L17: Faces

Case study P.S.

• Had prosopagnosia (couldn’t recognise races, even her own)

Damage to what part of the brain presents prosopagnosia

Fusiform face area

• Region specific to processing faces

• In the inferior temporal lobes and fusiform gyrus

Bruce & Young: Cognitive model of face processing

• How they thought we extract information from faces

• Two pathways:

  • Expression information

  • Identity information (i.e. who is that person)

• Two pathways suggested from prosopagnosia research

  • Prosopagnosic deficit located at face recognition unit stage

Prosopagnosia

• Damage to occipitotemporal cortex (acquired prosopagnosia)

• Unable to recognise familiar faces

• No impairments in identity of familiar people

  • Use other cues i.e. voice, name

• Ability to recognise and name other objects is spared

Ventral visual stream

• Concerned with identifying objects

• ‘What’ pathway

• Occipital to temporal lobes

• Information travels along the posterior parietal cortex

Dorsal visual stream

• Concerned with locating objects in space

• ‘Where’ pathway

• Occipital to parietal lobes

• Information travels along the inferotemporal cortex

What stream is largely concerned with face perception

Ventral visual stream

What are the three main face-sensitive areas in the brain

• Occipital Face Area (OFA)

• Fusiform Face Area (FFA)

  • Invariant aspects of faces (i.e. identity)

• Superior temporal sulus (STS)

  • Responds to face and bodies

  • Dynamic stimuli, changeable aspects of faces (e.g. expression)

Where is the Occipital Face Area (OFA) located

Inferior occipital gyrus

OFA role in facial recognition

• Early stage of face perception (occipital lobe)

• Sends input to fusiform and superior temporal regions

• Defined by greater response to faces than non-face categories

• Responds to upright and inverted faces

• Sensitive to physical changes in stimulus

FFA (fusiform face area) role in facial recognition

• Greater response to face vs. other non-face categories

• Responds more to upright faces

• Sensitive to changes in identity, relatively insensitive to physical changes

• Acquired prosopagnosia usually due to lesions in or around FFA

Where is the FFA located

Fusiform gyrus

What study shows us the OFA is sensitive to physical changes vs the FFA being sensitive to identity changes

Rotschtein et al (2005)

• Use face adaptation

• Greater signal change in OFA when physical properties were changed

  • Visa versa for FFA

Where is the STS (superior temporal sulcus) located

Superior temporal sulcus

STS (superior temporal sulcus) role in facial recognition

• Responds to faces, BUT also other stimuli like bodies, eye gaze, etc.

• Changeable aspects of faces (viewpoint, gaze direction, expression), but not identity

• Responds to moving bodies and to changes in gaze direction

Haxby, Hoffmann & Gobbini (2000): Neural Model for Face Perception

• ‘Replace’ Bruce and Young’s cognitive model

• Distributed neural system for face perception

• Proposed separate pathways for processing expression and identity

What pathway did Haxby, Hoffmann & Gobbini (2000) argue we extracted identity information

OFA (inferior occipital) to the FFA (temporal lobes)

What pathway did Haxby, Hoffmann & Gobbini (2000) argue we extracted identity information

OFA to the STS

What is the question of FFA being expertise of specificity referring to?

Are faces special (specificity) OR is the reason we have areas of the brain dedicated to faces is because we are exposed to them so frequently (expertise)

• Domain-general (expertise) vs. Domain-specific (face specificity)

What is the hypothesis that FFA expertise follows?

• Expertise hypothesis argues that face-specific mechanisms are highly specialised for distinguishing between exemplars of a category

  • If you had similar expertise in a non-face category, then same processing mechanisms would be engaged

What is the hypothesis that FFA face-specificity follows?

Kanwishes

• Face-specificity hypothesis argues that face perception is a process occurring in dedicated, specialized cognitive and neural mechanisms

Evidence for face-specificity hypothesis in FFA

• FFA responds more to faces than non-face categories

• Activation to faces is more consistent and much more robust than dace selective activity in OFA and STS (other face areas)

  • HOWEVER, on their own, neither of these conclusively provide evidence that FFA activation to faces supports face specificity hypothesis

Study that tested FFA expertise: Greebles

Gauthier et al. (1999): FFA sensitive to expert within-category visual discrimination

• Trained P’s to distinguish between greebles

• Results

  • Found increased response in FFA to upright vs inverted greebles in fusiform area

• Suggests some sort of holistic processing (hallmark of face perception)

  • Proof as Greebels NOT facial perception BUT object perception

Real-world expertise: Gauthier et al. (2000)

• Group of P’s that were bird watchers OR car experts

• Results

  • Highest response in FFA to faces

  • BUT in bird experts, see increased FFA activity compared to car experts when looking at birds

  • HOWEVER, not significant difference for cars and car expert

Counterevidence for FFA expertise hypothesis

Case study R.M: Sergent & Signoret (1992)

• Prosopagnosic

• Unable to identify himself or his wife

• Collected over 5000 miniature cars

  • No problem in discriminating his cars

  • THEREFORE, fusiform cant be ONLY region important in facial recognition

• Summary

  • No neuropsychological cases where facial recognition AND objects of expertise have BOTH been impaires, while recognition of nonexpert objects is spares (or visa versa)

Evidence against Kanwishes argument that the FFA is only important for face processing

• Ventral stream also important for object processing

• Haxby et al argued role in object recognition is also important

  • Distribution of responses may contain critical information for discriminating between objects

  • Patterns of activity important, not just one is higher one is lower

Why is object recognition in the fusiform (FFA) area unlikely to be crucially important

Prosopagnosia and visual agnosia

• P’s with prosopagnosia can have damage to fusiform area but object recognition remain intact

• Visual agnosia (unable to recognise object categories) but not prosopagnosia

  • Would expect some crossover if objection perception was important in the fusiform area

What was the conclusion reached when Calder & Young (2005) revisited both the cognitive and neural model of face perception

• Both models argued for early separation of pathways for changeable (i.e. expression) and unchangeable (i.e. identity) aspects

• On revision, Calder & Young (2005) concluded this could NOT happen so early

What evidence did Calder and Young (2005) use to conclude there couldn’t be such a clear separation between identity and expression so early on

• If processes are separate, should see double dissociation for identity and expression

  • Don’t see this

How can emotional response to neutral familiar faces be measured - automatic and conscious

Skin conductance response (SCR)

• Familiar faces elicit greater SCR than unfamiliar faces

SCR response to familiar faces with prosopagnosic individuals

• P’s with prosopagnosia also show increased SCR to familiar faces

  • Suggests two pathways to face recognion, one conscious (impaired in prosopagnosia) and one unconscious

What is the capgras delusion

Believe familiar people have been replaced by an imposter

SCR response to familiar faces with capgras delusion individuals

• No increased SCR for familiar faces (no emotional response)

• More evidence for dissociation between emotional response to faces and face recognition

• BUT P’s with ventromedial frontal lobe damage have no SCR to familiar faces, and no Capgras delusion