Harris and Chomsky at the Syntax-Semantics Boundary

Abstract

Both Harris and Chomsky formulate questions about language structure in terms of algebraically formalizable tools of description, e.g. transformations. Harris and Chomsky differ on specifying the goals of research and in defining transformations. In Harris’ system, a transformation relates one surface string with another under certain conditions of paraphrase and cooccurrence. In Chomsky’s system, a generative transformation applies to abstract structures that often bear no close resemblance to actual surface strings. In Harris’ system ‘Does T _x change meaning?’ is a focal point of research, and the question ‘Which aspect of the syntactic structure correlates with which aspect of the semantic interpretation?’ makes no sense. In Chomsky’s system, however, the latter question is the focal point of research, while the former makes no sense. It is shown that one can impose an interpretation on the meaningless questions and force an answer, that is, one can use Harrisian ideas to formulate problems in Chomsky’s system and Chomsky’s ideas to formulate questions in Harris’ system. The answers to the forced questions are misleading. An example of a generative grammar with a surface structure interpretive rule is presented to assign readings to sentences containing the elements either/or and not. The conclusion discusses the sense in which Syntactic Structures offered a new conception of explanation in linguistics.

Material support for this research derives from several sources: This work was supported by National Endowment for the Humanities grant RO-7837–73–205. This work was initiated while working under an American Council of Learned Societies grant. Some of the research was performed through (partial) support of the National Science Foundation Institution Grant to New York University. Planet Earth, Inc. provided some clerical assistance.

For comments on this paper thanks must go to many persons, but especially to Joan Bachenko, Roy Byrd, Noam Chomsky, Evelyne Delorme, Mike Fetta, Karen Flynn, David Halitsky, Michael Helke, Justin Lieber, Kathleen Riordan, and Virginia Sterba.

Certain terms are abbreviated as follows: transformational generative grammar, TGG; transformational taxonomic grammar, TIG; phrase marker, PM; phrase structure rule, PS rule; deep structure, DS; surface structure, SS. Therefore, deep structure phrase marker will be DSPM, etc.