Negation in gapping (review)

Sophie Repp’s new book contains two things in one. It is an in-depth investigation of negation that uses the gapping construction as a window into its true syntactic and semantic nature. At the same time, it presents a novel and original analysis of gapping based on scope facts of negation in gapping. Ch. 1 provides a concise introduction of the phenomena and critical reviews of the previous analyses of gapping. There are three basic scope patterns with negation in gapping, as exemplified by 1.

1. 1.

   1. a. Pete hasn’t got a video and John __ a DVD.

   2. b. Pete didn’t clean the whole flat and John __ laze around all afternoon.

   3. c. Pete wasn’t called by Vanessa but John __ by Jessie.

   4. d. Pete wasn’t called by Vanessa and John __ only by Jessie.

Among the three scope options, the first two, the distributed scope (1a: ⌝A & ⌝B) and the wide scope (1b: ⌝ (A & B)), have been widely discussed in the past. The third reading (1c and 1d), in which negation is confined within the first conjunct while the second conjunct is understood to be positive (i.e. ⌝A & B), has not been noted before, and plays an important role in distinguishing R’s proposal from the previous analyses.

In Ch. 2, ‘The syntax of clausal negation: The distributed readings in main verb gapping’, the distributed scope reading becomes the centerpiece of R’s syntactic proposal. The crucial observation is that, unlike in English, the distributed scope reading is not readily available in German. This crosslinguistic difference is attributed to the syntax of negation: the English not is a functional head and projects NegP, which is a part of the extended verbal projections, whereas the German negation nicht is adverbial and simply adjoins to VP. The main ingredient of R’s syntax for gapping is the ‘sideward’ movement operation of Nunes 1995, 2004. Take a simple gapping sentence like Andy read The Tim Drum, and Billy __ Momo. The second conjunct begins its numeration only with what is overtly expressed, {and, Billy, Momo}. The two DPs have Casefeatures that must be checked off, but as it stands, this numeration contains no material that can perform the feature-checker role. Therefore, the relevant material from the first conjunct is copied to the second conjunct via sideward movement. This copying process proceeds incrementally (i.e. the lexical verb first, then the small v, and ultimately T), and all the material that is a part of the extended verbal projections will be copied. This is how the difference between English and German is derived. The English negation not heads a NegP, which is selected by T. Thus, if negation is contained in the first conjunct, it must be copied to the second conjunct along with other verbal projections. In contrast, the German nicht is an adjunct, and being an unselected item, its presence is not required for the tree-building process. Hence, negation is not copied to the second conjunct. The notion of ‘need’, however, must be interpreted more broadly than just in terms of syntactic geometry of tree building. R notes that if the gapped predicate is a negative polarity item (NPI) (e.g. austehen können), negation can distribute to the second conjunct. The emergence of the otherwise unavailable reading is due to the necessity of negation for NPI licensing, which is semantically motivated and unrelated to the tree-building process.

Ch. 3, ‘The right kind of contrast: Narrow scope readings’, presents extensive discussion on the negation-narrow-scope reading with but and the focus adverb only, and the theoretical focus shifts to the semantic/pragmatic aspect of gapping. It has been noted in the past (cf. Kuno 1976, Sag 1976) that gapping requires the right kind of contrast between the conjuncts. R formalizes the condition of contrastiveness and lets it work as a licenser of the negation-narrow-scope readings. The analysis of the negation-narrow-scope with but (= 1c) can be summarized as follows: (i) the remnants are contrastive topics...