The Fifth Way

The Fifth Way, as often remarked, depends on final causality, explanation in terms of an end, of what is the point of things. The aim is to show that the world, like its parts, requires an explanation in terms of what itís for, and what the world is for is set by God. It is a mistake to say that this argument is based on the worldís being for a purpose: it is rather based on the worldís having a point, which is at first sight quite another thing. The circulation of my blood has a point, but it does not have a purpose: nor do even I have a purpose in circulating my blood, since circulating my blood is not something in the strict sense which I do. I canít help it. It is true that in order to reach God Aquinas has to argue that the point of the world has to be a purpose set for it by God, but that is a further step in the argument. We can see this clearly from the text:

"The fifth way is taken from thingsí being directed.

We see that there are things that have no knowledge, like physical bodies, but which act for the sake of an end.

This is clear in that they always, or for the most part, act in the same way, and achieve what is best. This shows that they reach their end not by chance but in virtue of some tendency.

But things which have no knowledge do not have a tendency to an end unless they are directed by something that does have knowledge and understanding. An example is an arrow directed by an archer. Therefore there is some being with understanding which directs all things to their end, and this, we say, is God." This is the briefest of all the Five Ways, and its structure is apparently clear.

1. We see that there are things that have no knowledge, like physical bodies, but which act for the sake of an end.

2. But things which have no knowledge do not have a tendency to an end unless they are directed by something that does have knowledge and understanding.

3. Therefore there is some being with understanding which directs all things to their end, and this, we say, is God.

But this simplicity is deceptive, and the history of natural theology and apologetics has made it yet more difficult for us to read this argument as we should. Most people would want nowadays to object to the first premiss, and to deny that we see that physical bodies act for the sake of an end. But they would grant that if we saw physical bodies acting for the sake of an end, we should conclude immediately to the existence of God. Aquinas does not agree, I think. He thinks that it is almost unquestionable that e.g. physical bodies act for the sake of an end: what is tricky is the step from the unconscious end-directedness which we see all around us to the conscious end-directedness which he needs to assert if he is to prove the existence of God.

One of the things that has happened between Aquinas and ourselves has been the growth of a general disbelief in explanation in terms of what things are for. This is partly the result of a failure to understand what it is to explain something in terms of what itís for, and partly the result of the rather curious psychological phenomenon of the near-universal acceptance of what is really a rather poor argument for the existence of God, the argument from design.

The argument from design had its heyday between the time of Newton and the time of Darwin, say, a time in which most people apparently came to see the world as a minutely designed piece of craftsmanship, like a clock. It is no coincidence that the most famous presentation of the argument from design actually compares the world to a clock: it is known by the name of Paleyís watch. It is also worth noticing that according to the great computerised Index Thomisticus, in the 8,000,000 words Thomas Aquinas definitely wrote, and the 3,000,000 he may have written besides, the universe is never compared to a clock.

The argument from design takes as its basis the perceived mis-match between the detailed dovetailing of the different parts of the mechanism of the world and any story we could tell about how this comes to be as the result of blind chance. A favourite example of Paleyís is the eye: how could the eye be so perfectly adapted to its function if it came about merely by chance? It must have been designed.

It is worth noticing that this is not so much an argument about final causality, an argument in terms of what things are for, as an argument about efficient causality, an argument in terms of how things have come to be the way they are. This is because the notion of teleology, of thingsí being for a point, which is used in the argument from design is indeed a notion of purpose, of conscious teleology. Unconscious teleology or final causality, as we shall see, does not conflict with efficient causality: indeed, a decent explanation of what things are for actually demands a decent explanation of how they come to do what they do. Conscious teleology, or purpose, or design, is different. Conscious teleology supplants any alternative explanation in terms of efficient causality and supplies its own. An account in terms of conscious teleology, purpose or design says "Some mind conceived the idea of the end, and set about to arrange things so as to bring it about". When there is a designing mind involved, end or point becomes purpose, and as it were gets in behind the chain of efficient causality and sets it going.

There is no doubt that this argument is or was immensely attractive. Reid regards the human tendency to infer from detailed arrangement some designing mind as one of the principles of common sense; and even Hume regards the argument from design as the chief reason for believing in God.. But the argument from design has two enormous drawbacks.

The first is that it is, roughly speaking, an argument from ignorance, and is thus immensely susceptible to advances in science. Paley could not see how the eye could have come to be without a Designer; Darwin comes along and does no more than sketch a sort of answer of how things like eyes might more or less have come to be, and no-one finds it possible to take Paley seriously any more. This is not good philosophy, but it is a pretty clear picture of the fate of eighteenth-century natural theology in the early part of the nineteenth century.

The second objection to the argument from design is that it does not get us to God, but only to a Designer, a Demiurge, as Plato would say; or, as the eighteenth century loved to say, the Great Architect. The Being whose existence is revealed to us by the argument from design is not God but the Great Architect of the Deists and Freemasons, an impostor disguised as God, a stern, kindly, and immensely clever old English gentleman, equipped with apron, trowel, square and compasses. Blake has a famous picture of this figure to be seen on the walls of a thousand student bedrooms during the nineteen-seventies: the strong wind which is apparently blowing in the picture has blown away the apron, trowel and set-square but left him his beard and compasses. Ironies of history have meant that this picture of Blakeís is often taken to be a picture of God the Creator, while in fact Blake drew it as a picture of Urizen, a being who shares some of the attributes of the Great Architect and some of those of Satan.

The Great Architect is not God because he is just someone like us but a lot older, cleverer and more skilful. He decides what he wants to do and therefore sets about doing the things he needs to do to achieve it. God is not like that. As Hobbes memorably said, "God hath no ends": there is nothing that God is up to, nothing he needs to get done, nothing he needs to do to get things done. In no less lapidary Latin, Aquinas said "Vult ergo Deus hoc esse propter hoc; sed non propter hoc vult hoc". In definitely unlapidary English we could say: The set-up, A-for-the-sake-of-B is something that God wants; but it is not that God wants B and for that reason wants A. We know that the set-up A-for-the-sake-of-B is something that God wants, because it is something that exists, and everything that exists, exists because of Godís will. But it is simply profane to think that you can infer from that the unfathomable secrets of the inside of Godís mind and will. Acorns for the sake of oak trees, to repeat an example of Geachís, are definitely something that God wants, since that is the way things are. But it is not that God has any special desire for oak trees (as the Great Architect might), and for that reason finds himself obliged to fiddle about with acorns. If God wants oak-trees, he can have them, zap! You want oak trees, you got íem. "Let there be oak trees", by inference, is one of the things said on the third day of creation, and oak trees are made. There is no suggestion that acorns have to come first: indeed, the suggestion is quite the other way around. To "which came first, the acorn or the oak?" it looks as if the answer is quite definitely "the oak". In any case, whatís so special about oak trees that God should have to fiddle around with acorns to make them? God is mysterious: the whole objection to the great architect is that we know him all too well, since he is one of us. Whatver God is, God is not one of us: a sobering thought for those who use "one of us" as their highest term of approbation.

