Lexical Functions and the Translation of Collocations

This paper discusses the lexicographical concept of lexical functions (Mel' cuk and Zolkovsky, 1984) and their potential exploitation in the development of a machine translation lexicon designed to handle collocations. We show how lexical functions can be thought to reeect crosslinguistic meaning concepts for collocational structures and their translational equivalents, and therefore suggest themselves as some kind of language-independent semantic primitives from which translation strategies can be developed. 1. Introduction Collocations present speciic problems in translation, both in human and automatic contexts. If we take the construction heavy smoker in English and attempt to translate it into French and German, we nd that a literal translation of heavy yields the wrong result, since the concept expressed by the adjective (something likèto excess') is translated by grand in French and stark in German. The example is a good illustration of the fact that languages diier in how they express the concepts mediated in collocational structures, e.g. a heavy smoker in English is a `large' smoker in French (grand fumeur) and a `strong' smoker in German (starker Raucher). We observe then that in some sense the adjectives stark, grand and heavy are equivalent in the collocational context, but that this is of course not typically the case in other contexts, cf: grande boite, starke Schachtel and heavy box, where the adjectives could hardly be viewed as equivalent. It seems then that adjectives which are not literal translations of one another may share meaning properties speciically in the collocational context. How then can we specify this special equivalence in the machine translation dictionary? The answer seems to lie in addressing the concept which underlies the union of adjective and noun in these three cases, i.e. in-tensiication, and hence establish a single meaning representation for the adjectives which can be viewed as an interlingual pivot for translation. The CEC project ET-10/75 : Collocations and the Lexicalisation of Semantic Operations, is concerned with investigating the lexical functions (LF's) of Mel' cuk (Mel' cuk and Zolkovsky, 1984), as a candidate interlin-gual device for the translation of adjectival and verbal collocates. In this paper we will attempt to provide an overview of the key areas of research entailed by the project, including a characterisation of collocational structures , an evaluation of Lexical Functions and proposals for representation 1