LINGUIST List 20.4080: Syntax: Mavrogiorgos: 'Proclisis and Enclisis in Greek'

LINGUIST List 20.4080

Sun Nov 29 2009

Diss: Syntax: Mavrogiorgos: 'Proclisis and Enclisis in Greek'

Editor for this issue: Di Wdzenczny <dilinguistlist.org>

1. Marios Mavrogiorgos, Proclisis and Enclisis in Greek

Message 1: Proclisis and Enclisis in Greek
Date: 29-Nov-2009
From: Marios Mavrogiorgos <mm476cam.ac.uk>
Subject: Proclisis and Enclisis in Greek
E-mail this message to a friend

Institution: University of Cambridge Program: Department of Linguistics Dissertation Status: Completed Degree Date: 2009

Author: Marios Mavrogiorgos

Dissertation Title: Proclisis and Enclisis in Greek

Linguistic Field(s): Syntax
Subject Language(s): Greek (ell)
Dissertation Director:
Ian Gareth Roberts
Dissertation Abstract:

My dissertation provides a principled analysis for two interrelatedphenomena in the morphosyntax of Greek clitic pronouns: proclisis (1) andenclisis (2), respectively:

(1) When the verb is in the indicative or the subjunctive, the cliticprecedes the verb, and nothing may intervene between it and the verb.

(2) When the verb is in the imperative or the gerund, the clitic followsthe verb, and nothing may intervene between it and the verb.

In my dissertation I argue that object clitics are topicalizers, i.e.optional determiner heads merged on top of the proper Determiner Phrase,which mark the direct or indirect object of the clause as topic/oldinformation. I further argue that syntactic cliticization follows fromsyntactic agreement between the clitic pronoun and a phase head (seeChomsky 2001 for the assumption that syntactic derivation is computed inphases). For Greek and for other languages which have adverbal clitics Ipropose that the relevant phase head is v*-transitive and not Tense (T),contrary to Kayne a.o. Syntactic agreement between the clausal head v* andthe clitic leads to movement of the clitic to the left edge of v* andincorporation of the clitic into v*. Incorporation yields proclisis andtakes place for two interrelated reasons: first, because the cliticcontains only a subset of the features contained in v*, and second becausethe edge of v* is still open/accessible due to the fact that v* has notchecked all of its features.

I further propose that in enclisis person agreement (on T) is defective,while the verb must check verbal inflection on the higher ComplementizerModal (CM) head. The clitic targets v*, as in proclisis, however it doesnot incorporate into it, because CM is the new phase head by being thehighest inflectional verbal head in a chain of verbal heads. The verb movesto CM and the clitic merges with it from the lower specifier of vp(assuming that T is not projected in enclitic environments, i.e.imperatives and gerunds). In this way we get the generalization thatenclisis obtains when the verb moves across the cliticization site to aV-related site, i.e. to a site where a verbal inflectional head is found.This allows us to differentiate between imperatives which have V-to-Cmovement and take enclitics, from interrogatives, which in some languageshave V-to-C movement, but take proclitics.

The advantage of this analysis is that it can be generalized acrossconstructions and across languages without great difficulty, since it isbased on the general principles of (i) subset of features, and (ii) edgeavailability/accessibility. Moreover, by basing cliticization on agreementwith phase heads, and given Chomsky's theory on phase heads, it is flexibleenough to be able to account for a plethora of distinct cliticconstructions both within a single language and cross-linguistically, whichis something previous theories lacked.