CORPUS LINGUISTICS AND TRANSLATION STUDIESIMPLICATIONS AND APPLICATIONS.doc

Corpus Linguistics and Translation StudiesImplications and ApplicationsMona BakerCobuild, BirminghamAbstractThe rise of corpus linguistics has serious implications for any discipline in which language plays a major role. This paper explores the impact that the availability of corpora is likely to have on the study of translation as an empirical phenomenon. It argues that the techniques and methodology developed in the field of corpus linguistics will have a direct impact on the emerging discipline of translation studies, particularly with respect to its theoretical and descriptive branches. The nature of this impact is discussed in some detail and brief reference is made to some of the applications of corpus techniques in the applied branch of the discipline.0. Introduction A great deal of our experience of and knowledge about other cultures is mediated through various forms of translation, including written translations, sub-titling, dubbing, and various types of interpreting activities. The most obvious case in point is perhaps literature. Most of us know writers such as Ibsen, Dostoyevsky and Borges only through translated versions of their works. But our reliance on translation does not stop here. Our understanding of political issues, of art, and of various other areas which are central to our lives is no less dependent on translation than our understanding of world literature.Given that translated texts play such an important role in shaping our experience of life and our view of the world, it is difficult to understand why translation has traditionally been viewed as a second-rate activity, not worthy of serious academic enquiry, and why translated texts have been regarded as no more than second-hand and distorted versions of real texts. If they are to be studied at all, these second-hand texts are traditionally analysed with the (233) sole purpose of proving that they inevitably fall short of reproducing all the glory of the original. A striking proof of the low status accorded to translated texts comes from the young but by now well-established field of corpus linguistics. A recent survey commissioned by the Network of European Reference Corpora, an EEC-funded project, shows that many corpus builders in Europe specifically exclude translated text from their corpora.1 This is presumably done on the grounds that translated texts are not representative and that they might distort our view of the real language under investigation. It is perhaps justifiable to exclude translated texts which are produced by non-native speakers of the language in question, but what justification can there be for excluding translations produced by native speakers, other than that translated texts per se are thought to be somehow inferior or contrived? Biased as it may be, this traditional view of translation implies, in itself, an acknowledgement of the fact that translational behavior is different from other types of linguistic behavior, quite irrespective of the translators mastery of the target language.The starting point of this paper is that translated texts record genuine communicative events and as such are neither inferior nor superior to other communicative events in any language. They are however different, and the nature of this difference needs to be explored and recorded. Moreover, translation should be taken seriously by related disciplines such as linguistics, literary theory and cultural and communication studies, not least because these disciplines can benefit from the results of research carried out in the field of translation. At the same time, as a phenomenon which pervades almost every aspect of our lives and shapes our understanding of the world, the study of translation can hardly be relegated to the periphery of other disciplines and sub-disciplines, those listed above being no exception. What is needed is an academic discipline which takes the phenomenon of translation as its main object of study. For many scholars, this discipline now exists. Some refer to it as the science of translation, other as translatology, but the most common term used today is translation studies.Eco (1976:7) distinguishes between a discipline and a field of study. The first has “its own method and a precise object” (my emphasis). The second has “a repertoire of interests that is not as yet completely unified”. It could be argued that translation studies is still largely a “field of study” in Ecos terms. The vast majority of research carried out in this, shall we say emerging discipline, is still concerned exclusively with the relationship between specific source and target texts, rather than with the nature of translated text as such. This relationship is generally investigated using notions such as equivalence, (234) correspondence, and shifts of translation, which betray a preoccupation with practical issues such as the training of translators. More important, the central role that these notions assume in the literature points to a general failure on the part of the theoretical branch of the discipline to define its object of study and to account for it. Instead of exploring features of translated texts as our object of study, we are still trying either to justify them or dismiss them by reference to their originals.It is my belief that the time is now ripe for a major redefinition of the scope and aims of translation studies, and that we are about to witness a turning point in the history of the discipline. I would like to argue that this turning point will come as a direct consequence of access to large corpora of both original and translated texts, and of the development of specific methods and tools for interrogating such corpora in ways which are appropriate to the needs of translation scholars. Large corpora will provide theorists of translation with a unique opportunity to observe the object of their study and to explore what it is that makes it different from other objects of study, such as language in general or indeed any other kind of cultural interaction. It will also allow us to explore, on a larger scale than was ever possible before, the principles that govern translational behavior and the constraints under which it operates. Therein lie the two goals of any theoretical enquiry: to define its object of study and to account for it.Section 1 below offers an overview of the emerging discipline of translation studies and explains why translation scholars are now in a position to use the insights gained from corpus linguistics, and some of the techniques developed by it, to take translation across the threshold of field of study and into the realm of fully-fledged disciplines.1. Translation studies: the state of the art1.1 Central issues: the status of the source text and the notion of equivalenceUntil very recently, two assumptions dominated all discussions of translation and were never questioned in the literature. The first is that of the primacy of the source text, entailing a requirement for accuracy and faithfulness on the part of the translator. The second is a consequence of the first and is embodied in the notion of equivalence which has been the central concern of all discourse on translation since time immemorial. Translations should strive to be as equivalent to their originals as possible, with equivalence being understood, (235) mainly as a semantic or formal category. The