UC Santa Cruz

This paper centers on a phenomenon known as interrogative continuation, which encompasses a class of fragment questions that depend on a linguistic antecedent and contain at least one apparently unraised interrogative element, as illustrated by the

example in (1).

(1) A: John bought a picture.

B: A picture of whom? [Bechhoffer 1976:45]

First appearing in the literature via Ross’s (1969) discussion on sluicing, the construction in (1) has received little attention to date and all investigations thus far have been largely inconclusive, inconsistent, and independent of one another. The goal of the current investigation is therefore to provide the most detailed account of interrogative continuation by building upon previous works about the phenomenon and drawing on aspects from all major subfields of linguistics. The widespread usage of this construction both within and across languages and the fact that it has been so understudied make this an all the more worthwhile endeavor.

In this paper, I survey a wide range of naturally-occurring data and lay out the core principles of interrogative continuations, which include their commonly sub-sentential fragment structure, discourse dependence, unembeddability, and apparent lack of whmovement. Using two acceptability judgment surveys, I resolve one major disagreement in the limited existing literature pertaining to the constructions’ dependence on preceding discourse material. I also discuss the distinction between interrogative continuations and echo questions, which I attribute primarily to the presence of a [Q]-related [F] feature that is inherent to genuinely interrogative elements but absent in echo questions (following Truckenbrodt (2012)). This featural difference in the syntax contributes to both prosodic and semantic differences between echo questions and continuations, where interrogative continuations follow an intonational pattern similar to that of regular constituent questions and function as genuine information-seeking questions. To account for these facts in the syntax, I propose a fragment-based analysis which builds on an account proposed by Weir (2017) and makes use of a new copying mechanism called Reuse-XP. I posit that this mechanism copies syntactic material from a linguistic antecedent into the continuation site, wherein an interrogative element is adjoined or realized as an overt complement to a copied predicate. Although many puzzles remain, this paper should serve as a starting point for future investigations into the nature of fragment questions.