Simple Foreknowledge vs. Open Theism on Providence, Round 1

But Zeus and Jove are co-refering terms, so they must know all and only the same things. :P

Clarificatory question: are you thinking of the simple foreknowledge theorist as considering God in time or timeless?

Trent,

I was tended to think of simple foreknowledge as involving God being in time--this was just primarily to have only 1 significant difference between Jove and Zeus. But I don't think that anything hangs on this assumption, as I think that everything I said about Zeus could be stated in a way that is consistent with eternalism--it would just lead to awkward locations (e.g., "it is true at t1 that Zeus timelessly knows that X does in fact occur at t2"). And while that wouldn't technically be foreknowledge, I don't think that God's relationship to time is the key issue here.

Dear Kevin,

This issue is a bit foreign to me, so sorry. As I understand things you aim to contrast two views of foreknowledge and then want to contrast the entailments of these views.

I'm not sure I understand the views. Open Theism seems to be a view where God, at some time t, does not know what will happen at t. This is to say that if there is an event at t' after t, then it is false that at t, God knows that event will happen at t'. One motivation for this view seems to be that there are not facts, at t, about what will happen after t. If it is a fact, then the probability, at t that this event will occurs at t' is 1, but indetermism is true and that entails that the probability cannot be 1.

Anyway, do I have Open Theism correct? I understand Simple Foreknowledge to deny the view about indeterminism and to accept that there are fact, now, about what will happen at a later time, and these facts now have a probability of 1. Is that Simple Foreknowledge?

Christian,

There are a number of different permutations of Open Theism, but the central idea is that God does not have exhaustive foreknowledge of undetermined future events (either because there are no truths about the future in this way, or if they are they are in principle unkowable). God can have exhaustive foreknowledge of what will happen if it is already determined to happen, and can have knowledge of what is likely to happen in the future. But He can't know whether or not E will happen in the future if E is indeterminate.

The proponent of simple foreknowledge thinks that God can, in fact, have exhaustive foreknowledge of the future in its entirety, including undetermined events and (if there are any) free actions.

Hi Kevin,

Interesting post.

As far as I know, those who advocate the simple foreknowledge view think that God has *exhaustive* foreknowledge of the future. Maybe I’m wrong about this though; David Hunt is the only person I’ve read on the view. But you do seem to indicate that this is what the simple foreknowledge position amounts to in your response to Christian; “The proponent of simple foreknowledge thinks that God can, in fact, have exhaustive foreknowledge of the future in its entirety…�

However, you say, “At t1, if Zeus extends his foreknowledge to include the fact that X does in fact occur at t2, then He can know that the probability of Y occurring at t3 is .97.�

I’m a bit confused, because it seems like Zeus doesn’t have exhaustive foreknowledge in your example. The cases you describe rely on Zeus’ being able to “extend� his foreknowledge to cover certain future contingents, but not others. That is, Zeus extends his foreknowledge to include X’s happening at T2, but *not* Y’s happening at T3. Zeus leaves Y’s happening at T3 epistemically open, and assigns it only a probability of happening, and thus plans accordingly (in better way than Jove).

But if Zeus has *exhaustive* foreknowledge, then this foreknowledge includes whether or not, as a matter of fact, Y does or does not occur at T3. Either he foreknows that it will take place or he foreknows that it will not take place, but in neither case can he do anything to change this fact. In other words, he doesn’t know it as a probability (97% likely) – he knows it simply as a matter of fact: Y will occur at T3 (or won’t). And so on for all of the future.

Anyhow, perhaps exhaustive foreknowledge isn’t essential to the simple foreknowledge view. But if we are thinking in terms of exhaustive foreknowledge, then as far as I can see, this eliminates the crucial component of your examples, namely, Zeus’ extending his foreknowledge just so far (to include X’s occurrence at T2) and no farther (leaving out Y’s occurrence at T3).

(Perhaps incidentally, this sounds like the voluntary nescience view that Alan described in his Four Versions of Open Theism post; presumably, Zeus *could* extend his foreknowledge to cover all future events but doesn’t. Thus, I wonder if what you’ve described here isn’t more properly called a version of open theism. The future is, after all, epistemically open to Zeus: he doesn’t know that Y will or will not occur at T3.)

Patrick,

You're supposed to be able to read my mind!

Sorry about being less than clear. Zeus, in the example above, is able to have exhaustive foreknowledge, as I indicated in my response to Christian. But if Zeus already knows that X will happen, then it is true that X will happen, and thus Zeus can't (won't?) be able to prevent it. You're correct here--and I don't see anyway to get around this.

So, yes, I was thinking along the same lines as Hunt, and that while Zeus canexhaustively foreknow the future, he brackets part of what he could know in order to exercise providence.

But I don't think that such a view is best described as a version of Open Theism, but maybe this will just come to semantics and how we define Open Theism. This view certainly is different than Open Theism as open portrayed, such that God can't have exhaustive foreknowledge of future indeterminate events and actions.

Kevin, you want a case in which if one knows what will happen rather than what will likely happen, then that individual will have some ability we take to be important. In some sense, probabilistic knowledge will not help as much, with respect to trying to achieve some outcome, as determinate knowledge.

But what I don't understand is that "if" this knowledge is in principle impossible because, as you say, "either because there are no truths about the future in this way, or if they are they are in principle unknowable," then how exactly can you describe a possible case?

I'm sorry, I'm a little confused.

Christian,

Like Patrick, you're supposed to be able to read my mind as well!

Here's a bit more of the dialectic between proponents of simple foreknowledge (SF) and Open Theism (OT). Proponents of OT say that God can't, in principle, have exhaustive foreknowledge of the future (namely He can't know indeterminate future events) because(i) there are no truths about the future to be known, or (ii) if there were, they are unknowable even for an omniscient God. Proponents of SF, obviously, reject (i) and (ii).

But proponents of OT often also make a further claim: (iii) even if God could have foreknowledge of future indeterminate events, such knowledge would be providentially useless (for the reasons laid out above). In addition to proponents of OT, a number of Molinists (Flint, etc...) also agree with (iii) and thus think that middle knowledge is needed to get one a strong view of providence. In this posting, I'm granting OT and SF their own views about (i) and (ii), and am just trying to address (iii).

I see. I'm understanding the debate better now. I'm still having a hard time making sense of the following two claim though:

1. "Proponents of OT say that God can't, in principle, have exhaustive foreknowledge of the future."

2. "Proponents of OT often also make a further claim: (iii) even if God could have foreknowledge of future indeterminate events, such knowledge would be providentially useless."

(2) is a claim about what God could do, per impossible. According to OT, there are no possible worlds in which God has foreknowledge of indeterminate future events. It is really hard for me see how the cases you describe are consistent with this fact. Let me explain.

A premise in Case 1: At t1, Jove could only know that there is an overall probability of .497 that Y will occur at t3.

The idea is that , if OT is true, then God must multiply objective probabilities to figure out what the probability is that Y will occur at t3.

Another premise in Case 1: At t1, if Zeus extends his foreknowledge to include the fact that X does in fact occur at t2, then He can know that the probability of Y occurring at t3 is .97.

It seems to me, and this is where I get confused, the idea is that, if SF is true, then God need not multiply objective probabilities to figure out what the probability is that Y will occur at t3. God can instead "extend his foreknowledge" to include the claim that "X does occur at t2" and then just infer that it's .97 likely that Y will occur at t3.

But, a premise in this argument is that "At t1, there is only a .17 probability that X will occur at t2." This means that we are evaluating the argument under the assumption that indeterminism is true, we are considering the premises as true in some world in which the probabilities are as described. Then we are considering two beings, in this same world, and what they could know.

First, God could know, if he extended is foreknowledge, that Y will will occur at t3, just like he knew that "X does occur at t2" by extending his foreknowledge.

