Perception

Based on

• Sensory Date
• Previous Knowledge
• Developed Hypothesis

'Depth cues' can create expectations of size and distance.

- Visual illusions are caused by misapplied hypothesis i.e usually work in real work perception

Supports Gregory's Theory...

• The Necker Cube: An example of 'hypothesis testing' - we cannot decide which hypothesis is right due to inadequate cues.
• The Muller-Lyer Illusion: Look different lengths due to our experience with depth cues, e.g. buildings.

+ Khorasani et al. found illusions had less effect when participants knew about them.

+ Bruner and Minturn (1955) digits 1,3 perceived as B in letters and 13 in numbers.

+ Segall et al. (1963) found people living in round huts less likely to perceive line length differences in Muller-Lyer illusion as they have no experience with corners of rooms and edges of buildings.

+ Theory explains how we can continue to perceive in ambiguous situations of where the retinal image is poor.

- Based largely on studies in artificial settings i.e illusions in lab settings (lacks ecological validity)

- Theory fails to explain why we keep seeing illusions when we know our brain is misleading us. 

Information is gathered from simply what we see

We have sufficient information from the...

Optic Array: Bundle of light rays moving towards observer from any point of view.

Optic Flow: Objects directly infront remain stationary while objects to the side appear to move.

Invariants: Don't change as we move 

- Texture gradient: Closer the object the rougher, further away objects finer.

- Horizon ratio: If position on the horizon is the same then objects are same size.

Affordance: Objects offer action potential

Resonance: How we make sense of the broadcast information. 

+ Explains how perceptual abilities evolved.

+ Correctly highlights information in the optic array.

+ Practical applications - optic flow concept used to train pilots.

+ Johansson (1973) placed small lights of persons body when they moved children and animals responded appropriately.  Suggest its innate.

+ The Ames Room - without a horizon we assume it runs across.

- Ignores situational and cultural factors.

- Explains species perception but not individuals perception.

e.g how a bushman would know you put things in a postbox.

- Must involve some top-down processes (knowledge and hypothesis)

Depth Perception:

- Monocular Cues: Visual cliff (Gibson and Walk 1960) - Most infants refused to crawl across to their mothers, so can perceive depth.

- Binocular Cues: (Bower et al 1970) - 1 week old infants shown 2 objects (one for each eye.) They grabbed out showing they can perceive 3D objects.

Visual Constancy:

- Shape Constancy: Bower (1966) operant conditioned infants to prefer slanted rectangle shape, shown other shapes but still preferred rectangle.

- Size Constancy: Slater et al. (1990) infants habituated to cubes, different sizes.  2 cubes placed, 1 larger and further away to make look equal size.  Infants looked longer at unfamiliarised cube.

Suggest perception is innate and bottom-up.

BUT

'Perceptual completion' needs previous knowledge.  E.g Infants cannot perceive Kaniza's square.

- Hard for infants to communicate accurately.

- Experimenter bias more likely to occur.

- Poor concentration of infants.

Depth Perception:

- Turnbull (1963) pygmy man perceived buffalo as insects as used to living in dense forest and seeing them from far away so they looked small.

- Hudson (1960) European and Bantu children shown picture of elephant in background, man and antelope in foreground. Bantu less likely to use correct depth cues.

Visual Constancy:

- Shape Constancy: Allport and Pettigrew (1963) zulu people in South Africa could not perceive window illusion due to no experience with windows.

- Size Constancy: Segall et al. (1963) zulu's living in round huts could not perceive Muller-Lyer illusion.

- Language barrier for cultural studies.

- Imposed etic may occur as a result of a study.

- Natural experiments are hard to repeat so are not very reliable or generalisable.

Nature

- Gibson and Gibson (1955) Differentiation Theory.  We can learn the differences between objects from simply just sensory data.

- Infant Studies: The Visual Cliff.

- Cross-cultural studies: Johoda and McGurk (1974) found no difference between Scottish and Gahnian when testing depth perception using pictures.

Nurture

- Jean Piaget (1974) Enrichment Theory.  Sensory data is not sufficient.  We need cognitive schemas (expectations) to enrich the data.

- Infant Studies: Yonas et al. (2001) older children reached out more to objects that appeared closer due to shadows used as depth cues.

- Cross-cultural studies: Pygmy man perceiving buffalo as insects.

Blakemore and Cooper (1970) kitten in drum.

- Both require learning, so a combination...

• Gopnik and Meltzoff (1997) 'theory' theory.  We have innate knowledge but cognitive theories require learnt knowledge aswell.

• Development of the brain (nature) relies on outside world experiences (nurture).

Model:

Remember order of:

Structual Encoding

View-Centered Descriptions

Expression - Independent Descriptions

Expression Analysis - Facial Speech Analysis - Directed Visual Processing

Face Recognition Units - Person Identity Nodes - Name Retrieval Units

          FRU                                    PIN                               NRU

THE COGNITIVE SYSTEM

+ Shows precise predictions which can be easily tested.

+ Shows differences in how familiar and unfamiliar faces are detected.

- Details of processing are vague.

- Cognitive system not clearly specified.

- Each component not fully explained.

Young et al. (1985)

22 participants had to keep a record of the face recognition mistakes they made over 8 weeks.

They found that people could not recall names without personal identity information.

...supports that PIN comes before NRU.

If not then Bruce and Young are wrong.

Special:

Sergent et al. (1992) Fusiform face area (FFA) more active when identifying face then objects.

Not Special:

Gauthier et al. (2000) Participants asked to distinguish between bird types and car types. The FFA was also active.

Two Categories:

Apperceptive - cannot draw/match objects

Failure to recognise object because one cannot perceive it's form.

Associative - can draw/match but can't recognise what they are

Occurs at a later stage.  Cannot recognise an object because one cannot access stored knowledge in the brain.

Apperceptive Patient: Benson and Greenberg (1969) man could not name or match objects or faces but could identify numbers and colours.

Associative Patient: HJA could copy and match objects but was unable to name them.

Can describe face and expression but cannot recognise it.

Many patients can recognise objects but not faces, supporting that face recognition is special...

De Renzi and Di Pellegrino (1998)

VA could recognise objects but not faces.

Prosopagnosia supports Bruce and Young's model because participants can decode emotional expressions but not recognise faces.

However, overall evidence can not prove that face recognition is special.