Issues in Prolongational Theory: A Response to Larson

Abstract

ions and Transformations Citing Cohn and Dempster (1992), Larson distinguishes between representational and inclusional hierarchies. In the former, the leaves of the tree persist as leaves at underlying levels; in the latter, nonsurface structures are abstractions from the leaves. As he points out, the hierarchies in GTTM are representational. Thus a 13 chord at the musical surface persists in being a 13 chord at the underlying levels for which it continues to function. In contrast, underlying events in Schenkerian theory are in some sense abstractions, even though they are represented by musical notation. It is easy to say that underlying events are somehow abstract, and it is only somewhat harder to discuss how events might be abstract. The really difficult task is to develop a formal theory that assigns specific abstract structure to underlying events, and in a way that corresponds to the listener's presumed internal representations. Jackendoff and I chose to avoid this morass. GTTM is concerned with derivation of prolongational structure rather than with underlying abstractions of already assigned prolongations. Our strategy reflects developments in linguistic theory, Larson's asseverations to the contrary. Early generative grammar (Chomsky 1957) relied heavily on transformations, but by the 1970s the difficulty in constraining the application of became clear, and alternative formalisms were developed, some of which resemble GTTM's preference rule system (see the second Preface to the 1996 printing of GTTM). Larson conflates phrase-structure rules (in the technical linguistic sense) with two kinds of transformational rules, those that reorder elements in a string and those that alter the internal content of an element. Early generative grammar employed phrase structure rules, specifically the hierarchical rewriting of syntactic categories, and reordering transformational rules, such as the one that converts active into passive sentences. Recent phonological theory is often concerned with underlying cyclic derivations that transform the inner sound structure of words (Halle and Vergnaud 1987). It is interesting to catalogue Schenkerian transformations accord-