The argument from design fails, then, because it is an argument from ignorance, because it confuses the final and efficient modes of explanation, and because even if it succeeded it would not prove the existence of God but of some Masonic impostor. But like other bad arguments, its defeat and death has left it to wander the world like a ghost, oppressing the spirits of those who are looking for other and better arguments. The haunting we suffer from when we attempt to take a proper look at the Fifth Way means that people think that as soon as we admit any kind of teleology into the world, any kind of explanation in terms of what things are for, we are at once committed to God.

I have a heavy burden of prejudice to overcome here. Let me just say that as far as I can make out by far the hardest move in the Fifth Way is the move from unconscious to conscious teleology, and, for my part, I am not sure that St Thomas gives us conclusive reasons for making that move. I regard them as strong, but not conclusive, and the impression I have is that others regard them as not very strong at all. So bear with me: the case I am going to make for explanation in terms of what things are for is worth examining in its own right, and I give you my word that there is no booby-trap: you will not find yourself suddenly committed to believing in God merely because you have come to accept that at least some things in the world need to be explained in terms of what theyíre for.

"Having a point" or "being for something" or "displaying a tendency", or, as Aquinas would probably say, "directedness", is the feature X in this Way. I think that it is, as Aquinas says, perfectly obvious that some things in the world display this feature X, and it is clearly a feature X that requires an explanation in terms of a relation to something else: viz, their point, what they are for, what they have a tendency towards or what they are directed to.

"We see that there are things that have no knowledge, like physical bodies, but which act for the sake of an end. This is clear in that they always, or for the most part, act in the same way, and achieve what is best. This shows that they reach their end not by chance but in virtue of some tendency."

Aquinas clearly thinks that this feature X is generalisable: that if any part of the world displays this feature X, the world as a whole displays it. This step I find more doubtful, though I do find it plausible to infer that if any part of the world apparently displays this feature X, has a point, then it at least makes sense to ask whether the world as a whole displays feature X.

I shall not in fact stop with the obvious facts to which Aquinas alludes, but shall go on to argue that in fact every bit of the world displays feature X, at least to the same extent as that to which every bit of the world displays the feature X we looked at in the Second Way, that of being a subject of efficient causality. This move I shall make principally for its own sake, because it is philosophically important and deeply unfashionable. But this move also helps the Fifth Way a little: if every bit of the world displays this feature X it becomes a good deal more urgent and at least apparently a good deal more sensible to ask whether the world as a whole displays it; and, by parity of reasoning with the lumping-together move we performed in the Second Way, it even begins to look likely that the world as a whole does display feature X. But we still will not have reached God: there will be a further step, which Aquinas scarcely even sketches for us, which will say that if the world as a whole does display feature X, the way in which it displays this feature cannot be an unconscious way. In this respect the world as a whole is unlike most of the bits of the world which display feature X, since they do display feature X in an unconscious way. The point in terms of a relation to which the world is to be explained is thus going to be revealed as being indeed a purpose, a point conceived by a mind to which the world is directed by that mind. But it is this last step that brings us to God, and it is the most contentious step of all.

To an unprejudiced mind there are innumerable features of the world which display the relevant feature X, that have a point. All the different parts of an organic whole ó a human body, for example ó have a point in relation to the life of the whole body. Acorns we have already mentioned; and indeed all the paraphernalia of animal and plant reproduction actually resist definition except in terms of what they are for. Imagine, for example, the task of identifying the sexual organs of some animals that were very different from us, animals found on Mars, for example. You cannot hope to identify the sexual organs in terms of three-dimensional geometry, of what bits stick out and what bits stick in: you have to identify the function. People who study the ecology tend more and more to describe it in teleological terms: what else is meant by the metaphor of calling the whole ecosystem of the earth by a proper name, "Gaia", as if it were itself an organism?

These biological examples are particularly useful in that they show us that we are not here depending on ignorance. The way the sexual organs work is well understood. We even known quite a lot about how the ecosystem works, and it is significant that we have come to speak of the ecosystem in more and more organic terms ó that is, more and more in terms of what things are for ó the better we have come to understand how it actually works. Thus teleological explanations, explanations in terms of what things are for, are not in the least rivals to efficient explanations, explanations in terms of what things do or of how they do it. The idea that final causality and efficient causality are in some sense at loggerheads derives from arguments like the argument from design, where efficient explanations were lacking. Because efficient explanations were lacking, the obvious final explanation was postulated as pre-existing as an idea in someoneís mind, which would then provide the required efficient explanation. The mind of the Great Architect is a cover for ignorance of efficient causality: the existence of final schemes of explanation may have suggested that cover, but the reason why it suggested the cover is that a decent teleological explanation, far from usurping the place of a decent efficient explanation, actually demands one.

Unlike Paleyís watch, Geachís watch is worthy of a place in contemporary and non-historical works of philosophy. A good old-fashioned mechanical watch, like many other artefacts, is a paradigm of something that can and must be explained in terms of efficient causality. The minute hand rotates twenty-four times in each period of the earthís rotation, because it is moved by cog A; the hour hand rotates twice in every period of the earthís rotation, because it is moved by cog B; cog A and cog B move at such-and such a rate because they are moved by cog C via gearing D; .... because it is moved by the escapement, which moves because it is moved by the hairspring, which moves because it is moved by the mainspring. So far so good. If I had had a story like that to tell about the world my version of the Second Way would have been far more elegant.

But the same watch, like many other artefacts, is also a paradigm of something that can and must be explained in terms of final causality. The mainspring moves in order that the hairspring may move which moves in order that the escapement may move which moves .... in order that cog C via gearing D may move cog A and cog B, and cog A moves in order that the minute hand may rotate twenty-four times in each period of the earthís rotation, and cog B moves in order that the hour hand may rotate twice in every period of the earthĎs rotation. This story, in terms of final causality, of what is the point of what, of what things are for, is just as good as the last story, which was in terms of efficient causality, of how things come about. Indeed, it does not take much philosophical acumen to realise that the one story has to be as good as the other, since they are the same story read in opposite directions with suitable adjustments to the connectives. Neither is prior, neither makes the other redundant. Sometimes, as in the case of the sexual organs, we start by telling the final story and get round to developing the efficient story, and sometimes, as in the case of, I believe, the pineal gland, we first tell the efficient story and then get round to telling the final story ó working out what itís for.

