King, Jeffrey C. (2001) Complex Demonstratives: A
Quantificational Account. MIT Press, xiii+207pp, paperback
ISBN 0-262-61169-4, $18.00, Contemporary Philosophical
Monographs 2, A Bradford book.
Announced at http://linguistlist.org/issues/12/12-1125.html
Previous review at http://linguistlist.org/issues/12/12-2794.html
Edward J. Garrett, University of Virginia, Charlottesville.
Ever since the ground-breaking work of David Kaplan, the
standard view of demonstratives in philosophy of language
has been that they are devices of direct reference. A
demonstrative such as 'that' directly grasps an object and
places it into a proposition, without any mediating sense.
Like a name, and unlike a definite description, it is
thought to go straight to the object itself.
This orthodoxy is challenged in Jeffrey C. King's
interesting new monograph, Complex Demonstratives. King
claims that complex demonstratives, i.e. phrases of the
form 'that N', are in fact context-sensitive quantifiers
rather than referential expressions.
In Chapter 1, King presents a range of empirical data from
English which is difficult or impossible for direct
reference theorists to account for. In some cases,
demonstratives seem to be used when the speaker does not
have a particular referent in mind. In other cases,
demonstratives behave like quantifiers, taking non-rigid
and narrow scope readings.
King starts with so-called 'no demonstration no speaker
reference' (NDNS) uses:
(1) That student who scored one hundred on the exam
is a genius. (p. 3)
(2) That hominid who discovered how to start fires
was a genius. (p. 9)
These are NDNS uses because the speaker has no individual
in mind when she says them, nor is she talking about anyone
in the physical utterance context. Rather, says King, such
statements reflect the speaker's 'completely general'
belief that exactly one person satisfies the conditions
imposed by the noun phrase.
King's claims about NDNS are slippery. The examples he
discusses are all stilted. They sound better with 'whoever'
or 'the N' in subject position. Furthermore, the purported
generality of belief does not hit home: for a truly general
belief, we expect to hear 'whoever'. Here as elsewhere, the
monograph would benefit from a discussion of 'real' data,
such as extracts from actual conversations, corpuses, and
situated language use.
Other uses show 'that' as more of a quantifier, taking
narrow scope under another quantifier (3), or participating
in a Bach-Peters dependency with another demonstrative in
the same clause, as in (4):
(3) Every father dreads that moment when his oldest
child leaves home. (p. 10)
(4) That friend of yours who studied for it passed
that math exam she was dreading. (p. 13)
These are intriguing data, which admittedly challenge the
view under which demonstratives are necessarily rigid.
However, I would like to see the direct reference theorists
respond. Certain features of the examples should give us
reason to pause and reflect - for example the genericity in
In Chapter 2, which constitutes the bulk of the book, King
develops various competing quantificational accounts of
complex demonstratives. All these accounts share certain
features in common. First, they see 'that' as a multi-place
predicate heavily dependent on context. Specifically, they
take 'that' to be a predicate all but two of whose
arguments are saturated by the context of use. The
contextual arguments have to do with the speaker's
intentions, and what's left after these arguments are
saturated is an ordinary (two-place) determiner.
Second, the quantificational accounts agree that there are
two kinds of intentions speakers can have in using 'that'.
In having a perceptual intention, the speaker intends to
speak about a particular object which she presently
perceives or once perceived. Perceptual intentional uses of
demonstratives are the fully rigid uses of demonstratives
that we are all familiar with. In having a descriptive
intention, on the other hand, the speaker does not intend
to talk about a particular specific object, but rather, she
intends to talk about the unique object that satisfies the
relevant descriptive property. For example, consider:
(5) That person swimming across Lake Tahoe must now
be cold. (p. 34)
uttered where the speaker believes on general grounds that
there is a unique person swimming across Lake Tahoe at the
In a descriptive intentional use of 'that', the speaker's
descriptive intentions are usually redundant with the overt
condition found within the complex demonstrative itself
(i.e. the N, e.g. 'person swimming across Lake Tahoe'),
although this is apparently not necessary.
