Lexical representation is one of the `hottest' issues in current debate, both from the point of view of the lexicon as a theory of the smallest units of language to modern lexicographic databases with relational, object-oriented or hypertext structures.
On the one hand, principles of representation need to be sufficiently well-defined to permit computational implementations. On the other hand, lexical information of different kinds tends to require many different kinds of representation, from sequences of orthographic units through phonological lattices of parallel and sequential sound events, to complex syntactic frames and semantic networks. Further, practical considerations such as `typability' at the computer keyboard impose further constraints. Applicability to written and spoken system development is an additional consideration, with issues ranging from the representation of speech signal annotations, phonetic representations, to questions of the integration of lexica with speech databases in spoken language dialogue system development (cf. [Draxler (this volume), ]).
In actual implementations, it is often the case that for reasons of modularity and efficiency each element of the microstructure, i.e. each field of a lexicographic database, representing one property or kind of lexical information, is separated out and realised as a separate `lexicon'. A typical example is the speech synthesis lexicon described in [Quazza & van den Heuvel (this volume)].
In the following sections, some of the more theoretically determined constraints on representation will be discussed.