Toward a Uniform Analysis of Short Answers and Gapping

As Steedman (1990, 248) puts it “[...] even the most basic gapped sentence like Fred ate bread, and Harry, bananas is only really felicitous in contexts which support (or accommodate) the presupposition that the topic under discussion is Who ate what?.” In this paper, I want to argue that Steedman’s remark hits the spot: given an adequate analysis of short answers, this analyis can be straightforwardly generalized to gapping within any model of discourse that is based on the assumption that discourse is structured by explicit or implicit wh-questions (e.g. Roberts, 1996; Büring, 2003). This, of course, is essentially equivalent to saying that the only difference between short answers and gapping boils down to the following: whereas in case of short answers the relevant wh-question is always present explicitly, this question typically needs to be reconstructed on the basis of the information structure of the initial conjunct in case of gapping. To substantiate the claim of a uniform analysis of short answers and gapping, I will first point at some well-known, but nevertheless suggestive commonalities between short answers and gapping. I will then sketch Rooth’s (1992) theory on question/answer congruence, and show that the constraint Rooth imposes on congruent sentential answers turns out to be an empirically adequate condition for ellipsis in short answers, in fact a more adequate condition than Merchant’s (2001) notion of e-Givenness. In a second step, I will generalize the short answer analysis to gapping along the lines suggested by Steedman’s quote. Actually, I will present two versions of the uniform approach: one version within alternative semantics, the other one within Schwarzschild’s (1999) theory on F-marking and accent placement. The paper closes with a discussion of some interesting implications of a uniform approach to short answers and gapping.