The focus of this thesis is the description and analysis of clausal complementation in Sundanese, an Austronesian language spoken in Indonesia. The thesis examined a range of clausal complement types in Sundanese, which consists of (i) yén(/wi)/réhna 'that' complements, (ii) pikeun 'for' complements, (iii) sangkan/supaya/ngarah/sina 'so that' complements, (iv) raising complements, (v) crossed control complements, and (vi) nominalizations. This varied set of complement structures display distinct properties in terms of the sort of elements admitted in the complements.
The theoretical aspect of the thesis is the examination of two important generalizations: (i) that complementation is a universal feature of human languages (Noonan 1985, 2007); and (ii) the well-accepted precept that finiteness plays a role in the world's languages. This thesis provides evidence that Sundanese evinces (syntactic) complementation and that any claim to the contrary is unfounded. In terms of finiteness, despite the lack of overt morphological manifestations of finiteness, the thesis argues that finiteness seems to be at work in Sundanese and that it operates as it does in other languages to account for the distribution of overt subjects.
In addition, the body of data presented herein is also germane to a host of other theoretical issues, especially with regard to Austronesian languages. The first is inclusion of VoiceP in a clausal structure. Following (Sukarno 2003, Son 2006, Son & Cole 2008, Cole et al. 2008, Ko 2009 and Legate 2011), the thesis adopts an additional functional layer above vP, i.e. VoiceP, to harbor voice marking. It is proposed that in Sundanese transitives, both actor DPs in active sentences and actor PPs in passive counterparts are arguments and are therefore merged in the same slot, i.e. Spec,vP.
The second theoretical point investigated in this thesis is whether Raising to Object and Proleptic NP constructions are alike or different. In this thesis, I claim that the two types of constructions should be analyzed as instantiations of two distinct structures, mainly due to structural properties: Raising to Object involves movement, while prolepsis does not.
The next theoretical issue has to do with a subset of control predicates, which exhibits behaviors atypical of canonical control. I propose a slightly different analysis that draws upon earlier accounts (Polinsky& Potsdam 2008; Fukuda 2008; Nomoto 2008, 2011, Sato & Kitada 2012). On the basis of (a) the presence of a plural marked-verb inside the crossed control complement and (b) the apparent parallelism between the ordinary control and the crossed control, I postulate that the structure for the two types of control of the same predicates is identical, in which case their complement includes VoiceP.
The last theoretical concern is related to the fact certain nominal structures display verb-like properties. Following Alexiadou (2001), the present thesis proposes that, like verbal structures, some nominals contain functional projections such as AspP, VoiceP and VP. This naturally explains why nominals exhibit verbal properties that they do.