Cognitive, Retina & Processing, WEEK 2 Flashcards
1
Q
Complication of perception
A
- We are somehow able to perceive certain features of objects (e.g. colour, structure etc) and use the information of these features to understand what it represents as a whole (e.g. a black coffee mug i perceived in terms of its differing features but how do we know it is a black coffee mug?)
- How can we go from photons (particles of light) bouncing of the screen + hitting our retina to knowing this is coffee and other information about the image > some cognitive process must be involved
2
Q
Eye to cortex
A
- We need info from the outside world to enter our sensory system then travel to our brain so that we can perceive it > eye to cortex (surface of brain)
Process of eye to cortex explained: - Reception: physical energy of photons enters the eye, hits the retina, interacts w/ photo receptors > info is then received
- Transduction: Here, the energy from the photon is converted into a electrochemical pattern which get sent down neurons, brain cells into our brains which enables the next step
- Coding: There must be a correspondence between what is in the world + what is happening in our brain > there has to be a physical representation of what is happening in the external EV and what this physical stimulus is represented by in the firing pattern in our brains
3
Q
Eye and the retina
A
- Pupil is where the light enters the eye and this light hits the back surface of the eye, called the retina
- The retina is covered in photoreceptors called rods or cones > Rods are photoreceptors which are interested in dim light, don’t care about colour but how much light/photons are hitting the photoreceptor. > Cones are interested in processing colour, lots of different kinds at the back of the eye which form the retina.
- Mammalian eyes have evolved “inside out” > light has to travel through blood vessels and things before getting absorbed by photoreceptors > this is a problem because all these signals from photoreceptors need to get out of the eye so they can be sent to the brain (via optic nerve), to get out of the eye there must be a hole for signals to leave > consequently there is a hole at the back of the eye where the signals get send down the optic nerve
- Optic nerve is where electrical signals are sent from interaction between the photons + photoreceptors
- The hole at the back of your eye is called the blind spot > in this hole, there is no processing + no light hitting it so we cannot see what is happening in this blind spot, but in everyday life we don’t experience this “not seeing”, we can only tell if purposely try to find it
4
Q
Retina & top-down constructive perception
A
- Perception is a constructive process where the way the world appears to be is not necessarily how it is (top-down processing) > this is happening in our retina due to our blind spot
- Retina comes from the way cones and rods are distributed (this comes into play when thinking about constructive perception)
- most cones are in the fovea whilst most rods are in the outer region (periphery)
- This evidence would indicate that you can only see things in colour if they are directly in front of you (in your fovea) + would not see things in your periphery in colour > but we can see things in the corner of our eyes in colour (even though no colour processing is happening in the periphery, as there’s no cones)
- Somehow, the brain gives the perception of colour in the periphery despite there not being colour processing happening > similarly to how our eyes fills in blind spots even though there is no information directly at our eyes, our brain fills it in > helps the constructive process.
5
Q
Colour vision
A
- We can only see colours because that is a specific part which our eyes respond to on the electromagnetic spectrum > we respond to visible light which we interpret as colours
- The electromagnetic spectrum is wide but we just respond to certain wavelengths of electromagnetic energy in visible light > important because we cannot see radio waves or x-rays even though this is possible (they can be seen), it is just that our eyes cannot see them > Human eyes are most sensitive to green light
6
Q
Trichromatic theory
A
- Young found you can produce any colour by mixing the 3 primary colours > when you shine a blue, red and green light onto a black surface, any visible colour can be created > this is additive colour mixing (subtractive is when you mix paints)
- Helmholtz suggested the most efficient way for the eye to work is to have 3 types of colour receptors (red, blue and green) because the different amount of input into these colour receptors could combine to give all possible colours of light
- Research found this is right and there are 3 types of colour receptors which prefer different wavelengths > Blue like short wavelengths, green likes medium and red likes long
