Consciousness is not sempiternal, it has a history, a natural genesis. Once upon a time the universe contained no consciousness; then it sprang up here and there; and now the planet is flooded with the stuff. This is not to make the trivial observation that what people think and feel changes over time and generations, sometimes quite radically; it is a point about the deep biological roots of consciousness. Just as animal bodies are products of a long evolutionary process, in which chance variation is rigorously winnowed by natural selection, so animal minds must have a remote genesis in the mechanisms of differential survival as they worked on the available materials. Eyes gradually emerged as engines for exploiting the information contained in light, relying on the given chemical and optical properties of matter; and consciousness likewise must have emerged for some good biological reason, building on the prior properties of organisms. The question is how and why this happened: how did mentality arise from cell tissue? Answering this question would tell us not merely about the aetiology of consciousness: it would also help us to understand the nature of consciousness – particularly its relation to its physical substrate. If we knew the history of mind, then we would have effectively solved the mind-body problem, since we would understand how consciousness arises from matter.
Nicholas Humphrey’s book is a bold and speculative attempt to reconstruct mental history and hence to develop a theory of consciousness. He has a good project, and he is bracingly undaunted by its difficulty. He has a number of interesting and sensible things to say about a variety of topics, from the affective dimensions of colour to the nature of blind-sight. And he writes in a fresh (if jaunty) style. But in the end, I fear, the theory he proposes is a dismal failure: it doesn’t work at all. Don’t blame the author, though; blame the problem – it is just so hard. Like most attempted theories of consciousness, Humphrey’s looks like a contender only by trading on a mixture of obscurity and circularity. What is instructive about it are the manifest contortions needed to offer something with even the appearance of a decent theory. You see Humphrey being driven from pillar to post, alternating confidence with aporia, in a doomed attempt to lasso his quarry. Consciousness still swims out of reach, flaunting its mysterious gleam.
The book starts encouragingly enough, by locating the problem in the nature of basic first-order sentience. Humphrey tells us, dis-armingly, that in his earlier work he ‘came in at too high a level and left the fundamental problems unsolved’. ‘Too high a level’ was the level of self-reflection – knowledge of one’s states of consciousness. This leaves quite untouched the prior question of the nature of the mental states themselves – the pains, the tickles, the seeings of red, the smellings of roses. How do these spring from mere irritations of nervous tissue? Humphrey now sees the problem, correctly, as that of ‘explaining how states of consciousness arise in human [sic] brains’: how do we get from brain cells to subjective sentient fields?
Humphrey’s theory has two main parts: 1. a distinction between sensation and perception, with consciousness attaching directly only to the former; and 2. the suggestion that to have a sensation is for the brain to initiate a feedback loop from its core to its periphery. Both parts of the theory strike me as fundamentally flawed and crucially unclear.
Many theorists of perception have felt the need to distinguish between a component of sensory experience that acts as a sensation in or for the subject and a component that corresponds to the way the objective world appears to the subject. When we smell a rose, we have a sensation in our nose, it seems, as well as perceiving something to be so in the environment. Awareness of our own body thus seems somehow implicated in awareness of the external world: we perceive the world from a specific body and our experience seems to reflect this fact. Something happens in us when the outer world presents itself to our senses. The difficulty has been to formulate this intuition, or set of intuitions, in a way that does not misdescribe the character of sensory experience. When we see a scene, which aspects of the experience constitute the sensation we feel and which depict the outside world? What is strictly inside us, experientially, and what points outward? This kind of question has occupied philosophers for centuries and the pitfalls have been diagnosed. In particular, the need to preserve the essential intentionality of sensory experience has long been recognised.
Unfortunately, Humphrey shows little awareness of, or sophistication in, the conceptual and other issues that arise at this point, and often writes in ways that indicate a good deal of error and confusion. He fails to give any account of the representational character of sensory experience. It is totally unclear, in particular, whether he takes his category of sensation to be intrinsically non-representational, or whether it is simply pre-judgmental. Is he making a division within how things look to the subject, as he sometimes seems to be, or is he trying to identify a level of experience that has no world-directed intentionality written into it? Humphrey says repeatedly that sensations ‘represent what is happening to me’, at my bodily surface. It is not clear how he is using ‘represent’ here (itself a crucial issue), but it is surely false to suggest that in typical visual experience the way in which my retina is being physically stimulated is part of how things seem to me – I have no experience of my retina.
