Towards the end of our first series of Gifford Lectures Professor Waddington complained that we had none of us given a definition of mind. To give such a definition is not in fact hard, though of course a definition by itself will not solve many problems. I want to start my contribution by offering a definition which I hope will be sufficiently broad to capture our intuitive notions of what mind should be, while sufficiently precise to provoke philosophical disagreement.
The mind is the capacity to acquire intellectual abilities. This definition, it will at once occur to you, does not get us very far unless we add to it a definition of ‘intellectual’. I will do that in a moment, but first I want to draw your attention to some features of the definition as it stands. I define mind as a capacity, not as an activity. Thus it is possible to say that babies have minds even though they do not yet display intellectual activities of the appropriate kind. Secondly, mind is not only a capacity, but a capacity for capacities. Knowledge of a language such as English is itself a capacity or ability: an ability whose exercise is the speaking, understanding, reading of English. To have a mind is to have a capacity at a further remove from actualisation: to have the capacity to acquire such abilities as a knowledge of English.
By intellectual activities I mean activities involving operation with symbols. On this definition, clearly the use of language, the solution of a mathematical problem, the painting of a portrait are, as we would expect, intellectual activities. The definition leaves many things uncertain at the borderline, but this is all to the good, for so does our ordinary concept of mind that we are trying to capture in a definition. We are left in doubt by this definition whether Washoe and Sarah have minds, because we are in doubt whether they are using language, and therefore whether they have acquired the capacity for intellectual activity; and therefore whether they possess the capacity, for acquiring such abilities, which is what I am calling mind. On the other hand all the things which are well within the realm of what we regard as mental activities—like physics, philosophy, and poetry—are well within the bounds laid down by this definition. (Music is an exception to this, as to many other generalisations about the arts and the mind.) And according to this definition, I would want to say, computers very definitely do not have minds. They do, in a sense, operate with symbols, but with our symbols; they are not symbols for them, it is we and not computers who confer the meaning on the symbols. So I have to add something to the definition. To have a mind is to have the capacity to acquire the ability to operate with symbols in such a way that it is one's own activity that makes them symbols and confers meaning on them.
Not only does the definition have a broad borderline, it is in a way infinitely open. For many different things are called operating with symbols, and we cannot set any bounds to the possibility of the invention of new types of symbolic operation. At no time can we draw a boundary around the forms of symbolic activity current at that time and say: ‘Those activities and no others, count as intelligent activity’. Nor can we be sure that we shall ever be able to isolate from them some common element and say: ‘Anything which is intelligence must have this element’, unless we leave the notion of element as open as the notions of ‘symbol’, ‘meaning’ and ‘activity’ as I am using them.
In these lectures we have often stressed the importance of the possession of autonomous, long-term goals as a constituent of mentality. It may appear that my definition neglects this, but in fact it does not. The pursuit of self-selected goals that go beyond the immediate environment in space and time is not possible without the use of symbols for the distant, the remote, and the universal. And on the other hand, the use of symbols itself involves purposes which go beyond the temporal and spatial present. First of all, meaning something is a matter of intending, and intending involves having purposes. Secondly, to use something as a symbol and not as a tool is to use it in such a way that any effect which it may have on the environment lacks the immediacy and regularity characteristic of physical causality. So the mind, as I have defined it, is a volitional as well as a cognitive capacity: it includes the will as well as the intellect.
There are two ways in which my definition, traditional though it is, departs from familiar approaches to the definition of mentality. First, I do not take the making and using of tools as by itself an exhibition of mentality. Assisting oneself by inert instruments in the performance of an activity may or may not be a manifestation of mind; that depends, among other things, on whether the activity itself is. Thus the use of clocks to tell the time or the programming of computers to simulate natural languages is an intellectual activity; the use of a stick to shake a banana from a tree is not.
