|Title:||Clauses without 'that': The case for bare sentential complementation in English||Add Dissertation|
|Author:||Cathal Doherty||Update Dissertation|
|Email:||click here to access email|
|Institution:||University of California, Santa Cruz, Department of Linguistics|
|Abstract:||This dissertation argues for the `IP-hypothesis' of the structure of finite subordinate clauses without complementizers, such as the complement and relative clauses below:
I said [it was true.]
the chest [the key opened]
That is, it argues that the bracketed constituents above are bare finite clauses (IP) as opposed to CP with a phonologically null head (the 'CP-hypothesis').
This proposal has some broad theoretical consequences which are outlined in Chapter 1, e.g. it supports the claim of Chomsky 1986a and Gazdar et alia 1985 that 'S' is a maximal projection, contra the position that S is non-maximal (Bresnan 1982). The more specific theoretical and empirical consequences of the IP-hypothesis are investigated in the three major chapters of this dissertation.
In Chapter 2, evidence for the IP-hypothesis of the structure of \\f2that\\f1-less argument clauses is presented and the question of their distribution addressed: one possible objection to the IP-hypothesis is that it is incompatible with the ECP account of the distribution of these clauses (Stowell 1981). However, it is argued that the ECP account faces enough conceptual and empirical problems that its loss is not an objection to the IP-hypothesis.
Chapter 3 presents evidence for the IP-hypothesis of 'that-less' relatives (contact clauses). It is argued that the relative head directly A'-binds the gap in these constructions, without the apparatus of operator movement. This proposal forms the basis of an explanation of the similarities and differences between contact and other `full' relative clauses and also provides an explanation of the adjacency condition to which these clauses are subject.
In Chapter 4 the implications of the IP-hypothesis for extraction theory, specifically for the 'that-trace' effect, are explored. It is argued that the proposal allows for a maximal reduction of previous accounts based on the head-government relation (Rizzi 1990).
Finally, Chapter 5 contains some concluding remarks and a preview of what the extension of the IP-hypothesis to the nonfinite domain might entail.