Catalan Journal of Linguistics 7, 2008 9-40
Exclamatives: A Thematic Guide with Many
Questions and Few Answers*
Xavier Villalba
Universitat Autònoma de Barcelona. Centre de Lingüística Teòrica
08193 Cerdanyola del Vallès (Spain)
Xavier.Villalba@uab.cat
Abstract
In this paper a snapshot is offered of the state of the art in the research on exclamatives, review-
ing our current understanding of “classic’’ issues, like factivity, high degree, or the relationship
between exclamatives and interrogatives, and highlighting new theoretical and empirical chal-
lenges, such as the contribution of exclamative sentences to a dynamic model of discourse or the
placement of exclamative phrases in the left-periphery.
Key words: exclamative sentences, factivity, degree, widening.
Table of Contents
1. Introduction 4. The form of exclamatives
2. The meaning of exclamatives 5. Conclusions
3. The use of exclamatives References
1. Introduction
The interest in exclamative sentences in modern linguistic theory is customarily
traced back to the work by Dale E. Elliott in the early 70’s (Elliott 1971, 1974).
Elliott designed a pool of tests for determining the exclamative character of English
sentences in close contrast with interrogative sentences. Even though of limited
application beyond English and burdened with quite heterogeneous a nature, Elliott’s
approach to the issue in general and his tests in particular have been influential
enough to determine the research agenda of the area for nearly two decades.
* I am grateful to Elena Castroviejo for comments that contributed to improve this article. The author
want to acknowledge the finantial support from the grants HUM2006-13295-C02-01/FILO
(MCyT/FEDER), BFF2003-08364-C02-01 (MCyT/FEDER), and 2005SGR00753 (Generalitat de
Catalunya) awarded to the Grup de Gramàtica Teòrica (now Centre de Lingüística Teòrica) of the
Universitat Autònoma de Barcelona. Usual disclaimers apply.
,10 CatJL 7, 2008 Xavier Villalba
On the positive side, the work by Elliott raised attention to exclamative sen-
tences, which, when considered at all, had received a parochial and secondary role
with respect to interrogative sentences. It can be said with justice that Elliott set
the grounds for a fruitful development of the area in the subsequent years. Consider
for instance, his observation that exclamative clauses can only be selected by fac-
tive predicates, so that the following contrast arises –note that very makes the sen-
tence necessarily exclamative, another fact pointed out by Elliott:
(1) a. It’s amazing how very expensive this wine is.
b. *I asked how very expensive this wine was.
Even though Elliott just notes this contrast without raising any theoretical con-
sequence, this observation is at the basis of a major debate in the field concerning
the proper semantic characterization of exclamative sentences. For instance,
Grimshaw (1979) took Elliott’s observation a step further and argued that excla-
mative constructions are factive –that is, they presuppose the truth of the proposi-
tion they denote–, so that the illformedness of (1b) results from the clash between
the presupposition associated with the exclamative sentence (“This wine was very
expensive’’) and the semantic requirements of the predicate ask, which presup-
poses ignorance concerning the validity of the proposition denoted by its comple-
ment.
This line of research has been formalized in several ways. Gutierrez-Rexach
(1996, this volume) derives factivity from the denotation of an illocutionary excla-
mative operator, which includes a null emotive predicate. A different technical
solution –even though with an equivalent import– is pursued in Zanuttini and
Portner (2003), where it is argued that the factivity of exclamatives is syntactical-
ly encoded by means of a factive operator in CP.
Yet, a different line of research exists that questions the factivity analysis of
exclamatives (see Castroviejo 2006 and Mayol this volume). We will consider this
issue in more detail in 2.3, but we hope that this brief discussion suffices to make
apparent the long-lasting influence of Elliott’s pioneering work in the research on
exclamative sentences.
Important as it has been, one must acknowledge that Elliott’s influence hasn’t
been always positive, for two reasons. First of all, some tests are of a very limited
coverage. For instance, he remarks that subject-verb inversion is a clear test to sep-
arate matrix interrogative sentences, which involve inversion, from exclamative
ones, which don’t:
(2) a. How many books you read!
b. How many books did you read?
