Essays on the Representational and Derivational Nature of Grammar: The Diversity of Wh-constructions (review)

Abstract

Reviewed by: Essays on the representational and derivational nature of grammar: The diversity of wh-constructions by Joseph Aoun and Yen-hui Audrey Li Jamal Ouhalla Essays on the representational and derivational nature of grammar: The diversity of wh-constructions. By Joseph Aoun and Yen-hui Audrey Li. Cambridge, MA: MIT Press, 2003. Pp. xi, 289. ISBN 0262511320. $30. This book is an attempt to identify the nature and range of diversity of wh-constructions and to explore their implications for current theoretical issues, in particular the representation versus derivation debate. Although wh-movement (or operator movement) is involved in more constructions, as is well known since at least Chomsky 1977, the discussion is restricted to wh-questions and relatives. However, the need to bring out all of the major patterns of diversity meant having to investigate these constructions in at least three languages: Chinese, English, and Lebanese Arabic. The complexity that arises from involving three languages is skillfully dealt with in a clear style of writing and presentation that makes frequent use of reminders, summaries, detailed lists of properties, and other helpful devices. The main theoretical conclusion reached is that grammar should incorporate both derivational and representational mechanisms. In other words, chains can be formed derivationally in terms of Move or representationally in terms of Match. The risk of redundancy, often invoked as a major argument in favor of an exclusively derivational or representational model (see Brody 1995), is partly addressed with an attempt to demonstrate that the two mechanisms are independent of each other: Move(ment) is a bottom-up process while Match is a top-down process. The book is divided into two parts. Part 1 deals with wh-interrogatives and includes three chapters on superiority and movement, superiority and the minimal match condition, and superiority and interpretation. The first two chapters outline evidence against reducing superiority to movement and in favor of a representational approach in terms of the minimal match condition (MMC). The third chapter, which focuses on the interpretation of wh-phrases, shows that although an approach in terms of MMC is desirable, it still does not account for all available interpretations, of which three are identified: functional, distributive, and pair-list. A proper account of all three readings would need to appeal to the internal constituent structure of wh-phrases. To be precise, if wh-interrogatives are assumed to have a tripartite structure that consists of question, quantification, and restriction, the three readings can be shown to be a function of the distribution of the three constituents. For example, the availability or not of the pair-list reading reduces to whether quantification is base-generated along with restriction or not. More generally, variation can operate on the basis of whether question, quantification, and restriction are all base-generated together within a nominal projection or separately from each other. In English and Lebanese Arabic, all three parts are base-generated together. In Japanese, only quantification and restriction are base-generated together, with question located higher in the structure. In Chinese, restriction alone is base-generated inside the nominal projection, with both quantification and question located higher in the structure. In addition, variation can operate along a different dimension having to do with wh-movement: whether the three constituents move together as one unit (wh-movement in English and Lebanese Arabic), whether quantification alone moves (Japanese), and whether all three constituents remain in situ (wh-in-situ in Lebanese Arabic and English). The classification allows for the possibility that a language may lack anything corresponding to wh-movement, which is apparently the situation in Chinese. Part 2 deals with relatives and consists of four chapters on head-initial relative constructions, head-final relative constructions, adjunction structure and derivation, and typology of relativization. The most surprising aspect of this part is that it proposes and defends radically different structures for head-initial and head-final relatives. Head-initial relatives have a complementation structure where the relative clause is the complement of D. The head-initial order is derived in one of two ways depending on whether the head moves to the initial position from the relativized position (the promotion analysis) or is base-generated in the...