TMA01: Are faces ‘special’ in terms of how we process and recognise them? Critically discuss with reference to theoretical models of recognition.

TMA01: Are faces ‘special’ in terms of how we process and recognise them? Critically discuss with reference to theoretical models of recognition.

Faces can be viewed as representations of stimuli that we rely on when making social interactions. They provide us with the information regarding an individual’s mood, age, identity, attractiveness, approachability and race among others. This tends to be remarkable more so if people have the notion that faces possess similar composition features. Their uniqueness with regard to the social signals that they convey illustrates why face perception has played a critical role in social interaction for millions of years. Due to its importance, face processing has gained prominence in most cognitive science areas of research that include behavioural, neuro-imaging and neuropsychology studies (Rivolta, 2013). The essay tries to show how faces are “special” in terms of how they are processed and recognised compared to other objects.

Humphreys and Bruce (1989; cited in Kaye, 2010) postulate that recognition of the face and other objects fit into a wider context of cognition, which involves perception, categorization and naming. Marr, (1982) developed a primal sketch with five stages on how objects are processed.

Early Visual Processing



 Recognition involves the process of comparing objects descriptions as they are generated in the retinal image with other descriptions that are stored in the long-term memory. There are different types of recognition based on what is being recognised.  Naming an object is not necessarily a vital component of being able to recognise objects. A good example is how animals do not have the capacity for language, but they can still recognise objects. However, the names that are allocated to things give clues to the existence of different types of recognition (Humphreys and Bruce 1989; cited in Kaye, 2010).

In an attempt of trying to ascertain whether faces are special with regards to how they are processed and recognised, the essay examines three theoretical models of recognition. The models include Marr and Nishihara’s theory of object dimension, behavioural experiments and neurological research. Marr and Nishihara’s theory is based on the establishment of a canonical coordinate frame that helps in defining the central axis of an object. This is easy to identify if the object being examined has a line of symmetry or a length that is greater than its depth and width. Establishing the central axis is a critical stage for this recognition process. This means that if it is difficult to establish the central axis, then it is also difficult to recognise an object (Kaye, 2010). Some of the evidence to support this model have emanated from various neuropsychological case studies. Warrington and Taylor (1978; cited in Kaye, 2010) ) postulated that patients who had damage on the right hemisphere of their brain had the ability to recognise objects presented in a typical view, but failed to recognise them when presented in unusual views. When it comes to the neurological research model, emphasis is placed on how an individual recognises objects when they have various brain disorders. In this case the essay focuses on prosopagnosia disorder. On the other hand, behavioural experiments involve trying to ascertain the credibility of two opposing views with regards to face processing and recognition. The views involved here are those of domain specificity and expertise hypothesis. Domain specificity hypothesis is of the view that face recognition uses different neural and cognitive processes compared to other object recognition. Expertise hypothesis on the other hand, postulates that expertise plays a vital role in creating holistic mechanisms for faces. It argues that humans are better when processing faces compared to objects just because of the increased experience that they have with the faces (Rivolta, 2013).


The main debate in the essay is whether object and face recognition differ in any way. Presence of differences would make one of them “special” compared to the other. To begin with, Marr and Nishihara’s theory of object dimension is more concerned with the identification of a central axis for there to be a possibility of recognising an object. It is easy to identify various objects by use of this model, more so if the object is 3D. However, faces have 3D models, but it is rather problematic recognising them by use of this model.  The model has concentrated on the perceptual classification stage of object recognition process. The stage might provide useful information, but there is need for more interaction when it comes to faces. For example, when an individual is confronted by another person, they not only want to know whether it is a human face, but also want to know whose face it is (Kaye, 2010). Face processing and recognition can be categorised at different levels. At the first level it is where we decide that the stimulus is a face and not another object. At the next level it is decided whether the face is female or male, and maybe derive other information like ethnic origin and attractiveness. The recognition process is then followed by the decision whether the face is familiar or not. When the face is familiar, now there is the need to identify to whom it belongs to. It is through this level where we identify that face processing and recognition is rather different from that of other objects (Kaye, 2010). A good example that can provide evidence to this is the research study conducted by Kemp et al. (1997), which was aimed at ascertaining people’s ability to recognise unfamiliar faces. From the research, it was observed that people have problems matching unfamiliar faces even when presented simultaneously. The experiment involved observing cashiers on how well they would match shoppers with photographs from their credit cards. The research identified that the cashiers had the tendency of accepting credit cards from people that bore resemblance to the credit card photograph. Only 36% of the rejected credit cards in this study did not have photographs of the shoppers holding them at that moment. In cases where the photograph was of someone else who bore some resemblance to the shopper, the correct decision rate in rejecting the cards was only 66% (Kaye, 2010). Therefore, this study was an indication that the usage of Marr and Nishihara’s theory of object dimension is not appropriate for face recognition. This means that faces are special with regards to how they are processed and recognised.


