The Presumption of Atheism1

Antony Flew

1. What it is, and why it matters

At the beginning of Book X of his last work The Laws, Plato turns his attention from violent and outrageous actions in general to the particular case of undisciplined and presumptuous behaviour in matters of religion:

We have already stated summarily what the punishment should be for temple-robbing, whether by open force or secretly. But the punishments for the various sorts of insolence in speech or action with regard to the gods, which a man can show in word or deed, have to be proclaimed after we have provided an exordium. Let this be it: "No one believing, as the laws prescribe, in the existence of the gods has ever yet performed an impious action willingly, or uttered a lawless word. Anyone acting in such a way is in one of three conditions: either, first, he does not believe the proposition aforesaid; or, second, he believes that though the gods exist they have no concern about men; or, third, he believes that they can easily be won over by the bribery of prayer and sacrifice" (§885B).[1]

So Plato in this notorious treatment of heresy might be said to be rebuking the presumption of atheism. The word 'presumption' would then be employed as a synonym for 'presumptuousness'. But, despite the interest of the questions raised by Plato, the term has in my title a different interpretation. The presumption of atheism which I want to discuss is not a form of presumptuousness. Indeed it might be regarded as an expression of the very opposite, a modest teachability. My presumption of atheism is closely analogous to the presumption of innocence in the English law; a comparison which I shall develop in Section 2. What I want to examine is the contention that the debate about the existence of God should properly begin from the presumption of atheism, that the onus of proof must lie upon the theist.

The word 'atheism', however, has in this contention to be construed unusually. Whereas nowadays the usual meaning of 'atheist' in English is 'someone who asserts that there is no such being as God', I want the word to be understood not positively but negatively. I want the originally Greek prefix 'a' to be read in the same way in 'atheist' as it customarily is read in such other Greco-English words as 'amoral', 'atypical', and 'asymmetrical'. In this interpretation an atheist becomes: not someone who positively asserts the non-existence of God; but someone who is simply not a theist. Let us, for future ready reference, introduce the labels 'positive atheist' for the former and 'negative atheist' for the latter.

The introduction of this new interpretation of the word 'atheism' may appear to be a piece of perverse Humpty-Dumptyism, going arbitrarily against established common usage.[2] 'Whyever', it could be asked, 'don't you make it not the presumption of atheism but the presumption of agnosticism?' It is too soon to attempt a full answer to this challenge and this suggestion. My justification for introducing the notion of negative atheism will be found in the whole development of the present chapter. Then in Chapter Two I intend to argue for a return to the original usage of the word 'agnosticism', as first introduced by Thomas Henry Huxley. In the meantime it should be sufficient to point out that, following the present degenerate usage, an agnostic is one who, having entertained the proposition that God exists, now claims not to know either that it is or that it is not true. To be in this ordinary sense an agnostic you have already to have conceded that there is, and that you have, a legitimate concept of God; such that, whether or not this concept does in fact have application, it theoretically could. But the atheist in my peculiar interpretation, unlike the atheist in the usual sense, has not as yet and as such conceded even this.

This point is important, though the question whether the word 'agnosticism' could bear the meaning which I want now to give to the word 'atheism' is not. What the protagonist of my presumption of atheism wants to show is that the debate about the existence of God ought to be conducted in a particular way, and that the issue should be seen in a certain perspective. His thesis about the onus of proof involves that it is up to the theist: first, to introduce and to defend his proposed concept of God; and, second, to provide sufficient reason for believing that this concept of his does in fact have an application.

It is the first of these two stages which needs perhaps to be emphasised even more strongly than the second. Where the question of existence concerns, for instance, a Loch Ness Monster or an Abominable Snowman, this stage may perhaps reasonably be deemed to be more or less complete before the argument begins. But in the controversy about the existence of God this is certainly not so: not only for the quite familiar reason that the word 'God' is used -- or misused -- in many different ways; but also, and much more interestingly, because it cannot be taken for granted that even the would-be mainstream theist is operating with a legitimate concept which theoretically could have an application to an actual being.