Secondly, in my definition of mind I have not said anything about consciousness. There are at least two sharply distinct things which may be meant by that slippery term. The first is the consciousness which is more or less the same thing as perception: the awareness of, and ability to respond to, changes in the environment; the senses like hearing, seeing, smelling and tasting. The second is self-consciousness: the knowledge of what one is doing and of the reason why one is doing it. In human beings self-consciousness presupposes sense-consciousness but is not identical with it. Self-consciousness presupposes something else also, I should maintain: it presupposes language; one cannot know how to talk about oneself without knowing how to talk, and one cannot think about oneself without being able to talk about oneself.
I say this last not because of any general thesis about the relationship between talking and thinking, but because of a particular reason connecting talk about oneself and thought about oneself. A dog may well think that his master is at the door: but unless a dog masters a language I cannot see how he can think that he is thinking that his master is at the door. There is nothing that the dog could do that could express the difference between the two thoughts: ‘My master is at the door’, and ‘I think that my master is at the door’. If I am right that self-consciousness is thus intimately connected with language, then I can take account of the tradition that regards self-consciousness as closely linked with mentality without mentioning it specially in my definition. On the other hand, by distinguishing between mentality and sense-consciousness I am able to do justice both to my admiration for Descartes and my affection for my dog. I can agree with the former that animals do not have minds while according to the latter a full measure of non-mechanical consciousness.
Consciousness is sometimes identified with the acquaintance with a private world within oneself. I said several times last year that I think that this picture involves a philosophical confusion, and I won't bore you by repeating my reasons for thinking so. I will just say that the confusion seems to me to arise from people's being over-impressed with their ability to talk to themselves without making any noise, and their ability to sketch things before their mind's eye instead of on pieces of paper. I think that the acquisition of the ability to talk about oneself is enormously significant; the acquisition of the ability to talk to oneself is by comparison merely a matter of convenience. A society which differed from ours only in that everyone thought aloud all the time instead of thinking silently would be perfectly conceivable, equally intellectual, merely unbearably noisy.
My account of mind is, as I say, a very traditional one. It traces its ancestry back at least to Plato. In the tradition beginning with Plato the mind is thought of as being above all the ability to know universal ideas and eternal truths. Many philosophers have thought that Plato's belief, that beyond our symbols in mathematics and philosophy there lay such sublime entities, was due to a misunderstanding of the nature of symbolism. Perhaps so: but I hesitate to take sides definitely against Plato since the greatest philosopher of logic in modern times, Gottlob Frege, thought that Plato was basically right.
But my definition leaves the question open. If the mind is the capacity to operate with symbols, and if the way we operate symbols is correctly understood in the way that Plato and Frege understood it, then the Platonic account of the mind is to that extent correct.
In several places Plato argues, and in this he has had many followers throughout the centuries, that if the mind can know the eternal and changeless then the mind must itself be essentially eternal and changeless, and at the very least be an immortal entity that can survive the death of the body. This strand of the Platonic tradition seems to me to involve a fundamental philosophical mistake which I shall try to unravel in a moment: but one can see immediately that to get this conclusion Plato needs not only the premise that the mind is the capacity to know the eternal, but also the premise that like can only be known by like. There seems to me to be some truth in this archaic dictum if suitably reinterpreted: for example, if something is to be said to have knowledge of the multiplication table, it need not necessarily have the structure of the multiplication table (it may not even be clear what this would mean), but it must be capable of an output which is isomorphic to the multiplication table. Chomsky's argument that, since English allows unlimited embedding, no finite-state automaton could internalise English grammar would be a modern instance of the theorem that like must be known by like. But the thesis that the knower must have the same properties as the known is plausible only so far as it applies to the structural properties of the knowledge and its object; it concerns the content of knowledge and not the mode of knowing. The Platonists have never provided any good reason for thinking that there cannot be fugitive acquaintance with unchanging objects and temporary grasps of eternal truths. There seems no more reason to deny mortal knowledge of immortal verity than to deny the possibility of a picture in fireworks of the Rock of Gibraltar.
If the Platonist argument for the immortality of the mind embodies a fallacious inference, its conclusion seems to embody a conceptual confusion. If the mind is a capacity, we must ask what it is a capacity of; and the answer seems to be that it is a capacity of a certain kind of body. If this is so, then the notion of a disembodied mind is the notion of a capacity which is not anything's capacity. This seems to be as nonsensical as the notion of the Cheshire cat's smile without the Cheshire cat.
