Knowing our petitionary prayers have been answered
February 17, 2010 — 8:59

Author: Alexander Pruss  Category: Religion and Life  Tags:   Comments: 25

Assume we know theism is true. Can we theists ever tell that our prayers have been answered? I pray for E and E occurs. Can I ever know that God acted on my prayer rather than E occurring completely independently of my prayer? It turns out that the answer is simpler than one might think, and that we can know this much more often than one might think.

Consider the property of Reasons Maximalism (RM) that an agent might have. An agent has RM if and only if whenever she chooses an action A, she chooses it on account of all the unexcluded reasons she is aware of in favor of A. Suppose, for instance, that I have a duty to visit a sick friend and I enjoy her company even when she is sick, but, on the other hand, it’s a long drive and the hospital is depressing. Nonetheless, I do visit her. If I don’t have RM, I might be visiting her only out of duty or only for pleasant companionship. But if I have RM, I am visiting her because of both duty and pleasant companionship. And if I have RM and decide not to visit her, then I will decide to do that because of both the long drive and the depressingness of the hospital.

I submit that God has RM. Being perfectly morally good and perfectly rational, in every decision God takes into consideration all the unexcluded reasons he has. Of course, in the end, it may not be possible for him to act on all the reasons, because some of the reasons will pull in different ways. But his choice will have been made on the basis of all the reasons he is aware of in favor of it. Moreover, in the case of an omniscient being, the reasons she is aware of in favor of A is the same as the reasons she has in favor of A. Thus, God chooses A on the basis of all the unexcluded reasons he has that favor A.

Now, that I’ve requested something good and grantable is always a reason to grant the request. In rare cases, it will be an excluded reason–perhaps I earlier authoritatively commanded the person to stop granting my requests for a day. But I cannot think of an exclusionary reason God might have against considering our requests for good things. (If God promised not to hear our requests, that would be an exclusionary reason, but he made no such promise.)

I don’t know exactly how to analyze “grantable”. One class of non-grantables are states of affairs ruled out by divine promises. Another class of non-grantables are states of affairs that cannot be brought about, whether because they are metaphysically impossible or because they are metaphysically necessary. It may also be that people’s free choices are non-grantables. However, perhaps when we pray that x (where x is not God) might freely do A, God reinterprets our prayer charitably as a prayer that x be given lots of reason to do A, and that is a grantable. I do not know whether things that God has already promised are grantables, but I am inclined to think they are (cf. the sick friend visit case).

So, our requests for grantable good things are always an unexcluded reason for God to grant the request, and God being omniscient is aware of this. Moreover, God is a concurrent cause in all good events (in fact, in all events, because evil is a mere privation, but nevermind that), so that all good events count as caused by God. Therefore, by RM, if I pray for grantable good, and God brings about the request, then God produces the good in part because of the request. So, a sufficient condition for my knowing that an event has happened as a result of my request is that (a) I prayed for it, (b) it was good, (c) it was grantable and (d) it occurred.

In particular cases, these conditions are very commonly satisfied. If you pray for someone’s safety during a trip, and she returns safely, she does so in part because of your prayers. If you pray that you find a lost object, and you do find it, you find it in part because of your prayers. If you pray that a friend might recover from an illness, and she recovers, she recovers in part because of your prayers.

Now, you might say that because of the “in part” this is unsatisfying. You might want to know when it is that God grants it solely on account of your prayers. Assuming the thing you prayed for was good, the answer is: never. If it was good, then God had a reason to bring it about, and by RM if he brought it about, he brought it about in part because it was good. The one exception would be if there were an exclusionary reason, such as a divine promise that he will only bring this good about as a result of prayer. But Revelation does not, I think, tell us that God has such exclusionary reasons, and we can reasonably presume he doesn’t. So it is never the case that a good you prayed for was granted solely because you prayed for it.