This last suggestion is not really as new-fangled and factitious as it is sometimes thought to be. But its pedigree has been made a little hard to trace. For the fact is that, traditionally, issues which should be seen as concerning the legitimacy or otherwise of a proposed or supposed concept have by philosophical theologians been discussed, either as surely disposable difficulties in reconciling one particular feature of the Divine nature with another, or else as aspects of an equally surely soluble general problem of saying something about the infinite Creator in language intelligible to His finite creatures. These traditional and still almost universally accepted forms of presentation are fundamentally prejudicial. For they assume that there is a Divine Being, with an actual nature the features of which we can investigate. They assume that there is an Infinite Creator, whose existence -- whatever difficulties we finite creatures may have in asserting anything else about Him -- we may take for granted.

The general reason why this presumption of atheism matters is that its acceptance must put the whole question of the existence of God into an entirely fresh perspective. Most immediately relevant here is that in this fresh perspective problems which really are conceptual are seen as conceptual problems; and problems which have tended to be regarded as advanced and, so to speak, optional extras now discover themselves as both elementary and indispensable. The theist who wants to build a systematic and thorough apologetic finds that he is required to begin absolutely from the beginning. This absolute beginning is to ensure that the word 'God' is provided with a meaning such that it is theoretically possible for an actual being to be so described.

Although I shall later be arguing that the presumption of atheism is neutral as between all parties to the main dispute, in as much as to accept it as determining a procedural framework is not to make any substantive assumptions, I must give fair warning now that I do nevertheless believe that in its fresh perspective the whole enterprise of theism appears even more difficult and precarious than it did before. In part this is a corollary of what I have just been suggesting; that certain difficulties and objections, which may previously have seemed peripheral or even factitious, are made to stand out as fundamental and unavoidable. But it is also in part, as we shall be seeing soon, a consequence of the emphasis which it places on the imperative need to produce some sort of sufficient reason to justify theist belief.

2. The presumption of atheism and the presumption of innocence.

One thing which helps to conceal this need is a confusion about the possible varieties of proof, and this confusion is one which can be resolved with the help of the first of a series of comparisons between my proposed presumption of atheism and the legal presumption of innocence.

(i) It is frequently said nowadays, even by professing Roman Catholics, that everyone knows that it is impossible to prove the existence of God. The first objection to this putative truism is, as my reference to Roman Catholics should have suggested, that it is not true. For it is an essential dogma of Roman Catholicism, defined as such by the First Vatican Council, that "the one and true God our creator and lord can be known for certain through the creation by the natural light of human reason" (Denzinger, H., Enchiridion Symbolorum, §1806). So even if this dogma is, as I myself believe, false, it is certainly not known to be false by those many Roman Catholics who remain, despite all the disturbances consequent upon the Second Vatican Council, committed to the complete traditional faith.

To this a sophisticated objector might reply that the definition of the First Vatican Council speaks of knowing for certain rather than of proving or demonstrating; adding perhaps, if he was very sophisticated indeed, that the word 'demonstrari' in an earlier draft was eventually replaced by the expression 'certo cognosci'. But, allowing that this is correct, it is certainly not enough to vindicate the conventional wisdom. For the word 'proof' is not ordinarily restricted in its application to demonstratively valid arguments, that is, in which the conclusion cannot be denied without thereby contradicting the premises. So it is too flattering to suggest that most of those who make this facile claim, that everyone knows that it is impossible to prove the existence of God, are intending only the strictly limited assertion that one special sort of proof, demonstrative proof, is impossible.

The truth, and the danger, is that wherever there is any awareness of such a limited and specialised interpretation, there will be a quick and illegitimate move to the much wider general conclusion that it is impossible and, furthermore, unnecessary to provide any sufficient reason for believing. It is, therefore, worth underlining that when the presumption of atheism is explained as insisting that the onus of proof must be on the theist, the word 'proof 'is being used in the ordinary wide sense in which it can embrace any and every variety of sufficient reason. It is, of course, in this and only this sense that the word is interpreted when the presumption of innocence is explained as laying the onus of proof on the prosecution.

