There's a war going on in the face processing literature, fought not with bullets but with case studies and journal publications. The bone of contention is whether there's something unique about our face processing ability, or if face processing is just like any other form of expertise.
Supporting the expertise account are brain imaging findings showing, for example, that the so-called fusiform face processing region of the brain (seen by some as a dedicated face processing module) is also activated by other forms of expertise, such as when bird watchers view pictures of birds.
On the other hand, there's ample evidence that face processing really is unique. For example, brain damaged patients have been described who have lost their ability to process objects for which they previously had expertise (e.g. toy soldiers), but whose face processing abilities have remained intact.
A key issue in this debate revolves around the so-called face-inversion effect. This is the finding that our ability to process faces, but not other objects, is seriously impaired when they are turned upside down. One explanation is that our specialised ability for processing faces is based on a global or "configural" approach, which breaks down when a face is inverted. Infants as young as four to five months - too young to have built up much face "expertise" - show signs of the face-inversion effect, providing further evidence that face processing has an innate component and is different from other forms of acquired expertise.
Now Laura Schmalzl and colleagues have joined the battle by describing an unusual case of a teenager, born with brain abnormalities, who showed superior ability at processing inverted faces compared with upright faces. In other words, he showed an inverted face-inversion effect!
The boy, known as J.M., was born with hydrocephalus (too much cerebrospinal fluid in the brain) and epilepsy, and among his impairments he's always had a deficit processing faces - a condition known as prosopagnosia.
Schmalzl's team had J.M. complete a wide-ranging battery of neuropsychological tests and a key pattern to emerge was his bias towards local processing at the expense of global processing. For example, he could identify line drawings of objects when local detail was visible, but he failed when he was forced to rely on silhouettes.
When it came to faces, J.M. was generally impaired as the researchers expected. For example, he struggled to recognise photos of people he knew. However, with specific facial features like noses and mouths, some strange patterns of performance emerged. Consistent with his local-processing bias, J.M. could distinguish between various noses and mouths when they were shown in isolation. He could also do this when they were embedded in inverted faces, yet crucially, and somewhat bizarrely, he couldn't distinguish between facial features in upright faces.
This inverted face-inversion effect is the opposite of what you'd expect in a healthy person. The researchers aren't entirely sure what's going on but they think the case may be a rare example of a faulty face-processing module being activated automatically by the sight of an upright face, thus hindering J.M.'s ability to process the local facial features. This is very much an argument from the camp that sees face-processing as a unique, innate ability. They're saying it's possible that J.M.'s innate face processing module still has some uses - for example, it may aid the perception of eye gaze or speech perception - but that in other situations it's dysfunction actually impairs performance.
"In future studies it would be of great interest to formally test cases such as J.M. from early infancy, and longitudinally, in order to shed further light on how innate versus experience-based factors influence the development course of face-processing abilities," Schmalzl's team concluded.
Schmalzl L, Palermo R, Harris IM, & Coltheart M (2009). Face inversion superiority in a case of prosopagnosia following congenital brain abnormalities: what can it tell us about the specificity and origin of face-processing mechanisms? Cognitive neuropsychology, 26 (3), 286-306 PMID: 19657795