Date: Tue, 05 Nov 96 13:41:15 EST
From: Howard Lasnik <LASNIK@UConnVM.UConn.Edu>
Subject: Comments on Pesetsky's comments
To: linconf@tamvm1.tamu.edu
Message-Id: <961105.134312.EST.LASNIK@UConnVM.UConn.Edu>
In-Reply-To: <v03007800aea2fcfdb157@[18.162.0.29]>
Pesetsky's commentary is, as always, both penetrating and useful.
I find myself in substantial agreement with the points he raises,
so I will simply take this opportunity to elaborate on a few of
them.
In summarizing the three possibilities I considered for
Uriagereka's example
(15) There arrived two knights on each other's horses
Pesetsky's (B) is slightly misleading in attributing to
Uriagereka the idea that the formal features of the associate
move to a position higher than the reciprocal. Uriagereka's
actual proposal was in terms of Chomsky's (1986) 'expletive
replacement', under which the entire associate moves in LF to
subject position. Under the recent idea that in LF only features
move, we might render that proposal as Pesetsky's B, and that is
presumably what Pesetsky had in mind.
Next, Pesetsky is correct about my caution in rejecting a
Larsonian shell solution to the problem of (15). For all the
paradigms I actually discuss, the overt raising analysis gives
the correct results while the Larsonian approach does not do so
invariably. But, as Pesetsky notes, there are further phenomena,
such as his
(X) Mary gave candy to the children on each other's birthdays
where the overt raising account encounters difficulties. Of
course, as Pesetsky shows in Zero Syntax, the Larsonian approach
also runs into trouble with the range of binding out of PP
configurations represented by (X). See Zero Syntax for extensive
discussion, which, unfortunately, I did not have time to consider
in my paper.
Pesetsky's observation about double object asymmetries and
Larsonian shells is also on target. I rejected Larson's actual
proposal in part for Hoekstra's reason: Larson's derivations had
the effect of reversing the c-command relations between theme
(originally higher) and goal (subsequently higher). Given the
Belletti-Rizzi 'anywhere' approach to anaphor binding, which
Larson embraces, we predict not asymmetry but symmetry. However,
if goal is always higher than theme, asymmetry is correctly
predicted. Thus, as Pesetsky says, "Larsonian shells will offer
an account..." This conclusion is valid, if rather ironic, since
Larson's whole paper is an extended argument for the specific
structures he proposed, and the point of departure was the
binding asymmetries. But if his other arguments were to be
accepted, he would left with no account of the binding facts.
Further, just considering the binding facts, if the deep height
relations are preserved throughout the derivation, it is not
clear that we need shells in Larson's sense.
The fascinating backwards binding out of datives noted by Burzio
(Pesetsky's (Z)) does, when placed alongside (Y), suggest a
Belletti-Rizzi approach under the assumption that in dative
constructions (unlike in double object constructions), the
relative heights of the complements have been reversed.
(Y) I showed the professors to each other's students
(Z) I showed each other's students to the professors
If, on the other hand, anaphora is taken as strictly an LF
phenomenon, it is not obvious how these facts can be captured.
Possibly, this argues that LF (i.e., the interface with
semantics) is not one specific syntactic representation, but
rather scattered throughout the derivation, as in early work of
Jackendoff, and recent work of Epstein and some of his students.
One further factor to be considered, though, is that negative
polarity item licensing does not show the same symmetry in these
constructions:
(i) I showed none of the professors to any of the students
(ii)*I showed any of the students to none of the professors
Finally, with regard to the claimed parallel behavior of scope
and binding in dative constructions reported by Aoun and Li
(1989), I remain agnostic (along with Pesetsky, as a matter of
fact, who called the scope data "rather murky" in Zero Syntax).
--Howard Lasnik