Analytical Table of Contents (Full Text)
- Whatever some people may think, Christians do pray for things to happen. I am here concerned only with prayers for ordinary observable happenings, not for grace or for miracles. p. 86
- Christians believe, moreover, that God gives us some things not only as we wish but because we wish. This does not make prayer magically compulsive. p. 87
- But it does imply that what happens because of prayer is not something that would have been given by God anyhow. Since the prayer might not have been offered, the thing prayed for might have never come about; but it also could come about, since it did so; therefore, before it happened it had two-way contingency. pp. 88-89
- What is past at the time of a prayer cannot then have two-way contingency. So we cannot rationally pray for something to have happened; for in using the imperative of prayer we represent a situation as still to be brought about, which is incompatible with representing it as past, as a fait accompli. pp. 89-91
- ’Time does not exist from God's point of view' is a muddled way of saying that time is a delusion. But time cannot consistently be treated as a delusion; and so to treat it would be ruinous to the Judaeo-Christian doctrine of creation. pp. 91-93
- God is unchangeable; but because of changes in creatures we have to use different propositions at different times to say what God knows or can do; and God sees the world as changing, because it does change. p. 93
- It is rational to pray only concerning future contingent issues. Are there any such issues? Or are all events already determined in their causes? The applications of probability theory go to show that what we should commonly count as chance events (e.g. the fall of dice) really are contingent in relation to created causes: though they fall under the order of the Divine Reason, and this accounts for our ability to find a pattern even in chance events. pp. 93-95
- Contingency in the natural world is necessary in order that human choices may have scope to affect the natural world. It is in this realm of contingency that we may hope for prayers to be granted. pp. 95-97
- Can we say that after a prayer God brings about what he was not going to bring about before the prayer? It is not clear that saying so would involve a ‘real' change in God's will. pp. 97 99
- But the notion of ‘real' change, even as applied to creatures, is obscure. p. 99
Chapter Full Text
- Christians have a custom of praying to get things: not only supernatural things like grace and glory, but ordinary observable things like daily bread or recovery from sickness or success in an examination. I am not going to consider how this custom might be justified, or whether indeed it can be justified by reasons that do not presuppose Divine revelation; I shall confine myself to drawing certain consequences from the supposition that the practice is justified; and to answering some objections to the practice.
- It is merely odd that some people have doubted whether this is a really Christian practice. I remember a philosopher affirming in discussion that the Lord's Prayer was an example not of the ‘lower' sort of prayer in which one prays for something to happen but only of the ‘higher' sort in which one expresses acceptance of the Divine Will. He was, however, clearly wrong, even if we forget the petition for daily bread. For ‘Thy will be done' in the context of the Lord's Prayer is not expressing resignation; it is a petition that God's will may be willingly obeyed by the inhabitants of Earth as it is already by the blessed in Heaven; and like the rabbi in the Jewish story, we need only open the window to see that this Messianic hope is not yet fulfilled. Those who say it is fulfilled already are just uttering incantations, like Christian Scientists who say men are never sick or sorry.
- I shall be concerned only with prayers for ordinary observable happenings. Miracles would of course by definition be observable; but the concept of miracle is very difficult, and anyhow I have no idea what to say on the question when it is right to pray for a miracle.
- It is part of Christian tradition, not only that people ought to pray for things to happen, but also that they sometimes get things by praying for them; theologians call this ‘impetration'. Moreover, belief in impetration by prayer can certainly not be reduced to the bare thesis that what people pray for sometimes comes about, and comes about by God's will as all things do; such a reduction would, for example, make a mockery of Christ's emphatic words about prayer in the Gospels. Even if the duty of petitionary prayer could be established by natural reason, I at least do not see how the impetratory role of prayer could be thus established. But Christians, who rely on the word of their Master, are confident that some prayer is impetratory: that God gives us some things, not only as we wish, but because we wish. I shall do my best to clarify these causal propositions, and to remove certain difficulties in the way of our accepting them.
