Generalized domain widening.
Domain widening has come to play a central role in the analysis of the English determiner any. Follow-
ing Kadmon and Landman (1993; KL), many authors analyze any as an existential indefinite with a
widened quantifier domain restriction. Such analyses necessarily interweave two threads--domain
widening and existential quantification. If, however, domain widening is an independent semantic
mechanism, it should be possible to disentangle these threads by finding domain wideners that
aren't tied to existential quantification. This is what this paper does, in providing an analysis of the
German adverb überhaupt. Überhaupt combines with a variety of categories that involve contextually
supplied arguments similar to quantifier domain restrictionsfor instance comparison classes in
APs headed by gradable adjectives, or domain restrictions as expressed by domain adverbs.
To initially illustrate the parallel to the familiar any-series, the data in (1) on the attached page
show how überhaupt combines with the German existential indefinite etwas, yielding a meaning just
like English anything (at all). The translations illustrate how the distribution of the combined expres-
sion überhaupt etwas is just like that of anything as well.
One very appealing insight in KLs proposal and those following it is to link any to domain
'
widening, and to derive its distribution in an intuitive way, by making domain widening subject to
licensing by semantic strength. However, since überhaupt is independent of existential quantifica-
tion, its distribution is expected to be wider than that of the any-series. This prediction is correct.
Any, for instance, is licensed under the scope of negation, as it is there that the existential receives a
universal interpretation (by the well-known fact that ¬x x ¬), in which case quantification
over a widened domain leads to a stronger statement. Überhaupt, by the same token, is predicted to
combine not only with existentials under negation, but also with universals in positive contexts as
well. This prediction is borne out as (2a) illustrates. Hand in hand with this comes the prediction
that überhaupt should be infelicitous with universals in downward-entailing contextsa prediction
that is borne out as well, as illustrated by the example in (2b).
The data in (3) illustrate überhaupt's ability to target not only quantifier domain restrictions, but
other contextually supplied domain restrictions as wellin (3B) it combines with an AP headed by a
gradable adjective, and in (4B) with a VP that was previously modified by a domain adverb.
(3) and (4) not only illustrate part of überhaupt's flexibility as a general domain widener, but
show something about its licensing conditions as well. In general, I follow the insights of current
work on domain widening that domain widening `should be for a reason' (Kratzer and Shimoyama
2002, in describing KL). Examples (3) and (4) illustrate however that several factors can license
domain widening: there for instance a challenge to a narrow domain restriction in a discourse. This
discourse orientation can be seen in the example in (5) below as well. In this case, überhaupt has a use
described in König (1983) as `targeting presuppositions', which I argue can be analyzed as widening
the common ground.
Summing up, this paper provides arguments in favor of an independent mechanism of domain
widening. It supports the idea that domain widening has to be licensed, but shows that there are sev-
eral ways to do this. It also provides a uniform semantics for the apparently diverse uses of überhaupt.
As items like überhaupt appear in a variety of languages (Migron 2005), this analysis will hopefully
provide some insight into the semantics of these items as well.
(1) a. Davon kann man ganz selten überhaupt etwas erkennen.
of that can one whole rare überhaupt something recognize.
`You can only rarely recognize any of that at all.'
b. Davon kann man ganz oft (*überhaupt) etwas erkennen.
of that can one whole often überhaupt something recognize.
`You can quite frequently recognize some (/*any) of that.'
(2) a. Meine Mutter kennt überhaupt jeden in Mindelheim.
my mother knows überhaupt everybody in Mindelheim.
`My mother knows absolutely everybody in Mindelheim.'
b. Meine Mutter kennt nicht (*überhaupt) jeden in Mindelheim.
my mother knows not überhaupt everybody in Mindelheim.
`My mother doesn't know everybody in Mindelheim.'
(3) A: Peter ist groß für einen fünfjährigen. B: Peter ist überhaupt groß.
Peter is tall for a five-year-old. Peter is überhaupt tall.
`Peter is tall for a five-year old.' `Peter is tall in general.'
(4) A: Politisch kann man da einiges machen.
Politically can one there some things do.
`Politically, there are a number of things you could do in this situation.'
B: Da kann man überhaupt einiges machen.
There can one überhaupt some things do.
`There are quite generally a number of things you could do in this situation.'
(5) A: Du hast dir von Fritz viel Geld geliehen.
You have to you from Fritz much money borrowed.
`You borrowed a lot of money from Fritz.'
B: Ich kenne Fritz überhaupt nicht.
I know Fritz überhaupt not.
`I don't even know Fritz.'
References
Kadmon, Nirit and Fred Landman (1993): Any. In Linguistics and Philosophy 16(4), 353-422.
König, Ekkehard (1983): Polysemie, Polaritätskontexte, und überhaupt. In Harald Weydt
Partikeln und Interaktion, 160-171. Tübingen: Niemeyer
Kratzer, Angelika and Junko Shimoyama (2002): Indeterminate Pronouns: The View
from Japanese. In Yukio Otsu (ed.) The Proceedings of the Third Tokyo Conference on
Psycholinguistics (TCP 2002), 1-25. Tokyo: Hituzi Syobo.
Migron, Hagrid (2005): Expressions Similar to Hebrew `bixlal'. In: Linguist List Issue16-999.
http://www.linguistlist.org/issues/16/16-999.html#1