I would like to develop this by some considerations about capacities in general, in which we can hope to study the philosophical nature of the problem free of the emotional perturbances which attend discussions of survival and immortality. There are two things which a capacity must be distinguished from: its exercise, and its vehicle. Take the capacity of whisky to intoxicate. The possession of this capacity is clearly distinct from its exercise: the whisky possesses the capacity while it is standing harmlessly in the bottle, but it only begins to exercise it after being imbibed. The vehicle of this capacity to intoxicate is the alcohol that the whisky contains: it is the ingredient in virtue of which the whisky has the power to intoxicate. The vehicle of a power need not be a substantial ingredient like alcohol which can be physically separated from the possessor of the power, though it is in such cases that the distinction between a power and its vehicle is most obvious (one cannot, for example, weigh the power of whisky to intoxicate as one can weigh the alcohol it contains). Take the less exciting power which my wedding ring has of fitting on my finger. It has this power in virtue of having the size and shape it has, and size and shape are not modal, relational, potential properties in the same way as being able to fit on my finger is. They are not the power but the vehicle of the power. The connection between a power and its vehicle may be a necessary or a contingent one. It is a contingent matter, discovered by experiment, that alcohol is the vehicle of intoxication; but it is a conceptual truth that a round peg has the power to fit into a round hole.
Throughout the history of philosophy there has been a tendency for philosophers—especially scientifically-minded philosophers—to attempt to reduce potentialities to actualities. But there have been two different forms of reductionism, often combined and often confused, depending on whether the attempt was to reduce a power to its exercise or to its vehicle. Hume wanted to reduce powers to their exercises when he said that the distinction between a power and its exercise was wholly frivolous. Descartes wanted to reduce powers to their vehicles when he attempted to identify all the powers of bodies with their geometrical properties.
But if it is important to distinguish between powers and their exercises and vehicles it is important also not to err on the other side. A power or capacity must not be thought of as something in its own right, for instance as a shadowy actuality or insubstantial vehicle. The difference between power and its exercise or vehicle is a category difference, not a difference like that between solid and shadow. To hypostatise a power, to think of a power, say, as something which, while remaining numerically the same, might pass from one possessor to another is one way of erring on the anti-reductionist side, an error which has recently been aptly named by Michael Ayers ‘transcendentalism’. In the Andersen fairy-tale the witch takes the old wife's gift of the gab and gives it to the water-butt. Less picturesque, but equally absurd, examples of transcendentalism can be found in the pages of many great philosophers.
The mind is a capacity, and the philosophical errors which occur in dealing with capacities in general occur in a particularly vivid way with regard to the mind. Behaviourism, when it takes the extreme form of identifying mind with behaviour, is a form of exercise-reductionism: treating the complex second-order capacity, which is the mind, as if it was identical with its particular exercises in behaviour. Materialism, when it takes the extreme form of identifying mind with brain, or with the central nervous system, is a form of vehicle-reductionism: reducing my mental capacities to the structural parts and features of my body in virtue of which I possess those capacities. The Platonist belief in immortality, on the other hand, is a form of transcendentalism: for only a transcendentalist can believe that a capacity can be separated from its possessor, or pass from one possessor to another by incarnation in successive different bodies.
The title of this lecture is ‘The Origin of the Soul’. Despite this I have been speaking about the mind, not about the soul; and I have been talking about its nature, without saying anything about its origin. But the points I have been making are in fact very germane to my main topic. ‘Soul’ is often used as equivalent to ‘mind’, though usually with a different emphasis to stress one or other of the mind's features; for example, to insist that it includes will as well as intellect. Sometimes it is a synonym for ‘disembodied mind’; when we speak of All Souls Day it is only disembodied souls that are quantified. To that extent what I have been saying about minds applies without modification to souls also. In the Aristotelian tradition, however, a soul is defined as the form of the body and as the principle of life, and this definition is rather different from the one I have been using of mind.
