Predicate Clefts in LFG

[Seth Alfred Cable]

Hi Andrew,

Actually, I'm not familiar with the arguments that "so" is a verbal (or
VP) pronoun.  If there is strong evidence that it is, I suppose the
proposal could we weakened to "there is no verbal (or VP) pronoun in
languages with predicate clefts" or "the 'gap' seeks out the null element,
and barring that goes with lexical elements in that catagory."  However, I
suppose my first move would be to (perhaps wrongly) challenge the claim
that "so" is a VP pronoun.  For example, couldn't the examples in (1)
simply involve VP ellipsis with the same "so" as appears in the sentence
bellow:

	Mary did so read the book.

Now, the "so" above has a kind of "contrastive" interpretation that isn't
apparent in the sentences in (1), but that might have to do with the
contrastive meaning of "did" here.  Moreover, note the degraded status of
(a) as compared with (b).

	 (a) * Bill read the book, and Mary did so.
	 (b)   Bill read the book, and so did Mary.

This suggests that "so" needs some kind of special focus-related
positioning even in situations where it might be analyzed as a VP
pronominal (in (1) below, it is adjacent to a focused "too"), bringing it
closer to the "so" witnessed in the first sentence above.

But, then again, I'm just riffing here...

Best,

Seth.





On Fri, 14 Nov 2003, Andrew Koontz-Garboden wrote:

> Hi, Seth.  I have nothing terribly substantive to contribute,
> unfortunately, just a quick question about the proposal you outlined,
> perhaps due to my misunderstanding/misreading.
>
> > "gap"; there is no "verbal trace".  VP has no "pronominal elements" at
> > all; thus, for verbs, this feature seeks out only full Vs.
>
> Do you not regard the english 'do so' as a VP pronominal element?
> E.g.,
>
> (1)  	a.  Kim ate the cake, and Sandy did so too.
> 	b.  Kim gave Sandy a book, and Pat did so too.
> 	etc...
>
> Perhaps there's some literature I'm unfamiliar with arguing against the (I
> thought standard view) that it is a VP pronoun of some sort?  Or, is the
> claim/empirical observation that in languages that have predicate clefts,
> there is nothing like the English 'do so', which perhaps forces them to
> use the verb itself instead?  Or, were you restricting the discussion to
> "silent" VP pronominal elements?
>
> Just curious,
>
> Andrew
>
>
>