But perhaps you want to know something else: You want to know if it is the case that the good would not have been granted had you not prayed for it? Well, sorry: this can only be known if Molinism is true and God reveals it to you. But I think Molinism is false. Given the falsity of Molinism, there will be no facts of the form: The good would not have been given had you not prayed for it. For had you not prayed for it, God would still have had the reason in favor of it given by the fact that it was good, and he still might have acted on the reason. That said, sometimes you can know that the event would still have been given–for instance, when the event was promised by God. However, you can never know that the event would not have been produced had you not prayed for it, when the event is good. (I leave open some questions about praying for neutral and bad things.)

Can we ever know that our prayers have not been granted? Maybe not. For it seems reasonable for God to grant our prayers by giving us something greater than what we prayed for, something we weren’t wise enough or knowledgeable enough to ask for. Suppose George has a flu he knows about and an undiagnosed cancer he doesn’t know about. He prays to be cured of the flu, but instead God cures the cancer. That’s better than what George asked for, and George cannot complain that his prayer has been unanswered. Any substantive good in curing the flu is there in the curing of the cancer, and if George had known he had the cancer and had any sense, that is what he would have prayed for the healing of. In that case, George’s prayer was, arguably, answered, but George cannot know how it has been answered. Though if he is a Christian and reflects on Scripture, he can know that it has been answered somehow. And of, course, healing faults in the soul would be even better than curing the flu or healing a cancer.

None of the above provides an atheist with a reason to think prayers have been answered. But that wasn’t the question I was asking.

  • Joshua Rasmussen

    Those are helpful considerations.

    February 17, 2010 — 12:54
  • Kevin Timpe

    “So, a sufficient condition for my knowing that an event has happened as a result of my request is that (a) I prayed for it, (b) it was good, (c) it was grantable and (d) it occurred.”
    I’m not so sure about this. Say I pray that the sun rise tomorrow (or, more properly, that the earth continue to rotate so that it appears that the sun rises tomorrow). I take it that such a thing is good and grantable. Lo and behold, come tomorrow, the sun rises.
    This doesn’t seem to happen as a result of my request if it would have happened even had I not prayed. For in that case, it looks like my prayer is irrelevant to the sun’s rising.
    I’ve worked with this definition of impetrations before:
    (i) the prayer is offered by an agent A at time t2 ;
    (ii) the prayer requests that God bring about some state of affairs S at time t1;
    (iii) the prayed-for state of affairs S is brought about by God, at least in part, as a result of A’s prayer; that is, God’s knowledge of A’s prayer is one of the reasons He has for bringing S about; and
    (iv) God desires to bring about S only if A prays for S, such that if A does not pray for S, then God will not bring it about.
    You seem to reject (iv), in part because it requires Molinism. I don’t see that it requires Molinism, but it does require that there be certain counterfactuals about what God would do in various circumstances. Why can’t there be such facts, even if Molinism is false?
    And in what sense is the prayer ‘answered’ if God would have done the exact same thing even had the prayer not be given?

    February 17, 2010 — 18:28
  • Mike Almeida

    And in what sense is the prayer ‘answered’ if God would have done the exact same thing even had the prayer not be given?
    Here’s the sense. According to (iv), A’s prayer cannot be a reason for God to (want to) bring about S if the prayer is such that, had A not performed it, God would have brought about S anyway. But that can’t be right. Suppose I poke Smith in the arm with a knife, and Smith cries. Clearly, the reason Smith cried was because of what I did. But suppose it is true that, had I not poked Smith in the arm, he would have cried anyway. Does that mean that what I did was not the reason that Smith cried? I don’t think so. Suppose that had I not poked Smith in the arm, Jones would have done so. True, Smith would have cried anyway, but it hardly follows that the reason Smith cried had nothing to do with what I did. Similarly for God and prayers.

    February 17, 2010 — 19:16
  • Mike,
    Your examples are cases of causal overdetermination. There are a couple of kinds of overdetermination that I can imagine. One is where X causes Y, where had X not caused Y, Z would have caused Y. Another is where both X and Z are independently sufficient for causing Y, both are instantiated, and both take place in a causal chain that causes Y.
    Pruss’s picture requires that God have multiple simultaneous independently sufficient reasons for doing the same action. But it seems that for a petition to truly be “answered,” the petition should be causally necessary, not just sufficient. Otherwise, I can find all the good things God is likely to do without my asking, ask for them, and then say God answered all my prayers. That seems to just miss the sense of “answered.” But making prayers causally necessary is, it seems to me at least, more accommodating to the concept of prayer. I think this way of phrasing matters avoids Pruss’s worries about Molinism.
    If Pruss is correct, then a prayer can’t be sufficient for God’s action (the action must be independently good, for example), nor can it be necessary (God already will have sufficient reasons for doing it).

