The Problem of Fragment Answers

Abstract

In this paper, I discuss the status of fragment answers to yes-no questions, based on facts from European Portuguese. I argue, along the lines of argumentation in Merchant (2004), that there is reason to believe that at least some of these fragments are derived through deletion. However, I show that data from EP does not support Merchant’s analysis of fragments as constituents moved to the left periphery before deletion. This leaves us with the problem of non-constituent deletion, which I argue is not a problem for a phonological deletion theory of ellipsis.