Consciousness - philosophical deadends. Homunculi, dualism, epiphenomenalism, solipsism and panpsychism.
Strange ideas

In grappling with deep philosophical issues, we are sometimes pushed towards theories we should never have adopted for their own sake. The ideas on this page are classic cases of this kind - if you find your theory leads you into one of these positions, it's a sign that you need to look again at your theory...

Homunculi home


An homunculus is a 'little man'. It's evident that we don't literally have a small but fully-formed person in our heads controlling what we do, but some explanations of consciousness fail because they include a central observer which itself has all the mental properties a conscious person would have, in effect an homunculus. Explaining such a central person is obviously as difficult as explaining consciousness in the first place, and the hypothesis therefore does not get us very far. The normal argument, in fact, is that homuncular arguments actually fall into an infinite regress, with the consciousness of each homunculus explained by ever smaller homunculini, nested like Russian dolls: but in fact this only applies to the hard-line homuncularist position that homunculi are the only possible explanation of consciousness. One can still believe that there is, as a matter of fact, an homunculus who is responsible for our consciousness, but that his consciousness is explained on some other basis. Most would agree, after all that thereis a central entity inside us which does all the thinking - namely the brain.

Generally, however, homunculi are regarded as evidently absurd, and used mainly as a smear against other people's theories. Daniel Dennett is an unexpected exception here. No-one, in principle, is more hostile than Dennett to the idea of a central observer: Consciousness Explained repeatedly denounces the idea of the 'Cartesian Theatre', whose audience must surely be an homunculus (although the same book proposes a 'Joycean Narrator', and one might question whether the substitution of an Irish novel for a French play makes all that much difference). At the same time, Dennett has claimed that cognitive scientists frequently use homunculi in their tentative theories, and that their practice demonstrates that it is alright to do so. The apparent inconsistency is explained by the loose interpretation which Dennett gives to 'homunculus': he really means something more like a 'black box': part of a process which cannot yet be explained, but which we take for granted for the time being in order to get on with developing the rest of a theory in the meantime. Strictly, to be an homunculus, a component has to be fully conscious in its own right, or at least have a full set of conscious mental abilities, not just a particular narrowly-defined function.


Dualism home


Dualism  is the belief that there are two different worlds, or two fundamentally different kinds of stuff, in existence: matter and spirit, for example. In spite of being the received wisdom for hundreds of years, it is actually unsustainable inits strict form.  The problem arises when you consider how the two kinds of stuff, or the two worlds, interact.

If they interact directly, they are in effect parts of a single world after all, or different forms of essentially similar stuff. If spirits affect the behaviour of material objects, they can best be thought of as another aspect of the physical world, albeit one which to date remains relatively obscure. Matter and energy, after all, are  regarded as parts of the same world. It does not help to postulate intermediary stuff linking two worlds or two substances because the same problems recur for the relationship between the stuff in each world and the intermediary stuff. If, on the other hand, the two worlds don't interact, one of them is effectively null. We, the observers, must exist in one or other of the worlds (or be made of one or other kind of stuff): since the other stuff or the other world cannot affect us, we can never be aware of it and talking about it at all is a waste of time.  

For reasons of this kind, and because it has sometimes been a refuge for people who would prefer things to remain unexplained, dualism has become a kind of smear which philosophers sometimes hurl at each other's theories. Most people believe the world is complex to some degree, and most accept that different levels of explanation are required in order to say all that is worth saying about it, which means there is always some hook in anyone's theory on which an accusation of dualism can be hung. There are, in any case, many different varieties of dualism, of differing degrees of radicalism. A distinction is normally drawn, in particular, between full-blown substance dualism, and property dualism, which recognises only one underlying kind of stuff, but asserts that it has two unrelated sets of properties.

In recent years, Descartes has rather unjustly come in for particularly heavy criticism over his dualism, and it seems to have become almost mandatory to start any popular book about consciousness by criticising him for it (while borrowing one of the quaint illustrations from his books). He is often baselessly accused of originating the whole idea or of having invented, rather than perpetuated, the distinction between body and soul. Of course he was a dualist, but in that respect he was merely giving a slightly more secular expression to a belief which was pretty well universal for centuries before (and after) his time. It might be fairer to see him as restricting the role of the spirit; he believed that most actions of the human body were purely mechanical, rather than directly actuated by spiritual influence, so he deserves respect from materialists.

Dualism still has its appeal: the Platonic form which gives numbers and abstract entities a real existence in a world of their own seems to appeal to mathematicians, and physicists talk about other worlds, though in both cases it is often unclear how far this is a metaphor, or a dramatic way of talking about levels of explanation rather than a genuine ontological commitment. Two of the thinkers touched on in these pages - Roger Penrose and John Eccles - advocate theories of three worlds. In spite of the hostility of most (not all) philosophers towards dualism, moreover, discussions of alternative worlds remains a popular method of argument about possibility, identity, and indeed, consciousness. Dualism also remains a basic component of much religious and paranormal theory, and hence an influential part of the intellectual background.