Second, more importantly, this world is "not" one in which 1 above is true. God can have foreknowledge of the future because, by hypothesis, Zeus does. If this world is possible, then (OT) is false. If that is right, then (OT) will not accept that you have provided a counterexample to (OT).

In a different way, in order to evaluate the argument we need to consider its premises under two hypotheses, that (OT) is true, and that (SF) is true, but then we are not considering a possible world and Case 1 cannot function as a counterexample to (iii).

Again, I don't know the literature on this stuff, but (iii) strikes me as a very odd claim for one to make, not because it is false, but because it isn't clear how it can possibly be evaluated.

Kevin,

I think case-2 might smuggle in middleknowledge. Zeus knows that

(1)If reason Y occurs to Black at t2, then he will share argument X with Smith at lunch.

But this already presumes Smith is going to be at lunch which in turn assumes that God has already prevented the car from starting. Presumably, Black will not share the argument with Smith at lunch if Smith is not at lunch even if reason Y does occur to him at t2. In order to have the requisite knowledge to violate POLICY, Zeus would have to know that the counterfactual:

(2) If reason Y occurs to Black at t2 and if Smith is at lunch, then Black will share argument X with Smith.

Then and only then, it seems to me, would he have the sort of knowledge needed to violate POLICY. But by hypothesis he doesn't have this knowledge.

However, what do you think about this case (I think Dave Hunt shared this with me in conversation): God wants to determine if he is going to create hell. He will only do so if there will be people (non-human and/or human) who deserve to go to such a place. Via his foreknowledge he sees that there will be people (if only Satan and his minions) who deserve this and thus he creates hell.

Justin,

I don't think the case smuggles in middle knowledge, but you do point out a legitimate problem with how I described case 2. I said above: "The divine being also knows, by knowing Black's moral character, that he will share argument X with Smith today at lunch at t2 only if reason Y occurs to him at t2." And I stipulated that both Jove and Zeus know this. But it should have read: "The divine being also knows, by knowing Black's moral character, that he will share argument X with Smith today at lunch at t2 only if reason Y occurs to him at t2 and provided Smith is there." So your (2) above is right.

But I don't think that this requires God to have middle knowledge, despite looking like it might. For one, keep in mind that the divine being (whether we're talking about Jove or Zeus) doesn't know the truth of (2) pre-volitionally. As described, he would know this only at t1, which is after he already done a lot of things (including willing to create Smith and Black). But I think that even the Open Theist would grant that sometimes God could know what an agent would freely do in a particular situation in virtue of knowing that agent's present moral character. For example, I think that even the Open Theist would allow God to know (modifying one of Eleonore Stump's examples) that if I was offered a nickle to chop up my daughter into little pieces 5 mintues from now, I would freely reject. God needn't know this via middle knowledge. Rather, he knows that this is what I would do by knowing the contents of my moral character right now. So I would think that the Open Theist would grant that in some cases, God could have the kind of knowledge described by your (2). Any Open Theists out there listening want to confirm whether or not this is true?

If the account of how simple foreknowledge could be providentially useful at all works, then it might also help in the hell case that you suggest. But what I'm trying to develop here is the mechanism by which such knowledge could be providentially helpful. Do you think that your example shows this more clearly than either of the two I've suggested?

I'm finding SF problematic. If God really knows all future events, then he knows them from creation on (at least). He doesn't come up with them on the fly, as history unfolds. And I think he would therefore have no use for probabilistic knowledge, except as a kind of intellectual curiosity.

Also, he would know his own future decisions (assuming he makes those in time). If he knows before t that Y will occur at t because he makes it occur then, is that pre-volitional knowledge or not? I don't know.

Likewise, it can't be that God knows "X will happen unless I do something" and then makes decisions about whether to do something. That would be middle knowledge. He has to know that he will do something and X won't happen, or know that he won't do anything and X will happen. But I'm not sure that qualifies as providential control.

Kevin,

With respect to the hell case, if all we are trying to do is provide (at least) one case in which simple foreknowledge allows God some providential advantage over and above what he might have if he only had probabilistic knowledge of the future, then, yeah, I think the hell case works. But whether it is more effective than your examples, I'm not sure. I'm tempted to think it works better than case-1 insofar as the really interesting cases involve human agency. However, I suspect your second example is more dialectically advantageous insofar as it is more "realistic". Perhaps the more examples the merrier for the propoent of SF.

Kevin, you say above,

But I think that even the Open Theist would grant that sometimes God could know what an agent would freely do in a particular situation in virtue of knowing that agent's present moral character. For example, I think that even the Open Theist would allow God to know (modifying one of Eleonore Stump's examples) that if I was offered a nickle to chop up my daughter into little pieces 5 mintues from now, I would freely reject. God needn't know this via middle knowledge. Rather, he knows that this is what I would do by knowing the contents of my moral character right now.

"OT's hold, I would have thought, that only that which is causally or logically necessitated by the past or present is capable of being known"

That's an unusual condition on knowledge. Why is it that, when it comes to a proposition p concerning the future, God (or anyone else for that matter) knows p only if p is causally (or more strongly) necessitated? Why not the well-received condition on knowledge that God knows p only if p is true?
God has strong epistemic credentials, for sure, and it is true that proposition p might have been false. What then? What about the world in which p is false? We say here about that world that had p been false (had that world been actual) it would *have to have been* the case that God knew ~p instead. I can't think of a more proper set of conditions for a backtracking conditional. It is precisely what happens in Newcomb Problems with 100% accurate predictors; and God is (if anyone is) a 100% accurate predictor.

The question is a bit more specific, namely, how is a God who possesses simple foreknowledge able to make decisions based on simply-foreknown events. The problem is that foreknown events and choices are by definition already the result of whatever was done to influence them. Basinger (The Case for Freewill Theism: A Philosophical Assessment) remarks, “Since there can never be a time when a God who possesses complete SFK [simple foreknowledge] does not know all that will occur, and since foreknowledge can be utilized in a providentially beneficial manner only if there is a time at which what is foreknown can influence a divine decision that is itself not also already foreknown, there can exist no conceivable context in which SFK would enable God to make providentially beneficial decisions that he would not be able to make without this knowledge.�

If God simply foreknows all of history in the sense argued, then he knows his own as well as our choices, and he knows them in one single and (often argued) timeless sweep. He doesn’t acquire his knowledge in stages or bites. In this case, simple foreknowledge provides God no providential advantage in governing the universe for the simple reason that God cannot intervene into the story line of history on the basis of such knowledge. What is foreknown, according to this view, is what actually happens and by definition is already the result of whatever has or has not been done to influence it. Such knowledge can’t then also be the basis of such interaction for the simple reason that this knowledge comes to God too late (logically speaking) for God to use in influencing outcomes. Thus proponents of the open view claim there is nothing a God who possesses simple foreknowledge can do that a God who does not possess such foreknowledge cannot also do (relative to foreknowledge that is). It seems to me that whatever might be the reciprocal nature of the relationship between God’s actions and ours within the foreknowledge of God, the simple foreknowledge model offers us no help whatsoever in understanding it.

To dismiss this objection, one would need to show specifically how God could base a decision about event S on the knowledge that “S will occur.� But foreknown-S already IS the result of what was or was not done in an attempt to prevent or bring it about.

It’s not the case that open theists all deny that God knows the future ‘exhaustively’. The question is whether or not God’s knowledge of the future is exhaustively ‘definite’ or not. That is, whether God’s exhaustive knowledge of the future is limited to knowledge of what ‘will’ and ‘will not’ occur or whether it also includes knowledge of what ‘might and might not’ occur. So we’d agree that there are no ‘will-’ or ‘will not’-type truths that posit future contingent events. But we can agree that there are ‘truths’ describing such events, namely, ‘might and might not’ propositions. So open theists can argue that God’s knowledge is exhaustive in scope, infallible, immediate, and dynamic (not unchanging). We don’t have to deny exhaustive foreknowledge unless you assume “exhausitive knowledge� excludes knowledge of future contingents as contingent. But why believe that? God can know that “S might/might not occur at t.� That counts as foreknowledge in my view. We certainly base our choices and planning on knowledge of such contingencies don’t we?

