69.

Existential sentences 1829

Ross, John Robert 1972. Act. In: D. Davidson & G. Harman (eds.). Semantics of Natural Language.
Dordrecht: Reidel, 70–126.
Rothstein, Susan 1995. Small clauses and copula constructions. In: A. Cardinaletti & M. T. Guasti
(eds.). Small Clauses. San Diego, CA: Academic Press, 27–48.
Rothstein, Susan 2001. Predicates and their Subjects. Dordrecht: Kluwer.
Russell, Bertrand 1919. Introduction to Mathematical Philosophy. London: Allen and Unwin.
Schlenker, Philippe 2003. Clausal equations (a note on the connectivity problem). Natural Lan-
guage and Linguistic Theory 21, 157–214.
Sharvit, Yael 1999. Connectivity in specificational sentences. Natural Language Semantics 7, 299–339.
Sharvit, Yael 2003. Tense and identity in copular constructions. Natural Language Semantics 11,
363–393.
Williams, Edwin 1983. Semantic vs. syntactic categories. Linguistics & Philosophy 6, 423–446.
Zimmermann, Thomas E. 1993. On the proper treatment of opacity in certain verbs. Natural Lan-
guage Semantics 1, 149–179.

Line Mikkelsen, Berkeley, CA (USA)

69. Existential sentences

1. What is an existential sentence?
2. The existential proposition
3. Analyses of the definiteness restriction
4. Analyses of the predicate restriction
5. Conclusion
6. References

Abstract
The term ‘existential sentence’ is used to refer to a specialized or non-canonical construc-
tion which expresses a proposition about the existence or the presence of someone or
something. Because of their special structural and interpretive characteristics, existential
sentences have offered a rich ground on which to test theories concerning the semantics of
noun phrases and of predication, as well as theories concerning the role of non-canonical
constructions in information packaging. This chapter begins by reviewing the basic struc-
tural, semantic and discourse functional properties of existential sentences. Since, across
languages, existential sentences resemble copular, possessive and locative sentences, consid-
erable debate has arisen about the extent to which their semantics are similar. The chapter
therefore continues with an overview of the different analyses that have been proposed for
the core existential proposition. The remainder of the chapter is devoted to two distinctive
features of these sentences which have generated substantial discussion in the semantics
and pragmatics literature: 1) the so-called definiteness restriction, which limits the ability of
definite and quantificational nominals to appear as the ‘pivot’ of the construction; and, 2)
the predicate restriction, which has been claimed to restrict the expressions that can appear
as the ‘coda’ to so-called stage-level predicates.

von Heusinger, Maienborn and Portner (eds.) 2011, Semantics (HSK 33.2), de Gruyter, 1829–1848
1830 XIV. Sentence types

1. What is an existential sentence?

The term ‘existential sentence’ is used to refer to a specialized or non-canonical con-
struction which expresses a proposition about the existence or the presence of someone
or something. Thus, the sentence in (1a) is considered existential because it is specialized
(insofar as it has an expletive subject, whose distribution is highly restricted in English)
and entails nothing other than the existence of one even prime number. In contrast, (1b)
will not be considered existential for the purposes of this article because, even though it
entails nothing other than the existence of one even prime number, there is nothing spe-
cialized about its syntax: it has the canonical subject-predicate structure used in English.
Finally, (1c), though closely related to the construction in (1a), differs from it in syntactic
details, in its use in discourse, and in the fact that, thanks to the verb, it entails something
more than mere existence or presence. This latter type of sentence, an example of the
‘presentational-there’ construction (see e.g., Aissen 1975), will not concern us further
here.

(1) a. There is one even prime number.

b. One even prime number exists.
c. There stood in the corner an empty coat rack and umbrella stand.

Existential sentences vary somewhat in structure, semantics, and pragmatics from

language to language, though they generally share certain characteristics as well. This
article will include examples from a variety of languages, though the discussion of their
semantics and pragmatics will be based primarily on English.

1.1. Structural properties

Syntactically, existential sentences typically manifest most or all of the following five
characteristics (see Francez 2007 for a particularly clear discussion and examples). First,
an expletive subject, like English there or French il in (2a–b), may be required; however,
in languages without expletives no such subject appears, as illustrated by the Catalan and
Maori examples in (2c–d). Note that hi in (2c) is a locative clitic equivalent to French y
in (2b):

(2) a. There are bugs eating the corn.

b. Il y a eu une reunion.
it loc has been a meeting
‘There has been a meeting.’

c. Hi ha espai a l′ armari.
there has space to the closet
‘There is space in the closet.’

d. Āe he taniwha.
yes a taniwha
‘Yes, there are taniwhas.’ (from Bauer 1993, cited in Chung & Ladusaw 1994: 42)
69. Existential sentences 1831

Second, though existential sentences do not always contain a verb (see e.g., (2d)), if there
is one it is often homophonous with a verb meaning ‘to be’ (2a) or ‘to have’ (2b–c), or
with some other verb related to possession (such as geben ‘give’ in German in (3a)) which
is ‘bleached’ of its content; it may also be a special lexical item such as Hebrew yeS (3b).

(3) a. Es gab ein Kind in dem Garten.

it gave a child in the garden
‘There was a child in the garden.’

b. yeS harbe tisot ad xacot.

EX many flights until midnight
‘There are many flights until midnight.’ (Francez 2007: 60)

Curiously, the survey in Clark (1978) indicates that the verb in existential sentences is
rarely, if ever, homophonous with a verb that literally means ‘exist’, although in English
one does find sentences like (4):

(4) There exists one even prime number.

However, Aissen (1975), building on Ross (1974), has argued that such sentences, which
are also attested with a limited set of verbs including remain, ensue, and follow, do not
share all of the properties of existential sentences with be, but rather manifest some
characteristics of the latter and some of the so-called ‘presentational-there’ construction
illustrated in (1c).
Third, in all existential sentences there is a ‘pivot’ nominal which describes the indi-
vidual whose existence is under discussion (bugs in (2a), une reunion in (2b), etc.).
The pivot is typically subject to certain semantic restrictions which will be discussed in
section 3.
Fourth, in most languages, a ‘coda’ phrase may appear (such as eating the corn in (2a)),
which is external to the pivot noun phrase. As with the pivot, only predicates meeting
certain semantic conditions are licensed as codas, as will be discussed in sections 1.2
and 4.
Finally, in many, though by no means all, languages, a locative expression appears
which may be obligatory and ‘bleached’ of content (e.g., there, y, hi), though the syntactic
role this expression plays in the construction may vary from one language to another.
For instance, in English it is the subject, while in French and Catalan it is not. The pres-
ence of such a locative expression has resulted in the frequent grouping of existential
sentences together with locative constructions (see e.g., Lyons 1967; Kuno 1971; Kimball
1973; Clark 1978; Freeze 1992; Rigau 1997; Zeitoun et al. 1999 and several more recent
references cited in Francez 2007), although others (e.g., Milsark 1974) have argued that
the similarity between existential and locative sentences is superficial or spurious, at least
in some languages.
A crucial and very difficult question is how all of these elements in an existential
sentence are related syntactically and semantically. On this point there has been con-
siderable disagreement among linguists. Existential sentences are often truth condi-
tionally equivalent to copular sentences ascribing a property or location (see article 68
1832 XIV. Sentence types

(Mikkelsen) Copular clauses for more on such sentences). For example, (2a) is
synonymous with (5), though it differs in the contexts in which it is used.

