Introduction to Behavioral Neuroscience · Vision
Extrastriate Cortex
On this page 9 sections
In 30 seconds
V1 is the first cortical stage of vision, but it is not the last — or even the largest. Surrounding it, and stretching forward into the temporal and parietal lobes, is a patchwork of visual areas collectively called the Extrastriate cortex All visual cortical areas beyond V1 Full entry → ("beyond the striate cortex"). Each area re-analyzes V1's output for a specialized job: V2 refines contours and depth, V4 and the inferior temporal cortex handle color and object identity, MT (V5) Middle temporal motion area Full entry → handles motion, and the posterior parietal cortex uses visual information to guide attention and action.
The organizing insight of the last half-century is that vision is not a single pipeline but two broad streams. The Ventral stream The "what" pathway from V1 to inferior temporal cortex Full entry → ("what") runs toward the temporal lobe and builds object recognition; the Dorsal stream The "where/how" pathway from V1 to posterior parietal cortex Full entry → ("where/how") runs toward the parietal lobe and supports spatial analysis and action guidance. Understanding these streams — their inputs, their destinations, and what happens when they break — is the core of this topic.
Why this matters
- Most visual problems are extrastriate problems. A patient with a damaged V1 is blind in part of the field; a patient with damaged extrastriate areas can often see perfectly well in terms of acuity yet fail to recognize faces, colors, motion, or objects. These dissociations are some of the most striking evidence in all of neuroscience.
- The streams explain real symptoms. Prosopagnosia Inability to recognize faces Full entry → (face blindness), Akinetopsia Motion blindness from MT damage Full entry → (motion blindness), and Visual agnosia Inability to recognize objects despite intact vision Full entry → (inability to recognize objects) each map to a specific part of the extrastriate network — and each tells us what that region normally contributes.
- It grounds the "how does the brain do it" question. Extrastriate cortex is where perception meets memory, attention, and action. Brain–computer interfaces, visual prosthetics, and machine-vision algorithms all borrow ideas from how these areas work.
- It is the standard answer to "where is vision processed?" On exams and in clinical reasoning, the two-stream model and its area-by-area specialization are high-yield.
The college version
Core Concepts
The two-stream architecture
From V1, visual signals branch into two broad pathways (an influential framework from Ungerleider and Mishkin, later refined by Milner and Goodale):
- Ventral stream (the "what" pathway): V1 → V2 → V4 → inferior temporal (IT) cortex. Builds representations of object identity, shape, color, and faces. Damage produces recognition failures (agnosias).
- Dorsal stream (the "where"/"how" pathway): V1 → V2 → V3 → MT (V5) → posterior parietal cortex. Processes motion, spatial location, and the visual information needed to guide reaching and grasping. Damage produces spatial and action failures.
The streams are not sealed boxes — they interact constantly — but the behavioral dissociations they produce are real and reproducible.
V2: the first extrastriate area
V2 surrounds V1 and receives its output directly. Its neurons are organized into repeating stripes visible with cytochrome oxidase staining: thick stripes (disparity/depth and motion), thin stripes (color), and pale stripes (orientation and form). V2 neurons respond to more complex features than V1 cells, including illusory contours — you can "see" the edges of a Kanizsa triangle that are not physically drawn. V2 is where the cortex begins constructing surfaces and boundaries from local cues.
The ventral stream: from V4 to object recognition
- V4: an intermediate color and form area; neurons respond to color, curvature, and complex shapes. Damage here is classically linked to cerebral achromatopsia — losing color vision after brain injury while form vision remains.
- Inferior temporal (IT) cortex: the end stage of the ventral stream. IT neurons respond selectively to complex objects — faces, bodies, places — and even to specific categories. Functional imaging reveals clusters such as the Fusiform face area (FFA) Ventral cluster selectively responsive to faces Full entry → (faces), the parahippocampal place area (PPA) (scenes), and the extrastriate body area (EBA) (bodies).
