterman and Arrojo. Focusing on cause and effect is ‘a tad pedestrian’; to formulate problems to be solved would be more interesting and challenging (ibid.: 336). Pym’s views expressed elsewhere (Pym 1999, 2000b) about the inadequacy of various dichotomies (cultural studies vs. linguistics, descriptive vs. prescriptive) to represent the recent past of translation studies or to give direction to the future of the discipline are echoed here when he reminds us again about the (lack of) relevance of theoretical deliberations such as these for the issues of intercultural communication that arise in the real world and for expanding our very limited understanding of training and educational aspects of translation.

Edwin Gentzler (2001: 160), also responding to Chesterman and Arrojo’s paper, sees its shortcoming in the fact that it excludes a ‘multitude of theories, methodologies, and discourses being used to discuss translational phenomena around the world today’. He makes an important point about the very nature of the debate: namely, that in striving for consensus, we ‘tend to exclude voices at the periphery … The desire by prominent scholars in the field to focus the debate on the center, to prescribe a methodology for discussion, and to dictate the discourse of the discussion, has limited the field’s growth for too long’ (ibid.: 161).⁶ Gentzler argues for a ‘multiple-model approach, not one that presents only the consensus of scholars, but one that includes the differences’ (ibid.: 163). He believes that our theories of translation should be open to insights from other disciplines and to developments in the role of translators and their activities.

This brief outline of the debate highlights the problems of labels, dichotomies, dualisms and diametrically opposed poles. It could be argued that there will always be people - ‘the crazies in the black tails’ of Pym’s z-curve (Pym 2000a: 335) - who cling to one or other end of the scale. Similarly, there will be people who seek reconciliation and shared ground. In an attempt to move away from the traditional linguistics vs. cultural studies divide, Mona Baker (2001) focuses on the similarities between translation studies and other areas of intellectual endeavour in the humanities. She takes the example of linguistics and its current preoccupations to highlight issues and concerns that are also of relevance to translation studies, with both disciplines seen as part of a broader intellectual movement. Baker summarizes these elements as follows:

• increased interest in the impact of phenomena on human life … rather than on their internal workings or structures
• a marked tendency towards self-reflexivity
• increased recognition of the role of subjectivity in scientific research
• legitimation of scientific research as a form of political action in its own right
• growing concern with methodology, particular with the need to balance subjectivity and objectivity in order to conduct credible academic research
• rejection of neat categorizations and idealised constructs and increased recognition of the complexity inherent in all phenomena
• growing stress on inter/multidisciplinarity […]
• a questioning of normative approaches and increased attention to human agency
• increased interest in questions of ideology and ethics
• increased interest in and respect for qualitative (as opposed to quantitative) methods of research.

(Baker2001: 17-18)

The appearance of some of these points in the ‘shared ground’ debate of translation studies bears witness to their broader relevance and the commonality of elements of intellectual endeavour across disciplinary boundaries.

The relevance of these perspectives for a discussion of corpora in translation studies relates to the fact that, on the one hand, research using corpora is generally seen as linguistic, empirical, quantitative, data-based or data-driven. However, it has also been billed as an approach with the potential to unify opposing perspectives (e.g. Tymoczko 1998). Rather than seek to align the use of corpora with a particular approach, perspective or paradigm within translation studies,⁷ it is more fruitful to focus on corpus work as a research methodology. The use of corpora in translation studies research is thus seen as the application of corpus analysis techniques, both quantitative and qualitative, to the study of aspects of the product and process of translation. As a research method, it may therefore be employed in researching questions and issues in conjunction with a range of theoretical frameworks, assumptions, tools and concepts. Moreover, as a research method introduced relatively recently into translation studies, it requires development, adaptation and refinement, through discussion, application, assessment and recognition of strengths and limitations. This process is now being undertaken by an ever-growing number of scholars in translation studies.

1.3 Ches...