(ii) A second element of positive analogy between these two presumptions is that both are defeasible; and that they are, consequently, not to be identified with assumptions. The presumption of innocence indicates where the court should start and how it must proceed. Yet the prosecution is still able, more often than not, to bring forward what is in the end accepted as sufficient reason to warrant the verdict 'Guilty'; which appropriate sufficient reason is properly characterised as a proof of guilt. The defeasible presumption of innocence is thus in this majority of cases in fact defeated. Were the indefeasible innocence of all accused persons an assumption of any legal system, then there could not be within that system any provision for any verdict other than 'Not Guilty'. To the extent that it is, for instance, an assumption of the English Common Law that every citizen is cognisant of all that the law requires of him, that law cannot admit the fact that this assumption is, as in fact it is, false.

The presumption of atheism is similarly defeasible. It lays it down that thorough and systematic inquiry must start from a position of negative atheism, and that the burden of proof lies on the theist proposition. Yet this is not at all the same thing as demanding that the debate should proceed on either a positive or a negative atheist assumption, which must preclude a theist conclusion. Counsel for theism no more betrays his client by accepting the framework determined by this presumption than counsel for the prosecution betrays the state by conceding the legal presumption of innocence. The latter is perhaps in his heart unshakably convinced of the guilt of the defendant. Yet he must, and with complete consistency and perfect sincerity may, insist that the proceedings of the court should respect the presumption of innocence. The former is even more likely to be persuaded of the soundness of his brief. Yet he too can with a good conscience allow that a thorough and complete apologetic must start from, meet, and go on to defeat, the presumption of atheism.

Put as I have just been putting it, the crucial distinction between a defeasible presumption and a categorical assumption will, no doubt, seem quite obvious. But obviousness really is, what some other things nowadays frequently said to be are not, essentially relative: what is obvious to one person at one time may not have been obvious to that same person at an earlier time, and may not be obvious now to another. There is no doubt but that many do find the present distinction difficult to grasp, especially in its application to exciting cases. Indeed one reason why I decided to write the lecture on which the present chapter is based is that I had found even the most acute and sympathetic critics of my God and Philosophy faulting me for asking everyone to start from my own notoriously atheist assumptions. It was clear that a more lucid and more adequately argued statement was needed. For in that book I had recommended only the present methodological presumption, not a substantive assumption.

I cite another example from a quite different sphere, an example which is again the more salutary since the offender was above suspicion of any dishonourable intent wilfully to misunderstand or misrepresent. Lord Attlee, once Leader of the British Labour Party, reproached the "general assumption that all applicants are frauds unless they prove themselves otherwise" (Young A.F. and Ashton, E.T., British Social Work in the Nineteenth Century, p. 111). But, we must insist, to put the onus of proof of entitlement upon the beneficiary is not to assume that all, or most, or even any of those who apply for welfare benefits are in fact cheats. Such presumptions are procedural purely. They assume no substantive conclusions.

(iii) However -- and here we come to a third element in the positive analogy -- to say that such presumptions are in themselves procedural and not substantive is not to say that the higher-order questions of whether to follow this presumption or that are trifling and merely formal rather than material and substantial. These higher-order questions are not questions which can be dismissed cynically as 'issues of principle as opposed to issues of substance'. It can matter a lot which presumption is adopted. Notoriously there is a world of difference between legal systems which follow the presumption of innocence, and those which do not. And, as I began to indicate at the end of Section 1, to adopt the presumption of atheism does put the whole argument into a distinctive perspective.

(iv) Next, as a fourth element in the positive analogy, it is a paradoxical consequence of the fact that these presumptions are procedural and not substantive that particular defeats do not constitute any sort of reason, much less a sufficient reason, for a general surrender. The fact that George Joseph Smith was in his trial proved guilty of many murders defeats the original presumption of his innocence. But this particular defeat has no tendency at all to show that even in this particular case the court should not have proceeded on this presumption. Still less does it tend to establish that the legal system as a whole was at fault in incorporating this presumption as a general principle. It is the same with the presumption of atheism. Suppose that someone is able to prove the existence of God. This achievement must, similarly, defeat our presumption. But it does not thereby show that the original contention about the onus of proof was mistaken.

Etymologically the word 'defeasible' ( = defeatable) does imply precisely this capacity of survive defeat. A substantive generalisation -- such as, for instance, the assertion that all persons accused of murder are in fact innocent -- is falsified decisively by the production of even one authentic counterexample. But a defeasible presumption is not shown to have been the wrong one to have made by being in a particular case in fact defeated.

