Superior Temporal Sulcus (STS)

Updated 11 December 2025

STS is a prominent fissure in the primate temporal lobe that integrates social perception, audiovisual signals, and attentional control.
It contributes to a distinct third visual pathway by linking dorsal and ventral streams to process biological motion and dynamic social cues.
Neuroimaging and lesion studies reveal that specific STS subdivisions underpin mental state attribution and spatial attention in complex cognitive tasks.

The superior temporal sulcus (STS) is a prominent anatomical fissure within the primate temporal lobe that plays central roles in social perception, attentional control, audiovisual integration, and the mediation of higher-order cognitive programs. Situated between the superior temporal gyrus and the inferotemporal cortex, the STS bridges classical dorsal (“where”) and ventral (“what”) cortical visual streams. Functional, anatomical, neuroimaging, and lesion evidence converge on a model in which the STS constitutes the apex of a third, dedicated visual pathway (distinct from dorsal and ventral routes) specialized for the dynamic processing of biological motion, faces, eye gaze, social signals, and context-dependent attention (Pitcher, 10 Dec 2025, Ramezanpour et al., 2021).

1. Anatomical Architecture and Connectivity

In primates, the STS occupies a zone between the superior temporal gyrus dorsally and inferotemporal cortex ventrally. Cytoarchitectonic subdivision yields posterior, middle, and anterior banks—though the exact borders and cross-species homologies remain under investigation (Ramezanpour et al., 2021). In monkeys, tracer and single-axon studies have established a direct, lateral feed-forward pathway from primary visual cortex (V1) through the motion-selective complex (V5/MT, MST, FST) terminating in the fundus and dorsal bank of the STS (Pitcher, 10 Dec 2025). Human diffusion tractography reveals this trajectory as the “middle longitudinal fasciculus” (MdLF), distinct from the inferior longitudinal fasciculus (ILF) of the ventral stream; it supplies motion and social cues robustly to posterior STS (pSTS). Further connections involve the arcuate and superior longitudinal fasciculi, extreme capsule (EmC), and uncinate fasciculus (UF), linking STS to orbitofrontal and anterior temporal regions (Pitcher, 10 Dec 2025).

STS receives bidirectional inputs from early and intermediate visual cortices (V1–V4), inferotemporal object/face patches, lateral intraparietal cortex (LIP), frontal eye fields (FEF), ventral prearcuate (VPA, homologous to human IFJ), and subcortical structures including the pulvinar and superior colliculus. This architecture positions the STS as a critical integrative hub for both feed-forward sensory signals and top-down executive control (Ramezanpour et al., 2021).

2. Functional Subdivisions and Specializations

Functionally, the STS comprises multiple discrete loci, each with selective roles:

pITd (posterior inferotemporal, dorsal STS): Acts as a feature-blind priority map; encodes attended locations but is not tuned to specific stimulus features. Electrophysiology and fMRI demonstrate a spatial “spotlight” effect, error-predictive signaling, and the capacity to drive attention via microstimulation (Ramezanpour et al., 2021).
Mid-STS: Exhibits loss of attentional enhancement after superior colliculus inactivation; direct mid-STS silencing induces contralateral neglect, defining it as a core node in spatial attention control.
Gaze-following patch (GFP, posterior STS): Neurons here encode the locus of another individual’s gaze, conditional on its behavioral relevance; tuning is dynamically modulated by task rules (Ramezanpour et al., 2021).

Beyond attentional roles, pSTS is selectively responsive to dynamic not static social cues—such as facial motion, eye-gaze shift, hand and body movements—yielding significant and robust BOLD activity compared to ventral face regions (Pitcher, 10 Dec 2025). Convergent evidence from TMS, lesion studies, and neuropsychological testing demonstrates that STS supports integrated processing of facial expressions, gaze, and intentions, separate from static-based object and identity recognition (Pitcher, 10 Dec 2025).

