A Critical Review of Bruce and Young’s (1986) Face Recognition Model

Bruce and Young (1986) proposed a stage model of facial processing which is also

noted as Bruce and Young’s Face Recognition Model. It is a stage model because it presents

facial recognition as a sequence of stages. The stages are accessed serially; hence, one after

the other. There are two main paths within the model - one for recognizing familiar faces and

another for recognizing facial expressions. The model suggests three main assumptions – that

familiar and unfamiliar faces are processed differently, that facial identity and facial

expressions are processed separately, and that for familiar faces information about familiarity

is accessed before their names. The model is applied only to facial recognition but not general

object recognition (Bruce & Humphreys, 1994, Duchaine & Nakayama, 2005).

The earliest stage of the model, where face recognition begins as soon as a face is

seen, is called Structural Encoding. The first stage consists of View-Centered Descriptions

and Expression-Dependent Descriptions, where a very basic representation/pattern of the face

is built by analyzation of the expressions of each feature. For instance, if someone is smiling

the View-Centered Descriptions is going to analyze a smiling mouth. At the same time, the

Expression-Dependent Descriptions is going to process individual features which do not

change with expression.

After that basic first analysis the rudimentary information achieved in the earliest

stage is refined into three ways/sections. The first way/section is called Expression Analysis;

where at first comes a conclusion about the emotional state of the person being looked at, with

the use of cues from their facial expressions. Then, the focus is being moved on to the lips, to

process lip and facial movements in speech perception – to what is being said, if anything is

said. This is the second way/section, called Facial Speech Analysis. In parallel to the

Expression Analysis and Facial Speech Analysis, particular characteristics of the face are also
being noticed, like facial hair, glasses, etc. This is an important check for the face recognition

and it is happening through the third way/section, which is called Directed Visual Processing.

Then, the face is compared to the Face Recognition Units (FRU). The FRU contain

structural information about known faces, stored representation of familiar faces; or, hold

every face ever stored in memory. The FRU work with structural similar similarities and look

for something particular that is recognizable, such as the shape of nose of someone. The FRU

work fast to compare the incoming information to the familiar faces stored in the memory

system, and if there is a match there comes a sense of recognition. Also, if there is a match the

Person Identity Nodes (PIN) are accessed. The PIN contain biographical information about

faces. For example, what job this person has, or hobbies, or partner, etc. Then, finally, a name

is generated.

After that, what is known as the last stage of the facial recognition model is the

Cognitive System. The Cognitive System holds all information about faces, adds, and

combines new information. It is vast and complicated because it has prior knowledge relevant

to face recognition. It holds all the visual memories of faces, the environment they are usually

seen in, the feelings these faces hold and the feelings they leave (Pascalis et al., 2011). The

cognitive system is also used throughout various stages and due to that research suggested

that it should not be viewed as the ‘last’ of the stages (Young et al., 1985; Ellis, 1992).

The cognitive system was also seen as able to play far more integrative role in the

face recognition process, which made it one of the main reasons for the model to become a

target for criticism over the last three decades. More specifically, due to lack of clarity

specified concerning the localization of the cognitive system or how it is accessed, parts of the

process seem as vague.

commendatory. Most of the evidence used to support the stage face recognition model came

from Cognitive Neuropsychology studies involving sample with impaired facial processing

function. For example, Bruce and Young (1993) conducted investigations to soldiers who

have been brain damaged. It was found that some experienced difficulties in matching

unfamiliar faces, whereas others had difficulties with familiar face recognition. Evidence

showed that one part of the face recognition system may be damaged, but another one will

stay intact. For instance, the soldiers could recognise that a face is familiar but they did not

know where from, or they could recognise somebody from their facial expression, but not

entire face. Bruce and Young’s (1993) study found supporting evidence for Modularity; or

that the face recognition system is being a process of modules. As it appeared that the soldiers

had problems with a specific part of the system but not the entire system, and that deficits can

occur at different stages.

More evidence supporting the Modularity of the model was presented by Brunsdon

et al. (2006) that conducted testing on Al, an 8-year-old child with Prosopagnosia.

Prosopagnosia is a cognitive disorder in which the ability to recognize faces is impaired

(Bornstein & Kidron, 1959). Al showed deficits at the level of structural encoding; thus, he

was unable to achieve primal basic perception of faces. Al was treated with a programme

focused particularly on perception, which resulted in excellent face naming for familiar faces.

Thompson (1980) also presented further supporting evidence for the modularity of

the facial recognition model, though through a more holistic approach to viewing faces - with

the ‘Thatcher illusion’. In his experiment an inverted picture of Margaret Thatcher was used

in which the eyes and mouth remained in normal perspective. It was possible and easier to

recognize the facial expressions of Margaret Thatcher in that case than with a truly inverted
face. This suggested that facial features are processed separately, and faces are not seen in as

much detail as thought.