3. The case for the presumption of atheism

What does show the presumption of atheism to be the right one is what we have now to investigate.

(i) An obvious first move is to appeal to the old legal axiom: "Ei incumbit probatio qui dicit, non qui negat." Literally and unsympathetically translated this becomes: "The onus of proof lies on the man who affirms, not on the man who denies." To this the objection is almost equally obvious. Given just a very little verbal ingenuity, the content of any motion can be rendered alternatively in either a negative or a positive form: either, "That this house denies the existence of God"; or, "That this house takes its stand for positive atheism". So interpreted, therefore, our axiom provides no determinate guidance.[3]

Suppose, however, that we take the hint already offered in the previous paragraph. A less literal but more sympathetic translation would be: "The onus of proof lies on the proposition, not on the opposition." The point of the change is to bring out that this maxim was offered in a legal context, and that our courts are institutions of debate. An axiom providing no determinate guidance outside that framework may nevertheless be fundamental for the effective conduct of orderly and decisive debate. Here the outcome is supposed to be decided on the merits of what is said within the debate itself, and of that alone. So no opposition can set about demolishing the proposition case until and unless that proposition has first provided them with a case for demolition: "You've got to get something on your plate before you can start messing it around" (Austin, J.L., Sense and Sensibilia, p. 142).

Of course our maxim even when thus sympathetically interpreted still offers no direction on which contending parties ought to be made to undertake which roles. Granting that courts are to operate as debating institutions, and granting that this maxim is fundamental to debate, we have to appeal to some further premise principle before we become licensed to infer that the prosecution must propose and the defence oppose. This further principle is, once again, the familiar presumption of innocence. Were we, while retaining the conception of a court as an institution for reaching decisions by way of formalised debate, to embrace the opposite presumption, the presumption of guilt, we should need to adopt the opposite arrangements. In these the defence would first propose that the accused is after all innocent, and the prosecution would then respond by struggling to disintegrate the case proposed.

(ii) The first move examined cannot, therefore, be by itself sufficient. To have considered it does nevertheless help to show that to accept such a presumption is to adopt a policy. And policies have to be assessed by reference to the aims of those for whom they are suggested. If for you it is more important that no guilty person should ever be acquitted than that no innocent person should ever be convicted, then for you a presumption of guilt must be the rational policy. For you, with your preference structure, a presumption of innocence becomes simply irrational. To adopt this policy would be to adopt means calculated to frustrate your own chosen ends; which is, surely, paradigmatically irrational. Take, as an actual illustration, the controlling elite of a ruling Leninist party, which must as such refuse to recognise any individual rights if these conflict with the claims of the party, and which in fact treats all those suspected of actual or potential opposition much as if they were already known 'counter-revolutionaries', 'enemies of socialism', 'friends of the United States', 'advocates of free elections', and all other like things bad. I can, and do, fault this policy and its agents on many counts. Yet I cannot say that for them, once granted their scale of values, it is irrational.

What then are the aims by reference to which an atheist presumption might be justified? One key word in the answer, if not the key word, must be 'knowledge'. The context for which such a policy is proposed is that of inquiry about the existence of God; and the object of the exercise is, presumably, to discover whether it is possible to establish that the word 'God' does in fact have application. Now to establish must here be either to show that you know or to come to know. But knowledge is crucially different from mere true belief. All knowledge involves true belief; not all true belief constitutes knowledge. To have a true belief is simply and solely to believe that something is so, and to be in fact right. But someone may believe that this or that is so, and his belief may in fact be true, without its thereby and necessarily constituting knowledge. If a true belief is to achieve this more elevated status, then the believer has to be properly warranted so to believe. He must, that is, be in a position to know.

Obviously there is enormous scope for disagreement in particular cases: both about what is required in order to be in a position to know; and about whether these requirements have actually been satisfied. But the crucial distinction between believing truly and knowing is recognised as universally as the prior and equally vital distinction between believing and believing what is in fact true. If, for instance, there is a question whether a colleague performed some discreditable action, then all of us, though we have perhaps to admit that we cannot help believing that he did, are rightly scrupulous not to assert that this is known unless we have grounds sufficient to warrant the bolder claim. It is, therefore, not only incongruous but also scandalous in matters of life and death, and even of eternal life and death, to maintain that you know either on no grounds at all, or on grounds of a kind which on other and comparatively minor issues you yourself would insist to be inadequate.

