The following claims seem to be true:
(1) God not be in any way caused to act by anything outside of himself.
(2) Providence requires that God in his actions respond to events in the world.
If one adds the plausible claim:
(3) If God responds to events in the world, then these events are partial causes of his action
one concludes that God cannot have both aseity and providence. This conclusion, however, is theologically problematic: aseity seems to follow from transcendence, and providence is affirmed by all the major monotheistic religions.
Strong Sovereignty (the doctrine that God determines the truth value of every contingent proposition) and Molinism can be seen as providing ways to deny (2). One could also take them as ways to reinterpret (2) in a way that denies (3). I don't think it matters which option one takes. Anyway, the idea is that given Strong Sovereignty, God simply decides everything ahead of time in such wise that he has no need to "respond" to events, since all the events are always already part of his plan. Strong Sovereignty has a difficulty, however, with free will and with the deductive problem of evil. Molinism, on the other hand, holds that God knows prior (I always understand "prior" as "prior in the order of explanation" here) to any decision what to create how any stochastic processes (including free choices) would turn out in any possible circumstances, and using this knowledge, he can make a complete plan of creation without needing to "respond" to events. Thus, God knows that
(*) were Adam and Eve placed in the garden, they would freely sin.
Thus he does not need to "respond" to their sin with their expulsion. He can simply strongly actualize Adam and Eve in the garden, and strongly actualize their expulsion, knowing that they would sin. Molinism faces two major problems: (a) that the subjunctive conditionals like (*) arguably make no sense, and (b) the problem that Robert Adams raised that it appears to commit one to explanatory loops.
In this post I want to describe a model of creation developed in discussion with Grant Matthews and Sarah Coakley, based on some ideas of theirs, that offers a way to reconcile aseity with providence without making use of Strong Sovereignty or Molinism. I think the model has theological problems, which may be insuperable, but it should be on the table. Interestingly, the model does not even require foreknowledge (though, of course, I believe in foreknowledge).
First, consider a simple foreknowledge based account of providence. God sees what people will do, and responds to these doings, sometimes ahead of time (e.g., inspiring a prophet to prophecy the doings, or setting up another chain of events that harmonizes with or counters the doings) and sometimes simultaneously. It really seems, now, as if the doings of people cause God to act in certain ways. Maybe there is a way of avoiding this within the bounds of a foreknowledge account, but it is not clear how.
But now consider a different approach. A chess grandmaster plans out responses to potential moves of the other player, and does so some distance into the future. Imagine that, prior to creating, God considers all the possible ways that stochastic processes--especially human free actions--could go, and plans out a response for each one. Moreover, just like the grandmaster when playing against a weaker opponent will be able to have a plan such that whatever the opponent does, a satisfactory end-state will be reached (viz., the grandmaster wins), so God will be able to have a plan such that regardless of how the stochastic processes turn out, God's plan will lead to a result satisfactory to God. (This is, I think, how far the Coakley-Matthews account goes. I think they both endorse it so far, but it is the stuff that comes now that they probably do not endorse. Nor do I really.)
When God creates, he strongly actualizes some states of affairs, and others flow out of the strongly actualized states of affairs and the outcomes of stochastic processes, and these are called "weakly actualized". Thus, God may strongly actualize Adam and Eve's being in the garden, but their sinning follows from their free choice, and is only weakly actualized.
Without invoking Molinism or even foreknowledge, we can say that God knows all the possibilities: all the ways things could turn out. Suppose, then, God strongly actualizes certain initial states of affairs (e.g., Adam and Eve being in the garden, the Big Bang, etc.), and also strongly actualizes a large bunch of subjunctively conditional states of affairs specifying what happens in the case of any possible set of outcomes of stochastic processes.
Thus, God might strongly actualize facts like (actually, the antecedents and consequents would be much more detailed, of course):
- Adam and Eve's being placed in the Garden.
- The fact that were Adam and Eve to sin, they would leave of the garden and be unable to return.
