We have tried to make out what St Thomas thought he was trying to do in the Summa theologiae, and to some extent we have tried to make out why he did it. Also we have tried to see why the attempt is not a wholly pointless one. Our aim is to elucidate the first substantive question he asks in this study, and to do this we needed to examine what kind of a study it is. St Thomas was trying to construct what he thought of as a science: that is, a systematic body of knowledge, a part of wisdom, as he says, out of all the materials available to him. Among the materials available to him are the authorities, and, as we have seen in the last chapter, it came naturally to a medieval thinker to begin his scientific investigations with an examination of material derived from authority, as it comes natural to us in our ordinary occupations and thoughts. The two authorities whom Aquinas most respects and most uses are Aristotle and St Augustine. The choice of these two is not arbitrary. St Augustine in some sense provided the model, Aristotle the detailed method.

The striking feature of the work of St Augustine is his attempt at harmonisation, at reconciliation. This is recognised by any encyclopedia, which will explain Augustine’s achievement as that of bringing about a synthesis between Christian and pagan wisdom. But there is more to be said. The synthesis, the harmony, which Augustine first sought was a harmony between his thought and his life. This harmony was something he had been seeking over half a lifetime, but which he had only achieved through making the submission of humility required for his conversion. The well-known story of the moment of his conversion is extremely relevant here, as in the discussion of many other stages of Augustine’s thought. The great scholar was only able to harmonise his life and his thought through taking the half-overheard inconsequential babblings of a child as a voice from Heaven. It was this that enabled his will to respond to his intellect, and to achieve the harmony of life that he had been seeking, the lack of which harmony had been torturing him for years.

From this fact stems Augustine’s belief that in some sense the understanding of faith precedes the understanding of reason: a paradoxical claim, since it seems rather obvious that while God gives reason to all, he gives faith only to some. But for Augustine reason on its own will wander blindly, incapable of understanding even its own truths, unless the reasoner submits to the radical conversion to the faith which Augustine himself had undergone, a conversion which threw new light on all his previous learning. At last, he thought, he had achieved an understanding in which his life and his thought could be united. This meant a wholesale re-appraisal of his previous thought. At last, he felt, he could understand Plato correctly, better even than Plato had done. He could at last see why and how Plato had been right (when he had been right) and why and how he had been wrong (when he had been wrong). All human learning could now be seen in the light of faith: it could either be pressed into service to illuminate the understanding of faith, or be definitively rejected as inconsistent with the faith.

This view of Augustine’s developed into a tradition, which enlightened the succeeding centuries. Harmonisation, reconciliation, and synthesis became the ideals pursued at the University of Paris, as they had been at the Augustinian-inspired schools which had preceded it. In Aquinas’s time this ideal was facing its greatest challenge.

The slow, argumentative discussion and development of this unified tradition of human and divine learning had been thrown into turmoil by the re-appearance in the West of the works of Aristotle. Aristotle was already, to the early medievals, a name to which respect was due, an auctor, one with authority. The neo-Platonic philosophy which Augustine had made the basis of the human part of his synthesis of wisdom had already, before Augustine’s time, adopted Aristotle’s logic to provide it with a structure. Moreover, some of Aristotle’s fundamental metaphysical categories had become known in the West through the writings of Boethius, and had provided a framework for the development of theological speculation about the inner life of God, and God’s relation to the world.

For this reason, the newly discovered writings of Aristotle could not merely be brushed off as unimportant, as trivial and erroneous philosophising. Moreover, the general philosophy of Aristotle fitted far better, unsurprisingly, with the logic of Aristotle which the medievals were used to working with, than it did with the general neo-Platonic philosophy adopted by Augustine. Above all, Aristotle’s work presented a fully developed, coherent view of the world as a whole, and of the human being’s place in it, which was in important respects inconsistent with the view upheld by the Augustinian synthesis. It naturally presented itself as a rival to that view.

