What sluicing can do, what it can't and in which language - On the cross-linguistic syntax of ellipsis

The study of ellipsis in current generative grammar is still strongly—perhaps too strongly—construction oriented. Every introductory article on the subject recognizes at least sluicing, VP-ellipsis, NP-ellipsis, gapping, stripping, pseudogapping, conjunction reduction, and a handful of other constructions as falling under the general rubric of ellipsis (cf. e.g. Merchant 2009). On the one hand, this diversification is not surprising, as it is well-known that not all of these elliptical phenomena behave alike. For example, Lobeck (1995) shows in detail that sluicing, VP-ellipsis, and NP-ellipsis share certain properties that set them apart from gapping, stripping, and pseudogapping. On the other hand, however, such properties might simply be telling us what are—or rather, what are not—good diagnostics for identifying a particular elliptical construction. A revealing example in this respect is the line of reasoning initiated by Jayaseelan (1990), who tries to reduce pseudogapping to VP-ellipsis (see Gengel 2007 for recent discussion and references). To the extent that this analysis is on the right track, it suggests that whatever properties set apart pseudogapping from VP-ellipsis (e.g. sensitivity to the Backwards Anaphora Constraint) is not a distinctive trait of VP-ellipsis and hence should not be used in the identification of this construction. A strong indication that this approach is worth exploring comes from crosslinguistic research into ellipsis. What emerges from such studies is that independent syntactic differences between languages can cause the elliptical constructions of those languages to come out differently as well. This implies that those aspects that differ can no longer be seen as defining characteristics—that is diagnostics—for that particular elliptical construction. A case in point is the study of VP-ellipsis in OUP UNCORRECTED PROOF – Page Proof, 28/3/2013, SPi