- The fact that were Adam and Eve to sin, a means of salvation would be available to them.
- The fact that were Adam and Eve not to sin, they would remain in the garden.
- The fact that were Adam and Eve not to sin, various other goods would befall them.
- The fact that were Peter to betray Jesus, Jesus would have predicted a betrayal.
- The fact that were Peter to betray Jesus, Jesus would not have predicted a betrayal.
- The fact that were Jones to lose all his money in the indeterministic casino, he would feel inspired to rethink his past life.
- The fact that were Jones to lose some of his money in the indeterministic casino, he would feel inspired to quit gambling.
- The fact that were Jones to win some money in the indeterministic casino, he would feel inspired to give it to charity.
- The fact that were Moses to raise his staff and stretch his hand over the waters, the waters would part.
With sufficiently many such conditionals being willed, a complete plan of action is actualized, accounting for all possible contingencies. Creaturely choices, then, determine which of the planned-for pathways are actually taken. If Adam and Eve sin, they leave the garden, but later humankind is redeemed. If not, something else happens.
This is compatible with providence: every contingency is provided for, and God, like the grandmaster, has ensured that no matter what happens, an outcome satisfactory to him occurs.
Moreover, this seems compatible with aseity. For it is not that, say, Adam and Eve's sin causes God to kick them out of the garden. Rather, God has eternally willed, independently of how Adam and Eve would actually choose, that if they were to sin, they would leave the garden.
Now one might wonder here about the mechanics about some of these things. How does it come about that once Moses raises his staff and stretches his hand over the waters, the waters part? Is it not, surely, that in response to Moses' obedience, God parts the waters? But why say that? Only an occasionalist would feel much temptation say that when the water is heated, in response God boils the water, and even an occasionalist might balk at the "in response".
God speaks, and it is so. God wills a possible state of affairs, and the possible state of affairs coems into existence. Why cannot some of the states of affairs be conditional? After all, presumably God is responsible for the laws of nature, and the laws of nature imply lots of conditional states of affairs.
Philosophical objection: Subjunctively conditional states of affairs are never basic. There must always be something behind them, a law of nature, an essence, etc.
Response: Note that a Molinist cannot make this objection. Note, also, that we can imagine that God creates further realities to ground the conditionals in the master plan, if that is needed. Thus God can ensure there is a law of nature that says that when Moses raises his staff and stretches his arm over the waters, the waters part. If one thinks that laws of nature cannot make reference to particular individuals, then all that God has to do is ensure there is a highly specific law of nature that makes no reference to particular individuals but is very specific: "If an individual, x, named 'Moshe' is raising his staff and stretching his arm over a sea, with a bunch of people in chariots pursuing x and x's compatriots, and if x is exactly 5.44747318998182828885912 feet tall, then the waters part". If laws of nature are grounded in essences, God can ensure the existence of combinations of essences that ensure the right conditionals hold.
Theological objection 1: It seems that God becomes incarnate in response to human sin. Sure, we can posit a conditional like: "If humans sin, God will become incarnate". But such a conditional does not save one from the idea that God's action of becoming incarnate is a response to a human action. It was essential to the example of Moses and the parting of the sea that the consequent of the conditional not be a divine action. But where the consequent is a divine action, the account is of no help.
Respone 1: It is not necessary to hold that the incarnation is a response to human sin. It may be that God would have become incarnate no matter what we did, redeeming us if we sinned and sharing in our joy if we did not. It is necessary to hold that the crucifixion is a response to human sin. But Christ's earthly actions such as not abandoning his disciples and running away come from his human will, Christ having both a human and a divine will.
Response 2: One can say that God willed the conditional: "Were humans to sin, there would come into existence a human being who is the same person as the second person of the Trinity." I don't know how satisfactory this is as a solution, and it can't be grounded in anything like laws of nature, but maybe something can be made of it.
