| |
Abstract:
Background
: Are the words and rules of language subserved by distinct
mechanisms (Pinker, 1991) or a common mechanism (Elman et al.,
1996)? On a dual-mechanism view, regular past tense forms (
look-looked
) are computed by the application of an -
ed
-suffixing rule, whereas irregular forms (
dig-dug
) are dependent upon memory. On a single-mechanism view, both past
tense types are retrieved from associative memory. Neurological
double dissociations between regulars and irregulars would
strengthen the dual-mechanism case. Patients with temporal lobe
damage (posterior aphasia, amnesia, Alzheimer's disease) are worse
at generating irregular than regular past tense forms; those with
frontal/basal-ganglia lesions (anterior aphasia, Parkinson's
disease) show the opposite pattern (Ullman et al., 1997; Ullman
& Corkin, 1997).
Motivation & Method
: Can these findings be replicated with a different task and a new
population of aphasic patients? We asked 6 Broca's aphasics (all
with left anterior lesions) and 5 Wernicke's aphasics (3 with left
posterior lesions, 2 with unlocalized left lesions) to
read
out loud 17 regular and 17 irregular past tense forms matched
pairwise for stem and past tense frequencies and for final
consonant structure (e.g.,
slipped-swept
).
Results
: The Broca's aphasics had more difficulty reading regular (mean
30% correct) than irregular (44%) forms. The Wernicke's aphasics
showed the opposite pattern (61% vs. 55%). The interaction between
Irregular/Regular Verb and Posterior/Anterior Aphasia was
statistically significant by item (p < .05), and approached
significance by subject (p = .08).
Conclusions
: The results dissociate the reading of regular and irregular past
tense forms, and are consistent with the dual-mechanism hypothesis
that the mental lexicon is part of a temporal lobe declarative
memory system for facts and events, whereas grammatical rules are
processed by a frontal/basal ganglia procedural system for motor
and cognitive skills (Ullman et al., 1997). Grant support: Army
DAMD17-93-V-3018 and NIH DC00494
|