- May conclude that if you can create any colour of light from this relative combination + if we have blue, green + red cones, presumably your experience of colour just comes from the relative activation blue, green + red cones in your retina
7
Q
Opponent-process theory
A
- The assumption above^ has issues because there are problems when people describe colours > sighted people never describe a colour as “blueish-yellow or reddish-green” > may be associated w/ this theory
- This theory has the idea that the inputs from the 3 different cones are processed oppositely > doesn’t give us how blue something is for example, but rather says where something falls on a scale of red-green, blue-yellow and dark-light > sighted people cannot describe things like blueish yellow but that is how our eye sees it
• Opponent processes work by, the input from the red and green cones would be subtracted (find difference) and send this electrical signal to the brain, this results in a signal which falls somewhere between red and green (cannot be red + green at the same time) - Blue-Yellow: can make yellow by mixing red + green light > so add inputs from red + green cones and subtract the difference between this and blue (difference between blue + yellow)
- We can find how light/dark something is by adding all cone inputs together, we have a measure of how bright something is (how many photons are hitting the retina)
- Dual process theory took this opponent-process idea and connected that to the trichromatic theory > evidence that there are neurons in low level parts of the brain which perform the above calculations > before the red, blue or green signals get processed by the brain, they are re-processed and re-described in the opponent process manner + this info enters the brain for higher processing > basically how much blue, green or red light there is (trichromatic theory) gets processed in the retina then before this colour is hierarchically processed, opponent processes occur where it says where the colour falls on a scale of blue-yellow/red-green > these scales because colour blind people tend to have deficits in identifying between blue + yellow, red + green
8
Q
Colour constancy
A
- Tend to perceive something to have same colour regardless of the colour light which is shining on it (even though in the real world, this is not the case)
- If there is a blue object (reflects blue light) and we shine a red light on it, the actual colour reflecting would be black > but we do not perceive it as this
- Our brain creates a construction of our perception > we often perceive things as colours they are even when lit with light, which doesn’t even let that colour hit our retinas
- Beneficial from an evolutionary view > e.g. colour of sun changes through the day (light) but we can still see the genuine colour of fruit and know if it is safe to eat
- Our brains are involved in filling in gaps to try and assume colours (e.g. A looks grey and B looks white but when separated they are actually the same which is what the cones see. But our brain fills in the gaps and assumes a light is shining on A while B is in shadow)
- Perception is not what our cones perceive (because they can see the genuine colour) but we do not know this > perception is a constructive process (top-down)
- How we perceive the world is also not how the world necessarily is (e.g. human eyes may see a flower as yellow if it reflects that kind of light, but a butterfly may not see this same as their eyes don’t focus on light but on UV light so they see it in a way we don’t. what we see may not be intrinsically true)
- Perception is a result of a combination of top-down and bottom-up processing
9
Q
What happens after the retina?
A
- Signals from the retina towards the brain (via optic nerve) travel on two parallel pathways > these pathways are two different types of neurones sending info in parallel
- Parvocellular pathway has most info on cones and send this to the brain (sensitive to colour + fine detail)
- Magnocellular pathway has most info on rods + sends to the brain (sensitive to motion + light)
10
Q
Pathway from eye to brain
A
- The info your eye takes in travels to your retina through the above parallel pathways towards the back of your brain
- Info from eyes (retina) travels down the optic nerve (1), goes through optic chiasm, this is where the optic nerves cross over (3), optic nerve carries on into a part of the thalamus called lateral geniculate nucleus (4), after info is processed in LGN, it gets sent to the back of the brain to the occipital lobe to V1 (5)
- Signals from the left side of both retinas is processed in the left side of the visual cortex, information hitting the right side of both retinas gets processed in the right side of the visual cortex
- Signals reaching the left visual cortex come from BOTH left sides of the TWO retinas, and signals reaching the right cortex come from BOTH right sides of the TWO retinas
- Retina, 2. Optic Nerve, 3. Optic Chiasm, 4. Lateral geniculate nucleus, 5. cortical area primary visual cortex (V1)
11
Q
How is information processed in reality?