Theological objection 2: The notion of God responding to us is essential to our having a relationship with him, engaging him in conversation. A conversation requires a back-and-forth. In an intervention with an alcoholic, the interveners have scripted contingency plans for whatever they can think of the alcoholic saying. Such an intervention--for all its value--is not a conversation. A conversation requires making decisions in response to the other's actions. Quite likely, we would make a different conditional decision ahead of time than the unconditional decision we would make at the given time.
Response: Well, here I am not very happy with my response, and I think this is a fairly serious problem for the model. But maybe we just have to say that a relationship with an omniscient and omnipotent being will have to be different from other human relationships.
As I say more than once, I am not endorsing the model. But it's an interesting mdoel.
Alex,
I don't think that A's responding to B requires B to be a partial cause of A's actions.
I follow Stump on this. She claims that responsiveness is a counterfactual beast. Something like the following: (A responds to B) iff (if B hadn't ___ed, then A wouldn't have ____ed). I know that's not enough, since there are counterexamples (I wouldn't be typing this if my great-great-grandfather hadn't had sex, but I'm not typing this in response to his sex [eww]). But, something in the ballpark is right, I think. Perhaps we'd have to include a clause like this: "A acts because A thinks B ____ed."
On this model of responsiveness, responding doesn't require temporal priority (or any temporal relations at all), which is a plus for Divine Atemporalists.
Also, it doesn't require that B exist while A exists. Consider this example (which I steal from Stump): suppose I want to start a foundation called the St. Thomas Aquinas foundation, and I want to do so because I value his contributions to the Church. I'm responding to his saintly life and philosophy, but he isn't a cause (partial or otherwise) for my naming the foundation after him. Or, to make the example more vivid, suppose I'm naming an organization after a dog of my great-great grandfather's that saved his life, but died before I was born. That dog doesn't now exist to be a cause of my action, but I'm still responsive to the dog and its actions of saving my great-great grandfather's life.
So, I'd deny 3 and keep 1 and 2. I think in doing that, one gets around the problem.
That dog doesn't now exist to be a cause of my action, but I'm still responsive to the dog and its actions of saving my great-great grandfather's life.
Tim, the fact that the dog is not alive now does not entail that he did not cause your action. Suppose I poison Smith and die before the toxin takes effect. Nonetheless, though dead before any harm occurs, I did cause the harm. Similarly, the dog's actions might cause you to do all sorts of things, though you learn about those actions after the dog is dead.
My intuition is to say the same about the Aquinas example. To the question "could you explain to me why you created this foundation" the likely answer is "because of the life and work of STA". That is at least part of the causal explanation of all this effort on your part. Wouldn't you say?
Mike,
I like your poison example and do think that you are right that I can set in motion a causal chain which has an end effect after I die.
I disagree that anything STA did or the dog did sets in motion a causal chain which leads to my so naming the foundation.
I'd say STA's work is part of the explanation for why I name the foundation so, but not that it is part of causal explanation for why I name the foundation so.
I think that A can respond to B's xing without B's xing having any causal influence on A. If that's at least possible, then Alex's 3 is false.
I disagree that anything STA did or the dog did sets in motion a causal chain which leads to my so naming the foundation. . . I'd say STA's work is part of the explanation for why I name the foundation so, but not that it is part of causal explanation for why I name the foundation so.
That's fine, but (if you will) do more than disagree. Tell me how this could be? Assume determimism first. Some event takes place in the past--the dog does something wonderful. You read about the event. You then name your organization Fido Inc. Remove Fido's actions from the historical facts, and your naming event never takes place. How does that not count as causing your action? Of course I agree, we could name other causes as well, but this is true in virtually every case in which some event is called "the" cause.
But now assume indeterminism. In that case we get the same series of events, but the result--your naming the organization Fido Inc.--is probabilistically caused by the Fido's actions.
I can't see how we avoid either a probabilistic or deterministic cause in these cases. But, you tell me.
