Corpus-based cognitive semantics: Extended units of meaning and their implications for Translation Studies

Abstract

Traditional lexical semantics focuses on the meaning of individual lexemes. Firth (1957) brought our attention to collocations and the fact that meaning is not isolated in the lexeme. In 1996 Sinclair argued for the existence of extended units of meaning which, as the expression indicates, go beyond the lexeme. In recent years Stubbs (2001b), and other corpus linguists have convincingly shown that meaning is a phraseological phenomenon to a high degree. Corpus searches allow us to study lexemes in their immediate context, study their most frequent collocates and thus help us reveal their semantic preferences (Sinclair 1987, 1996) and semantic prosodies (evaluation) (Louw 1993). Some of the findings confirm intuitions, whereas some make us aware of connotations which we have never before consciously known the existence of. In this article, I shall argue for the application of corpus-based cognitive semantics as a tool for researchers within translation studies (TS) who are particularly interested in revealing evaluative aspects of the units of meaning of source texts and their translations. What may formerly have been described as something intangible like an ‘atmosphere’, now becomes tangible because of the patterns emerging from large numbers of examples. I shall provide empirical examples in various langua ges of such evaluative patterns which are of course not automatically generated but come about as the result of computer-generated concordance lines and thorough manual analysis.