Last night, I criticised Longuet-Higgins' Central Dogma on the grounds that the concept of randomness was negative and equivocal. I said that to be random was to be inexplicable, and there were many different kinds of explanations. Tonight, I want to go a bit further in distinguishing and correlating two kinds of explanations.
You will remember that Longuet-Higgins said that in his youth he had been persuaded by his elders and betters to believe that there was an important distinction between ‘how’ and ‘why’ questions, and that the scientist could only answer ‘how’ but not ‘why’; but had since abandoned this belief, with some qualifications. Tonight, I, who am neither elder nor better than Longuet-Higgins, cannot ask him to accept my say-so; but I shall merely ask to be allowed to adduce some reasons which may persuade him that there are different types of explanations, and that these explanations need not be necessarily incompatible. Indeed, this is an obligation which I inherit from last year (The Nature of Mind, p. 3), when I said I was a dualist in believing that there were different types of reason, and tried to maintain, partly as a long-stop for Longuet-Higgins' own arguments against reductionism, a thesis that one sort of reason could not be reduced to another.
I now want to purge myself of a suggestion he made last year (p. 118), that if I said this I was thereby putting up ‘keep off’ notices round the phenomenon of mind, and saying that we just could not explain minds at all. I want to show how it is perfectly possible to have explanations of the mind, and indeed explanations of different sorts, both of the mind and of other things, without their necessarily getting in each other's way. That is my second object tonight. My third object is to legitimise Professor Waddington. That is to say, I want to show that the author of the Strategy of the Genes and many other works is perfectly entitled to use the personal language and the functional explanations which he naturally wants to use, and that biologists are entitled to follow their own instincts in giving the explanations of the type which seem to them to be the most rational and illuminating.
We heard yesterday that the minds which we now possess derive their origin from the practical needs of our remote ancestors. Let me distinguish two different things which our ancestors may have done and which we definitely do, which could be counted as explanations: the first paradigm of explanation—the one which we found most natural when we explain ourselves—is when we explain why we do whatever it is that we have done, or why we should do something we have in mind. These explanations of actions, given characteristically in the first or the second person, have certain important characteristics. The most important thing about them is that they should be rational and seen to be rational. One has got to be able to see why these features are relevant to this proposed course of action. The second point about them is that they are seldom conclusive. There is almost always room for further considerations. It is a matter of pro and con. We know in our own experience how often we are torn by wanting to go one way and wanting to go another way, and we have to decide between them, recognising that there are weighty considerations on either side of the case—we ponder, we deliberate; and it is because of this second point that we are able fairly easily to understand other people's minds. It is temptation rather than actual experience which widens the bounds of our understanding. I understand what it is to want to murder someone because, sometimes, although I never actually have, I have myself wanted to do away with a colleague or a pupil. I find it more difficult to understand the activities of some of the later Roman Emperors, but one can try. We draw a sharp distinction between those cases where we can enter into the mind of someone else and see why he does it, and those which seem to be entirely opaque, and we just have to take it for an established fact that this type of animal emits that sort of behaviour.
The other paradigm of explanation is where we are trying to explain how an event happens. For this very different ideal requirements are put forward. I don't have to see why arsenic should poison someone. If poisoning is what I want to do, all I need to know is that if I do this sort of thing, namely, give the rich uncle some arsenic, then the desired result is accomplished and my inheriting his money will ensue; that is why this paradigm has very different characteristic features. It is important that it should be universalisable in a much stronger sense than when we are just simply explaining our own actions. It is not important that the features we adduce should be either rational or seem to be rational or even apparently relevant. All that is required is that there should be some reasonably sure-fire connection between bringing about those things which are means and under my control, and the subsequent occurrence of the ends that I desire. This second paradigm I call the regularity paradigm. It was first put forward by Hume, and has become the standard explication of the concept of cause. That is, the official philosophy of science agrees entirely with what Longuet-Higgins said yesterday, and is based entirely on the regularity paradigm, although occasionally with slight qualifications.