Wanna say something about hell too, but I gotta run!

Tom

Sorry 'bout all the italics! I stink at formatting.

Tom

John,

I agree with you that OTs think that God can have foreknowledge about what will happen (as opposed to what might happen) only in those cases where some feature or set of features about the present is causally or logically sufficient for the future event in question. But I take it that sometimes a person's moral character is such that it provides the needed link in the conditional. This is what I was trying to get at with the example above. Given my present moral character, I think that either your or God could know that the following conditional is true: if I was offered a nickle to chop up my daughter into little pieces in 5 mintues, I would freely reject. Of course, if I had developed a worse moral character during the past 30 years, then this conditional might be false. And I might even be able to change my moral character so that in the future it would be false. But I'm pretty sure that it's true right now. Now, maybe you think that there is a reason why one's moral character can't even underwrite such claims. If that's the case, then I guess it's back to the drawing board.

Mike,

Spoken like a true Ockhamist! You ask: "Why is it that, when it comes to a proposition p concerning the future, God (or anyone else for that matter) knows p only if p is causally (or more strongly) necessitated? Why not the well-received condition on knowledge that God knows p only if p is true?" Most OTs think that there just aren't any true propositions about what will happen in the future for truly indeterminate events/actions. So they just don't think that this condition is satisfied in such cases. At least one OT thinks that there are such truths, but they are in principle unknowable, and thus God wouldn't be justified in believing them even if He did, so such beliefs wouldn't count for knowledge.

Tom,

Is there a reason that SFK must be understood as God everlastingly knowing exhaustively about the future all at once? In other words, is there some reason that the line of thought that I'm exploring here, and that David Hunt has argued for elsewhere, doesn't even deserve to be on the table? Or are you instead suggesting that this view isn't a view involving simple foreknowledge? Also, if you note above, I said that on OT God has probabilistic knowledge of the future--or "knowledge of what ‘might and might not’ occur" as you put it. Not only do I not deny that, but I build it in to the various cases discussed above. I'm not here arguing that there are 'will'-type truths about indeterminate events in the future. Rather, I'm trying to argue that if there were, then they would not be providentially useless (see my comments above dated May 19, 2006 09:48 AM). Unless one can say why the two cases I've sketched out above fail to give Zeus more providential control than Jone, or a principled reason why the type of knowledge Zeus has doesn't count as simple foreknowledge, it seems (to me anyway!) that I've shown how such knowledge could make a providential difference.

Kevin, thanks,

Is OT based on the assumption that there are no true (contingent) propositions concerning the future? (I take it there aren't any false ones either). I thought rather that this was a position they were forced to accept on the basis of some argument about the incompatibility of foreknowledge and freedom. But I could certainly have that wrong.

Hi Kevin,

I’ll take a stab at saying why your examples don’t show that SF helps God providentially govern the world.

Your examples presuppose that there are facts about what will be the case that allow Zeus to foreknow the future in part. That is, there is an exhaustive set of facts about the future, and Zeus has the power to extend his foreknowledge to cover some of these facts and not others. On the basis of knowing some of these facts, Zeus is in position to providentially act in order to bring about some desired result.

The problem is that the same set of facts that allows Zeus to know that (to use your first example) X will occur at T2 also contains a fact about whether or not Y will occur at T3. In other words, Zeus could, in principle, extend his foreknowledge to include whether or not, as a matter of fact, Y does or doesn’t occur at T3. Zeus just chooses not to in order to providentially act with respect to Y’s occurrence. But since we’ve granted that there are truths about the future, there are also truths about Y’s occurrence, and unless we are thinking that Zeus can somehow make it so that Y doesn’t occur at T3 when it’s always been true that Y *will* occur at T3, then Zeus can’t do anything with respect to the facts concerning Y’s occurrence at T3.

For your examples to work, it seems to me that we’d have to suppose that there is no fact of the matter concerning whether or not Y occurs at T3. But then, if there is no fact of the matter about this indeterminate event, why should there be one concerning X’s occurring at T2? In other words, your examples presuppose that there are facts concerning what will happen at all times up to T2 so that Zeus can know these facts. But your example also presupposes that there aren’t facts about what will occur at T3 since Zeus stands in position to providentially act with respect Y’s potential occurrence at T3. This seems odd.

So here is an example in which SF *would* give Zeus more providential control. There are facts concerning what will happen at times T1 – T20 *but not* facts concerning what will happen at times T21 and onward. Zeus extends his foreknowledge to cover the facts concerning what will happen at times T1 – T20, but since there are facts about what will happen at these times, Zeus does not stand in position to providentially act with respect to these events. However, there are no facts about what will happen at times T21 and onward, and so on the basis of knowing what will transpire at T1-T20, Zeus is in better position than Jove (who only has probabilistic knowledge of what will occur at T1-T20) to providentially bring about certain results at times T21 and onward. (Zeus wouldn’t be changing any facts about what occurs at T21 onward; there are only probabilities concerning what will occur at such times.)

This would be a quite strange universe, however. Why should there be facts about indeterminate events only up to a certain time, after which there simply are no facts? I’ve no idea, but as I see it, this is what would be needed for SF to give God more providential control than a God without SF.

Kevin, if it takes logical or causal necessitation to make knowledge of the future possible, I don't see how the OT will allow anything less in the case of a counterfactual of freedom. That's what worries me here. I would expect the OT to say here just what they say about the future. At best, God can only know what is likely to happen, and at best, the link between antecedent and consequent is probabilistic.

Jon –

This is admittedly just a suggestion, but at present I don’t see why it shouldn’t work: The OT can say that certain decisions we make are psychologically necessary. This is, I take it, what Kevin was getting at in his example: it would be psychologically impossible for Kevin to kill his daughter for a nickel. For the OT, it remains metaphysically (or logically) possible that Kevin do so, but given facts about Kevin’s psychology (or, as he put it, his moral character) it is not psychologically possible that Kevin do so. Nothing about what the OT must say about freedom precludes the OT from saying that some decisions we make could be psychologically necessary (supposing, of course, that some decisions we make can actually be so.)

An even more tenuous suggestion is that this psychological necessity could ultimately be analyzed in terms of causal necessity, e.g. perhaps certain brain states etc. that reflect my psychology causally necessitate that I do X in Y. Maybe I’m missing something, but this seems alright to me. If this could work, then it seems the OT can get around your worry.

I think the OT should say something like this, ultimately. Certain facts about my psychology make true certain counterfactuals about me, and on the basis of sufficient causal factors being in place for the antecedents of such counterfactuals obtaining, God can know in a non-probabilistic fashion what I’ll freely do in the future. For example, supposing that its psychologically necessary that I do X in Y, and God sees that there are sufficient causal factors for my being in X in 5 minutes, God knows that I’ll do X in 5 minutes. (Sufficient causal factors are in place for my doing X in 5 minutes.)

God would then be a sort of quasi-molinist God who can generate foreknowledge of what we’d freely do, not in virtue of middle knowledge, but in virtue of those counterfactuals that are made true because of what our psychology is like. So long as certain decisions we make really are psychologically necessary (and why shouldn’t some be?) the OT God can have foreknowledge in this way.

Patrick, here's a dilemma for your suggestion if you really want to be an OT. Either psyc. imp. is a kind of causal impossibility or not. If it is, then everything is fine, except that you can't be a libertarian anymore about the action in question. If it's not, then take any claim about the future concerning my actions that psychologically I have to perform (suppose there's a law of nature to undergird my continued existence until then). Now, why isn't mere psychological necessity enough to make that truth knowable for the open theist?

My point has been that parity of reasoning should apply both to the OT's attitude toward the knowability of the future and to the knowability of a conditional.

