Elżbieta Górska ON THE EVALUATION OF WORD FORMATION COMPETENCE MODELS WITHIN THE LEXICALIST FRAMEWORK. THE STATE OF THE ART REPORT. PART ONE: THE 1970s AND THE EARLY 1980s. As a point of departure for the present discussion, 1 let me take the require- ment of explanatory adequacy placed on a linguistic theory by generative gram- marians, which states that a given theoretical framework must be capable of selecting the model that best represents the intuition of native speakers. To say the least then, the development of any generative theory towards the achievement of the requirement mentioned, is considered necessary. The problem of selecting the best model confronted generative morphology as soon as the theory had started to develop into the lexicalist framework. By "lexicalist framework" I mean here Word Formation (WF) lexicalism. The basic tenet of this approach is, to use Humboldt's terminology, that "energeia" (i.e. WF-processes) rather than "ergon" (a list of stored [and/or exceptional] complex items) is the object of WPstudies in generative grammar. The first model built according to the process approach was that of Aronoff (1976). The author as- sumes that there is a separate WF-component with its own types of regularities which is to be a model of the native speaker's WF-competence. The "energeia" approach to WF was a reaction against an earlier idea of the lexicon of generative grammar (as in Chomsky 1965), which, in fact, followed closely the Bloom- fieldian dictionary - a repository of irregularities or an appendix to the grammar - and which, at the same time did not postulate any concept of a Word Formation Rule (WFR). 2 Yet, having accepted the notion of a WFR, lexicalism faced new categories of problems. And one of them is the topic of the present paper. In the 1970s and the early 1980s several different models of word formation were postulated within the lexicalist framework (cf. e.g. Aronoff 1976, Booij 1977, Allen 1978, Lieber 1981, Strauss 1982, Wolff 1983, Szymanek 1985, and Malicka-Kleparska 1985). Thus, the lexicalist framework of those days should try to solve the problem of how to select the model that would be the best represen- tation of WF intuitions of native speakers. Even a look at two models, namely those of Aronoff (1976) and Malicka- -Kleparska (1985), shows that the state of the art in the WF theory of the 1970s and early 1980s was far from being settled (cf. also the discussion in Górska 67