Lexis in Contrast: Corpus-Based Approaches (review)

This volume is composed of papers presented at the conference on Contrastive Linguistics and Translation Studies—Empirical Approaches (Belgium, 1999), plus an introduction by the editors and another contribution (Teubert’s). The introduction provides a useful survey of theoretical and practical contexts in which contrastive lexical study has emerged, the types of corpus evidence available, theoretical and methodological concerns, and directions for further research. Following this introduction, the chapters are grouped into thematic sections.

In the first section (‘Cross-linguistic equivalence’), Raphael Salkie argues for ‘Two types of translational equivalence’ and illustrates these with (just) two cases from English/French and German/English translations. Elena Tognini Bonelli (‘Functionally complete units of meaning across English and Italian’) offers a three-step methodology for the use of parallel and comparable corpora in comparative lexical study and illustrates it by investigating three idioms involving the words case/caso in English/Italian. Bengt Altenberg contrasts ‘Causative constructions in English and Swedish’ in order to investigate differences of usage between native speakers/learners and translations/sources.

In the ‘Contrastive lexical semantics’ section, Åke Viberg uses Swedish få and English get to investigate ‘Polysemy and disambiguation cues across languages’. Lan Chun compares up/down and shang/xia metaphors in Chinese and finds mostly similarities, plus some differences (although one difference is, I believe, based on a misinterpretation of the English data), thus making a case for the universality of some aspects of metaphor. Michel Paillard contrasts the use of hypallage and metonymy in French and English.

The next section concerns ‘Corpus-based bilingual lexicography’. Wolfgang Teubert’s ‘The role of parallel corpora in translation and multilingual lexicography’ argues that conceptual ontologies (widely pursued in service of machine translation) are not actually useful for translation; instead, parallel corpora can serve as ‘repositor[ies] of re-usable translation units’ (203). Victòria Alsina and Janet De C esaris investigate the usefulness of monolingual corpora and dictionaries in preparing bilingual dictionaries and find that corpora could aid in the organization of information in dictionary entries. The contributions by Sylviane Cardey and Peter Greenfield (‘Computerised set expression dictionaries’) and Christine Chodkiewicz, Didier Bourigault, and John Humphries (‘Making a workable glossary out of a specialised corpus’) highlight the importance of linguists’ analytical and intuitive contributions to corpus-based, computer-assisted lexicography.

The final section concerns ‘Translation and parallel concordancing’. Olivier Kraif reviews methodological problems of processing parallel corpora below the sentential level and distinguishes between ‘translation alignment and lexical correspondences’. François Maniez demonstrates various ways in which solutions to translation problems can be at tempted through automated analysis of corpora and the limits of these approaches. Finally, Patrick Corness describes the merits of ‘Multiconcord: A computer tool for cross-linguistic research’.

The individual papers in this volume will appeal to researchers interested in similar lexical phenomena or languages, but as a whole the book is valuable for its many discussions of methodological issues in comparative work with corpora. The sometimes very narrow foci of the papers (often two or three words or constructions) occasionally raise questions about the validity of the broad conclusions made about the nature of language or translation, and many of the authors inappropriately assume that the average reader can understand the relevance of untranslated examples from European languages. The specificity of the research highlights the fact that, in spite of the computational and corpus resources nowav ailable to us, contrastive work is still time- and resource-expensive. Nevertheless, the contributions in this volume prove repeatedly that such work has great practical rewards.