Neural responses to meaningless pseudosigns: evidence for sign-based phonetic processing in superior temporal cortex

Abstract

To identify neural regions that automatically respond to linguistically structured, but meaningless manual gestures, 14 deaf native users of American Sign Language (ASL) and 14 hearing non-signers passively viewed pseudosigns (possible but non-existent ASL signs) and non-iconic ASL signs, in addition to a fixation baseline. For the contrast between pseudosigns and baseline, greater activation was observed in left posterior superior temporal sulcus (STS), but not in left inferior frontal gyrus (BA 44/45), for deaf signers compared to hearing non-signers, based on VOI analyses. We hypothesize that left STS is more engaged for signers because this region becomes tuned to human body movements that conform the phonological constraints of sign language. For deaf signers, the contrast between pseudosigns and known ASL signs revealed increased activation for pseudosigns in left posterior superior temporal gyrus (STG) and in left inferior frontal cortex, but no regions were found to be more engaged for known signs than for pseudosigns. This contrast revealed no significant differences in activation for hearing non-signers. We hypothesize that left STG is involved in recognizing linguistic phonetic units within a dynamic visual or auditory signal, such that less familiar structural combinations produce increased neural activation in this region for both pseudosigns and pseudowords.

Figure 1

A) Illustration of the VOI analysis showing greater activation in left STS for deaf signers compared to hearing non-signers for the contrast pseudosigns minus fixation baseline (peak coordinates: −57, −30, −3). The statistical map was rendered on a single subject T1 image in MNI space using MRIcron. B) Illustration of brain regions that were more active for pseudosigns than ASL signs for the deaf participants. The statistical map is from a one sample t test of pseudosigns minus ASL signs and was rendered on a single subject T1 image in MNI space using MRIcron.