A C-stranding Analysis of Null Complementizers

This paper is purposed to explain the distribution of null complementizers in English (and Korean) in terms of an extension of the analysis of preposition stranding (Abels 2003). In English the complementizer that is basically optional in canonical positions like complement clause. This is, however, not always the case when we consider non-canonical contexts like subject clauses or dangled complement clauses. The complementizer that is obligatory in those contexts. If we consider more non-canonical (derived) positions, the situation gets still more complex. If the complementizer clauses are topicalized to the front, that is obligatorily required. Similar ungrammaticality is found in other non-canonical/derived positions if no overt complementizers show up before the subordinate clause. This paper tries to find a solution to the paradigm through an extension of Abels (2003) generalization about C. This paper extends his observation to null complementizers. The gist of the paper is that TP does not strand C, whether overt or covert. The proposal of this paper can be extended to Korean. This paper also argues that C to V movement analysis of null complementizers is not on the right track, contra Pesetsky 1992, Bošković and Lasnik 2003, An 2004, Ahn and Cho 2005. In addition, this paper will eventually claim that a subordinate clause without overt C is still CP rather than TP, contra Doherty 2001, Déprez 1994, Ishii 1999/2004.