Blinded by blindsight?

In 1917, George Riddoch, a temporary medical officer in the British army, wrote a remarkable article. He described how soldiers who had been blinded by gunshot wounds that had destroyed their visual receptive cortex could still see motion in their blind fields, though not much else. Crucially, he wrote repeatedly that they were conscious of having seen the motion. This was an embarrassment for scientists at the time, as it was thought that the visual receptive cortex, which we now call V1, was the brain's only visual area. How to deal with it? Dismiss it and keep it out of mind.

This is in effect what happened until the 1970s, when the phenomenon was rediscovered. By then, it had become clear that V1 is only the first visual area in the brain that feeds visual signals on a selective basis to many other visual areas, among them area V5, which is specialised for the perception of visual motion. But unlike in 1917, it was argued that the affected person could discriminate the direction of motion correctly, but at the same time was not conscious of having seen anything at all. This seemingly bizarre phenomenon - in which a cortically blind subject can see without being aware of having seen - was given the name blindsight; and it excited the interest of many, not least philosophers and others interested in the problem of consciousness. It also led to the supposition that visual consciousness is not possible without activity in V1.

Studies of blindsight have been largely based on a single patient, GY, who is, this book tells us without a hint of irony, "a very experienced subject". GY became hemi-blind following damage to V1 of one hemisphere in a car accident at the age of seven. Over the past 20 years, he has been studied on an almost industrial scale, indeed so much of the material here is based on experiments with him that the book could have been dedicated to him.

This is despite the fact that in 1993, study of GY dealt the field of blindsight a grievous blow when it was shown that he was capable of conscious vision even in the absence of V1. Parallel brain-imaging experiments demonstrated that his conscious awareness of visual motion correlated with activity in visual area V5 (specialised for visual motion) by receiving a direct input from the retina, bypassing area V1.

Post-1993, in what seems like a damage-limitation exercise, the proponents of blindsight have admitted that conscious vision is possible without V1 and have divided it into four categories, only one of which appears to correspond with the original definition of blindsight. Almost imperceptibly, the argument has shifted from the question of whether GY is conscious of visual stimuli in his "blind" field to whether this consciousness is "visual" consciousness.

This year there was another grievous blow, in the form of an article by Petra Stoerig and Eberhardt Barth showing that GY's consciousness is indeed visual consciousness. Why, it may be wondered, should such a demonstration in a single patient compromise the whole theory? The answer is simple: the theory of blindsight derives much of its force from the study of this one patient.

The uninitiated reader of this book, lost in the praises lavished on the phenomenon, will not be able to glean just how controversial blindsight is, and how many doubt its existence. Those impressed by the earlier claims for blindsight will be perplexed to learn that GY was usually aware of the stimulus in the experiments described by Alan Cowey and his colleagues; and even more perplexed to find that, in some of the other articles, the original definition of blindsight (that there is no awareness for visual stimuli that are correctly discriminated) is still being used. This is true, for instance, of the highly interesting discussion of the unconscious processing of fearful stimuli by John Morris and Ray Dolan, who write of blindsight patients as "lacking conscious awareness of visual stimuli in their blind field".

Elsewhere, in an otherwise excellent review of neglect following parietal lobe lesions, Jon Driver and Patrik Vuilleumier are at pains to draw a distinction between that syndrome and blindsight - but they too use the term as originally defined, and write that "it may come as no surprise that destruction of [V1] can remove conscious vision".

This is unfortunate because it distracts from important issues. Why is it that with parietal lesions, awareness is impossible even with an intact V1, whereas with blindsight, awareness is possible even with a damaged V1? A more rewarding and insightful discussion has thus been stifled. The simpler but perhaps more interesting questions are lost in erudite discussions of signal-detection theory, second-order motion and whether the conscious visual awareness that GY has is of visual motion or of the displacement of a visual stimulus from A to B.

However, the much-battered phenomenon of blindsight has served a useful purpose in focusing attention on the fact that a great deal of processing occurs in the brain before matters are brought to conscious awareness. This, of course, is not new. Leibniz wrote in his Monadalogie that "PerceptionI is to be distinguished from Apperception or Consciousness... In this matter the Cartesian view is extremely defective, for it treats as non-existent those perceptions of which we are not aware." But the conditions under which a conscious correlate occurs have only recently become the subject of intensive study, thanks largely to human brain-imaging techniques. Hence the more interesting part of this book is to be found in the articles about conditions in which brain activity is not correlated with conscious awareness, in the chapters on emotion, memory and attention. It is particularly interesting to note what is now a common theme, that exposure to a specific stimulus can lead to activity in specific cortical areas even when subjects are not aware of the stimulus; a point made especially clear in the article by Driver and Vuilleumier. Hence a more fruitful discussion, however speculative, might perhaps have focused on why activity at a given locus can have a conscious correlate under one set of conditions but not under another.

It is a pity that a phenomenon that has attracted so much attention should not have been more critically discussed in a multi-contributor book. This would have meant soliciting contributions from doubters and sceptics, as well as from aficionados, but the result would have been rewarding. Such an approach is not the hallmark of this book, which banishes critical examination of blindsight. But in doing so, it does not infringe the Trades Description Act - for the book is titled, aptly, Out of Mind .

Semir Zeki is professor of neurobiology, University College London.