| |
Abstract:
A controversial issue is whether specialized neural
mechanisms exist that are activated only by speech. A case in point
is how temporal information is processed in the brain when signaled
in a linguistic versus nonlinguistic context. Thai is a language in
which changes in vowel duration signal differences in word meaning.
In a functional magnetic resonance imaging study, Thai and English
subjects were presented binaurally with pairs of Thai monosyllabic
pseudowords and nonspeech "hummed" versions of the same stimuli,
and asked to make same-different judgments of syllable-final
consonants, vowel length, and hum duration. Data were acquired for
three active tasks (speech: vowel length, consonant; nonspeech:
duration) and one passive listening task (nonspeech: duration).
Paired-image comparisons were: 1) vowel length vs.
duration(active), 2) consonant vs. duration (active), and 3) vowel
length vs. duration(passive). On vowel length and consonant tasks,
Thai listeners showed significant activation in a part of the left
premotor area which extended ventrally into the inferior frontal
sulcus and the inferior precentral sulcus (BA 6/44), whereas
English listeners showed activation in this same region on the
consonant task only. Both groups of listeners showed an activation
pattern including the supplementary motor area (SMA; medial BA 6)
and both posterior superior temporal gyri (BA 22). Our findings
support a task-dependent view, i.e., regardless of acoustic cue(s)
or phonological unit (segmental vs. suprasegmental), complex
auditory stimuli serving a linguistic function will selectively
engage left hemisphere speech processing mechanisms. The left
posterior inferior prefrontal cortex may be specialized for
phonological processing.
|
