Greg Detre
@6 on Thursday, 18 May, 2000
B&B - III
Parietal cortex
parietal = dorsal
when ventral damaged � but blindsight still applies
why don�t you have consciousness of them?
don�t have to reason with syntactic args re actions � just a sensori-motor mapping
in
contrast, you need to reason about complex plans re objects
reaching out to an object is relatively simple, once you�ve decided on that object (i.e. what it is)
can you do
object-plans without knowing where they are
may be that consciousness = tied up to things you need to reason about � consciousness comes in when you have to correct object-based plans
what/where = conscious/unconscious visual systems???
a bit vague, not good basis for distinction � better to distinguish by functioning
correlate of conscious = interesting
it�s really only in primates with their big temporal lobes that have the ventral stream
only ventral �/span> ability to reason about objects and are conscious at all � can do quite a lot unconsciously
can do a lot with a dorsal system without knowing anything about it
lots of hand-manipulation neurons in the parietal (Mountcastle, 1970)
contrary to Milner + Goodale, ventral = just as necessary for action, not just perception
role of ventral in action = generalising/flexibility
action = towards a particular object � need a target for action
how transfer the information about target-object to target-place
are the 2 streams more in contact than we think?
dennett would
say that they don�t have to
but only one of these is conscious
but what you
are (and are not) conscious of is no indication of what�s going on or what you
were briefly semi-conscious of
form from motion (for example) is evidence that there is some cross-connection
may not be done neurally, but through
the environment � use ventral to identify the object in your fovea, then
dorsal knows where that is
can pick things up without focusing visual attention on them, i.e. peripheral vision
blindsight people � can�t select between 2 objects
can guess �x� vs �o� � rough vs smooth
behavioural goals
---
coordinate transforms
how far do you go?
unique rep in head-centred coordinates???
is the parietal rep viewer-centred
kolb & whishaw � egocentric patients following parietal lesions
retinal coordinate frame � head coordinate frame
combining information from eyes
advs���� can still see things when you move your eyes, multi-modal
auditory � visual mapping
since auditory = in head-based coordinates because your ears don�t move
need both to be in the same coordinate frame
evidence
area 7 � electro recording
one fires when cooridnate with tongue, sound etc.
map of auditory space around the head, some can be activated by visual stimuli (Hyvarinen)
never been followed up
not grained at all
Andersen (1988)
egocentric head based framework
one neuron for left 25deg from head with eyes at zero, one for left 25deg when eye is pointing that way
neurons code for retinal angle and eye position
is what you would find in a neural network
body based coordinates
take posture into account
know where the object is in relation to arm
arms = connected to body, not head
don�t you need a zero point? no, you don�t need an absolute position, only need to compare objects in extra-personal space relative to each other (done by the neural network learning)
proprioceptive input
most important area = neck (especially for arm movements)
useful for actions in space � it�s a fixed transform, which you need to translate into motor commands for you arms
area 5 (receives from 1,2,3) in the superior parietal lobule
neurons there = combinations of the neurons at the previous state
the neurons are giving you the spatial position of the arm relative to the trunk = similar to our trunk/body-based coordinates � can nearly match up a visual and non-visual somatosensory/motor representation in the parietal
allows us to make actions in space
so far, all egocentric � now allocentric
surroundings/ground/room centred
need to know head direction etc. but also need to know my x-y-z position relative to rest of environment
is no evidence for allocentric coordinate frame:
except topographical agnosia � can�t draw maps
is actually not parietal, but much more ventral towards hippocampal gyrus (siriqu)
lurye
subjects lost track of where they are, when they were where
could be hydrograde amnesics (he might not have known where their lesions were) or lost object permanence
allocentric � hippocampus
john o�keefe - place cells � code in allocentric space
only fires when the rat is in a particular position in space relative to room cues, or angle relative to body axis
Rolls � similar finding in primates
the parietal cortex doesn�t seem to have anything to do with allocentric space
Giles Brindley � would just stare up into space when asked about evidence � can�t think of any evidence so there probably isn�t any
what about relative to ground � e.g. when standing, weight distribution, falling
object centred coordinate frame
e.g. agnosia where you lose the left half of every object, implying visuospatial scratchpad
has there been a review of Brain and Emotion?
how old evolutionarily and how advanced in us is the parietal cortex?
how great are individual differences in anatomy between people?
for the big sulci, there�s a lot of similarity
but for other areas, you can�t rely on the sulci to tell you where you are
can�t just map respective to the cranium or some of the smaller sulci
have to use a few standard tasks, like global motion or gratings, then identify key visual ares, then identify the new visual area respective to those � display on flattened sheets which ignore the sulci
why does the brain tend to fold at certain spaces
sudden transition in cytoarchicture and packing densities between different areas, linked to sulci and foldings etc.
also e.g. at the junctions/meridians between V1 and V2 and the connections stretch depending on how far away they are
monkeys � there�s 1 principle sulcus, and the archiweight
in humans, the principle sulcus gets mixed up amongst the others
and they are more variable
is there vision processing inside the brain, i.e. non-cortical?
rolls says that you can have anasagnosia from parietal lobes
why is it so surely probable that there are specific active, neuronal processes correlating with consciousness?
that�s not sure at all
don�t take them too seriously
cycling binding synchronisation � anaesthetised cats synchronise best
lateralisation
contralateral neglect
right parietal lobe is better for spatial, so why do we have equal spatial representations for each half?
would probably only be asymmetrical for higher-order
the least egocentric aspects would be less affected
language just happens to be in the left half
is the wrong way round for handedness
visual attention
selecting one object from a background
some patients need things pointed out
memory
impairment of STM � 3 digits
held in parietal lobe, or damages language functions?
simultanagnosia � visual STM deficit � could be just a visuospatial scratchpad, with some sort of memory component to hold them together
if you have to perceive a spatial input while processing new ones, shunt it off to frontal � otherwise, if it�s not tied up you can use your parietal
pre-frontal attractor � necessary for remembering sample stimulus while others are being perceived
hence the need for separate STM system in prefrontral
phi phenomenon
parallel pathways
complex neural operations going on underneath � different readouts get to different bits at different times
different outputs going to different places
is dennett right about the fuzziness/degrees of consciousness?
how do you turn a bicycle? turn the opposite direction slightly at first to shift your CoG
pre-frontal cortex
Does the pre-frontal cortex have a unitary function?
does it do one thing � functional specialisation of different bits of the prefrontal
incl the orbitofrontal � 4 lines (emotion)
humans & monkeys equally