Conscious of the need to talk

All serious neurobiologists agree that consciousness is inextricably linked to the physical substance of the brain and to physiological processes going on inside it. It should therefore be subject to the laws of physics and be accessible to study by scientific means and methods. Neural Correlates of Consciousness , edited by Thomas Metzinger, contains a collection of essays by leading neuroscientists that sum up current views on "what" and "where" in the brain consciousness is.

However, the book starts off with a section that deals with philosophical issues. First, how can the subjective, personal experience of consciousness be investigated with objective experimental methods? A minimum requirement is the reproducibility of the so-called phenomenal content of a certain state of consciousness. Metzinger suggests that most philosophers would indeed agree that any conscious experience is fully determined by a complete physical description of the internal state of the brain at any given moment in time. What is not clear is the nature of the relationship between physical state and subjective experience. Is it a mutual causal interaction, as dualists would argue? Or is consciousness an emergent property of a complex brain, an epiphenomenon without causal effect? Or are, indeed, conscious experiences and brain states two sides of the same coin, two descriptions of the same reality? This leads on to the second question: what, as a biologist, does one actually look for, what is a "neural correlate of consciousness" (NCC)? David Chalmers addresses this question in his contribution, starting from the definition that: "A neural system N is an NCC if the state of N correlates directly with states of consciousness." In order to avoid a trivial answer such as "the brain is an NCC", Chalmers refines his definition of NCC to mean a minimal neural system sufficient for a corresponding conscious state. As Antti Revonsuo points out, the premise for biological research into consciousness has to be that consciousness literally is in the brain. This implies that certain brain states or neural activity patterns are isomorphic with phenomenal experience or, in other words, are neural correlates of consciousness. An open issue is whether neuroscientists today possess the tools to reveal those patterns.

The second section deals with the big question of what it might be that we should look for as NCC. Christof Koch and Francis Crick, at the forefront of consciousness research, revisit the concept of a homunculus inside the brain and make an interesting proposal about how to avoid the infinite regression implicit to it. In contrast, Antonio Damasio rejects the idea of a homunculus outright and "assembles" consciousness and the self in two steps. First-order maps or representations of objects are made in sensory brain areas and reactions are generated in core structures, giving rise to an unconscious proto-self. The autobiographical self, which gives us our sense of identity and makes us the conscious owners of our thoughts, in Damasio's view, is dependent on second-order maps that produce mental images from first-order inputs. These second-order maps need not be strictly localised in specific brain structures.

Wolf Singer tackles the central question of how representations are formed at the neuronal level. The most intuitive answer is that the activity of individual neurons is equivalent to certain subjective experiences. One would have liked to see, in this section, a modern evaluation of this neuron doctrine, which has retained a lot of explanatory power to this day, a good example being the perception of visual motion. However, the brain does not contain enough "grandmother neurons" to encode a near-infinite richness of sensations. Singer proposes that representations are formed through the activity of small groups of neurons, and that these fluid assemblies are tagged by means of synchronous activity of the cells that constitute them. This attractive model, which is supported by considerable experimental evidence, particularly from the visual system, combines very high combinatorial power with fast processing speed.

The third and largest section of the book deals entirely with neural correlates of consciousness in the visual system, one of the best-understood subdivisions of the brain. Two chapters give examples of processing of visual stimuli or stimulus attributes that the observer is not aware of. Such accounts are often used to support a modular view of consciousness, and even to confine consciousness to certain parts of the brain to the exclusion of others. If we are not aware of the activity in some area of the brain in response to some stimulus, then consciousness cannot have access to that area. As Dominic ffytche points out in his essay, there is a long tradition, going back to Gall's 19th-century phrenology, of the localisation of functions in various brain areas. While this has proved an extremely useful concept in the study of sensory perception, it seems to me somewhat naive with respect to consciousness studies. The endeavour to localise brain functions appears to have proved particularly counterproductive in recent imaging studies of perceptual awareness. In monkeys, very often the same cortical areas are active when responding to a sensory stimulus irrespective of whether the animal is awake or anaesthetised. Ffytche suggests that "the difference between conscious and unconscious processing is not coded in the location of the activity but in the type of processing performed". Unwittingly, his hypothesis raises doubts over whether functional imaging in its current technological state can reveal neural correlates of consciousness at all. Methods such as Pet (positron emission topography) and FMRI (functional magnetic resonance imaging) measure brain activity only very indirectly, as an increase in metabolism, at spatial and temporal resolutions that are far too low compared with the size and electrical firing rate of neurons. They are therefore highly unlikely to reflect the phenomenal level of organisation that constitutes consciousness.

Interestingly, Erik Lumer uses FMRI data to support a conclusion in direct contrast to ffytche's: he holds a rather unitary view of consciousness as being the result of an integration of anatomically distinct circuits that are distributed throughout a large part of the brain. Lumer has studied binocular rivalry, which has become one of the most popular paradigms among neuroscientists who are interested in neural correlates of subjective perception. Binocular rivalry is observed when we view two very different patterns with the two eyes, say vertical stripes with the left and horizontal stripes with the right eye. In this case, fusion is impossible and one perceives the two gratings in continuous aperiodic alternation, but never a criss-cross. So while the retinal images and therefore the patterns of electrical activity in the two optic nerves are stable, the phenomenal content of visual awareness constantly changes. A good deal of research has been directed at establishing at which stage of visual processing brain activity corresponds to the visual percept rather than the physical stimulus. It seems a pity not to have given more space in this book to the variety of approaches and interpretations, as these could have served to illustrate nicely many of the concepts introduced in the earlier chapters.

The final part of the third section looks for NCC at a cellular level. Hans Flohr argues that anaesthesia, which eliminates consciousness, is ultimately achieved by blockade of a particular receptor that plays a crucial role in inter-neuronal communication, and that consciousness must therefore be instantiated by processes occurring at this receptor. This hypothesis is conceptually reminiscent of lesion studies that were widely used in early investigations of brain function but have fallen into disrepute more recently, not least because they presuppose a highly modular working of the brain. A certain part of the brain is removed; whichever function is then compromised will be attributed to that area. Flohr's hypothesis is critically reviewed by the other three contributors to this section.

The book concludes with an outlook that places consciousness and subjective experience of self in a social context, arguing that consciousness is the critical step that has taken mankind forward from biological to cultural evolution.

It is one of the strengths of Neural Correlates of Consciousness that it brings together philosophers and neuroscientists, and one can only hope that its readership includes both of these groups as well. It seems that interdisciplinary communication is rather scarce, owing both to mutual prejudices and to the lack of a common language. Most of the contributions in this book are highly readable and should give all scholars of consciousness studies food for thought. However, readers should not expect definitive answers as to what neural correlates of consciousness are, simply because few hard facts are known. Contradictory views are presented side by side, with Metzinger's section introductions providing a useful context. Despite some omissions, the important questions are raised in this book, and both conceptual and experimental solutions are proposed, some of which will certainly be controversial.

Frank Sengpiel is senior research fellow in neuroscience, Cardiff University.