I find the perceptual-descriptive distinction to be
illuminating, but I wish aspects of it had been further
explored. An understandable but regrettable omission is
King's silence on past perceptual uses of 'that'. A more
glaring omission is the absence of any discussion of
discourse factors which might affect the descriptive
intentional use. It seems to me, for example, that the
speaker of (5) is not merely intending to talk about the
person uniquely satisfying the condition of 'swimming
across Lake Tahoe'. In addition, (5) seems to invoke 'that
person of whom we spoke'; in other words, the demonstrative
is being used anaphorically to pick out someone already
mentioned in prior discourse. Sentences such as these may
belong to the category of demonstratives qua anaphors
rather than demonstratives qua demonstratives.
In Chapter 3, King goes out on a limb and argues that
'that' is so much like other quantifiers that it can
scopally interact with modals, negation and propositional
attitudes. Going very much against the ideas of direct
reference theory, he argues that demonstratives may take
narrow scope in relation to these operators.
Here I believe King is just wrong about the facts.
Consider, for example, his cases of complex demonstratives
taking scope under negation:
(6) At any rate, that supermodel who told Alan he is
handsome isn't in there. (p. 96)
(7) That diamond isn't real. (p. 107)
We sense that (6) is true even if no supermodel told Alan
that he was handsome. Since the existence of the supermodel
is not presupposed, argues King, 'that supermodel ...' must
scope under negation.
In fact, King's observation about (6) has nothing to do
with the demonstrative. The same effect can be reproduced
for any rigid nominal:
(8) At any rate, God/his guardian angel/Tom [an imaginary
friend]/Shrek [that monster] isn't in there.
Are we to conclude from this that names, too, are
Similarly, the fact that (7) seems true even if 'the
diamond' is not a diamond is unrelated to the presence of
negation. Consider the following non-negative variation on
(8) That diamond is a fake diamond.
Since the same intuitions hold here, King is forced to
either abandon his claim or to assert that 'fake diamond'
is also grammatically (and not just conceptually) negative.
In my opinion, Chapter 3 suffers from King's silence on
discourse anaphoric approaches to demonstratives. This is
somewhat surprising, given the important role played by
discourse semantics in influential theories of definite
determiners. For example, the notion of discourse
'familiarity' plays a crucial role in both Heim's file
change semantics and Kamp's Discourse Representation
Chapters 4 and 5 conclude by taking up assorted loose ends,
including interesting differences between 'that' and 'the'
(pp. 133-34), and the question of whether or not King's
treatment should be extended to 'that' occurring on its own
without a restricting noun. He does not decide matters one
way or the other, but it looks quite likely that his
account could be extended with similar success to simple
demonstratives. Finally, he concludes the monograph with an
appendix on his formal semantic apparatus.
Overall, King's monograph is an interesting contribution to
philosophy of language, whose key points are sure to be
debated whenever the semantics of demonstratives is raised.
At the same time, King's monograph suffers from various
inadequacies which are likely to limit its usefulness to a
broader audience. First off, in an area of philosophy ripe
for naturalization, the text has little empirical bite.
Many examples are awkward, and the only demonstrative to be
discussed is 'that'. There is absolutely no mention of
'this', even though we know that 'that' has a contrastive
role, and that its semantics cannot be disentangled from
the semantics of 'this'.
Similarly, since all the data in the monograph is from
English, we find no discussion of the rich linguistic work,
especially typological work, done on demonstratives in the
world's languages. There was a day when such omissions were
forgivable in linguistics, and no doubt they are still
forgivable in philosophy, but as linguists we know that
this day must end.
In conclusion, although I enjoyed King's book, I still
eagerly await the final naturalizing blow, the text on
demonstratives in which philosophy and linguistics coalesce
into one, the text which makes it impossible for the two
disciplines to go on ignoring each other.
ABOUT THE REVIEWER
I am especially interested in topics at the intersection of
philosophy, semantics, and pragmatics, and have done
extensive field work on Tibetan. I recently received my
doctorate from UCLA.