Interview by Jeremy Stangroom

Stangroom: How did you become interested in the philosophy of mind?

Honderich: By way of determinism, about which we spoke before. It involves three large problems: the first is formulating a theory of determinism, getting a conceptually decent theory; the second is the problem of the truth of the theory; the third, the problem most attended to by philosophers, is the consequences of determinism, its upshot for our lives, say for moral responsibility or hopes.

So step one in writing my tome A Theory of Determinism: The Mind. Neuroscience, and Life-Hopes, was trying to get an adequate theory to operate with, and that really involved arriving at a whole philosophy of mind. Determinism got me going in my present direction, on the way to the World Congress of Philosophy in Boston, pleased with myself, to tell my new tale of what consciousness is to some of the 5,000 philosophers that are to be on hand.

 

Stangroom: How do you see the philosophy of mind?

Honderich: There are two fundamental questions within the real philosophy of mind - and the lower-order or computerized theorizing about the mind that threatens to engulf us. The first question is about the nature of the consciousness, or, more specifically, the nature of the events which make up our streams of consciousness - call them conscious or mental events. The second is about how consciousness is related to the brain - the mind-body problem. Of course, you can answer the first question in a way that also gives an answer to the second question. If you say that the nature of consciousness is that events of consciousness are neural events, you also specify the relationship between mind and brain - conscious or mental events are identical with neural events.

For anybody of a naturalistic turn of mind, of course, this physicalism or materialism or identity theory has always been a theory with an attraction. If you sign on, you can stop being baffled about how there can be causation between thoughts or feelings and physical events like arm movements, which obviously there is. I was certainly attracted to physicalism, but thought it needed some clarification. The clarification seemed to me - it still does - to turn it into two things: call them strict identity theory and dualistic identity theory. You get to this clarification by asking some questions.

If somebody says that a certain conscious event was identical with a certain neural event, the very first thing that has to be done is to settle what is meant by the claim of identity - saying the one thing was identical to another. It is a remarkable fact that a lot of contemporary philosophy of mind, in its great hurry, never really asks that question.

If you do ask it, there are lots of possible answers, but it seems to me that the only tolerable answer is Leibniz's. To say that two things are one thing - or, better, to say that the thing designated by one name or description is one and the same thing as the thing designated by another description - is to say that each of them has all and only the properties of the other. To say a is b is to say a has all and only the properties that b has. And to say that a mental event was identical to a neural event is to say that the first had all and only the properties of the second.

So an identity theory is a claim to the effect that when you took in this green lamp a moment ago, the conscious event of your visual experience was identical with a neural event. The thing picked out by the first description, about seeing the lamp, had all and only the properties had by the thing picked out by the second description, where that is some description of a neural event - in terms of axons, transmitter-substances and so on. M, your past conscious event, had all and only the properties of N, the neural event. But then another question arises.

What properties did M have? Something comes up here, in one way or another. Almost everybody thinks that conscious events are different, that they have a character that can be called their subiectivity. Most of us are inclined to this kind of idea, even if it turns out to be the most difficult thing in the world to explain.

Suppose that the identity theorist says, in answer to the question about the properties of the conscious event, that they were subjective properties. If he does say it, his view becomes a disastrous nonsense. because he is saying that N had all and only the properties of M. The supposed neural event had only subjective properties. This line of reasoning turns the brain, neural structures, transmitter-substances, into only subjectivity - we mentalize or subjectivize the brain.

Things are as bad, of course, if you start at the other end of his equation. You ask the identity theorist what he means by N, the neural event, and he answers, let's say, that it has only neural properties, only the properties of transmitter-substances and so on. Then the upshot is that M, the mental event has only such properties. You neuralize the mind. You leave out consciousness entirely, another disaster.

So the identity theory, that M was identical to N, interpreted in Leibnizian fashion, and with essential questions answered in certain ways, reduces to either the crazy strict identity theory that the brain is just mental, absolutely unlike what we know it is like, or to the other crazy strict identity theory that the mind, or consciousness, is only cortex, neurotransmitters and the like.