    February 17, 2010 — 21:27
  • Kevin Timpe

    I agree with the second of the two cases you present. Overdetermination makes counterfactual accounts difficult, as I note in a footnote in the article of mine I refered to earlier.
    But I’m still not sure about the former case. Say that Sally is has apendicitis. I poke Sally hard in the arm. (And no one else does, nor is there anyone else waiting in the wings to poke her iff I do not, etc…) She cries out in pain. Is my poking her the reason why she cries? It’s still not clear to me that it is, if the apendicitis is reason enough to bring her to cry.

    February 17, 2010 — 21:33
  • Kevin:
    a. Overdetermination, by causes or reasons, is possible. If an action is overdetermined by reasons R1 and R2, then the agent would have done the action even had R2 not been there, but nonetheless it is correct to say that R2 was one of the causes or reasons for the action.
    b. Suppose in your case that the following two counterfactuals are true: (i) Were God not to have wanted to maintain uniformity between the past and the future (replace this with your favorite description of the reason he makes the sun come up everyday), he (still) would have made the sun come up (because of your prayers); (ii) Were you not to have prayed, he (still) would have made the sun come up (because of his desire for uniformity). In this case, things are symmetric, and it is just as true that God made the sun come up because of the desire for uniformity as that he made it come up because of your prayer. But the possibility of this conjunctive case (at least, it’s possible given Molinism, which for the sake of this argument I’m assuming) shows that (ii) is quite compatible with his having made the sun come up because of your prayer.
    c. I don’t know for sure whether the sun’s coming up is a grantable. This is because it might be argued that the sun’s coming up is something he’s committed to doing regularly (until the eschaton?), and things that God is committed to doing maybe aren’t grantables. Maybe the fact that one has promised to do A not only excludes reasons against doing A, but excludes all other reasons for and against A? I don’t know.
    d. I agree that some counterfactuals of freedom can be true if Molinism is false. I actually think I even have a story I can tell about which ones (cf. principle (**) in this post). However, I have a strong intuition that these ones aren’t it. If God in fact did A for R1 and R2, but R1 by itself could have sufficed to justify God in doing A, then on my view, the following might-counterfactual will be true: Were God to have had only R1, he might still have done A on its account. And so (by the might/would duality): not(Were God to have had only R1, he would not have done A).

    February 17, 2010 — 21:37
  • It seems plausible that if I poke someone who has appendicitis, and she would have cried out even had I not poked her, then it is false that the reason she cried out was the poke. But it is true that a reason she cried out was the poke. She cried out because of the poke. And she cried out because of the appendicitis.
    And cases of prayer for goods are always, or at least typically, cases of overdetermination, because God has independent reason to bring the goods about.

    February 17, 2010 — 21:41
  • Note, also, that there is typically another form of overdetermination always present. My friend suffers from disease D. I pray that she be healed. But… I am never alone in praying for her. After all, there are tons of people who pray general prayers like: “God, please send your healing on all those who need it.” (May he indeed!)

    February 17, 2010 — 21:43
  • Kevin Timpe

    I see now that overdetermination is much more of a than I’d earlier thought (particularly for b in your first post and then your second and third posts). Thanks for that.
    I wonder what you’d say about the following case. Assume that RM is true of me. I buy my son a desired object, X, on Monday, for the reason that I think it would contribute to his happiness; but I do not plan on giving it to him until wednesday. On Tuesday, Jameson asks me to buy, at some point, and give him X on wednesday. On wednesday, I give him X. Was his asking a reason of my giving him X?