3. Attentional and Cognitive Control Mechanisms

STS forms the functional core of a proposed “Temporal Cortex Attention Network” (TAN), which integrates bottom-up feature conspicuity from sensory cortices and subcortical structures with top-down control signals from prefrontal cortex (notably IFJ/VPA and FEF) (Ramezanpour et al., 2021). Although explicit equations for TAN are not provided, the framework is informed by the Selective Tuning model: hierarchical feed-forward propagation, winner-take-all selection, top-level pruning, and feedback suppression of non-selected signals.

Within this network, each STS locus (pITd, mid-STS, GFP) may instantiate a mid-level priority map, gating the locus of attention based on both spatial position and task rule, then relaying selection signals backwards to earlier visual areas (e.g., V4/V2/V1). These modules collaborate with frontoparietal systems to implement task-dependent attentional operations—cue interpretation, priming, discrimination, disengagement, and rapid shifting—supporting the flexible sequencing required for cognitive programs (Ramezanpour et al., 2021).

Recent neuroimaging with naturalistic paradigms highlights STS as a core component in social perception and Theory of Mind (ToM). Parcel-based fMRI contrasts demonstrate that correct attribution of mental states in animated shapes (i.e., “mentalizing”) robustly recruits a region adjacent to pSTS, with activation levels tracking individual mentalization ability (Srivastava et al., 2023). Although precise stereotactic coordinates and conventional PPI modeling were not implemented, parcel-level subtraction and correlational analysis reveal network-level coactivation among pSTS/TPOj, anterior/association auditory areas (AA), lateral temporal cortex (LTC), and medial prefrontal cortex (mPFC). High functional connectivity among these nodes (e.g., TPOj–LTC r=0.911, p<0.001) supports integrated perceptual and affective contributions to ToM (Srivastava et al., 2023).

Consistent with prior research, pSTS hypoactivation is observed in neuropsychiatric conditions such as ASD and schizophrenia, paralleling the observed correlation between pSTS recruitment and mentalization accuracy in neurotypical individuals (Srivastava et al., 2023). This underscores the role of the STS as a gateway, transforming dynamic social sensory input into higher-order inferential processes.

5. Audiovisual Integration and Shared Emotional Experience

STS serves as a central processor for dynamic audiovisual information and has been implicated in the synchronization of shared emotional perception, particularly in ecologically valid settings such as movie-watching (Muller-Rodriguez et al., 22 Oct 2024). Using small-volume–corrected fMRI analysis with Automated Anatomical Labeling version 3, bilateral STS, and especially the left STS, exhibit robust and significant activation for moments of high inter-observer agreement on valence (both positive and negative) and high arousal (e.g., MNI coordinates [–58 –22 6], Z=5.21, pFWE=0.016 for POS>LAE; [–60 –4 2], Z=5.28, pFWE<0.001 for NEG>LAE).

A generalized linear model, with stick functions at event onsets and parametric modulators reflecting musical features (tempo, volume, spectral bands), indicates that STS activity is sensitive to both emotional consensus and auditory complexity. No significant activation was observed in the amygdala or IFG under similar contrasts. This suggests that STS acts as a hub for rapid audiovisual binding—supporting both perceptual mirroring and inter-subject synchrony—rather than encoding emotional valence exclusively (Muller-Rodriguez et al., 22 Oct 2024).

6. Theoretical Conclusions and Integration with Visual Systems

The classical two-stream model posited orthogonal dorsal and ventral pathways, responsible for spatial/action and object/identity processing, respectively. Anatomical, neuroimaging, and neuropsychological data now substantiate a “third visual pathway” through V5/MT into STS, dedicated to social perception, dynamic biological motion, and action understanding (Pitcher, 10 Dec 2025). The pSTS displays bilateral, panoramic receptive fields, equal responsiveness irrespective of stimulus hemifield, and independence from ventral face networks. Furthermore, lesion and TMS studies confirm selective impairment of dynamic social cue processing with STS disruption, but not with ventral (OFA/FFA) damage.

This third pathway’s computational role is to transform visual motion input into high-level social inferences, providing essential perceptual evidence for downstream mentalizing circuits (TPJ, medial PFC) and supporting adaptive social interaction. The integration is accomplished through distributed circuits along the STS’s posterior–anterior axis, with gradations for biological motion, facial/body cues, audiovisual integration, language, and high-level cognition.

References