Unaccusativity in HPSG
Stefan Müller
Stefan.Mueller at cl.uni-bremen.de
Mon Nov 29 20:41:25 UTC 2004
Hello Raúl,
> I think for the molst
> part we agree, and I am glad to see that there is so much work on the
> fine-grained grammatical aspects of unaccusativity in HPSG.
Yes, I also think that the proposals do not differ that much in the end.
> The most interesting piece of data is that change of state verbs
> (i.e. get sick, rot, grow, etc.) do not seem to behave like other
> 'unaccusatives.'
German `erkranken', `faulen', `wachsen' do not allow passivization:
* Damals wurde öfter erkrankt.
* Leider wurde von der Pflanze gefault.
* Diese Nacht wurde von allen Pflanzen kräftig gewachsen.
(There are strange exceptions where unaccusatives are passivized. But
this is only possible in certain pragmatically restricted contexts
(Ruzicka, 1989, 350)).
`erkranken' and `wachsen' these verbs take `sein' as an auxiliary (as I
said, this test has exceptions) and they allow prenominal participles.
`faulen' takes `haben':
Das Wasser hat gefault.
The first two verbs do have prenominal participles:
die erkrankten Studenten
the ill.got students
die gewachsenen Zentren der Städte und Gemeinden
the grown centres of.the towns and ...
`faulen' does not:
* die gefaulten Zähne
the rotten teeth
> I have seen this unexpected behavior in many other
> tests, and I am curious to know whether in the German tests you talk
> about (prenominal participles and passive) aren't also sensitive to
> this group.
Many of the tests that have been suggested for the unaccusativity in
German also apply to so-called theme verbs (Wegener, 1990). For instance
they do not permit passivization.
> For me, the question is this: do we need a syntactic representation
> for the split in the class of intransitives, or can we by-pass that,
> attributing the contrasting behacior of 'unergatives' and
> 'unaccusatives' tho differences in semantics/pragmatics directly?
> Maybe in the cases you are talking about the generalization has to do
> with theta-roles, and not with a syntactic feature that distinguished
> internal from external arguments.
Well, in the case of passive, many reasearchers claim that we have to
have an agent that can be demoted in passivization. Some note that one
has to have a broad conception of `Agent' to get all the cases. For
instance, Hentschel and Weydt (1995, p. 175) explicitely allow for
inanimate agents.
I think the issue is really dificult, since there are several cases that
do not allow for a simple statement like: ``The participle in a `werden'
passive has to have a agent role''.
For instance the subject of `sehen' (`see') is usually considered to be
an experiencer:
(1) a. Er sah den Einbrecher.
he saw the burglar
b. Der Einbrecher wurde gesehen.
the burglar was seen
For so-called impersonal passives in German there seems to be a strict
restriction that the suppressed subject has to be animate (Paul, 1919,
p. 40; Jung, 1967, § 429; Zaenen, 1988). But this restriction does not
hold for the so-called personal passive (the examples show that the term
is a misnomer):
(2) a. Die Schneeflocken beeinflußten meine Entscheidung.
the snowflakes influenced my decision
`The snowflakes influenced my decision.'
b. Meine Entscheidung wurde durch die Schneeflocken beeinflußt.
my decision was by the snowflakes influenced
`My decision was influenced by the snowflakes.'
(3) a. Die Grammatikalisierung überlagert sie.
the grammaticalization overlies them
`The grammaticalization eclipses them.'
b. da sie von der Grammatikalisierung überlagert werden.
since they by the grammaticalization overlain are
Maybe one gets these examples with some proto-role approach.
The most problematic piece of data is the following:
(4) a. Es regnet die Stühle naß.
it rains the chairs wet
`The chairs were soaked by the rain.'
b. Die Stühle wurden naß geregnet.
the chairs were wet rained
`The chairs got rained on.'
(Literally: `The chairs got rained wet.')
The examples are due to Dieter Wunderlich (1997), who discusses them in
the context of resultative predication. The problem is that the subject
in (4a) is an expletive. Therefore all theories I am aware of (including
my own) exclude the example in (4b). I have no idea about what is going
on in (4) and how this could be allowed while excluding (5):
(5) * Heute wurde wieder geregnet.
today was again rained
It is possible that some broad conception of agent gets (1)-(3), but
there is no agent in (4).
> Maybe in the cases you are talking about the generalization has to do
> with theta-roles, and not with a syntactic feature that distinguished
> internal from external arguments.
> If one still needs to do so, however, I think an alternative is to
> play around with the configurations that we already have. In my paper
> I suggests that one class of postverbal subject in Spanish stays
> inside the VP. This is what in GP approaches was taken to be a
> manifestation of an internal argument: these are unaccusative
> internal arguments that never make it to SPEC-IP. I suggested that
> these are arguments that stay in the COMPS list, not in the SUBJ
> value. In these verbs, SUBJ is empty, and a non-canonical NP occupies
> the first slot in COMPS. So, one way to represent unaccusative verbs
> is to have an identity between SUBJ and one member of ARGST, but not
> the first one: the first ARGS member is a non-canonical NP. In other
> words: unaccusative subjects are subjects, but not a-subjects.