Yesterday, Longuet-Higgins allowed that, even with a straightforward scientific explanation, some considerations of coherence and economy were to be taken into account, as well as the uniform and reliable regularity with which the laws of nature were instantiated. Even David Hume, the Arch-Priest of the regularity paradigm, in an unguarded moment pays homage to the requirement that causes should be in some respects rationally explicable, when he allows that it is one of the criteria of there being a causal connection that the cause and effect should be contiguous. But these qualifications apart, the official philosophy of science has standardised itself on the regularity paradigm; and this has generated a world view which is characteristic, but distorted. In this world view, we see how things happen merely by virtue of knowing certain general regularities, and having some initial conditions which in conjunction with those general regularities must produce the result we are trying to explain; and so when we ask why it is that we are here and that there are minds engaging in lots more mental behaviour than was happening, say, two million years ago, the official philosophy of science says that this is a question which cannot be answered. It is just mere chance, just the result of a random walk through a space of biological possibilities. We feel somewhat affronted at this. It is a question we want to ask, indeed we may not feel only affronted but somewhat shrivelled in the face of the apparent pointlessness and uncaringness of the universe, as the official philosophy of science portrays it. For our normal assessments of what is important, it is very relevant what states of mind people have; and if when we are trying to consider our own position we are told that this is totally irrelevant, then we feel, as it were, orphaned, shrivelled and lost.
The question I want to raise is how far this view of the universe depends entirely on our adopting one paradigm. To take a traditional metaphor, people often say that our philosophy and philosophical concepts provide the spectacles with which we look upon the world. I do not say that we have, as it were, two pairs of spectacles; that was Kant's solution. If you push it through, it produces a sort of philosophical schizophrenia; you sometimes see things from the standpoint of the scientist and sometimes from that of the more engaged agent. What I want to say, instead of this, is that we have bifocal spectacles, and that normally we shift quite easily and unconsciously from one focus to another, but when we try and look at the world as a whole, we have difficulty in integrating the images which the two lenses are presenting to us.
Consider the difficulties biologists feel about functional explanations. Biologists have divided loyalties. They are scientists; they stand in awe of their physical and chemical colleagues. They are afraid that if they were to give a teleological or functional, or even a psychological, explanation of why the nerves pass through a certain pathway, Christopher might be cross. But biologists are human too. They are men; and they feel the force of explanations which seem to explain why we have eyes. ‘Why do animals have eyes?’ ‘In order to see’; one would have to be very certain of one's convictions to believe that this is not an explanation one is entitled to put forward. Biologists like to put forward functional explanations, and they do so; they think these are important, not only for the reason which came out yesterday, that sometimes a functional explanation will give us the circumstances in which some more basic regularity explanation can be put forward, but for two other reasons. First, they think of these functional explanations as being somehow schemata which enable them to gather together a whole lot of features and make them into one coherent and intelligible whole; and secondly, they use them because they have a certain homoeostatic quality which is what they discover in the biological world around them. By ‘homoeostatic’ I mean that if circumstances were slightly different something happens in order to remedy the situation, even in the face of adventitious adverse changes of circumstance. For example, if I get rather heated in this lecture hall, then I sweat a bit more in order to keep my blood-temperature constant. That is to say, a functional explanation doesn't lie just simply on the surface; it is also an exercise in the hypothetical subjunctive mood; it suggests, not always very explicitly, the sort of thing that might happen, were circumstances to be different. To borrow a phrase of John Austin's, it is ‘constitutionally iffy’. And for these two reasons, functional explanations are going to be important in biology; to go back to the points that were being made yesterday, they set the scene in which other explanations, explanations according to the causal paradigm, can be sought for, and they make some sort of rather vague predictions at the periphery of what we know. Take the first point in the antithesis that Longuet-Higgins gave us yesterday, the ‘how’ and the ‘why’. The ‘why’ questions are important because the ‘why’ explanations, among other things, set the questions to which some scientists will be able to give the answers. To take another case besides the maintenance of body-temperature; if I am in a bright light, then my iris expands and the pupil contracts in order that the amount of light falling on the retina shall stay roughly the same. This is a functional explanation which is illuminating, and is also going to suggest a whole line of physiological research: ‘How is it that the increase in the stimulus on the retina will affect certain muscles which will cause the iris to expand?’