A
- Light can only travel in straight lines > this means light from the left hand side of the screen travels into the right side of the retina because it can only travel in straight lines > light on the right hand side of the screen travels in a straight line and hits the left side of the retina
- Keeping in mind info entering the left retina is processed by the left side of brain and info entering the right retina is processed by the right side of the brain, this means in reality, what we see from the left side of the visual world is processed by the right side of the brain + what we see from right side is processed by the left hemisphere
12
Q
Properties of visual neurones
A
- Visual neurones are neurones which process visual information > we process this info using these
- Info travels down the optic nerve + entering the brain where visual neurones process the info received
- Visual neurones are important in helping construct our perception of the visual world
- Receptive field: this is the region of sensory space (i.e retina) within which light will cause a neuron to fire > neuron in the brain fires when something happens in a certain area of the visual world, this only fires when something happens in that particular part of the world > this part of the world is that neurons receptive field
- Retinotopy: Things which are close together in space (in front of you) are also processed by neurons which are physically close together in your brain > if one neuron has a certain receptive field, the neuron next to it will have a certain receptive field very close to the original neuron (neurons next to each other process physical space next to each other) > retinotopic map where neurons are a map of what we see ahead
- Lateral inhibition: neurones next to each other can inhibit each other > if one neuron is firing a lot because a lot is happening in it’s receptive field, it can reduce the likeliness that the neurons around it will fire > leads to the neurons in the similar space to fire a lot in a way of trying to inhibit the other from firing which is useful for enhancing contrast at edges of objects
13
Q
Lateral geniculate nucleus
A
- LGN is part of the thalamus > thalamus is a subcortical (below cortex) relay for most brains sensory input (i.e. what we see) + motor output
- LGN contains neurons which process visual info (receives info from retina) > these neurons have receptive fields which are called centre-surround receptive fields
- Centre-surround receptive fields really care about differences in light between the centre and edges of the receptive fields (e.g. light in centre and dark in surrounds)
- Maintains a retinotopic map
- Rapidly processes visual info, particularly can combine signals across space and time to give an early + quick representation of whether something is moving or not > first point of processing info from eyes in thalamus can tell if something is moving towards you as well as coordinate motor output (because thalamus is responsible for this) > very quick reaction > evolutionary benefits
- If something is moving towards you rapidly, you can process + react to this very quickly w/o the info being processed higher up in the brain
14
Q
Primary visual cortex (V1)
A
- After the thalamus, the info gets sent to the back of the brain to the V1 (part of the occipital lobe)
- V1 cares about low level basic info from the visual scene such as edges, orientation (vertical/horizontal lines) > V1 processes this info at a low level by saying for example there appears to be a line here + sends this info forward in a hierarchical way > sends up the hierarchy so it can be processed in more depth (shape, colour, movement etc..)
- Maintains retinotopy
- Single cell recordings by Hubel + Weisel suggest you have some neurones in V1 which respond to simple features (e.g. only care about whether there is light coming into your visual field or not) + other neurones in V1 which combine info from multiple cells which have receptive fields which are close together > these cells may determine if there is a line using light and where there is darkness + connect the info to establish what is there
15
Q
Damage to primary visual cortex: Blindsight
A
- Damage to V1 leads to diagnosis of cortical blindness where patient cannot consciously report on things in front of them in this region of space (no conscious vision)
- But they are able to make judgements about things in that blind part of their visual field > patients have some level of processing happening in that part of their visual world even if they cannot report what is happening there > 2 routes of vision (conscious/unconscious)
- The patients can make visual discriminations such as orientation or movement direction
- There may be an unconscious visual processing route where visual info is still processed + sent throughout the brain to be useful but this cannot be accessed consciously + another route for visual info which can be accessed consciously
- Patients w/ damage to V1 may lose ability to have conscious experience of visual info in the damaged part of their visual field but the info in that visual field is still available to the brain to help make discriminations (e.g. know what direction things are moving)
- Geniculostriate route may be responsible for conscious vision while other routes act unconsciously > these other routes seem to enter the thalamus and skip the primary visual cortex which still allows some level of unconscious processing of what is happening in the visual world
16
Q
Blindsight
A
- Filling in of blind regions is similar to that of filling in our blind spots > your brain creates a perception of what the world is most likely to be in the blind region
- If a person w/ blindsight is presented with a semi-circle in the visual field they can see they will know it is a semi-circle, but if the semi-circle is put on the edge of their blind region, their brain creates the perception that it is a circle because that is the most simple answer. If the semi-circle is in their blind region, they won’t see it at all
- We can have visual info which isn’t consciously available to the person, but the info is still available to the brain + can be used to change behaviour
17
Q
Visual processing beyond V1
A
- After signals leave V1, they move up the hierarchy which are well connected > info can move up but also down at times
- Information can be shared between different nodes in the visual system before going further up into more complex parts of the brain > visual info is processed by steps
- V1, V2, V3, V4 + V5 are at the back of the brain where info gets sent towards the top of the brain in the parietal lobe, STS or IT
- Information gets more complex as we move through the hierarchy.