First of all, there are many living things that do not have minds, so that if a soul is a principle of life it should belong not only to kings but also to cabbages, a conclusion which Aristotle accepted. Secondly, a principle of life doesn't seem to be the same kind of thing as a capacity for the acquisition of certain abilities. Perhaps that is only because of the difficulty of specifying what are the activities which constitute life; perhaps a principle of life is after all not much different from a capacity for metabolic change. But more seriously, if the soul is the form of the body, then it seems that a body cannot be identified as the kind of body it is unless it has the capacities in question. This seems to be perfectly correct as applied to the souls of cabbages and animals; it is a truistic consequence of being willing to accept the terminology of matter, form, and soul in their case. But in the case of human beings the soul in this sense cannot be identified with the mind: it seems that a body may be perfectly identifiable as a human body, while being unable, through some innate defect, to exhibit any intellectual activity at all.
What, since Descartes, it has been most natural to call the mind was called by Aristotelian theologians the intellectual soul. It was this alone for which they claimed immortality (sometimes on the Platonist grounds I earlier dismissed), and this which they took to be the form of the human body. Concerning this intellectual soul they raised a number of questions which, though they strike a modern audience as bizarre and archaic, do in fact concern interesting, and live, conceptual and empirical issues. One was whether the intellectual soul was the only soul of the human being; another was where it came from, and whether it was created or generated. I want to discuss the second of these for a moment.
The theological debate about the origin of the individual soul began at a very early stage of ecclesiastical history. St Jerome and St Augustine held opposite views on the question. Their disagreement was dramatised by an anonymous author who uses fragments of their writing to construct a lively and ill-tempered dialogue between them, which is published with the correspondence of St Jerome in Migne's Patrologia Latina. St Jerome held that each individual soul was created by God; St Augustine held that each human being inherited his soul from his parents, or rather that each human soul was generated by the souls of the parents in the way that each human body was generated by the bodies of parents.
St Augustine's theory was clearly motivated by an attempt to do justice to the biblical doctrine of original sin. In one man, St Paul had said, all men sinned; and St Augustine reasonably enough thought that men could hardly inherit sinful souls from Adam unless they inherited souls from Adam; souls, like bodies, must all belong together on a family tree with its roots in the Garden of Eden. Otherwise it seems that we must think of God as somehow gratuitously stamping sinfulness on the new souls he mints afresh with each new conception of a human infant.
I don't wish to suggest that we should this evening discuss the doctrine of original sin. Some recent authors, like Iris Murdoch, have spoken sympathetically of it out of disgust with liberal optimism about human perfectibility, and some ethological writers have claimed that the human race is a uniquely murderous species by comparison with its evolutionary ancestors. But I do not think that such authors mean to suggest more than that the human nature we inherit is in some way diseased or vicious; I do not think they mean to say, as St Augustine and other theologians did, that this inheritance of a crippled nature is in some way a punishment for a historical voluntary action of a common ancestor of the race. In theological jargon, the modern versions are interested only in peccatum originale originatum, not in peccatum originale originans; in original sin but not the original original sin.
Only those who accept the doctrine of original sin in a fairly fundamentalist form have precisely the same motivation as Augustine for speaking of the inheritance of souls. The majority of scholastic theologians in fact accepted Jerome's view rather than Augustine's. Strictly speaking, those of them who were Aristotelians should have given the answer that the question about the origin of the soul was a senseless one. For the soul was the form of the body, and according to sound Aristotelian doctrine forms don't really come into existence in the way that babies do—as the Latin tag had it: forma nec est nec fit nec generatur.
This answer seems to me right: there can only be a question of the origin of the soul in the sense that we can seek an explanation of why human beings have the intellectual capacities they have. Later we shall have to enquire what can be said about this question with regard to the origin of the intellectual capacities of the species as a whole: at present we are interested in the ontogenetical question of the transmission of intellectual abilities from one generation to the next.