It is by reference to this inescapable demand for grounds that the presumption of atheism is justified. If it is to be established that there is a God, then we have to have good grounds for believing that this is indeed so. Until and unless some such grounds are produced we have literally no reason at all for believing; and in that situation the only reasonable posture must be that of either the negative atheist or the agnostic. So the onus of proof has to rest on the proposition. It must be up to them: first, to give whatever sense they choose to the word 'God', meeting any objection that so defined it would relate only to an incoherent pseudo-concept; and, second, to bring forward sufficient reasons to warrant their claim that, in their present sense of the word 'God', there is a God. The same applies, with appropriate alterations, if what is to be made out is, not that theism is known to be true, but only -- more modestly -- that it can be seen to be at least more or less probable.

4. Objections to the presumption of atheism

Once the nature of this presumption is understood, the supporting case is, as we have just seen in Section 3, short and simple.

(i) One reason why it may appear unacceptable is a confusion of contexts. In a theist or post-theist society it comes more easily to ask why a man is not a theist than why he is. Provided that the question is to be construed biographically this is no doubt methodologically inoffensive. But our concern here is not all with biographical questions of why people came to hold whatever opinions they do hold. Rather it is with the need for opinions to be suitably grounded if they are to be rated as items of knowledge, or even of probable belief. The issue is: not what does or does not need to be explained biographically; but where the burden of theological proof should rest.

(ii) A more sophisticated objection of fundamentally the same sort would urge that our whole discussion has been too artificial and too general, and that any man's enquiries have to begin from wherever he happens to be: "We cannot begin with complete doubt. We must begin with all the prejudices which we actually have.... These prejudices are not to be dispelled by a maxim" (Peirce, C.S., Collected Papers, Volume V, pp. 156-157). Professor John Hick has urged, in Theology Today for 1967: "The right question is whether it is rational for the religious man himself, given that his religious experience is coherent, persistent, and compelling, to affirm the reality of God. What is in question is not the rationality of an inference from certain psychological events to God as their cause; for the religious man no more infers the existence of God than we infer the existence of the visible world around us. What is in question is the rationality of the one who has the religious experiences. If we regard him as a rational person we must acknowledge that he is rational in believing what, given his experiences, he cannot help believing" (Hick, loc. cit., pp. 86-87).

To the general point drawn from Peirce the answer comes from further reading of Peirce himself. He was in the paper from which I quoted arguing against the Cartesian programme of simultaneous, systematic, and (almost) universal doubt. Peirce did not want to suggest that it is impossible or wrong to subject any of our beliefs to critical scrutiny. In the same paragraph he continues: "A person may, it is true, find reason to doubt what he began by believing; but in that case he doubts because he has a positive reason for it, and not on account of the Cartesian maxim." One positive reason for being especially leery towards religious opinions is that these vary so very much from society to society; being, it seems, mainly determined, as Descartes has it, "by custom and example". The phrase occurs, in Part II of his Discourse on the Method, almost immediately after the observation: "I took into account also the very different character which a person brought up from infancy in France or Germany exhibits, from that which ... he would have possessed had he lived among the Chinese or with savages."

To Hick it has at once to be conceded: that it is one thing to say that a belief is unfounded or well-founded; and quite another to say that it is irrational or rational for some particular person, in his particular time and circumstances, and with his particular experience and lack of experience, to hold or to reject that belief. Granted that his usually reliable Intelligence were sure that the enemy tank brigade was in the town, it was entirely reasonable for the General also to believe this. But the enemy tanks had in fact pulled back. Yet it was still unexceptionably sensible for the General on his part to refuse to expose his flank to those tanks which were in fact not there. This genuine and important distinction cannot, however, save the day for Hick.

In the first place, to show that someone may reasonably hold a particular belief, and even that he may properly claim that he knows it to be true, is at best still not to show that that belief is indeed well-grounded, much less that it constitutes an item of his knowledge.