18
Q
Functional specialisation theory
A
- Evidence for different functions of V1,V2,V3,V4 + V5 comes from this theory (Zeki,1992)
- They recorded from individual neurons using intracranial electrophysiology in these parts of the visual cortex of macaque monkeys
- Each of these parts (V1,V2,V3,V4,V5) are all specialised for a particular part of visual processing
- V1 + V2 is the earliest stage which care most about the outlines of objects, shapes and orientation
- V3 prefers things which move
- V4 has neurons which fire a lot and care a lot about colour > colour processing may occur here
- V5/MT neurons process motion in our visual field
- The central assumption made is that colour, shape and motion are processed by different parts of our brain > we have a binding problem again as we don’t perceive things as separate (colour separate from shape for example) > we see these as bound together while in our brain they are processed separately
- BUT, evidence comes from macaque monkeys, what about humans?
19
Q
Zeki et Al. (1991) in human V4 + V5
A
- Did a brain imaging study + took a PET scan > PET shows neurons which are firing more because ppt eats radioactive sugar which shows these active areas
- If healthy participants are shown images in black or white in a greyscale or images in colour + compare the two > found V4 was more active for coloured images than for greyscale images > V4 may be specialised for colour like in macaque monkeys
- Similarly, V5 is associated w/ motion so to test this we compare moving dots with static dots and evidence shows V5 is more active for the moving dots which indicate it is specialised for motion
20
Q
V4 Evidence: the colour centre of the brain:
A
- Could look at neuropsychology for more evidence
- Patients with cortical achromatopsia cannot see colour primarily due to damage of V4 but this deficit also happens with damage to V2/V3 > this suggests that V4 may not be the ultimate centre of colour processing as initially thought
- Some case studies also indicate intact implicit colour processing in patients w/ achromatopsia > they may perceive colours for the things they know the colour of (e.g. they may experience a banana as yellow if they came across it before their damage)
- Therefore, V4 is important for colour processing but may not be the only place where colour processing occurs in the brain
21
Q
V5 Evidence: the motion centre of the brain & akinetopsia
A
- Brain damage to V5/MT leads to Akinetopsia > after damage to V5, these patients have trouble with processing motion in the world
- Patient LM had perfectly fine colour vision + could locate stationary objects but their motion perception was highly impaired
- This supports the view that V5 is involved with processing motion
22
Q
Challenge to Functional Specialisation: The Binding Problem
A
- Sighted people don’t perceive the colour of things separately to their shape yet in their brain, these are processed separately in our brain > so where in the is this full thing perceived?
- Conscious experience is not just the colour of something, the movement nor shape. It is all these things combined so where or how is the info bound together?
- One explanation may be temporal coding > several neurons from these different regions fire at the same time perhaps the simultaneous firing of the neuron creates our conscious experience > perhaps attention is involved in solving the binding problem
23
Q
Beyond the visual cortex: two important pathways
A
- Info continues to travel in the brain after going through the visual cortex > there is info which travels over the top of the brain, this pathway is called the Parietal or Dorsal processing pathway.
