Inner-eye exploration

The scientific study of consciousness is now almost a decade old. Colin Clifford hails the work of one of its pioneers.

The discussion of consciousness was, until recently, strictly the province of philosophers. Psychologists and neuroscientists would happily talk in terms of mental representations and neural information-processing, but the mention of consciousness was frowned on - it was almost taboo. While it was scientifically respectable to say that one conducted experiments on the workings of the brain or the mechanisms of the mind, to say one worked on consciousness was to risk derision and dismissal as a crank.

The situation has now changed, and the study of consciousness has become a multidisciplinary enterprise. The philosophers still have their role, but they have been joined by an increasing number of psychologists and neuroscientists using experimental means to advance our understanding of what has been described as one of science's last great unsolved mysteries: the material basis of consciousness.

More and more books on consciousness are appearing, as though everybody from philosophers and psychologists to psychiatrists, and even physicists, thinks they have something to offer on the subject. But The Quest for Consciousness: A Neurobiological Approach stands out from other books for two main reasons. The first is its author, Christof Koch, who in collaboration with the late Francis Crick (who provides a foreword) has been in the vanguard in making the scientific study of consciousness a respectable pursuit in academic circles. The second is the book's empirical grounding, whereby philosophical speculation takes a back seat to neurobiological evidence.

Koch writes with real passion tempered by a disciplined scientific mind, so that the reader is easily able to distinguish between opinions based on solid empirical foundations and the necessarily speculative leaps that must be made to go beyond the data and visualise a bigger picture. He has a very personal style, often slipping in references to his love of rock climbing and giving the strong impression of being someone who really lives his science rather than merely has a career as a scientist. Koch describes his approach to the study of consciousness as "a pragmatic, empirical one".

Unlike other authors, he does not pretend to explain consciousness.

Instead, he builds a framework within which to understand and study it.

This is a valuable exercise. It gives researchers an empirically grounded conceptual basis from which to develop and refine knowledge in the area. No doubt there will be aspects of the framework that will in time have to be dismantled, but as Koch himself acknowledges: "A good framework is one that is reasonably plausible - relative to the available scientific data - and that turns out to be largely correct. It is unlikely to be correct in all details."

Koch argues persuasively that a vital step towards understanding consciousness is to identify the neural correlates of consciousness (NCC).

He defines the NCC as the minimal set of neural events sufficient for a specific conscious percept. Establishing where in the brain the NCC lie and what form they take is the essence of the mind-body problem.

One aspect of the book that might at first appear surprising is that it focuses almost entirely on the visual system. Koch's principal justification for this is typically pragmatic when he writes that "humans are visual creatures". Vision is also one of the attributes of consciousness most readily amenable to scientific investigation. Modern computer graphics technology offers tight control of the visual images generated and, in turn, the subjective visual experience produced. In addition, much is known about the workings of the visual systems of humans and other species throughout the animal kingdom. So, while the book is definitely biased in its content towards the quest for visual consciousness, that bias reflects the way consciousness is currently being scientifically explored as much as any bias the author himself may have.

As a consequence of the focus on vision, several of the earlier chapters in the book are in essence tutorials on the neurobiology of the primate visual system. These are necessary to provide the reader with the background to evaluate much of the scientific evidence presented in later chapters.

Interspersed with this more didactic material are lively chapters on the NCC and, in particular, which areas of the brain do not constitute NCC.

Koch reviews evidence from a range of neurobiological techniques that leads him to the controversial view that the NCC are not to be found in the primary visual cortex (V1). Much of the controversy surrounds the observation that damage to V1 can lead to the phenomenon of "blindsight", whereby the affected patients report no visual experience in at least part of their visual field, yet paradoxically are able to indicate, for example, the position of the stimulus within their "blind" region. The capacity of V1 lesions to produce blindsight has been taken as evidence that activity in V1 is necessary for visual awareness. However, as Koch points out, while V1 is clearly necessary for normal vision, the existence of blindsight does not mean that V1 must be an NCC.

To understand the distinction between being necessary for normal vision and being an NCC, consider the case of the retina. At least one functioning retina is necessary for normal vision - without the retina no signals would reach the cortex. Yet the retina is not an NCC. This is elegantly demonstrated in the phenomenon of binocular rivalry. To generate this, two dissimilar images are presented, one to each eye, so as to occupy a single region of the visual field. What the observer typically experiences is "rivalry" between these two images, such that one is perceived and the other suppressed. Over time, perception flips between the two images, even though the retinal stimuli themselves remain constant. Activity in each retina is determined by the image presented to that eye and is unaffected by flips in perceptual dominance between the two images.

Because retinal activity does not correlate with the conscious percept, the retina cannot be considered an NCC. Electrophysiological evidence suggests that the activity of the majority of neurons in V1 also fails to correlate with perception, although this correlation increases as we proceed up the visual processing hierarchy beyond V1 and into the inferior temporal cortex. By the time visual signals reach the inferior temporal cortex, about 90 per cent of neurons show modulations in their activity correlated with the dominant percept. Koch argues that the NCC consist of regions of the brain such as the inferior temporal cortex that project directly to the executive and planning centres of the brain in the prefrontal cortex and do not include early sensory areas such as V1.

A fundamental question about consciousness is that of function - what is consciousness for? Put another way, what can our conscious brains do that could not be achieved equally well by a set of unconscious neural mechanisms? Questions of this type take us into the territory of the philosopher's "zombie". The zombie is a hypothetical being, the product of a thought experiment, able to perform the same range of mental operations as a normal person but lacking any conscious experience. For example, a zombie would be able to recognise a lemon by its yellow colour but it would have no idea of what it feels like to experience the yellowness of yellow, the so-called qualia that comprise the subjective content of consciousness.

What advantage do qualia give us over zombies? One possibility is that the theoretical zombie could never exist, that any system of neural mechanisms as complex as the human brain would inevitably possess consciousness. In this case, we need not look for a function of consciousness - consciousness would simply be what it feels like to be a complex neural system. Alternatively, perhaps zombies could exist but we have evolved to be conscious because qualia themselves offer some functional benefit.

It is the latter position that Koch favours. He likens the content of qualia to executive summaries of sensory information made available to the planning stages of our brains. However, as Koch acknowledges, there is nothing in his account of qualia to explain why it should feel like anything to be conscious; he simply asserts that "qualia are too structured to be an irrelevant by-product of the brain". Understanding the subjective content of consciousness is one area where, as yet, the neurobiological approach has little to offer. While it remains an open question whether the secrets of qualia will ever be unlocked by neurobiology, Koch is certainly keen to entertain the possibility.

This book is well suited to an audience of postgraduate researchers in the fields of psychology, neuroscience and philosophy. I also plan to recommend certain chapters as "further reading" to accompany my final-year undergraduate lectures in perception. I have been going through the book two chapters at a time as part of an interdisciplinary group of psychologists and robotics engineers. It has inspired us to many lively discussions on the nature of human consciousness as well as the best ways to design and build intelligent robots.

Although I find myself disagreeing with some of Koch's conclusions, this is by no means a criticism. They are eloquently justified with full reference to the relevant literature. This enables the critical reader to pinpoint where the logic of an argument or the interpretation of data is unconvincing and to formulate an alternative theory or to design a decisive experiment.

Overall, this is the most accessible and the most insightful book on consciousness among the many that I have read in recent years. I recommend it to anybody who wants to know more about the state of knowledge in the scientific study of consciousness and is willing to work through the introductory chapters on neurobiology.

Colin W. G. Clifford is a fellow in the School of Psychology, University of Sydney, Australia.