The uniquely human ability to perceive and produce speech sounds is among the most complex cognitive functions to which the auditory system contributes.
The precise role of auditory cortical structures in speech is not fully understood, but considerable progress has been made in the past decade. Studies in monkeys have uncovered neural response features that may be considered as evolutionary precursors to human speech analysis—for example, the sensitivity of certain neurons to formant transitions, found in both human speech and monkey vocalizations, or the fact that some neurons respond to specific abrupt changes in sound envelope that correspond to some consonant speech sounds. However, the complexity of speech decoding goes beyond these low-level features since speech is combinatorial in nature, meaning that words in all natural languages consist of permutations and combinations of a limited set of sounds, called phonemes, which are by themselves not meaningful. The ease with which all normal humans understand the complex sound patterns inherent to speech and the fixed developmental sequence for language acquisition, strongly suggest that a specific neural specialization underlies this ability.
It has been clear for more than a century that one aspect of this specialization is that the left cerebral hemisphere plays a more important role in speech processing than the right in the majority of individuals. Classical studies of aphasic populations not only revealed a lateralization but also indicated that cortical regions in the STG posterior to HG were especially important for the perception of speech sounds since damage to these areas, traditionally referred to as Wernicke's area, typically results in disruption of speech comprehension. The precise role of these areas has not been clear, however, because the precise location of the damage was often not well-known and because patients with such damage often display complex linguistic disorders that go beyond perception of speech sounds. Moreover, damage to frontal lobe language regions can often lead to speech perception deficits. The boundaries of the speech-related areas in the temporal lobe were further defined in the 1950s by Penfield, who tested neurosurgical patients under local anesthesia with the cortical stimulation technique. Penfield and others identified that disruption of speech could often be elicited by stimulation not only of the left posterior STG but also of regions in the middle temporal gyrus as well as the inferior parietal lobe (Fig. 5).
Recently, functional imaging studies have begun to reveal the organization of speech-related auditory cortical areas. The results of most of these studies are in agreement that speech sounds engage auditory areas both anterior and posterior to HG, but that the greatest differential response is in the left posterior temporal area (Fig. 4). The extent to which these specific areas respond uniquely to speech sounds per se, as opposed to certain acoustic characteristics that are present in speech, remains a matter of current research. It has also been proposed from imaging research that regions in the left STS and middle temporal gyrus respond specifically to auditory words.
Depending on the nature of the task, different patterns of activity have been noted. In particular, imaging studies have indicated that in addition to posterior temporal areas, left premotor areas may also be active when the task requires discrimination of speech phonemes. This effect has been attributed to the engagement of articulatory processes that may be involved in some aspects of speech processing. According to this view, which is not universally accepted, speech decoding is not solely dependent on analysis of the acoustic waveform but also depends on comparison of the auditory input with internal motor representations of the vocal gestures necessary to produce the sound.
Conversely, functional neuroimaging studies have also suggested that the area of the planum temporale can also be active during articulation even in the absence of auditory stimulation, for example, when auditory input is blocked by masking noise or when viewing lip and mouth movements, as mentioned previously. These findings thus challenge the conventional view that the posterior temporal speech zone is exclusively dedicated to speech perception, and that the frontal region is devoted to speech production. These regions may instead form part of an interactive
functional network, including other brain areas, that is involved in speech processing.
Much recent attention has been devoted to the idea that speech perception may depend on cortical mechanisms specialized for processing rapidly changing acoustic patterns. Because the perception of certain consonants in speech depends on small temporal differences (on the order of 20 msec), it has been proposed that the temporal resolution of the left auditory cortex is higher than that of the right, and that this factor underlies the hemispheric specialization for speech processing. Evidence in favor of this view derives from electrophysiological and functional imaging studies, which tend to show a greater response from the left than the right auditory regions to sounds that are separated by very brief time intervals, that contain short, abrupt changes in stimulus energy, or that contain rapid changes in formant transitions. Also, in at least some studies, patients with damage to auditory regions on the left have increased temporal discrimination thresholds, as do populations of children with certain developmental language acquisition disorders. It is thus possible that at least some aspects of hemispheric specialization for language arise from differences in processing at the level of primary and surrounding auditory cortices in the left and right hemispheres.
Was this article helpful?