Dorsal vs ventral

Haxby et al (1995)

- showed that these parallel streams also exist in humans, despite most of the research being done on monkeys it can be generalsie to humans

processing vision

30% of the cortex is taken up to process vision in primates

modularity in the extra striate cortex

A particualr structure which repsonds selectively to a particular quality- this shows specialisation of the brain for certain things. This is useful as it means the brain can become very good at processing vision.

Ungerleider and Mishkin (1982)

Dorsal and ventral streams- Ventral (Temporal cortex)-- what- V1- v2- V4- IT Dorsal (parietal cortex)- where / how- V1-V2-v3-MT (v5) Expeirments on animals where they lesioned monkeys and gave them objects- Object discrimination- saw an abject- then saw 2 objects- were rewarded if they pushed the familiar object away (showing recogntion)- monkeys with lesions to the temporal lobe could not do this but those with lesions to the parietal lobe could. Then in a spatial landmark discrimination task- monkeys saw a well, which they had to remove the cover of to get food- those with parietal lobe lesions could not locate the well but those with lesions to the temporal lobe did not struglle with this. This shows the double disassociation- they are parallel pathways

Hierarchical visual system-

Each area gets bigger- more convergence as you go up, but less sensitivity- larger areas converge many different properties e.g. colour, form (in IT)

Kanwisher et al (1997)

Face recognition is very complex- this occurs in the FFA a area which is specific to faces- it is not responding to indivudal features, it doesn't respond to scrambled features. This suggests that the higher up in the ventral stream you go the larger the visual field becomes and objects and faces are processed as a whole.

Tanaka (1996)

IT responds specifically to certain shapes, not just components of shapes, but the whole thing. This suggests that the higher up in visual processing we are the more complex visual processing becomes

MT- a module for movement- Newsome and Pare (1988)

Monkey's look at dots moving on a screen- motion coherance measured- monkeys could discriminate the motion the dots were moving in when they correlated as little as 10-20%. But then when the MT was lesioned, this meant they could no longer see tell the direction of the dots. This shows that mt is specific for motion. These neurons are direction selective so chose fire depending on specific direction and then distrubuted coding is used.

MT neurons respond to movement- Gallesse et a (1996)

This is repsonsive to how we do things- this is evident in the fact that mirror neurons are found in the MT which respond to watching others do things and when we perform them ourself. This supports the claim that the dorsal stream is the how stream as these neurons respond to how others do things and how we do things as well.

V2-

V2 is involved in processing curves and textures within a visual field- they ahve a alrger visual field

Kohler et al (1995)

measured the brain activity in humans and showed that there is a double disassociation between object memory task and spatial tasks- in the spatial task the parietal cortex was activated but in the object recognition task the IT was activated

Milner and Goodale (1995)

suggested that the dorsal stream is instead the how stream- this is because it isn't involved in where Lesion patients- DF- lesion to the dorsal stream- couldn't recognise objects but could grab them- visual form agnosia- where you can't identify objects based on visual information. She also showed difficulty matching the prientation of a card to a slot, yet when when asked to mail somehting throguh a slot she was able to. This supports the fact that the dorsal stream (in the parietal lobe) is responsive to 'how' a person carries out actions and tracking other's movements

Allman and Kass (1975)

suggested that the v6 area is involved in self motion perception- in the dorsal stream, and includes a topographical map of the entire visual field

Cross talk between dorsal and verntal streams- Ungerleider and Haxby (!990)

v4 recievs input from both the magnocellular stream and the parvocellular stream- magnocellular stream is concerned with motion- this suggests that they are not seperate streams- this is because information will flow backwards from higher areas to v1 for 'feedback and predictive coding'

Quiroga et al (2005)

very specific neurons which fire for specific things in the parietal cortex-distrubuted coding- using the ocmbination of many neurons and combines their firing to identify objects.