Towering inference

There are two types of philosopher of mind: those who eschew the messy, real-life brain in favour of old beer cans as an equally likely vehicle for consciousness, and those who strive to place consciousness in a biological context. Daniel Dennett adopts the second approach by looking to animals for clues to the frustrating and compelling questions about the nature of our mental activities. In itself this is hardly novel: over 20 years ago Thomas Nagel started the ball rolling with the now-famous question: What is it like to be a bat? However, Dennett derides this question as a nonstarter: we humans are not bats, therefore we will never experience at first hand the bat perspective on life.

Dennett's launch pad is, then, the idea that different species have different types of consciousness which cannot be subjectively experienced at first hand, yet which he can describe objectively in the appropriate nested setting of the physical body, environment and lifestyle of the animals in question. So far so good. But this side-stepping of first-hand experience might still be a trip wire for the unwary, despite the pace and eloquence of Dennett's narrative. How does his third-person type exploration of animal mentality relate to that elusive, subjective feel of the first-person, personal world? If we join with Dennett and admit that we are mere voyeurs not inside the skin of our simian or feline cousins, have we not thrown the baby of subjective consciousness out with the bath water of unwarranted anthropomorphisation?

The nagging contrast between what we ourselves feel and what we can study in others, could perhaps have been cleared up early on, had Dennett fully considered the dichotomy in the subtitle of his final chapter, "Our consciousness, their minds". When we reserve "consciousness" for the first-person, inside-the-skin sensation, and "mind" for the objective and observable factors that constrain and determine the quality of that first-hand sensation, then the journey immediately becomes more clearly charted. Dennett's explorations can then be viewed as a survey of "mind", a collective of different, objective factors that influence the type of consciousness one might have, but do not explain our actual experience.

In focusing on these objective factors, the "brain design" of different animals, Dennett explores not so much the differences in interior organisation of the brain, but rather the resultant behaviours. He reworks the ancient nature-nurture opposition into a question of degree. In brief, the more primitive the brain, the more predictable the action according to the dictates of prewired instinct (rather confusingly referred to as "free-floating"). The idea is that a more sophisticated brain will liberate its owner from single-minded genetic tyranny and allow us to develop an individual, ontogenetic agenda as we interact with the environment. This gradual shift from phylogeny to ontogeny can be marked by the flexibility of the behavioural repertoire available.

Dennett talks of a "tower" of increasingly impressive brain design. At the "lowest" level are "Darwinian" creatures with no choice in their course of action since they are entirely preprogrammed to act in one way or another: their survival depends on the efficiency of a genetically determined strategy, not on personal experience. At the next "level" up, the "Skinnerian" brain can benefit from personal experience: it is, as its sobriquet suggests, capable of being moulded by trial and error. As a further improvement, in the "Popperian" brain some potential strategies can be analysed in advance and dismissed before even being tried. Finally comes the "Gregorian" stage (not a monk-like, meditative stance but a way of thinking named after Richard Gregory), in which we sophisticated brain owners can see a meaning to objects that is not immediately apparent from their physical properties.

Needless to say, for brains that are able to generalise from garden shears to include nail scissors, the opportunities are ever greater. Such powers of abstraction are enormously aided by, if not dependent on, language, which enables humans to soar above the rest of the animal kingdom in problem-solving and hypothesis formation.

But note that power within the tower is measured by problem-solving agility, not in feelings. Surely feelings, not only pain, but the full impact of the senses, can be dissociated from prowess at problem solving or from appreciation of the "significance" of external objects. The young child, the skier or the dancer caught up, however transiently, in a literal and vivid present is not accessing a host of reflective and associative processes that would merit a place high in the tower. "Thinking" is not a prerequisite for consciousness to be at full throttle.

None the less, every so often Dennett does make forays into the question of sentience, more specifically the capacity to feel pain. He introduces the helpful and interesting idea that sentience is not a matter of all or nothing, but a "ramp", a continuum spanning different species in varying degree. But sadly, the implications of this scenario for our understanding of first-person consciousness in general, remain undeveloped. On the other hand, Dennett uses the issue of pain to show how consciousness cannot be divorced from the physical brain and body of which it is part. For example, you cannot download a headache as computer aficionados might download a problem-solving strategy. First-person consciousness is locked into the biological body, but cognitive processes are not.

Dennett's scheme of towered brains is based on the experience-based problem-solving beloved by those working in artificial intelligence, not on depth or subtlety of feeling. This means that his strategies for characterising the "mind" - but not the feel of first-person consciousness - might be tractably exported into synthetic devices. And yet the reader may well feel confused: how can it be possible to explore the "mind" objectively, while retaining the exclusively biological infrastructure that Dennett advocates? While his emphasis on the integral nature of consciousness, with biological systems living interactively in specialised environments, is a far cry from the world of the computer modeller and the panpsychic - his tower of escalating strategies, hypotheses and associations is assailable by functionalists building beer-can brains.

There is far more to the mind than problem-solving agility, not least its chemical basis and its manipulation by mood-modifying drugs. Dennett makes no use of this obliging bridge between first-hand sentience and the objective, physical factors which control it. Perhaps such neurochemical excursions do not fit in with his consideration of animal minds. But at the very least they must set the reader musing over the validity of a distinction between sentience, the slippery "feel" of consciousness itself and the physical infrastructure that will constrain that experience. How might the ramp of consciousness be built into the tower of mind? Though we are offered no clear, overarching design, Kinds of Minds is a thought-provoking, well-written read. After all, philosophers, by Dennett's own admission, traffic in questions, not answers.

Susan Greenfield is professor of pharmacology, University of Oxford.