Here it seems we find in the contemporary situation something very parallel to the ancient debate between Jerome and Augustine. I am not referring to the controversy about the importance of inherited factors in the determination of individual differences in human intelligence, though this controversy has been carried on with a dogmatic fervour and a sourness of temper not too unlike that of the two Church Fathers I mentioned. I refer not to the question of whether a human being inherits a particular IQ, but how he inherits intelligence at all from the previous generation. Here, it seems to me, the disagreement which we have seen in these lectures between Christopher and John bears a more than accidental resemblance to the quarrel between St Augustine and St Jerome.
Let me explain. On the one hand it seems obvious that we inherit our soul, our mind, our capacity for the intelligent activities characteristic of human beings. Surely, it is a part of the human nature we inherit from our parents. On the other hand, John has produced a very ingenious argument, based on Gödel's incompleteness theorem, to show the human mind does not operate algorithmically. Now it seems to me that if the mind does not operate algorithmically there cannot be any algorithm for constructing it either. But according to Christopher's account last week—and indeed according to most popular presentations of the mechanisms of inheritance—human beings are constructed by algorithms. Christopher, if I remember rightly, described the chromosomal DNA as containing ‘the program for constructing the human being’, but if John is right there cannot be any such program. And in his dialogue ‘The Seat of the Soul’ Christopher suggested that the soul itself might be viewed as ‘a special program in charge of all our various subroutines’ (Towards a Theoretical Biology, 3, p. 241). So presumably Christopher, like St Augustine, takes the view that our soul is inherited from our parents. If he wanted to, he could no doubt make his system broad enough to accommodate original sin as the inheritance of a bugged program. Whereas for John, as for St Jerome, every new human being represents in some way a completely fresh start, inexplicable by the history of heredity up to that moment.
John has not, so far as I know, claimed, as St Jerome did that each human soul was a fresh divine creation. He has offered his argument from Gödel's theorem as a proof of the freedom of the will, but not, so far as I am aware, as a proof of the existence of God. Indeed he has claimed several times that his views do not involve setting up any ‘keep off signs for science. But it seems to me that in consistency he must set up such a sign in the case of the origin of the soul. He is compelled to reject the mechanism of heredity as commonly explained. Last year he committed himself to the view, that any causal explanation must be capable of expression in an algorithm (The Nature of Mind, p. 74). It seems to follow that he must reject the possibility of any causal explanation of the possession of intelligence by any human being. I don't know whether this involves him in postulating a direct divine creation of the individual human soul; but if it isn't a’ keep off sign to science, then my name is Trespassers William.
I must leave it to Waddington to say how much is metaphorical and how much literal in popular presentation of the mechanism of inheritance. I would like to learn from him whether current genetic theory does indeed imply that hereditary features are passed on by means of an algorithm. Only then will I be able to tell if the identification of Christopher with St Augustine and of John with St Jerome is as apt as it appears superficially to be.
Certainly we humans are always inclined to explain things we only imperfectly understand in terms of the most advanced technology of the age we live in. Nowhere is this more true than in the case of the mind. One of the most bizarre, as well as the most ubiquitous, misunderstandings of the nature of the mind is the picture of mind's relation to the body as that between a little man or homunculus on the one hand and a tool or instrument on the other. We smile when medieval painters represent the death of the Virgin Mary by showing a small scale-model Virgin emerging from her mouth: but basically the same idea can be found in the most unlikely places. The great Descartes was one of the first people explicitly to warn against the homunculus fallacy, but he fell into it himself when he tried to explain the relation between mind and body in terms of transactions at the pineal gland. The soul was supposed to read off images on the pineal gland, and then use the gland as a sort of tiller to steer the body on its way by means of subtle fluids called the animal spirits. The homunculus fallacy is still with us, and it is not difficult to find in the writings of distinguished contemporary psychologists passages which suggest that there is a man within a man who reads the information stored in the brain and triggers off impulses which set the body in motion.