(5) Bugs are eating the corn.

For this reason, many analyses have related existential and copular sentences derivation-
ally or have attributed to them the same basic predicational structure, in which the pivot
nominal serves as the argument to the coda phrase, which in turn effectively serves as
the main predicate for the sentence – this was the intuition behind the ‘There-insertion’
transformation in transformational grammars (e.g., Burt 1971; Milsark 1974; see e.g.,
Stowell 1978; Safir 1985; Pollard & Sag 1994 for related analyses for English).
However, other linguists have treated the similarity between existential and copular
sentences as accidental. For example, Barwise & Cooper (1981), Chung (1987) (for
Chamorro), and McNally (1992) argue that the pivot is the only complement to the exis-
tential predicate. On such analyses the coda either has to be treated as a pivot-internal
modifier (as assumed by Barwise & Cooper), or else as some sort of adjunct. McNally
(1992) specifically proposes that those codas which cannot be plausibly treated as pivot-
internal modifiers should be analyzed as predicative adjuncts similar to the depictive
adjunct in (6).

(6) They ate the vegetables steamed.

Finally, Williams (1984, 1994), Hazout (2004), and Francez (2007) all argue that the pivot
nominal is in fact the main predicate of the existential sentence while the coda serves
as a pivot-internal modifier or as an adjunct. What constitutes the subject of the exis-
tential predication is a matter of disagreement: Williams and Hazout argue that it is the
expletive, though they do not provide an explicit semantics for this predication rela-
tion; Francez argues that the pivot denotes a property of an implicit contextual domain
(something intuitively similar to a location), whose identity can be restricted by the
coda phrase. The variety in the proposals for the basic syntax and argument structure of
existential sentences makes it difficult to directly compare the semantic analyses that
have been proposed in the literature for what Francez refers to as the ‘existential propo-
sition.’ Nonetheless, a brief overview and comparison of these analyses is provided in
section 2.

1.2. Semantic and discourse functional properties

Although it is unlikely that one single semantics and discourse function can be assigned
to existential sentences cross-linguistically, certain semantic and discourse functional
properties are consistently associated with these sentences across languages. Perhaps
the most important of these is the intuition that existential sentences serve primarily to
introduce a novel referent into the discourse – one fitting the description provided by
the pivot nominal. This function is, in turn, almost certainly related to two other charac-
teristics commonly attributed to existential sentences. The first of these is the so-called
definiteness restriction on the pivot nominal (an early exhaustive description of which
appears in Milsark 1974), to be discussed in section 3. Though the exact characterization
of the facts is complex, as a first approximation the definiteness restriction amounts to
69. Existential sentences 1833

a restriction on the acceptability of definite, demonstrative, and necessarily quantifica-

tional noun phrases, including proper names and personal pronouns, in the pivot (see
Chapter IX for more details on these various kinds of noun phrases):

(7) a. ??There is each/every first-year student present.

b. ??There are most first-year students in that class.
c. ??There is the neighbor’s dog barking.
d. ??There is that carpet under the table.
e. ??There are them / Anna and Bob waiting outside.

As will be discussed in section 4, Milsark related the definiteness restriction to another

semantic restriction which is evident in existential sentences in English and other lan-
guages. Specifically, the coda has to be what Carlson (1977a) referred to as a stage-level
predicate, as exemplified in (8); so-called individual-level predicates are not licensed, as
shown in (9). Though the precise definition of what counts as an individual- or stage-level
predicate is a difficult matter which cannot be dealt with here, stage-level predicates
tend to describe accidental properties, while individual-level predicates tend to describe
essential ones.

(8) a. There were many people in line already.

b. There were a few people waiting for hours.
c. There was a live pig roasted.
d. There were no taxis available.

(9) a. ??There were many students anarchists.

b. ??There were lots of donors generous.
c. ??There were inmates psychotic.

It is also often claimed of existential sentences that they are thetic (term due to the phi-
losophers Brentano and Marty; see Kuroda 1972; Sasse 1995 for modern discussions),
purely rhematic, or topicless; or, alternatively, if they are assumed to have a topic-
comment or theme-rheme structure, the topic or theme is hypothesized to be a location
rather than the referent of the pivot (see e.g., Babby 1980 for Russian; see also Kim
1997, and Leonetti 2008 for Romance and a general survey): Since the only obligatory
nominal in an existential sentence is the pivot, and the pivot introduces a novel referent
(whether directly or indirectly), there is no other candidate expression except perhaps
the coda that can provide the topic of the sentence or serve as the subject of a categorical
(logical subject-predicate) or theme-rheme propositional structure. Note that the failure
of the pivot to serve as a topic would also follow directly if the pivot were in fact the
main predicate of the existential sentence. (See Chapter XV for more on notions such as
theme, rheme, and topic.)
These semantic and pragmatic generalizations, though frequent, are not entirely
unproblematic, however. For example, Abbott (1992, 1993, 1997) and Ward & Birner
(1995) have argued that existential sentences serve not only to introduce novel discourse
referents but also to reintroduce or focalize referents that have already been mentioned.
Consider, for example, (10) (from the Challenger commission transcripts, cited in Ward
& Birner 1995: 727):
1834 XIV. Sentence types

(10) I think there was one flight where

we had one problem. It wasn’t ours,
but there was that one flight.

In this example a demonstrative noun phrase appears in the pivot position which
is directly anaphoric to an expression that appears in the previous sentence. A more
familiar example is provided by the so-called list use of existentials (see e.g., Rando &
Napoli 1978), illustrated in (11), in which proper names are often found:

(11) A: Who showed up?

B: Well, there was Alex.

Such examples raise the question of whether a uniform semantics and discourse function
can be given for everything that looks formally like an existential sentence, or whether in
reality there are several subtypes of existential sentence, perhaps with distinct semantics
and pragmatics.
The claim that the pivot cannot be a topic is similarly questionable. As Leonetti (2008:
Fn 21) points out, existential sentences in which the pivot is topicalized are attested; the
English, Catalan and Spanish examples in (12a), (12b), and (12c), respectively, illustrate.
In the Catalan example, the presence of the partitive clitic n′ (glossed as ‘some’) is ana-
phoric to pintura and marks the sentence clearly as a clitic left-dislocation construction.
In such constructions the left-dislocated expression has been argued to be topical (see
e.g., Vallduví 1992):

(12) a. They told us there was a solution, and indeed a solution, there was.

b. Pintura, n′ hi ha dins l’armari.

paint some there has in the closet
‘Paint, there is in the closet.’

c. Ardillas, hay en el bosque.

squirrels have in the woods
‘Squirrels, there are in the woods.’