Moreover, the Augustinian position itself demanded that it should be able to give some account of this rival. The aim of the Augustinian project was a synthesis of all human wisdom. Any proposition put forward for inclusion in that synthesis, in principle, could be judged in the light of that wisdom. If true, it could be incorporated; if false, it could be shown to be false and rejected. It looked, at this time, as if the philosophy of Aristotle could neither be incorporated nor rejected. Medieval thinkers were all too prone to accuse their opponents of teaching that there could be two separate kinds of truth, a religious truth and a truth of reason. Any thinker who did so hold would have been guilty of abandoning the Augustinian project (as well as the Aristotelian logic which structured it), and it is in fact hard to pin down any medieval thinker as actually having made such a radical claim. But we do have, from the period, the notes of an anonymous student who was confused enough by the problems of the age to note "The above [Aristotelian] propositions are true in the Faculty of Arts, but not in the Faculty of Divinity".

It was into this world, facing this problem, that Aquinas came: and he did not shirk the problem. He attempted to establish the framework for, or to make the first step in, creating a new complete synthesis, a synthesis of the Augustinian tradition and the wisdom of Aristotle. The joint importance of these two authors to Aquinas can be brought out in a crude but effective way: in the index of a recent edition of the Summa Theologiae, references to both run to over thirty columns, while references to their nearest rival, St Leo, run to only ten.

We could say, putting it crudely, that the aim and spirit of St Thomas’s project was Augustinian. Aquinas was seeking an understanding which a Christian could live by, and the holiness of his own life may be held by some to have confirmed that he achieved this aim. But in many ways the structure was Aristotelian.

Thus in the case at issue, of the discussion of the nature of science: Aquinas is seeking for wisdom, wherever it may be found, and if the beginning of wisdom is the fear of the Lord, a necessary preliminary is to establish that the Lord is God. But this wisdom is understood by Aquinas as being an Aristotelian science: a science of the kind for which Aristotle laid down principles in his Posterior Analytics, and which he tried to develop in his Metaphysics.

The Posterior Analytics is a curious work. It seems to present as the ideal science a system of deductive inferences from stated axioms. A body of knowledge on these lines was well known for centuries in Euclid’s geometry. We may pause to marvel at Aristotle’s perspicuity in laying down rules for what was not to be achieved until several generations after his death, but our principal reaction is usually one of impatience. Surely no other body of knowledge is, or could be, articulated in such a way. If no other body of knowledge can be an Aristotelian science, then the notion of an Aristotelian science is of very little interest to us.

Moreover, there is a very apparent contradiction between Aristotle’s account of science in the Posterior Analytics and his actual practice in the Metaphysics, the Physics, the Ethics, or the De Anima, to say nothing of his practice in his magisterial works on natural history. We find nothing in these that is parallel to the structure of Euclid’s geometry. Instead, Aristotle examines common experience, as found either in folk-wisdom or in the writings of the poets or of other philosophers; or, where common experience is lacking, he makes his own observations. He then brings arguments, his own or those of others, against the obvious or usual explanations of these experiences. If these objections can be rejected, as they often are, the common view is held to stand, perhaps suitably modified by criticism, or perhaps with a more developed explanation provided by Aristotle himself. And then on to the next topic.

We can make of this what we will. One bizarre, but possible, reaction, might be to reject all of Aristotle’s substantive work, for not fitting in with his professed methodology. Another, less bizarre reaction, which actually occurred among later scholastics, would be to reduce all of Aristotle’s work to the structure suggested in the Posterior Analytics. This, as I say, has actually been done: it is an operation which has in great part been responsible for the bad name that Aristotelian scholasticism still has in some quarters. Another reaction is to reject the Posterior Analytics: a course which has been taken by authors as recent, and as favourable to Aristotle, as G.E.M.Anscombe.

