In contemporary linguistics, the term feature is used in several ways. Two main uses can be distinguished:
- The term feature is sometimes used for a specific dimension of classification along which a given linguistic item is categorized. The specification of this dimension is then called a value. For example, the English noun bread carries the value '+' for the feature [countability].
- More generally, the term feature is also used as an equivalent of 'property', esp. when used for the classification of linguistic items. For example, the English phoneme /d/ can be said to have the feature [voiced] (i.e. the property of being voiced).
Features in Phonology
In phonology, a feature can be regarded as a property of a sound segment. Segments can be considered to be composed of more elementary characteristics, i.e. a finite set of features with (preferably) a phonetic correlate.
Distinctive features are used to explain that phonological rules apply to natural classes of sounds, i.e. sounds which share certain (phonetic) properties. Features were introduced into phonological theory by Trubetzkoy and Jacobson ('the Prague School'); Chomsky & Halle (1968) (SPE, The Sound Pattern of English) proposed a major revision of the theory of distinctive features. In SPE features are considered to be binary, i.e. a feature has two values + (present) or - (absent). For instance, [p] is (among other things) [ -voiced] and [ -nasal] while [m] is [ +voiced] and [ +nasal]. After SPE different feature inventories have been proposed.
Some features have been replaced with structure (for instance [stress] and [syllabic]). Furthermore, the binarity of features is under debate: multi-valued features and single-valued or unary features have been proposed. The development of feature geometry (cf. Clements 1985), in which natural classes are represented by hierarchical structure as well as by features themselves has been a major revision of the theory proposed by Chomsky & Halle (1968). For a detailed summary of various segmental features and their current status, see Keating (1988) and references cited there.
Features in Syntax
The syntactic features encompass lexical and grammatical features. The lexical features '±' and '±' define the four lexical categories (N=[+N,-V]; V=[-N,+V]; A=[+N,+V]; P=[-N,-V]; see also X-bar theory). Among the grammatical features we find features for person, number and gender (so-called 'Phi-features'); the verbal features [±past], [±tense]; and the binding features [±anaphoric] and [±pronominal] introduced in Chomsky (1981).
Features in Lexical Semantics
In lexical (esp. decompositional) semantics, features stand for meaning components of (semantically non-primitive) lexical concepts. They can also be regarded as indicating category membership. For example, the Engl. noun 'boy' can be decomposed into the features [+ male] and [- adult]. Alternatively, we can say that it is contained in both the category corresponding to the feature [+ male] (i.e. the category of male entities) and the category of [- adult] (non-adult) entities.
- Chomsky, N. 1981. Lectures on Government and Binding, Foris, Dordrecht.
- Chomsky, N. and M. Halle 1968. The Sound Pattern of English, Harper and Row, New York.
- Clements,G.N. 1985. The geometry of phonological features,, Phonolgical Yearbook 2,, 225-252
- Jacobson, Fant & Halle 1963. Preliminaries to speech analysis, MIT Press, Cambridge MA.
- Kerstens,J.G. 1993. The Syntax of Number, Person and Gender; A Theory of Phi-features, Linguistic Models 18, Mouton de Gruyter:Berlin/New York.
- Löbner, S. 2003. Understanding Semantics. London: Arnold.
- Muysken & Van Riemsdijk 1986. Features and projections, Foris, Dordrecht.
- Sagey, E.L. 1986. The Representation of Features and Relations in Nonlinear Phonology, PhD diss. MIT, Cambridge, Mass.