Lynsey Wolter (U.
California, Santa Cruz)
This is a talk
about the distribution and interpretation of
demonstrative descriptions (e.g. that cat,
this dog) and definite descriptions (e.g.
the bird, the highest mountain in California).
Demonstrative and
definite descriptions both require that there be
a uniquely identifiable referent satisfying the
descriptive content. Thus, in (1), a context in
which there is no uniquely identifiable cat,
neither kind of description is acceptable.
|
(1)
|
|
There were three
cats in the room. *The/*That/*This cat meowed.
|
But there are also
striking differences between definite and
demonstrative descriptions. For example,
demonstrative descriptions are sensitive to
speaker demonstrations, while definite
descriptions are not, as illustrated in (2).
Definite descriptions, on the other hand, are
compatible with content that necessarily denotes
a singleton set, while demonstrative
descriptions are not, except with a special
emotive reading, as shown in (3). Differences
such as these have led some researchers to
conclude that the referents of demonstrative and
definite descriptions are identified in
fundamentally different ways.
|
(2) |
a. |
I like that
painting [pointing at a painting] but
not that painting [pointing at another
painting] |
|
|
b.
|
*I like the
painting [pointing at a painting] but
not the painting [pointing at another
painting] |
|
(3) |
a. |
The center
of the universe is moving. |
|
|
b. |
??That
center of the universe is moving. |
|
|
c. |
That mother
of John Smith is quite a woman! |
In this talk, I
will propose an analysis in which definite and
demonstrative descriptions are not as dissimilar
as the contrasts in (2–3) might suggest. I will
argue that definite and demonstrative
determiners require uniqueness relative to a
contextually restricted domain, and that
demonstrative determiners place constraints on
how domain restriction takes place. The
analysis, while implemented in a formal
framework, makes room for insights from both
formal and functional approaches, resulting in
greater empirical coverage than previous
approaches have achieved. More broadly, the
analysis has implications for the
semantics-pragmatics interface, suggesting that
the lexical semantics of determiners is tightly
interwoven with pragmatic factors.
back