    February 17, 2010 — 22:28
  • Kevin Timpe

    Slight correction: “…much more of a PROBLEM than I’d earlier thought…”

    February 17, 2010 — 22:45
  • Russ Dumke

    It seems to me that one’s praying (given that it is the case that I request a good that God was disposed to grant anyway) is irrelevant, because that which is prayed for would be granted without one’s praying. Furthermore, I think that if the might/would duality were granted, those who perform petitionary prayer per the strictures you delineate would determine God’s actions such that God’s omnipotence would be abridged.

    February 17, 2010 — 22:58
  • Kevin:
    In the example, his asking probably was a reason. After all, you were still free not to give it to him. You could have put it on craigslist instead.
    Suppose you add: you made up your mind when you bought it, and you no longer had a choice about it. Then, his asking probably wasn’t a reason.
    It may be that you get more in the way of non-trivial relevant counterfactuals about God’s actions in the case of open theism than in the case of a God who knows future free acts. For the God of open theism may permanently make up his mind earlier. So then we might distinguish two cases.
    Case 1: At t0, God decides to give you G at t2, and unchangeably makes up his mind. At t1, you ask him to give you G at t2.
    Case 2: God doesn’t decide until t1. At t1, you ask him to give you G at t2, and he there and then decides to do it.
    In Case 1, had you not asked him to give you G, he’d still have given it to you. In Case 2, had you not asked him to give you G, he might not have given it to you. So you can distinguish the two cases with counterfactuals, and say that in Case 2, your prayer was more relevant.
    This differs from what you did with your counterfactual. The counterfactual you proposed was: “If you had not asked, God would not have given it.” In Case 1, that counterfactual is false. But it’s also false in Case 2, assuming God has RM. For at t1, in Case 2, God acts on both the value of G and the request, and so there are no grounds (apart from Molinism) to state “If you had not asked, God would not have given it.” Instead, you have the weaker claim: “If you had not asked, God might not have given it.”
    So we might propose, as a weaker account, that the might-counterfactual has to hold. The open theist, then, will have cases like Case 1 and Case 2 which differ in the might-counterfactual, but I don’t know that the non-open theist is going to have such cases.
    But all this assumes that the open theist can have cases like Case 1 where God unchangeably makes up his mind. The only cases like that I can think of are ones where God makes a promise. But it doesn’t seem likely that God very often makes relevant promises of this sort, so Case 1 will be fairly rare–most of the time we’ll have Case 2, so we can assume that most of the time if we pray for G and we get G, then we got G because of our prayers.

    February 18, 2010 — 0:50
  • Ross:
    To give an agent a reason to act a certain way is not the same as to determine the agent to act in that way.

    February 18, 2010 — 0:50
  • Mike Almeida

    Joshua, you write,
    But it seems that for a petition to truly be “answered,” the petition should be causally necessary, not just sufficient. Otherwise, I can find all the good things God is likely to do without my asking, ask for them, and then say God answered all my prayers
    Seems mistaken. The following’s perfectly possible. I pray for X and because I did so, God gives me X. My prayer was answered. Had I not prayed for X, my friend would have done so for me, and I would have recieved X. God would then have answered his prayer, not mine. So, to get a prayer answered it is not necessary that my failing to pray would have resulted in my failing to recieve X. As far as I can tell, that’s contrary to Kevin’s condition (iv) and your claim above.

    February 18, 2010 — 7:15
  • Mike Almeida

    But I’m still not sure about the former case. Say that Sally is has apendicitis. I poke Sally hard in the arm. (And no one else does, nor is there anyone else waiting in the wings to poke her iff I do not, etc…) She cries out in pain. Is my poking her the reason why she cries? It’s still not clear to me that it is, if the apendicitis is reason enough to bring her to cry.
    We’ll run into big problems if we try to retain condition (iv), I think, and not merely overdetermination cases. There are underdetermination cases too. Suppose Jones is connected to a device that conveys electric shocks to him. Each of us turns the dial on the machine up a single notch. It is true of each of us that had we not turned the dial upward, then Jones would still have suffered the same perceptible pain (I assume that ‘perceptible pain’ is redundant, but I guess we can argue about that). So none of our actions is the reason that Jones is suffering. But certainly we do not want to say that no one did anything wrong: after our actions Jones is in extreme pain! But, given your conditions, we cannot say that what the group of us did was wrong (or brought about the pain). Let G be the conjunction of our actions. According to (iv), what we did was the reason for the pain only if it is true that Jones would not have perceived the same pain had ~G been true. But that’s false. He would have perceived the same amount of pain, had ~G been true. We are forced to conclude that what we did together is not the reason for the pain.
    Obviously, we want that conclusion to be false in the case of prayer. We want it to be the case that what we did together, namely collectively pray or G, by way of praying for an outcome is efficacious. We want that to be true even if it is false that, had some of us failed to pray (i.e., had ~G been true) we would have gotten the same outcome. Otherwise we have no individual prayer as efficacious and no collective prayer as efficacious, in the sorts of cases I’ve described.

