A problem with strong actualization

Wow. heavy duty, but great stuff!

This is nice, Alex. You write,

God strongly actualizes each of two states of affairs whose conjunction entails Eve's freely refaining

So, the argument is supposed to go like this.

1. If Eve doesn't freely refrain, God will create life on Pluto.

2. There is no life on Pluto.
--------------------
3. :. Eve freely refrains.

And we are assuming that strong actualization transmits across entailment, (1) and (2) are strongly actualized, and so God strongly actualizes (3). But there is no question, I think, that Eve strongly actualizes her refraining from eating the apple. She does that. So, there is an argument at least as good that Eve + God together strongly actualize that there is no life on Pluto.

1. If Eve doesn't freely refrain, God will create life on Pluto.
3. Eve freely refrains.
---------------------------
2. :. There is no life on Pluto.

A better conclusion, I think, is to deny transmission of strong actualization across entailments. The fact that P entails Q does not itself entail that in any world where P obtains Q does. Consider,

1'. I strongly actualize this sentence.
2'. If I strongly actualize this sentence, then 2+2= 5.
3' :. I strongly actualize that 2+2 = 5.

So, when God actualizes T, it is evident that T entails that 2+2 = 5, but the latter state of affairs does not obtain. So not everything that is entailed by what we strongly actualize is something we strongly actualize or something that so much as obtains.

An even better conclusion is to hold that God does not strongly actualize the state of affairs of Pluto not having life on it, and neither does Eve. Presumably, he's supposed to strongly actualize (cause) that state of affairs by refraining from bringing about (causing) life on Pluto. But it follows immediately from this position that God causes all free action by refraining from causing it to be otherwise. So refraining from causing some contingent state of affairs p to be otherwise does not in general entail that God causes p to be as it is.

S1: God tells Gabriel that he'll refrain from creating a ringed planet in the alpha Centauri system if Eve doesn't freely refrain. S2: God creates a ringed planet in the alpha Centauri system. In W, both S1 and S2 obtain.

1. If Eve doesn't freely refrain, God does not create Saturn
2. God does create Saturn (say, to give it a name).
3. :. Eve freely refrains.

Now (1) and (2) entails (3), therefore (assuming transmission) Eve is not free. That's curious, since (1) and (2) are true, the argument is valid, but(3) is false. But the falsity of (3) entails (along with (2)) that (1) is false. Contradiction! God asserted (1), so (1) cannot be false. Your premises entail a contradiction, so I don't think you are describing a possible case.

The problem is that the FWD requires transmission of strong actualization across entailments.

Can't be right, at least not perfectly generally. For any state of affairs P that God strongly actualizes, P entails (Q & ~Q). But the state of affairs Q & ~Q is obviously not strongly actualized by God or anyone else, since it obtains at no worlds. So it is not in general true that entailment preserves strong actualization.

Does God cause Eve to freely take or freely refrain from taking? One might argue that there's a third option. If God allowed Satan to coerce in a way that violated her freedom, she might take without sinning. Does what God causes preclude that? I'm not sure it does on a typical libertarian account.

Hi Alex,

(1), (2), and (3) are not compossible, as I think the comment above shows. G's strong actualization is incompatible with E's libertarian freedom. Let [Gs] be the God strongly actualizes operator. We have assumed that entailment preserves strong actualization. The argument looks like this,

(I)
1. [Gs](Eve doesn't freely refrain -> God does not create Saturn)
2. [Gs](God does create Saturn)
3. :. [Gs](Eve freely refrains)

But it is impossible that [Gs](Eve freely refrains), since no one can strongly actualize a state of affairs in which Eve freely refrains except Eve. If on the other hand you time index the operators, then there are no worries.

(II)
1. [Gs]t-1(Eve doesn't freely refrain -> God does not create Saturn) (T)
2. [Gs]t+1(God does create Saturn) (T)
3. :. [Gs]t(Eve freely refrains) (F)

Here, (1) and (2) do not entail (3). God cannot strongly actualize a state of affairs that already obtains in past, so God's failing to create Saturn at t+1 together with his strongly actualizing the conditional in (1) at t-1, does not entail that he strongly actualizes Eve's freely refraining at t. The propositions that are consistent are these, letting [Es] stand for Eve's strong actualization.

(III)
1. [Gs]t-1(Eve doesn't freely refrain -> God does not create Saturn). (T)
2.[Gs]t+1(God does create Saturn) (T)
3. [Es]t(Eve freely refrains) (T)

It is explicit in Plantinga that if God strongly actualizes S and S includes S*, then God strongly actualizes S*. It is implicit that entailment is sufficient for inclusion.

Well, right, as I mentioned above the transmission principle would have to be restricted, but the restriction to contingent propositions is not enough. There are contingent propositions describing free actions that cannot be strongly actualized by anyone other than the agent of the action, as illustrated in argument (I) above.

(1)-(3) are ambiguous, on some readings they are compossible. I gave one of them, and it was (III),

(III)
1. [Gs]t-1(Eve doesn't freely refrain -> God does not create Saturn). (T)
2.[Gs]t+1(God does create Saturn) (T)
3. [Es]t(Eve freely refrains) (T)

On (1*)-(3*):

1*. Alex doesn't freely refrain from eating an apple → God doesn't create Saturn
2*. God does create Saturn.
3*. Alex freely refrains.
They are compossible because they are true.

