empty verbs

[Tibor Kiss]

Hi,

I am quite surprised about the discussion concerning empty elements vs.
constructions or other vehicles. I always thought that an empty element in
itself is absolutely innocent. The question is not: should we have empty
elements or not?, but rather: How abstract should syntactic structures be?
This question motivated much research in the early 1980s where people
provided answers to the question: how can we provide an adequate grammar
which is less abstract than a transformational grammar (e.g. by eliminating
transformations)?

What I think is currently lost in the discussion is the obvious  conclusion
that a construction (in the sense of construction grammar or Ivan's work on
questions and relatives) is in no sense less abstract than an empty element.
It is only abstract on a different dimension. To frame it in old (european)
structuralist terms: a trace or empty element is syntagmatically abstract
while a construction is paradigmatically abstract. In sum: they are both
abstract and stand in the way of making syntax less abstract.

That a trace is abstract only on a syntagmatic level has been made explicit
in both GKPS and PS (1994), where a trace does not differ in its
paradigmatic properties from any other element. (BTW: This seems to be the
position in MP as well, since traces do not longer exist here.)

So in my view, if you eliminate a single trace in favour of a single
construction, you have gained nothing. But you haven't lost either. However,
if you replace a single trace by a 100 constructions, the gain is -99, and
that's a loss.

There has always been a second issue lurking in the back, viz. the
explanatory power of the abstract weapons we use. How many phenomena can we
explain by making use of an abstract entity and how many phenomena do we
sacrifice by not using it. I would claim that replacing a single trace by a
100 constructions does not actually put the burden of proof on the side of
the trace.

Finally, and back to the straight and narrow, it seems that with regard to
the empty verb in verb second constructions, Stefan Müller's solution of
using domains is equivalent to my suggestion of using an empty element
(Kiss/Wesche 1991, Kiss 1995b, Batliner et al. 1996).
I make use of an empty element and a lexical rule, Stefan needs domains.

However, in broader terms, the empty verb analysis fits nicely with analyses
of quantifier scope, modal scope, negation, and adverbial scope, scrambling
and extraposition. It has to be shown that a domain analysis captures the
same facts (and again, by not making use of more abstract entities). In
Andreas Kathol's book on domains, scrambling is not addressed. If his
analysis of extraposition were combined with a vanilla (Reape-style)
analysis of scrambling, than both extraposition and scrambling are basically
handled as PF-phenomena. For a refutation of such analyses, cf.
Culicover/Rochemont (1990) or Kiss (forthcoming). Quantifier scope and
scrambling is discussed in Kiss (2001). Adverbial and modal scope is
discussed in Kiss (1995b). Ok, written in German. But it contains a lovely
refutation of Kasper's claim (Kasper 1994) that adverbials can take scope
independently of position (he did not consider extraposition). Modal scope
is discussed in my thesis (1995a), again, sorry, written in German. Perhaps
the discussion is a further motivation to eventually provide an English
paper on verb-second in Germanic languages.


Best regards

T.


P.S.: References:

Batliner, A., A. Feldhaus, S. Geißler, A. Kiessling, T. Kiss, R. Kompe, E.
Nöth: Integrating Syntactic and Prosodic Information for the Efficient
Detection of Empty Categories. In: Proceedings of the In-ternational
Conference on Computational Linguistics (Coling), Kopenhagen, 1996.

Kathol, A. (2000): Linear Syntax. Oxford: Oxford University Press.

Kiss, T.: Infinite Komplementation – neue Studien zum deutschen Verbum
infinitum. Linguistische Arbeiten, Band 333. Tübingen: Max Niemeyer Verlag,
1995a.

Kiss, T.: Merkmale und Repräsentationen. Eine Einfühung in die
merkmalsbasierte Syntaxanalyse. Opladen/Wiesbaden: Westdeutscher Verlag,
1995b.

Kiss, T.: Configurational and Relational Scope Determination in German. In:
Meurers, W.D./T. Kiss (Eds.): Constraint-Based Approaches to Germanic
Syntax. Stanford: CSLI Publications, 2001, S. 141-176.

Kiss, T.: Semantic Constraints on Relative Clause Extraposition. Forthcoming
in Natural Language and Linguistic Theory.

Kiss, T./B. Wesche: Verb Order and Head Movement. In: Herzog, O./C.-R.
Rollinger: Text Understanding in LILOG: Integrating Computational
Linguistics and Artificial Intelligence. Berlin: Springer Verlag, 1991, S.
216-240.

----------------------------------------------------------------------------
Prof. Dr. Tibor Kiss -- Sprachwissenschaftliches Institut
Ruhr-Universität Bochum, D-44780 Bochum
+49-234-3225114 // +49-177-7468265 // +49-234-3214137 (fax)
What is present now will later be past.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Tibor Kiss.vcf
Type: text/x-vcard
Size: 139 bytes
Desc: not available
URL: <http://listserv.linguistlist.org/pipermail/hpsg-l/attachments/20010830/997b2913/attachment.vcf>