Functional explanations in this sense are going to gather together a whole lot of different physiological phenomena and put them into a coherent pattern. They enable us to see the point of the physiological process and also to spot topics of research the physiologist can prosecute in order to fit it all into a rational pattern of explanation. Thus far, then, the ‘how’ and the ‘why’ questions are on, as it were, two different levels, and are compatible. Where they differ is in the way they generalise. This is where the difference between the person who has in mind a regularity paradigm, whom I might call at the moment a mechanist, and a person who has a functional paradigm in mind, take different standpoints. If I look at a homoeostatic mechanism from a mechanist point of view, I shall expect to be able to explain, up to a point, how it works in purely mechanical terms, and then I shall stop. This is the way we understand a thermostat. We see how the negative feed-back mechanism works, and then we are finished. A biologist, however, is less limited in what he looks for, and often will expect a whole lot of other phenomena to come into play even outside the range of what he already knows. That is to say, when he gives a functional explanation he has at the back of his mind an expectation—which may not always be fulfilled—that even if circumstances are different in some unforeseen way, one way or another the desired goal will be achieved. He is issuing a large number of undated and partially blank cheques on future research; and he tends to express this by saying that Nature is very cunning, Nature with a capital N.
We have been considering, thus far, specific cases of the functioning of organisms, and I want now to illustrate the difference between the approaches with regard to the course of evolutionary development, in order to pick up yesterday's theme and to toss a catch to Waddington. I will take one of the examples from last year's Gifford Lectures where he described (The Nature of Mind, p. 128) how the horse developed the way it did, because its ancestors made a choice. They chose an environment, as much as being formed by their environment. They chose to start eating grass instead of shrubs, and to deal with predators not by turning and kicking them with their front legs like giraffes, nor by getting heavier and heavier like buffaloes, so as to be able to charge their predators out of the way, but by running; and this choice being made, a certain line of evolution is thereby going to be adopted. The horse starts by being about as big as a dog (Eohippus), and then grows longer and longer legs and so gradually evolves into the breed of horse we have now. Another example comes from Sir Alister Hardy's set of Gifford Lectures (The Living Stream, London, 1965, p. 170). Tits, as some of you are probably all too well aware, have learnt how to peck through the aluminium foil of milk bottles and help themselves to the creamy top. This has happened in our own time; it is a piece of evolution we can observe now. It is largely learnt; one tit imitates another, and Hardy argues that it gives a different complexion to our whole picture of evolution, for it is now apparent how enterprise on the part of one or two birds is copied by their friends and rivals, and produces a new lot of behaviour, and results in a new set of environmental conditions being important and a new set of evolutionary pressures operating. If you are a tit and are going to peck bottle tops, then you have got to be able to recognise aluminium foil, your beak has got to be toughened up to get through it; if, as is conceivable, the dairyman retaliates by putting on thicker and thicker foil, then the advantage will go to the tits who have stronger and stronger beaks, and perhaps soon they will be able even to get through the plastic caps that some milkmen put on milk bottles.
I am not competent to talk about the biological moral of this, but I want to draw a philosophical one. Neither Waddington nor Sir Alister Hardy are anti-Darwinians. On the contrary, they are ultra-Darwinians. What they are doing is what very often happens in the natural sciences; they are making a reverse take-over bid for the truths that Lamarck had discovered, to incorporate them into Darwinian theory. They are showing how, if we look at the course of evolution as a whole, we keep on seeing new ways which we hadn't anticipated and which don't really lie within the compass of what we should have expected, if we believed only the Central Dogma we heard yesterday. One way or another, organisms develop new ways of doing things which are going to be good for them.
To parallel this with another sort of explanation which might be given, I said ‘why do we have eyes?’, and answered ‘to see with’. It is a problem which biologists often face. Why have very different organisms who haven't inherited their eyes from a common ancestor developed very comparable organs? The sort of answer that I like to give is to say that the earth's atmosphere is opaque to all forms of electro-magnetic radiation except for the visible spectrum and for wireless waves. Wireless waves are no good for direction-finding or for detecting things at a distance, because they are too long; they just would be confusing. Therefore, if you are going to do anything of this sort, you are going to need to have eyes which will respond to the visible spectrum. This much granted, then the rest of the story follows. We can understand the different ways insects and mammals have developed their eyes, but it follows not so much from the chances of evolutionary development but from the over-arching conditions—the way in which the pinboard (p. 10) is set up.