>
> SUBJ <NP1>
> COMPS < >
> ARGST <NP-noncanon, NP1>
The structure you gave seems to contradict the text. Do you mean the
following?
SUBJ < >
COMPS < NP1 >
ARGST < NP-noncanon, NP1 >
In the paper you had NP_gap at the first position of ARG-ST. The nice
thing is that the GAP-NPs are not mapped onto SUBJ or COMPS, but the
problem is that corresponding LOCAL values will be introduced in SLASH.
Since this is not what you want, you have to also mention this argument
in the BIND feature or introduce a special kind of non-cannonical NPs
that are not mapped onto SUBJ, COMPS, or SLASH. These special arguments
have to be mentioned in the restrictions that do the mappings.
If you go for the BIND solution, you use a mechanism that was introduced
for nonlocal dependencies to account for noncannonical argument
structure mappings.
> I don't know if there are any advantages of this approach over the
> one you suggest. I avoids having to introduce additional features for
> the purpose of distinguishing internal from external arguments. But
> the price is another mismatch between the valence lists and ARGST.
I think that the approaches are not so different: You say that there is
a non-cannonical mapping between ARG-ST and valence lists for certain
lexical items. You put an element on the ARG-ST list, that is marked in
a certain way. The proponents of DA, ERG also say that one element is
special and define a special mapping: Participle forms are derived from
a stem by a morphological lexical rule that also blocks the DA:
+- -+ +- -+
| | | |
| DA [1] | -> | DA [1] |
| SUBCAT [1] + [2] | | SUBCAT [2] |
| | | |
+-stem -+ +-word -+
This rule works for both unaccusatives and unergatives. We also could
append the DA at the end of [2] or another list in the output of the LR,
if this is needed for binding or we could place it somewhere further to
the right in the SUBCAT (ARG-ST) list of the stems, if this turns out to
be important for binding.
Best wishes
Stefan
@InCollection{HW95a,
author = {Elke Hentschel and Harald Weydt},
title = {{Das leidige {\em bekommen\/}"=Passiv}},
pages = {165--183},
editor = {Heidrun Popp},
booktitle = {Deutsch als Fremdsprache. An den Quellen eines Faches.
Festschrift f"ur Gerhard Helbig zum 65.\ Geburtstag},
address = {M"unchen},
publisher = {iudicum verlag GmbH},
year = 1995
}
@Book{Jung67a,
author = {Walter Jung},
title = {{Grammatik der deutschen Sprache}},
publisher = {VEB Bibliographisches Institut},
address = {Leipzig},
edition = 2,
optnote = {erste Auflage 1966},
year = 1967
}
@Book{Paul1919a,
author = {Hermann Paul},
title = {Deutsche Grammatik. Teil IV: Syntax},
publisher = {Max Niemeyer Verlag},
address = {Halle an der Saale},
volume = 3,
note = {\2nd\unchangedEdition 1968, T{\"u}bingen: Max Niemeyer
Verlag},
year = 1919
}
@Article{Ruzicka89,
author = {Rudolf R\r{u}\v{z}i\v{c}ka},
title = {{Lexikalische, syntaktische und pragmatische
Restriktionen des unpers\"onlichen
Passivs. (Eine komparative Et\"ude)}},
journal = daf,
volume = 26,
number = 6,
pages = {350--352},
year = 1989
}
@InCollection{Wegener90a,
author = {Heide Wegener},
title = {{Der Dativ -- ein struktureller Kasus?}},
booktitle = {Strukturen und Merkmale syntaktischer Kategorien},
editor = {Gisbert Fanselow and Sascha W. Felix},
series = {Studien zur deutschen Grammatik},
number = 39,
publisher = {Gunter Narr Verlag},
address = {T\"ubingen},
pages = {70--103},
year = 1990
}
@Article{Wunderlich97c,
author = {Dieter Wunderlich},
title = {Argument Extension by Lexical Adjunction},
optaddress = {Oxford},
journal = {Journal of Semantics},
volume = 14,
number = 2,
pages = {95--142},
year = 1997
}
@TechReport{Zaenen88a,
author = {Annie Zaenen},
title = {Unaccusative Verbs in {Dutch} and the Syntax-Semantics
Interface},
institution = {Center for the Study of Language and Information},
address = {Stanford},
type = {Report},
number = {{No. CSLI-88-123}},
year = 1988
}
--
Stefan Müller
Universität Bremen/Fachbereich 10 Tel: (+49) (+421) 218-8601
Postfach 33 04 40
D-28334 Bremen
http://www.stefan-müller.net
http://www.cl.uni-bremen.de/~stefan/Babel/Interaktiv/
More information about the HPSG-L
mailing list