Australasian Science: Australia's authority on science since 1938

Face Off

By Tim Hannan

Do motoring enthusiasts recognise cars in the same way people recognise faces?

A recent investigation by a team of cognitive neuroscientists from Vanderbilt University, Tennessee, has challenged long-held assumptions about the visual perceptual systems involved in face and object recognition.

It has been widely accepted that the ability to recognise faces involves the function of a highly specialised cognitive system located in the fusiform gyrus, which extends across the occipital and temporal lobes. This face recognition system is presumed to be damaged in cases of the developmental or acquired condition termed prosopagnosia, in which a person experiences a marked deficiency in recognising and discriminating others’ faces yet retains the ability to recognise other objects, such as chairs, dogs and cars. The existence of separate systems for face processing and object recognition has been supported by neuroimaging research, which has demonstrated that part of the fusiform gyrus in the temporal lobe – labelled the fusiform face area (FFA) – is activated by faces but not by other objects.

The assumption that face processing relies on a genetically determined system with, as one researcher expressed it, “its own private piece of real estate in the brain”, has now been put under the microscope by Rankin McGugin, Isabel Gauthier and their colleagues at Vanderbilt University. In the Proceedings of the National Academy of Sciences they observed that neuroimaging research has not settled the question of the specificity of the FFA to faces.

Most prior studies used only standard resolution functional magnetic resonance imaging (fMRI), and those few studies that employed high resolution methods did not control for the possibility that perception is affected by experience. This is critical, as previous research has demonstrated that significant experience with objects (e.g. cars, planes, birds) influences the speed and accuracy of recognition. In particular, while most people direct their gaze to parts of an object and not to the object as a whole, those with extensive expertise with an object employ a more global, holistic perceptual process that results in swifter and more accurate recognition. This enhanced visual recognition may be observed in the behaviour of radiographers, chess masters and motoring enthusiasts identifying cars.

Using high resolution fMRI, the team recorded FFA activity in individuals viewing randomly presented images of cars, planes, animals and human faces. Consistent with previous research, they found markedly increased activity in the FFA when faces were presented. However, when participants with demonstrated expertise in the visual identification of cars were examined, increased activity in the FFA was found, similar to what is observed for face recognition.

McGugin and colleagues concluded that, while the study does not demonstrate that the same groups of neurons are involved in both face and object recognition, the findings argue against a brain region exclusively devoted to facial recognition. Rather, the FFA also assists the chess masters, radiographers and car enthusiasts to recognise objects within their particular area of expertise.

While the localisation of cognitive systems is of intrinsic interest to neuropsychologists, the study’s findings have other potential implications. If the neural systems underlying face recognition also serve the recognition of other objects, and if expertise enhances access to this system, then it may be that the apparent specialisation of the system for face recognition is not innate as generally assumed. Rather, face recognition and discrimination may be a skill substantially acquired through experience, in the same manner and using the same neural systems as the recognition of objects by experts.

Further, it is well-established that individuals with social cognitive disorders such as autism are deficient at recognising and discriminating faces. The Vanderbilt study suggests that the poor face recognition may not reflect a congenital disorder, but may rather be due to the reduced experience of looking at others’ faces resulting from avoidance of social interaction.

This and other studies by Gauthier and colleagues at Vanderbilt University have raised the stakes in the continuing debate over whether the FFA is a genetically determined, hard-wired mechanism for face recognition due to the evolutionary importance of this skill, or whether it is a system more generally serving the rapid recognition of objects with which we are very familiar.