Superior Temporal Sulcus—It’s My Area: Or Is It? Grit Hein and Robert T. Knight Abstract Downloaded from http://mitprc.silverchair.com/jocn/article-pdf/20/12/2125/1759317/jocn.2008.20148.pdf by guest on 18 May 2021 & The superior temporal sulcus (STS) is the chameleon of the portion, mainly involved in speech processing, and a poste- human brain. Several research areas claim the STS as the host rior portion recruited by cognitive demands of all these dif- brain region for their particular behavior of interest. Some see ferent research areas. The latter finding argues against a strict it as one of the core structures for theory of mind. For others, functional subdivision of the STS. In line with anatomical evi- it is the main region for audiovisual integration. It plays an dence from tracer studies, we propose that the function of important role in biological motion perception, but is also the STS varies depending on the nature of network coactiva- claimed to be essential for speech processing and processing tions with different regions in the frontal cortex and medial- of faces. We review the foci of activations in the STS from temporal lobe. This view is more in keeping with the notion multiple functional magnetic resonance imaging studies, focus- that the same brain region can support different cognitive ing on theory of mind, audiovisual integration, motion pro- operations depending on task-dependent network connec- cessing, speech processing, and face processing. The results tions, emphasizing the role of network connectivity analysis in indicate a differentiation of the STS region in an anterior neuroimaging. & INTRODUCTION often interpreted in line with this assumption. For In the last decade, the human superior temporal sulcus example, the left STS region is seen as a prime area (STS) and surrounding regions have been widely studied. for speech processing (Wernicke, 1874), with the MTG Paradoxically, the exploding numbers of findings have being more strongly involved in word comprehension, made the role of the STS region in the human brain even and the anterior STG, STS, and angular gyrus being more mysterious. The STS is a major sulcal landmark in more important for sentence processing (Dronkers, the temporal lobe and adjacent cortices on the surface of Wilkins, Van Valin, Redfern, & Jaeger, 2004). At the same the superior and middle temporal gyri (STG and MTG, time, a particular region in the upper bank of the left respectively) and the angular gyrus at the intersection to STS, intersecting the parietal lobe, has been proposed to the inferior parietal lobe posteriorly. The STS is claimed be the host of ToM because patients with a lesion in this to be the host brain region for theory of mind (ToM) and region showed a deficit in a false belief task (Samson, social perception (Saxe, 2006; Zilbovicious et al., 2006; Apperly, Chiavarino, & Humphreys, 2004). The anterior Gallagher & Frith, 2003), but also for audiovisual (AV) inte- portion of the right STG was suggested to play a role in gration (Amedi, von Kriegstein, van Atteveldt, Beauchamp, spatial awareness because it was shown to be the center & Naumer, 2005; Beauchamp, 2005; Calvert, 2001). For of lesion overlap in neglect patients (Karnath, 2001). some, it is the prime brain structure for biological mo- Other authors claim that the right STG hosts biological tion processing (Puce & Perrett, 2003; Allison, Puce, & motion processing because a patient with a lesion cir- McCarthy, 2000), whereas others discuss it in the con- cumscribing the entire right STG showed a deficit in bio- text of speech perception (Price, 2000) and face pro- logical motion perception (Akiyama, Kato, Muramatsu, cessing (Haxby, Hoffman, & Gobbini, 2000). Saito, Nakachi, et al., 2006; Akiyama, Kato, Muramatsu, What is the secret of the multifunctionality of the STS Saito, Umeda, et al., 2006). The observation of a behav- region? One assumption is that a large structure as the ioral deficit can make a strong case for the function of a STS and its adjacent cortices has some regional special- particular damaged brain region. However, such conclu- ization. According to this assumption, for example, one sions have to be drawn with caution. It is fair to say that subsection of the STS region might mainly be associated lesions exclusively focusing on the STS region are rare in with ToM and social perception, whereas another dis- humans and that most patients exhibit deficits in more tinct STS region might host AV integration. The results than one function. For example, Akiyama et al.’s patient of studies on patients with lesions in the STS region are showed spatial neglect in the acute stage, besides the impairment in biological motion processing. Samson et al.’s patients had deficits in speech comprehension in University of California at Berkeley addition to impairment in ToM. One explanation is that D 2008 Massachusetts Institute of Technology Journal of Cognitive Neuroscience 20:12, pp. 2125–2136 Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/jocn.2008.20148 by guest on 26 September 2021 the patients’ lesions