If God strongly actualized (1*) and (2*)--say, by asserting (1*) sincerely and creating Saturn--then you did not freely refrain this morning. But if God did not assert (1*), then you strongly actualized (1*) by strongly actualizing it's antecedent.So, we can see that (1*)-(3*) are compossible in this way.

1*. [As](Alex doesn't freely refrain from eating an apple → God doesn't create Saturn).
2*. [Gs](God does create Saturn).
3*. [As](Alex freely refrains).

What are not compossible are the following:

1*. [Gs](Alex doesn't freely refrain from eating an apple → God doesn't create Saturn).
2*. [Gs](God does create Saturn).
3*. [Gs](Alex freely refrains).

But if God did not assert (1*), then you strongly actualized (1*) by strongly actualizing it's antecedent

i.e., by strongly actualizing the negation of its antecedent.

Right, I agree with this, which I guess is the rational thing to do. Let me know what he says.

We're now puzzled about this simplified puzzle. Suppose God says: "Eve will freely do A." Let S be the state of affairs of God saying "Eve will freely do A."
Then, if God strongly actualizes S, God strongly actualizes everything contingent entailed by S, and that includes Eve freely doing A. So that's absurd

It's not a serious problem. If God utters at t-1, 'Eve freely does A at t', then it is perfectly possible that Eve freely does A at t. God's utterance is a soft fact (Plantinga takes it this way, and for what it's worth so do I). So Eve could, in spite of the utterance, do ~A at t. If she were to do so, .....well, you know the rest of the story.

Surely, God is strongly actualizing his utterance. I don't think there is any way around that. What we have to conclude I think is that possibly, God's strongly actualizing p and [](p -> S does A), God does not strongly actualize S does A. That is to deny that transmission holds generally.

But Plantinga does not need (what might be called) a strong transmission principle. He does not need it to be true in all cases that strong actualization is closed under entailment. What he needs, really, is a restricted entailment principle which governs inferences from atomic propositions (states of affairs) to more structured propositions. (e.g. ([s]p & [s]q) -> [s](p & q), but not the converse). God strongly actualizes a set of atomic states of affairs. From those nothing untoward follows for His strong actualization of the larger states of affairs they compose. The problems come the other way around, from strongly actualizing larger to strongly actualizing smaller. So transmission needs to be restricted in one direction.

It would be interesting to work out the logic of strong actualization which already looks like it would have a minimal model.

Plantinga's argument in Profiles needs the claim that if T(W) entails p, where p is atomic and contingent, then God strongly actualizes p. Since T(W) is not atomic, a restriction to atomic antecedents will not do the job.

Yes, it will. Note that T(W) is just shorthand for [s]p, [s]p1, . . ., [s]pn, for all p in W. So we get the inference without a problem, [s]p -> [s]p. No worries.

Rather I should make that a conjunction,

1. T(W) = [s]p0 & [s]p1 & . . .& [s]pn

and

2. [][([s]p0 & [s]p1 & . . .& [s]pn)-> p0]

Morever, if p is the proposition that Eve freely did A, then Said(God,y,p) ("God said p to y") seems to be atomic, so the restricted entailment principle should hold for it.

Neither 'Said(God,y,p)' nor '("God said p to y")' are atomic propositions, though of course every proposition has the form of an atomic proposition.

Or, conversely, from:
2'. T(W) does entail p0 and p0 is a contingent proposition
to: 1'. God strongly actualizes p0
In neither form do we get to assume that p0 is one of the conjuncts in T(W).

I'm proposing that to say God strongly actualizes T(W) is just to say that God causes EACH atomic state of affairs in W. I honestly don't know how that could be false. It would then follow p0 is a conjunct. Things are a bit more complicated, of course, since [s]p0 will also be a conjunct along with p0. God causes both states of affairs to obtain when he strongly actualizes p0. But I think this complication is manageable.

Alex, you say this above,

Plantinga's argument in Profiles needs the claim that if T(W) entails p, where p is atomic and contingent, then God strongly actualizes p. Since T(W) is not atomic, a restriction to atomic antecedents will not do the job.

Your worry (re-quoted here) was apparently that T(W) is not atomic and so the proposed restiction on tramission won't help. I showed you how T(W) could entail p given the restriction I proposed. The restriction states that ([s]p & [s]q) -> [s](p & q), but not vice versa. The restriction is observed in this case under the assumption that God strongly actualizes T of W only if for all atomic and contingent propostions p in T, [s]p. So I think this problem is solved. I'm not sure what other problem I'm supposed to be solving.

Plantinga now needs to argue that T(W) does not entail e.

I'm not sure wht Plantinga needs to show this, since we've agreed (haven't we?) that it is possible that [s]e. But suppose we now disagree about that. In that case, the basis of our disagreement is that for any free action A and agent S, necessarily, if is it strongly actualized that S does A, then S strongly actualizes that S does A. In that case, T includes the restriction that no state of affairs describing the free actions of created agents is in T, otherwise a contradiction follows.

Plantinga needs to argue that T(W) doesn't entail e, because he needs to argue that T(W)→e is contingent. If T(W) entails e, then T(W)→e is a necessary truth, and the FWD fails.

The idea is then that necessarily, God can actualize any possible world? Is that it? This sounds like Otte's PPR argument. It does show that FWD fails. It is easy to show that Otte's argument fails.

It's part of a large project. I hope I can send you a copy for comments on the Otte/Pruss objection to FWD!!