Nor is it acceptable to predicate colour words of my sensations themselves, supposing that when I see something green there is literally something green inside me – a ‘green sensation’. This is the old mistake of transferring to the experience what properly belongs to its intentional object. In fact, Humphrey unwittingly (I assume) treads the old path of the sense-datum theorists, with its debates about the status of secondary qualities. Maybe such theories are more defensible than they have seemed to recent philosophers of perception, but Humphrey is too naive about the philosophical issues involved to persuade us of that. The central question he needed to work harder on is this: is the sensation/perception distinction, as he wishes to draw it, a distinction within the way the world appears to the subject? That visual experiences have affective corollaries does not show, pace Humphrey, that there is a sensational component to them which can be hived off from the way they represent the world, since this might be extrinsic to their content.
Equally problematic is Humphrey’s conception of what the perception side of his contrast is supposed to be. Sometimes it seems to consist in the judgments the perceiver makes on the basis of his experience, in which is not a component of experience at all, since it goes beyond how things appear. At other times it seems intended to capture anything about experience that represents the world outside the subject – say, its looking to one as if there is a red round thing there – in which case its distinctness from sensation becomes problematic. Matters are not helped by saying, as Humphrey does, that perception is not modality-specific, which is analytically false, and by an alarming tendency to conflate percepts with the objective facts they represent. Nor is it clear what he could mean when he argues that perception, as distinct from sensation, does not involve consciousness, since surely how the world appears to me enters into the determination of my state of consciousness.
The trouble is that Humphrey is playing with a number of distinctions and failing to pin down precisely which one he has in mind. Talk of two ‘parallel channels’ is either unhelpfully metaphorical or downright mistaken if taken to imply that there is a dual representation (in the proper sense of the words) in every sensory experience. I don’t, after all, see my retina every time I see something that isn’t my retina.
The point of this distinction, for Humphrey, is to pave the way for his theory of consciousness: it is offered as a theory of the sensation component. I find this theory bizarre, unmotivated and inadequate. The idea appears to be this: to have a sensation, say a pain or a visual experience, is for the brain to send a signal to the periphery of the body, so creating a physical disturbance there, and for this disturbance to be registered, via a feedback loop, in the initiating segment of the brain. The theory is that a sensation resembles, and descends from, such actions as the wiggle of an amoeba in response to an impinging stimulus: in other words, it’s an activity that originates centrally and has effects at the body’s surface. A visual sensation of red is thus (is nothing more than) the action of causing the retina to fire in response to incident light – as it were, wiggling the retina. Consciousness reduces, then, to the neural causation of peripheral bodily disturbances.
To be sure, Humphrey is compelled to modify the original (intuitive!) statement of his theory to handle the fact that we can have sensations without anything occurring in the body, as with phantom limbs: his amended claim is that the body finds a surrogate in a ‘cortical map’, so that sensations become instructions to cause physical disturbances at the surface of the cortex – ‘cerebral sentiments’, he calls them. Feeling something consists in making your cortex wiggle. But this does nothing to make the theory any more palatable – quite the contrary. To repeat, a conscious sensation just is the physical action of the brain as it causes changes in outlying portions of itself or in the body if there is one, where these changes are themselves kept track of by means of a feedback loop. The immediate presence of sensations to the conscious subject is held to consist in such ‘loopiness’. Since brains can do these physical things, and since that is all a sensation is, we have an explanation for how brains generate sensations. Success!
This is a very disappointing solution to the original problem. First, the theory is really just a variant of familiar physicalist theories. Like behaviourism, it sees the essence of a mental state in the disposition to cause bodily changes, though in the sensory receptors, not the motor system. Like central-state materialism, it ends up identifying mental states with neural events in the brain. Like functionalism, it stresses the ‘software’ descriptions of the underlying brain processes, thus allowing for different physical realisations of the reverberating feedback loops which constitute sensations. And it faces all the standard problems which beset these doctrines, without making any real advance on them.