A sounder reaction is one which has more recently gained favour. This is to observe that what Aristotle is recommending in the Posterior Analytics is not what he thinks he is doing in the rest of his works. In the Posterior Analytics Aristotle is laying down what he takes to be the correct articulation of a body of knowledge once achieved: it has nothing to do with the way in which that body of knowledge is acquired. The model in the Posterior Analytics is just that, a model: an ideal. It bears the same relation to what Aristotle does in the rest of his works as a management consultant’s flow-chart of the operations and relations aimed at in the operation of a work-space to the processes of designing and constructing an office layout, and hiring and training the staff who will carry out the operations and bear these relations. It is not, in fact, a theory of finding out truths, which, crudely speaking, is what we think a theory of science should be; it is a theory of the relations which should be seen to exist between truths once found out.

Whether the theory in the Posterior Analytics is in fact a good account of the relations between the truths of a science once found out, is a disputable question, which will be examined shortly. What is important to notice here is that Aristotle should not be criticised for doing badly something he never set out to do — giving an an account of finding out — and that therefore neither should St Thomas be criticised for imitating him.

We might observe, though, that even if Aristotle’s account of the structure of a completed science is correct, we may fault him for failing to draw our attention to the difference between this account and an account of the way in which we can find out. Certainly we can criticise him for not giving us a sufficient account of finding out. He gives us no theory of science in the modern sense, no theory of finding out, though he held such a theory, at least tacitly, as we can see by examining his actual practice. This criticism can also be made against St Thomas, but with less force; in his Commentary on the Posterior Analytics St Thomas does give more explicit attention than his master to the process of finding out.

Nevertheless, we can still criticise St Thomas for a lack of explicitness in distinguishing between a theory of a completed science, and a theory of finding out. And we also need to look at whether the theory of the ideal, completed, model science is accurate or not. If it is not accurate, then, we might think, just in so far as St Thomas succeeds in making his new divine science like the model, he will be failing to produce a good science. In the same way, the management consultant’s flow charts may be perfectly well-constructed, but they may not help us to achieve what we need to achieve.

This reaction would be an exaggerated one. St Thomas would not have thought that what he left at the end of the Summa Theologiae, even if that work had been completed, would have conformed to the model in any detail. The construction of a divine science is not work for one individual. Even in geometry, Euclid’s achievement was to put together in a coherent structure the work of innumerable predecessors. Aristotle was proud of being the first to produce a science of logic, but it was clearly not complete, he knew it was not complete, and St Thomas, who himself provided some interesting developments and corrections of Aristotle’s logic, did not take it to be complete. St Thomas would have held that no science, with the possible exception of geometry, was at his time complete: all required more development. The more developed a science was, the more one could hope to see in it the features of the model: and St Thomas does often draw attention to the way in which Aristotle’s work often does match the model in important respects. We can expect, then, to find St Thomas holding that in so far as his work is complete it will match the model: but the fact that the model is based on a misunderstanding, if it is a fact, need not mean that Aquinas’s work, when valuable, is so only by mistake.

Clearly, though, some examination of the model is in order. Crudely, the model is articulated in a top-down, deductive manner, though any science which has been built up to fit it will in fact have been developed in a bottom-up, inductive or dialectic manner. The process of finding out starts with individuals, and therefore with the contingent: but it eventually reaches the universal and the necessary. A completed science, therefore, has to do with the universal, and it consists of necessary truths.

The conception of "necessity" involved here need not be a very strong one: often it seems to be equivalent to no more than "universality" or "everlastingness". But, for an Aristotelian, there are, strictly speaking, no universal and everlasting contingent truths: there is nothing that just happens to be always the case. The necessity involved in a completed science is at least natural necessity. Though it need not be as strong as logical or mathematical necessity, it is never as weak as pure contingent universality or everlastingness, as what just happens to be always or everywhere the case.

It is important to recognise, though, that for a theistic Aristotelian such as Aquinas, all the natural necessities in the world, though genuine necessities, are in a sense conditional necessities, and in that sense are infected with contingency. Aquinas thought, with Aristotle, that the uniform circular motions that had been observed in the heavenly bodies were necessary, both in the sense that they continue for ever, and in the sense that given that there are the heavens that there are, the movements could not have been otherwise. But it was a debatable point whether there could not have been more or fewer heavenly bodies: and, for Aquinas, it was undisputably true that the world, with its heavenly bodies, might not have existed at all. What Aristotle would have thought on this question is again disputable, but the medieval thinkers who assimilated Aristotle were unequivocal on this question.

