Faces Flashcards
Why is face processing important in developmental psychology?
Faces are complex and dynamic stimuli that are salient to infants.
Identifying different people and recognising faces as the same person with a different expression or from a different angle retired fine discrimination within and between people.
Faces have great social importance.
Key theme pervasive to theories, data and methods in developmental psychology is whether developmental changes are domain-general or domain-specific.
Question infant’s sensitivity to salient stimuli: people’s faces.
Investigating babies’ early biases towards human faces and emotional expressions combined with later changes in face processing abilities helps to evaluate the innateness and construction of social cognition.
There is evidence to suggest that children with autism process faces differently.
How do adults process faces?
Face-responsive neurons that fire more to face than non-face stimuli have been found in the temporal and frontal cortex of monkeys. Indicative of specificity.
Using fMRI, face-responsive brain activation in humans has been found in the fusiform gyrus which is activated more to face than non-face stimuli and allows them to recognise faces in more detail than similarly complex inanimate objects.
Adults can detect faces in less than 100 milliseconds.
Adults with acquired prosopagnosia are unable to recognise familiar faces and it is specifically associated with the fusiform gyrus. They become reliant on the less sensitive object-recognition system to recognise faces. Represents a double dissociation as it shows that the two functions are localised in different areas of the brain.
Faces are treated differently by the brain than any other stimuli.
What is a domain- specific learning?
Humans learn different types of information differently and have distinctions within the brain for many of these domains.
Infants have specific neural circuits that are sensitive to face-like configurations.
What is domain-general learning?
born with mechanisms that exist to support and guie learning on a broad level regardless of the type of information being learned e.g starts off with the development of general functions related to perception and memory, and gradually refines into specialised facial recognition mechanisms over exposure
What face skills do neonates have?
Seems to be some representational bias that the neonate brings to the learning situation for faces - Karmiloff-Smith (2006)
The classic evidence to support this is that neonates like to look at faces.
However it could be argued that this is just because faces are complex and dynamic stimuli so are interesting to look at so that a more general principle is at play.
What did Johnson’s (1991) studies about infants’ face preferences find?
Experiment 1
Tested 24 neonates who were 30 minutes old.
Used a tracking procedure measuring head and eye turns in response to three schematic stimulii.
Measured the degrees of angle that the neonates turned their head, indicative of them tracking the stimulus.
Conditions: face-like stimuli, scrambled face, blank stimuli.
Results were consistent with the hypothesis that faces are special stimulus.
Head and eye turns were both greatest for the face-like stimuli and showed more interest in the scrambled stimuli than the blank stimuli.
The stimuli have too many differences to conclude that it is because faces are special stimuli.
Experiment 2
Used neonates of a similar age as before but with some additional conditions.
4 Schematic stimuli: face, top heavy dot configuration of a face, inverse dot configuration of face, linear scrambled face.
No effect on head turn.
For eye tracking, it was found that babies track onger to face stimuli rather than the inverse and linear stimuli.
Looked at the face somewhat longer than the configuration but no statistics were used in the official analysis.
It was concluded that neonates are more interested in faces than other stimuli even when the saliency and amount of information is the same.
Across studies, differential attention shown to face-like stimuli by neonates.
Further consistent evidence that faces are special stimuli.
What did Reid et al.’s (2017) study into fetal face preferences find?
Studied 39 fetuses in the third trimester of development - eyes open.
Face-like or inverted stimuli were projected across the maternal abdomen, moving across the fetal visual field five times.
Measured the number of fetal head turns in response to stimuli by coding using ultrasounds.
Found that there were more looks towards stimuli that were face-like and this higher level of engagement is consistent with the hypothesis that faces are special stimuli.
Face preference by neonates seems to replicate
What is the discontinuity in development in face preference?
There is a reported decline in face preference between 1-2 months (Johnson et al, 1991)
This preference is found to reemerge at 2-3 months.
This is consistent with the view that infants gradually construct representations of faces over the first few months of life.
However it is a surprising contradiction that the preference should come and go.
This has been explained as a reflection of two different systems
System 1: governs the behaviour of younger infants.
There is a period of decline in system 1 as system 2 develops.
System 2: governs the behaviour of older infants.
What is Johnson and Morton’s (1991 ) two process model for development of face processing?
Process 1 CONSPEC: the subcortical visuomotor pathway is responsible for preferential tracking of faces in newborns
similar to fixed action patterns - which are species specific behavioural sequences that occur because of a neural network.
Suggestion that this region of the brain receives input only from ovals with inverted triangles.
Predisposes infants to be interested in faces and to attend to them which fuels the development of process 2, which in turn may inhibit process 1.
Process 2 CONLERN: adult-like cortical pathways specialised for faces emerge over time as a consequence of experience.
Suggestion that the stage 1 bias for inverted triangles causes the brain region to receive only face-like input and the region gradually becomes specialised for faces.
Thus there is a gradual emergence of a face area in the brain.
Faces are special because infants show preference over other non-face-like objects and these preferences are directed by the face-specific processing mechanisms CONSPEC and CONLERN.
What are the problems with CONSPEC?