- The doctrine that prayer is impetratory must be sharply distinguished from a magical view of prayer. In Southey's poem the wicked Maharaja Kehama manages to sacrifice a hundred stallions in a ritually flawless way on a hundred successive days, and thereby makes it impossible for the gods not to grant his wishes. It is certainly not open to a Christian to think of impetration like that; nor need he feel himself pushed towards such an idea. It makes sense to approach God in the style of a petitioner only if one conceives of God as a rational agent who acts by free choice. Now if a rational agent does what is asked of him, this does not mean that he lacked freedom to have acted otherwise; yet we may truly say that he acted as he did because he was asked. To say that God brought something about because of a man's prayers is not at all to say that, once the prayers had been said, God could not but grant them; for this is not at all what we mean when we use similar language about petitions men address to other men.
- Let us, however, consider a certain implication of this ‘because' language we use about human granters of petitions. If a man would have done something, was going to do something, even unasked, then we cannot compatibly with this say he did the thing because he was asked. A father, let us say, has already decided to give his son a pony for his birthday, and so we can say he was anyhow going to give him a pony; we then cannot compatibly with this say that the father gave the pony because the boy asked for it. At least, this is not quite right; as was remarked to me, the father might have decided as follows: ‘I'll give the boy anything he asks for, within reason; but if he doesn't ask for anything, I'll give him a pony'. In this case the two propositions could both be true—that the father would have given the boy a pony even if he hadn't asked for anything, and that the father did give the boy a pony because he asked for it. But even so we can say these two propositions are incompatible:
… (1) The father was going to give the boy a pony for his birthday whatever the boy asked for or didn't ask for:
… (2) The father gave the boy a pony because he asked for one.
- It is this implication that raises difficulties about the impetratory role of prayer. In speaking of a human granter of petitions, I have been switching about between ‘he was going to give' and ‘he would have given', as though it were all one which expression I used. I think this is justifiable, but there are problems here; and for the moment I don't want to discuss how we can use tensed language about the eternal God, let alone such a philosophically puzzling tense as the past future, ‘was going to'. (This is only a postponement; we shall find that in this discussion we cannot shirk the problems of tensed language about God.) But even if we stick to the ‘would have' way of speaking, I submit that we get a straight incompatibility between ‘God brought about situation S because of X's prayers' and ‘God would have brought about situation S regardless of X's praying or not praying'. Or, to put this another way: ‘God brought about situation S because of X's prayers' implies ‘If X had not prayed, or had prayed otherwise, God would not have brought about situation S'.
- The logic of such subjunctive conditionals is notoriously obscure. But one thing seems clear: from ‘It could have been the case that p' and ‘If it had been the case that p, it would have been the case that q' there follows ‘It could have been the case that q'. Now whether and how a man prays is within his own power: in so far as this is not so, if a man cannot help praying—e.g. if he is hypnotized or brainwashed into uttering the words of a prayer, or if he utters them in some brevis furor of anger or fear—then the prayer just does not count as a prayer. But if the man had not prayed then God would not have brought about the situation S that he prayed for, and therefore S would not have come about. So S could have never come about; but also S could come about, because S did come about—the prayer was granted. The upshot is that if we are to be justified in saying that a state of affairs S came about from somebody's impetratory prayer, then at the time of the prayer S must have had two-way contingency: it could come about, it could also not come about.
- The first and most obvious conclusion from this is that there can be no impetratory prayer in regard to things already past at the time of the prayer. The contrary opinion, along with various other theologically suspect ones, was held by Gerard Manley Hopkins. In his verse, at any rate—as he made plain in his explanatory letters to Robert Bridges—he had prayers for people not to have gone to hell, though he firmly believed it was no good praying for people already in hell:
And that prayer thou hearst me making
Have, at the awful overtaking,
Heard: have heard and granted
Grace that day grace was wanted.
Not that hell knows redeeming,
But for souls sunk in seeming
Fresh, till doom-fire burn all
Prayer shall fetch pity eternal.
I do not know Hopkins's prose remains, so I do not know whether he offered any rationale of these ex post facto prayers. Anyhow, such prayers relate to the granting of Divine grace, which as I said I cannot discuss here. But we find a rationale of ex post facto prayers expounded in C.S. Lewis's works, The Screwtape Letters and Miracles; and he is concerned with observable past issues, such as: whether my friend has come safely through some exposure to danger that is now over. C.S. Lewis argues that God timelessly sees the whole pattern of events in time and the whole pattern is subject to God's will: so God can shape an event that comes earlier to fit in with a prayer that comes later.
