Heidi Harley: "light" verbs in English (reply to Carson Schutze)

[Martha McGinnis]

I agree that in a late-insertion approach "be"-support makes pretty
good sense, at least when "be" is tensed. Distinguishing between
"be"-support on the one hand and "have" and "do" support on the other
is pretty easy, I think, as Carson suggests:
-T is realized as "be" when the complement to TP is a small clause,
-you get "do" when the complement to TP (or NegP) is a vP (and hence eventive),
- you get "have" when the complement to TP is a saturated PP headed by P(have).

(That last one isn't so nice; presumably a saturated PP is itself a
small clause, so it really ought to get "be" -- but maybe "have" is
really just a suppletive spell-out of "be" in a certain environment,
i.e. with a P(have) complement).

But: of course "be" can behave like it's not "support" at all; it's
got a past participle form, etc. How do you get "be-support" to work
out in "John has been swimming"? (this applies to "have" as well, of
course). Seems likely that this is the reason that most folks have
stuck with the more theoretically cumbersome and intuitively strange
notion that auxiliaries in English are generated as main verbs and
then move. And a late-insertion approach that realizes "be" as a v or
V node will have to explain why those v or V node s can raise but
ones that don't get realized as "be" can't -- so we're back to our
same old problem of explaining why "underspecified" v or V can raise
but contentful v or V can't.

But there must be a better way!

Cf a semi-interesting recent discussion by Lightfoot (I think) about
how "be" isn't really a verb at all.

Thoughts?

best, hh

-- Original Message --

>Now that the list has sprung back to life, here's a question I've been
>wondering about for a while. It has to do with possible explanations for
>the fact that in English, 'be' and (some uses of) 'have' are the only
>verbs (as opposed to modals) that can appear to the left of negation.
>
>Almost the entire literature on this topic follows the line that these verbs
>can "raise across" negation in some way that other verbs cannot, a view
>that seems to me to raise more questions than it answers.
>
>There is an alternative, under which one can maintain that the Head Movement
>Constraint (or whatever underlies it) actually blocks all head movement
>across Neg in English, and what's special about 'be' and 'have' is that
>they can be inserted/generated above Neg. The only place I can recall
>seeing this pursued at all is in Ouhalla's work, including his 1991 book
>on functional categories. My question for the list is: why haven't more
>people followed this approach? Does it have some fatal flaw that Jamal (and
>I) have not noticed?  [I would grant that he doesn't work out many details,
>including perhaps crucial ones, but my impression was that they could be
>worked out.]
>
>Why is this a question for the DM list? Well, it seems particularly natural
>on a late insertion view that follows a 'be-support' approach and claims
>that be' is the last resort (default) spell-out when you need a V node and it
>has no properties, i.e. no encyclopedic meaning. (Getting 'have' is perhaps
>a little trickier, but conceivable if one believes that 'have'3D'be+P'.)
>
>I would be happy to hear about either arguments against this approach or
>references to where it is worked out.
>
>	Carson
>

---------------------------------------------------------------------
Heidi Harley
Department of Linguistics
Douglass 200E
University of Arizona
Tucson, AZ 85721
Ph: (520) 626-3554
Fax: (520) 626-9014
hharley at u.arizona.edu