The whole question of the necessity of the truths of a completed science is yet more complicated by an observation of Aquinas that the link from effect to cause is necessary, while the link from cause to effect is contingent. His point is that this individual effect could not have come to be without this cause — since had it come about any other way we would not count it as "this individual effect", but as another, qualitatively indistinguishable, individual effect — while what in fact caused this effect, in a different overall context, could have had a different effect. Since in the realm of physics, for example, many of the explanatory links in a science will consist of the efficient causality of which he is speaking here, the distinction between the dialectical, inductive and contingent process of building up a science, and the necessary, deductive nature of the science once constructed, seems doubly threatened, in that the links of bottom-up, effect-to-cause reasoning employed in the construction of a science, which should be contingent according to the theory, appear to be necessary; while the links of top-down, effect-to-cause reasoning that will be found in a completed science, which should be necessary, appear to be contingent.

The problem is intractable. We can perhaps make a start by pointing out that there is no reason to suppose that every useful, interesting or important piece of knowledge which we acquire in building up a science will eventually find its place in the completed science. Certainly, not every necessary connection which we make use of in building up a science will have to form a part of the completed science. While it is clear that St Thomas, following Aristotle, regards what he calls a "science", a completed science, understood as we have said, as a paradigm of knowledge, it is also clear that there is plenty of important knowledge which fits this paradigm only very imperfectly. We need to take a step back, and try to understand, first, why their notion of "science" is important to Aristotle and to Aquinas, and how one science is related to another in the overall structure of speculative wisdom.

Plato, before Aristotle, had drawn attention to the difference between knowledge and true belief, and the distinction may go further back than that, to Socrates. Clearly, we are unwilling to grant that a person knows this or that fact simply because she or he very strongly believes it. We are all acquainted with strong, unfounded belief, and it may very well be false. Even when it happens to be true, that it is true is contingent. We expect someone who claims to know some fact to be able back the claim up by reasons, as Plato said in the Meno. We do not, in general, accept that people know that so-and-so is the case, even if so-and-so is the case, unless they have not only reasons, but the right reasons, for claiming that so-and-so is the case.

The tag that Plato uses in the Meno to mark out the difference between true belief and knowledge — that knowledge is "tied down by calculation of reasons" — comes close to Aristotle’s own definition of episteme, science, already mentioned, that it is "definite knowledge through reasons" or "through explanations". (The more classical translation, "certain knowledge through causes", is ambiguous with regard to the "certain", and too restrictive in modern philosophical English as regards "causes". "Cause" in modern philosophical English tends to mean what Aristotle called the "efficient cause" or "efficient mode of explanation", an explanation in terms of how a thing came about. Aristotle famously also recognises explanation in terms of matter — of what a thing is made of; in terms of form — of what makes what it’s made of into what it is; and in terms of end — of what it’s for.)

Aristotle, and, following him, St Thomas, regard science as the fullest kind of knowledge because the reasons given in science are the right reasons. This appears to mean the following. Let us suppose, to take an example of Aristotle’s own which St Thomas discusses, that there is an eclipse of the Moon. I can know this by observation. I may go further, and discover, building up a science, that there is an eclipse of the Moon because the Earth is obstructing the light of the Sun. Is this all I need to know? Are my reasons for knowing that there is an eclipse of the Moon, and that the Earth is obstructing the light of the Sun, as good as they might be?

Aristotle holds that they are not. In this case I know that p — that there is an eclipse of the Moon and I know that q — that the Earth is obstructing the light of the Sun. I also know that p-because-q, that there is an eclipse of the Moon because the Earth is obstructing the light of the Sun. Aristotle clearly thinks that until the order of reasons for my knowledge matches the order of reasons for reality, my knowledge is still imperfect. His ideal is that I should know that p because I know that q, that I should know that there is an eclipse of the Moon because I know that the earth is obstructing the light of the Sun. To know that the sun is obstructing the light of the sun because I know that there is an eclipse of the Moon is an imperfect kind of knowledge, one that falls short of complete science.