So: prejudices aside, we can ask what this or that bit of the world is for; and we can, in some cases, get an answer. We get it, at least sometimes, by considering efficient causality: what this or that bit of the world actually does. There is, thus, no conflict between final explanations and efficient explanations, and it might be possible to maintain, as in the cases we have considered, that the one requires the other and vice-versa. They require each other, they even live off each other, given that the final explanation is just the efficient explanation read in the opposite direction, with the "becauses" substituted by "in order tos", while the efficient explanation is just the final explanation read off in the other direction, with the "in order tos" substituted by "becauses". This is certainly Aquinasís doctrine: as he puts it, rather more elegantly, "causae sunt ad invicem causae", one kind of explanation explains the other.

Aquinas thinks that the world as a whole is a system of efficient causes, as we saw when considering the Second Way. This is not the sort of claim that many people, except possibly neo-Pythagoreans, would bother to question. But a corollary of this belief, combined with this account of the relationship between efficient and final modes of explanation, is that the world as a whole is a system of final causality; and this claim is unlikely to be even considered by the majority of our contemporaries. It implies, for example, that of almost anything ó certainly of anything that forms a part of a system of efficient causality ó we can sensibly ask "Whatís it for? What is the point of it?", and can hope one day to be able to give an answer which is more or less correct.

Pick up any book of the history of the philosophy of science and you will find this sort of view guyed unmercifully. Aristotle ó and, by implication, St Thomas, though few of those who work on the history of the philosophy of science even bother to look at St Thomas, confining themselves to remarks about angels on the points of pins ó is represented as a sort of primitive animist. Stones fall to the ground, Aristotle says, because they have a tendency to return to their proper place, which is the centre of the earth. Add to this the fact that the word he uses for "tendency" is "orexis", "appetitus" in Latin, which can equally well be translated as appetite, desire, or even lust, and you have a rather charming but deeply ridiculous picture. Things like stones can be represented as having a little and rather simple soul within them, a soul which only knows where its rightful place is an only wants to return there. Remove constraints from the stone and this passionate desire takes over, and it flies to the bosom of the earth panting with lust. Why this picture, which I find rather exciting, should be considered as ridiculous in the century of Freud, I am not quite sure. I would have thought that a picture of the Universe as brimming with quasi-sexual desire would be rather attractive to our contemporaries. Perhaps this tells us something about the way in which philosophy has become detached from the concerns and interests of ordinary people in the last couple of hundred years.

The picture just drawn of Aristotleís world-view, though worthy of serious poetical consideration, is in fact no more than a parody, aimed at making the notion of final causality ridiculous. But the reality of the doctrines of Aristotle and St Thomas, someone might want to say, is already ridiculous enough. The idea that we can seriously ask of almost any bit of the universe, at least in so far as it forms a part of a system of efficient causality, "Whatís the point of it?" is an idea which has been seriously out of fashion for about the last three hundred years.

Well, many ideas which have been out of fashion for the last three hundred years are quite good ones. About three hundred years ago it became unfashionable for preachers to tell their flocks that the goods they happened to have in their power over and above what they needed for their own support and that of their families, and for the discharge of any civic duties they might have, did not belong to them at all, but belonged to the poor, to those who needed it, and was owed to them as a debt of justice. Someone who had more than he needed and did not give to the poor was not guilty of a lack of charity or philanthropy but was simply a thief. That idea became unfashionable three hundred years ago, but it seem to me a pretty good one. The same might well be true of the notion of final causality.

Unsurprisingly, I should like to approach this question of final causality through the question of efficient causality. We have already mentioned that the standard modern account of causality regards it as a relation between two events. This is entirely false to the practice of science, which deals with relations not between events but between things. As was perhaps mentioned when discussing the notion of the per accidens, and the young Russellís pretence that people and things were logical constructions out of events, making events prior to things is not so much to put the cart before the horse as to put the mule before the horse and the donkey. Events cannot be identified except as in terms of their relations to things. The Battle of Waterloo, an event which we quite frequently refer to, at least if we live to the south-west of London, can only be identified in terms of Napoleon and a lot of his chaps meeting Wellington and (rather later) Blücher and a lot of their chaps, on the slope of a hill about twenty miles south of Brussels and spending most of the day beating all hell out of each other. There is no chance of our identifying Wellington dependently, in terms of his being the victor of Waterloo. The death of Socrates is the death that it is because it is Socratesís death: Socrates is not who he is because he is the person who died that death. Which death? we ask, and the only answer that we can give which makes no reference to Socrates is that it was the death that occurred early in the morning of (say) the fourth day of the month Boedromion, in the prison of the Eleven, in the archon year of (say) Agathocles. But this identification of the event depends on the prior identification of all kinds of things and persons: the sun, to give us the hour, the moon, to give us the day of the month, Athens, whose officers the Eleven were, and where the prison was situated, and Agathocles. "399 B.C." is even less help, since it contains a reference to the birth of Jesus of Nazareth in Bethlehem, a reference which which we know was established by one Dionysius Exiguus in terms of a reference to Rome and its foundation by Romulus, which Dionysius derived from the histories of Livy, a reference which we know (by reference to the account given by Josephus of the life and death of Herod the Great) must have been mistaken. In a word: things are prior to events, and the world, as a system of efficient causality, is a system of things. The world is the totality of things, not of facts.

I believe that this claim is already likely to be unpopular enough, but having started I might as well continue. These things which are related by efficient causality are things with powers, and it is in virtue of their powers that they exercise causality. This thesis is usually subject to as much ridicule as is the picture of Aristotleís highly sexed physical universe. As soon as the word "power" appears on the page, someone will already be penciling in the margin a witticism about "virtus dormitiva". This joke derives from Molière and is probably the poorest joke ever used in any philosophical context: certainly the poorest joke to be used more than once. I have a certain right to speak here, as many of the jokes I make in philosophical contexts are very poor indeed. Some of them have no point, others have a very slight and unimportant point. But at least my jokes donít miss the point altogether, and donít get used to make exactly the opposite point to the one they really represent. And the really poor ones I feel embarrassed about, and donít use more than once.