Yet, although this test works fine in English or French, it does less clearly in
German or Italian, fails altogether in Catalan or Spanish, or is simply unavailable
in SOV languages like Japanese. This limited coverage reaches its peak with the
,Exclamatives: A Thematic Guide with Many Questions and Few Answers CatJL 7, 2008 11
very test –i.e. English wh-exclamatives may be appended with the modifier very:
How (very) tall Mary is! –, which is a highly idiosyncratic feature of English excla-
matives. Obviously, it would be unfair to charge Elliott for these shortcomings, for
he designed the tests for English exclamatives. The responsibility for devising new
and more far-reaching tests is to be put on the burden of the community of
researchers, and a quick glance at the recent literature shows how difficult this task
can be.
Secondly, even though some tests allow for a wider application, and can be
arguably taken to reveal some general property of exclamatives, Elliott failed to
make clear the exact import of his tests, particularly whether the presumed feature
underlying them is syntactic, semantic or pragmatic –which, to be fair with his
contribution, is a matter of investigation nowadays. Take for instance, the coordi-
nation test, which shows that interrogatives are typically associated with a dis-
junctive reading in terms of alternatives, whereas exclamatives associate with an
additive reading:
(3) a. How expensive is this book? Ten euros {or rather/#and even} twenty in
some bookstores?
b. How expensive this book is! Ten euros {#or rather/and even} twenty in
some bookstores.
Intuitively, the disjunctive option seems fine with typical interrogative sen-
tences, for it nicely correlates with the exclusive nature of the process of assign-
ing a value to the variable introduced by the interrogative wh-element. Exclamative
sentences, instead, do not seem to involve this process, for reasons to be explained
–a particularly cogent problem for the proponents of a semantic analysis of excla-
matives based on that of interrogatives; see Abels (2004), d’Avis (2002), and
Gutiérrez-Rexach (1996, 2001, this volume). However, what is still a matter of
debate is the exact signification of this test.
In this paper we will offer a snapshot of the state of the art in the research on
exclamatives, reviewing where we are with respect to “classic’’ issues, like factiv-
ity, high degree, or the relationship between exclamatives and interrogatives, and
highlighting new theoretical and empirical challenges, such as the contribution of
exclamative sentences to a dynamic model of discourse, the placement of excla-
mative phrases in the left-periphery, and the relation between exclamatives and
subject-verb inversion. It goes without saying that the list of discussed issues is
not exhaustive, but we believe that it is representative enough to give an accurate
picture of the lively research area of exclamatives.
The organization of the article is a classical tripartite division between mean-
ing (section 2), use (section 3) and form (section 4).
2. The meaning of exclamatives
In this section we will be concerned with the main aspects involved in the meaning
of exclamatives, and the (divergent) solutions devised to account for them. From the
, 12 CatJL 7, 2008 Xavier Villalba
literature, one can set the following interpretive issues, that we consider in detail in
the following subsections:
— high/extreme degree (2.1)
— emotional content (2.2)
— factivity (2.3)
— semantic type (2.4)
2.1. High degree
It is a well-attested fact that exclamatives involve somehow the placement of an
individual on a scale, and particularly on the extreme of such a scale, as the fol-
lowing example and the paraphrases try to show:
(4) How expensive this wine is!
“This wine is extremely expensive’’/“This wine is expensive to an extreme
degree’’
The immediate question that arises concerns the proper treatment of this inter-
pretive feature, and the answers are, as usually, divergent.
One solution is to derive this meaning from the presence of a maximality oper-
ator over degrees, following Rullmann (1995). This is the path that takes Gutiérrez-
Rexach (1999) for Spanish degree neuter constructions:
(5) a. ¡Lo caro que es el vino!
the.NEUTER expensive that is the wine
‘How expensive the wine is!’
b. lo caro ⬅ MAX(λdλx.expensive’(d)(x))
Villalba (2004) extends this analysis to degree wh-exclamatives in order to
account for the quantificational restrictions that affect exclamatives, namely the
very restricted distribution of negation and the necessity of quantifiers over indi-
viduals to be specific (see González this volume for a different look at negative
exclamatives).
Yet, treating the high/extreme degree meaning as a semantic presupposition
following from the maximality operator does not seem an adequate solution. For one
reason, there is evidence that this interpretation is not always present in exclama-
tives, as argued for by Castroviejo (2006), and Gutiérrez-Rexach (1996, this volume).
One nice argument comes from degree exclamatives associated with adjectives
denoting a closed scale, like full or dry (Castroviejo 2006). Crucially,
(6) How dry the towel is!
does not entail that the towel is dry (cf. wet: How wet the towel is! ⇒ “The towel
is wet’’), which makes the extreme degree interpretation (“The towel is dry to an
