Right Node Raising and Gapping: Interface Conditions on Prosodic Deletion (review)

This revised version of the author’s 1998 dissertation investigates the role that focus and accent play in ellipsis phenomena in coordinated structures in German and proposes a natural view of the syntax-phonology interface that, coupled with the right semantics, allows a unified analysis of right node raising (RNR) and gapping. Hartmann derives the properties of RNR and gapping from the interplay of theories of the informational structure and of the tonal structure of a sentence. In Ch. 1, ‘Basic assumptions’ (1–20), she outlines the syntax, semantics, and phonology relevant in this study. Working within a (pre-/non-minimalist) version of the principles and parameters theory, she adopts the framework of alternative semantics for a semantic theory of focus, and prosodic phonology to help with the organization of the syntax-phonology interface.

Ch. 2, ‘Coordinated structures’ (21–52), offers a discussion of a number of issues relating to coordination, in particular the most appropriate structure. With a battery of tests at her disposal, H decides on the functional projection &P as the main ingredient: &P is headed by the conjunction, which takes the second conjunct as its complement. The entire &P is then adjoined to the first conjunct so that the first conjunct is the element that fulfills selectional requirements, binary branching is respected, and all binding data fall out as desired. H also assumes that the grammar makes available two strategies: Base-generation of ‘small’ conjuncts and deletion from base-generated ‘large’ conjuncts. Finishing off this chapter with an illustration of the first strategy (DP-coordination), H pursues the second strategy in the subsequent chapters as both RNR and gapping support this large conjunct hypothesis: Large coordination structures that involve some form of deletion.

Deletion plays a major role in this theory. In Ch. 3, ‘Right node raising’ (53–142), H sets out to make her case against a movement analysis in terms of extraposition or right dislocation, for example, concluding that RNR violates all of the traditional diagnostics for movement. Instead, H proposes the PF-reduction theory: What looks like right dislocation of some constituent actually results from reduction of identical material at PF. Two components are needed to turn this claim into a falsifiable theory. First, H turns to the phonology of RNR to determine the phonological prerequisites that license the purported deletion. She identifies two possible contours which both employ accent tones immediately preceding the target in each conjunct. But the target of RNR itself contains additional accents in one contour while it is unaccented in the other. With government by prosodic phrasing established, H then argues that un-accented targets prosodically incorporate into the preceding phonological phrase. On the semantic side, H develops a version of alternative semantics that allows her the determination of focus assignment in coordinated structures. The interplay of syntactic, semantic, and phonological properties now allows H to apply the PF-reduction theory to RNR-constructions. In Ch. 4 (143–74), she extends the empirical validity of this theory to gapping, which differs from RNR mainly in terms of syntactic constituency, as the remnants must be maximal projections.

A one-page ‘Conclusion’ (175) wraps up the book and is followed by the bibliography, a name index, and a brief subject index. This book is an excellent addition to the ever-growing literature on ellipsis. It should therefore be compulsory reading for anyone working on such issues. But H’s monograph is also an exemplar for connecting different components of a grammar and developing a coherent, empirically testable theory.