- Fortunately we need not discuss the mysteries of God's eternal knowledge and power in order to see that C.S. Lewis's theory will not do. For it is not a matter of what God knows or can do, but of what we in prayer can sensibly say. If a prayer does not make good sense, then equally it does not make good sense to speak of God's granting it. And a prayer certainly does not make sense if we try to use a past tense, of the imperative mood and pray that a state of affairs may have come about. In using the imperative we represent the situation as still to be brought about, and in using the past tense we represent it as already a fait accompli one way or the other. These representations will not fit together; such a prayer makes no better sense than a schoolboy's prayer for pi to have the value he gave it in his maths test--3.1416 or 3 1461, as the case may be.
- C.S. Lewis admits that it is no good praying for a thing not to have happened when we know it has happened; for in that case we should be setting our own will against God's will as manifested in the actual issue. But if my own objection to praying about the past is a sound one, then it is not a matter of our ignorance, any more than it is a matter of God's omniscience. A prayer for something to have happened is simply an absurdity, regardless of the utterer's knowledge or ignorance of how things actually went; just as the prayer for pi to have a certain numerical value is absurd, whether or not the schoolboy has learned the actual value of it after doing the test.
- I am not arguing that an imperative in the perfect tense never occurs in ordinary language or is necessarily meaningless. As Hopkins pointed out in a letter to Bridges, this would be quite wrong; we get such an imperative when a tutor says to his pupil ‘Have your essay written by Tuesday morning'. But this does not make sense of the sort of imperative that would be used in praying about the past. The indicative corresponding to the tutor's imperative is ‘You will have written your essay by Tuesday morning'; and this is not in the past tense. It is in the tense aptly called by old grammarians 'paulo post futurum' : it relates to a time somewhere between the time of speaking and the following Tuesday morning—and thus to a time which is future, not past, when the tutor speaks, though by the Tuesday morning it will be a past time.
- Now suppose the tutor had said: ‘Have your essay written by last Tuesday morning'. The undergraduate could make no sense of this; there is no coherently constructible indicative to correspond. For ‘You will have written your essay by last Tuesday morning', read so as to correspond to this imperative, implies a contradictory time-specification, of a time which is both future at the time of speaking and past by the previous Tuesday morning. And it is equally incoherent to pray to God ‘Let my friend's aeroplane not have crashed last Tuesday morning'—supposing that I've just learned my friend was flying that day, but have not learned whether he landed safely. No doubt God, who knows our necessities before we ask and our ignorance in asking, sometimes does in his mercy do something about the most confused prayer; but there's no sense in saying he grants such a prayer.
- C.S. Lewis more than hints that time does not exist ‘from God's point of view' but only exists ‘from ours'; in Miracles he has a parable of a designer who can see a whole pattern of lines on a piece of paper as simultaneous, whereas to the consciousness of a given line in the pattern the various bits of the pattern appear as successive. This way of talking is quite common in popular theology, and it is important to see why it will not work. If time does not exist from God's point of view, if God sees all things as simultaneous, then there are only two alternatives. First, that God is unaware of the temporal aspect of things, which is so important to us; much as the God of Empedocles (if we may trust Aristotle's Metaphysics) was wholly unaware of Strife, which men know only too well. I need not linger on this alternative. The other alternative is that things really are all simultaneous, just as God sees them to be, and that our consciousness of things as successive is a misperception. C.S. Lewis seems to favour this alternative; but it is equally inadmissible.
