The present paper examines the microstructure of entries devoted to prepositions in bilingual English-Polish dictionaries. Given the problems inherent in the lexicographic treatment of function words, it seems worthwhile to ask:
* what are the strategies employed in cases of lack of interlingual equivalence?
* what is the preferred type of sense-structure (i.e. source-language based, target-language based or mixed)?
* does the entry highlight links between related senses?
* how much phraseology is deemed necessary to present the properties of the preposition in question?
* how does the choice of entry organisation affect its usefulness and user-friendliness?
* what possible improvements could be introduced?
Answers to the above questions are believed to have implications for lexicography in general, not merely bilingual lexicography in the English-Polish context.
1. Prepositions as a problem area
As argued in detail in Adamska-Sataciak (2008), prepositions constitute a major problem for lexicographers, grammarians, theoretical linguists, and foreign language learners. Due to their focus on so-called lexical words and their semantically-based microstructure, dictionaries may not be the best place for describing function words (Sinclair 1991: 81). Reference grammars do not fare much better, since, as admitted, e.g., by Quirk and Greenbaum (1973: 143), "it is difficult to describe prepositional meanings systematically". Theoretical linguists tend to agree: "the individual meanings of the prepositions overlap, creating a lexical nightmare for anyone trying to represent prepositional meaning on the basis of semantic contrast and a syntactic nightmare for anyone trying to characterize their occurrence on the basis of lexical meaning or grammatical category alone" (Rice 1992: 90). As if this were not enough, foreign learners have to cope with the fact that different languages conceptualise space in different ways, so that even seemingly equivalent prepositions cease to act as such in contexts other than those involving the most basic spatial configurations. There are also, as is well known, differences at the level of grammatical structure, so that, for example, relations which in one language are expressed by prepositions in another one may be coded by case inflection. All this makes reliance on the native language a poor guide to mastering the behaviour of prepositions in a foreign language.
2. Prepositional polysemy and monolingual dictionaries
There exist small-scale studies of the presentation of prepositions in monolingual dictionaries of English, especially in learners' dictionaries (henceforth MLDs). Swanepoel (1998) notes that prepositions are normally presented in dictionaries as long lists of consecutively numbered senses, which creates the impression that those senses are completely arbitrary and must therefore be learnt by heart. (2) In a recent monograph on the semantics of English prepositions, Tyler and Evans (2003) have called this "the homonymy position". The name is of little importance, for the approach is indistinguishable in practice from what could be dubbed "unrestricted/unconstrained polysemy", a policy most conspicuous in period dictionaries. DOE is a prime example, presenting the prepositions for and from as having more than 100 senses each (Healey 2002: 139). (3) Whatever label we attach to the approach in question, there is little doubt that adopting it makes it difficult to show that the various senses of a prepositional headword may be related.
Coffey (2006) examines the pedagogical usefulness of the treatment of function words (including the prepositions at, by, for, from, in, of on, to, with) in "the big five". (4) He claims that, from the point of view of the advanced learner, detailed polysemic analysis is superfluous, as is information on the basic uses, which such learners are unlikely ever to look up. …