To: linconf@tamvm1.tamu.edu
Subject: reinhart's remarks on verbal reflexives
Date: Wed, 30 Oct 1996 12:49:44 -0500
From: Jeffrey Lidz <lidz@louie.udel.edu>
In-Reply-To: <Pine.SOL.3.91.961030020953.23488A-100000@cclsun01>
Message-Id: <9610301749.aa04267@monk.ling.udel.edu>
Reinhart says that the fact that verbal reflexives like _koL_ are used
in apparent non-reflexive constructions "does not mean necessarily
that koL is not a reflexivizer. Chierchia has suggested, based on
Italian, that, in fact, lexical reflexivity and ergativity are derived
by the same semantic operation on a theta role, in the lexicon. (I
pursue the same line in a work in progress. The general correlation
between the two was observed also by Marantz.) The basic idea is that
the ergative version (in ii) is derived in fact from the transitive
'close', by an operation which absorbs one of the roles. The only
question is whether the remaining argument will realize as the subject
(reflexive) or as the object (ergative). In many cases the realization
is just identical. E.g. in German, 'the door opened' happens to be
'Die Tur offnet sich')."
Without seeing the details of this proposal, it is hard to evaluate.
However, there are reasons to doubt that the verbal reflexives involve
the same semantic operation in all constructions. There are several
pieces of evidence which support this position. First, we know that
in Kannada sentences which are semantically reflexive, both the
anaphor and the verbal reflexive are possible:
(i) Hari (tann-annu) hode-du-koND-a
Hari self-acc hit-pp-refl.pst-3sm
'Hari hit himself'
If ergative sentences arise due to the same semantic operation that
yields reflexive sentences, we would also expect the anaphor to be
possible in ergatives (all else being equal). It is not:
(ii) kiTiku (*tann-annu) tere-du-koND-itu
window (self-acc) open-pp-refl.pst-3sn
'the window opened'
Second, Reinhart suggests that the only difference between ergatives
and reflexives lies in the syntactic position filled by remaining
argument. The reflexive subject is a d-structure subject and the
ergative subject is a d-structure object. However, tests for
unaccusativity (in Kannada and Italian) suggest that even the
reflexives are unaccusative. In Italian, this can be seen through
auxiliary selection (both reflexives and ergatives take 'BE'). In
Kannada a test for unaccusativity is the inability to be embedded
under causative (see Chandrashekar 1994). Neither ergatives nor
reflexives can be embedded under causative.
(iii) *hari elea-gaL-annu biddav-is-id-a (unaccusative)
hari leaf-pl-acc fall-caus-pst-3sm
'Hari caused the leaves to fall'
(iv) *hari kiTiki-yannu tere-du-koLL-is-id-a (ergative)
hari door-acc open-pp-refl-caus-pst-3sm
'Hari caused the door to close'
(v) *hari raaju-vannu hogaL-i-koLL-is-id-a (reflexive)
Hari raaju-acc praise-pp-refl-caus-pst-3sm
'Hari made Raaju praise himself'
This suggests that the d-structure position of the reflexive subject is
syntactically equivalent to that of the ergative subject.
Third, not all ergatives are reflexive-marked. In Italian the
ergatives fall into two classes (see Rosen 1981), some of which are
reflexive-marked some of which are not. In Kannada, all ergative
verbs allow reflexive marking optionally (or at least, I have yet to
find an ergative verb which doesn't show this optionality). Further,
even the non-reflexive-marked ergatives behave like unaccusatives.
So, the only parallels I can see between the reflexive verbs and the
ergative verbs are the presence of the morpheme and their
unaccusativity. Even the former does not hold up fully, since some
ergatives are not reflexive-marked. In the places where it seems like
the Chierchia/Reinhart theory would predict differences between these
classes, it looks like they behave the same. In the places where they
predict similarity, it looks like they behave differently.
Finally, Reinhart makes no mention of the third class of sentences
which cross-linguistically show the presence of the reflexive marker:
(vi) Hari kannu-gaL-annu tere-du-koND-a
hari eye-pl-acc open-pp-refl.pst-3sm
'Hari opened his eyes'
It is hard to see how these sentences involve the absorption of an
argument (if that is indeed the Chierchia story (see also Grimshaw
1982; Burzio 1981) since the very argument that is allegedly absorbed
(the object) is present in the syntax.
Again, if this involved exactly the same operation as the semantically
reflexive sentences, then we would expect an anaphor to be possible
here in an argument position. This is not possible:
(vii) *Hari tan-age kannu-gaL-annu tere-du-koND-a
Hari self-dat eye-pl-acc open-pp-refl.pst-3sm
'Hari opened his eyes'
note that dative anaphors in argument position are perfectly grammatical:
(viii) Hari tanage pustakavannu koTTukoNDa
Hari self-dat book-acc give-pp-refl.pst-3sm
'Hari gave himself a book'
Furthermore, there is a systematic interpretive difference
cross-linguistically between sentences like (vi) with the reflexive
marker and those like (ix) without it:
(ix) Hari kannu-gaL-annu tere-d-a
Hari eye-pl-acc open-pst-3sm
'Hari opened his eyes'
When the reflexive marker is absent, we get the interpretation that
hari opened his eyes in the natural way that eyes are opened. When
the reflexive marker is present, we get the interpretation that Hari
pried his eyes open, perhaps with some instrument. Interestingly the
presence of an instrumental phrase is possible only with the reflexive
marked sentences:
(x) Hari kai-gaL-inda kannu-gaL-annu tere-du-koND-a
Hari hand-pl-instr eye-pl-acc open-pp-refl.pst-3sm
'Hari opened his eyes with his hands'
(xi) *Hari kai-gaL-inda kannu-gaL-annu tere-d-a
hari hand-pl-instr eye-pl-acc open-pst-3sm
'Hari opened his eyes with his hands'
Now, if the reflexive marker is possible only when there has been some
sort of lexical absorption of an argument, it is mystery why this
third role can occur only when the reflexive-marker is present, i.e.,
why it can occur only when it has been absorbed.
I agree with Reinhart, of course, that there is a similarity in the
lexical representations of the ergatives and the reflexives. This
similarity, however, does not seem to involve any kind of lexical
absorption. (for more detailed arguments and a proposal, see Lidz 1995, 1996)
-Jeff Lidz
References:
Burzio(1981) "Intransitive Verbs and Italian Auxiliaries" MIT Dissertation
Chandrashekar (1994) "The Syntax of Nominative Case in Kannada"
University of Washington Dissertation
Grimshaw (1982) "on the lexical representation of Romance Reflexive
clitics" in Bresnan, ed. "the mental representation of grammatical
relations" MIT Press.
Lidz (1995) "On the Non-existence of Reflexive Clitics" CLS
parasession on clitics.
Lidz (1996) "Dimensions of Reflexivity" University of Delaware Dissertation
Rosen (1981) "The Relational Structure of Reflexie Clauses: Evidence
from Italian" Harvard dissertation.
-Jeff Lidz