- What damage looks like: lesions in this territory produce visual agnosia (seeing an object but failing to recognize what it is) and prosopagnosia (failure to recognize faces despite intact vision). The person can describe shape and color but cannot attach meaning — evidence that "seeing" and "recognizing" are separable stages.
The dorsal stream: motion, space, and action
- MT (middle temporal area, V5): the motion area. A large proportion of MT neurons are direction-selective, responding to motion in one direction but not the opposite. Damage to MT on both sides produces akinetopsia — motion blindness, in which the world appears as a series of still snapshots (e.g., pouring tea appears frozen, then suddenly overflowing).
- MST (medial superior temporal area): responds to optic flow — the expanding/rotating patterns of motion produced as you move through the world — supporting self-motion perception.
- Posterior parietal cortex: uses visual (and other) signals for spatial attention and for guiding actions. Damage here can produce Optic ataxia Misreaching for seen objects Full entry → (misreaching for objects in sight) and spatial neglect (ignoring one side of space). Milner and Goodale's famous patient DF, with ventral-stream damage, could not name or match objects but could still reach out and grasp them with correct hand posture — action guidance survived while recognition failed, the signature dissociation of the two streams.
Specialization vs. distributed processing
The area-by-area story is tidy, but the full picture is more nuanced: areas within a stream also interact, and some neurons respond to combinations (e.g., a motion- and form-tuned cell). Modern views treat the streams as weighted specializations rather than absolute divisions: dorsal areas are relatively more concerned with motion/action, ventral areas with identity/color — with cross-talk everywhere. This matters because it predicts graded deficits after damage rather than perfectly clean "modules."
Common Confusions
| Do Not Confuse | With | Difference |
|---|---|---|
| Extrastriate and striate cortex | Each other | Striate = V1, first cortical stage; extrastriate = everything beyond it, specialized for higher features |
| Ventral stream processing "where" | Ventral = what, dorsal = where | Ventral ("what") = identity/color; dorsal ("where"/"how") = location and action guidance |
| Seeing and recognizing | Each other | Vision can be intact while recognition fails (agnosia) — perception and recognition are separable stages |
| Achromatopsia and color blindness | Each other | Color blindness is a retinal/photoreceptor condition; cerebral achromatopsia is a brain (V4-region) condition |
| Akinetopsia and blurred vision | Each other | Blur is optical; akinetopsia is the absence of motion perception with normal static vision |
| MT damage causing blindness | MT damage causing motion loss | MT lesions spare static vision; they specifically remove motion perception |
| The streams being fully independent | The streams interacting | They specialize but cross-talk constantly; deficits after damage are graded, not perfectly modular |
Eli explains
The same idea, in plain words
Explain it like I’m 10
After the picture wall (V1) sorts the lines, the pieces go to two big workshops. The workshop in the back-bottom of the brain (ventral) asks, "What is this thing?" — it recognizes your friend's face, your dog, a banana. The workshop in the top-back (dorsal) asks, "Where is it, and how do I grab it?" — it tracks the ball flying toward you and aims your hand. If the "what" workshop is hurt, you can see the banana but can't tell it's a banana. If the "where/how" workshop is hurt, you know it's a banana but keep missing it when you reach.
Worked example
Patient A had a stroke affecting the fusiform gyrus on the right. Her eye chart acuity is normal. She can describe a photograph: "there are two people, one has a hat." But she cannot say who the people are, even family members, and she sometimes mistakes strangers for friends — she has prosopagnosia. Her ventral stream cannot bind the face's features into an identity, yet her dorsal stream still lets her track and point to the faces. She can see; she cannot recognize.
Patient B, after bilateral damage near MT, reports a bizarre symptom: "when I pour tea, the stream looks frozen; then suddenly the cup is full." He can name objects, colors, and faces perfectly — but the world's motion is missing. He has akinetopsia. His ventral stream is intact; his dorsal motion area is not. The contrast between A and B is the two-stream model in miniature: same eyes, same V1, different extrastriate damage, completely different worlds. (Educational illustration of well-documented dissociations; not medical advice.)
