| |
Abstract:
We focus on comprehension in Broca's and Wernicke's aphasics
in terms of two semantic operations: aspectual and complement
coercion. Aspectual coercion is purely semantic in nature (i.e, not
syntactically encoded) and its purpose is to make elements within a
verb phrase agree in their intrinsic temporal constraints. In the
contrast (a) The girl slept until dawn vs. (b) The girl jumped
until dawn, the interpretation of (a) is obtained via syntactic
composition alone (i.e., the meaning of the sentence comes from the
meaning of the lexical items put together via syntactic processes).
However, the iterative interpretation of (b) (i.e., The girl jumped
repeteadly) is not signaled by any overt-morphosyntactic element.
Nevertheless iterativity cannot be avoided. The requirement that
iterativity be added to the meaning of the sentence is called
aspectual coercion.
Complement coercion is also semantic in nature. In the sentence
(c) The boy began reading/writing the book, interpretation is
obtained straightforwardly via simple syntactic composition. The
boy is initiating an activity reading/writing with an object
noun-phrase the book. This represents a case of syntactically
transparent semantic composition: the meaning of the sentence comes
directly from the meaning of the lexical items put together by
syntactic processes. The verb begin selects for an activity, and it
is the activity writing/reading that semantically licenses the
object-NP the book. By contrast, the interpretation of (d) The boy
began the book is that the boy began doing something with the book.
This sentence contrasts with (c) because the interpretation of what
the boy is actually doing with the book does not come from begin
and even though it is associated with the object NP the book, it is
not signaled by any overt morpho-syntactic means. Nevertheless it
cannot be avoided. Notice also that begin is, by nature, not
compatible with any kind of object that is not an activity. This
semantic incompatibility should in principle yield an ungrammatical
sentence. To explain this incompatibility, and in a way analogous
to aspectual coercion, it is posited that in these cases semantic
well-formedness is achieved through the introduction of a piece of
content in the semantic representation of the verb-phrase that
changes its meaning from referring to an object to one referring to
an activity. In sentence (d) this piece of information is
introduced to achieve compatibility between the head of the verb
phrase begin and its syntactic complement the book (also inside of
the VP) (see Pustejovsky, 1991, Briscoe et al. 1990; Jackendoff,
1997).
Experiment 1 examined aspectual coercion via a two-choice task.
The patient listened to the experimental sentences over headphones.
After each sentence, the tape was stopped and the experimenter
would ask the patient whether the activity had occurred "once" or
"many times": Nine subjects were recruited (3 normal controls, 3
Brocas patients, 3 Wernickes patients). Experiment 2 examined
complement coercion using a picture-matching task. Each correct
depiction was accompanied by a semantic foil in which only the
activity performed by the character(s) was different from the
correct depiction. For this second experiment, eight subjects were
recruited (3, normal controls, 3 Broca's patients, 2 Wernicke's
patients). For both experiments, results indicate that whereas
Brocas patients have little or no trouble understanding sentences
requiring either aspectual or complement coercion operations
(performance was above-chance for all conditions), Wernickes
patients performed at normal-like levels only for the conditions
that did not require these coercion operations. Findings suggest
semantic operations of this sort rely very specifically on the
integrity of the left posterior cortical region associated with
Wernickes aphasia and not on the left inferior frontal region
implicated in Brocas aphasia.
|
