To recap: on Justin Taylor’s blog, a commenter called ‘Arminian’ took issue with an article by John Piper by contending that Calvinism is incompatible with the claim that our prayers can be “genuine causes” of God’s decisions about how to answer those prayers. As he put it, “the person’s request for God to do the thing cannot reasonably considered a cause of God doing the thing.” I responded (here and here) that (1) this is correct, but Piper wasn’t making that claim in the first place, and (2) it’s hard to see how our prayers could be “genuine causes” (in the sense intended by ‘Arminian’) on the classical Arminian view either. This post is an elaboration on (2).
1. Let GC be the claim that human prayers are “genuine causes” of divine decisions. I’ll assume that the modifier ‘genuine’ is designed to exclude any understanding of causation that is far removed from our normal understanding of the term, as found in claims like “The moon causes the tides” and “Frank caused the commotion in the dinner hall”.
2. Let CA be the conjunction of the two distinctive claims of classical Arminianism, viz., that humans have libertarian freedom and that God has exhaustive infallible knowledge of all human choices (past, present, and future). The question before us is whether CA and GC can both be true.
3. I take it that God is either timeless or temporal, that is, either (a) God transcends time altogether and therefore does not exist in a series of moments or (b) God inhabits time therefore and exists in an endless series of moments. In the latter case, God may or may not be time-bound prior to the moment of creation. Arminians take different views on this point, but both variants can be treated together for the purposes of this discussion.
4. Consider first the case in which God is temporal. On this view, God infallibly knows at the moment of creation not only every future human prayer but also every response He will make to each prayer. It follows that God’s decisions take place prior to the prayers. After all, it makes little sense to say that God only makes the decision at the time of the prayer, given that He already knows what He will decide! (Just try to imagine yourself in that situation: “I always knew infallibly what I would decide to do today, but now, at long last, the time has come. So, what shall I do?”)
But if God’s decision is an effect, the cause of which is a future human prayer, then the effect temporally precedes the cause. The Arminian who takes this position is thus committed to backward causation. Backward causation is counterintuitive at best and metaphysically impossible at worst; it certainly raises a number of paradoxes that are very difficult to resolve. If Arminianism commits one to backward causation, then so much the worse for Arminianism.
5. Some Arminians, following the lead of Boethius, have appealed to divine timelessness to resolve the tension between divine foreknowledge and human libertarian freedom. ‘Arminian’ seems to take this route in his first reply: “God can do this [i.e., foreknow free human acts] because he is not bound by time.”
Does this move help explain how CA and GC are compatible? I don’t think so. In the first place, our normal understanding of causation is that effects are events, that is, they are temporal occurrences. An effect is what happens because of its cause (or causes) — and happenings are events. Events, of course, take place in time. But if God is timeless, His decisions are also timeless; therefore His decisions cannot be effects (since they cannot be events). Alternatively put: if God’s decisions are effects then God isn’t timeless after all.
6. Here’s a slightly different route to the same conclusion. According to our normal understanding of causation, a cause either temporally precedes or temporally coincides with its effect. So we expect both cause and effect to be temporally related in a certain way, such that both take place in time.
This raises an immediate prima facie problem for Christians who hold that God is timeless. How then could a timeless God cause the world (including time) to come into existence? Now, I’m not raising this as an objection to classical Arminians who understand divine eternity as timelessness. After all, I agree with them on this issue! My point is that we must both concede that some causes can be timeless, even though their effects are temporal. I’d argue that our intuition that causes must be events is not nearly as strong as our intuition that effects must be events, although in this dialectical context I don’t need to. (Notably, many defenders of libertarian freedom make a point of denying that causes must always be events.)
So let’s grant that there can be timeless causes with temporal effects. The more relevant question is whether there can be temporal causes (human prayers) with timeless effects (divine decisions). If the former is coherent, why not think that the latter is equally coherent?
Here’s one reason why not. For those who take God to be timeless, time-boundedness is typically considered a limitation. (This is, after all, one of the main reasons we take God to be timeless.) It makes considerable sense, I think, to hold that a unlimited, time-transcending being has the power to affect limited, time-bound beings. Does it make just as much sense to hold that limited, time-bound beings have the power to affect an unlimited, time-transcending being? The greater has control over the lesser, but does the lesser have control over the greater? God can “reach into” a time-bound universe with His causal powers. But can I “reach out” of a time-bound universe with my causal powers?
There clearly isn’t a parity between the two. So the burden of proof, I suggest, remains firmly on the one (such as ‘Arminian’) who claims that temporal causes can bring about timeless effects.
7. Here’s a final argument against the claim that CA and GC are compatible. I’ve already noted the difficulty in reconciling CA and GC on the assumption that God is in time. But the problem doesn’t go away merely by placing God outside time, since a timeless God still has the power to cause events within time, which once again raises the specter of backward causation.
Consider this scenario. Suppose I offer a prayer to God today that somehow causes His (timeless) decision to respond in a certain way. (Remember that this is what our commenter ‘Arminian’ claims actually happens in practice.) On this view, God not only knows (timelessly) about my prayer, but His response is caused by the prayer.
