| |
Abstract:
It has recently been proposed that abnormally low lexical
activation could be at the root of many of the lexical, syntactic,
comprehension and production problems associated with Broca's
aphasia (Blumstein and Milberg, 1999). This proposal assumes a
distributed model of language processing, in which lexical access
is crucial for all levels of language comprehension - from
identifying individual words in the speech stream, to understanding
word meanings, to determining a word's most likely syntactic role,
to producing meaningful, fluent and grammatical speech. In a series
of semantic priming studies, Blumstein and colleagues showed that
in general Broca's aphasics exhibit semantic priming in fewer
situations than normal subjects, suggesting that Broca's aphasics
have abnormally low levels of lexical activation (e.g. Milberg,
Blumstein & Dworetzky, 1988). However, since interpreting
semantic priming data requires inferring a prime word's activation
from the amount of time subjects take to make a lexical decision
about the target, semantic priming is a relatively indirect measure
of lexical activation. Furthermore, it fails to provide time course
information.
In the past few years, eye movement data have been used to more
directly measure lexical activation, and as a source of time course
information (e.g. Allopenna, Magnuson & Tanenhaus, 1998). A
pilot study compared eye movements of 8 normal subjects with those
of 3 Broca's aphasics. Subjects were shown an array of four
pictures and instructed to point to one of them. The target item
(e.g., "candle") was either in the company of (a) an onset
competitor (e.g., "candy") and 2 phonologically unrelated items,
(b) a rhyme competitor (e.g., "handle") and 2 unrelated items, or
(c) 3 unrelated items. Eye tracking results for normal subjects
showed more eye movements to both onset and rhyme competitor items
than to unrelated items (replicating Allopenna et al., 1998).
For Broca's aphasics, reduced lexical activation should cause
target words to be initially less active than in normal subjects.
Therefore, the reduced activation hypothesis predicts that these
patients should require more input (and consequently more time)
than normal subjects to activate a target word. Furthermore, a
target that is initially only weakly active will have a smaller
inhibitory influence on competitors. Eye tracking results showed
that Broca's aphasics indeed took longer than normal subjects to
disambiguate the target from unrelated items. In addition, like
normal subjects, Broca's aphasics exhibited onset and rhyme
competitor effects. However, for Broca's aphasics the onset and
rhyme competitor effects lasted longer than for normal subjects,
and the rhyme competitor effect was greater than for normal
subjects. These results are consistent with the hypothesis that
Broca's aphasics have abnormally low levels of lexical activation.
[Supported in part by NIH grants.]
Allopenna, P.D., Magnuson, J., & Tanenhaus, M.K. (1998).
Tracking the time course of spoken word recognition using eye
movements: Evidence of continuous mapping models. Journal of Memory
and Language, 38, 419-439.
Blumstein, S.E., & Milberg, W.P. (1999) Language deficits in
Broca's and Wernicke's aphasia: A singular impairment. In Y.
Grodzinsky, L. Shapiro, and D. Swinney (Eds.). Language and the
Brain: Representation and Processing. New York: Academic Press.
Milberg, W., Blumstein, S., & Dworetzky. (1988). Phonological
Processing and Lexical Access in Aphasia. Brain and Language, 345,
279-293.
|
