| |
Abstract:
Studies of semantic memory in brain-damaged patients have
revealed greater deficits in knowledge of living things than of
non-living things in some patients, and the opposite pattern of
deficits in other patients. This double dissociation suggests that
different categories of information may be represented in distinct
brain regions. Martin, Wiggs, Ungerleider and Haxby (Nature, 1996)
reported neuroimaging evidence from normal subjects that is
consistent with this possibility: tool naming selectively activated
a left premotor region, whereas animal naming selectively activated
the left medial occipital lobe. If neural activity in premotor
areas serves an obligatory and functional role in tool naming, then
the efficiency of tool naming should be compromised when subjects
engage in a concurrent manual-motor task. In contrast, the latency
to name animals should not be affected. We tested this prediction
in a dual-task experiment in which subjects named a series of
pictures with and without a concurrent key-pressing task. The
manual task produced a disproportionate increase in tool-naming
latencies, suggesting that tool naming requires the engagement of
manual-premotor areas to a greater extent than animal naming. The
implications of these results for neurocognitive models of semantic
memory are discussed. Studies of semantic memory in brain-damaged
patients have revealed greater deficits in knowledge of living
things than of non-living things in some patients, and the opposite
pattern of deficits in other patients. This double dissociation
suggests that different categories of information may be
represented in distinct brain regions. Martin, Wiggs, Ungerleider
and Haxby (Nature, 1996) reported neuroimaging evidence from normal
subjects that is consistent with this possibility: tool naming
selectively activated a left premotor region, whereas animal naming
selectively activated the left medial occipital lobe. If neural
activity in premotor areas serves an obligatory and functional role
in tool naming, then the efficiency of tool naming should be
compromised when subjects engage in a concurrent manual-motor task.
In contrast, the latency to name animals should not be affected. We
tested this prediction in a dual-task experiment in which subjects
named a series of pictures with and without a concurrent
key-pressing task. The manual task produced a disproportionate
increase in tool-naming latencies, suggesting that tool naming
requires the engagement of manual-premotor areas to a greater
extent than animal naming. The implications of these results for
neurocognitive models of semantic memory are discussed.
|
