Language and Information (한국언어정보학회지:언어와정보)
- Volume 9 Issue 2
- /
- Pages.69-83
- /
- 2005
- /
- 1226-7430(pISSN)
Where a Null C Fails to PF-merge
Abstract
This paper discusses the distribution of a null complementizer C, as opposed to an overt counterpart 'that', presenting empirical support both for and against the PF-merge analysis of C proposed by Boscovic and Lasnik (2003, henceafter B&L) who in turn attribute to the proposal in Pesetsky (1992) and Halle & Marantz (1993). In Section I, as a background, I discuss B&L's proposal that a null complementizer C is a PF-affix which undergoes a PF-merger operation at the PF component. In Section 2, after a brief sketch of the distribution of a null C mostly in bare-relatives, I explore the possibility of extending B&L's analysis to accomodate the null C's in the bare-relative constructions. In Section 3, I argue that despite some empirical difficulties, B&L's analysis of a null C as a PF-affix can still be maintained, if Adverb Fronting is an operation to Spec-C position. Furthermore, I propose a rule - PF Spell-Out Constraint - to account for the C-trace (i.e. that-trace) effect in relative constructions. With the PF Spell-Out Constraint and B&L's PF-merge account, the distribution of a null C can better be analyzed.
Keywords
- Adjacency;
- Adverb Fronting;
- bare-relative constructions;
- a null complementizer;
- PF merger;
- Edge;
- C-trace effect;
- extraposition;
- stacking