Of course the whole story can go differently. You say M was identical to N; you suppose that that has to mean Leibnizian identity; you ask what properties did M have; and the answer you give, different from the one considered a moment ago, is not that M had only properties of subjectivity, but that M had both those properties and also neural properties. It had both the character of subjectivity and an electrochemical character. Most contemporary identity theorists take something like this line. Donald Davidson does, for one. He specifically states that when he says that the mind is the brain, that mental events are physical events, he does not mean that mental events are nothing but brain events. What he means is something that follows from answering, when asked what properties did a mental event have, that it had both mental and neural properties. This position immediately escapes the two earlier disasters. The mind is not neuralised and the brain is not mentalised. You have a dualistic identity theory. 

 

Stangroom: Why is this still an identity theory at all?

Honderich: You might well ask! What it says is that M was N, and that turns out to mean something like this: that there was one thing that had two different properties. If you ask why is this an identity theory, I guess it's because it says that there is one event in question, with the two properties. But obviously, if you might call it an event-monism, it is also a property-dualism. Davidson's identity theory as I mentioned, his anomalous monism, seems to be just such a thing. I don't think that fact of classification is important. What is a lot more important, as I see it, is that this view of his amounts to an epiphenomenalism because it doesn't involve any other relation between the two properties - say the mental one being necessary to the neural one.

 

Stangroom: What is your objection to epiphenomenalism, by the way? You seem to consider it axiomatic that it is false.

Honderich: When we're actually put on the spot and asked to prove that epiphenomenalism is false, it is not easy to do. However, we can say something. I do really take it that it is an axiom that an explanation, a full explanation, of your being here now in this room can't possibly leave out your recent desires, intentions, beliefs, plans and the like - where those things were a matter of your consciousness, of subjectivity. It would seem to me to be mad to think that your being here now could be fully explained without reference to any of that stuff. But epiphenomenalism is precisely the view that the explanation doesn't need to have in it anything at all at all about consciousness in this real ordinary sense. It is presumably the case that in any kind of enquiry - think of formal logic if you want - there are ground-level intuitions from which the enquiry starts. Not everything gets to be proved. I think mental causation or mental efficacy - the denial of epiphenomenalism - is that sort of thing.

 

Stangroom: Is there a possible defence here? Might one argue that whilst consciousness, because it is an emergent property of the neurological, will always be present, it plays no causal role in action - because what is causal is necessarily physical, and consciousness is not physical, and thereby one can have an epiphenomenalism that allows for the presence of consciousness but denies to it any causal role?

Honderich: Well you have provided a good and accurate definition of epiphenomenalism. It is essentially Huxley's nineteenth century view, which is not a denial of consciousness, but a denial that it does anything, that it is explanatory. It is the view that although the mental property exists, it is just a side-effect. And that, I put it to you, is unbelievable. I'm not surprised that hard-nosed philosophers of mind don't buy epiphenomenalism. Take Davidson again. He's extremely concerned to resist the charge that Anomalous Monism is epiphenomenalist - that, as you can put the point, it's a doctrine that denies that our reasons, in a real sense of reasons, have anything to do with our actions.

 

Stangroom: To go back to the matter of consciousness, and the mind-body problem, does Davidson make any attempt to describe what consciousness is, what subjectivity might be?

Honderich: No he doesn't, so far as I can see. His characterisation of consciousness or the mental avoids really facing up to the question of what it is. He gives a kind of linguistic criterion of a mental event, which he allows doesn't work very well, but the main point about it is that it isn't even an attempt to give the real nature of such an event, or of course of consciousness generally. He doesn't go in for what can be called mental realism - trying to get to an answer to the fundamental question rather than avoid it in one way or another.

 

Stangroom: Is Functionalism a good answer? First, what do you take it to be?

Honderich: It can begin, I take it, from what is dead obvious. If anybody wants to give an adequate account of what it is, say, to want or desire something, maybe a glass of wine from that bottle over there, they could not adequately do it without mentioning what leads up to the wanting and what follows from it. It would be impossible to characterize the wanting without mentioning the prior perceptual experience and then something about subsequent behaviour. So it does seem perfectly obvious that adequate definitions of particular mental events will include a reference to the prior causes of those events and their subsequent effects. Wanting, among other things, is something that comes in between the seeing and moving.

