The superior temporal polysensory cortex (STP), particularly in the superior temporal sulcus (STS), responds selectively to biological motion and body movements. This region integrates form information from the ventral stream with motion information from the dorsal stream to perceive animate agents. STP provides input to mentalizing regions, linking biological motion perception to social cognition.
You already know from the middle temporal area (MT/V5) that the dorsal visual stream handles motion processing, detecting speed and direction of moving stimuli. MT responds vigorously to a random dot field moving coherently — it extracts the motion signal regardless of what is moving. But biological motion poses a distinct problem: a person walking, dancing, or throwing is both a moving thing and a recognizable agent whose movements carry social meaning. The superior temporal sulcus (STS) and the surrounding superior temporal polysensory cortex (STP) are where the brain begins to answer the question "is that moving thing alive?"
The classic demonstration of STP's selectivity uses point-light displays: a person wearing reflective markers on their joints in a dark room, producing only 10-15 moving dots. Humans are immediately and automatically able to recognize these as a walking person, even in brief presentations, even upside-down versions cause recognition failure — suggesting the brain has a strong template for upright human motion specifically. Neuroimaging consistently shows that STS responses to point-light biological motion are stronger than to scrambled versions of the same dots moving with identical local motion statistics. This selectivity requires precisely the integration of form knowledge (where joints are relative to one another in a human body) with motion signals — a computation that MT alone cannot perform.
The "polysensory" in STP is not accidental. This region receives input from auditory cortex as well as visual motion and form pathways. It responds to the sound of footsteps paired with the visual motion of walking more strongly than either alone, suggesting it integrates multimodal cues about animate agency. This makes sense: in the real world, biological agents produce correlated visual and auditory motion signals, and detecting animacy from multiple channels is more reliable than from vision alone.
The downstream consequence is what makes STP theoretically important beyond motion perception. STP sends projections to the temporoparietal junction (TPJ) and medial prefrontal cortex — core nodes of the mentalizing network responsible for attributing mental states to others. The chain from motion detection (MT) → animate agent detection (STS) → mental state attribution (TPJ/mPFC) maps a plausible functional pathway from "something is moving" to "that agent has intentions." Consistent with this, disruptions to STS responses to biological motion are a reliable finding in autism spectrum conditions, with potentially cascading effects on social cognition that depend on smooth automatic detection of biological motion as the starting point.
No topics depend on this one yet.