implied aim of all studies on translation was never to establish what translation itself is, as a phenomenon, but rather to determine what an ideal translation, as an instance, should strive to be in order to minimise its inevitable distortion of the message, the spirit, and the elegance of the original.The essentialist question of how equivalence per se might be established in the course of translation has gradually been tempered by experience and by an explosion in the amount and range of texts which have come to be translated in a variety of ways on a regular basis. Hence, we now have a massive amount of literature which attempts to classify the notion of equivalence in a multitude of ways, and the question is no longer how equivalence might be achieved but, increasingly, what kind of equivalence can be achieved and in what contexts. This in itself is a noticeable improvement on the traditionally static view of equivalence, but it still assumes the primacy of the source text and it still implies that a translation is merely a text striving to meet the standards of another text.1.2 Developments which support a move towards corpus-based researchThe attempt to extend and classify the notion of equivalence has brought with it a need to explore not only the source text as the modal to be adhered to but also the target language, and the specific target language text type, in order to give meaning to such categories as stylistic equivalence and functional equivalence. If the idea is not simply to reproduce the formal structures of the source text but also to give some thought, and sometimes priority, to how similar meanings and functions are typically expressed in the target language, then the need to study authentic instances of similar discourse in the two languages becomes obvious.There have been other developments which have played a more direct role in preparing the ground for corpus work. One such development is the decline of what we might call the semantic view of the relationship between source and target texts. For a long time, discourse on translation was dominated by the idea that meaning, or messages, exist as such and can, indeed should, be transferred from source to target texts in much the same way as one might transfer wine from one glass to another. The traditional dichotomy of translating word-for-word or sense-for-sense is a product of this view of meaning. At about the same time that the notion of equivalence began to be reassessed, or perhaps a little earlier, new ideas began to develop about the nature of meaning in translation. Firth (1968:91) was among the first to sug(236)gest that, difficult though as it may appear, an approach which connects structures and systems of language to structures and systems in the context of situation (as opposed to structures and systems of thought) is more manageable and “more easily related to problems of translation”. Similarly, Haas (1986:104) stresses that, in practice, correspondence in meaning amounts to correspondence in use and asserts that “unless we can succeed in thus explaining translation, the mystery of bare and neutral fact will continue to haunt us”. Two expressions are equivalent in meaning if and only if “there is a correspondence between their uses” (ibid). The importance of this change in orientation, from a conceptual to a situational perspective and from meaning to usage, is that it supports the push towards descriptive studies in general and corpus-based studies in particular. Conceptual and semantic studies (in the traditional sense) can be based on introspection. Studies which take the context into consideration, and even more so, studies which attempt to investigate usage, are, by definition, only feasible if access is available to real data, and, in the case of usage, to substantial amounts of it.Apart from the decline of the semantic view of translation, another, and very exciting, development has been the emergence of approaches which undermine both the status of the source text vis-à-vis the translated text and the value of the very notion of equivalence, particularly if seen as a static relationship between the source and target texts. The move away from source texts and equivalence is instrumental in preparing the ground for corpus work because it enables the discipline to shed its longstanding obsession with the idea of studying individual instances in isolation (one translation compared to one source text at a time) and creates a requirement which can find fulfillment in corpus work, namely the study of large numbers of texts of the same type. This is precisely where corpus work comes into its own.1.2.1 New perspectives: polysystem theoryIn the late seventies, Even-Zohar, a Tel-Aviv scholar, began to develop a theory of literature as a polyststem, that is as a hierarchical and dynamic conglomerate of system rather than a disparate and static collection of texts. A given literary polysystem is seen as part of a larger cultural polysystem, itself consisting of various polysystems besides literature, for example politics and religion. These polysystems are structured differently in different cultures. Polysystem theory has far-reaching implications for the status of translated literature in general and for the status of the source text vis-à-vis the target text in particular. First, the theory assumes a high level of inter-dependence among the various systems and sub-systems which underlie a (237) given polysystem, as well as among the polysystems of literature in various cultures. This means that, for instance, “literature for children would not be considered a phenomenon sui generis, but related to literature for adults” and, similarly, “translated literature would not be disconnected from original literature” (Even-Zohar 1979:13). As a consequence, the status of translated literature is elevated to the point where it becomes worthy of investigation as a system in its own right, interacting with its co-systems and with the literary polysystems of other cultures. By recognising translated literature as a system in its own right, polysystemists shifted the attention away from individual literary translations as the object of literary studies to the study of a large body of translated literature in order to establish its systemic features.Second, one of the main properties of the polysystem is that there is constant struggle among its various strata, with individual elements and systems either being driven from the centre to the periphery or pushing their way towards the centre and possibly occupying it for a period of time (ibid:14). This constant state of flux suggests that no literary system or sub-system is restricted to the periphery by virtue of any inherent limitations on its value. Thus, the approach stresses that translated literature may, and sometimes does, occupy a central position in the polysystem and is therefore capable of providing canonised models for the whole polysystem. Moreover, given that polysystem theory recognises that intra- and inter-relations exist within both systems and polysystems, leading to various types of interference and transfer of elements, models, canons, and so on, it becomes obvious that “semiliterary texts, translated literature, childrens literatureall those strata neglected in current literary studiesare indispensable objects of study for an adequate understanding of how and why transfers occur within systems as well as amon