Epiphenomenalism home



Epiphenomenalism is the view that one's actions are not caused by one's thoughts: that we are really just passive spectators under the illusion that we control our own behaviour. The idea appeals to those who want to accept that physics explains the causes of all events, including the behaviour of human beings, while still regarding consciousness as something over and above this merely mechanical process. The idea is lent some (much-needed) plausibility by Libet's famous experiments which seem to show that decisions are effectively made before we become aware of having made them: and by many others which demonstrate people's remarkable tendency to invent reasons, and accept responsibility for, behaviour which was actually caused by a post-hypnotic suggestion, a brain malfunction, or other factors they were not consciously aware of.
Strictly, of course, speech is another form of behaviour, so a rigorous epiphenomenalist would actually  have to hold that these rationalisations and confabulations were also nothing to do with the person in themselves.

Philosophically, epiphenomenalism is hard to maintain because it makes the person-in-themselves so completely irrelevant. It seems better to locate personhood somewhere in the material process, or even do without it altogether. Epiphenomenalists are faced with the task of explaining how our conscious thoughts and our physical actions remain in such close harmony, which is impossible to do without raising further difficulties. Philosophically, therefore, the idea is generally regarded as untenable (though it has some appeal for David Chalmers, for example): in psychological discussions, a looser sense of the word is sometimes used, to mean a doctrine that people merely have much less conscious control over their actions than they believe.

We could illustrate the distinction between these two readings of epiphenomenalism by drawing an analogy between a human being and a large corporation. A 'corporate epiphenomenalist' would assert that the chief executive actually had no control over the organisation. On the strict philosophical reading, this would have to mean he had no influence on the organisation at all; on this view we should find ourselves forced to admit he drew no salary (because that would affect the balance sheet, however slightly), had no office, and, in fact, was effectively invisible and intangible. On the looser reading, these difficulties would not arise: the chief executive could be a substantial member of the corporate community: he could even write the Annual Report and issue convincing press releases; he just wouldn't normally be able to get any internal memos or instructions acted upon.  


Solipsism home


Solipsism is the belief that nothing really exists except oneself. So far as I know there are no actual solipsists (then again there might be hundreds of silent ones - solipsism rather undercuts the incentive to publish); nevertheless the idea is interesting in that it represents one extreme version of the strategy of putting the boot on the other foot. This strategy is the one adopted by those idealists who, rather than accepting the physical world and trying to explain how consciousness goes with it, begin instead with consciousness, and demand a justification of belief in the physical world. The ultimate position in this direction must surely be to deny the existence of anything outside your own mind.

Solipsism also functions as a kind of black hole which threatens to draw in other radical and sceptical viewpoints, especially epiphenomenalism. If you're an epiphenomenalist, you must believe that any other conscious entities in the world are incapable of communicating with you, or exerting any influence on the world - so what grounds can you have for believing in them? It would be a more economical hypothesis to believe that they, and the material world, are really just figments of your own imagination.

Ultimately it's impossible to prove that solipsism is wrong; we can only point to the regularities in our experience which are most easily explained by the existence of other entities, some of them conscious. The strongest objection is that, taken seriously enough, solipsism doesn't make any difference. So everything but me is just my dream - so what? Even dreams are entities and require explanation, and to ignore them, while we are dreaming, would be as irrational as ignoring reality while awake.  


Panpsychism home



Panpsychists believe that everything is to some extent conscious, or looking at it another way, that conscious entities are all there really is. This neatly reverses the usual argument by taking consciousness as given and forcing opponents to explain and justify their belief in the reality of inanimate objects. Those who lean towards relativism may find this a comfortable position because it allows them to conceive of the world as consisting only of points of view.

Equally, if you are one of the people who think real, phenomenal consciousness has no place in the chain of physical cause and effect, why shouldn't everything be conscious? It wouldn't make any perceptible difference, so there's no way you can prove it isn't the case: and if everything were conscious, it might help to make the prescence of consciousness in human beings seem less anomalous. David Chalmers speaks rather wistfully about the attraction of panpsychism, without quite comitting himself to it. Equally, however, there can be no good reason to believe it is the case; and the usual principle, known as Occam's Razor , is not to believe in any additional entities (such as the consciousness of stones and anvils) which aren't strictly necessary. 

There are a few issues to deal with, too: if everything is conscious, what counts as a thing? If this stone is conscious, is the left hand side of it separately conscious too? Is the strange but surely valid entity consisting of my left foot and a small portion of Tower Bridge conscious?

But the ultimate problem is that panpsychism does not really revolutionise the argument the way it promises to do. The real issue facing us is to explain the difference between being conscious and not being conscious, and in another form this remains the problem even for those who think everything is conscious. If panpsychism is true, an anvil has a form of consciousness, but one which does not influence its behaviour. My body, which after all is a physical object too, must have this same sort of ineffectual consciousness, but I also have another kind, the kind which causes me to speak, move, and act. Explaining the relationship between these two kinds of consciousness comes down to much the same thing as explaining my consciousness from a non-panpsychist materialist position.


Blandula I'm afraid this won't do at all. Far from being a 'dead end', panpsychism and panexperientialism are enjoying a rising tide of popularity. See this piece and the comments on it for a more balanced perspective...