Jon –

To me, one horn of the dilemma you suggest is perfectly plausible: the action in question isn’t free in the libertarian sense. That is right. If it is psychologically impossible for Kevin to kill his daughter for a nickel, then if put in that situation, Kevin isn’t free to do otherwise than reject the nickel. (Again, it’s metaphysically possible that he does so, but psychologically impossible.)

But, crucially, and as Kevin has pointed out, Kevin has freely (in the lib. sense) made himself into the sort of person for whom it is psychologically impossible to take a nickel to kill his daughter. So long as this is true, Kevin is morally responsible for his decision to reject the nickel.

So I’m fine with saying that I’m not free in the libertarian sense with respect to at least some of my decisions: my psychology may causally necessitate that I do X in Y, so if I find myself in Y, then I’m not free to do otherwise than X. But I’m responsible for my psychology being that way and not some other way, so still responsible for doing X in Y even though I can't do otherwise.

Patrick,

Your last response is a great expression of restirctivism--van Inwagen would be so proud! I prefer to think in terms of direct/indirect freedom in addition to direct/indirect moral responsibility, and that in the character formation case I'm both indirectly responsible and indirectly free. But I don't think that anything of significance hangs on this distinction between the restrivist and myself on this point. It may just be terminological, and I've been told on more than 1 occasion that I should concede to the restrivist here.

I still need to respond to your previous comment. But I'm going to take my dog to the park and finish marking some exams, so will get to it later (maybe tomorrow). Thanks for the fun and productive exchange.

Patrick, you've lost sight of the conditional in question: it requires libertarian freedom in the consequent. You just took it away with what you said. So the counterfactual is false on your account.

I stepped away a day and fell behind! So let me quickly post this without having read the last few posts, then I'll play catch up when I can.

Kevin: I agree with you that OTs think that God can have foreknowledge about what will happen (as opposed to what might happen) only in those cases where some feature or set of features about the present is causally or logically sufficient for the future event in question.

Tom: Right. A presentist/open theist would argue this. The present grounds all truths, including truths about the future. Present realities (physical laws, dispositional realities, etc.) may be such they entail a certain coming-into-being, viz., they’re such that some event or state of affairs ‘will’ occur. Or they may be such that they rule out certain states of affairs (and so, ‘will not’). An open theist can argue that some present realities are such that certain future states are neither entailed (‘will’) nor ruled out (‘will not’). They ‘might/might not’ happen.

Kevin: But I take it that sometimes a person's moral character is such that it provides the needed link in the conditional. This is what I was trying to get at with the example above. Given my present moral character, I think that either your or God could know that the following conditional is true: if I was offered a nickle to chop up my daughter into little pieces in 5 mintues, I would freely reject. Of course, if I had developed a worse moral character during the past 30 years, then this conditional might be false.

Tom: I’d agree. In this case a person’s moral character just is that present reality that grounds truths about her future (whether she ‘will’ or ‘will not’ or ‘might/might not’ do something). Mother Theresa had a sufficiently determined character that rendered this or that conditional true or false. She freely chose her way over time into a predictable way of behaving. Eventually the moment came when “Mother Theresa might abuse small children� was false. Similarly, Hitler chose his way into an equally predictable way of behaving, only with regard to evil. So moral character counts as one of the present realities that may determine this or that about our future. But our characters may also be insufficiently developed so that a good deal about our futures is indeterminate (‘might/might not’).

Kevin: And I might even be able to change my moral character so that in the future it would be false.

Tom: So just include in the conditional the state of moral character that determines this or that about some future choice (or include some time limit as you do, as in “in five minutes� which would reflect your moral character at that time). So you’d have something like, “If Keven freely solidifies into morally corrupt ways of behaving, he ‘would’ abuse his daughter in situation Z.�

Kevin: But I'm pretty sure that it's true right now. Now, maybe you think that there is a reason why one's moral character can't even underwrite such claims. If that's the case, then I guess it's back to the drawing board.

Tom: I think moral character can be sufficiently solidified over time. I think Mother Theresa probably BECAME her choices so that eventually she loved ‘compatabilistically’ rather than ‘libertarianly’. This is fine with me so long as the compatibilism was libertarianly determined, if you get my meaning. As Augustine (?) said (or something close to it), “Sow a thought, reap a choice, sow a choice, reap a habit, sow a habit, reap a character, sow a character, reap a destiny.� Libertarian free will would then be finite. It gives way over time to compatibilism. But we freely determine which way we end up—compatibilistically loving or compatibilistically selfish and evil.

Mike, you may find something of interest for Ockhamists in this article: www.alanrhoda.net/ papers/opentheism.pdf. Some open theists are Ockhamists. Bill Hasker, I think, is. He’d either deny bivalence and argue that “Bill ‘will’ do S in Z in 10 years� is (where S in indeterminate) neither true nor false, or he’d agree that it’s either bivalent but logically impossible to know. I don’t find either or these the best avenue for defending an open theist semantics. With Alan (and other open theists), I’m a Piercean (www.alanrhoda.net/papers/In_Defense_of_Prior's_Peircean_Tense_Logic.pdf).

Kevin: Most OTs think that there just aren't any true propositions about what will happen in the future for truly indeterminate events/actions.

Tom: If you’re using ‘will’ in the Peircean sense, yeah. As OVers work it out, I think most are/will settle into the conviction that there are true propositions about future indeterminate events. It’s just that those truths are three (as opposed to the traditional two):

X will do Y in Z
X will not do Y in Z
X might/might not do Y in Z

Kevin: Is there a reason that SFK must be understood as God everlastingly knowing exhaustively about the future all at once?

Tom: Most advocates of simple-foreknowledge will argue that God’s being omniscient is the reason. God is at all times omniscient. And at all times there are truths about the future. Open theists are the only people around that I know of who are arguing that God knows some of that future as indeterminate, that is, knows what might/might not happen (without also knowing about the same events that they ‘will’ or ‘will not’ happen).

Kevin: Or are you instead suggesting that this view isn't a view involving simple foreknowledge?

Tom: Right. I’m not seeing simple-foreknowledge in your examples.

Kevin: Also, if you note above, I said that on OT God has probabilistic knowledge of the future--or "knowledge of what ‘might and might not’ occur" as you put it. Not only do I not deny that, but I build it in to the various cases discussed above.

Tom: It only makes sense to me to say God has knowledge that an event is more or less probable to occur if one is an open theist. This sort of knowledge is precisely what makes one an open theist. Okhamists approach ‘will’ and ‘will not’ with a different semantic and attempt to depict God as epistemically closed with regard to the future (that is, God knows the future exclusively in terms of what shall and shall not occur, as ‘definitely this’ and ‘definitely not that’) while at the same time suggesting that God knows the relative probabilities of indeterminate events. So God knows that 10 years from now I’ll definitely have a latte at 9:02 AM (and so on) and he knows this ‘will’ occur with a 75% probability. That’s all self-contradictory to me.

Kevin: I'm not here arguing that there are 'will'-type truths about indeterminate events in the future. Rather, I'm trying to argue that if there were, then they would not be providentially useless (see my comments above dated May 19, 2006 09:48 AM).

Tom: Then I don’t understand your examples, because I don’t see that what you’re suggesting IS a providential advantage. I’m looking for an example of God simply-foreknowing that “S ‘will’ occur� and then basing some decision about S on this knowledge of S. That would be a providential advantage. But you haven’t shown that so far as I can see.

Tom

Thanks Tom. Incidentally, which Tom?

Hi Jon –

Here is the conditional in question, right?

If Kevin were offered a nickel to kill his daughter, he would freely reject it.

I’m unclear on why the freedom in the consequent of this conditional has to be libertarian freedom. I’m not sure whether Kevin originally intended for it to be a counterfactual of lib. freedom, but as I see it, it isn’t one. It is a counterfactual of some other sort of freedom; I’ll go with a ‘counterfactual of source-incompatibilistic freedom’.