Key takeaways
- Extrastriate cortex = visual areas beyond V1, spanning occipital, temporal, and parietal lobes.
- Ventral stream ("what"): V1 → V2 → V4 → IT; object identity, faces, color. Lesions → agnosias, prosopagnosia, achromatopsia.
- Dorsal stream ("where/how"): V1 → V2 → V3 → MT → parietal; motion, space, action guidance. Lesions → akinetopsia, optic ataxia, neglect.
- V2 stripes: thick (disparity/motion), thin (color), pale (orientation/form); V2 responds to illusory contours.
- MT/V5 is the motion area (direction selectivity); MST handles optic flow.
- FFA (faces), PPA (places), EBA (bodies) are category-selective clusters in the ventral stream.
- Milner & Goodale's patient DF: ventral damage → cannot recognize objects but can grasp them — the classic stream dissociation.
- The streams are weighted specializations with cross-talk, not isolated modules.
- Exam favorite: match the symptom (can't recognize faces vs. can't see motion vs. can't reach) to the area.
Check yourself
6 review questions from the chapter. Try each one, then open the answer.
Name the two visual streams, their destinations, and the "question" each answers.
Show answer
Ventral stream ("what"): V1 → V2 → V4 → inferior temporal cortex — object identity, faces, color. Dorsal stream ("where/how"): V1 → V2 → V3 → MT → posterior parietal cortex — motion, location, guiding action.
Where in the cortex is motion processed, and what happens if that area is damaged bilaterally?
Show answer
MT (V5) and neighboring motion areas in the dorsal stream. Bilateral damage produces akinetopsia (motion blindness): static vision stays intact but moving objects appear as a series of still frames.
What do the FFA, PPA, and EBA have in common, and in which stream do they sit?
Show answer
They are category-selective clusters in the ventral stream: FFA responds to faces, PPA to places/scenes, EBA to bodies — evidence that object categories are mapped in temporal cortex.
Patient DF could grasp objects she could not recognize. What does that dissociation demonstrate?
Show answer
It demonstrates the two-stream dissociation: her ventral ("what") stream was damaged, so recognition failed, but her dorsal ("how") stream was intact, so visually guided grasping still worked — recognition and action guidance use separate pathways.
What is an Illusory contour A perceived edge not physically present (e.g., Kanizsa triangle) Full entry →, and why is it evidence that the brain constructs boundaries?
Show answer
An illusory contour is a perceived edge that is not physically drawn (e.g., the Kanizsa triangle). Because V2 neurons respond to it, the cortex must be constructing boundaries from surrounding cues — perception is an active process.
A patient can name colors and objects but cannot tell you who a face belongs to. Which stream or area is most likely affected, and what is the deficit called?
Show answer
The ventral stream, most likely the fusiform face area region — the deficit is prosopagnosia (face recognition failure with otherwise intact vision).
Study tools & related lessonsKey vocabulary · Related
Key vocabulary
- Extrastriate cortex
- All visual cortical areas beyond V1
- Ventral stream
- The "what" pathway from V1 to inferior temporal cortex
- Dorsal stream
- The "where/how" pathway from V1 to posterior parietal cortex
- Visual agnosia
- Inability to recognize objects despite intact vision
- Prosopagnosia
- Inability to recognize faces
- Akinetopsia
- Motion blindness from MT damage
- Optic ataxia
- Misreaching for seen objects
- Fusiform face area (FFA)
- Ventral cluster selectively responsive to faces
- MT (V5)
- Middle temporal motion area
- Illusory contour
- A perceived edge not physically present (e.g., Kanizsa triangle)
Sources & references
This lesson was adapted from the open educational references above; their licenses and attributions are preserved. See Copyright & Licensing.
Educational content only. It is not medical, legal or professional advice. Found an error? Tell us.