Who Goes There?

respond preferentially to complex sounds (e.g., bandOur understanding of the visual system has benefited pass noise or species-specific vocalizations) relative to markedly from the hypothesis that visual processing tones. In contrast, neurons in the core region respond occurs in two separate streams—one for what (or who) preferentially to tones relative to complex sounds. It is and one for where (Ungerleider and Mishkin, 1982). A important to note that neurons in the same noncore dorsal processing stream mediates spatial processing region can show comparable increases, relative to core through a circuit including the visual cortex (which conregions, in their sensitivity both to sound source location tains the primary visual cortex and spatial processing and to complex sounds. Thus, while there is hierarchical areas, such as area MT) and the parietal cortex. A ventral processing between the core and belt regions, these processing stream mediates object processing through areas do not process preferentially spatialand objecta circuit including the visual cortex (which contains the related information. primary visual cortex and object processing areas, such Neuroimaging studies also indicate that in human auas area V4) and the inferotemporal cortex. These two ditory cortex the extent of activation within core and processing streams converge in the prefrontal cortex. belt/parabelt regions increases as stimulus complexity However, in recent years, it has become clear that these increases (Wessinger et al., 1998, Soc. Neurosci., abtwo streams are not strictly parallel. For example, cells stract). Activation for tones is primarily limited to the in the lateral intraparietal (LIP) area, an area associated core regions. However, additional noncore regions are traditionally with the dorsal stream, are also modulated activated by more complex sounds such as band-pass by the shape of an object, a computation associated noise bursts, frequency-modulated sweeps, and spoken traditionally with the ventral stream (Sereno and Maunsell, syllables. Importantly, the same noncore regions are 1998). activated by speech-like stimuli and other complex Recently, processing streams for audition, comparasounds. This observation suggests that this hierarchical ble to those in the visual system, have been proposed processing is not language specific but represents gen(Raushecker, 1998). In this model, a dorsal stream, which eral processing necessary for the identification of comconsists primarily of the auditory cortex, parietal cortex, plex sounds. and prefrontal cortex, processes information about the The role of the human auditory cortex in spatial prolocation of a sound source. A ventral stream, which cessing is unclear. Some data indicate that the auditory consists primarily of the auditory cortex, the rostral sucortex is not preferentially involved in spatial properior temporal cortex, and prefrontal cortex, processes cessing. For example, the activation in the auditory corinformation about auditory objects. Given this interesttex is statistically similar when participants listen pasing model, we think it is reasonable to ask how indepensively to sounds and when they actively localize sounds dent these two streams are. In this minireview, we will (Bushara et al., 1999; Weeks et al., 1999). Similarly, the describe recent findings that probe the role of these auditory cortex is not modulated, relative to baseline different areas in spatial and object processing. Our controls, when participants listen passively to moving focus will be on those studies that examine this cortical sound sources (i.e., auditory motion; Griffiths et al., processing in areas beyond the primary auditory cortex 1998). In contrast, Baumgart et al. (1999) suggested that, in awake, behaving nonhuman primates with neurophysin the posterior auditory cortex, there is preferential actiiological techniques or in humans with neuroimaging vation when subjects listen passively to auditory motion, techniques. relative to tasks in which they listen passively to stationAuditory Cortex ary sound sources. These contrasting results may be The nonhuman primate auditory cortex is organized into due to differences in analysis techniques or methodolothree distinct regions: the core, the belt, and the paragies. For example, the finding of Baumgart et al. (1999) belt. Each of these three regions contains a number of may be due to their use of tonotopic organization to anatomically and physiologically defined auditory fields; parcel auditory cortex into several functional regions of the primary auditory cortex is a tonotopic field in the interest, which afforded them the ability to differentially core region of the auditory cortex. The details of the probe the functional role of these regions. organization of the auditory cortex are out of the purview Superior Temporal Cortex of this minireview, but the reader can refer to Kaas and The superior temporal cortex is important for auditory Hackett (1998) and Rauschecker (1998) for further inforobject processing. Lesions in the superior temporal cormation on this topic. tex severely impair a monkey’s performance on an audi-