- Information can either travel through the dorsal route which goes over the top of the brain towards the parietal lobe, OR, it gets sent down along the ventral pathway towards the temporal lobe
- Where route: Parietal/Dorsal processing pathway – primarily concerned with movement for processing – often called “vision for action” pathway because visual info is sent to the motor areas of the brain > sends useful visual info to parts of the brain which helps you move
- What route: Temporal/Ventral processing pathway – concerned w/ colour + form processing – often called “vision for perception” > primarily looks at what we are processing or looking at, this pathway may be more related to our conscious experience
- What happens in the ventral pathway may relate more to our experience while what happens in the dorsal pathway is less related to experience + a more unconscious level of processing
24
Q
Patient DF: Vision for perception/Vision for action
A
- Patient DF had a lesion to her lateral occipital cortex which is where her ventral pathway is (where info gets sent to her temporal lobes)
- Information still travelled through the dorsal pathway to the parietal lobe without problem but the information travelling along the ventral pathway was impaired
- Patient DF was compared to a control + had to line up an envelope with the changed angle of the letterbox (it could rotate), this was the perceptual orientation matching task. In the next task, they just had to post the letter as quickly as they can
- Control ppt could do these tasks very easily but patient DF could not
- All the lines shown are the conscious perception of where patient DF thought the post-box angle was (even if it was just vertical, she would say it was all over the place) > she didn’t have a conscious experience of where the orientation of the post box was
- In the second task, DF behaves similarly to controls
- This suggests patient DF cannot tell us where the angle of the post box is or where it is facing, but her motor system can tell us this (as she is able to post it accurately) > but when asked to think about the orientation of post box (consciously think), she cannot do this.
- Indicates the ventral pathway is involved with perception + our conscious experience of what is in front of us while the dorsal pathway seems to allow processing unconsciously > even if DF cannot say what is in front of us, the motor processing enables us to react/move without even thinking about (unconscious
25
Models of object recognition
- How do we get from processing things like shapes and colours (certain parts of visual field) to knowing exactly what is in front of us
1. Early visual processing > colour, motion, edges involving V1,V2,V3,V4,V5
2. Perceptual segregation > grouping of visual elements, so we put together the lines and other factors, we use this to segregate/separate one object from another (distinction)
3. Structural description: Next, we match this visual description with representation (info in our brain which represents this). E.G. if we see lines which resemble a book + match this to our representation of a book in our minds, we will then recognise it is a book
4. After this, the meaning is associated with the object because we have recognised it. E.G. we know books are for reading
26
Perceptual segregation
- Refers to separating visual input into individual objects > occurs before object recognition
- Gestalt Psychology: The law of Prägnanz (law of simplicity) > suggests that what you will perceive to be out there is the simplest solution of what it could be (what you see in front of you can be many things but what you perceive will be the simplest possibility) > more simple to have rules which help describe what things are likely to be in the world + these are the things we perceive rather than what they may actually be > bottom-up approach
27
Gestalt laws of perceptual organisation
- The rules which perceptual organisation is assumed to follow include the law of proximity, law of similarity, law of good continuation and law of closure
- Law of proximity: For A, most likely to perceive there are 3 rows of 4 > we somehow combine the rows because they are close to each other > we organise things based on how close they are to each other
- Law of similarity: we organise things in our visual field based on if they are similar to each other > For B, we may organise the circles + squares to be part of the same object
- Law of good continuation: you assume things that move in the same direction, continue to move in the same direction > For C, may see 2 lines overlapping but it may be that there are 4 lines which intersect
- Law of closure: Lines continue how they have previously + that those form a whole > For D, this circle is not complete but we would perceive there is a circle, we just cannot fully see it (could actually be that it is not a circle + something else) > more simple to assume that there is a full circle where we cannot fully see it
28
Figure-Ground Segregation
- Another aspect of segregation comes from this > idea that whenever we perceive something in our visual field, we prefer to perceive it as being in front of something else > object + background
- Face-goblet illusion: Drawing which can be seen as two faces or as a goblet > whichever figure you perceive seems to be in front of a background (e.g. if you see a face, you see a black background, if you see a goblet you see a white background.