Understanding such counterexamples in the face of the strong intuitions concerning the
discourse referent introduced by existential sentences and the information structure of
these sentences entails first making a decision as to the basic semantics of existential
sentences, a difficult issue to which we now turn.

2. The existential proposition

The space of possibilities for the basic semantics of existential sentences includes various
options, five of which will be mentioned here. On the first option, proposed in Barwise &
Cooper (1981), an existence predicate serves as the main predicate. The pivot is analyzed
as a generalized quantifier and thus takes the existence predicate as its argument, as in
69. Existential sentences 1835

(13b) for the sentence in (13a) (though one might also consider treating the pivot as
non-quantificational when possible and having the existence predicate take it as its argu-
ment, as in (13c), where the pivot is analyzed for the purposes of illustration as denoting
the value of a choice function, along the lines of Reinhart () or Kratzer (1998); see also
Chapter IX).

(13) a. There is one good answer.

b. (one(λx[answer(x) ∧ good(x)]))(λx[exist(x)])
c. exist(f(λx[answer(x) ∧ good(x) ∧ |x| ≥ 1]))

Either way, such an analysis runs into trouble in cases where the coda phrase cannot be
plausibly treated as part of the pivot (see Keenan 1987 for arguments that such cases do
in fact exist). However, Zucchi (1995) suggests a variant of this analysis on which the
coda serves to contextually restrict the domain of quantification of the pivot, eliminating
this criticism concerning the syntactic and semantic function of the coda.
The second option involves treating the coda, rather than an existence predicate, as
the main predicate. Again, the pivot could in theory be treated either as a generalized
quantifier (following e.g., Keenan 1987, see (14b)) or as non-quantificational (14c).

(14) a. There was a room available.

b. (a(λx[room(x)]))
(λx[available(x)])
c. available(f(room))

Such an analysis effectively equates existential sentences semantically with copular

sentences. However, Keenan (1987) adds the proposal that what makes existential sen-
tences special is the possibility of what he calls an existential reading, formulated as in
(15), where ‘1’ is the universal property, that is, a property all individuals have, roughly
analogous to the property of existing:

(15) 1 ∈ Det(NP ∩ XP)(Keenan 1987: 301, minor details modified)

Keenan observes that an existential sentence with a logical form like that in (14b) will
have an existential reading just in case the determiner in the pivot has the property
of being what he calls existential. Keenan’s notion of an existential determiner will be
discussed in section 3, as it plays a role in his account of the definiteness restriction.
Among the criticisms that have been leveled against analyses like that in (14),
two stand out. First, there is evidence that the coda is not the main predicate. For
example, Williams (1984) argues that if the coda were the main predicate, it should be
able to extract like a main predicate. But as the contrast between (16a–b) shows, it does
not.

(16) a. How sick were the children?

b. *How sick were there the children?

Such reduced acceptability under extraction is characteristic of adjuncts.

1836 XIV. Sentence types

McNally (1992: Ch. 2) offers an additional argument against treating the coda as the
main predicate, based on asymmetries between extraction from within complements vs.
adjuncts. Huang (1982) showed that extraction of an adjunct from within a complement
or main predicate is possible, whereas extraction of an adjunct from within an adjunct
is not. (17), which shows the contrast in acceptability in the extraction of a manner
adverbial from a main predicate vs. a depictive adjunct, illustrates this asymmetry.

(17) a. How badly was she coughing _?

b. *How badly did you leave the concert hall coughing _?

The impossibility of such extraction out of the coda indicates that it is an adjunct, and
not the main predicate.

(18) *How badly were there people coughing _ in the audience?

As yet another argument against treating the coda as the main predicate, Francez (2007:
Ch. 4) observes that quantificational expressions within the coda always take scope over
the pivot, while quantificational predicates in main clauses need not take wide scope
over their subjects; contrast (19a) and (19b):

(19) a. Some drummer I know is in every rock band.

b. There is some drummer I know in every rock band.

While the scope facts follow on Francez’ analysis of the coda as a modifier, this argument
is weakened by the fact that the pivot can be independently shown to take narrowest
scope (see section 3, below).
A second criticism of analyses like that in (14) is that there are a number of types of
existential sentences that do not have acceptable copular sentences as counterparts (see
e.g., Kimball 1973 and the examples in (20)).

(20) a. There was space in the room.

b. ??Space was in the room.
c. There was a fire in the school.
d. ??A fire was in the school.

If the compositional semantics for the contrasting pairs in (20) is the same, it is difficult to
explain why the existential sentences are acceptable while the copular sentences are not.
A third possible analysis for the existential proposition is that proposed in Milsark
(1974, 1977), on which the existential predicate contributes an existential operator; the
pivot denotes a property that serves as its restriction (with any determiner that might
be present serving as a cardinality predicate); and the coda serves as the scope of the
existential quantifier. Milsark’s semantic ‘E rule’ is as follows (1974: 206):

(21) there AUX (have-en) be Q NP X is interpreted: the class C denoted by NP has at

least one member c such that P(c) is true, where P is a predicate and P is the reading
of X and the set of such members c is of cardinality Q.

See Landman (2004) for a variant on this property-based analysis of the pivot.
69. Existential sentences 1837

The analysis in McNally (1992), closely related to Milsark’s, constitutes a fourth pos-
sibility for the existential proposition. This analysis differs from Milsark’s on three points.
First, the pivot is analyzed not as a property looking to be bound by an existential oper-
ator, but rather as the entity correlate of a property (Chierchia 1984) – the reification
of a property that allows it to serve as the argument to another predicate (just as the
nominalization goodness of the adjective good allows us to say things like Goodness is a
virtue). Second, and relatedly, the main predicate in the sentence is the property of being
instantiated, which applies to the pivot’s denotation. Though these two aspects of the
analysis might appear to be simply notational variants of Milsark’s analysis, see McNally
(2009) for arguments that they make distinct empirical predictions. Finally, the coda, as
mentioned above, is analyzed as a VP-internal adjunct modifier which stands in a con-
trol relation to the pivot and which serves to restrict the spatiotemporal index at which
the main predicate holds by forcing it to be included in the time the adjunct predicate
holds of its argument. The semantics for (14a) on this analysis is sketched in (22), where
(i) constitutes the core existential proposition, and (ii) and (iii), the ultimate effect of the
adjunct predicate (see McNally 1992: Chapter 4 for a slightly different formulation and
further details).

(22) (i) instantiate(∩λx[room(x)]);

(ii) the token individual y who supports the truth of (i) at a time t is in the
extension of available at t′ ; and
(iii) t ≤ t′ .

Milsark’s and McNally’s analyses share the fact that they remove any quantificational
force from the pivot. Though both argue that this leads to a natural account of the defi-
niteness restriction (see below), the proposal to treat the pivot as a property can be criti-
cized. For example, the semantics proposed in (21) and (22) initially runs into difficulties
when the pivot contains a monotone decreasing determiner such as no, few, or at most
two, or a non-monotone one like exactly three (see article 43 (Keenan) Quantifiers). Take
(23) as an example:

(23) There was exactly one cookie left.