The example, though Aristotle’s own, is in some ways badly chosen, and makes the whole idea of Aristotelian science look more absurd than it need. Aristotle’s science is not concerned with individual occurrences: he does not, therefore, have to hold that scientific knowledge is a better way to know whether the Moon is in fact at present eclipsed than is looking at the Moon, though the way he sets up the example seems to suggest this. What he does have to hold is that if I know that at times the Sun’s light is obstructed by the Earth, and that therefore there are, at those times, eclipses of the Moon, I have a better reason for holding that there are at times eclipses of the Moon than I would have through having noticed the Moon eclipsed from time to time; and, a fortiori, it is to have a better and more complete kind of knowledge than if I only know that the Earth sometimes obstructs the light of the sun because I know that the Moon is sometimes eclipsed.

This may still seem a little odd: surely, some would say, I could not have better warrant for believing that the moon is eclipsed than my having seen the moon eclipsed? I think this is a mistake. I have seen rainbows, and water on tarmac roads on hot days. But I do not believe that there is a physical object called a rainbow, and I do not believe that there is more water on tarmac roads on hot days than on cool rainy days. My having observed eclipses is far from being the best reason I could have for believing that there are physical process called "eclipses". If I am to believe this, I need stronger warrant, which brings my belief in eclipses into an organised system of explanation of how the physical world works. It is because I have no such warrant for a belief in a physical object called the Rainbow, or for the existence of water on the road on a hot day, that in the end I come to disbelieve in their existence.

Put in this context, Aristotle’s position does not look quite so absurd. Moreover, it draws attention to the fact that each piece of knowledge in a scientific system is only as good as the system of science as a whole that supports it. If the best reason I can have for holding that the moon is eclipsed is my knowledge that the Earth is obstructing the light of the Sun, then it matters a great deal what is the best reason I can have for holding that the Earth is obstructing the light of the Sun. And the best reason here would be to do with some optical thesis about the way light travels in straight lines, and some astronomical thesis about the relative movements and positions of Earth, Moon and Sun.

This explains why the notion of science is important to Aristotle and to Aquinas. A body of science, if achieved, would provide one with the best foundation or warrant for a knowledge-claim that one could have. In the same way, the best foundation or warrant for a claim to know some theorem in Euclid is to be able to demonstrate it from Euclid’s axioms.

This connects with the next point of explanation. Each "science" — each articulated body of explanatory knowledge, within a given subject-matter — should, like Euclid’s geometry, be traceable back to certain axioms. These will be truths which are taken as basic for the science in question. They will be few and of universal scope. Thus, certain definitions and statements about properties of bodies as such will form the axioms of physics, certain definitions and statements about properties of living beings will form the axioms of biology, and so on for each individual science. It is important to notice that these axioms are taken as fundamental for the purposes of the science concerned. It is not the task of a science to prove its own axioms: what a science has to do is to draw out the reasonings from them. The axioms can be, and usually are, the conclusions of a more fundamental science: indeed, it is the question of what conclusions are used as the basis for what science that settles the question of which science is more fundamental than which. Thus, strictly, no science is complete until all sciences are complete. It is hardly to be wondered at that the practice of Aristotle and St Thomas is so different from what they lay down in their model. If there were no reasonable process of constructing a science which differed markedly from the structure laid down for how a completed science should be, there would be no possibility of starting: nothing could be known until everything were known.

Though the above remarks may have done something to make the notion of Aristotelian science less thoroughly alien to our conceptions, less absurd, and therefore to that extent more acceptable, little has been done in direct defence of the notion. In fact, little can be done. I shall continue by drawing attention to some other features of the articulation of science which have their effect in the theory of how a science is to be built up, in the hope of showing at least that the theory of a completed science, though perhaps in this day and age indefensible overall, is at least comprehensible, and is not likely to give rise to great distortions in the building up of a science in detail.