We are presented with a figure, a background + a context which it is in > we pay more attention to the figure than the ground
29
Problems with Gestalt Psychology
- Ignores top-down contribution to perception: Segregation process is not always bottom-up which means it doesn't always follow the laws of perceptual organisation > evidence shows that when participants are asked whether X is on the same object or not (using a button). RT's are faster for the top row than bottom (RT's are a reflection of how quick perceptual organisation is) > Top row are letters which we are familiar with so we are quicker here while the bottom row are random shapes > because we have encountered these letters before so we are quicker at segregating objects (top-down)
30
Object recognition, brain & deficits
- Ventral pathway is where we expect object information to be processed
- If object recognition occurs at a hierarchical level (where early visual processing happens in V1, V2 etc while segregation and other processes happen higher up), then we would expect different kinds of impairments to object recognition depending on where along this stream you have the injury (e.g. V2 or ventral pathway)
- Agnosia: impairment in object recognition that does not come from damage to primary visual cortex (not from damage to areas which process low level visual features) > damage to higher up processes
31
Dissociating two types of Agnosia
- Apperceptive Agnosia: impairment in processes which constructs perceptual representation from vision (grouping + segregation processes) > in our "what" (ventral) pathway
- Typically seen in damage to lateral occipital lobe > Damage to this area tends to affect ability to group visual features into an object
- Associative Agnosia: Impairment in process that maps perceptual representation on the knowledge of object (meaning of object) > in "what" pathway but further down than ^
- Typically seen in damage to occipito-temporal damage (where occipital + temporal lobe join together)
32
Apperceptive Agnosia: seeing parts but not the whole
- Patient HJA had bilateral damage to lateral occipital lobe (side of the occipital lobe) > HJA could recognise objects from touch but had an impairment in what that object was (recognising it) + also in recognising if what they are looking at is 1 object (e.g. a paintbrush looked like 2 separate objects when it is one)
- HJA seems to have a problem with grouping together information (perceptual segregation) + recognising that certain low level features are apart of the same object
- People w/ apperceptive agnosia seem able to copy from pictures + draw from memory but are unable to decide if an object is real or not (unable to workout that a certain combination of lines together form one object + this object maps onto a structural representation we have (meaning)
- Also impaired in naming an object, saying what an object is (e.g. say they see a carrot)
33
Associative Agnosia: seeing the whole but not the meaning
- Damage further up the ventral pathway, in the occipito-temporal region (where occipital + temporal region join
- Patients of this lack access to the meaning of what they are looking at
- Patient LH was able to copy drawings of objects but was unable to name them > had no access to semantics of objects
- Visual object agnosia: one kind of associative agnosia (there are several kinds)
34
Neuropsychology of object recognition
- Essentially, we assume processing happens from a low level (basic visual features) which then undergoes perceptual organisation/segregation processes where the info is put together to establish if it is actually an object, this then goes further up the hierarchy to establish what the object is
- Neuropsychological evidence supports this model > damage to V1 removes conscious visual perception (if you don't process things in V1, you aren’t aware they happened
- Damage towards the lateral occipital causes apperceptive agnosia > deficits in ability to segregate features of the visual world into individual objects
- Damage further down towards the temporal lobe produces associative agnosia > patients struggle to understand what an object is + what it is for, even though they know it is an object
35
Caveats of object perception
- Most psych research comes from white male western researchers who tend to study on white western participants > need to be cautious about generalisability of things found from psychology
- This is more important when looking at low level processing, in this case, object perception > as we know culture impacts how our mind works
- E.G. evidence indicates westerners tend to prioritise processing the objects in visual scenes while east Asians are more likely to prioritise the context or relationship in what they are seeing (Nisbett & Masuda, 2003)
- For example, if westerners are shown an image of 2 elephants, they will prioritise processing the fact there are 2 individual objects while east Asians process the relationship between the 2 elephants + the context they are in (measured using eye-tracking > means there are cultural differences in how we even look at scenes)
- Evidence also shows less activation of object perception regions in east Asians than westerners during scene viewing (Gutchess et al, 2006) > but object perception associated regions are only found in western ppts, so these regions may not necessarily be associated w/ object perception in other cultures
36
Different types of objects
- Different categories of objects are processed in different places along the ventral stream > if you have damage to a certain part of the ventral visual stream, you may just be impaired in one particular type of visual object
Lesions to different areas of the what stream can be associated w/ agnosia of different types of objects like naming faces, animals or tools > faces as a type of object seem to be special in our visual stream
37
Problems with face recognition
- Face processing is problematic because we do it everyday (if normal sighted)
- Face recognition is within-category discrimination > all faces look relatively very similar + we rarely do within-category discrimination, most things are between-category discrimination (e.g. you will rarely look at a room full of pens + look for a certain pen)
- Is it a case of processing faces is different to other types of objects? Do they have a separate processing mechanism?