Nor, second, is to accept the presumption of atheism as a methodological framework, as such: either to deprive anyone of his right "to affirm the reality of God"; or to require that to be respectable every conviction should first have been reached through the following of an ideally correct procedure. To insist on the correctness of this presumption as an initial presumption is to make a claim which is itself procedural rather than substantive; and the context for which this particular procedure is being recommended is that of justification rather than of discovery.

Once these fundamentals are appreciated, those for whom Hick is acting as spokesman should at first feel quite content. For on his account they consider that they have the very best of grounds for their beliefs. They regard their "coherent, consistent, and compelling" religious experience as analogous to perception; and the man who can see something with his own eyes and feel it in his own hands is in a perfect position to know that it exists. His position is indeed so perfect that, as Hick says, it is wrong to speak here of evidence and inference. If he saw his wife in the act of intercourse with a lover then he no longer needs to infer her infidelity from bits and pieces of evidence. He has now what is better than inference; although for the rest of us, who missed this display, his testimony still constitutes an important part of the evidence in the case. The idiomatic expression, 'the evidence of my own eyes', derives its paradoxical piquancy from the fact that to see for oneself is better than to have evidence (Austin, J.L., Sense and Sensibilia, pp. 115-116).

All this is true. Certainly, too, anyone who thinks that he can as it were see God must reject the suggestion that in so doing he infers "from certain psychological events to God as their cause". For to accept this account would be to call down upon his head all the insoluble difficulties which fall to the lot of all those who maintain that what we see, and all we ever really and directly see, is visual sense-data. And, furthermore, it is useful to be reminded that when we insist that knowledge as opposed to mere belief has to be adequately warranted, this grounding may be a matter either of having sufficient evidence or of being in a position to know directly and without evidence. So far, therefore it might seem that Hick's objection was completely at cross-purposes; and that anyway his protégés have no need to appeal to the distinction between actual knowledge and what one may rationally and properly claim to know.

Wait a minute. The passage of Hick which has been under discussion was part of an attempt to show that criticism of the Argument from Religious Experience is irrelevant to such claims to as it were see God. But on the contrary: what such criticism usually challenges is just the vital assumption that having religious experience really is a kind of perceiving, and hence a sort of being in a position to know about its putative object. So this challenge provides just exactly that positive reason, which Peirce demanded, for doubting what, according to Hick, "one who has the religious experiences ... cannot help believing". If therefore he persists in so believing, without even attempting to overcome this criticism, then it becomes impossible to vindicate his claims to be harbouring rational beliefs; much less items of authentic knowledge.

(iii) A third objection, of a different kind, starts from the assumption, mentioned already in Section 2 (i), that any programme to prove the existence of God is fundamentally misconceived; that this enterprise is on all fours with projects to square the circle or to construct a perpetual motion machine. The suggestion then is that the territory which reason cannot inhabit may nevertheless be freely colonised by faith: "Faith alone can take you forward, when reason has gone as far as it can go"; and so on.

Ultimately perhaps it is impossible to establish the existence of God, or even to show that it is more or less probable. But, if so, this is not the correct moral: the rational man does not thereby become in this area free to believe, or not to believe, just as his fancy takes him. Faith, surely, should not be a leap in the dark but a leap towards the light. Arbitrarily to plump for some particular conviction, and then stubbornly to cleave to it, would be -- to borrow the term which Thomas Aquinas employed in discussing faith, reason and revelation in the Summa contra Gentiles -- 'frivolous' (I(vi): his Latin word is 'levis').[4] If your venture of faith is not to be arbitrary, irrational, and frivolous, you must have presentable reasons: first for making any such commitment in this area, an area in which by hypothesis the available grounds are insufficient to warrant any firm conclusion; and second for opting for one particular possibility rather than any of the other available alternatives. To most such offerings of reasons the presumption of atheism remains relevant. For though, again by the hypothesis, these cannot aspire to prove their conclusions they will usually embrace some estimation of their probability. If the onus of proof lies on the man who hopes definitively to establish the existence of God, it must also by the same token rest on the person who plans to make out only that this conclusion is more or less probable.

I put in the qualifications 'most' and 'usually' in order to allow for apologetic in the tradition of Pascal's Wager, which I shall discuss more fully in Chapter Five. Pascal makes no attempt in this most famous argument to show that his Roman Catholicism is true or probably true. The reasons which he suggests for making the recommended bet on his particular faith are reasons in the sense of motives rather than reasons in our previous sense of grounds. Conceding, if only for the sake of the present argument, that we can have no knowledge here, Pascal tries to justify as prudent a policy of systematic self-persuasion, rather than to provide grounds for thinking that the beliefs recommended are actually true.