My claim isn’t that the OT God can know what we’ll freely do in the lib. sense, but that he can know what we’ll freely do in some other sense. This is because, as I see it, there are no true ‘would’ counterfactuals of libertarian freedom (and thus no way for God to know what we would lib. freely do). Here’s why.

If I would do X in Y, then if in Y, I will do X. If I will do X in Y, then if in Y it is causally necessary that I do X. (This is, of course, going to be the point of contention.) If it is causally necessary that I do X in Y, then I can’t do otherwise than X in Y. And if I can’t do otherwise than X in Y, then I don’t have libertarian freedom with respect to X in Y.

So if it is true that I would freely do X in Y, this can’t be a counterfactual of lib. freedom. It’d have to be a counterfactual of some other sort of freedom (I’ve suggested source incompatibilistic freedom, though I’m quite open to suggestions here!), namely, *whatever sort of freedom it is that I’ve got with respect to X when I’ve lib. freely made myself into someone for whom it is psychologically necessary to do X*.

On the basis of the truth of this sort of counterfactual of freedom, the OT God can know what we’ll freely do in the future – though not what we’ll freely do in the lib. sense.

Again, I’m not quite sure what to call this sort of freedom, but unless one thinks that libertarian freedom is the only freedom on offer, or that in such a situation one isn’t actually free with respect to X, then I think this can work.

One last thought: I chose source incompatibilistic freedom because the freedom in question is still incompatibilistic since it requires one to have lib. freely made oneself into a certain sort of person. And one’s own will is still the ultimate reason for why one is the sort of person for whom some decisions are psychologically necessary.

Hope this gets at your last post, though maybe I’m missing something.

Patrick, I think this is an objectionable way to talk, enough to make my philosophical skin crawl. My objection is to the idea that, for each theory of X, there is a kind or type of X, or sense of 'X' that we can employ in conversation. Down this path lies the practice of talking about coherentist justification, or foundationalist justification, or contextualist knowledge or invariantist knowledge, or Kantian rightness or utilitarian badness. Such language can often be reinterpreted in terms of intended reference to the theories in question, but in other cases, nonsense results. There is knowledge, there is justification, there is freedom, and there are various theories of each. When attempt are made to talk about types of each, when the reference is clearly to a particular theory of each, we should always insist that the type talk be replaced with accurate descriptions in terms of the theories in question. If you think source incompatibilism is true, that's fine. But there isn't a source incompatibilist sense of freedom or compatibilist kind of freedom or libertarian type of freedom. There's freedom and there are various theories.

Moreover, OT's agree with Molinists that the correct theory of freedom is libertarian. That's settled enough in the literature to be part of common philosophical parlance, and having a settled vocabulary here is necessary for productive conversations about the issues. There are perhaps interesting close cousins of these views that might endorse some other theory of freedom, and I expect that's the direction you'd prefer to pursue. But when Kevin suggests an OT endorsement of a counterfactual of freedom, any OT defense will have to assume the theory of freedom that is distinctive of the OT position. Otherwise it's like some contemporary theologians claiming to defend some aspect of traditional Christianity by reinterpreting the words so that, on that reinterpretation, the sentence expresses a true proposition (one, of course, which is contrary to some aspect of traditional Christianity). The word 'obfuscation' comes to mind...

Patrick says of the conditional

“If Kevin were offered a nickel to kill his daughter, he would freely reject it�

that he’s unclear as to why this has to be libertarian free will.

From this OVer’s perspective, if one understands the freedom in question to be libertarian, then there are no true CCFs and the proposition about Kevin would be false. There are no ‘will’ or ‘will not’ (from the Piercean perspective) type propositions that posit future contingents. To posit a conditional involving some indeterminate act, then (as I see it) the only way to express this propositionally is “If Kevin were offered a nickel to kill his daughter, he might and might not reject it.�

Patrick continues: “It is a counterfactual of some other sort of freedom; I’ll go with a ‘counterfactual of source-incompatibilistic freedom’.�

I see your point, though (even as an open theist) I’d just call it compatibilistic freedom (though libertarianly determined over time by the agent). Mother Theresa became (or rather chose to become) psychologically incapable of abusing children. Was she physically capable of committing such crimes? Sure. But she chose to become a slave (in her mind, emotions, judgments, etc.) to goodness. Evil increasingly became less and less likely for her. Eventually some evils became psychologically impossible for her.

Patrick: My claim isn’t that the OT God can know what we’ll freely do in the lib. sense, but that he can know what we’ll freely do in some other sense.

Tom: If I’m understanding you, then your thinking along the lines of sufficiently developed moral character, in which case open theists can agree. The future of some characters can be predictable in certain situations. My point is that those characters are not predictable until agents render them over time sufficiently predictable through the exercise of libertarian free will (at least in morally significant terms). So the only way to predict what an agent will do based on psychology/character development is based on that character at the time the prediction is made. But this would rule out divine pre-creational planning, for example, or even God’s knowing what we’d freely do during our lifetimes before we rendered ourselves sufficiently predictable. But this “non-libertarian sense� seems to me to be nothing but compatibilistic free will and not a third sort of freedom.

Patrick: This is because, as I see it, there are no true ‘would’ counterfactuals of libertarian freedom (and thus no way for God to know what we would lib. freely do).

Tom: I’m with you here. So IF there’s true ‘would’, it cannot posit a libertarian action. (Ah, I see you already say this a paragraph or two later.)

Patrick: So if it is true that I would freely do X in Y, this can’t be a counterfactual of lib. freedom. It’d have to be a counterfactual of some other sort of freedom (I’ve suggested source incompatibilistic freedom, though I’m quite open to suggestions here!), namely, *whatever sort of freedom it is that I’ve got with respect to X when I’ve lib. freely made myself into someone for whom it is psychologically necessary to do X*.

Tom: I’ve been trumpeting this for several years now.

Patrick: On the basis of the truth of this sort of counterfactual of freedom, the OT God can know what we’ll freely do in the future -though not what we’ll freely do in the lib. sense. Again, I’m not quite sure what to call this sort of freedom…

Tom: It’s compatibilistic freedom straight-up. It’s just that were most theistic compatibilists use compatibilism to help defend the idea of God’s determining our actions and our also being morally accountable for them, an open theist at this point would still argue that there’s no long-term foreknowledge available through the sort of *freely determined compatibilistic behavior your talking about. This is true because the agent herself is the one who shapes and determines the direction her compatibilism will take (for good or for evil), and that is libertarianly determined and so unpredictable, and b) the situations in which we’re predictable (like somebody’s offering Kevin a nickel to kill his daughter) are usually themselves indeterminate events and involve genuine libertarian choices (other than our own choices). So predictable charaters offer some providential advantage for a God who knows our characters perfectly. Any open theist concedes this. But they don’t offer long-range forecasting for individuals not yet existent or existent but not yet of sufficiently developed character.

Patrick: One last thought: I chose source incompatibilistic freedom because the freedom in question is still incompatibilistic since it requires one to have lib. freely made oneself into a certain sort of person. And one’s own will is still the ultimate reason for why one is the sort of person for whom some decisions are psychologically necessary.

Tom: Then it’s just compatibilistic freedom, not incompatibilistic freedom (or so it seems to me). Sure, it was libertarianly determined by the agent herself over time, but once determined, once settled into, the exercise of freedom is quite compatibilistic. If I follow Jon, I tend to agree. One is either libertarianly free or not. So I’m using “compatibilistic� freedom to refer to the non-libertarian exercise of the will. Libertarians have traditionally said no such exercise of the will may be considered ‘morally significant’ or accountable behavior. But I think such exercise of the will can be morally significant on one condition--namely, that the agent himself libertarianly chose to develop the sort of character that became fixed, or solidified, with respect to good or evil. At some point Mother Thersea was not libertarianly free to canabalize children. But we take this to me a morally admiralbe trait because SHE chose libertarianly to become that sort of person. Had some force OTHER than her own libertarian will determined her this or that way, then libertarians would say her development and character were not her own and therefore are not morally significant (accountable to her).