However, as time passes and technology progresses the tool or instrument which the manikin is fancied to control gets more and more sophisticated. Thus Plato thought that the soul in its relation to the body could be compared with a sailor in a boat or a charioteer holding the reins. Many centuries later Coleridge said that what poets meant by the soul was a ‘being inhabiting our body and playing upon it, like a musician enclosed in an organ whose keys were placed inwards’ (Letters, i, 278). More recently the mind has been compared to a signal-man pulling the signals in his signal-box, or the telephone operator dealing with the incoming and outgoing calls in the brain. Finally, for Christopher, the boat, the chariot, the railroad and the telephone exchange give way to the electronic computer. The soul is to the body as the programmer is to the computer, so that he can describe his difficulties in handling decimal coinage as failures in programming.
I trust that Christopher means his talk of people programming their thinking as nothing more than a metaphor. As a metaphor, manikin talk may be no more than harmless necessary fancy. But it is much harder than one thinks to keep such fancies out of the constructions of one's theories, and once a manikin gets into a theory it turns into a very devil. An explanatory theory bedevilled by a homunculus is a failure as an explanatory theory: because whatever needs explaining in the behaviour of a man turns up, grinning and unexplained, in the shape of the manikin.
I'd like to take Tony up first of all on the things he said almost at the very end of his lecture, when he attributed to me the view that the body is to the soul as the computer is to the programmer. In ‘The Seat of the Soul’ I was trying to express the idea that the body is to the soul as the computer is to the programs which it is capable of implementing, and it is capable of implementing these programs by virtue of having been primed in advance with the requisite software; some very clever person actually prepares the computer so that it can implement the programs which we feed into it. Now I think it's very important in discussing the origin of the mind to make a clear distinction between the things a computer can do when it has been programmed, or a human being can do when it has developed into a mature creature, and the question of how it got that way. I think Tony has in fact rather blurred the distinction between these two issues. The question of ontogeny, as it faces the biologist in relation to the phenomenon of mind, is the question of how do the DNA blueprints actually get obeyed? How did they get consulted and how did the instructions get carried out? These are very difficult and very intricate questions in developmental biology, but once the creature is mature then it seems to me we have a completely different set of questions about the relation between its mature mind and its mature body, and I would want to say that we shouldn't make the mistake of obliterating the comparison between software and hardware on the one hand, or between minds and bodies on the other, because we happen to know, as a matter of fact, that computers are put together by human beings and primed by human beings with the relevant software.
So I think I must go and ask Tony whether he wouldn't allow that there's a sense in which a successful project in artificial intelligence would, in some respects, get around the problem of the manikin or homunculus. Is it not possible in principle for a man-made system to manifest intellectual activity, or at least to manifest behaviour which we ordinarily take to be perfectly clear evidence of intellectual activity? I believe that the computer is not just another inadequate technological model of the mind, though of course it is extremely inadequate in many respects; because even if existing programs miss the mark in detail, or indeed, in important matters, I still think that we have here a very useful and important paradigm, on the basis of which we can hope to understand a little bit more about the relation between minds and bodies. So doesn't Tony allow that there is a real difference between the computer paradigm for mental activity and all previous models; and doesn't he agree with me that the origin of a system should not affect our interpretation of its behaviour when constructed? Last year indeed, he allowed that if a man from IBM arrived and said: ‘I want to open you up and service you’, he might be surprised, even horrified, but would not regard the incident as annihilating his own claim to consciousness or to having a mind. If, in such bizzare circumstances, Tony would resent the suggestion that he had not really been using symbols, why should he deny that a more ordinary computing system manifests mental activity when it is operating with symbols, even though we are responsible for its doing so?
At the monent I just want to try an ad homunculum argument against Tony. He defined mind as something which the body possesses; and that, of course, underlaid the main thrust of his argument to clear the swipes he was going to make against me. But later on when he was talking about his own mind, he didn't talk about a particular body possessing that mind but the little word ‘I’ slipped out—‘I possess’, ‘the mind that I possess’—and I think this in effect shows that his usage, as often, is better than his theory; and that we should regard minds not only as being possessed by bodies but as being possessed by persons; and that it's best to see both a body and a mind as something which characteristically is mine, yours, his, hers, theirs and so on.