- Even if a man's impressions as to which realities are past, present, and future are illusory, the fact that he has in that case different and uncombinable illusions shows that at least his illusions are really successive—that they are not all present together, but now one illusion is present and now another. Whether entities A, B, and C really are events in succession or are only misperceived as being such, if an observer has the experience of remembering A, perceiving B to happen, and expecting C, and also has the experience of remembering that A happened before B happened and perceiving C, then these two experiences of the observer are uncombinable, and cannot occur in him simultaneously, but only successively. But in that case temporal succession itself cannot be an illusion, since the so-called illusion of successiveness is already a real succession of experiences: just as misery cannot be an illusion, because to be under the illusion of misery would be real misery. Parmenides and Mrs. Eddy alike can neither deny nor fit into the universe what they call ‘the error of mortal mind'.
- Even if there were not decisive reasons for rejecting as logically incoherent the theory that time and succession do not exist ‘from God's point of view', we should still need to notice that no Christian apologist can use this theory. For the one thing that is clear about Divine creation is that the things created have to be temporal, even if the Creator and his creative act are held to be eternal. If time is an illusion, then there is not the sort of mutable world that can be supposed to be a creation; so in this case the Judaeo-Christian doctrine of creation has to be rejected. Significantly the doctrine was rejected by great metaphysical thinkers, Spinoza and Bradley and McTaggart, who denied the reality of time. And again, if time is an illusion, this strikes at something that is central to the Christian creed: its reference to a datable event in time—the crucifixion of Christ under Pontius Pilate.
- I am not saying that God is changeable, that he has different information available at different times; but I am saying that God sees creatures as changing and as being in different states at different times, because that is the way things are and God sees things as they are. Moreover, though God's knowledge is unchangeable, we have to use different propositions at different times in order to say what God knows. As Aquinas remarked, it is wrong to say ‘Whatever God once knew he does know', if the ‘whatever' relates to propositions (enuntiabilia); for example when Socrates was sitting it was right to say ‘God knows Socrates is sitting', but it is wrong to say this now if Socrates is no longer sitting.
- Similarly, though the arm of the Lord is not made short, it is false to say ‘Whatever God was able to bring about he still is able to bring about', or ‘Whatever prayer God could have granted he still can grant'. For though whatever state of affairs is conceivably accomplishable can be accomplished by God, nevertheless when once a situation belongs to the past, to the region of fait accompli, it no longer comes under the notion of the conceivably accomplishable. Here, as in regard to what God knows, the change involved is a change in creatures not in God.
- It is idle and vain, then, to pray concerning an issue envisaged as past; we can sensibly pray only concerning future contingent issues, where things still can go either way. (It is irrelevant that a past issue once was contingent, if we know that it is now decided and there is no longer any contingency about it.) The contingency of what we may sensibly pray for arises because it is foolish to try to obtain by prayer what is either impossible or inevitable; we sensibly pray for S when S may yet come about because our prayer impetrates it, though S may also not come about—after all, our prayer does not magically constrain God.
- Now a future issue cannot be thus contingent if, miracles apart, it is already determined in its causes. I said at the beginning of this paper that I don't know in the least when one ought to pray for a miracle; let us rather consider an issue, say the result of an examination, about which it would not be justified to pray for a miracle. Suppose the examiners have completed their formalities and are bound by university statutes to make no further alterations: then the result is determinate in its causes, and at this stage it is unreasonable to pray with a view to affecting the issue, if one knows things have got thus far. We may reasonably pray about the weather if we regard the weather as future contingency; but astronomical events are already determined in their causes, and if we are not praying for a miracle it is senseless to pray about sunsets or eclipses.
- The distinction I have just drawn, between issues already determined in their causes and contingent issues, was of course regularly drawn by medieval philosophers. The actual place where they drew the line was fixed by obsolete Aristotelian physics and astronomy. But the distinction itself was sound; at least, so it appears to me.
- Men of science have indeed often asserted that ‘in principle' the movements of the stars and of the atmosphere are on the same level of predetermination, and that the appearance to the contrary is just our ignorance. Well, they've kept saying this since Queen Victoria's days; but a Shropshire farmer still acts with good sense if he treats the weather forecast on the radio a bit less seriously than the time of sunset printed in his diary. As for the matter of 'principle': even in principle the weather is not predictable from material causes, for it is notoriously affected by the voluntary actions of men; the rainfall of the Holy Land, for example, has been affected by its deforestation in past centuries under the Arabs and its reafforestation under the Jews. So there is not the least reason to distrust the prima facies of contingency in the weather.