Second, what about bodily changes initiated internally and subject to feedback that manifestly doesn’t involve any consciousness – healing of the skin, muscle growth, digestion, blushing? Since these involve essentially the same physical processes yet don’t generate consciousness, the feedback loop theory cannot provide sufficient conditions lot our being in a conscious state. Indeed, it is hard to see why, according to Humphrey, a thermo-statically-controlled heating system doesn’t have conscious sensations, since it meets his conditions for consciousness. Physical feedback loops come too cheaply to add up to mentality.
Third, the distinctive sense in which sensations are owned by the subject can hardly be captured by this theory, since all bodily states are similarly ‘owned’. No special link with the introspecting self has been established.
Fourth, it is quite implausible to maintain that the phenomenal type of a sensation is explicable in terms of the bodily characteristics of the site of peripheral disturbance. That would imply that the type seeing red is constituted by the fact that my retinal rods and cones are firing in a certain way – as if that were how things seem to me when I see something red! The physical properties of my receptors, e.g. the inside of my nose, are not what individuate the phenomenal type of my sensations, e.g. the smell of a rose. This kind of physicalist reductionism is no more plausible when the physical facts obtain at my surface than it is when they occur further in. And calling the bodily disturbances the ‘adverbial style’ of the cerebral action does nothing to make the theory more attractive. We are still being told that the feeling of pain is just one kind of physical wiggling among others.
In fact, after a lot of preliminary stage-setting Humphrey spends a breathtakingly small amount of time explaining how his theory is meant to capture the characteristic properties of sensations, and his explanation is obscure and unpersuasive. What he seems to be offering, at bottom, is a kind of peripheralist identity theory: a sensation is identical with a bodily perturbation of a certain sort. I see no good reason for this variant on familiar central identity theories, save for a kind of half-suppressed behaviourist urge somehow to get the wiggle into the act.
Towards the end of the book, Humphrey quotes me on the difficulty of the mind-body problem and the inadequacy of our current modes of thought, and he issues this challenge: ‘If McGinn still wants to deny that it’ – Humphrey’s theory – ‘is the wine of consciousness, let hirn taste it and say what is missing.’ Well, I found the taste elusive at first, though finally it revealed itself as the usual old plonk. What was missing? Oh, not much – just the presence of any real grapes. Seriously: despite some interesting incidental reflections, and an admirable breadth of reference, Humphrey leaves the mind-body problem exactly where it was. His excursus into speculative mental history has turned up nothing to alter the basic geography of the issue.
A puzzling question I would like to have seen discussed is why consciousness is so prevalent in the biological world; beyond the simplest organisms all animals seem to have some. This could either be because it has great and unique biological utility or because it is written deep into the nature of matter and can’t, so to speak, help emerging when particles coagulate in certain ways. The first alternative is hard to reconcile with the fact that it seems quite possible to imagine even complex organisms reproducing efficiently without then behaviour being guided by sentience – so why aren’t there any (complex) robot species on the planet? But if the second alternative is the case, then there is something amiss, after all, with trying to understand consciousness biologically, as if its emergence must have a direct biological rationale. It might, on the contrary, odd as this may sound, be simply a by-product of traits that do have such a rationale. One of the great puzzles of evolution is why sentience seems to be the preferred method for handling adaptivity to the environment. Why not process information without any inner feeling at all? Why, that is, does conseriousness exist?
It is common to hear theorists insist that consciousness must be viewed as a natural phenomenon with a natural history, subject to the rules that govern other evolved characteristics. Fine. But the same biological perspective should encourage a more sceptical thought: namely, that the human power to understand the world is itself a natural biological phenomenon, subject to the usual constraints and limitations, There is no empirical or a priori reason to suppose that our capacity to understand nature extends to all the things that puzzle us; it would be amazing if it did. Consciousness may be one of the subjects that our biology has not equipped us to understand. This should be regarded at least as a live possibility by anyone who takes the biology of mind seriously. For all his vaunted naturalism, Nicholas Humphrey is, like so many others, unwilling to take his naturalism the whole way. The reason he can’t produce a good theory may be that his brain won’t let him.
Send Letters To:
London Review of Books,
28 Little Russell Street
London, WC1A 2HN
Please include name, address, and a telephone number.