Key elements in the structure of a completed science are definitions. Definitions, for Aristotle and St Thomas, are always real rather than nominal definitions: they are always definitions of what the thing is, not definitions of what the word is. I notice that the distinction between the two is well brought out by the difference between, say, the Oxford English Dictionary and the Spanish Diccionario de la Real Academia. The English dictionary seeks to tell you how the word is used, and thus tell you what it applies to. The Spanish dictionary tells you what the thing is that the word applies to, and thus tells you how the word should be used. Both kinds of dictionary are useful in different ways: and since the word often means what the thing is which the word is used to apply to, the definitions often coincide. But there is an important notional difference at stake.

A definition, then, for St Thomas, is a formula of words expressing what a thing is. This, too, is likely to be misunderstood. For Aquinas, as for Aristotle, what a thing is is its true essence: what explains why it is the way it is. Thus, to choose a geometrical example, a circle is a plain figure bounded by a line called the circumference, which is such that all straight lines drawn to the circumference from a single point within it are all of equal length one to another.. This definition explains why a circle has all the features that it has. Equally, a human being is a rational animal: this definition, if it is both true and complete, will explain why human beings have the features which they essentially have. Time is the measurement of movement, according to before and after. And so on.

We will see at another moment somethiing of how these definitions are arrived at, in the stage of constructing a science. What is important here is to see what is their role in a complete science. They are the key steps in its structure. It is clear that very little about the real world of moving bodies follows from those first principles of physics, those axioms we mentioned above, which define what it is to be a physical entity subject to change. For those first principles to give reasons for what bodies actually do in the world, we have to add in information about what bodies there are in the world, and what these bodies are. This information is given by the definitions. I have just said: "what bodies there are" and "what these bodies are", as if distinguishing the question of existence from the question of essence. It is natural to make such a distinction, and indeed it is made by both Aristotle and St Thomas: but it is important to notice that in their minds they are linked. If there are no planets, for example, then there is no "what" for the planets to be: there is no essence, and therefore no definition, of what does not exist. Equally well, to say that the planets have no essence is to say that the planets do not exist, perhaps in the sense that Rainbows do not exist; or, if they do exist, that they are mere coincidental phenomena, as we might say the weather is, and not appropriate studies of science.

Here we come up against the harder questions which I have been dodging. Do not the limits imposed on the notion of science by Aristotle and St Thomas make it too restricted to be interesting? Indeed, do they not make it too restricted to have any application at all outside the realms of Euclidean geometry and its offshoot, Ptolemaic astronomy? (And since it turns out that Ptolemaic astronomy is false in large and important respects, this is a limited field indeed.)

It is true that for St Thomas and Aristotle it seems that there could be no such science as what we call the science of meteorology. For these two authors the phenomena which make up the weather are just that, phenomena, coincidental existents which obey no laws, which have no essence. This is not just ignorance on their part: the fact that we can point to regularities and laws in the weather would probably make no difference to them. The weather, and the entities that make it up, are not substances, things with their own nature or essence, and therefore there can be no unchanging truths about them.

For it is a well-known fact that Aristotelian science is about eternal, or at least everlasting and unchanging truths. This follows from the "necessity" of the truths of science, which again follows from the aim of science to provide as good a reason for one’s beliefs as one could possibly have. We cannot have a better reason for our belief that p than that it could not possibly be that not-p. Hence science aims at necessary truths. And what is necessary is always the case: since what is sometimes not the case can be not the case. Hence science is of unchanging truths.