Mike,
I thought I did more than just disagree in my first comment on this post. I gave what I thought was a decent first shot at defining responsiveness which didn't require causal influence, only counterfactual dependence. If it is possible to have counterfactual dependence without causal influence, then I think we have good reason to believe that there are cases where A responds to B's xing, and B's xing doesn't have a causal influence on A's responding. [I'm not sure you'll grant that there is counterfactual dependence without causal influence, though, since you say: "Remove Fido's actions from the historical facts, and your naming event never takes place. How does that not count as causing your action?"]
I tried to give one or two examples where I think there is no causal influence even if there is responsiveness. I think that my appreciation for Fido's actions are part of a causal chain that leads to me naming the foundation Fido Inc. But, those actions themselves, even if they help explain why I name the foundation as I do, don't cause me (partially or otherwise) to name it such.
Here's another example of responsiveness without causation. Some folks think that abstract objects can't stand in causal relations. However, it makes sense to say that one can respond to an abstract object's having a property. One can respond with tears to the plights of creatures of fiction, or respond with wonder at the properties of mathematical objects. So, A can respond to something without that thing having a causal influence on A.
However, I think we may have different notions of what it is to be a cause or a partial cause. As a matter of fact, I'm sure we do, since my notion is ill-formed and incomplete. But, as far as I can see, I don't need a notion of causation to say what I've been saying. My point has been that 3 assumes a notion of responsiveness that we aren't forced to affirm. We can understand responsiveness as a counterfactual dependence relation. If so, I don't see why we must affirm 3. But, perhaps someone can show why 3 follows if we understand responsivenss as a counterfactual dependence relation.
The case of abstract objects is interesting. I don't think that you're responding to the mathematical objects etc. themselves, since you'd have the same responses if fictionalism about such objects turned out to be true.
I didn't mean to sound critical, I actually can't see how the counterfactual dependence you described is not an instance of causal dependence, too. I agree that there are other counterfactual dependencies which are not causal. For instance, if x were not a polygon, x would not be a triangle. But x's being a polygon does not cause x to be a triangle. So I don't see all counterfactual dependencies as causal dependencies, though I suspect the converse is true.
I don't think that you're responding to the mathematical objects etc. themselves, since you'd have the same responses if fictionalism about such objects turned out to be true.
Mike, if you really have "responses"--just not responses to mathematical objects-- in a case of fictionalism, then isn't that a really good example of responsiveness without causal influence? I mean, if fictionalism is true, there aren't really mathematical objects. If there aren't really mathematic objects, then they can't causally influence me. So, either way, whether they are abstract or they don't exist, we can have responsiveness without causal influence.
I don't know what to say about the last point you make. You don't see how it could be that my naming the foundation Fido Inc. can counterfactually but not causally depend on Fido's actions. It seems to me that it does, and that I'm still responding to Fido's actions. But, again, this must be your having a more thought out understanding of what it is to be a partial cause than I do. Maybe if I knew more about what it is to be a partial cause I'd see why it can't be that I respond to Fido's actions by naming the foundation after him without being caused to do so by either him or his actions.
Tim, you write,
You don't see how it could be that my naming the foundation Fido Inc. can counterfactually but not causally depend on Fido's actions. It seems to me that it does, and that I'm still responding to Fido's actions.
Let me try to make this clearer to myself. Here's one idea. Maybe you mean that given that Fido did exist and performed certain actions, you still might not have named the foundation after him? You might not have responded as you did? But in that case you're saying that your action does not really depend counterfactually on Fido's actions. You are saying, in effect, that you did respond to Fido's actions in a certain way, but that you might easily have failed to do so. I can see how that would be a response that was not caused. You can get this by denying centering for worlds. And that's one way to clearly deny the causal connection between Fido's actions and your actions.