I’ll try to interact more specifically with Kevin’s examples regarding providential advantage in the future.

Tom (Belt)

I hope this brings me up to date on Kevin’s examples.

Kevin: It seems to me that in both of these cases, Zeus is better able to exercise providence because he has more knowledge in the very same situation than does Jove. If this is true in either case, than the antecedent to CONDITION is fulfilled and Sanders’ argument fails. (Now, in these cases Zeus might not have much more providential control than would Jove, but that’s not the issue here.)

Tom: Just to make sure I’m following you. You begin by saying Zeus is “better able to exercise providence� but qualify the claim with “Zeus might not have much more providential control than would Jove� but that this isn’t the issue. I thought how much providential control foreknowledge provided *was the issue. But maybe you just meant to say that Zeus would only have slightly more providential control. Not sure.

Kevin: Case 1 (involving indeterminate events): At t1, there is only a .17 probability that X will occur at t2. If X occurs at t2, then there is a .97 probability that Y will occur at some later time, t3. If X does not occur at t2, then the probability that Y will occur at t3 is .24. At t1, Jove could only know that there is an overall probability of .497 that Y will occur at t3. At t1, if Zeus extends his foreknowledge to include the fact that X does in fact occur at t2, then He can know that the probability of Y occurring at t3 is .97. Zeus is a better epistemic situation at t1 with regard to Y occurring at t3 than is Jove, and thus is better equipped to providentially control whether or not Y occurs.

Tom: So are you suggesting that God’s foreknowledge is volitional (or, that he’s not all times omniscient)? He can choose which future events to foreknow by “extending his foreknowledge� to include them? I’ll scan the posts again to catch up, but I’d just say that as I understand the simple-foreknowledge view, the only probabilities God knows are 1 (for events he knows ‘will� occur) and 0 (for logically possible events he knows ‘will not’ occur). So what you’re suggesting doesn’t look like the simple-foreknowledge view I’m familiar with. You want to say God knows X ‘will’ occur (which to my mind means ‘will with a probability of 1�) but that God also knows X is indeterminate. But to my mind indeterminate means “with a probability of greater than 0 and less than 1.� As I understand it, once you posit simple-foreknowledge, the only probabilities God attaches to possible events are 0 and 1. Once you introduce God knowing objective probabilities for future events, you start sounding like an open theist. But then you revert with “extending his foreknowledge to X� to God’s knowing only one probability for future events—1 (that which ‘will’ certainly occur).

I’m sure an Okhamist like Craig would work the semantics differently so that he could allow for objective possibilities AND have God knowing the entire future exclusively in terms of what shall and shall not occur. But still he recognizes there has to be some *wiggle room for God to make use of what he knows will happen in providentially advantageous ways, which is part of the reason why Craig is a Molinist. He realizes that with the standard simple-foreknowledge view (where there’s never a time at which God does not know ALL that will occur) there’s not the needed wiggle room to base decisions that are not themselves foreknown upon events which are foreknown in an attempt to influence the latter. So he posits middle knowledge. It looks to me like you’re just positing an open future of indeterminate events but then simply supposing that God can choose to foreknown them. But I really think the supposed providential advantage it gives is vacuous. If God simple foreknows all that will occur, there are no probabilities of greater than 0 and less than 1 that God also foreknows.

Let me ask this, Kevin. In Case 1 you mentioned no objective probability for X. So what is X’s probability of occurring? Is X’s occurring a historically contingent event? Is its probability of occurring 1 or less than 1? If less than 1 (viz., if it’s indeterminate), then what are the grounds for God’s knowledge that X ‘will’ certainly occur?

Kevin: Case 2 (involving free will): In virtue of knowing Smith's present moral character, the divine being knows that if Smith doesn’t consider argument X at t2, then Smith will remain an atheist. But by also fully knowing Smith's present moral character, the divine being also knows that if Smith were to consider argument X at t2, there is a .87 probability that he would convert to theism. The divine being also knows that the only human being that is capable of articulating argument X is Black, who normally has lunch with Smith everyday at the university. The divine being also knows, by knowing Black's moral character, that he will share argument X with Smith today at lunch at t2 only if reason Y occurs to him at t2. But the divine being knows that the probability that reason Y will occur to Smith at t2 is only .14. The divine being also knows that Black is planning on leaving the university early today to prepare for tonight’s television address. At t1 (where t1 is earlier than t2, which is in turn earlier than t3), Black gets in his car to drive home. The divine being could prevent Black from driving home, and thus ensure that he has lunch with Smith at t2, by making Black's car fail to start. But the divine being has a general policy, which he calls POLICY, of interfering with cars only if he can be more confident than not that it will result in the conversion of at least one person.

Tom: Sounds perfectly open theistic!

Kevin: But Zeus, if he has foreknowledge that despite the low probability of .14, knows that reason Y will occur to Smith at t2 and that Smith will therefore share argument X with Black, knows that there is a .87 probability that Black will convert to theism. On the basis of this knowledge, Zeus knows that the antecedent of POLICY is fulfilled. Zeus thus makes Black's car fail…

Tom: On the simple-foreknowledge view, there’s no time at which God foreknows that there’s an 87% probability Black will convert to theism but does not also foreknow that he’ll cause Black’s car to fail. That’s the problem.

You obviously see that when it comes to foreknowledge, providential advantage requires some temporal gaps in (what shall we call it?) the divine acquisition of truths about some future event. You need these gaps, this wiggle room, to allow God to do the “basing of decisions not foreknown� upon other events "already foreknown" which is at the heart of your examples regarding Zeus. But the “on the basis of this knowledge, Zeus [then] knows so-and-so and THUS [then] decides to make Black’s car fail� actually denies what simple-foreknowledge affirms, that there are such gaps in God’s knowledge of the future; that is, that there are ever times when God does not know all that shall occur in the future. On the simple-foreknowledge scheme, Zeus could not reason his way through to the decision you suppose. To make those kinds of providential moves, Zeus either (a) has to be ignorant of the outcome he desires to influence (Smith’s choice to convert or not at hearing Black’s argument) at the specific time he makes his choice to have Black’s car fail, or (b) has to make all such decisions based on infallibly knowing at all times the objective probabilities of the indeterminate events in question, probabilities that are grater than 0 and less than 1 (which is the Open View). Zeus simply doesn't have the sort of wiggle room you suppose IF he's omniscient and IF such omniscience entails he's always knowing all that shall occur in the future (i.e., if the simple-foreknowledge view is true).

Tom

In my comment about Craig it was of course "objective PROBABILITIES" (not "possibilities") that I meant.

Tom

One last Q for Kevin-

Forgetting for the moment just what the “simple-foreknowledge� view actually entails, let’s just consider a view, call it UF for ‘undivided foreknowledge’, and let’s suppose that this view makes the following claim:

“There is never a time when God does not know all that will occur in the future.�

Would you agree that foreknowledge of this sort offers God no basis upon which he is able to decide how to act providentially in the world (in influencing foreknown events, in bringing them to pass, in preventing them, etc.)?

Tom

I've been doing end of the term grading, but hopefully now can catch up on the discussion.

Patrick,

In your May 23, 2006 01:38 PM post, you suggest the following dilemma: either (i) Zeus knows all the facts about the future or (ii) Zeus knows knows the facts about the future up until a certain time. With regard to (ii), I agree that it would be a strange world. But if it is even possible that the world is like that, then we have a case where foreknowledge would be providentially useless, and the argument that I started this post with would fail. With regard to (i), I don't see why I couldn't appeal to the standard Ockhamist strategy with regard to God's knowledge about whether Y occurs at T3. Say that Zeus could know that, but simply chooses not to. Sure there is a truth about whether or not Y occurs, but the truth is grounded in what Zeus chose to do on the bases of the knowledge that he did have. So I don't think that my case does require, as you suggest, "that there aren’t facts about what will occur at T3 since Zeus stands in position to providentially act with respect Y’s potential occurrence at T3." Instead, the truth of what occurs at T3 just counterfactually depends on how Zeus providentially acts at T1 given his foreknowlege of what will transpire at T2.

