A yes/no question fragment (YNQF) is an incomplete sentence that is
nonetheless interpreted as a full yes/no question. I argue that those YNQF's
are derived from an underlying syntactic structure through a process of deletion.
First, I shall adduce evidence for the claim that ellipsis is involved by
looking at connectivity effects (Merchant 2004), such as case-matching, binding
and control. Next, I will consider the nature of the process of deletion in
these fragments. I claim that the deletion in question is phonological in
nature. Evidence comes from cases of nonconstituent deletion and contrasts
between Dutch and English YNQFs.