We propose to continue to study the neurophysiological basis of speech perception with particular emphasis on the role in auditory and phonemic perception of middle and posterior fields in and near the left superior temporal sulcus (STS), using a comprehensive neuroimaging approach in healthy adults and children. The middle portion of the left STS (mSTS), just anterior and ventral to Heschl's gyrus (HG), is responsive to speech utterances including prelexical segments such as isolated syllables and phonemes, as well as non linguistic familiar sounds such as nonspeech human vocalizations and animal vocalizations. Our principal hypothesis is that the ventral pathway from HG to mSTS mediates the highly-efficient process of mapping the sequential, chronological acoustic information in speech waveforms onto linguistically relevant, highly-abstract and categorical representations stored in mSTS. We propose to examine the spatiotemporal organization of this pathway and its specificity for phonemic perception. The left STS posterior to HG (pSTS) has been implicated in the learning of non-native speech and other novel sounds, in the retrieval and rehearsal of speech (phonological processing) and other auditory sequences and in articulatory planning. The pSTS, bilaterally, also responds to the perception of biological motion. Our principal hypothesis is that a region of the left pSTS acts as a working-memory buffer, transiently storing veridical representations with low-levels of abstraction, that retain the sequential, chronological properties of sounds necessary for phonological processing and articulatory planning. Similarly, regions of the pSTS are responsive to biological motion because of the temporal dimension central to its perception. We propose to examine the common role of this region in speech and biological motion perception and in forming transient veridical representations for novel sounds. Children still developing phonemic perception compared to adults display shallower boundaries between phonemic categories. Children also display a pattern of activation for speech sounds that is more distributed, including left frontal and premotor cortex and a more posterior involvement of the temporal lobes. However, children demonstrate greater plasticity in the acquisition of non-native phoneme categories compared to adults. We propose to compare the ability of children and adults to learn a non-native phonetic contrast, presumably initially through formation of transient veridical representations with low-levels of abstraction in pSTS and eventually through formation of long-term highly-abstract, categorical neural representations in mSTS. We will examine the developing and mature brain with regard to the rate and limitations in the development of phonemic category representations and the neural mechanisms underlying acquisition of phonemic perception.
In the proposed experiments, we provide a theoretical framework for defining the functional differentiation of middle and posterior temporal regions for speech and auditory perception. Understanding the respective role of these regions may suggest means of improving the treatment of disorders associated with deficits in phonemic perception and in particular dyslexia. Delineating the cortical areas involved in different aspects of speech processing is an important building block in the effort to develop a functional map of the cortex and for optimizing presurgical language mapping procedures. The proposed studies will also inform developmental models of speech acquisition.
Showing the most recent 10 out of 16 publications