Jon,

I agree with Patrick about the conditional in question. There are numerous incompatibilists who think that alternative possibilities aren't required in the case of determination by one's moral character. In these cases, one can still act freely so long as one freely set one's moral character (which will require alternative possibilities at some point in the character-forming-process). This is, I take it, one of the central lessons of Kane's use of the Luther example. Stump also seems to agree. I think that van Inwagen would disagree on the basis of his restrictivism, but that's simple because he defines free will as requiring the ability to do otherwise at that time. We could rework the whole problem in such a way that it's about actin in a way for which one could be morally responsible, and then I think he'd agree with the character cases. But I don't think we need to do that. But contrary to what you seem to think, and what Tom clearly thinks, the freely involved in such a conditional is NOT compatibilism--because it does not think that free will is compatible with causal determinsm as it is usually expressed in the free will literature. Free will might be, in certain cases, compatible with local determinism of a certain sort, but spelling out the "of a certain sort" is going to ensure that we're still talking of an incompatibilist view. And if libertarianism is understood, as it usually is, as the conjunction of incompatibilism and 'we are sometimes free', I don't see how this isn't a form of libertarianism.

Tom,

I'm going to respond to you still--but you've raised so much, it will take me a bit. Thanks for your patience.

Tom,

I don't have the time to respond to every issue you raise--good though the are--but here is a start. Thanks for pressing me on some of these issues.

Tom: our characters may also be insufficiently developed so that a good deal about our futures is indeterminate.

Kevin: Sure, but I just need a case where our moral character is set with regard to one particular aspect of it to get a case to work. And you endorse that this is possible. So I think we're in agreement here.

Tom: I'm not seeing simple-foreknowledge in your examples.

Kevin: Here's where the foreknowledge is. Zeus foreknows at T1 that an indeterminate event will happen at a future time T2. And since this knowledge isn't based on middle knowledge, it's simple foreknowledge. And such knowledge is providentially useful to Zeus because he is in a better position, at T1, to guide what happens at T3 than is Jove (who lacks foreknowledge). Because of his foreknowledge of what will (not a Percian 'might' but a real 'will') happen at T2, he is able to have a higher degree of providential control about what happens at T3 than is Jove.

Tom: Then it’s just compatibilistic freedom, not incompatibilistic freedom (or so it seems to me).

Kevin: I agree with Jon that there's just one sort of freedom, though the compatibilist and incompatibilist disagree with what is required for one to be able to have such freedom. But that aside, I disagree that this is compatibilism for the reasons indicated in my comments to Jon above. Source incompatibilism is clearly a form of incompatibilism, not compatibilism. And if libertarianism is simple the conjunction of incompatibilism and the existence of free will, then the view under discussion is still a kind of libertarianism. And I don't think that I'm unusual in my understanding it this way. Kane, Stump, McKenna, Fischer, Campbell, Clarke, Levy would all, I'm pretty sure, agree with me here. And while I'm not suggesting that agreement entails truth, I do think that this pedigree is quite substantial. I'd be happy to argue for this at greater length, but I don't think that the present post is the best place to do so.

Tom: You begin by saying Zeus is "better able to exercise providence" but qualify the claim with "Zeus might not have much more providential control than would Jove" but that this isn't the issue. I thought how much providential control foreknowledge provided *was* the issue. But maybe you just meant to say that Zeus would only have slightly more providential control.

Kevin: Yes, if Zeus has any more providential control than does Jove, then it is false that simple foreknowledge offers no providential advantage.

Tom: Let me ask this, Kevin. In Case 1 you mentioned no objective probability for X. So what is X’s probability of occurring? Is X’s occurring a historically contingent event? Is its probability of occurring 1 or less than 1? If less than 1 (viz., if it’s indeterminate), then what are the grounds for God’s knowledge that X ‘will’ certainly occur?

Kevin: Yes X is contingent. The probability of an indeterminate event depends what time you're asking about. At T1, the probability that X will occur is, as described above, .17. At T2, if X occurs, the probability is now (and forever after) 1 and if X does not occur, the probability is now (and forever after) 0.

Tom: Would you agree that foreknowledge of this [i.e., what you called undivided foreknowledge] sort offers God no basis upon which he is able to decide how to act providentially in the world (in influencing foreknown events, in bringing them to pass, in preventing them, etc.)?

Kevin: Yes, I don't see a way to resist affirming this (though I'd like to!). But I'm trying to see if there is a way that a version of foreknowledge that is less complete than this could be providentially advantageous. Are you familiar at all with David Hunt's work in this area? I think I'm trying to press a similar argument to the one that he does.

Kevin, Open Theism isn't just the position that denies exhaustive foreknowledge. There are a lot of positions that do that. What makes Open Theism special is that it sacrifices complete providence in order to preserve human freedom, and the motivation for the conflict here is because they adopt a libertarian account of freedom.

Also, compatibilism isn't the position that freedom is always and everywhere accompanied by deterministic causes. So, if you get, as Patrick argued, causal necessitation in the consequent of the conditional you asked about, Tom is right that this is compatibilism. To get all the way to Thomism, you have to add that all free actions are determined so as to provide a basis for complete providence, but that extension is not part of compatibilism proper. And to ask an Open Theist whether freedom is compatible in any case whatsoever with causal determination will be to get a clear "no".

Of course, there are related positions one might introduce here based on alternative positions to libertarianism and compatibilism. That's an interesting project to pursue, but it's confusing to label these as versions of Open Theism.

This point touches on your earlier post where you first defined OT and then identified four versions of it. I think your definition of OT fails precisely because it doesn't include the notion of freedom that is so central to everything open theists want to maintain. It is a position that endorses restrictions on omniscience and providence for the purpose of retaining a libertarian conception of human freedom. The restrictions are interesting in their own right, but they don't define the position. To see this, note that skeptics who think there is an omnipotent being who created everything out of nothing are not open theists.

Jon,

I agree with almost everything that you say in the first and last paragraph. I'm getting ready to leave for a 2 week vacation, so I don't have much time. But let me quickly say this. I agree that OT assumes a libertarian position. I didn't spell this out earlier when defining OT, but I was making that same assumption--maybe I should have made that clearer. But it doesn't seem right to me that libertarianism can never allow an action to be free and determined at all. Two quick reasons. First, causal determinism is usually defined as a global feature of a world, and it is this form of determinism that is usually involved in defining compatibilism and incompatibilism. So, incompatibilism is the thesis that it is not possible to be free if the conjunction of the state of the universe at one moment in time and the laws of nature entail the state of the universe at any other moment of time. Libertarianism is incompatibilsm plus the thesis that we are at least sometimes free (hence entailing the falsity of causal determinism).

Second, on the way that you define libertarianism, neither Bob Kane nor Eleonore Stump would qualify as libertarians (for they both allow that free actions can be causally determined by one's character, so long as one's coming to have that kind of character was free and not determined--see The Significance of Free Will and "Augustine on Free Will" respectively). Do you really mean to say that niether Bob nor Eleonore are libertarians? If their views are forms of libetarianism as I think they are, then I don't see a why an OT couldn't have this sort of view of free will in mind. Now, maybe they don't; that is, maybe no extant OT view allows for determination by one's freely chosen/formed moral character. But I don't see why such a position is in principle not an option.

