Even more on COMP

[Chris Culy]

Hi all, 

Just a couple small comments in response to TMA (or ATM??) on COMP. 

First, I'm not at all adverse to getting rid of COMP. In fact, I'm well on the
way to being convinced that it should be. However, ... 

Second, I agree that COMP as a GF is not available at a-structure, so that when
I referred to using COMP to block passive, I was using that as a shorthand for
introducing a new feature (a proposal dismissed by {T,M,A} in their original
message. I'm not adverse to adding new features *where necessary*, since I
added features galore in that extraposition paper I mentioned :). It is now at:

http://www.uiowa.edu/~linguist/faculty/culy/papers/index.html 

Third, the ungrammaticality of 

(1) *That they could reach the cookies was found (by the children) 

remains unaccounted for. Even if you think that the extraposed clause in the
alternate version: 

(2) It was found (by the children) that they could reach the cookies. 

is a subject, (and I don't) it doesn't explain (1). 

I also don't share {T,M,A}'s intuition about the meaning difference between
"find" and "realize". In particular, I don't think "realize" or "find" carries
any implication about the speaker's knowledge. 

The claim is 'that "realise" carries the meaning that the speaker knew prior to
the act of realisation that the proposition is true.' "find out" is claimed to
have a similar meaning. But consider (3-4). These should be odd on this view,
but they don't seem to be. (5) is given for comparison, to show that "find" has
the same status as the other two. The presupposition of all three verbs, I
believe, is simply that the clause is true, not anything about the time of the
speaker's acquisition of that knowledge. 

3. I quickly realized that I could reach the cookies by standing on tip-toes.
4. I quickly found out that I could reach the cookies by standing on tip-toes.
5. I quickly found that I could reach the cookies by standing on tip-toes. 

This is not to dispute the possibility of a semantic explanation. But until
some solution to (1) is found, the case against COMP is not airtight. 

Finally, I think that {T,M,A} are correct in raising the issue of the status of
COMP, and just what is being selected for by the verbs in question. The idea
that many of the properties can be explained on categorial or semantic grounds
is a good one. I'd also point out that XCOMP is subject to the same objections
as COMP if LMT defines grammatical functions only in terms of [r] and [o]. But
of course, this is a matter of definition, not an argument against COMP and
XCOMP. 

Regards, 

Chris 
chris-culy at uiowa.edu