It is always possible to reject this sort of explanation altogether. One reason why it should not be rejected is that biologists have tried to reject it and failed. Biologists, you see, are not only human, but being human have a strong sense of sin. They feel they want to use functional explanations, therefore they feel they ought not to use them; and have often feared that if they used any such explanation they would be thought to be vitalists or accused of worshipping entelechies or something of the sort. Nevertheless, they keep on coming back again and again, offering functional explanations, with a long stream of apologetic talk pleading forgiveness for doing it, but still doing it. This seems one reason why we should accept these explanations, not only as being obviously illuminating, but as being in some sense necessary. But I can see that a person could remain blind to these blandishments. It is always possible, just possible, logically possible, to refuse to allow this type of explanation at all. It is the opposite case to that of the vitalists who refused to allow that there were any casual explanations of certain particular parts of biological behaviour, and just simply said that it was due to an entelechy. In each case, the position is tenable, logically tenable, but only just; and I should say that just as the evidence of the development of physiology provides a powerful argument against the vitalists, so the evidence of ethology provides a powerful argument against those who want to reject utterly the use of rational explanations in biology. This brings us back to where we were yesterday, and allows us to raise the question about the whole process of evolution, which seems to have its own rationale. Constantly there seem to be developing more and more complicated organisms, organisms which are in one way or another independent of their environment, organisms which are homeostatic and come to be autonomous. We want not merely to raise the question in terms of the consequences of some vast Schrödinger equation which represents the whole earth irradiated by the sun, but to ask ‘why does this happen?’ We are impelled to ask the question ‘why?’, and if we are told that this is a question which cannot be asked within the limitations of the regularity paradigm, we are entitled to conclude from this that it is showing us the limitations of the regularity paradigm rather than that the question, or any answer that we might give, is necessarily inappropriate.
This is what we ask. But in asking this question, we needn't alarm the scientists, who are concerned, quite properly, with answering ‘how’ questions; for, as I've tried to show, the fact that ‘why’ questions can be asked and may perhaps be answered, does not in the least bit show that the ‘how’ questions are in any way to be ruled out of court or to be regarded as redundant.
I do not want to question the main points made by John Lucas. After all, he said he wanted to legitimise me and the kind of arguments I shall be putting forward in my lecture. At my age, being handed a certificate of virginal purity is an acceptable if unexpected experience, even if it only applies in the domain of higher philosophical thought. And, of course, I quite agree with John that it is entirely legitimate to answer ‘why’ questions about living organisms by referring to their goals or ends. This is so because the theory of natural selection allows us to see how organisms can be moulded or programmed to act in ways which would tend to bring about a certain state of affairs at a later time, that is to say to reach a certain goal. Under these circumstances the concept of a goal escapes the objections which makes the Aristotelian idea of teleology or finalism unacceptable. Many years ago I suggested that, within the realm of things subject to natural selection, the idea of a goal is only ‘quasi-finalistic’; more recently it has become usual to refer to it, in a neater but perhaps less easily interpretable way, as ‘teleonomic’ rather than teleological.
I should like to raise the question whether we should not accept that there is a third type of causation operative in biology, which we might refer to as the historical, or, as I should prefer to call it, the epigenetic. Christopher referred to this briefly in his talk yesterday, when he mentioned the old biological slogan that ontogeny recapitulates phylogeny; translated into plain language, this means that the development of an individual repeats, in some form or other, evolutionary changes that have taken place in the ancestry of that individual. It implies that the cause for the appearance of some organ or some process during the development of the present-day individual, is to be found in the occurrence of that organ or process in its ancestors.
For the benefit of the non-biologists perhaps I should give a simple example. During the evolutionary history of the vertebrates, from the earliest primitive fish to the most recently evolved mammals such as ourselves, three types of kidney have been developed. The primitive fish operated only with the first type of kidney, and this was the organ which filtered waste materials out of their blood stream into urine, which could be excreted to the outside world. By the time evolution had proceeded far enough for the appearance of birds, this function of purifying the blood had been taken over by a second type of kidney, which is presumably more efficient than the first. We ourselves run mainly on a still later, third type of kidney. However, I want to mention only what happens in the birds.