Aristotle, indeed, may even have held that what is always the case is necessary, that there is nothing that just happens to go on for ever, though the attribution of this view of Aristotle is disputed. But at this point we strike a difference with Aquinas. Part of the reason for holding that Aristotle believed that what is always the case is necessary, is the view that he believed in the principle of plenitude: that whatever can happen, at some time does happen. (There is an interesting parallel with the view of necessity and possibility which is nowadays called extreme modal realism, that whatever can happen does happen somewhere, in some real but non-actual "possible world".) Thus, whatever can stop happening, at some time does stop; thus, whatever always happens, happens of necessity. Aristotle was thus committed to a view, which we find him frequently mocked for, of the eternity of e.g. animal species. Since it was clear that it is necessary for kittens to come from cats, in some sense of "necessary", Aristotle seems to have thought that the series of cats producing kittens must have existed from eternity. But it is also arguable that all Aristotle meant by "the eternity of species" was that a natural kind has no tendency to stop existing, as such, while the individuals of a given natural kind generally do have such a tendency.

Be that as it may, this is where we begin to see important differences between Aristotle and St Thomas. The latter’s metaphysics is creationist: the former’s is not. St Thomas believed that the world at some time came into existence and will at some time cease; Aristotle did not. Thus St Thomas cannot have held the principle of plenitude in the form that, it is alleged, Aristotle may have held it. That form of the principle depends on the existence of infinite time in which all real possibilities might be actualised. St Thomas did not believe in an infinite extent of time.

St Thomas thus could not glibly identify the necessary, in any sense, with what always happens. Moreover, he had given some thought to the question of mules. Mules had been discussed by Aristotle, but he did not see the metaphysical implications. St Thomas did see the metaphysical implications, and accepted them. A mule is born from a horse and a donkey, but is neither a horse nor a donkey. It is recognisably a third equine species, another natural kind of the horse family — though an imperfect one, as one cannot breed mules from mules. Even if one grants that horses and donkeys might have existed for ever, as far back as the world exists breeding and giving birth to the next generation of horses or donkeys, this cannot be true of mules. It must be the case that species can come into existence; and it must have been obvious to St Thomas that if one took the trouble to keep horses and donkeys apart, that species would cease to exist as well.

St Thomas, then, does not identify the necessary with what always happens. Nor, given his creationist metaphysics, could he hold that anything created could be in any very strong sense a necessary existent: that it could not not have existed.

Be that as it may, there seems to be room in St Thomas’s theory for a genuinely necessary science of the contingent. If this is so, he might be able to admit what we call the science of meteorology as a genuine science, in his sense — albeit one which is still in process of completion, and not yet completed. The key text for this is the introduction to his Commentary on the Nicomachean Ethics of Aristotle.

The subject matter of ethics, in general, is human action. But human actions are individual, contingent, temporal and temporary. How then can they be the subject-matter of an Aristotelian science, which has to do with the necessary and everlasting? The answer given is to distinguish between a broader and a narrower sense of "subject matter". It is true that in general the subject matter of ethics is human actions. Human actions form its "material object", to use scholastic jargon. But clearly this is going to be in any case an insufficient account. Other studies, such as history, or, in our day, psychology and sociology, also study human actions. What is it that distinguishes ethics from these?

The distinguishing mark of any science is its "formal object": the precise aspect under which its material object is studied. Clearly the formal object of ethics will have to do with human actions in so far as they are good or bad. St Thomas prefers to say that the formal object of ethics is the ordering of human actions towards the end of human life, or, since one and the same science deals with contraries, the extent to which human actions fail to be ordered to their end. There are clearly good reasons for holding that this ordering will be in the relevant sense necessary and unchanging. The end of human life is what human beings are for; and it is a plausible claim, as well as one which Aristotle and St Thomas would certainly endorse, that what human beings are for depends on the very nature which is expressed by the definition of the human being. It is arguable that this end, what human beings are for, does not, will not and cannot change so long as human beings exist. The manner in which human actions are directed towards or away from this end will also be radically unchangeable. Not, of course, that there may not be more or less wicked periods of history: but such a historical study falls outside the scope of ethics. All ethics has to tell us is, for example, that if there are periods in which such-and-such is done, they will be more wicked periods, to the extent that such actions are directed away from their proper end.