If exactly one is treated as a cardinality predicate, (21) and (22) predict that (23) will be
true not only when there is exactly one cookie left, but also when there is more than one
left: (21) and (22) only require finding an individual cookie with a cardinality of exactly
one, but fail to make the sentence false if more than one such individual can be found. To
solve this problem, McNally (1998) suggests a decompositional analysis for no and few
(equivalent to a sentential negation plus a and many, respectively), following a proposal
in Ladusaw (1992) (see also Chapter XIII); she follows Krifka (1999) in arguing that at
most and exactly should be factored out from the semantics of the pivot and treated as
focus-sensitive operators whose semantics is incorporated at the clause level. Francez
(2007) points out that such an analysis is inelegant, though he provides no conclusive
empirical arguments against it. Francez also observes (as did McNally) that it is diffi-
cult on the analysis in (22) to treat the coda as a controlled adjunct predicate in a fully
compositional fashion.
1838 XIV. Sentence types

Francez’s (2007) alternative constitutes a fifth proposal for the existential proposition.
Francez argues that the pivot – and not an existence predicate or the coda – should be the
main predicate of the existential proposition. However, he also maintains that the pivot
should denote a generalized quantifier, that is, a property of sets. This leads him to con-
clude that the logical subject of the existential proposition must be a set, rather than an
individual. Specifically, he proposes that this set is a contextual domain – a set of salient
entities in the context. Hypothesizing that the subject of the existential proposition is a
contextual domain is way of capturing the intuition that existential sentences are ‘about
the context.’ This domain, as its name suggests, is not directly denoted by any specific
constituent in the existential sentence; however, the coda can contribute to restricting
its identity. Francez’s definition of contextual domains appears in (24). It defines this
domain as a set of individuals who stand in a contextually specified relation R to another
individual α.

(24) For every element α of type τ, let dα be the contextual domain of α, where dα = def
λyτ′[R<τ′,<τ, t>>(α,y)]

Francez’s analysis of a sentence like (25a) is thus as in (25b). The pivot contributes the
negative existential quantifier, the rest of the sentence contributes nothing else, and thus
the quantifier is applied to the contextually-supplied domain dα.

(25) a. There was no coffee.

b. λP<e,t>[no(λx[coffee(x)],P)](dα)

What might dα be? It must be a set of individuals, in order to serve as the scope of no
coffee. And by (24) it will have to be a set that stands in contextually-determined relation
R to some type of object. Francez (2007: 74–75) suggests for this particular example that
α is the spatio-temporal parameters of utterance, and R is the relation of being located
within those parameters. Thus, (25b) ends up being equivalent to (26), which can be para-
phrased as saying that no coffee has the property of being in the set of things located
within the spatiotemporal parameters of the utterance (represented in (26) as stu).

(26) no(λx[coffee(x)],λy[loc(stu,y)])

When there is an explicit coda, the analysis treats the coda as a modifier which as a rule
serves to restrict the relation R.
(25b) has the virtue of treating the pivot as the main predicate in a technical sense, an
analysis for which there is increasing syntactic support. However, it is not at all usual for
the subject of a predication to be a set and, moreover, to effectively serve as the nuclear
scope of a quantifier. It is therefore not clear to what extent this analysis really preserves
the spirit of the claim that the pivot is the main predicate. On the other hand, unlike
Keenan’s analysis it manages to treat the coda as a modifier, and it avoids the criticisms
that have been made of property-based analysis of the pivot.
As can be seen, each of these analyses has pros and cons. Given the comparatively
little content of existential sentences, it might even be the case that different languages
choose different options for expressing the existential proposition (see McNally 2009 for
examples and discussion of this possibility).
69. Existential sentences 1839

3. Analyses of the definiteness restriction

The definiteness restriction, illustrated in (7), is the characteristic of existential sentences
that has received greatest attention in the semantics and pragmatics literature. In addi-
tion to the restrictions on necessarily quantificational noun phrases and certain defi-
nite and demonstrative noun phrases (including personal pronouns and proper names),
Milsark claimed that there are restrictions on the indefinite noun phrases that are
licensed as well, which limit the acceptability of partitives (including covert partitives,
identifiable by stress on the determiner) and generically interpreted indefinites, though
Hoeksema (1989) observes that the restriction on partitives holds for English only when
there is no coda phrase.

(27) a. ??There are five of the prime numbers less than 10. (cf. Five of the prime num-
bers are less than 10.)
b. ??There were SOME teachers on strike, but not others.
c. ??There are dinosaurs extinct. (cf. Dinosaurs are extinct.)

Milsark (1977) referred to the noun phrases/determiners excluded from the pivot posi-
tion as strong, and to those licensed in the position as weak. An important aspect in the
development of subsequent analyses of the definiteness restriction has been the attempt
to arrive at a more precise and empirically adequate characterization of the weak/strong
distinction.
A number of different syntactic, semantic and pragmatic analyses of the definiteness
restriction have been proposed. Since space precludes reviewing all of these here, only
the most representative of these analyses will be reviewed. Strictly syntactic analyses of
the restriction will not be discussed, as the facts clearly indicate that even these analyses
must ultimately rely on semantic or pragmatic notions. For example, Milsark himself,
Woisetschlaeger (1983), Holmback (1984), and others have observed that morphologi-
cally definite noun phrases such as those in (28) are perfectly acceptable in existential
sentences when they are ostensibly semantically indefinite.

(28) a. There was the most amazing painting in their collection!

b. There was the mother of a student waiting outside.

Thus, no analysis that appeals exclusively to the form of the pivot will account for the
definiteness restriction.
To an important extent the divergence in the analyses of the restriction reflect
divergence in the starting assumptions concerning the data. These divergences will be
introduced as the different analyses are reviewed.
Although Milsark (1974, 1977) did not provide a fully formalized, compositional
semantic account of the definiteness restriction, the essence of his proposal was clear.
If we posit that the existential predicate contributes an existential operator, we would
expect the pivot to license only those expressions whose denotations could combine with
such an operator. In other words, the pivot has to denote a property. Pivots containing
necessarily quantificational determiners such as each and every should be excluded
because they are already quantificational. The operator contributed by such determiners
will bind any variable within the logical form for the pivot. In contrast, Milsark suggested,
1840 XIV. Sentence types

weak indefinite noun phrases are systematically licensed as pivots because indefinite
determiners can be analyzed as cardinality predicates, making the noun phrase amenable
to a property-type analysis. Heim (1987: 23) characterizes the restriction in similar, if
negatively defined terms:

(29) *There be x, when x is an individual variable.

(29) would exclude, for example, the bound variables left behind by quantifier raising.
Heim also argues that the generalization in (29) can account for another fact that has
been related to the definiteness restriction, namely that when the pivot is relativized,
only a so-called amount relative reading appears to be available.
Amount relative clauses can be distinguished from ordinary relative clauses both
syntactically and semantically. The semantic difference is that the amount relative
describes a quantity rather than a specific individual. This is evident in examples like
(30), an adaptation of one of Heim’s amount relative examples:

(30) It would take days to drink the champagne they spilled that evening.