extended from, for example, the On the one hand, these data provide evidence for a ‘‘ToM region’’ to the ‘‘speech processing region’’ of the functional subdivision of the STS region in nonhuman STS. Alternatively, the finding that lesions in similar STS primates. On the other hand, the variety of reciprocal regions cause different functional deficits could argue connections between certain STS regions and distinct against a strict functional subdivision. higher-order areas could be the anatomical basis for Lesion studies or single-cell recordings in nonhuman an alternative assumption. Instead of a strict functional primates are spatially more precise than lesions in the subdivision, the multifunctionality of the STS region human brain. Although the comparability of human and might be based on coactivations with other brain re- monkey STS is uncertain, there are similarities in ana- gions, such as the frontal, parietal, and mesial temporal tomical projections, permitting assumptions about the cortex. In this case, STS region activity would be deter- Downloaded from http://mitprc.silverchair.com/jocn/article-pdf/20/12/2125/1759317/jocn.2008.20148.pdf by guest on 18 May 2021 human STS on the basis of nonhuman data (Keysers & mined by the nature of the network interactions with Perrett, 2004; Petrides & Pandya, 2001). There are a num- other brain regions it is coactivated with. For example, ber of findings indicating that a region in the anterior coactivation with ventral and medial frontal regions, pro- portion of the upper STS is selectively involved in the jecting to the limbic system, might determine a role of processing of body movements, including gaze direc- the STS in ToM, whereas coactivation with premotor tion (Oram & Perrett, 1994, 1996; reviewed in Keysers areas would make it part of the network for motion pro- & Perrett, 2004). Other single-cell studies report neu- cessing, and so forth. The findings from patient studies rons in the STS region, which selectively respond to and studies on nonhuman primates, to date, are insuffi- moving images, faces, or nonface complex visual stimuli cient to clarify the function of the STS region in the hu- (Desimone & Ungerleider, 1986; Van Essen, Maunsell, & man brain. Bixby, 1981; Seltzer & Pandya, 1978). Most of these neu- In this review, we consider the contribution of func- rons are reported in a multimodal region (TPO) in the tional magnetic resonance imaging (fMRI) studies to this upper bank of the STS (Baylis, Rolls, & Leonard, 1987). question. We review foci of activation in the STS region This multimodal region could be distinguished from a of multiple fMRI studies, investigating ToM, AV integra- more dorsal area involved in auditory processing (TAa) tion, motion processing, speech processing, and face and a more ventral region which mainly processes visual processing. If there is regional specialization within the stimuli (TEa and TEm) (Baylis et al., 1987). Apart from STS region, one would predict that foci of activations preferences for input modalities and visual properties, from studies investigating the same cognitive function the different STS regions are also characterized by the are clustered in a particular STS subregion. A distribu- pattern of anatomical connections with other brain re- tion of foci of activations independently of the inves- gions (Seltzer & Pandya, 1989a, 1994). The auditory and tigated cognitive function would be more in line with multimodal sections in the upper bank of the STS (TAa the assumption that STS functions are determined via and TPO) are reciprocally connected with the frontal network coactivations. cortex. The anterior portion of the multimodal TPO projects to and receives inputs from ventral and medial areas of the frontal lobe, which themselves project to METHODS the limbic system (Barbas, 2000; Carmichael & Price, 1995). The middle portion of the TPO is mainly con- The integration of foci of activation from multiple fMRI nected to lateral prefrontal regions around the principal studies is challenging and has its own inherent problems sulcus. The posterior portion of the TPO projects to (see also Duncan & Owen, 2000). One issue concerns dorsal prefrontal and premotor areas. Moreover, all STS statistical power. Statistical power is limited in any given subregions project to the parietal cortex. The three TPO study. Consequently, only a part of the activation pattern subregions differ not only in intercortical connectivity passes a certain conventional threshold. It is possible butalsoincytoarchitecturalandchemoarchitectural that noise differences between two studies investigating properties. The posterior TPO is characterized by large the same cognitive function might lead to activations pyramidal neurons in layer III and clearly separable lay- in different subregions of the STS, even if they actually ers V and VI. The middle TPO contains a well-developed recruit a common STS subregion. The noise, which can layer III with medium large pyramidal cells, but no dis- contribute to such a pattern of results, is increased if the tinction between layers V and VI.