To date, no real consensus has emerged among syntacticians about how to derive verbinitial order (V1); but the two main approaches, V0-raising and VP-raising, receive particularly widespread support in the literature. The syntax of Niuean pseudo noun incorporation (PNI) has played an important role in the propagation of the VP-raising analysis (Massam 2001), especially for VSO languages and languages with a VSO option. In this thesis, I present an analysis of the prosody of Niuean PNI and show that the PNI verb and incorporated argument form a prosodic constituent. While this result is consistent with the syntactic analysis of Massam (2001), it is also consistent with a prosodic restructuring analysis that explains the VOS order of PNI by appealing to prosodic well-formedness. I take the second approach. Specifically, the principle behind Selkirk’s (1984) Sense Unit Condition requires that the verb and its internal argument(s) form a unique phonological phrase. In order to satisfy this requirement, the incorporated argument moves into a position adjacent to the verb at PF. Positionally motivated categorical feature sharing (Adger and Svenonius 2011; Pesetsky and Torrego 2007) allows PF to reference the head-argument relationship between the verb and its internal argument, even though they are not sent to PF in structurally adjacent positions. The main result for the syntactic analysis of Niuean is that V0-raising replaces VPraising. The benefits of the V0-raising approach include i) less phonologically vacuous structure in places where Niuean has overt morphology, e.g., a perpetually null T0 in the face of overt tense markers; and ii) observance of the idea that thematic roles are correlated to structural positions. Thus, the prosodic analysis of Niuean PNI has a number of positive outcomes for Niuean syntax, as well as the potential to simplify the derivation of VSO cross-linguistically.

Prosodic Noun Incorporation and Verb-Initial Syntax

V1 or V2?: On the left in Niuean

http://ir.lib.uwo.ca/cgi/viewcontent.cgi?article=1065&context=frenchpub

External possession meets bare nouns in Malagasy

Inspired by Kayne’s (1994) Antisymmetry hypothesis and Linear Correspondence Axiom (LCA), there have been over the last decade numerous analyses employing extensive Remnant Movement1 (RM). While many of these are controversial, the RM analysis of Malagasy found in Pearson (2001) and Rackowski & Travis (2000) (R&T) in particular solves several outstanding puzzles of Malagasy syntax, although it does run into a possible problem with regard to the proper placement of arguments and adverbials, discussed in Thiersch (2004). Nevertheless, this approach to Malagasy seems promising, and confronted with apparent counterexamples to the analysis, one would like to see if they bear up under scrutiny. A welcome result would be that under a proper analysis, the approach actually predicts the judgments in question. This paper proceeds as follows. We give a brief summary of Malagasy syntax and the Remnant Movement approach, and give by way of motivation three sample problems which this analysis explains. We then discuss an alleged counterexample, which turns out to be a problem for languages other than Malagasy as well, and show that under the correct analysis it is in fact consistent with R&T’s analysis.

A note on the scope of adverbs in Malagasy

Inspired by Kayne's Antisymmetry hypothesis, there have been over the last decade numerous analyses employing extensive remnant movement, among which the Koopman & Szabolcsi (2000) analyis of Hungarian and Dutch and the Rackowski & Travis analysis of Malagasy In particular, R&T's approach solves several outstanding puzzles of Malagasy syntax, but runs into a possible contradition with regard to the proper placement of arguments and adverbials, namely needing extraction out of deeply embedded specifier positions in some cases and wanting to forbid it in others It turns out that this problem is found in K&Sz's analyses as well, and appears to be common to Remnant Movement approaches in general This article discusses the two different ways of solving this problem in these frameworks

/pdf/two-systems-of-remnant-movement-ii-and-extraction-from-1sacufyhe7.pdf

P. Law

Papers

Two systems of "Remnant Movement II" and extraction from Specifier position