The relative clause in (30) refers to the quantity of champagne that was spilled, not to the
exact same liquid that was spilled. Turning to syntax, amount relatives are different from
ordinary relatives in that they require a definite or universal determiner on the head
noun they modify, and can only be introduced by that or a null relative pronoun in some
dialects of English (see Carlson 1977b). Note that the amount reading of (30) disappears
if the determiner is not universal or the relative pronoun is changed to which:

(31) a. It would take days to drink some champagne they spilled that evening.
b. It would take days to drink the champagne which spilled that evening.

Heim observes that relativization of the pivot syntactically resembles amount

relativization, as the examples in (32b–c) show:

(32) a. You’ve eaten every cookie there is in the house.

b. *You’ve eaten some/two/many cookies there are in the house.
c. *You’ve eaten every cookie which there is in the house.

She argues that the contrast between examples like (32a) and (32b–c) follows if in
ordinary relativization of the pivot an individual variable is left behind in the position
of the gap; she suggests that in amount relativization, the variable left behind does not
correspond to the gap but rather to a subpart of the gap. Thus, the logical forms for the
noun phrases containing the amount and restrictive relatives in (32a) and (32c) could be
informally represented as in (33a) and (33b), respectively. Heim suggests that the uni-
versal quantifier in the amount relative is interpreted as a maximality operator over
degrees d:

(33) a. Max d: there were d-many cookies in the house

b. every x: cookie x ∧ there was x in the house.
69. Existential sentences 1841

Despite the initial appeal of this proposal, there is one important respect in which relativ-
ization of the pivot differs from amount relativization: the former imposes an identity of
individuals requirement. That is, in order for (34) to be true, it is not enough that I read a
quantity of books identical to the quantity on the table in question; I must have read the
exact same books that were on the table.

(34) I read all the books there were on the table.

Grosu & Landman (1998) propose a semantics for amount relatives which attempts
to account for this fact. However, McNally (2008) argues that there are problems with
Grosu & Landman’s analysis and that, despite the superficial similarities between amount
relatives and relativization of the pivot, we cannot conclude that the latter is necessarily
amount relativization. Nonetheless, McNally fails to provide an alternative analysis
for all of the facts, and relativization of the pivot remains an understudied and poorly
understood phenomenon.
Returning to the definiteness restriction, what exactly Milsark’s (or Heim’s) basic
account of the restriction predicts for proper names, pronouns, definites, and strong (e.g.,
partitive) indefinites depends on the assumptions one makes about the semantics of
these latter expressions. Partee’s (1987) theory of noun phrase type shifting leads one to
expect that all of these expressions should be able to shift to well-defined and felicitous
property-type denotations, for example those in (35):

(35) a. [[Anna]] = λx[x = a]

b. [[the dog]] = λx[x = ιy[dog(y)]]
c. [[themi]] = λx[x = yi]

But if such denotations are available, we might expect these sorts of noun phrases to be
acceptable in existential sentences.
In fact, as noted in section 1.2, there is good evidence that true definite noun phrases
are indeed acceptable (see, in addition to the references cited elsewhere in this article,
Ziv 1982 and Lumsden 1988). But let us maintain for a moment the view that they are
not, or at least that such noun phrases are attested only with an alternative kind of inter-
pretation, such as the presentational interpretation illustrated in (1c), or the list interpre-
tation illustrated in (11). Such a view is assumed in both Barwise & Cooper’s (1981) and
Keenan’s (1987, 2003) accounts of the restriction.
Both Barwise & Cooper and Keenan assume that noun phrases as a rule denote gen-
eralized quantifiers. They take as their first task a proper semantic characterization of
those quantifiers that appear in (English) existential sentences vs. those that do not – in
other words, a more precise account of the weak/strong distinction. Barwise & Cooper
(1981: 264) begin by defining strong and weak determiners as in (36).

(36) A determiner D is positive strong or (negative strong, resp.) if for every model M =
〈E, [[ ]]〉 and every A ⊆ E, if the quantifier [[D]](A) is defined, then A ∈ D(A) (or A
∉ D(A), resp.). If D is not (positive or negative) strong then D is weak.

A quantifier is strong if it is headed by a strong determiner. However, other factors

may make a quantifier strong. Even though proper names lack a determiner altogether
1842 XIV. Sentence types

in English, they always behave like quantifiers constructed out of strong determiners;
similarly, partitive noun phrases, even though they may contain weak determiners, also
behave as if strong. In both cases, this behaviour can be shown to be follow from the
existence presupposition associated with the noun phrase.
Barwise & Cooper’s explanation of the definiteness restriction is pragmatic: Under
the assumption that the existential proposition is of the form in (13b), existential sen-
tences with strong nominals as pivots will be tautologous or contradictory, but never
contingent, and thus will be systematically uninformative. However, this account of the
restriction falls apart if the coda is external to the pivot, as Keenan (1987) and others
have argued: Even if it is not new information that the denotation of the pivot exists, the
ascription of the coda property to the pivot’s denotation could well be informative.
Keenan’s characterization of both the class of noun phrases licensed in existential
sentences and his account of the definiteness restriction are slightly different. Rather
than claiming that definite and universal noun phrases are unacceptable as pivots, he
simply assumes that they lack the existential reading in (15), above. Keenan (1987: 291)
defines the class of determiners that yield noun phrases with the existential reading as
existential (see (37a)); Keenan (2003) proves that the existential determiners are equiva-
lent to those defined as intersective as in (37b). The intersective determiners, in turn,
are just those that are conservative on both their first and second arguments (see
Barwise & Cooper 1981 on the notion of conservativity; note that Keenan 2003 also
generalizes these definitions to n > 1-place determiners and Boolean combinations of
determiners).

(37) a. A function f from properties to sets of properties is existential iff for all
properties p, q, p ∈ f(q) iff 1 ∈ f(q ∧ p).
b. [A function denoted by a determiner] D from PE into GQE,X is intersective iff
for all subsets A, A′, B, B′ of E, if A ∩ B = A′ ∩ B′ then DAB = DA′B′.

Intuitively, an existential determiner is one whose semantics can be calculated without

making a comparison between sets, as is necessary with, for example, proportional deter-
miners such as most. To know whether e.g. each student in a given class read Hard Times
it is necessary to compare the set of students in the class who read Hard Times with the
set who did not, or with the set of students as a whole. In contrast, in order to determine
the truth of a proposition involving an intersective determiner such as two, as in Two
students read Hard Times, we need only identify the set of students who read Hard Times
and determine its cardinality. This makes the contribution of an intersective determiner
more like that of a property than like that of a true quantificational operator.
However, the facts are not so simple. As Lumsden (1988) observes, necessarily quan-
tificational noun phrases are perfectly acceptable in existential sentences when they
quantify over kinds, as in (38a), or similar higher-order objects, as in (38b), where what is
asserted is the existence of an instance of every one of a set of types of reasons:

(38) a. There was each of the three kinds of chocolate available.

b. There was every reason to believe that it would rain.