    February 18, 2010 — 7:33
  • Mike,
    I guess I agree that there are some situations where an overdetermined event is an answer to prayer.
    But wouldn’t you say there is something funny about my other scenario, that “I can find all the good things God is likely to do without my asking, ask for them, and then say God answered all my prayers”?
    I don’t a good theory will include overdetermined answers in your example, but exclude the answers in mine. Surely there is some middle ground?

    February 18, 2010 — 8:28
  • Heath White

    Kevin, if you buy a toy for your kid on Monday, and he asks you to buy it on Tuesday, then you should not put the toy on Craigslist. You’ll get your full refund if you just return it to the store. 🙂

    February 18, 2010 — 8:39
  • Heath:
    Thanks for correcting my unhelpful advice.
    One issue that we haven’t discussed is that even though there may not be appropriate true counterfactuals, there may be conditional probabilities one could use. So maybe one could talk about the cases where the request significantly increased the probability of the gift, and say that those more fully count as “answers”. This is going to give a continuum, since the probabilities are on a continuum, and that’s not too satisfactory.
    We might in some cases be able to to use inductive data to estimate probabilities. For instance, the instant-eye-regrowth rate for one-eyed patients not prayed for by anybody is close to zero (well, as observed, it’s probably exactly zero). Granted, God has the ability to do this without being asked for it, but it seems the probability is so low that it hasn’t happened to our knowledge. So when Padre Pio prays for someone and their eye comes back, it seems reasonable to say that his prayers significantly raised the probabilities over the base rate.
    But strictly speaking even in these cases what you learn is not that the prayers have been answered, but how significantly the prayers have contributed.
    Take “answered” in the ordinary conversational sense of the word. I ask you what you were doing yesterday. You have two reasons to tell me–one, you really want to brag about it, and two, you have a reason to answer my question. So you tell me. Even though your resulting speech is overdetermined, I think we would still count it as an answer. And I think we would still count it as an answer if the increase in probability due to the question being asked was small (I am inclined to think that the probability that a good agent would do something also goes up as one adds reasons that the agent is aware of). But we wouldn’t count it as an answer if RM is false for you and you do not take the fact that I asked you into account at all (in which case the probability stays unchanged).

    February 18, 2010 — 11:50
  • Mike Almeida

    But it seems that for a petition to truly be “answered,” the petition should be causally necessary, not just sufficient. Otherwise, I can find all the good things God is likely to do without my asking, ask for them, and then say God answered all my prayers.
    Is this the problem you’re referring to? I guess I’d reiterate that, as a perfectly general requirement, the causally necessary condition seems false. So what if you can find all of the good things without asking? Does that mean that you’re asking was not efficacious? No, it doesn’t. That view generates a reductio.