Although by the time a young bird has hatched from the shell and starts excreting its waste products, it is operating by means of second-type kidneys, at a much earlier stage during its development the first-type kidneys put in an appearance. Why should this be so? Well, it can be shown that although these first-type kidneys never actually filter anything out of the blood stream, they do nevertheless fulfil a useful function during development. From that kidney there grows out a tube or duct, which in the fish is used to conduct the urine from the kidney to the exterior. In the embryo of the bird this duct acts as a stimulus or inducer which produces the formation of the second-type kidney in its neighbourhood. Now, it is a very common feature of embryonic development that when some organ needs to be produced in a definite location, some neighbouring structure which is already in the right place is used as a stimulus to induce it. So the bird embryo probably needs something to induce its second-type kidneys to develop in the right region. But we know that quite a lot of things could do so; for instance parts of the spinal cord which are not too far away. We have to conclude that the reason why the bird embryo uses the first kidney duct to perform this function, is just because its ancestors had developed first kidneys and ducts in order to carry out the necessary process of purifying their blood.
Here we seem to have three types of causation involved. The bird embryo forms a kidney for what John calls a ‘rational’ reason—in order to attain the goal of keeping poisons out of its blood—which has been built into it by natural selection. Then the duct of the first kidney plays a straightforward role in ‘regularity’ causality, by acting as the cause for the appearance of the second-type kidneys. But the reason why it is this duct and not something else which acts as that cause, requires a different type of explanation; it does so because the bird embryo is an evolutionary modification of something in which a duct of that kind happened to be present, all ready to be pressed into service in this way.
Questions of this third kind only arise when there is some over-riding necessity which prescribes a goal to be reached, and to that extent they are perhaps nearer to John's second, rationality paradigm, than to the first, regularity, one. But I think they are different enough to require a category of their own. The sort of problem they raise reminds me of the story of the man who drove his car into Buccleuch Place, then stopped and asked a local inhabitant how to get to Tollcross. After a bit of thought about the various one-way traffic systems, and so on, in the way, the answer came, ‘Well now, if I was wanting to go to Tollcross, I would'na be starting from here’.
I should now like to turn to a second comment. John has spoken of his regularity and rationality paradigms as quite distinct and separate. Are they really as distinct as all that? I tend to think that they are distinctions of emphasis rather than of kind. I believe that one cannot operate wholly within the regularity paradigm of cause and effect without having used the rationality paradigm to get into it in the first place, and vice versa.
My basic reason for thinking in this way is that we are concerned wholly with the realm of human knowledge and understanding. I do not believe that we have any access to a totally objective world which is completely independent of us. Nothing can enter into our knowledge—into the arena of our discourse—which is independent of our nature, since it is through our nature—our sense organs and general perceptive apparatus—that we become aware of it in the first place.
I know that it was convention, at least until recently, to suppose that the methods of science, following a regularity paradigm, revealed to us the objective truth; but in recent years we have seen so many radical changes of view, standing the previously accepted objective truth on its head, that I think this can scarcely be upheld any more. I suppose the classical case when science was supposed to have reached the objective truth was the discovery of Newton's Laws of Motion. You may remember the confident couplet written by Pope as an epitaph to Newton:
‘Nature and nature's laws lay hid in night;
God said: “Let Newton be!” and all was light.’
But then a couple of centuries later there came the amendment by J. C. Squire:
‘It did not last, the Devil, howling “Ho!
Let Einstein be!” restored the status quo.’
In order to operate within the regularity paradigm we have to use defined concepts. What I should argue is that the concepts we choose to define are profoundly influenced by the ways in which rationality principles, dependent ultimately on our telenomic goals, have led us to envisage or conceptualise the world. Something can only enter into our knowledge when we can recognise it as having a definite character. That means, in effect, perceiving that its character is near enough to some standard for it to be assimilated into that standard, and allowed to take its name. This is a process of the same kind as the assimilation of an activity under the heading of a goal.
Well, I very largely agree with that. I certainly want to say that there are more than just the two paradigms that I brought forward. After all, Aristotle said there were four causes (or becauses), and surely by now we can do better than that. To take another example, the whole range of mathematical explanation is different from either of the two that I have mentioned.
