Introduction

The issue of conventionalized language (idioms, formulas, prefabricated patterns, etc.) has long been a problem in linguistic theory, albeit a relatively marginal one. Despite the high frequency and pervasiveness of these forms and the efforts of a few interested linguists, conventionalized language has never really found a clear place in theories of language (Jespersen, 1922; Katz and Postal, 1963; Austin, 1965; Chafe, 1968; Weinreich, 1969, 1972; Searle, 1969; Newmeyer, 1972; Fraser, 1970, 1971, 1976; Mitchell, 1971; Makkai, 1972, 1978; Ferguson, 1976; Bolinger, 1977; Fillmore, 1979, 1983; Wood, 1986; Wierzbicka, 1987). This literature will not be reviewed here; the theoretical background it provides, however, has greatly influenced the research reported here and the work of the other applied linguists that will be mentioned.

Probably because of their interest in language use, various kinds of applied linguists have also dealt with conventionalized language. Their individual interests and foci have been so disparate, however, that their work cannot be said to constitute a coherent background (Whitaker, 1971; Brown, 1973; Clark, 1973; Hakuta, 1974; Wong-Fillmore, 1976; Peters, 1977; Fraser, 1978; Scarcella, 1979; Coulmas, 1979; Yorio, 1980; Ellis, 1984; and others). Neither has recent work in the area of speech act acquisition in a second language, by researchers like L. Beebe, S. Blum-Kulka, C. Candlin, A. Cohen, M. Eisenstein, G. Kasper, E. Olshtain, N. Wolfson, and others, yet crystallized into an integrated theory (Schmidt and Richards, 1980).