- This is, of course, by no means the only example of prima facie contingency. And there is solid reason for believing that the appearance is not delusive. Upon the whole, men's conjectures as to the ways of things in the world—e.g. as to the laws of motion, the nature of the stars, the shape of the Earth, the interactions of substances, the workings of the human organism—have begun by being extremely wide of the mark and have long continued to be so. But now consider what happened when a serious theory was first devised for the kinds of events that have immemorially been reckoned as chance events. The theory of probability arose, Voltaire tells us, from a problem posed by an homme du monde to an austere Janseniste. The Chevalier de Mere, a seasoned gambler, had observed that there is a slightly worse than even chance of getting a double six in twenty-four throws of a pair of dice, but a slightly better than even chance of getting a six in four throws of a single dice. Pascal's calculations, based on assumptions about equal possibilities, exactly agreed with this experience1. Again, being kicked to death by a horse is just what people would count as a matter of luck, bad luck of course: and during the lifetime of the Hohenzollern Empire German cavalrymen perished in this way strictly in accordance with the laws of probability. People will account for such facts by a vague idea that ‘it's logically bound to work out in the long run'. This supposed logic, as Von Wright has elaborately proved, is a tissue of fallacies; anyhow, the fact to be explained is agreement in the short run—that is what makes probability theory practically important; in the long run we are all dead, like the Hohenzollern Empire and its cavalrymen.
- There is massive evidence, then, that classes of events that would naively be counted as chance events really do conform to an a priori standard of contingency, in which now this, now that, possible alternative is realized. Such events, a theist may hold, are in truth not determined by any created causes, but do not on that account escape the order of Divine Providence. ‘Lots are cast in the lap: but they are disposed of by the Lord.' Our reason, seeking a pattern in chance events, is able to find one even there, because these events too fall under the order of the Divine Reason of which our own is a faint reflection.
- The empty claim that all events in the material world are determined in their causes and are ‘in principle' predictable has had a deplorable influence on Christian apologetic; although men of science today are less confident about this claim than a man like Tyndall was, the mischief has already been done by their predecessors. It is often held with absolute certainty that events in the material world (at any rate large-scale events) are determined in certain created causes, so that (to use C.S. Lewis's example) the fair weather at the evacuation of Dunkirk was determinate at the formation of the solar system. Sometimes this certainly is based on a superstitious belief in the predictive feats of science—really, of course, a physicist cannot predict how dice will fall any better than I can. Sometimes, people appeal to a supposed metaphysical truth, ‘the principle of causality2', which is alleged to rule out two-way contingency. Those who thus swear by ‘the principle of causality3' are rarely forthcoming with any attempt at an exact statement of it; but I well remember one of them having a shot at this in discussion, and coming up with a formulation essentially like Spinoza's, which had the undesired consequence of ruling out all free choice, Divine as well as human.
- We often find Christian apologists committing themselves to strict determinism as regards the material world (at least in its large-scale aspects) and then by consequence to the desperate task of explaining how God, while ‘binding Nature fast in fate Left free the human will'. Pace Pope and Leibniz, I think the thing cannot be done. If Nature is bound fast in fate, then the human will is a chimera buzzing in a vacuum and feeding upon second intentions. How can I have any freedom of speech if the sound-waves impinging on your ears as I speak are determined in material causes going back to the origins of the solar system and having nothing to do with my thoughts and intentions? It is indeed a presupposition of rational human action that there should be a great deal of determinism in the material world; otherwise, as the Warden of Keble has remarked in his Gifford Lectures, we should be like Alice trying to play croquet with a live hedgehog as a ball and a live flamingo as a mallet —both creatures having ‘wills of their own'. But equally we could not play croquet if the balls moved like stars in their courses regardless of the players' wishes and plans.
- I argued that the impetratory role of prayer requires large-scale two-way contingency in the observable world. But this is no difficulty against Christian belief if we have good reason to admit such contingency anyhow, as I have argued that we have. There are many large-scale future events in the physical world which are contingent in regard to all created causes, and which we cannot predict (even leaving aside the possibility of miracles) from consideration of any created causes; there has to be this element of chance in things if human choices are to have any Spielraum, as they manifestly have. But such contingent events do not fall outside the order of Providence, which can arrange them so as to answer prayers. Whether and when prayers are so answered, it is not for philosophy to say.
