Bruce & Young’s model of face recognition (1986)
Errors in face recognition (Young, 1986) Failure to recognise a familiar person (e.g. if they had grown a beard) Mistakenly identifying one person as another (typically in poor viewing conditions) Recognising a person but being unable to place them or remember their names Inability to remember someone’s name, even when other personal details were recalled
Young et al. (1986) CONCLUSION: there are at least THREE separate systems involved in face recognition: A face recognition system A semantic system that includes general information about people (job, etc.) A name system
View-centred descriptions Expression-independent descriptions Structural Encoding WE LOOK AT A FACE: Information is first encoded: this information is expression INDEPENDENT, first because we recognise faces irrespective of their expression and second because there is clinical evidence that we process faces and emotional expressions separately. Some patients can recognise faces and not emotions and vice-versa.
View-centred descriptions Expression-independent descriptions Structural Encoding Face Recognition Units (FRUs) This structural encoding stimulates the FACE RECOGNITION UNITS. The closer the correspondence between the face that is seen and the information that is stored in a FRU, the stronger the activation will be.
View-centred descriptions Expression-independent descriptions Structural Encoding Face Recognition Units (FRUs) Cognitive System Person Identity Nodes (PINs) The FRUs are linked to both the cognitive system and to person identity nodes (PINs). Face recognition makes available other information about the person from which identification can be made, such as their job or hobbies. Information from the cognitive system is included because we may often recognise someone using information other than their face, such as their voice or the way they walk.
View-centred descriptions Expression-independent descriptions Structural Encoding Face Recognition Units (FRUs) Cognitive System Person Identity Nodes (PINs) Name Generation The PIN is the point at which recognising a person can be said to have taken place. After this, name generation can occur! This is the process most likely used whenever we are identifying a person with whom we are familiar.
Directed visual processing View-centred descriptions Expression-independent descriptions Structural Encoding Face Recognition Units (FRUs) Cognitive System Person Identity Nodes (PINs) Directed visual processing Name Generation The Directed Visual Processing Unit is included because we can choose to focus on certain aspects of a face – e.g. if we are meeting a friend in a crowded space we may choose to look for someone with a particular hair colour.
View-centred descriptions The Facial Speech Analysis unit is included because there is clinical evidence that lip reading is a separate ability from face recognition: some patients with brain damage can read lips but not recognise faces, and vice-versa. View-centred descriptions Expression-independent descriptions Structural Encoding Face Recognition Units (FRUs) Cognitive System Person Identity Nodes (PINs) Directed visual processing Facial Speech Analysis Name Generation
View-centred descriptions Expression analysis helps us to recognise the angry person in a crowd, etc. Like facial-speech analysis and directed visual processing, this relates more to the recognition of unfamiliar faces. View-centred descriptions Expression-independent descriptions Structural Encoding Face Recognition Units (FRUs) Cognitive System Person Identity Nodes (PINs) Directed visual processing Facial Speech Analysis Expression Analysis Name Generation
Using the model Failure to recognise a familiar face: There is insufficient stimulation of a FRU for the face to be recognised. Misidentifying one person for another: The person misidentified is sufficiently similar to the familiar person to activate a FRU inappropriately. Not being able to place a person: The relevant FRU has been activated but the stimulation was not sufficient to activate the Cognitive System or the PIN and so retrieve more information. Not being able to remember a name: A problem has occurred between the PIN and Name Generation system.
Evaluation of the model The model can predict and explain many research findings particularly relating to brain-damaged patients. It can account for the recognition errors reported by Young et al.’s (1986) research. E.g. almost 20% of P’s referred to instances where they knew a lot of information about a person but couldn’t think of their name, however no-one reported being able to name a face whilst knowing nothing else about the person.
Evaluation of the model (cont) Some of the components have been less well explained than others. The role of the cognitive system is unclear. The model focuses on the recognition of familiar faces and is of limited value in helping to understand how, for example, eyewitness identification of unfamiliar faces can be improved. It does not account for the processes involved in learning to recognise new faces and storing these images in memory. The model is neutral on the question of whether face recognition is a special perceptual process that is qualitatively different from other types of object recognition. Evidence from brain-damaged patients is more revealing in relation to this!