Although Wilkinson (1991) suggests accounting for examples like (38a) by reanalyzing
noun phrases of the form D kind of N as equivalent to an N of D kind, McNally (1992)
69. Existential sentences 1843

points out various ways in which these two types of noun phrases are not equivalent, thus
casting doubt on the viability of such a reanalysis.
Still, the acceptability of the noun phrases in (38) and the intuition that they are
somehow ‘covertly indefinite’ raises the possibility that the definiteness restriction is not
semantic but rather pragmatic in nature. Pragmatic analyses of the restriction fall into
two general groups. One group (e.g., Prince 1992; Ward & Birner 1995; Abbott 1992,
1993; Zucchi 1995) attribute the restriction to a condition on the discourse status of the
referent associated with the pivot. For example, Prince (1992) argues that the referent of
the pivot has to be ‘hearer-new,’ i.e., not part of the common ground at the time of utter-
ance. Ward & Birner (1995) argue that the notion of hearer-newness can be extended to
account even for cases of definite noun phrases in pivot position; specifically, such noun
phrases are licensed just in case there it is possible to construe the referent of the noun
phrase as hearer-new in the context (e.g., because it has been forgotten about by the
hearer). However, Abbott (1997) contends that it is difficult to maintain that the referent
of the pivot is hearer-new particularly in cases where the pivot is anaphoric, such as in
(10), above, and suggests that ‘it may not be the case that any single discourse-based
principle can account for the distribution of NPs in this construction.’ (1997: 107).
The second general group of pragmatic approaches to the definiteness restriction
builds in one way or another on the non-topical nature of the pivot position, often taking
into account that the existential construction in most languages exists in a paradigm
with one or more other constructions that are conventionally associated with a different
information structure (see Hannay 1985; Borschev & Partee 2002; Mikkelsen 2002;
Beaver et al. 2005; Francez 2007; Hu & Pan 2007; Partee & Borschev 2007). The intuition
behind these proposals is that indefinite noun phrases often make poor subjects, particu-
larly when there is a strong association between subject and topic in a language, while
definite noun phrases make good ones. Existential sentences in many languages serve to
get out of subject position a noun phrase which would otherwise have to be expressed as
a subject (compare e.g., (20a–b), above). Part of the argument for this view comes from
the observation, developed in detail in Beaver et al. (2005), that the cross-linguistic varia-
tion in the definiteness restriction is gradient rather than absolute. Beaver et al. present
a quantitative study which, rather than dividing noun phrases into two groups (weak vs.
strong), orders them on a scale according to how many subject properties they manifest.
They argue that this same scale can be used to predict the variation in the definiteness
restriction across languages. Existential sentences in different languages may be sensitive
to different points on the scale, but it should always be the case that if a language allows
a given noun phrase candidate for canonical subject in the pivot position, it should also
allow all worse candidates for canonical subject to appear as pivots as well. Similarly, if a
given noun phrase type is blocked from pivot position, all better candidates for canonical
subject on the scale should be blocked as well.
Finally, in contrast to the above-mentioned proposals, which offer a unified account
for the definiteness restriction, McNally (1992) argues that the restriction is part semantic
and part pragmatic in nature. Her analysis is grounded in the claim in (22) that the exis-
tential predicate selects for the entity correlate of a property. On this analysis, any noun
phrase that can plausibly be treated as an entity correlate of a property or as a quantifier
over entity correlates of properties is expected to have a well-formed interpretation in
the pivot position. If the notion of ‘weak’ is taken to describe the semantic restriction
on the type of the pivot and if the semantics in (22) is recast in set-theoretic terms (as in
1844 XIV. Sentence types

McNally 1998), the proposal also leads naturally to the definition of ‘weak’ as ‘property
denoting’ (see Ladusaw 1994 for a development of this idea).
When coupled with Partee’s theory of type shifting, McNally’s proposal directly
accounts for the contrast between examples like (38) and similar examples, such as (7a),
which quantify over ordinary token individuals. The latter are excluded because they
neither quantify over entity correlates of properties nor are able to type shift felici-
tously to denote entity correlates of properties (see McNally 1998 for details on which
kinds of noun phrases do not have a felicitous property-type denotation and why they
do not). However, since definites, demonstratives, partitive indefinites, proper names,
and pronouns can all felicitously shift to a property type under Partee’s theory and
then can shift to the entity correlate of that type, they are predicted to have well-formed
interpretations in existential sentences, which, as noted above, they do. To account
for the oddness of existential sentences with these kinds of noun phrases (when they
are in fact odd), McNally adopts Prince-style pragmatic account. However, a mixed
semantic-pragmatic account of the restriction would also be possible using a Beaver
et al.-style analysis of the conditions on definites instead of one based on hearer-
newness.
Though a non-unified analysis of the definiteness restriction might seem less attrac-
tive than a unified one, McNally argues that it predicts that the patterns of variation found
in the restriction cross-linguistically will reflect the split between those noun phrases
which are excluded in English for semantic reasons and those which are excluded for
pragmatic reasons. This prediction is confirmed, for example, in Catalan, which system-
atically allows definites, demonstratives, names and pronouns in existential sentences,
but allows only necessarily quantificational noun phrases that quantify over kinds or
similar higher order objects.
The mixed semantic-pragmatic analysis also predicts an asymmetry in the scopal
behavior of the pivot. The great majority of noun phrases in pivot position systematically
have only narrowest scope with respect to other operators in an existential sentence
(though see Francez 2007 for possible counterexamples). This is seen in e.g., (39a), which
is unacceptable because some must take wide scope with respect to negation but cannot.
The narrowest scope requirement follows if the pivot denotes a (scopeless) entity cor-
relate of a property. However, there is no reason on this analysis to expect that pivots
denoting quantifiers like those in (38) should have to take narrowest scope, and indeed
they needn’t (see (39b)).

(39) a. ??There wasn’t some student at the meeting.

b. There wasn’t one particular variety of wine that we expected on the list.

Empirical studies such as Ward & Birner’s and Beaver et al.’s make clear that the facts
concerning the definiteness restriction are much more complex than what was suggested
by earlier studies of the phenomenon, and that additional cross-linguistic research is
needed to determine how best to parameterize a theory of the restriction so that it
can account for the attested variation while capturing what existential sentences have
in common across languages. It seems likely that some kind of gradient or non-unified
analysis will be inevitable; what remains to be determined is whether an analysis that is
purely pragmatic or one that mixes semantic and pragmatic conditions will prove more
insightful.
69. Existential sentences 1845

4. Analyses of the predicate restriction

The second restriction on existential sentences discussed in Milsark (1974) limits the
types of predicates that appear in the coda: As illustrated in (9), above, those allowed
in the coda correspond to Carlson’s (1977a) class of stage-level predicates, while those
excluded correspond to his class of individual-level predicates.
The predicate restriction has generated much less discussion in the semantics and
pragmatics literature than the definiteness restriction, in part because there is less con-
troversy over the facts. Perhaps the only point of contention has been whether noun
phrases are systematically excluded (which would be consistent with the behaviour
of nominal predicates under Carlson’s classification) or whether temporary state-
descriptive nominals are licensed, as Nathan (1978) and Hannay (1985) suggest, using
examples such as (40) in support.

(40) ??There was a woman the president.