I will only raise one consideration to try and persuade you that this is the right view, and the exact opposite of the mistake that Tony said was made by Platonists, transcendentalists, and for that matter by Pythagoreans, and the tellers of fairy stories, who allowed the possibility that the soul of my grandam might now inhabit a woodcock, or that the frog there might really be a beautiful prince. And it seems to me that these stories, although as far as we know untrue, and certainly very difficult to verify, are intelligible; one can conceive oneself waking up with an entirely different body; indeed, it is an important exercise in moral philosophy, particularly in the tradition of Kant and Hare, to conceive oneself not only in someone else's shoes but in somebody else's skin; and if this is intelligible, then it follows that mind shouldn't be conceived as being necessarily possessed by a body, but must be something independent of it.
I confess I feel rather out of my depth this evening. Tony defined ‘mind’ so that it relates only to things that use symbols, by which in practice he meant words. He defined it as essentially involving the use of language. Being a biologist who deals with the lower animals, most of my knowledge is about beings which do not have language, and therefore in these terms do not have minds. If we restrict ‘mind’ so that it refers only to a human characteristic, then we need some other word to fulfil a comparable function in relation to the non-language-using animals. But anyway, discussions of the meanings of words often make me feel that I am rapidly getting out of my depth, and attempts to define ‘mind’ and still more ‘soul’ soon take me into a region where I feel I can't get even one big toe down to a solid bottom. There is just one general remark I would like to make about Tony's definition. He says that mind is a capacity. Now I think that in much of modern science we have been trying to get away from speaking about capacities. Whisky has an intoxicating capacity; Benzedrin has a stimulating capacity; Barbiturates and other things have a dormitive capacity, as Molière pointed out. When I first started studying development, the orthodox answer to the question ‘Why does this part of an embryo develop into the nervous system?’ was that it had a neurogenic potency; and I spent quite a number of years finding facts and reasons which would persuade people that it would be more illuminating to give the answer ‘Because something happens within that region to switch on the genes which synthesise the proteins characteristic of nerve cells’—or at any rate some answer with that general character. ‘Capacities’ always make me feel uneasy. The natural sciences do not mind talking about what Tony calls the possessor of a capacity, or its vehicles, or its exercises-it is the capacities themselves that it finds difficult to make any sense of. I should prefer to say that ‘mind’ is the name for a category of activities, either observed, or more usually inferred, and has a logical status very similar to that of the word ‘digestion’ in the phrase ‘he has got a good digestion’.
Tony asked me one specific question: ‘Is it legitimate to say that inheritance operates by algorithmic processes, or is that statement just a piece of fashionable jargon?’ My short answer would be: ‘Yes, it is legitimate, it is not just fashionable’, but that needs a little expansion. Inheritance in the narrowest sense, just the passing on of hereditary factors from one individual to its offspring, is not an algorithmic, but a simple mechanical process involving such steps as the segregation of chromosomes, the union of sperm and ovum and so on. But Tony was using the word in a broader sense to refer to the inheritance of a complex characteristic of an organism from its ancestors; for instance, what is involved in the inheritance of mind? Recent biology does not say that the mind of your parent is, as it were, packed up small and stored away in some part of the egg or sperm. Nor is it good enough to say that the egg or sperm contains any thing which one might call a potency or capacity to develop into mind; what it does contain is a number of material structures (DNA genes) which specify rules for assembling amino acids into proteins. That is to say that they are programs or algorithms for carrying out some process which starts with a certain assemblage of raw materials (the amino acids) and leads to an end product (the protein) which can be defined in relatively simple terms. I think it is legitimate to say that each individual process controlled by a particular gene has an essentially algorithmic character. I do not feel quite so certain that this phrase is an adequate description of the global totality of all these part processes. In the first place all the many proteins that are produced go on interacting amongst themselves. In the second place another set of programs or algorithms put in an appearance and start operating, namely those that originate from factors in the environment. Thus the activities or characteristics of animals, or men, which might tempt us to say that they have minds are not at all straightforwardly simple readouts of DNA blueprints. They result at the end of long sequences of processes which start by operating the algorithms encoded in genes, but then go through the whole sequence of phases of interactions between these primary algorithms and between them and another set of external algorithms. I think that each individual step in this network of interactions has an algorithmic character, but I suspect that the overall result is of a kind which that phrase does not adequately suggest.