What functionalism does is to go mad with this true idea. It says that mental or conscious events have only such causal properties. That is all conscious events are. The mental or conscious event of wanting the glass of wine is only the event which has a certain ancestry, is the result of a certain input, comes from certain possibly complicated causes - and then issues in simple or complicated behaviour. A mental or conscious event is no more than a causal relatum. 

Built into this doctrine, indeed fundamental to it, is something called variable realisation. It's essentially this. We^’ve said that a mental event is only what stands in certain causal relations. And we haven't said any more about it than that. Indeed, we’ve claimed that you need no more than that to have the mental event. The mental event is whatever is in those causal relations. It can be that the event is 'realised' in neurons, as in our case. But it could be realised in silicon. One day there may be a computer which realises the desire to have a glass of that wine. Or there might be a creature that arrives from outer-space, neither silicon nor with our neurons, that realizes the same mental event in some third material or stuff. This is variable realisation.

So the idea is that when you say mental or conscious events are functional events, you say something which is not unswallowable in the way that the strict identity theory is unswallowable. Functionalism is supposed not to go against our conviction that mental events have more than neural properties. That doesn't seem to me right. Functionalism isn't really any more swallowable than that strict identity theory. It leaves as much out - the subjectivity.

In the case of human beings, functionalists say - although they deny its importance - that conscious or mental events are in fact neural events. They have those essential causal roles, but yes, they are neural events. Functionalists have to face this question: does a particular neural event come to have anything more to it, more properties or another character, if it's a fact that it could be replaced in its particular causal role by silicon? I put it to you that the answer is no. You don't get some further property assigned to a neural event with a causal role when somebody says of it that it could have been a silicon event. So in my view functionalism is just as unswallowable as the strict identity theory. It leaves out as much.

 

Stangroom: I get lost there. Why does that make functionalism unswallowable? And where does rejecting it, as well as a strict identity theory, leave you?

Honderich: Functionalists think that they give a more enlightened and a more credible account of conscious events than is given by the neuralizing identity theory. The idea is that the strict identity theory does indeed leave something out - what we are calling subjectivity - and that somehow functionalism gets something like it into the story when it identifies consciousness not with material stuff but with causal role. But, of course, they concede that in the case of human mental events - forget about the bloody computers and Martians - they are neural events. And, if you think about it, these events don't become something more than neural events in virtue of the fact that they might have been the same effects and causes but in different stuff.

As for where rejecting strict identity theory and functionalism left me, that also had to do with something else - essentially a commitment to naturalism. What such a commitment comes to, roughly, is a belief that what really or fundamentally exists is physical, and that the fundamental method of inquiring into reality is the scientific method and methods close to it, including a philosophical one. So consciousness really is somehow a matter of the physical. Of course something has to be said about what physicality is - about the realm of the physical, the nature of physical things.

There is a very bad current definition of the physical realm, if you can call it a definition. What it boils down to is that what is physical is whatever science lets in, what science says exists. The definition has got so many shortcomings that I wonder why so many people countenance it. In the first place, it's just uninformative, only a signpost. Also, psychology is presumably part of science, and it's a question whether psychology lets in the mind in a real sense. But then the definition of the physical is really unsettled - and at a crucial point. There is also the disability that the definition is presumably relative to current science, and so is going to be good, if it is any good at all, for only about ten years. 

A traditional view of the physical is better. It's set out, for example, by Anthony Quinton in his book The Nature of Things. Here the idea is that the physical consists in two lots of things, both of them things that take up space and time. The first lot of things also have perceived properties. The second lot of things lack perceived properties but are in causal or other lawlike relations with things in the first category - with spatio-temporal things with perceived properties. Sofas go into the first category and atoms go into the second.

 

Stangroom: So what about mental events?

Honderich: Well, given naturalism, they had to be physical. They had to be in one of the two categories. And obviously, since they themselves aren't perceived, they went into the second category. They had to be spatio-temporal things in causal connection with perceived things like our movements. So the view that seemed to me hard to avoid, until very recently, was roughly as follows. If you ask what a conscious event is, the answer is not an event with only neural properties. A conscious event has to be an event with two kinds of properties, neural properties and subjective properties - but those subjective properties have to be some kind of physical properties. And there is a lawlike connection between the neural and the subjective properties, by the way, which makes the thing different from Davidson and avoids epiphenomenalism.