    February 18, 2010 — 14:46
  • Russ Dumke

    I’m not sure I see any need to provide God with reasons to perform certain acts. The argument you offered, remember, is an argument for the efficacy of petitionary prayer. In simple language, the argument concludes that asking God for a certain class of things is an effective way to secure those things–we can know that our prayer has been answered. I wish to propose an alternative, however. Let us say that there is a something X that is good for us. God is omniscient, so he knows X. Furthermore, God is omnibenevolent and wants us to receive X. Let us also set the condition that there is no reason why we should not receive X. Let us then simply have God notify us is some suitable way that we have received or will receive X. Our petition is unnecessary.
    Strictly speaking, petitionary prayer is unnecessary because God knows our needs, and because God’s knowledge is perfect whereas ours is not, God knows our needs perfectly. And, because God is omnibenevolent, he will always do what is in our best interest. We can reliably count on him to do so. So why ask God for anything, whether we can know our request has been answered or not? If you try to counter this with a free will objection (God desires or requires our petition), then you run into an objection that is similar to my original one: God is somehow limited by our free will. But it sounds dubious to say that God’s doing good things for us is conditional upon our requesting them. So in my proposal God may additionally choose to communicate his activity wrt X to us, or he may simply do X. I suppose we could argue about the utility of his communicating with us. But I submit that my proposal is simpler. The bottom line is that I do not see how we might, in prayer or otherwise, give God a reason to act that God would not already possess.
    There is another problem as well. You say we can give God reasons to act. Is it possible that we could give God a reason that he could not refuse to act upon? Is there a class of reasons that might satisfy this condition? Then it is indeed possible to certainly know that our prayer will be answered. In this case, we can manipulate the God. I strongly suspect that this is not the way the divine works, however.

    February 18, 2010 — 15:15
  • Mike,
    I guess I just think there’s something problematic about finding all the things God will do anyway, praying for them, and then claiming that God has answered your prayers. If I’m the only person with this intuition, oh well.

    February 18, 2010 — 18:25
  • JB:
    Maybe that intuition can be handled by my probabilistic suggestions. In that case, the probability was pretty high already, so you didn’t raise it by much. So even though God did answer your prayer, his action did not come much from it.
    Your argument is assuming that there is a unique best for God to do, and that it is independent of what we ask. In regard to independence: if asking for something provides the person asked with a reason to grant it, it may change what is in fact best. In regard to uniqueness: there is incommensurability all over the place. In many cases, thus, there may not be a unique best for God to do, because incommensurable values pull in different directions. For instance, God always has a reason to heal a sick person, because health is a good. But God also, at least typically, has a reason to refrain from producing a miracle, because the nomic uniformity of the world is a good. The good of health and the good of nomic uniformity are incommensurable, and neither can be said to be objectively better than the other.
    Furthermore, there is a good in praying, since by praying we become, to a very little degree, cooperators with God. God heals George. If we prayed for it and if RM holds, then we are (partial) causes of that healing. That is a good thing.

    February 19, 2010 — 8:58
  • Mike Almeida

    I guess I just think there’s something problematic about finding all the things God will do anyway, praying for them, and then claiming that God has answered your prayers. If I’m the only person with this intuition, oh well.
    Maybe there’s an ambiguity in your concern. I agree that it would not be an answering of your prayers if it were true that had you failed to pray for X, and no one else prayed for X, and no one else gave you X, you would have recieved X.
    But suppose someone tells you that if you would please stop that praying nonsense, he’ll give you the baseball glove you’re asking God to deliver. You pray to God for the glove anyway, and he answers your prayer. That’s true despite the fact that, if you hadn’t prayed, your friend would have provided the glove. That is, you would have gotten the glove anyway.

    February 19, 2010 — 10:44
  • Incidentally, Leibniz would seem to attribute RM to God in order to secure knowledge of God’s purposes. See Discourse on Metaphysics 19 (and also several passages in the Theodicy). I take it that Leibniz is at least in part responding to Descartes’s teleological skepticism in the Fourth Meditation. Descartes says we can’t know God’s purposes because God is so far beyond us; Leibniz responds that God is (at least indirectly) responsible for everything that happens and God doesn’t do anything by accident, so every genuine good which occurs is one of God’s purposes. This strikes me as quite similar to your argument.

    February 25, 2010 — 1:08
  • This sounds like Leibniz.
    I also think RM either follows from or is made very plausible by divine simplicity. If RM is false, it seems to be possible that there be two worlds, in both of which creation is exactly the same, but in which God acts for different reasons. But then the difference between the two worlds must lie in God’s intrinsic properties. But God’s intrinsic properties cannot differ between worlds given divine simplicity. In fact, it is this that led me to RM.

    February 25, 2010 — 8:12