lexis (n.)

A term used in LINGUISTICS to refer to the vocabulary of a LANGUAGE, and used adjectivally in a variety of technical phrases. A UNIT of vocabulary is generally referred to as a lexical item, or LEXEME. A complete inventory of the lexical items of a language constitutes that language’s dictionary, or LEXICON – a term particularly used in GENERATIVE GRAMMAR: items are listed ‘in the lexicon’ as a set of lexical entries. The way lexical items are organized in a language is the lexical structure or lexical system. A group of items used to identify the network of contrasts in a specific semantic or lexical field (e.g. cooking, color) may also be called a ‘lexical SYSTEM’. Specific groups of items, sharing certain FORMAL or semantic features, are known as lexical sets. The absence of a lexeme at a specific STRUCTURAL place in a language’s lexical field is called a lexical gap (e.g. brother v. sister, son v. daughter, etc., but no separate lexemes for ‘male’ v. ‘female’ cousin). In comparing languages, it may be said that one language may lexicalize a contrast, whereas another may not – that is, the contrast is identified using lexemes, as in the many terms for the English lexeme ‘hole’ available in some Australian Aboriginal languages. Lexis may be seen in contrast with GRAMMAR, as in the distinction between ‘grammatical WORDS’ and lexical words: the former refers to words whose sole function is to signal grammatical relationships (a role which is claimed for such words as of, to and the in English); the latter refers to words which have lexical meaning, i.e. they have semantic CONTENT. Examples include lexical verbs (v. auxiliary verbs) and lexical noun phrases (v. non-lexical NPs, such as PRO). A similar contrast distinguishes lexical morphology from derivational MORPHOLOGY. HALLIDAYAN linguistics makes a theoretical distinction between grammar and lexis, seen as two subdivisions within linguistic FORM: lexis here is studied with reference to such formal concepts as COLLOCATION, and not in semantic terms. The mutual restriction governing the CO-OCCURRENCE of sets of lexical items is known as lexical selection (e.g. ANIMATE NOUNS being compatible with animate VERBS). Lexical density is a measure of the difficulty of a text, using the ratio of the number of different words (the ‘word types’) to the total number of words (the ‘word tokens’): the ‘type/token ratio’.

In GENERATIVE GRAMMAR, the insertion of particular lexemes at particular places in grammatical structures is carried out by a process of lexical substitution or lexical transformation, using lexical insertion rules. Lexical redundancy rules are used to simplify the specification of lexical entries, e.g. by omitting to specify SUB-CATEGORIZATION FEATURES which may be predicted on the basis of other features. Some generative models also recognize the so-called ‘lexicalist’ hypothesis, in which a class of lexical rules governing word-formation is distinguished from the set of syntactic transformations. Essentially, the hypothesis bans category-changing rules from the grammar – disallowing a verb or adjective from being transformed into a noun, etc. The terms lexical syntax and lexical phonology are also encountered in the generative literature: the former refers to an approach which incorporates syntactic rules within the lexicon; the latter is an approach where some of the PHONOLOGICAL rules are transferred to the lexicon, and integrated with the MORPHOLOGICAL component. In earlier GOVERNMENT-BINDING THEORY, N, V and A (but not P) were lexical categories, as their members were proper governors. In later work, the lexical categories are N, V, A and P, and C and I are non-lexical.