The trouble is that it's pretty outrageous. What the view amounts to is that as we carry on our conscious lives, there are two kinds of facts, both of them physical, going on in our heads. There are the neural properties of events, and there are the conscious properties of events, with the conscious properties being physical too. Awkward.

 

Stangroom: When you were thinking along these lines, how did you conceptualise the physical properties of subjectivity? How did you see them?

Honderich: Not very well. I thought about them as little as possible. But sometimes things like electro-magnetism came to mind - fields of force, that sort of thing.

 

Stangroom: But didn't all that beg a lot of questions? It doesn't specify the relationship between the neural and the subjective, and it doesn't say anything about what the subjective comes to, except that it's physical.

Honderich: Well, I did specify the relationship as lawlike connection, but you're right that the subjectivity didn't get explained. Something else also came over me after writing the piece in question. ('Consciousness, Neural Functionalism, Real Subjectivity’, American Philosophical Quarterly October 1995) It was bad enough to have two kinds of physical properties going on in the head, the neural ones and the conscious ones. But there was also this other thought. Suppose that in 50 years they actually discover other physical properties that are non-neural, and there are reasons for trying to identify them with consciousness. Won't a philosopher then say: 'Well, this isn't really consciousness - this leaves out subjectivity! This is as bad as the old strict identity theory and functionalism.

The result of my embarrassment was six months of struggle and a lecture in the Royal Institute of Philosophy series last year. ('Consciousness as Existence", Current Issues in the Philosophy of Mind, ed. Anthony O'Hear). It's a very different attempt to say what consciousness is. Or anyway perceptual consciousness - the experience of seeing, hearing and so on, which must be fundamental to a complete account of consciousness. This story of perceptual consciousness may be right or wrong, sane or crazy, but it's definitely not the same old stuff.

It has a fair amount to do with what is called the phenomenology of consciousness, what consciousness appears to be - and the idea that really there isn't any more to consciousness itself than the appearance of it, what we can non-inferentially report. Ask yourself the question 'What is it for me now to be perceptually conscious, conscious of the things around me? What does this come to?' A pretty good answer is that what it comes to is this: 'Things somehow exist, a world somehow exists, somehow things are out there in time and space and with other qualities.' What the new view comes to, very roughly, is that for you to be perceptually conscious is for a certain state of affairs to exist. It's a state of affairs that resembles the state of affairs that is, so to speak, the existence of the first category of things in the physical world, the things occupying space and time and having other perceived properties. Perceptual consciousness ceases to be a matter of something in your head.

 

Stangroom: To this idea that somebody's being perceptually conscious is for things somehow to exist out in space and time, for a world of perceptual consciousness to exist - couldn't it be objected that the world in question is just a mental world, one that depends on the perceiver in various ways? So you're no further ahead. You're just defining perceptual consciousness as involving a world of consciousness?

Honderich: It's true about the dependence. But what the objection overlooks, to go back to the traditional definition of the physical world, is that the perceived part of that world has related dependencies on perceivers. That's an undeniable fact. And - this is the crucial point - these dependencies don't lead us to put that part of the physical world inside heads. So why should the dependencies in the case of a world of perceptual consciousness turn that world into, so to speak, just a mental one?

I know this is a little baffling. Indeed, I find it a little baffling myself on occasion. But the view has a lot of recommendations. In addition to being true to the phenomenology - and there's a lot to be said for that – it really gives content to talk of subjectivity. What is subjective is, so to speak, a world, something different from the physical or objective world. But something that also has the reality of being spatio-temporal and so on. That last bit, the reality, is in good accord with our conviction that consciousness is somehow real - we aren't much affected by our idea of consciousness as ethereal or gossamer stuff. Finally, the view of consciousness as existence does well with the mind-body problem - causal relations between a world of perceptual consciousness and the physical world, anyway according to me, aren't baffling.

 

Stangroom: So have you just dispensed with your previous positions?

Honderich: Yes. I don't mind too much. In particular, it's good to be rid of the idea of a special kind of physical stuff in the head that is non-neural. That really had to go. And I do really think that the idea of consciousness as existence has a chance. It's fertile, and a kind of near-naturalism, and it seems to me to give us more of what we have to get into an account of perceptual consciousness than strict identity theory and functionalism and so on.