5. The Five Ways as an attempt to defeat the presumption of atheism

I have tried, in the first four sections of this chapter, to explain what I mean by 'the presumption of atheism', to bring out by comparison with the presumption of innocence in law what such a presumption does and does not involve, to deploy a case for adopting my presumption of atheism, and to indicate the lines on which two sorts of objection may be met. Now, finally, I want to point out that Thomas Aquinas presented the Five Ways in his Summa Theologica as an attempt to defeat just such a presumption. My hope in this is, both to draw attention to something which seems generally to be overlooked, and by so doing to summon a massive authority in support of a thesis which many apparently find scandalous.

These most famous arguments were offered there originally, without any inhibition or equivocation, as proofs, period: "I reply that we must say that God can be proved in five ways"; and the previous second Article, raising the question "Whether the existence of God can be demonstrated?", gives the categorical affirmative answer that "the existence of God ... can be demonstrated" (I Q2 A3). It is worth stressing this point, since it is frequently denied. Thus, for instance, in an article in Philosophy for 1968, Dr L.C. Velecky asserts, without, citation or compunction: "He did not prove here the existence of God, nor indeed, did he prove it anywhere else, for a very good reason. According to Thomas, God's existence is unknowable and, hence, cannot be proved" (p. 226). The quotations just made from Aquinas ought to be decisive. Yet there seems to be quite a school of devout interpretation which waves aside what Aquinas straightforwardly said as almost irrelevant to the question of what he really meant.

Attention usually and understandably concentrates on the main body of the third Article, which is the part where Aquinas gives his five supposed proofs. But, as so often, it is rewarding to read the entire Article, and especially the second of the two Objections to which these are presented as a reply: "Furthermore, what can be accounted for by fewer principles is not the product of more. But it seems that everything which can be observed in the world can be accounted for by other principles, on the assumption of the non-existence of God. Thus natural effects are explained by natural causes, while contrived effects are referred to human reason and will. So there is no need to postulate the existence of God."

(i) The Five Ways are thus at least in one aspect an attempt to defeat this presumption of (an Aristotelian) atheist naturalism, by showing that the things "which can be observed in the world" cannot "be accounted for ... on the assumption of the non-existence of God". and hence that there is "need to postulate the existence of God".

In this perspective it becomes easier to see why Aquinas makes so much use of Aristotelian scientific ideas in his arguments. That these are in fact much more dependent than is often realised on those now largely obsolete ideas is usefully emphasised in Anthony Kenny's The Five Ways. But Kenny does not bring out that they were deployed against a presumption of atheist naturalism.

Also one must never forget that Aquinas composed his own Objections, and hence that it was he who introduced into his formulation here the idea of (this Aristotelian) scientific naturalism. No such idea is integral to the presumption of atheism as that has been construed in the present paper. When the addition is made the presumption can perhaps be labelled Stratonician. (Strato was the next but one in succession to Aristotle as head of the Lyceum, and was regarded by Bayle and Hume as the archetypal ancient spokesman for an atheist scientific naturalism.)

By suggesting, a century before Ockham, an appeal to an Ockhamist principle of postulational economy Aquinas also indicates a reason for adopting such a presumption. The fact that Aquinas cannot be suspected of wanting to reach any sort of atheist conclusions can now be made to serve as a spectacular illustration of a point laboured in Section 2, that to adopt such a presumption is not to make an assumption. And the fact which has been put forward as an objection to this reading of Aquinas, that "Thomas himself was never in the position of a Stratonician, nor did he live in a milieu in which Stratonicians were plentiful" (Velecky, op. cit., pp. 225-226), is simply irrelevant. For the thesis that the onus of proof lies upon the theist is entirely independent of these historical and social facts. It is in the perspective provided by that thesis -- a thesis apparently accepted by Aquinas himself -- that we shall examine in Chapter Three and Chapter Five two famous attempts to defeat the presumption.