Since God transcends time, it is possible (given CA) for God to have revealed both my prayer and His response to some person, S, who lived 100 years ago. Thus it is possible for God to cause S to believe some proposition like the following:
(FP) Someone in 2009 will pray X, and God will do Y in answer to his prayer.
But in that case, my prayer would be a partial cause of S’s coming to believe FP. There would be a causal chain beginning in 2009, passing through eternity (via God’s knowledge and will), and ending in 1909. We’re faced with the possibility of backward causation again.
8. The scenario can be extended so as to make the problem even more acute. Suppose that God not only reveals FP to S, but also commands S to write it down. We now have a physical piece of paper in the year 1909 with the details of my prayer and God’s response written on it. Suppose further that this piece of paper is copied again, and again, and again.
The upshot is that if human prayers can be “genuine causes” of God’s timeless decisions, then my prayer today could be (or could have been) the partial cause of dozens of written records of that prayer existing 100 years ago. In fact, one of those papers — even the original written by S — could make its way through history into my hands today. Imagine the scene as I look down at the paper in my hand which tells me about the very prayer I will offer today and God’s answer to it. I wonder, would I feel at that moment that I had libertarian freedom to not offer that prayer?
In fact, it gets worse. For I might well decide to offer that prayer today because I know I will anyhow and because I know what God’s response will be (assuming it’s favorable). But then I’d be involved in a causal loop: my prayer would be the partial cause of itself! It seems that GC and CA together commit us to the possibility of circular causation as well as backward causation.
Furthermore (and this may be the crowning oddity) if I have libertarian freedom, as CA maintains, then at the point I freely offer my prayer I have the power not only to cause all those paper records to exist (or to have existed) but also to cause them not to exist (or not to have existed) — even though, due to the very pastness of the past, whether or not they existed has already been settled.
9. It won’t do to say, by way of objection, that in the scenario above I would only have the power to make false the propositions expressed in writing on the papers. The idea here, presumably, is that if I opted to pray otherwise (or to refrain from praying) the papers written in 1909 would be physically unaltered but would turn out to express falsehoods rather than truths. The problem with this claim is that it involves the falsification of divine revelation, something no classical Arminian would want to countenance. Remember that ex hypothesi these papers were written at God’s command, as a record of God’s revelation. If I have the power to falsify divine revelation, then Judas had the power to falsify Psalm 41:9 (cf. John 13:18). This way lies Open Theism, not classical Arminianism.
10. Neither will it do to say that while my hypothetical scenario is possible, it would never actually happen (perhaps because God just wouldn’t act that way) and therefore there would never actually be a causal chain from the present to the past. There are two reasons why this response is inadequate.
The first reason is that the mere possibility of the scenario is enough to show the problem in reconciling CA and GC. If the conjunction of CA and GC implies the possibility of backward/circular causation, then (by modus tollens) the impossibility of backward/circular causation implies that CA and GC cannot both be true.
The second reason this response won’t fly is because, as a matter of historical fact, a scenario very similar to the one I’ve described has in fact happened (minus the assumption that human prayers are “genuine causes” of divine decisions). Consider this prophecy from Jeremiah:
“For thus says the Lord: When seventy years are completed for Babylon, I will visit you, and I will fulfill to you my promise and bring you back to this place. For I know the plans I have for you, declares the Lord, plans for welfare and not for evil, to give you a future and a hope. Then you will call upon me and come and pray to me, and I will hear you. You will seek me and find me, when you seek me with all your heart. I will be found by you, declares the Lord, and I will restore your fortunes and gather you from all the nations and all the places where I have driven you, declares the Lord, and I will bring you back to the place from which I sent you into exile.” (Jer. 29:10-14)
Here we find:
- An infallible prediction of a future human prayer.
- An infallible prediction of the divine response to that human prayer.
- A physical written record of those predictions, with (most likely) multiple subsequent copies of that written record.
- The possession of that written record by some of the humans who would eventually offer the prayer (see Daniel 9).
The one thing we don’t find here is the suggestion that the human prayer is a “genuine cause” of a timeless divine decision. But it’s precisely that claim which introduces the philosophical problems I’ve spelled out above. So it seems that if the classical Arminian has to drop one claim to maintain coherence, it should be GC.
11. The last argument can be adapted, of course, to apply to the case in which God is time-bound. In other words, the problems arise whether or not one takes God to be “bound by time.” The difficulties arise not because of the relationship between God and time, but because of the implications of CA and GC taken together.
12. None of this raises any problems that I can discern for Piper’s original claim that our prayers can be the causes of the answers to those prayers, even if those prayers are foreordained. The same claim holds for classical Arminianism, with its weaker notion of foreordination.
But what doesn’t stand up to scrutiny, in light of the arguments I’ve offered above, is the claim that classical Arminianism has an advantage over Calvinism in that it can accommodate the idea that our prayers are “genuine causes” of God’s decisions about how to answer those prayers.