I should have noted that you can have everything that Stump wants, too. So you have Stump's counterfactual true in (S),
S. Had Fido not done X, then you would not have name the institution after him.
~X N-> ~I
And (C) true as well,
C. Were Fido to do X, you might not have named the institution after him.
X M-> ~I
With (S) and (C) you seem to have an uncaused response to the dog's Xing.
Mike,
I appreciate your help thinking through this. Thanks for doing some of my thinking for me :)
I do want (S) and (C).
But, I think I want centering too. Could you remind me how centering is defined? I thought it meant that, for any world W, no world is closer to W than W (or, maybe stronger: For any world W, W is the closest world to W, and, for any other world V, V is not as close to W as W).
I don't see how (S) and (C) together entail the denial of centering. Could you show why it does?
Also, you write: Maybe you mean that given that Fido did exist and performed certain actions, you still might not have named the foundation after him? You might not have responded as you did? But in that case you're saying that your action does not really depend counterfactually on Fido's actions.
What I meant by the counterfactual dependence of my naming the foundation after Fido is just that in any world where Fido doesn't do as he did, I wouldn't name the foundation after him. I didn't mean the stronger that, in addition to the previous counterfactual, in any world where Fido does his actions I name the institution such. I was using the term "counterfactual dependence" as shorthand for referring to the counterfactual responsiveness analysis I gave in my first comment. Perhaps I should have been clearer about that.
Thanks,
Tim
It seems pretty clear to me that the foundation case involves causal explanation. There is a chain of causal influences from the person for whom the foundation was named to the present. Without it, the idea of naming the foundation for that person would not exist. Moreover, this chain of causal influences is, plainly, explanatory.
Now, maybe you don't want to say this is causation, but only "causal connection" or something like that. But causal connection seems no less problematic vis-a-vis aseity.
Without a causal connection, the naming of the foundation for a famous person would be deeply mysterious.
I don't know what to make of the case of abstracta. I prefer to say that they are causally efficacious. Another thing that could be said is that the abstracta themselves structure our minds, in a way constitutive of, but not causal of, thought.
Constitutive relations are a nice case where one may have counterfactual connections without causal ones. But constitutive relations will not help with the case at hand.
Alex,
Do you agree with Mike that given (S) and (C) one can have an uncaused response?
I agree that between me and Tommy A, there is a chain of causal influences that is pertinent to my naming the institution. He has to do things, I have to hear about those things, and so on. But, I disagree that those things are a cause, partial or full, of my response. The exact same causal chain could be there and I not name the foundation.
Now, there's a disanalogy with God here, since God doesn't learn about things through the same causal chain story. So, even if, in the end, you think on this characterization of responsiveness any created being must have a causal chain linking him and the thing he responds to in order to avoid a deeply mysterious state of affairs, you needn't think that the same must be true of God. The way created things learn of stuff is categorically different than the way God does. So, that reason for a causal chain being necessary for the knowledge required to respond doesn't apply to God.
Hey Guys,
(1) God not be in any way caused to act by anything outside of himself.
I just don't the plausibility in (1) at all. A Christian shouldn't believe it. I mean, when Mary tickled Jesus's toes, did she not cause him to act, to wriggle and giggle?
Tim,
You're right that centering (actually strong centering) entails that for every w, no world as close to w as w. You have to give up centering to get (C) true. Recall that the antecedent in (C) is true (Fido did those wonderful things). So in order to get (C), you need another world (one as close to ours as ours is to itself) in which Fido does those things and you do not name an institution after him. That would make (C) true and a might-counterfactual. Otherwise (C) must be replaced with
X N-> I, and you have causal problems again.
Mike,
Sorry it has taken so long for me to get back to you on this.
You say: in order to get (C), you need another world (one as close to ours as ours is to itself) in which Fido does those things and you do not name an institution after him. That would make (C) true and a might-counterfactual.