Limited functionality of conspec may struggle to deal with the data supporting newborn’s preference for attractive faces (Slater et al., 2000) and a direct gaze over averted gazes (Farroni et al., 2002).
Compounded faces become more symmetrical as you average more of them together and are considered to be more attractive. Babies prefer averaged faces to more unusual ones.
Thus faces judged by adults as more attractive (note can’t project concept of attractiveness onto neonates) are looked at for longer by neonates.
However this is only true for upright faces and internal features seem to be important.
Not consistent with conspec, suggests more detail is processed than just an inverted triangle.
But it does support that top heavy stimuli are important.
Gaze experiments used photos of faces where gaze was either direct (at child) or averted.
They found that babies spend longer looking at a face with a direct gaze than if the same face had an averted gaze.
This is evidence of fine tuned processing which contradicts the very simplified discrimination conspec argues for.
But it can still be argued that this is not face specific (domain-specific) but domain-general processing.
Moreover there is more evidence to suggest that the preference is for top-heavy configurations as opposed to specifically face configuration.
What did Cassia et al.,’s (2004) study into infants preference for top-heavy stimuli find?
Investigated whether the preference for top-heavy configurations translates to non-schematic images of faces.
Found preference for an upright face over the same face inverted.
Scrambled the features of the same face and found a preference for the top-heavy configuration over bottom-heavy configuration.
Found no significant difference in preference for upright face or top-heavy configuration.
This suggests that newborns show preferences for top-heavy configurations but it is not specific to faces.
This appears to support the conspec mechanism of a bias to face prototype.
May be a more general perceptual bias that is not face specific at all.
A two-process theory may not be needed to explain processing of faces.
Data is also consistent with gradual development of a single mechanism reflecting an immature adult system.
Are neonates sensitive to humanness?
Heron-Delaney et al. (2011)
Investigate whether neonates are sensitive to humanness.
2 day neonates shown image of a human face and another of a non-human primate face
Showed a preference for the human face
Could indicate a special interest in human faces or something more basic.
How did Giorgio et al. (2002) critique and respons to Heron-Delaney’s (2011) study?
Studied 24 infants, 72 hours old
Controlled for lower level perceptual factors between the two images such as contrast, luminance and low spatial frequency
Experiment 1:
Found that there was no looking preference therefore contradicting previous findings
Experiment 2:
Repeated the experiment, adding a habituation phase
found that the babies can discriminate between the two classes of stimuli.
Experiment 3:
Compared looking preference for inverted and upright monkey faces
Babies looked longer at the upright faces, just as they had with human faces.
Conclusions
From this we can conclude that neonates preferences are not human-specific in the first week of life.
Neonates can discriminate the stimuli.
Controlling perceptual factors removes any preference for human face over primate face.
Suggestive that monkey faces are processed in the same way and the same effect of inversion is observed which is further evidence to support the hypothesis of a bias towards a top-heavy prototype.
Human faces do become special to infants.
Limited by the fact that only isolated faces or face-like schematic stimuli have been used but this is not representative of how they are viewed in the real world (Jayareaman et al. 2017)
Need to do more situated experiments to represent the complex input of faces.
Kelly et al. (2019)
Studied 241 infants who were 3-12 months old.
Monitored the eye movements of babies as they viewed a room where a person was either present or absent OR salient and not salient.
Machine image analysis enabled researchers to capture salience such as contrast.
Faces were classified as detected if the fist saccadic eye movement post-stimulus onset was to the face and was not detected if they were never looked at.
Results showed that more infants first looks were to faces than any other equivalently salient areas of interest.
Salient faces were more detected than non-salient faces.
Babies attend to faces regardless of silence, despite varying spatial location, small visual angle, inconspicuous sclera and much competing information.
Impressive face detection indiciating that infants may be better than previous studies concluded.
Further highlights how new technologies such as eye tracking can improve the precision of conclusions.
Conclusions
Infants are interested in faces from the outset and evidence that they are treated differently
This preference is perhaps driven by the unusual/ unique perceptual characteristics of faces, not face-specific or even species-specific)
A dual process model has been proposed to account for these changes CONSPEC-CONLERN
Face processing becomes increasingly specialized
The commitment seems to be driven by experience
What did Sugita’s (2008) studies into experience based learning find?
Separated 10 monkeys from their monkeys at birth.
The monkeys were not exposed to facs for a set time period of either 6, 12 or 24 months.
Then tested their looking preference to human faces, monkey faces and objects.
Found that the primates were interested in faces despite having no experience of faces, but this preference was not species specific.
Following the period of face deprivation they were either exposed to human faces or monkey faces.
One month of human experience led to human preference and one month of monkey experience led to monkey preference.
These preferences were maintained for at least 12 months post deprivation.
This is strong evidence of experienced-based learning.
Pascalis et al. (2002)
Studied adults, 6 month olds and 9 month olds
Used a visual paired comparison paradigm
Measured discrimination through looking preference
Habituated to one face
Investigated whether they developed a novelty preference to a new face.
Experiment 1:
Two different human faces
Experiment 2:
Two different monkey faces
Adults show novelty preference in the human test phase but not the monkey test phase
9 month olds show a similar pattern with longer looking times for the novel human face but no difference in looking times for the monkey faces.
