The indetermination of the concept of collocation in bilingual lexicography

One of the clear signs of good command in a foreign language is the appropriate use of lexical combinations. The usual term for these constructions is collocation, being impossible on this term to reach a total consensus on the type or types of combination that it covers. A regular user of bilingual dictionaries does not consider what collocation is, however, the analysis of the nature of lexical combinations is very useful for lexicographers, since they must adopt solutions in order to show in the bilingual dictionary the different types of combinations. In this paper a first section deals in a summarized way with the three main conceptions of this term (1. Anglo-Saxon tradition, which is based on frequency; 2. the linguist and phraseologist Melcuk and his followers, which emphasize lexical restriction over semantics; 3. Bosque, for whom most combinations are governed by the restriction) in order to highlight different types of lexical combinations (free or restricted, standard or non-standard, with lexical restriction or with semantic restriction). Next, we deal with the vagueness of the concepts of collocation and free combination in the lexicographical literature, especially as it regards bilingual dictionaries. This analysis shows that there are more or less predictable, more or less systematic or generalizable, more or less restrictive lexical combinations. Based on these theoretical considerations, the lexicographic components that can show the collocational phenomenon in its different forms are specified: the cotext indication for the most predictable and systematic, least restrictive combinations; the example for the most restrictive collocations and those that pose more difficulties from a bilingual perspective; and the infralemma (subentry) for fixed combinations that generally have a canonical equivalent in other languages.