This may just be a misunderstanding on my part, but I think that (C) doesn't need a world as close to ours as ours is to itself for (C) to be true. For a might conditional to be true,I just need a world where Fido does the wonderful things and I don't name the institution after him. As I see it, there could be closer worlds (including the actual world) where I DO name the foundation after him. But, so long as there is a world where he does what he does and I don't name it after him, then I have what I need for a might conditional (though, of course, not for a would conditional).
Concerning centering, I think that while there are no worlds closer to W than W, depending on which counterfactual we are analyzing, there may be worlds tied for closeness to W with W. So, for instance, if we're considering a counterfactual claim about two worlds, W and V, and W and V are duplicate worlds (or have identical temporal parts, or however your metaphysic has you say it) through t2, and we are considering a counterfactual about some time before t2 (t1), V can be tied for closeness to W. So, if centering doesn't allow V to be just as close to W under this closeness relation, then I guess I'm fine with rejecting centering.
Christian,
A Christan may thing that Mary caused Jesus to wiggle and giggle qua his human nature. But, qua his divine nature, nothing can have causal influence over him.
Tim,
You won't be surprised that I don't think that makes sense.
If the sheriff is also the deputy, then if I shot the man qua sheriff I also shoot the man qua deputy.
Though, the 'qua' strategy, if you were right, would save the inconsistent triad above easily. Moreover, something like it is required to respond to my objection.
Christian:
How the qua move works depends on the predicate. If Jones is both a taxi driver and a high school teacher, then one can fire Jones qua taxi driver without firing him qua high school teacher.
Alex,
I think that's just confusion though. Jones is a taxi driver and Jones is a teacher. No qua is needed. If Jones gets fired by his Principal, then Jones gets fired. Again, no qua is needed. Of course, Jones isn't merely fired, he is "fired from a certain job". We would say Jones lost is his teaching position but not his driving position. Again, no qua is needed. So we can say all of this without a qua, we can capture all of the relevant truths, and without distorting predication.
Moreover, the qua strategy doesn't seem to help with my example. If I shoot the man qua deputy, or if I shoot the man qua sheriff, then I cause the man qua "whatever" to have the property of being shot. I see no reason to treat tickling toes differently from shooting heads.
Finally, supposing all of that is just wrong, and I don't think it is, it doesn't matter. If a man is fired qua taxi driver, then that man qua teacher is fired qua taxi driver. That is still an effect on the man qua teacher. Similarly, if Mary tickles God qua human nature, then she effects God qua divine nature by causing it to have the property of being effected qua human nature. That is, God qua divine has the property of being effected by tickling qua human.
Christian:
Yes, there are difficulties here. One needs something like the notion of a property intrinsic to x qua F, and then one says God cannot be affected in respect to any property intrinsic to him qua God. How to define this, I don't know: I haven't worked on the metaphysics of the Incarnation.
Alex,
What about situations where the antecedent of your counterfactual conditional is something a creature (or creatures) do, and the consequent is something God does. Here are two Thomistic examples, but I'm sure that many who aren't Thomists would affirm them as well
1. The fact that: were Adam and Eve to copulate and the sperm and egg unite and everything else requisite for an embryo to form, then God would specially create a human soul.
2. The fact that: were Bob to stop putting impediments in the way of God's giving him sanctifying grace, then God would give him sanctifying grace.
Suppose Adam and Eve copulate. God, having strongly actualized 1, specially creates a soul for the offspring of Adam and Eve.
Is this a case of God being causally acted on--even partially--by his creatures?
If so, how, on the view we're discussing here, could one give a counterfactual concerning the special creation of a soul without violating aseity (aseity as it is understood here)? More generally, how could one give a counterfactual conditional where the antecedent is something that creatures do and the consequent is something that God does?
If this isn't a case where God is causally influenced by creatures, couldn't we say that God responds to Adam and Eve's action (or the the formation of the embryo, or what have you) when specially creating the soul? But then we have a scenario where "if creatures were to A, God would B" is true, creatures A, God B's, and the creatures' Aing doesn't cause God to B. That's responsiveness without causal influence.