I'm not so happy with the kidneys being so chreodic. It seems to me that this sort of epigenetic explanation is somewhat comparable to a great deal of what we have in history. Often the skill required is not in looking for further and more profound paradigms of explanation, but rather in fitting together various bits; it is a jigsaw exercise. ‘Well, there's that tube there, and what would have been here? Could there have been a membrane there? No, the spinal column would have been in a different position then.’ This seems to me to be an important sort of explanation.
Explanations often satisfy more than one paradigm. Kant and Professor Hare lay great stress on universalisability as a mark of rational explanation: if I explain why I do something or why you should do something, then there is some idea that this explanation should also hold good in all other cases. To this extent, the rationality paradigm is like the regularity paradigm. Or to go the other way, I entirely agree with Wad that even if we are giving a scientific explanation, regularity by itself isn't enough; there must be some sort of rational appeal too. I agree on both these points; I only want to pick out different facets which we need to emphasise in different ways. They are potentially incompatible, and if we try and generalise too far, then we get difficulties. As long as we remain practical men, we manage to use our bifocal spectacles quite easily; but when we try and survey the universe as a whole, unless we are rather careful, we shall be caught out by these two ways of emphasising in mutually incompatible directions.
I agree with most of Wad's comments on John's talk. John said a number of things which I certainly would agree with, but he attributed to me a number of opinions which I wouldn't want to express, in particular, the suggestion that the scientist, and especially the biologist, should reject or try to avoid using functional explanations. May I say how I see the relations between these different explanations of why we are standing here talking, using these funny things we call minds? One would want to say that our mental activity rests, among other things, upon the fact that we have brains of a particular structure. Our brains have this structure because our genes are the way they are, the message in the blueprint says ‘build it that way’. But the question now arises: ‘Why is the blueprint written that way?’, and here we come to what Wad calls the epigenetic explanation, in terms of the way things have evolved historically. Now the question: ‘Why did they evolve historically that way?’, can be answered up to a point, but only up to a point; once we have decided, as it were, to try and outwit our predators and competitors, then our genetic pool will naturally drift in the direction of programs which enable us to do that better. That's what I meant yesterday by saying that all organisms tend to get better and better at what they're already good at; and so far I'm more or less in line with John Lucas.
The point at which I would possibly depart from him would be in answering the question: ‘Supposing we were to go back to the year Dot, could we have foreseen any of this? Would it have been possible for anybody to say, “Aha, this must happen!”?’ Here I have to declare a complete agnosticism, because there would have been nothing to help us at all in the properties of atoms and molecules as such, and as there were no organisms around at that time there would have been nothing for them to be good at or to get better at; we would be reduced to saying ‘It is going to be one gigantic lottery’. From that point onwards, of course, as the lottery turns out, goals are developed and become selected for, and the thing acquires more and more meaning and momentum as it goes on. And now perhaps the greatest component of our evolutionary momentum is our own plans, in the ordinary sense of the word ‘plan’. In short, I wouldn't dream of suggesting that the humble ‘how’ questions are the only interesting ones; but they must be attended to if we are to understand not only the possibilities but also the limitiations of biological evolution.
With regard to teleology, I'd like to adopt a position somewhere between John's and Wad's. It seems to me that Wad doesn't allow quite enough scope to teleological explanation and John allows rather too much.
Wad says that it is all right nowadays to use teleological explanations because Darwin has shown us how to do so without lapsing into Aristotelian entelechies. I think that one has to distinguish between teleological explanations of the activities of organisms and teleological explanations of the existence of organisms. To take John's example of sweating, there is undoubtedly a teleological explanation of why I am sweating: it is in order to regulate the temperature of my body. Darwin comes in to point out that sweaty animals of a certain kind do better in the struggle for life than non-sweaty animals of the same kind, and that explains how there happen to be lots of sweaty such animals around and not many non-sweaty such animals. But that is not directly connected with why I am here and now sweating. That is to say, the Darwinian explanation is not in direct competition with the Aristotelian explanation. Aristotle, of course, had no explanation of the origin of species since he thought all species had been eternally there. The Aristotelian type of explanation was a teleological explanation of the activity of organisms, not of their origin, or the origin of the species to which they belonged.
However, I want to take up a position which is much closer to Wad's than it is to John's, because I think that when John says that there may be a teleological explanation of the whole course of evolution, he allows teleological explanation a greater scope than he has justified.