- My final problem concerns certain tensed propositions about the Divine will that need to be discussed in connexion with prayer. When we are speaking of a man who grants petitions, we may say pretty well indifferently ‘He was going not to bring about S'—having regard to a time before the petition—or ‘He would not have brought about S'. Can we use the past-future form with regard to God? To take a Bible example: can we say that God was going not to let Ezechias live (as indeed the Prophet Isaias declared to him) but that God in fact restored him to health because of prayers? I am not at all sure what the right answer is.
- The obvious objection to this way of speaking is that it would involve that God's will is changeable. If it was first true to say ‘God wills that King Ezechias should not recover' and later true to say ‘God wills that King Ezechias should recover' then surely God's will must have changed. But really the inference is not obvious. If we are to affirm God's changelessness (and I think there is very good reason to do that) then we must reject as invalid certain inferences about God's knowledge and power similar to this one about God's will. If God's knowledge is to be changeless, then a pair of premises like ‘It was once true to say: God knows that Socrates is sitting down' and ‘It is now no longer true to say: God knows that Socrates is sitting down' must not imply that God's knowledge has changed. If God's power is to be changeless, then we must not be allowed to infer a change in God's power from the premise that ‘God can stop Miss X from losing her virginity' was once true—before she was debauched—and is now true no longer. (I take this example, like the one about God's knowledge, from the Summa Theologica.) But if we disallow these inferences, we may equally disallow the inference from ‘God was not going to let King Ezechias recover' and ‘God did let King Ezechias recover' to a change in God's will. Traditional theology explains that the apparent change in God's knowledge or power is really a change in the creature, Socrates or Miss X, not in God: may not the apparent change in God's will be likewise explained away as a change in Ezechias? For Ezechias, as men say, was first of all in such condition that he was not going to recover, and then he did recover; and his condition, first and last, was by the will of God.
- Theologians, again, are wont to say that God's will is necessary ‘in its entity, but not in its term': that is, that propositions about what God wills in the created world are not necessarily true. But at that rate one might equally say that God's will is unchangeable ‘in its entity, but not in its term': that is, that propositions about what God wills in the created world are not unchangeably true. And surely we must say this; even apart from the matter of King Ezechias; for ‘God wills to bring so-and-so about' can no longer be true when so-and-so is already a fait accompli.
- I am well aware that we cannot hope in this life to resolve all such problems; but part of our difficulty here arises not from the mystery of God's eternal life, but from our inadequate concept of change. The only sharp concept of change that we have is what I have called the Cambridge concept4: viz. that A changes if of two propositions about A, relating to different times but otherwise alike, one is true and the other false. If the Cambridge concept were adequate as well as sharp, then each pair of premises that I have said does not imply a change in God does imply such change. But even in its application to creatures the Cambridge concept is clearly inadequate; for by that concept Socrates would change if Theaetetus grew and became taller than Socrates, or even if some schoolboy came to admire Socrates in the twentieth century. These, we should wish to say, are not real changes in Socrates. But I do not know of any criterion, let alone a sharp one, that will tell us when we have a real change in Socrates and not just a ‘Cambridge' change. The search for such a criterion strikes me as an urgent task of philosophy; even if you think philosophical theology is chimerical, this task is worthwhile for the general light we should gain about time and change if the search were successful; and in Christian dogmatic theology too a sound concept of real change would be a powerful tool. But I must end with this because I do not know what to say.
See "Geach (Peter) - God and the Soul (Analytical ToC)" for Geach's analysis of the rest of the book.
Footnote 1: I am told it was Fermat, not Pascal; this spoils Voltaire's epigram, but not my argument
Footnote 4: See Essay 5 (What Actually Exists).
Text Colour Conventions (see disclaimer)
- Blue: Text by me; © Theo Todman, 2020
- Mauve: Text by correspondent(s) or other author(s); © the author(s)