38
Are faces special? neuropsychological evidence
- Prosopagnosia: impairment in facial processing > specific impairment where damage to the temporal lobe (fusiform face area) of the ventral visual stream causes an impairment in face processing + doesn't come from damage to early visual processing > impairment at stage of matching stored information
- Means facial processing happens at a higher level
- De Renzi (1986) finds that a patient could not recognise his families faces but recognised their voices + could match different views of faces + name other objects
39
Neuroscientific evidence for faces being special
- If we put healthy ppt in a scanner + are shown faces and houses then use subtraction logic as houses are complicated visual objects which share features with faces > compare brain activity when look at faces + houses
- More activity in Fusiform Face Area when looking at faces, suggesting there is a certain part of the ventral stream which primarily responds to faces rather than other objects
40
Why are faces special?
- May be because face discrimination is harder than other tasks (deciding who someone is may be more difficult) > but can be ruled out using difficult discrimination tasks of objects yet there are still face effects
- Maybe faces are processed in a more holistic or configural way?
- Perhaps we are experts at processing faces as a normal sighted person sees a lot of faces so is used to it
- There may be domain-specificity where a part of the brain is dedicated to facial processing
41
Holistic processing
- Tests where ppts are shown faces then houses explain holistic processing > if we show a ppt a certain face (e.g. Jim) and then show other faces, then show a certain house (Jim's house) then multiple houses.
- After a short delay, you show either an entire face or a section of the face (e.g. eyes) and ask is this Jim? Similarly, show either an entire house or a part of the house (e.g. windows) + say is this Jim's house?
- Graph shows the accuracy of whether something was Jim's or not > orange line shown when shown the whole face/house, recognition is similar
- Purple line shows when shown a part of the face, recognition is much lower than when shown a part of a house
- Specific impairment in our ability to recognise faces based on parts unlike for other objects like houses > there is a requirement to see the whole face to recognise it (holistic processing)
- Perhaps the relationship between the features in the face is what we recognise rather than individual features
42
Facial recognition is disrupted by inversion
- We are worse at recognising faces when they are upside down whereas when looking at other objects like a chair, we can easily recognise them even if there is a slight delay > perhaps due to relationships between features in the face which may be more important for facial processing unlike other objects
- Face recognition is specifically impaired when that face is upside down (more than object recognition) > perhaps faces are not objects + processed differently
- Interpretation for this distinction is that faces requires holistic processing > to do with relationships between the features in a face, spatial-relational info + when you turn things upside down, the relationships between features in that face are disrupted.
- for example we are used to the nose being a certain distance below the eyes, but if we see an image with the nose above the eyes, we lose configural info + are less able to recognise the face
-Object processing however may rely less on this holistic/configural processing
Facial processing may be about the whole face w/ relations between features rather than individuals
43
Visual expertise
- Maybe there is a specific face processing part of the ventral stream because that might be where very complicated visual distinctions happen
- Classic studies looking at these use Greebles where ppts need to learn what different male + female greebles look like from differing families > when ppt become "experts" at visual discrimination between greebles, the fusiform face area becomes more active than other parts of the visual system
- Supports the fusiform face area may not actually be the face area, rather it may just be the fusiform complex visual processing area (where complex visual discrimination happens)
44
Criticisms of visual expertise
- Evidence from neuropsychology show that it is not the case that all prosopagnosic patients are impaired on within-category discrimination
- E.g. patient WJ had prosopagnosia but had a flock of sheep + could distinguish between them (is a difficult discrimination task which they could complete, yet couldn't recognise faces)
- If fusiform face area was specifically for complex discriminations they wouldn’t be able to do this
- Patient RM could distinguish between 5000 of his miniature cars but could not recognise his wife's face or own face > could complete a difficult discrimination task but not recognise faces
- This evidence indicates that difficulty of a discrimination task is not enough to explain why faces are processed differently to other objects
45
Conclusion: are faces special then?
- We do not know for sure but evidence does show that faces are special in the sense they are processed more holistically than other objects
- But, we do not know if the reason they are special is because they are faces or something else about what makes a face a face
- Evidence also shows a certain brain region is more active than others when processing faces (FFA) but this could be associated with visual expertise or within-category discrimination (not strong evidence)
- But we do know that if we have damage to a particular part of the ventral visual stream, then we can remove the ability to recognise faces more than ability to recognise other objects > suggests faces are special but we need more research to understand why