Molière represents for us an examination of a medical doctor, and the joke is that when the doctors canít answer difficult questions they disguise their ignorance with high-sounding Latin phrases. Thus, when asked why opium puts people to sleep, the candidate replies that it is because opium has a "virtus dormitiva", that is, a power to put people to sleep. It seems a little hard to pull such a joke to pieces, but since contemporary authors continue to misunderstand the point of the joke, I am going to have to. This joke is funny for the following reason: in his answer the doctor merely repeats what everyone knows in technical sounding language. Doing so is supposed to inspire confidence in the patient. This is a trick which doctors often play: in the first world war, what the troops called "trench fever" the doctors called "P.U.O.", pyrexia of unknown origin. The nature of this pyrexia is still unknown. The technique sometimes makes the patient feel more confident, which often helps towards a cure; and sometimes the mere giving of a label ó e.g. RSI, or repetitive strain injury, which means no more than that you are suffering discomfort or pain because you have strained your nerves, tendons and muscles, in some obscure way, through repeatedly performing slight movements such as typing ó can be of immense benefit to the patient because he or she is thus enabled to sue the pants off his or her employer. But when the doctor uses some high-sounding phrase to repeat what we already know, and we are sharp enough to spot this, and have no hope of a recovery or of compensation based on jargon alone, we laugh. So (let us hope) did Molièreís audience.

I feel deeply embarrassed flogging this joke to death, but it has to be done. The joke is funny because the answer is true but profoundly uninformative. Uninformativeness is a relative notion: an expression can be informative to one person, uninformative to another. When I point out that Glasgow is in Scotland, not England, this is uninformative to most of my readers, but when I visit Spain I have occasion to make this remark quite often, and it is informative. In the Molière case, we are supposed to be an intelligent and well-educated seventeenth-century audience who can work out that the dog-Latin sentence in question means that opium puts people to sleep because it has a power to put people to sleep. This is profoundly uninformative (and therefore, let me insist, mildly funny) because we knew it already. The answer is true, but tells us nothing we didnít know. We pay doctors to tell us things we didnít know, so when a doctor is represented as earning his money by telling us things we already knew well, covering his minor dishonesty in high-sounding language, we find it amusing.

Contrast, if you can bear to, the account given of this same joke by those who use it most, people of an empiricist cast of mind. They regard "opium puts people to sleep because it has a dormitive power" as not true but uninformative, but rather as tautological. What else, they ask, could "Opium has a dormitive power" mean other than "Opium puts people to sleep"? They think that the doctor in the play is to be laughed at because he repeats a tautology as if it were informative. This will not do. Informativeness and uninformativeness are, as I said, relative notions: relative to the degree of ignorance or education of the reader or listener. If Molière had wanted to give us a tautological answer, his own dog-Latin and that of his audience would surely have been up to it. "Why does opium put people to sleep?" "Quoniam dulce sugus papaveri inquantum sumitur omne animal dormitare facessit". That Latin is pretty doggy, but the worse the Latin, the easier to understand: "because the gentle juice of the poppy sends to sleep any animal that takes it". That is a tautology.

In contrast, "opium sends people to sleep because it has a dormitive power" is a perfectly sensible, if not very informative, scientific statement. It is not a tautology: and we can prove this in a nicely ironic way. A tautology is not significantly deniable: the negation of a tautology is a self-contradiction, which no-one can believe or even sensibly affirm. It is ironic that those very people of empiricist casts of mind who claim that "opium sends people to sleep because it has a dormitive power" is a tautology are just the people who deny this alleged tautology, who claim that it is false that opium sends people to sleep because it has a dormitive power. They deny, in fact, that opium has a dormitive power, because they deny that there are any such things as powers. They should not be allowed to have it both ways: they must not claim both that the famous sentence is a tautology and that it is false.

Someone like Locke would say that we have no way of knowing whether opium has, in its real essence,a dormitive power or not: we should limit ourselves to commenting on the fact that the (possibly quite disparate) kinds of thing which we group together roughly under the nominal essence of opium in fact send people to sleep. Though, for Locke, we shouldnít be surprised when they donít: a man who was not surprised at the story of the rational parrot, or the offspring of bulls with mares, or of apes with women, has no real right to be surprised at anything.

Or let us consider Hume. If you ask Hume whether it is the case that opium puts people to sleep because it has a dormitive power, he will answer No; at least if you catch him in the library rather than at the backgammon-board. There is no possible evidence on which we can base such an affirmation as "Opium puts people to sleep because it has a dormitive power": all we find is a constant conjunction between people taking opium and people going to sleep. To what extent we find this, and how constant the conjunction is, we had better not ask. In the interests of science and of annoying Humeans I am perfectly capable of taking a powerful emetic just before taking my opium, and thus showing that the conjunction between taking opium and going to sleep is by no means as constant as one might have thought. Of course it will be alleged that a constant conjunction which can be so easily falsified is not what Hume or the Humeans meant, and I am well able to believe it; but I have yet to find anyone capable of telling me just what the Devil they do mean.

Putting the rhetoric on the back burner for a little bit, the thesis I am trying to suggest is that the system of physical science with which we are all familiar, a system which most would be happy to see described as a system of efficient causality, is a system of things, principally bodies, and their powers, and the effects of these powers. It is true that science is more than this. I have claimed that the answer "Opium puts people to sleep because it has a dormitive power" is perfectly true, and, indeed, if the account of science I have just offered is in any way accurate, it is in some ways a model answer. But "model" answer is exactly right; or perhaps I should rather call it a blueprint for an answer, or the framework for an answer. It is, after all, wholly uninformative, it tells us nothing which we did not know already; and it can scarcely be claimed that science tells us nothing which we did not know already.

I am not a particular fan of uninformative answers, but I see nothing wrong with them when they prepare the ground for an informative one. And this answer is a case in point. Against the beliefs of any empiricists, or any believers in magic, it tells us that opium has a dormitive power and thus prepares us for an investigation into what that power consists in. Kenny would say: it prepares us for an investigation of the vehicle of that power, and the account he gives of the relationship between powers and their vehicles is a good one. What was wrong with the doctorís answer in Molière is that we expect the doctor to know more than we do, to be capable of going one step beyond the uninformative answer we already know, and to tell us what the vehicle of this dormitive power is: some feature of the chemical structure of opium which enables it to send us to sleep through some action within the chemical content of the brain (I expect). When the doctor fails to say something like this we feel cheated because we are paying him a lot for the years of study which are supposed to take him further in the investigation of the powers of natural things than we will ever get: and if he cannot tell us a suitable story about the vehicle we expect him to admit it and not fob us off with a lot of Latin. In my bedside table I have a packet of tranquillisers, and in the instructions that came with the pills I find the superb phrase "Alapryl, a preparation of halazepam, is a tranquilliser derived from benzodiazepine. As with other benzodiazepines, the exact mechanism by which it achieves its therapeutic effects is not clearly established." Thatís the way to talk: no comic dramatist in the world could take the mickey out of that kind of remark. Nor could any empiricist have any objection. But oddly enough the real scientists regard this kind of admission rather shame-facedly. I take this as evidence that my account of science, as an investigation of powers and their vehicles, though crude, is at least more accurate than that of the empiricists.

