PP-splits in Greek: Against scattered deletion

Nikos Angelopoulos

Keywords: PPs, splits, syntax

Abstract:

This paper examines previously unexplored splits in Greek involving possessor constructions embedded under Ps. In Greek possessor constructions, a genitive possessor, pjanu ‘whose,’ can be separated from a possessum, to podhilato ‘the bike’ (Horrocks & Stavrou 1987, Alexiadou et al 2007):
(1) Pjanui aghorase to podhilato ti?
whose.GEN bought.3SG the bike.ACC ‘Whose bike did she buy?’
I present a notable asymmetry split within PPs. While a possessor can never be separated from a possessum embedded under a P, (2, hereafter, possessor movement), some PPs allow what appears to be the movement of a non-constituent in (3). In this pattern (hereafter, possessum splitting), the possessor and the P co-occur in the left periphery to the exclusion of the possessum, which is left stranded postverbally.
(2) * Pjanui harike ja tin epitihia ti?
whose.GEN was.happy.3SG for the success.ACC. ‘For whose success was she happy?’
(3) [ Ja ti pjanu]j harike tin epithiai tj?
for whose.GEN was.happy.3SG the success.ACC ‘For whose success was she happy?’
Similar splits are allowed in different languages, prompting debate on their analysis. Some argue for scattered deletion (Fanselow & Ćavar 2002 i.a.), but I propose remnant movement as a more restrictive and explanatory account of PP-splits. Scattered deletion permits unattested patterns or rules out attested ones, unlike remnant movement, which accurately captures the observed data. This supports (4):
(4) *Scattered deletion: Scattered deletion is impossible in natural language.

Submission Number: 23