Kevin, my plea is for some terminological regimentation here, and without it there's no good way to answer the question who is a libertarian and who isn't. But, for what it's worth, I think your account of libertarianism is unfaithful to the history of the dispute. Note that on your view it takes only one contracausal possibility in the entire history of the world to make every causally determined action free. You're right that it is the possibility of global determinism that motivates compatibilism, but the dispute between compatibilists and incompatibilists is not about whether global determinism is true or even possibly true. Notice that your last sentence of the first paragraph formulated as it is as a sufficient condition, is correct (libertarianism is incompatible with global determinism); but that claim alone surely underrepresents the libertarian position since you can't replace the 'if' here with 'iff' and have it be plausible at all (since global determinism doesn't follow from the impossibility of freedom).

I doubt there is anything philosophically at stake in deciding whether Kane and Stump can appropriately be labelled libertarians, and my appeal wasn't intended to claim that there is some nature to these positions that requires the terms to be used in any given way. My complaint was that when one uses the term 'open theist' in an expanded way (since I think it is pretty clear that they claim to value freedom for actions in the sense that an action is free only if the everything strictly about the past plus the laws doesn't entail that the action is performed), confusion results. Better to let the term pick out the view as they maintain it and ask about related views, and here we get interesting questions such as: with Stump's or Kane's view of freedom, is complete providence possible? with a denial of the law of conditional excluded middle, does the endorsement of counterfactuals of freedom allow complete providence? I think the answer to the first question is "no" and the answer to the second is "yes, though not with necessity". But that's stuff for a different post/paper!

Oops! The "no" about Stump and Kane needs to be qualified in terms of non-Molinism. If Molinism is true, it doesn't matter whether you have standard libertarianism or the related incompatibilist views of Stump and Kane.

Hi all. A lot to catch up on here, so this is gearing up to be a long post!

It looks like this is a debate concerning what certain theories of freedom amount to, and I certainly don’t think I’ve got anything much to add here.

As merely a biographical note, though, I will say that my impression (and Kevin’s) was that compatibilism is the view that freedom is compatible with *universal* causal determinism, and that any view of freedom that denies this is a version of incompatibilism. Hence, when talking about the freedom in the consequent of the conditional we’ve been discussing, I didn’t say it was freedom as the compatibilist conceives it (since it is *not* compatible with universal causal determinism). But I certainly don’t want to argue that compatibilism has to be understood this way (though, right or wrong, this is how I’ve understood it).

I do want causal necessitation in the consequent of the conditional, but as everyone has pointed out, this is causal necessitation that results from what one has freely done in the past in a non-causally necessitated way. Thus, the freedom here is a freedom that is incompatible with universal causal determinism. That is the important thing, I think, and the thing that makes the view amenable to an open theist who thinks (1) we’ve got to be free to do otherwise with respect to some decisions at some times to be morally responsible agents (all OTs have to affirm this) and (2) God can’t know what we’ll freely do in this way but (3) it is also sometimes the case that in exercising freedom to do otherwise we can make ourselves into people for whom some decisions are causally necessary (b.c. of facts about our psychologies), and that with respect to these decisions we are still free. (Kevin indicates that this is a common incompatibilist view.) Moreover, (4) when we do make ourselves into this sort of person, certain ‘would’ counterfactuals of freedom become true, on the basis of which God can know what we’ll freely do in the future because he sees that sufficient causal factors are in place for the antecedents of such counterfactuals obtaining.

Tom,

From what I can tell, this is your view too, right? – except that you think the freedom involved with (3) and (4) above should be called (or is) compatibilistic freedom, and with Kevin, I’m not sure that it is best called this. But nothing much hangs on this, I don’t think. We both think that the freedom in (3) and (4) is of course incompatible with *universal* causal determinism.

Jon,

You say “And to ask an Open Theist whether freedom is compatible in any case whatsoever with causal determination will be to get a clear "no". Of course, there are related positions one might introduce here based on alternative positions to libertarianism and compatibilism. That's an interesting project to pursue, but it's confusing to label these as versions of Open Theism.�

This is interesting. Gregory Boyd and Tom think that the freedom required for moral responsibility is sometimes compatible with causal determinism (namely, in the sense described above). (Incidentally, I do too.)

But we also think that freedom to do otherwise rules out God knowing what we’ll do in this way, and that we are actually sometimes free to do otherwise, and so, as you say, we sacrifice complete providence in order to preserve freedom. You seem to think, though, that holding this first point, despite holding the second, warrants being called something else than an open theist.

But there are fewer more prominent open theists than Boyd, and the first point is very much in the spirit of what he often argues. Tom, I’m sure you’d be able to say much more about this than I, but Boyd often argues that we sometimes freely make ourselves into certain sorts of people, at which point our characters are sufficiently hardened so that God knows what we’ll do. (Boyd says just this to handle Jesus’ prediction of Peter’s denial, and Boyd thinks that Peter freely denied Christ.) Boyd and Tom call the freedom we have at this point compatibilistic, and hence, compatible with causal determinism.

I’ve just been putting it in terms of psychological necessity in order to make clear that this can be a sort of causal necessity, and, to my knowledge, Boyd doesn’t say just this. But it doesn’t seem to be out of line with what at least one prominent open theist says about certain decisions we make.

But, indeed, maybe we need a new label! I just wanted to point out that some people who are definitely known as OTs in the literature hold a position that says that morally responsible freedom is sometimes compatible with causal determinism. I agree, though, that it is confusing, since these OTs don’t hold strictly to a view of freedom that says that morally responsible freedom entails freedom to do otherwise (what I’ve been referring to as lib. freedom, but Kevin seems to have said that this isn’t essential to being called a libertarian, so I’ll remain agnostic about it). And this view of freedom is, of course, the view that drives OT. (What overall view of freedom it is that these OTs endorse is a question I'll leave for another day, important as it is!)

I don’t know how to clear this up, but since Boyd is definitely known as an OT, maybe it’d be easier to still refer to this view as OT, but just a modified sort. What do you think? More importantly, though, the view still retains what originally drives OT, namely, that freedom to do otherwise rules out God’s knowing what we’ll do in this way, and that we are sometimes free to do otherwise.

Well, that should do it for one post! Kevin, I’ll try to address your response to the dilemma I raised for your examples hopefully sometime in the next day.

Jon,

Two last points before I have to pack. First, you say "the dispute between compatibilists and incompatibilists is not about whether global determinism is true or even possibly true." Correct. But it is a debate about what would follow from the truth of causal determinism. Incompatibilists say that this would entail that there is no free will and compatibilists would say the lack of free will doesn't follow from the mere truth of causal determinism.

Second, you also say "Notice that your last sentence of the first paragraph formulated as it is as a sufficient condition, is correct (libertarianism is incompatible with global determinism)." I never said that the libertarian thinks that the falsity of causal determinism is a sufficient condition for free will, just a necessary condition. There may be (and very likely are) other necessary conditions. These might include: an epistemic condition, an authenticity condition, a sourcehood condition (it is this latter one, I think, that would disallow some kinds of 'local' determinism but not the kinds of determinism by character that I and Patrick have in mind here). Nor did I say that the falsity of causal determinism is a sufficient condition for libertarianism being true--again, it's just a necessary condition.

I feel like I must be missing at least part of what is motivating you here, but I'm not able to see what it is. Maybe this is just a drawback of holding such conversations via blogs and not in person. But regardless, I've both enjoyed and appreciated the conversation with you, Patrick, Tom and the rest. I hate to bow out of the conversation at this point, but London awaits. Maybe I can catch up with the rest of the conversation when I return.

As an open theist I’d agree with Jon on this—if “complete providence� be understood as exhaustive and unconditional determination of the minutia of creation by God’s will then open theists can’t (along with other non-open theists like traditional Arminians and Molinists) can’t be seen as believing in divine providence. We don’t think God relates to the world that way. But whereas Arminians and open theists agree that the future is (how shall I say it?), ontologically open (future indeterminism is real and not just epistemic), Arminians and Molinists share the Calvinist’s conviction that God knows the future in blueprint form as it were, so that