More on COMP

[Chris Culy]

This is a followup to Tara, Mohanan, and Alex's (TMA) interesting posting 
regarding COMP. Briefly, they argued that the grammatical function COMP has no 
place in current LFG, and that arguments that would have been categorized as 
COMP should be categorized in terms of [r] and [o]. Their argument, as I 
understand it, is essentially one of simplicity: Lexical Mapping Theory 
provides [r] and [o] for grammatical functions, and unless is there is some 
compelling reason to add to that inventory, we shouldn't.

I just want to mention a couple small points that I discussed in an unpublished 
paper a couple years ago. First, I also allowed arguments otherwise 
categorized as COMP to have the features [r] and [o] instead. The motivation 
came from extraposition: "it" seems to have the basic grammatical functions, 
and we understand "it" to "replace" (in some sense) the clause, so the clause 
should have the same basic grammatical function. To the extent that this 
reasoning can be maintained, it lends further support to TMA's position.

However, I did keep COMP around for one purpose, namely to prevent 
passivization with certain predicates. So we have:

1. The children found the marbles.
2. The marbles were found (by the children).
3. The children found that they could reach the cookies.
4. *? That they could reach the cookies was found (by the children).
5. The children realized that they could reach the cookies.
6. ?That they could reach the cookies was realized by the children.

1-2 show that when find takes an NP OBJ, it can be passivized. However, when 
find takes a clausal complement, it cannot be passivized, as in 3-4. In 5-6 we 
have realize, nearly synonymous with 3-4, suggesting that the 
non-passivizability of 3-4 is not due to semantics/thematic roles.

If there is some other way of blocking 4 (or if it can be argued that 4 
shouldn't be blocked), that would bolster TMA's position, which is otherwise 
weakened by these examples.

Finally, TMA say:

"Now, a possible argument that one might advance in support of COMP is the
observation that clauses cannot appear as complements of prepositions:"

They go on to argue that this observation is about c-structure, not GFs. We can 
also add that it has been arged that the observation itself is incorrect (thus 
supporting TMA). So we have examples like:

7. We didn't talk much about whether or not our shoes were muddy with Fred.
(Bresnan and Grimshaw 1978:333)
8. I didn't talk much about what a complete idiot I'd been.

TMA's proposal for the elimination of COMP seems eminently reasonable, with the 
one possible glitch mentioned above.

Regards,

Chris
chris-culy at uiowa.edu
P.S.
I can make the paper on extraposition available if there's interest.