But if the world, considered as a system of efficient causality, necessarily involves the notion of powers, then it also necessarily reveals itself as a system of final causality. For powers are defined and identified in terms of their ends ó their "exercises", as Kenny would say. We can thus perhaps see reason for identifying, (not merely this or that bit of the world), but the world as a whole as a system which requires explanation in terms of what itís for.

We can go further. Leaving behind the question of powers, it is possible to claim that the notion of efficient causality requires the notion of a tendency, and the notion of tendency, still more clearly than that of a power, needs to be defined and identified in terms of what itís a tendency towards. I shall attempt to show this by outlining a framework for understanding reasoning in causal contexts ó contexts of efficient causality ó which works by assimilating causal reasoning to practical reasoning. Following a suggestion of Geach, I shall argue that systems of causal reasoning display the logical feature of "defeasibility", a feature which is often held to be distinctive of practical reasoning. I shall suggest that the reason for this logical isomorphism between causal and practical reasoning is that just as practical reasoning starts from goals, which are conscious ends, so all causal reasoning involves the notion of tendencies, which are specified in terms of ends.

Causal set-ups, then, are to be described as stating tendencies. These tendencies are specified in terms of their end, of that to which they are tendencies. For example: substances A and B, when mixed, have a tendency to produce an explosion. Given the additional premiss that substances A and B have been mixed, we very happily conclude that there will be (or has been) an explosion. But suppose that we add in extra premisses, as for example that substances A and B have been mixed in a medium of substance C: and that substance C has a tendency to interfere with the operations of these tendencies of substances A and B. We conclude that there will be no explosion. This reasoning is as defeasible as practical reasoning is.

If causal premisses are taken to express tendencies, which can be interfered with, and if this interference may invalidate any conclusion that may be drawn from the premisses, the parallels between causal reasoning and practical reasoning seem to be clear. That which may defeat the conclusion of the causal reasoning is interference, that is, the action of other causal tendencies not mentioned in the original premisses. Meanwhile, that which may defeat the conclusion of practical reasoning is the addition of other goal-expressing premisses.

This parallel can be developed. The relevant premisses in practical reasoning are premisses which express goals: in causal reasoning, the relevant premisses are premisses which express or describe tendencies. The premisses in both cases involve an unavoidable reference to an end. This is clear in the case of practical reasoning: but it should be no less clear in the case of causal reasoning. A tendency is specified in terms of what it is a tendency towards: the notion of an end is thus unavoidably involved.

It is true that an explosion is not produced, in the imagined case: and to that extent the tendencies of substances A and B to produce an explosion are not fulfilled. But though these tendencies, under that description, are not fulfilled, there are surely other descriptions of the same causal tendencies of substances A and B under which the tendencies are fulfilled. We have to to tell some such story as that while substance A has a tendency to combine with substance B in such a way as to produce an explosion it also has a tendency to combine with substance C to produce some compound which does not have a tendency to produce an explosion when combined with substance B. Both these tendencies could perhaps be redescribed in general terms of e.g. valency etc., in such a way that both tendencies can be seen as species of one and the same generic tendency, just as both effects ó the combination with B to produce an explosion, and the combination with C to prevent an explosion ó can be seen as different fulfilments of the same tendency. We are not to think of tendencies becoming wholly inoperative when interfered with. A mixture of A and B in a medium of C is something with causal tendencies quite different from those of the same quantity of C alone: the tendencies of A and B continue to exist and to have some fulfilment. The same tendencies, we want to say, under certain descriptions may be fulfilled, and under other descriptions may not be fulfilled.

A question may arise of what right I have to speak of "the same tendencies under different descriptions". If a tendency is specified by its end, as it surely must be, then a tendency with a different end is a different tendency. I believe this problem can be circumvented. Every tendency has an end: but this end will be variously describable. Quite a lot of what one is first taught in science classes, I would say, consists in establishing alternative (and more general) descriptions of given tendencies. For example, in basic dynamics: attaching such-and-such a weight to this rope, running over a pulley, has a tendency to lift such-and-such other weight attached to the other end. An alternative description of this tendency, in terms of force, and alternative descriptions of the weights in terms of mass, are made available. It is clearly the same tendencies which are differently described here.

This may give the clue to an answer. We should not say that a tendency that is "frustrated" by interference is not fulfilled: we should rather say something like "This tendency is fulfilled in a non-paradigmatic way". Each tendency will be specified in terms of an end: the end will be an event, and hence the tendency will be specified by a description of an event. For example, we may describe a tendency as a tendency to produce an explosion, say. The paradigmatic fulfilment of the tendency will be the occurrence of an event that meets the description that was used in specifying the tendency: in this case, an explosion. There may well be no such occurrence, no explosion, in which case there will be no paradigmatic fulfilment of the tendency. But there will be some fulfilment of the tendency: a non-paradigmatic one. We will have to redescribe the tendency to produce an explosion in such a way that the failure of the mixture to explode is also a fulfilment of the tendency: e.g. by redescribing it as a tendency to combine with other substances in such-and-such ways. But a fulfilment of a tendency will be paradigmatic or non-paradigmatic relative to a certain description of the tendency: for every fulfilment of a tendency that is non-paradigmatic under a certain description, there will be another description of the tendency such that the fulfilment is paradigmatic under that description. Both the combination of substance A with substance B to produce an explosion, and the combination of substance A with substance C, and the consequent failure of substance A to produce an explosion despite the presence of substance B, will both be paradigmatic fulfilments of the same tendency of A under different descriptions. Moreover, as our understanding develops, we can offer a description of the tendency in A to the paradigmatic effect of producing an explosion in the presence of B such that is also a description of a tendency to the paradigmatic effect of failing to explode in the presence of B and C.

This will need tightening up. A first objection to this would seem to be that since any event can be described as "something happening", any tendency can be described as a tendency for something to happen. This would empty this doctrine of all content. "If there is a tendency for something to happen, something will happen" does not look helpful. (We may notice, though, that even this example is not wholly vacuous: it has enough content to be false. If the conflicting or interfering tendencies are equally balanced, then nothing will happen.) But the point is well made. We will need some kind of stipulations about levels of generality of descriptions of ends to be used in specifying tendencies. A first shot would be to stipulate something like the following: when we are dealing with conflicting or interfering tendencies, we are to seek for alternative, more general descriptions of these tendencies such that both tendencies are fulfilled paradigmatically. That is, the tendencies are to be specified in such a way that both are fulfilled paradigmatically. If we then stipulate that the least general such description is to be taken, we shall avoid at least the uninformativeness of attributing to agents "tendencies to bring it about that something happens".