(ii) What is perhaps slightly awkward for present purposes is the formulation of the first Objection: "It seems that God does not exist. For if of two contrary things one were to exist without limit the other would be totally eliminated. But what is meant by this word 'God' is something good without limit. So if God were to have existed no evil would have been encountered. But evil is encountered in the world. Therefore, God does not exist."

It would from my point of view have been better had this first Objection referred to possible difficulties and incoherencies in the meaning proposed for the word 'God'. Unfortunately it does not, although Aquinas is elsewhere acutely aware of such problems. The changes required, however, are, though important, not extensive. Certainly, the Objection as actually given is presented as one of the God hypothesis falsified by familiar fact. Yet a particular variety of the same general point could be represented as the detection of an incoherence, not in the proposed concept of God as such, but between that concept and another element in the theoretical structure in which it is normally involved.

The incoherence -- or perhaps on this occasion I should say only the ostensible incoherence -- is between the idea of creation, as necessarily involving complete, continual and absolute dependence of creature upon Creator, and the idea that creatures may nevertheless be sufficiently autonomous for their faults not to be also and indeed primarily His fault. The former idea, the idea of creation, is so essential that it provides the traditional criterion for distinguishing theism from deism. The latter is no less central to the three great theist systems of Judaism, Christianity, and Islam, since all three equally insist that creatures of the immaculate Creator are corrupted by sin. So where Aquinas put as his first Objection a statement of the traditional Problem of Evil, conceived as a problem of squaring the God hypothesis with certain undisputed facts, a redactor fully seized of the presumption of atheism as expounded in the present paper would refer instead to the ostensible incoherence, within the system itself, between the concept of creation by a flawless Creator and the notion of His creatures flawed by their sins. As for whether this incoherence is not only ostensible but also actual I shall have something to say in Chapter Seven.

NOTES

1. This and some later translations from the Latin are mine.

2. See Chapter VI of Lewis Carroll's Through the Looking Glass:

"But 'glory' doesn't mean 'a nice knock-down argument"', Alice objected.

"When I use a word". Humpty Dumpty said in rather a scornful tone, "it means just what I choose it to mean -- neither more nor less."

"The question is", said Alice, "whether you can make words mean so many different things."

"The question is", said Humpty Dumpty, "which is to be master -- that's all."

3. See the paper 'Presumptions' by my former colleague Patrick Day in the Proceedings of the XIVth International Congress of Philosophy (Vienna, 1968), Vol V, at p. 140. I am pleased that it was I who first suggested to him an exploration of this unfrequented philosophical territory.

4. The whole passage, in which Aquinas gives his reasons for believing that the Christian candidate does, and that of Muhammad does not, constitute an authentic revelation of God should be compared with some defence of the now widely popular assumption that the contents of a religious faith must be without evidential warrant.

A. C. MacIntyre, for instance, while he was still himself a Christian, argued with great vigour for the Barthian thesis that "Belief cannot argue with unbelief: it can only preach to it". Thus he urged: suppose religion could be provided with a method of proof ... since the Christian faith sees true religion only in a free decision made in faith and love, the religion would by this vindication be destroyed. For all possibility of free choice would have been done away. Any objective justification of belief would have the same effect ... faith too would have been eliminated" (MacIntyre, p. 209).

Now, first, in so far as this account is correct any commitment to a system of religious belief has to be made altogether without evidencing reasons. MacIntyre himself concludes with a quotation from John Donne to illustrate the "confessional voice" of faith, commenting: "The man who speaks like this is beyond argument" (MacIntyre, p. 211). But this, we must insist, would be nothing to be proud of. It is certainly no compliment, even if it were a faithful representation, to portray the true believer as necessarily irrational and a bigot.

Furthermore, second, it is not the case that where sufficient evidence is available there can be no room for choice. Men can, and constantly do, choose to deceive themselves about the most well-evidenced, inconvenient truths. Also no recognition of any facts, however clear, is by itself sufficient to guarantee one allegiance and to preclude its opposite. MacIntyre needs to extend his reading of the Christian poets to the greatest of them all. For the hero of Milton's Paradise Lost had the most enviably full and direct knowledge of God. Yet Lucifer, if any creature could, chose freely to rebel.

1 From his 1984 book : God Freedom and Immorality: A Critical Analysis