On the treatment of syntax in some generative proposals

This article intends to analyze and revise some aspects of the generative methodology when the latter is to be applied to describe correctly the organization of the syntactic component of a language. When considering a specific problem such as that of passivity, the analysis of the organizational principles of the standard (and extended) theory reveals the heterogeneity of the syntactic phenomena which are supposedly to be analysed by the same principles and an incapability to distinguish explicitly between relations based on meaning and on designation. Hence it is here claimed that the analysis of the same phenomena is made possible only if the general structure of the model be modified to include as underlying level of representation the type of relations proposed by Fillmore in his case grammar, and so be the procedures of categorization used with the adoption of the procedure of «partial matching » which as such does not require a rule or specific category for each and all of the syntactic phenomena. Upon such a modification, the possibility or convenience of a radical opposition between the levels of «competence» and «performace» becomes impracticable, as well as the claim to operativeness of the rules of the syntactic component either outside specific contextual situations or in consideration of other categorial alternatives