However, examples such as (40) are not generally considered acceptable in the literature
and there seems to be little or no evidence in favor of the claim that noun phrases are
licensed as codas.
The analysis of the predicate restriction is deeply bound up with the syntactic analysis
of existential sentences and with the nature of the existential proposition. Milsark (1974,
1977) observed a correlation between those predicates which are excluded from the
coda and those which disallow weak noun phrases as subjects of copular sentences. This
sort of correlation is exactly what is expected if existential sentences express the same
propositional structure as copular sentences. In contrast, if the coda is not considered an
independent constituent but rather simply a post-nominal modifier, as in Jenkins (1975),
Barwise & Cooper (1981) or Williams (1984), a non-stipulative explanation coda restric-
tion should follow from independently necessary restrictions on post-nominal modifiers.
Finally, those analyses on which the coda is a modifier must derive the restriction from
independently motivated restrictions on the kind of modification the coda provides. For
example, McNally (1992) proposes that the coda serves to restrict the spatio-temporal
parameters within which the referent of the pivot is instantiated; on this view, individual-
level predicates are ruled out because they lack the ability to provide the necessary
sort of spatio-temporal restriction. Francez (2007) makes a similar claim concerning
modification of the contextual domain.
It is obvious that a decision concerning the best analysis of the predicate restriction can
only be made by taking into account the syntax and argument structure of existential sen-
tences. It must take into the analysis of the definiteness restriction as well. For example, a
Milsark-style analysis of the predicate restriction predicts that when definite noun phrases
are (exceptionally) licensed as pivots, we should find violations of the predicate restriction
as well. However, this does not happen, as (41), an adaptation of (10), shows.

(41) I think there was one flight thatwas a problem. *It wasn’t ours, but there was that
one flight the problem.

This fact thus serves as another sort of argument, beyond the syntactic, semantic and
typological arguments that have been advanced with increasing frequency, against the
1846 XIV. Sentence types

view that existential sentences and copular sentences have the same propositional
structure.

5. Conclusion
Thanks to their special structural and interpretive characteristics, existential sentences
have offered and continue to offer a rich ground on which to test theories concerning
the semantics of noun phrases and of predication, as well as theories concerning the
role of non-canonical constructions in information packaging. Their close relation to
locative sentences raises interesting psychological and philosophical questions about the
relationship between the notion of presence and that of existence. Finally, the striking
similarities and differences in existential sentences across languages present interesting
challenges for efforts to develop theories of ‘cross-linguistic’ semantics.

6. References
Abbott, Barbara 1992. Definiteness, existentials, and the ‘list’ interpretation. In: C. Barker & D.
Dowty (eds.). Proceedings of Semantics and Linguistic Theory (= SALT) II, Columbus, OH:
The Ohio State University, 1–16.
Abbott, Barbara 1993. A pragmatic account of the definiteness effect in existential sentences.
Journal of Pragmatics 19, 39–55.
Abbott, Barbara 1997. Definiteness and existentials. Language 73, 103–108.
Aissen, Judith 1975. Presentational-there insertion: A cyclic root transformation. In: R. E. Grossman,
J. L. San & T. J. Vance (eds.). Papers from the Regional Meeting of the Chicago Linguistic Society
(= CLS) 11. Chicago, IL: Chicago Linguistic Society, 1–14.
Babby, Leonard H. 1980. Existential Sentences and Negation in Russian. Ann Arbor, MI: Karoma
Publishers.
Barwise, Jon & Robin Cooper 1981. Generalized quantifiers and natural language. Linguistics &
Philosophy 4, 159–219.
Bauer, Winfried 1993. Maori. London: Routledge.
Beaver, David, Itamar Francez & Dmitry Levinson 2005. Bad subject! (Non)-canonicality and NP
distribution in existentials. In: E. Georgala & J. Howell (eds.). Proceedings of Semantics and
Linguistic Theory (= SALT) XV. Ithaca, NY: Cornell University, 19–43.
Borschev, Vladimir & Barbara H. Partee 2002. The Russian genitive of negation in existential
sentences: The role of theme-rheme structure reconsidered. In: E. Hajičová et al. (eds.). Travaux
du cercle linguistique de Prague, vol. 4. Amsterdam: Benjamins, 185–250.
Burt, Marina 1971. From Deep to Surface Structure. New York: Harper & Row.
Carlson, Gregory 1977a. Reference to Kinds in English. Ph.D. dissertation. University of
Massachusetts, Amherst, MA. Reprinted: New York: Garland, 1980.
Carlson, Gregory 1977b. Amount relatives. Language 53, 520–542.
Chierchia, Gennaro 1984. Topics in the Syntax and Semantics of Infinitives and Gerunds. Ph.D.
dissertation. University of Massachusetts, Amherst, MA. Reprinted: New York: Garland, 1989.
Chung, Sandra 1987. The syntax of Chamorro existential sentences. In: E. Reuland & A. ter Meulen
(eds.). The Representation of (In)definiteness. Cambridge, MA: The MIT Press, 191–225.
Chung, Sandra & William A. Ladusaw 2004. Restriction and Saturation. Cambridge, MA: The MIT
Press.
Clark, Eve 1978. Locationals: Existential, locative and possessive constructions. In: J. Greenberg
(ed.). Universals of Human Language. Stanford, CA: Stanford University Press, 85–126.
Francez, Itamar 2007. Existential Propositions. Ph.D. dissertation. Stanford University, Stanford,
CA.
69. Existential sentences 1847