That is, we donít want to have laws of chemistry that say something like "When the substances A and B are put together, something happens." But nor do we want laws that say "When substances A and B are put together, an explosion occurs", which, given that there may be interference, would be false. We want a description of the tendencies of A and B such that the event described by "an explosion occurred" and the event described by "an explosion did not occur" are equally paradigmatic fulfilments of the tendencies.

We may want to add here a suggestion that the scientific project is one of determining appropriate levels of generality of description. The least general level is the place to start, but this, in our example, may get us no further than saying that A has a tendency to combine with B to produce an explosion and to combine with C to avoid an explosion. This is not yet a scientific explanation: but it seems to be the place at which to start a scientific investigation, to see what further more general and more explanatory, and hence more appropriate, descriptions of this tendency in A we can produce. Appropriateness here is a difficult notion to specify, but it is one which it is necessary to use. What we want is a level general enough for us to be able to say that the tendencies are fulfilled paradigmatically, but not so general as to be uninformative: and not so particular, meanwhile, as to be useless for explanation of other causal efficacy of the subjects.

The mysterious notion of explanatoriness seems to be making an entrance here. Explanatoriness is indeed a mysterious concept, but there are features which it is known to have which can be seen to be relevant here. The giving of an explanation appears to create a non-extensional context: to use Prof. Anscombeís example, to say "There was an international crisis because the President of the French Republic made a speech" is explanatory, while "There was an international crisis because the man with the biggest nose in France made a speech" is not, even if the President of the French Republic is the man with the biggest nose in France. That is, explanation is in some sense relative to a description: it should occasion no surprise, then, that causal reasoning in terms of tendencies, which is supposed to be explanatory, should turn out to have description-relativity built in to it. This is yet another parallel to practical reasoning: actions are said to be intentional only under certain descriptions, and here we have it that effects are attributable to a tendency only under certain descriptions. I am not able to enter into a full discussion of this.

This account clarifies a number of difficulties about non-voluntary causality, and allows us to make clear the difference between voluntary and non-voluntary causality. Non-voluntary causality needs to be described in terms of tendencies, tendencies to some paradigmatic effect. These tendencies may often be fulfilled non-paradigmatically, because of interference. That is, there may well be no possible description of the effect which is also a description of the paradigmatic fulfilment of the tendency. However,the tendencies will always be fulfilled, if only non-paradigmatically: there will always be some description of the tendency, in terms of some paradigmatic fulfilment, such that the tendency, so described, is fulfilled paradigmatically. The apparent vacuousness of this is limited by the fact that if the paradigmatic fulfilment does not occur, there will always be some interference: and the actual outcome will also be a non-paradigmatic fulfilment of the interfering tendency. Once we are committed to taking the least general descriptions of the two tendencies under which the fulfilments are paradigmatic, the account we give will be the starting place, at least, for the search for an appropriate explanation: it will describe the outcome as the paradigmatic fulfilment of the conflicting tendencies. There will always be such a description available, barring the case of miracles.

The case which has here been made, then, is that a description of the world in terms of its being a system of efficient causality, is necessarily a system which involves tendencies, and therefore also a system which involves finality: indeed, that a description of the physical world which ignores finality is as absurd as a description of human action which ignores human goals. Of course both kinds of description, which I here categorise as absurd, are possible, in the sense that they can be written out or talked about. What is more difficult to know, as Aristotle remarked about Heraclitus, is to what extent it is possible to believe in them.

If what has been said is accepted, or even half-accepted, it makes sense to see the world as a system of final causality, as displaying a tendency. This, it will be remembered, was the feature X for the Fifth Way. That "displaying a tendency" has the mark of a suitable feature X of requiring an explanation in terms of a relation to something else is clear. It is not so clear to what extent it is generalisable, to what extent we can infer from "part of this whole displays feature X" that "the whole displays feature X".

It is pretty clear that St Thomas believes that the feature is generalisable, since he does not take the trouble, as I have done, to claim that each and every part of the world displays a tendency. It is perhaps simple enough to understand why he thinks this. If I put my file on the floor, and then a book on top of it, and then my word-processor on top of the book, and then a flower-vase on top of the word-processor, and then put a stick in the flower vase, and then carefully wrap a small scrap of ribbon around the stick, a question arises of what the Devil Iím up to: more scientifically, what is the point of all that?

I answer: I put the file on the floor in order to put the book on top of it, and I put the book on the file in order to put my word-processor on top of the book, and I put my word-processor on top of the book in order to put the flower-vase on top of the word-processor, and I put the flower-vase on top of the word-processor in order to put a stick in it, and I put the stick in the flower-vase in order to tie a piece of ribbon around the end of the stick. I have given something of a sketch of a system of final causality. Each part of the system of final causality, until the last, has a perfectly good explanation.

(We might have given this example in terms of our stock group of Inuit, of course, and the same point could have been made. Ingekok is here to prevent quarrels between Kadlu and Kotuko, and so on. I have avoided using the example here because the point of any of the groupís presence will be in terms of conscious teleology, their purpose. When dealing with teleological relations within the world, we are interested principally in unconscious teleology.)

Something remains to be said: if the last step of tying the ribbon, or the system as a whole, fails to have a point, one is surely entitled to say that the apparent point of each previous stage is wholly illusory. Again, as in the case of the Inuit, there can be a variety of forms of final explanation which will suffice: maybe the point of each part other than the tying of the ribbon derives from the point of the ribbon in itself, as in the case where the ribbon is yellow, and lacking an old oak tree I wish to build some structure tall enough for my sweetheart to be able to catch sight of the ribbon from the window of the bus. Or it may be that the point of each part does not relate directly or indirectly to the point of the explanatorily privileged ribbon: I am merely seized by the aesthetic passion of making piles of incongruous objects. Or something between the two: the upper part of the structure is a booby-trap, and the lower part is to make sure that it stays put until my victim gets here. As in the case of the Inuit, no-one knows and no-one cares: but it is a plausible claim that unless the whole has a point no part of it has a point, all appearances of final structure notwithstanding.