For the behavioural experiments, more emphasis is placed on trying to ascertain the truthfulness of domain specificity and expertise hypothesis, which seem to contrast in opinion with regards to face recognition and processing. To begin with, let us examine the face-inversion, which was first described by Yin (1969). The effect shows that in experimental environments, if people were given a task of learning and remembering faces that they have not seen before, they would have between 20-25% better chances of obtaining the desired results when the faces involved are shown upright as opposed to when they are upside down. This form of inversion also affects other object recognition. However, the effect is smaller compared to that of the faces. For objects it is up to 8% (McKone et al., 2009). According to Yin (1969), the disproportionate inversion effect emanates because face recognition requires holistic processing and it is difficult to extract all the information with inverted faces. Evidence of holistic processing can be obtained from an experiment in McKone et al. (2009) conducted by Tanaka and Farah. In this experiment, people had to learn certain individuals like “Tim”, and later asked to recognise some of their features. Tim’s nose was put in isolation and people were to differentiate it from another person’s nose. Participants in the research were required to identify whether Tim’s nose was the one on the right or left side of the computer screen. The research indicated that face recognition is better when the features are within the face compared to when they are placed in isolation. The experiment shows the relevance of domain specificity hypothesis. This is an indication that faces processing and recognition is “special” since it is affected differently compared to other objects.

In an attempt of discarding the expertise hypothesis, which does not support the notion that faces are “special” in ways that they are processed and recognised an experiment has been reported in McKone et al. (2007). The authors of the experiment tested dog experts who had an average of 23 years experience and novices who had no such knowledge in dog recognition. Both of these experimental groups were experts in human face recognition. Following the expertise hypothesis, the manipulation involved should demonstrate holistic effects in both face processing and dog recognition for experts only. In one of the experiments, both the novices and dog-experts had to memorise upright faces and dogs. After a certain interval in which they had to complete another different task, they were shown the memorized stimuli on a screen together with other distracters that were not shown during the learning phase. Each participant was required to identify whether the stimulus they had learnt previously was on the right or left side. This procedure was repeated, and this time the stimuli were upside-down. As it would have been expected, novices showed the inversion effect. The memory performance for faces was affected more than that of dogs. However, despite the dog-experts’ experience, they recorded higher face inversion compared to dog inversion effects. This works out in favour of domain specificity that asserts “faces are special and processed by holistic processing” (McKone et al., 2007).

When it comes to the neuropsychological research model, prosopagnosia can be selected to identify whether faces are special in the way they are recognised and processed.  Prosopagnosia is a cognitive disorder whereby an individual’s ability to recognise faces becomes impaired while other intellectual and visual processing aspects remain intact. Most people suffering from this disorder find it difficult to recognise even their friends, family members and partners. The brain is not necessarily damaged for an individual to suffer from prosopagnosia. McKone et al. (2007) have included several studies that prove the prevalence of the disorder. Among the studies is one involving RM who was a car sales man with vast experience in this field. RM later on came to suffer from aneurysm, which impaired some of his abilities. He became extremely poor in face recognition, but he still retained his expertise with cars. RM still had the ability to recognise far more models, years and makes for different cars. This was an indication that he suffered from pure face deficit, but retained the recognition of other objects. From such an occurrence it would be prudent to conclude that object recognition differs from face recognition in various ways. If they are processed in a similar manner, RM would have recognised both or failed to recognise both of them after contracting the aneurysm (McKone et al., 2007).

In conclusion, faces are “special” in terms of how they are processed and recognised. Among the reasons for this conclusion is that faces look alike since they have similar features that appear in similar positions. Due to the similarity, there is need to use different forms of visual information in order to recognise a face from those used to recognise other objects like tables. Evidence supporting this stance has been presented above through various studies that included things like inverting or turning images upside-down. The visual stimuli impair people’s ability to recognise faces when compared with recognition of other objects. This has been identified as the inversion effect. Marr and Nishihara’s theory of object dimension that postulates identification of the central axis makes it possible for identification of objects does not work with faces. This is because it is not only about identifying that it is a face, but also ascertaining whose face it is. Several aspects need to be identified before a conclusion is made.  Evidence against the expertise hypothesis that argues against faces being “special” has also been supplied. There is also evidence to show that domain specificity holds some level of authenticity in its view.



Humphreys, G.W. and Bruce, V. (1989). Visual Cognition: Computational, Experimental and Neuropsychological Perspectives: Hove, Lawrence Erlbaum Associates Ltd.

Kaye, H. (2010). Cognitive psychology (2nd ed.). Milton Keynes: Open University.

Kemp, R., Towell, N. and Pike, G. (1997). When seeing should not be believing: photographs, credit cards and fraud. Applied Cognitive Psychology, 11(3), 211-225.

Marr, D. (1982). Vision: A Computational Investigation into the Human Representation and Processing of Visual Information, New York: W.H. Freeman and Company.

McKone, E., & Yovel, G. (2009). Why does picture-plane inversion sometimes dissociate perception of features and spacing in faces, and sometimes not? Toward a new theory of holistic

processing. Psychonomic Bulletin & Review 16, 778–797.

McKone, E., Kanwisher, N., & Duchaine, B. (2007). Can generic expertise explain special processing for faces? Trends in Cognitive Sciences, 11(1), 8–15.

Rivolta, D. (2013). Cognitive and Neural Aspects of Face Processing. Journal of Cognitive Systems, 10(11), 23.

Yin, R. K. (1969). Looking at upside-down faces. Journal of Experimental Psychology, 81,



Caramazza, A. and Coltheart, M. (2006). Cognitive Neuropsychology twenty years on. Cognitive Neuropsychology, 23, 3–12.

Do you need an Original High Quality Academic Custom Essay?