Grosu, Alexander & Fred Landman 1998. Strange relatives of the third kind. Natural Language
Semantics 6, 125–170.
Freeze, Ray 1992. Existentials and other locatives. Language 68, 553–95.
Hannay, Michael 1985. English Existentials in Functional Grammar. Dordrecht: Foris.
Hazout, Ilan 2004. The syntax of existential constructions. Linguistic Inquiry 35, 393–430.
Heim, Irene 1987. Where does the definiteness restriction apply? Evidence from the definiteness
of variables. In: E. Reuland & A. ter Meulen (eds.). The Representation of (In)definiteness.
Cambridge, MA: The MIT Press, 21–42.
Hoeksema, Jack 1989. Review of E. Reuland & A. ter Meulen (eds.). The Representation of
(In)definiteness (Cambridge, MA,1987). Language 65, 115–125.
Holmback, Heather 1984. An interpretive solution to the definiteness effect problem. Linguistic
Analysis 3, 195–215.
Hu, Jianhua & Haihua Pan 2007. Focus and the basic function of Chinese existential you-sentences.
In: I. Comorovski & K. von Heusinger (eds.). Existence: Semantics and Syntax. Dordrecht:
Springer, 133–145.
Huang, C. T. James 1982. Logical Relations in Chinese and the Theory of Grammar. Ph.D. disserta-
tion. MIT, Cambridge, MA.
Jenkins, Lyle 1975. The English Existential. Tübingen: Niemeyer.
Keenan, Edward 1987. A semantic definition of indefinite NP. In: E. Reuland & A. ter Meulen
(eds.). The Representation of (In)definiteness. Cambridge, MA: The MIT Press, 286–317.
Keenan, Edward 2003. The definiteness effect: Semantics or pragmatics? Natural Language
Semantics 11, 187–216.
Kim, Yookyung 1997. A Situation Semantic Account of Existential Sentences. Ph.D. dissertation.
Stanford University, Stanford, CA.
Kimball, John 1973. The grammar of existence. In: C. Corum, C. T. Smith-Stark & A. Weiser (eds.).
Papers from the Regional Meeting of the Chicago Linguistic Society (= CLS) 9. Chicago, IL:
Chicago Linguistic Society, 262–270.
Kratzer, Angelika 1998. Scope or pseudoscope? Are there wide scope indefinites? In: S. Rothstein
(ed.). Events and Grammar. Dordrecht: Kluwer, 163–196.
Krifka, Manfred 1999. At least some determiners aren’t determiners. In: K. Turner (ed.). The
Semantics/Pragmatics Interface from Different Points of View. Amsterdam: Elsevier, 257–291.
Kuno, Susumo 1971. The position of locatives in existential sentences. Linguistic Inquiry 2, 233–
278.
Kuroda, Shige-Yuki 1972. The categorical and the thetic judgment. Foundations of Language 9,
153–185.
Ladusaw, William A. 1992. Expressing negation. In: C. Barker & D. Dowty (eds.). Proceedings
of Semantics and Linguistic Theory (= SALT) II. Columbus, OH: The Ohio State University,
237–260.
Ladusaw, William A. 1994. Thetic and categorical, stage and individual, weak and strong. In: M.
Harvey & L. Santelmann (eds.). Proceedings of Semantics and Linguistic Theory (= SALT) IV.
Ithaca, NY: Cornell University, 220–229.
Landman, Fred 2004. Indefiniteness and the Type of Sets. Oxford: Blackwell.
Leonetti, Manuel 2008. Definiteness effects and the role of the coda in existential constructions. In:
A. Klinge & H. Hoeg-Muller (eds.). Essays on Nominal Determination. Amsterdam: Benjamins,
131–162.
Lumsden, Michael 1988. Existential Sentences: Their Structure and Meaning. London: Routledge.
Lyons, John 1967. A note on possessive, existential, and locative sentences. Foundations of
Language 3, 390–396.
McNally, Louise 1992. An Interpretation for the English Existential Construction. Ph.D. dissertation.
University of California, Santa Cruz, CA. Reprinted: New York: Garland, 1997.
McNally, Louise 1998. Existential sentences without existential quantification. Linguistics &
Philosophy 21, 353–392.
1848 XIV. Sentence types

McNally, Louise 2008. DP-internal only, amount relatives, and relatives out of existential
sentences. Linguistic Inquiry 39, 161–169.
McNally, Louise 2009. Properties, entity correlates of properties, and existentials. In: A.
Giannakidou & M. Rathert (eds.). Quantification, Definiteness, and Nominalization. Oxford:
Oxford University Press, 163–187.
Mikkelsen, Line 2002. Reanalyzing the definiteness effect: Evidence from Danish. Working Papers
in Scandinavian Syntax 69, 1–75.
Milsark, Gary 1974. Existential Sentences in English. Ph.D. dissertation. MIT, Cambridge, MA.
Reprinted: New York: Garland, 1979.
Milsark, Gary 1977. Toward an explanation of certain peculiarities of the existential construction in
English. Linguistic Analysis 3, 1–29.
Nathan, Geoffrey 1978. The Syntax and Semantics of the English Existential Construction. Ph.D.
dissertation. University of Hawaii at Mānoa, Honolulu, HI.
Partee, Barbara H. 1987. Noun phrase interpretation and type-shifting principles. In: J. Groenendijk,
D. de Jongh & M. Stokhof (eds.). Studies in Discourse Representation Theory and the Theory of
Generalized Quantifiers. Dordrecht: Foris, 115–144.
Partee, Barbara H. & Vladimir Borschev. 2007. Existential sentences, BE, and the genitive of nega-
tion in Russian. In: I. Comorovski & K. von Heusinger (eds.). Existence: Semantics and Syntax.
Dordrecht: Springer, 147–190.
Pollard, Carl & Ivan Sag. 1994. Head-Driven Phrase Structure Grammar. Chicago, IL: The Univer-
sity of Chicago Press.
Prince, Ellen 1992. The ZPG-letter: Subjects, definiteness, and information status. In: S. Thompson
& W. Mann (eds.). Discourse Description: Diverse Linguistic Analyses of a Fund Raising Text.
Amsterdam: Benjamins, 295–325.
Rando, Emily & Donna Jo Napoli 1978. Definites in there-sentences. Language 54, 300–313.
Reinhart, Tanya 1997. Quantifier scope: How labor is divided between QR and choice functions.
Linguistics & Philosophy 20, 335–397.
Rigau, Gemma 1997. Locative sentences and related constructions in Catalan: ésser/haver
alternation. In: A. Mendikoetxea & M. Uribe-Etxebarría (eds.). Theoretical issues at the
morphology-syntax interface. Bilbao: Universidad del País Vasco, 395–421.
Ross, John R. 1974. There, there, (there, (there, (there))). In: M. LaGaly, R. Fox & A. Bruck (eds.).
Papers from the Regional Meeting of the Chicago Linguistic Society (= CLS) 10. Chicago, IL:
Chicago Linguistic Society, 569–587.
Safir, Kenneth 1985. Syntactic Chains. Cambridge: Cambridge University Press.
Sasse, Hans-Jürgen 1995. ‘Theticity’ and VS order: A case study. Sprachtypologie und Universalien-
forschung 48, 3–31.
Stowell, Timothy. 1978. What was there before there was there? In: D. Farkas, W. M. Jacobsen &
K. W. Todrys (eds.). Papers from the Regional Meeting of the Chicago Linguistic Society
(= CLS) 14. Chicago, IL: Chicago Linguistic Society, 458–471.
Vallduví, Enric 1992. The Informational Component. New York: Garland.
Ward, Gregory & Betty Birner 1995. Definiteness and the English existential. Language 71, 722–742.
Wilkinson, Karina 1991. Studies in the Semantics of Generic Noun Phrases. Ph.D. dissertation.
University of Massachusetts, Amherst, MA.
Williams, Edwin 1984. There-insertion. Linguistic Inquiry 15, 131–153.
Williams, Edwin 1994. Thematic Structure in Syntax. Cambridge, MA: The MIT Press.
Woisetschlaeger, Eric 1983. On the question of definiteness in ‘an old man’s book’. Linguistic
Inquiry 14, 137–154.
Zeitoun, Elizabeth, Lillian M. Huang, Marie M. Yeh & Anna H. Chang 1999. Existential, possessive,
and locative constructions in Formosan languages. Oceanic Linguistics 38, 1–42.
Ziv, Yael 1982. Another look at definites in existentials. Linguistics 18, 73–88.
Zucchi, Alessandro 1995. The ingredients of definiteness and the definiteness effect. Natural
Language Semantics 3, 33–78.

Louise McNally, Barcelona (Spain)