The Dorsal Stream and Action Control

Core Idea

The dorsal visual stream (occipital → parietal cortex) converts visual information into motor commands for hand and eye movements. It encodes object location, orientation, and size needed for reaching and grasping. Posterior parietal cortex transforms visual coordinates into motor coordinates, while premotor cortex uses these signals to control action. Damage to dorsal stream regions produces optic ataxia—inability to reach toward visual targets despite normal vision.

Explainer

From your study of the visual processing pathway, you know that the visual system splits after primary visual cortex (V1) into two broad processing streams. The ventral stream projects toward inferotemporal cortex and handles object identification — the "what" pathway, answering the question "what is that object?" The dorsal stream projects toward posterior parietal cortex and has traditionally been called the "where" pathway. But that label is misleading. A better characterization is the "how" pathway — it does not simply represent spatial location for conscious awareness; it converts visual information into the format required for motor action.

The key insight is that vision-for-action requires different computations than vision-for-recognition. To identify an apple, you need categorical representation — shape, color, texture organized into a concept. To *grasp* an apple, you need precise metric information that updates in real-time: the current distance from your hand to the object, the exact width and orientation of the object's graspable surface, and the required finger aperture. These computations must happen quickly, automatically, and largely outside conscious awareness — every time you reach for your coffee cup, you don't consciously calculate its distance or adjust your grip aperture deliberately. The dorsal stream provides this service continuously.

Posterior parietal cortex (PPC), especially the intraparietal sulcus, is the critical integration zone. It receives visual input from dorsal visual areas and proprioceptive input about current limb position, and it performs visuomotor coordinate transformation — translating the retinal position of a target into a body-centered or hand-centered frame of reference that motor circuits can use. From PPC, signals project to premotor cortex and then to primary motor cortex, forming the complete sensorimotor loop. Crucially, this loop operates in parallel with (and largely independently from) the ventral stream's conscious object representations.

The clinical case of optic ataxia provides the sharpest evidence for this dissociation. Patients with damage to posterior parietal cortex (typically from bilateral occipitoparietal lesions) cannot accurately reach toward visual targets — they miss by several centimeters, despite correctly identifying the target and having normal primary visual acuity. The visual perception is intact; the visual-to-motor transformation is broken. The double dissociation is equally revealing: patients with ventral stream damage (visual form agnosia) cannot identify objects but can calibrate grip aperture correctly when reaching for them — demonstrating that the action system does not depend on conscious recognition. Together, these cases confirm that vision-for-action is a functionally and anatomically distinct system from vision-for-perception, running on the same visual input but processing it in fundamentally different ways for fundamentally different purposes.