Thus St Thomas thinks that if the world as a whole does not have a point, then the things in the world that seem to have a point donít have a point either. There is room here for a subordinate argument such as he gives us in the first three Ways, that we cannot go on to infinity in this line: but perhaps there is no need for him to give it. The point had already been made for him by Aristotle in the Nicomachean Ethics: if everything is for the sake of something else, there is no point in anything. A parallel example which I have sometimes given is the "Lottery in Babylon" of Borgesís short story of the same name. The story contains a number of deliberate inconsistencies, but one of the things that is said or suggested is that the lottery, which has come to embrace all aspects of Babylonian life, and doles out both good and bad fortune, has become infinite. Nothing ever happens as a result of any drawing of the lottery except some further drawing of the lottery. The governors of Babylon, or the gods of Babylon, have introduced and extended this lottery with the aim of keeping their subjects in a useful and agreeable state of apprehension and hope. The reader, though, wonders how long it will be before someone recognises that the lottery determines nothing but further drawings of the lottery, and announces, as it were, that the Emperor has no clothes on at all. And what will happen to the useful and agreeable state of hope and apprehension then?

St Thomas seems to put before us a stark choice: either the world requires an explanation in terms of what itís for, or nothing in the world in fact has any point, though it may seem to. The problem with this choice is that people may very well take the alternative St Thomas rejects. It is not uncommon for people to say that they regard the world as having no point, and regard any appearance that anything in the world has a point as purely delusive. Some theists, reacting to this answer, might make the kind of remark that Aristotle makes about Heraclitus: they say it, but they canít mean it. Having myself suffered in my time from clinical depression (that is what working in the Philosophy Department of Glasgow University can do for you), I think that this theistic retort is both cheap and inaccurate. One can perfectly well believe that neither oneís own life nor the life of the world as a whole has any point that one is capable of understanding, and still jog on without feeling logically, morally or psychologically compelled to go for the overdose of sleeping pills.

But this is where my stronger claim, that the world as a whole has to be seen as a system of final causality, makes it harder to refuse to take St Thomasís preferred alternative: that the world as a whole is something that requires explanation in terms of some point that it has, in terms of being for something. For if my account of science is anything like right, then the whole of science is entirely fallacious if the world is not the sort of thing that has a point. This is an option which will seem perfectly plausible to depressives, but will seem less so to others. It is not merely a question of failing to see much point in life, but one of suddenly being unable to see any intelligible structure to life at all. This would take us well beyond anything which we can find parallels for in depressive symptoms, and brings us into the country of people who mistake their wives for hats. And here Aristotleís point against Heraclitus does seem to have a certain validity. We know that there was at least one man who mistook his wife for a hat, but if someone tells us that it is a mistake he is prone to make we find it pretty hard to believe: at least when he is playing backgammon rather than elucubrating in the philosophy section of the library.

It begins to look, then, as if the lumping-together move with this feature X has some validity. Even if we are willing to be hard-nosed about St Thomasís claim that the bits of the world which apparently have a point will fail to have one if the world as a whole has no point, we may feel a bit more doubtful about being hard-nosed if we find out that if we refuse to ask what the point of the world is, we are left with a world whose structure as a system of final causality, and therefore as a system of efficient causality, turns out to be wholly illusory. Or perhaps not: perhaps this is the reason why people have been trying to drum into us for centuries that the world can be a comprehensible structure of (quasi-) efficient causality without implying any beliefs about tendencies, and therefore without implying any vision of the world as a structure of final causality. The question which remains, though, is not that of whether we are capable of seeing the world as a structure of quasi-efficient causality while ignoring the tendencies which are thus presupposed, thus enabling us to ignore the final structure of the world. Of course we can do it: weíve done it for centuries. The question is whether we should. Likewise, it is clear that we are capable of considering all the money and goods that come into our power by means which our legal system sanctions as being our own, to do what we like with. The question is whether we should: whether we would not be behaving with some kind of minimal honesty if we returned to the unfashionable old idea that what we have above and beyond what we need for our own support, that of our families, and the fulfilment of our civic duties, is not actually owed as a debt of justice to those who are in need of it.

Be that as it may, once we start wondering about what the point of the world is, about what the world is for, or even once we start wondering whether it might not after all make sense to ask such a question, we run into a curious little problem which takes us on to the last step in the Fifth Way.

"But things which have no knowledge do not have a tendency to an end unless they are directed by something that does have knowledge and understanding. An example is an arrow directed by an archer."

This is the crucial claim of the Fifth Way, the step that takes us to God: and it is a little curious. It can be paraphrased as: every unconscious teleology, every case of something being for something without awareness of what itís for, is dependent on some conscious teleology, on some mind which is aware what that thing is for. It seems obvious to me that just in so far as one is disposed to admit the existence of unconscious teleology, one will be disposed to deny this claim, and just in so far as one is disposed to admit this claim one will be disposed to admit the existence of unconscious teleology. Here I find myself isolated. Both Reid and Hume were inclined to admit the claim, and neither is much of a friend to unconscious teleology. Most of my contemporaries, I think, would agree with the two Scots here. St Thomas admits the existence of unconscious teleology and actually makes the claim: I feel rather friendless when I admit that to me the evidence for the existence of unconscious teleology seems overwhelming, but I cannot at first sight see any reason for holding the claim that it must in general depend on conscious teleology.

"In general", I say. In the case of the world as a whole, there is, as I have mentioned, a problem with the lumping-together move which may in this case, if in no other, justify the dependence of unconscious teleology on conscious teleology. We have agreed, at least for the sake of argument, to regard the world as a system of things which are for something, which have a point, which display unconscious teleology. And, at least for the sake of following St Thomasís argument, we have to agree to perform the lumping-together move and ask what is the point of the world as a whole.

This leads to a problem. Unconscious teleology seems always to be system-relative, if that is not too much of a neologism. A typical relation of unconscious teleology, of one thing being for the sake of another without consciousness of what itís for, will be the relationship of a part to a whole, as in the case of a bodily organ; or the relationship of one part of an interconnected teleological system to another part of the same system, as in the case of the acorn. The examples are the most obvious ones, but clearly if we agree to count anything which manifests a tendency as having a point, and any part of the system of efficient causes which we call the world as manifesting a tendency, this will be quite generally true. The bit of the world which displays unconscious teleology, which has a point, which is for the sake of something, is, in the end, for the sake of the world as a whole.

What then can we say of the world as a whole? If the world has a point, if it is for anything, this cannot be a case of unconscious teleology, since unconscious teleology is always relative to a system, is always a case of a part being for the sake of the whole. There is, by definition, no greater whole of which the world forms a part. Even if there were, we could just perform the "we cannot go on to infinity in this line" move, and quickly come to the limit. There is therefore going to be something, viz. the world, which has to have a point if anything within the world is to have a point. Apparently there are things within the world that have a point: indeed, apparently the whole structure of the world depends on things displaying tendencies and thus having points, being for something. Therefore the world as a whole does have a point, does display teleology. But this teleology cannot be unconscious, since there is no greater whole for the world to have as its unconscious point. Therefore the teleology of the world must be conscious: the point of the world must be conferred on it by some mind.