I have not yet been able to think out the implications of these ideas for John's argument derived from Gödel's theorem.
But the point I wanted to make is that in considering something like ‘the inheritance of mind’, I think Tony is justified in saying that we are talking about something with an algorithmic character, but I suggest we really need a theory of a different order of complexity than that with which we are used to dealing; something which is related to ordinary algorithm theory rather as population genetics is related to the genetics of individual matings.
I'll reply to the speakers if I may in reverse order. Of course, I agree with Wad's point that it gives one no scientific information whatever to be told that opium puts people to sleep because it has a dormitive power. It seems to me, however, to be a true though uninformative statement. The reason why it's worth trying to work out the philosophical relations between capacities, vehicles and their exercises, is not because philosophy alone is ever going to solve any scientific problems about the world. The aim of philosophy is rather to help to put the questions clearly, and I think Wad is absolutely right that when one is studying capacities, what one must do is to study their exercises and their vehicles, and that all the real progress is made by correlating particular exercises with particular vehicles. But unless one distinguishes at the outset between the three, then one doesn't get clear what are the things to be correlated, and I think the history of the study of the mind confirms this. I am grateful to Wad for confirming that behind the metaphors of the blueprint and the magnetic tape, there is indeed the firm thesis that the mechanism of inheritance works algorithmically; that seems to me to present a great difficulty for John's account of the human mind, and I look forward to hearing the defence of St Jerome's direct creation theory which he promises us. But I was surprised to hear John going rather further than allegiance to St Jerome, and allying himself with the Pythagoreans. The Pythagoreans, if I remember rightly, not only thought that the soul of John's grandam might inhabit a woodcock, they thought she might inhabit a bean, and for that reason they didn't eat beans in case they might be eating their grandmother. I must say that I find the thesis that a bean might be inhabited by the soul of my grandmother not just improbable, as John says, but absolutely meaningless.
If I did say that the possessor of a mind was a body, as John says I did, then I am perfectly happy to restate all the things I've said, about the relationship between the mind and its possessor, using the word person instead of the word body, as John would prefer. For me this would not make very much difference, because I think that human persons are bodies of a certain kind. I am a body of a certain kind with certain capacities which, I argued in my paper, constitute my mind. I think that John too is a body of the same kind, and so are we all.
Finally, I should like to turn to Christopher, and I'd very much like to correct any impression I gave that there wasn't a great deal to be learnt about the nature of the mind by the analogy between minds and computers. What I was complaining about, was that there seems to be an uncertainty, if not an inconsistency, in the way in which Christopher uses the analogy. He accused me of confusing hardware with software, and I agree that it is the relationship of the programmer to the program that he often uses as an analogy, rather than the relation of the programmer to the computer itself. If it is the relation of the programmer to the program in the computer, then my complaint is that he hasn't quite decided whether the mind is the program or the programmer.
A passage that illustrates his uncertainty, his indecision, rather better than what he said last week about programming himself with regard to decimal currency, is the end of ‘The Seat of the Soul’ where he says: ‘What kinds of thing do we really want to know about the brain? I suggest that what we would like is a detailed account, among other things, of the “software”. I mean what a computer scientist would mean: the logic of the master program which sees to it that the user's program is properly translated into machine code, and implemented according to his instructions.’ I don't see who the user can be here, if not the Cartesian soul. He goes on to say: ‘It's quite on the cards that there is a special program in charge of all our various subroutines, which must not conflict with each other if we are to behave in an integrated way. And possibly its instructions reside in quite a small part of the brain.’ The dialogue ends there, with the question: ‘The seat of the soul, in fact?’, but the philosophical parallel with Descartes seems at that point so close that one would not be surprised to see the seat identified as the pineal gland!