6 month olds’ looking times reveal novelty preferences to both human and monkey faces.
This suggests 6 month olds are better at discriminating monkey faces than 9 month olds and adults.
Further experiments (2005)
Infants exposed to monkey faces between 6 and 9 months of age maintained the ability to discriminate monkey faces at 9 months
What is perceptual narrowing?
Infants improve at dealing with stimuli that are meaningful for the environment while at the same time lose the ability to deal with stimuli that are irrelevant to the environment.
Perceptual biases to look at faces provide input into the system.
Experience with the input drives perceptual learning so that infants get better at processing stimuli they are often exposed to as this is needed to understand their world.
If infants are often exposed to human faces they will exhibit perceptual narrowing to human faces so they get better at processing human faces at the cost of becoming worse at processing other species’ faces.
However this is not exclusive to human faces, if exposed to monkey faces they will be better at processing those, or both.
This is indicative of both plasticity and experience-based learning
Thus perceptual narrowing is experience dependent.
However Sugita’s experiment suggests that there are limits to this plasticity as looking preferences were carried by the initial block of exposure.
What is the other-race effect (ORE)?
Human adults are expert at recognising faces but susceptible to errors when the target face is from an unfamiliar racial group. This is the other race effect (ORE).
This can be explained in part by experience.
Each individuals’ face prototype is an average of all the faces they have experienced and is unique to that individual.
Can be observed in infants to investigate learning from experience.
Kelly et al. (2005)
Visual preference task
Images of different face classes were paired.
All the participants tested were caucasian
Neonates exhibited no looking preference and spent 50% of the time looking at each image respectively.
3 month olds looked to caucasian faces for 60% of the time and 40% to other races. Showed no difference when two caucasian faces were paired.
Concluded that there is no preferential looking present in the first days of life but that this emerges by 3 months.
Supports the idea of a broad and unspecified face processing system at birth.
Early predominant exposure to one race (usually their own) tunes facial prototype to own-race dimensions.
This tuning leads to a preference to look towards own-race faces as more familiar stimuli.
Need to investigate infants in more heterogeneous environments.
Kelly et al. (2007)
Investigating discrimination
Habituated to a face in one orientation
Shown a test pair that differed in orientation
Test pair contained novel face and familiar face
Repeated for Africa, Asian, Middle Eastern and Caucasian stimuli.
All the infants studied were caucasian
The pattern of novelty preferences by face type varied across ages.
3 months showed novelty preferences across all four conditions
6 month olds showed novelty preference for chinese and caucasian faces only
9 month olds showed novelty preferences for caucasian faces only
From this we can conclude that there is no other-race effect at 3 months, partial ORE at 6 months and complete at 9 months.
Important to note that this is modifiable.
Consistent with perceptual narrowing account
Early predominant exposure to own race tunes facial prototype to own-race dimensions
This tuning leads to a preference to look toward own-race faces as a more familiar class of faces
As a result, superior recognition and discrimination develops for faces within own racial group/ more correct to say perceptual experience.
Kelly et al. (2010)
Investigates the cultural effects on adult eye movements and whether these are social norms or perceptual preferences.
Studies using western participants have found the typical scan pattern of people looking at faces focused on the eyes and mouth.
However south asian observers’ scan patterns focus most to the nose region.
If a human face is subbed with a sheep face the same scan path is observed.
Then repeated using greebles which are an invented category of novel objects which either differ from each other with a small number of common configurations.
Individuals can become expert at telling them apart after habituation
Culturally different scan patterns carry to greebles.
Perceptual preference as it is not influenced by the presence of other people but the result of perceptual tuning from experience of south asian faces.
Wheeler et al. (2011)
Studied caucasian infants 184-300 days old.
Observed the viewing patterns for 30 seconds when presented with caucasian and black african faces.
Subsequently examined the areas of interest.
Found that for own race only, interest in eyes increases with age while interest in mouths decreases (interest in noses is unchanged).
Repeated the study with chinese infants 122-284 days old.
Observed their viewing patterns for 30 seconds when presented with chinese and caucasian faces.
Subsequently examined areas of interest.
For own race, interest in noses increases with age, for other race, nose interest declines with age.
What is the link between perception and culture.
Cultural differences in scanning patterns in adults are not specific to faces and not simply the result of social norms.
In infants, scanning patterns reflect attention being directed to parts of the face that promote the most meaningful information in that environment.
Once adopted for faces this scanning pattern for identification is extended to other races faces and non-race stimuli.
What is autism and does it affect face processing?
Cultural differences in scanning patterns in adults are not specific to faces and not simply the result of social norms.
In infants, scanning patterns reflect attention being directed to parts of the face that promote the most meaningful information in that environment.
Once adopted for faces this scanning pattern for identification is extended to other races faces and non-race stimuli.
There is a large evidence base to suggest that people with autism process faces differently
Two explanations for differences in face processing.
Social hypothesis: reduced interest in, or motivation to, process faces (domain specific).
Non-social hypothesis: difference in general processing (domain-general)
Evidence is mixed and contradictory.