This kind of thought is not wholly alien to us, though some people nowadays may find its application to the field of ethics troublesome. (Though in this context it is worth remarking that distinguished contemporary moral philosophers in the English-speaking world, at least until very recently, have been willing to take the alleged special "universalisablity" of moral judgements as a mark of the special subject matter of ethics.) But in other cases we can see the application. Science, we tend to think, in some sense prescinds from the here and the now and aims at timeless validity. Token-reflexive expressions such as "here", "now", "I", "you", "yesterday", "tomorrow", "over there", have no place in what we would call a scientific discourse. This point is made clearly in Frege’s essay "The Thought", and Quine lays particular stress on it in e.g. Word and Object.Even experimental results — which are of course achieved in a particular time, in a particular place, by a particular scientist or group of scientists — are supposed to be intrinsically repeatable, and in so far as they fall short of repeatability or are suspected of falling short of repeatability, are to that extent ruled out as being serious scientific discourse. Thus this kind of Aristotelian thought is far from being wholly alien to us.

There can, then, be a true Aristotelian science of the apparently wholly contingent field of human actions, provided that the formal object of study is sufficiently clearly delimited to provide us with the necessity and unchangingness that we need. The same point can be made about physics. For Aristotle and St Thomas, most of the movements of terrestrial bodies are in themselves wholly contingent. But there is a necessary and unchanging ordering that they have, which we can make the object of scientific study. Aquinas will also allow us to express ourselves more loosely. We usually say that the subject matter of natural philosophy is that which is subject to change, rather than speaking more strictly and saying it is "the ordering of natural things"; and in the same way we can say that the subject-matter of moral philosophy is human performance in its ordering to its end, or human beings in so far as they act voluntarily for an end.

There are, indeed, more general considerations which help us to the same end. For Aquinas the per accidens, that which is composite or which exists coincidentally (see below, p. 000), cannot properly be the object of scientific study. But everything that is per accidens, that exists coincidentally, is made up of the per se, that which exists in its own right. And the per se is a proper object of scientific study. Thus there is nothing in the world that falls outside the scope of scientific study, even though not every description which is true of this or that part of the world sufficiently determines it as a possible object of a scientific study.

A passage early on in the Summa, later repeated, gives a general epistemological and logical background to this point. The way we understand the world need not be the way the world is, in the following sense: our structures of thought need not exactly match the world’s structures. To say that the cat is on the mat we need to use a number of words which cannot be all pronounced at once, though if the cat and the mat are not present all together, in the appropriate relation, the sentence will not be true. The relation between the cat and the mat is a spatial relation, while the relation between the words "cat" and "mat" is, in spoken English, a temporal one. In written English the relationship between the words is a spatial one, but not the same spatial relationship as that which exists between the cat and the mat. We do not need to perform typographical prodigies such as:

This, as Aquinas says, does not mean that our thought is false. Our thought would be false if it represented the world as being otherwise than the way the world is. It does not become false merely by itself being otherwise than the way the world is. This point, which is true of our thought in general, is true also of our scientific thought. It represents, in whatever way, necessary and unchanging aspects of a changing and contingent reality. According to Aquinas, the changeableness and contingency is made up of the coincidence of many strands of unchangeable and necessary causality or explanation. Our scientific thought does not misrepresent the world: it represents it to us in the only way that we can understand. That some descriptions that are true of the world are not descriptions relative to which we can understand the world adequately, or articulate our understanding, is not surprising. Understanding everything does not mean understanding everything about everything. When we read a book we do not need to know how many characters it contains, and when we examine the world of moving bodies we do not need to know how many moving bodies happen to have collided in the last half-hour between here and the end of the road.

The conclusion we can perhaps draw is that the Aristotelian notion of science employed by St Thomas is neither so bizarre nor so restricted as at first sight appeared. We can understand the desire for achieving a science which will be in some sense universal, necessary, and unchanging, and we can see that this desire will not rule out the possibility of genuine scientific knowledge of the contingent and changing world. Neither Aristotle nor St Thomas are partisans of the kind of a priori science that made the young Kant think he could make an accurate guess at the physical and indeed moral characteristics of the presumed inhabitants of other planets.

To next Section

Return to Contents

Return to Philosophy Page