Moreover, further Neuropsychological studies with people with Prosopagnosia

showed additional supporting evidence; or more specifically, concerning that facial and non-

facial recognition rely on separate mechanisms. Neuropsychological tests showed that those

with Prosopagnosia can not recognise faces, either through injury or disease, but are able to

recognise objects. This evidence suggested that face and object recognition happen in

different parts of the brain using different processes (Duchaine & Nakayama, 2005).

Bruce and Young’s (1986) model was also evidenced by a case study conducted on a

person known as Mr. W.J., whom was a 54-year old farmer with a Bilateral Occipital Lobe

Lesion, and profound Prosopagnosia (Bruyer et al., 1983). He could not recognise faces, even

his own face; but he could pick faces from other objects and animals, he could identify the sex

of faces, and he could recognise his sheep. The uniqueness of this case is that it showed

Prosopagnosia as a memory disturbance rather than a perceptual deficit; thus, supporting

further the modular way in which face recognition takes place. Mr. W.J.’s case showed that

while some parts of the facial recognition system can retain others can get lost.

However, further tests with Mr. W.J. made again Bruce and Young’s (1986) model a

subject of criticism. In a study of McNeil and Warrington (1993), the farmer showed that he

was able to identify faces but with no sense of recognition, which should not be possible

according to the model. Their evidence suggested that the face recognition model did not

explain familiarity without awareness - the so called Covert Recognition.

Besides that omission and the lack of information concerning the cognitive system

stated further above the text, there are some general limitations in the current face recognition

model. For instance, Bruce and Young (1986) did not present certainty about localization of
the face processing system. Moreover, supplementary research on face perception included

Visual Perception, which was not specified by Bruce and Young as well (1986) (Haxbi &

Gobbini, 2011).

Although, the face recognition model of Bruce and Young’s (1986) is generally

considered as highly useful as it stimulated research on face recognition over the last three

decades. The stage model stimulated research including positron emission tomography (PET)

scans, which have shown that different areas of the brain are accessed during the face

recognition process suggesting that it is Modular (Sargent & Signoret, 1992). Further

Neuroscientific evidence showed that there was an activation identified in the anterior middle

temporal gyrus when familiar faces are processed (Leveroni et al., 2000). In a study by

Rotshtein et al., (2005) with functional magnetic resonance imaging (fMRI), was found that

the right fusiform gyrus showed sensitivity to identity changes rather than physical changes,

suggesting the location of FRU.

Bornstein, B., & Kidron, D. P. (1959). Prosopagnosia. Journal of Neurology,

Neurosurgery & Psychiatry, 22(2), 124-131.

Bruce, V., & Humphreys, G. W. (1994). Recognizing objects and faces. Visual cognition,

1(2-3), 141-180.

Bruce, V., & Young, A. (1986). Understanding face recognition. British journal of

psychology,77(3), 305-327

Brunsdon, R., Coltheart, M., Nickels, L., & Joy, P. (2006). Developmental prosopagnosia:

A case analysis and treatment study. Cognitive Neuropsychology, 23(6), 822-840.

Bruyer, R., Laterre, C., Seron, X., Feyereisen, P., Strypstein, E., Pierrard, E., & Rectem,

D. (1983). A case of prosopagnosia with some preserved covert remembrance of

familiar faces. Brain and cognition, 2(3), 257-284.

Duchaine, B., & Nakayama, K. (2005). Dissociations of face and object recognition in

developmental prosopagnosia. Journal of cognitive neuroscience, 17(2), 249-261.

Ellis, A. W. (1992). Cognitive mechanisms of face processing. Philosophical

Transactions of the Royal Society of London. Series B: Biological Sciences,

335(1273), 113-119.

Haxby, J. V., & Gobbini, M. I. (2011). Distributed neural systems for face perception (pp.

93-110). The Oxford Handbook of Face Perception.

Leveroni, C. L., Seidenberg, M., Mayer, A. R., Mead, L. A., Binder, J. R., & Rao, S. M.

(2000). Neural systems underlying the recognition of familiar and newly learned

faces. Journal of Neuroscience, 20(2), 878-886.

Quarterly Journal of Experimental Psychology, 46(1), 1-10.

Pascalis, O., de Martin de Viviés, X., Anzures, G., Quinn, P. C., Slater, A. M., Tanaka, J.

W., & Lee, K. (2011). Development of face processing. Wiley Interdisciplinary

Reviews: Cognitive Science, 2(6), 666-675.

Rotshtein, P., Henson, R. N., Treves, A., Driver, J., & Dolan, R. J. (2005). Morphing

Marilyn into Maggie dissociates physical and identity face representations in the

brain. Nature neuroscience, 8(1), 107-113.

Sergent, J., & Signoret, J. L. (1992). Functional and anatomical decomposition of face

processing: evidence from prosopagnosia and PET study of normal subjects.

Philosophical Transactions of the Royal Society of London. Series B: Biological

Sciences, 335(1273), 55-62.

Thompson, P. (1980). Margaret Thatcher: a new illusion. Perception.

Young, A. W., Hay, D. C., & Ellis, A. W. (1985). The faces that launched a thousand

slips: Everyday difficulties and errors in recognizing people. British Journal of

Psychology, 76(4), 495-523.