The teleological explanations which we can give are given on two different bases. Many of them, including all those which we give for the activities of non-conscious agents, are based on regularities. Therefore, the contrast John drew between regularity-type explanation and teleological explanation was already misleading. For instance, it is not enough that an animal or a bird should do something which is beneficial to it, for us to accept that the animal did it in order to achieve that beneficial goal. Suppose that some migrant bird, at the time of year when it is due to migrate southwards, accidentally flies through the porthole of the Queen Elizabeth as it is about to take off for a Caribbean cruise. This may have an effect on it which is beneficial to the organism, indeed, it may achieve a goal which we know to be one of the goals that this organism has; but not even John, I think, is going to say that the bird flew in through the porthole for this purpose. When we do give teleological explanations of the migrations of birds, it is because of the regularity of these migrations.
Now, of course, in teleological explanations of the actions of human beings, we don't in the same way depend on regularities. Still thinking of movements on board ship, let us change from a migrant bird to the Emperor Nero. The Emperor Nero, you may remember, brought his mother, Agrippina, on board a ship in the hope that she would drown. He was getting rather bored with Agrippina, he had had the ship cunningly constructed so that it would founder, and he not unreasonably expected that she would drown though, being a strong swimmer like Bertrand Russell in a later age, she swam to shore. Now, the reason why we know that Nero put his mother on board ship in order to drown her, is not that there is any law of nature that whenever human beings put their mothers on ships this is always what it is for. On the contrary, so far as I know, it was a unique occurrence. The reason we know that this was why Nero did it, was that he was rash enough to tell some of his friends who let it out to one of those indiscreet writers, Suetonius or Tacitus.
So this is the second way in which we discover teleological explanations: through articulate communication and language. The reason for introducing these examples is to bring out that it isn't at all clear that either of these methods of discovering teleological explanations can be applied to the course of history of the universe as a whole. We haven't any laws of nature which say that ninety-nine per cent of worlds like ours sooner or later evolve human beings, and we don't claim—or at least I don't claim, I can't speak for John—to be in verbal communication with anyone who can tell us what the plan of the universe is.
But I was careful not to stress teleological explanation. Functional explanations are often classified as a type of teleological explanation and teleological explanations are indeed, a type of rational explanation. Aristotle thought that they were the paradigm type but he was wrong. This, of course, doesn't dispose of Tony's point, but I suspect that when he heard me say teleological he thought that I was getting ready to rehabilitate the Argument from Design. Well, I was only getting ready. His point would be fair if I was trying to bring forward a full teleological argument; I would then have to go a great deal further, the biological argument wouldn't be enough. Kant, when he produces his critique of teleological judgement, rests it ultimately not on any crucial piece of evidence, but upon the existence of moral consciousness, on the fact that men know that they can make moral choices; and I don't suppose that I shall be able to get by with fewer premises than Kant.
I want also to turn to Christopher, and express some sorrow that in the course of the last twenty-four and a half hours, he has become more reductionist than he was yesterday. He was making the point that I can't give a detailed prediction of the epigenesis of the kidneys, or of the whole universe; this is an important point to which I shall return. But yesterday, although I allowed that we couldn't make any detailed predictions or indeed any predictions—except that evolution was pretty well bound to produce something interesting—I said that there were two different types of explanation and these answer different questions. I want to qualify this now by making two further points; that each type of explanation uses its own concepts and will phrase the question in its own terms which are not always the terms that some other type of explanation would use, and that each type of explanation is not going to be able to give us all the answers. Thus to take up the exact point that Christopher was making against me, even if I produced lots and lots of paradigms of explanations, it doesn't mean that there aren't going to be unanswerable questions. There are, and I can't give an explanation, in the terms which he poses, of why some particular sequence of events, as described by him, should happen; for instance, of why we should be speaking here. This is something which couldn't be explained in terms of any sort of regularity paradigm, nor could it have been predicted at the beginning of time. It is of crucial importance as we look at explanations to see what are the terms in which they are posed. When Christopher made his one prediction, he used a quite different set of concepts from those which you would expect a person to use if he was concerned solely with the regularity type of explanation. ‘Something interesting’, he said, ‘would happen.’ I would like to end by saying this: we can't expect there to be any sort of explanation of all the features which we see around us; but we can hope for some sort of rational explanation of why something interesting should be going on.