Structure and meaning in cognizing

pp. 73-109

in: Charles E. Osgood, Lectures on language performance, Berlin, Springer, 1980

Abstract

The major work for which these Lectures are an "anticipation" will be titled Toward an Abstract Performance Grammar, this being taken from the well-known article by W. C. Watt ("On Two Hypotheses Concerning Psycholinguistics", 1970). He questions whether the deep mental grammar (MG) could be equated with a competence grammar characterized by any linguistic grammar, deriving from the intuitions of linguists about relations among sentence in a language. He then goes on to state that the deep mental grammar must be equated with an abstract performance grammar (APG) which is not characterizable by any purely linguistic grammar. His main argument is that the competence grammar which systematizes intuitions of linguists maximizes one economy criterion (deep syntactic relations between sets of sentences having different surface structures—e.g., x, y, where x are full passives and y truncate passives) and the APG that systematizes language performance maximizes another (" … a premium on economy of derivation of individual sentential paradigms' (p. 187), x and y independently, in order to most efficiently satisfy pragmatic demands of communication). But aside from casual reference to the language performances of ordinary speaker-hearers (particularly children), Watt is quite vague about how an APG is to be characterized (and thus we are left with MG = APG = ?). This book, and these Lectures, will be an attempt to reduce somewhat the "emptiness' symbolized by that ? mark.