"Therefore there is some being with understanding which directs all things to their end, and this, we say, is God."

I can never quite make up my mind as to whether or not I find this argument convincing. A parallel discussion in ethical arguments strikes me as equally perplexing, and perplexing in an interestingly similar way. The kind of moral philosophy I favour is broadly speaking Aristotelian, and has as its most fundamental concept that of human well-being, flourishing. I can argue to some effect about the content of this notion of well-being, about how it is related to human dispositions and actions, and what sort of things we can or should do to achieve it. That is, grant me that human well-being is a good thing and I can deduce for you a whole ethical system, offering on the way at least specious refutations of any rival system. But if you ask me to prove to you that human well-being or flourishing is itself good, I find myself at a loss. I donít even know how I could begin such a proof. The only way to start which occurs to me is to go all ecological and relate the life of human beings to the life of Gaia, or whatever. This has several difficulties. First, I donít believe in the existence of Gaia; secondly, itís pretty obvious that there are at least important aspects of the life of Gaia which we human beings are not very good for; thirdly, we cannot go on to infinity in this line: were I able to prove that human well-being is good for the life of Gaia you would be perfectly entitled to ask me what the life of Gaia is good for. No answer is obviously forthcoming.

At this stage the temptation is to throw up oneís hands and appeal to an intuition that human well-being is a good thing, or that the life of Gaia is a good thing. Intuitions are in any case suspect, and in this context, that of ultimate goods, I find so many different candidates for the post of being the ultimate good, offered to me on the basis of so many different peopleís different intuitions; and so many of these candidates I find absurd or disgusting. So the appeal to intuitions is doubly or trebly suspect.

Though I am a philosopher, and regard with the gravest suspicion any attempt to bring in theological considerations into my strictly philosophical work, I begin to wonder whether the only point of human well-being is that it seems to be something God wants. Perhaps here we are still in the realm of philosophical theology, natural theology. We spoke earlier of the truth which things have through their matching Godís mind. A closely related doctrine to this is that Godís knowledge of the world is causative: "Scientia Dei, causa rerum", Godís knowledge causes things to be. God does not have to observe the world to see how things turn out: this would make him passive to the world, and that contradicts the notion of God as Creator. Godís knowledge of the world is like our knowledge of our own actions: we know what we are doing without observation. The truth of things in the world depends on Godís knowledge, not vice-versa: it is not, that is to say, that the truth of Godís knowledge depends on things in the world.

A parallel point can be made for Godís will, and sometimes is. Things are the objects of our desire and goodwill because we find them good. But to transfer this kind of love to God would involve us in the same kind of mistake as would to think of God as having to learn from the world. Things are the object of Godís desire and goodwill and this makes them good. (I do not wish to enter into the famous voluntarist/anti-voluntarist controversy of the late Middle Ages here, and give serious consideration to the question of whether we might not wake up tomorrow to find that adultery was perfectly legitimate after all. The thesis I have maintained does not entail this kind of voluntarism. There might be aspects of Godís nature which set limits to what God can desire, and adultery might well fall outside these limits.) If human flourishing is a good thing, I want to suggest, in any ultimate sense, it can only be because God wants human beings to flourish. Likewise, if the world has a point, if it is for anything, it is because God has given it a point, has made it for something.

As I have said, I donít know whether or not the Fifth Way is a good proof. The alternative is pretty dreadful ó that thereís no point to anything ó but just because a thing is pretty dreadful doesnít have to make it false. Here the atheist cannot sensibly be accused of wishful thinking. But, as I have already commented, the theist need not be guilty of wishful thinking either, particularly if he or she subscribes to one or other of the traditional religions in our culture which embody classical theism. The atheistís conclusion, that thereís no point to anything, may be a pretty dreadful one; but a more dreadful one, surely, is the theistís conclusion that there is a very definite point to everything and that I and you and everyone you care for and countless others may well miss it.

If the worse comes to the worst, and you feel you have gained nothing either way from this book, draw this moral: donít make facile accusations of wishful thinking against people who disagree with you, and try to be consistent enough in your thought to make sure that any accusation of wishful thinking made against you is false.

So much for the irrelevant moral tacked on at the end of the tale: what has the real point of the tale been? The real point of writing this book has been to show that what I used to think of as philosophy can still be done. With modest attention to our own limitations, we still can gather remarkable new insights into important problems ó the existence of God, the nature of scientific inquiry, the notions of meaning and existence ó from a study of the wisdom of older authors. We are not confined to studying them as if they were members of an alien species, pegged out on a card maybe, or even going about their unfathomable business, whose thoughts can have no impact on our own. If we want to know about the existence of God, or about the nature of science, we should read Aquinas, not merely the writers of this century. If we want to study Aquinas we should pay him the compliment of treating as important what he thought of as important. To study Aquinas as Aquinas is a poor piece of flattery, since Aquinas cared very little for Aquinas, while he did care for God and for science.

Not that he cared for God and science in equal degrees: but he clearly thought that caring for God, for him, entailed developing a science of God. This meant he had to develop a theory of science. The first steps ó perhaps the most important step ó in his developed science is the question of the existence of God, and he answers it strictly according to his own rules for answering such questions. The Five Ways are not a brilliant jewel that can easily be wrenched out of its setting: to understand them we have to understand what kind of a question St Thomas thinks he is asking, why he asks it, how he thinks it is possible to set about answering it, and how the answers he gives meet up to his own standards ó yes, and to our standards too. Unless, of course, the careful reading of St Thomas has sometimes given us reason to abandon our own standards, which we have acquired in a severely limited environment, and taken on his standards.

The great benefit to be derived from reading pre-modern authors is to come to realise that after all we might have been mistaken up to now. This realisation is perhaps the chief and most general service I hope to have done to my readers.

There is a lesser, more modest and more specific aim that I have had: to make it clear that it is wise to attempt some understanding of St Thomasís theory of science before trying to understand his arguments for the existence of God. I also hope to have contributed something to the understanding of both. Equally well, I think it would be foolish to try to understand St Thomasís theory of science and explanation without relating it to his most famous scientific arguments in the context of explanation, which are the five ways. Lastly, I hope that some may have been able to develop new ideas of their own on the problems of the theory of science, on explanation, and of the existence of God, by being exposed to St Thomasís thought in a new way. That is the sort of thing I mean by "philosophy".

Return to contents

Go to Philosophy Home Page