Sanctified Agents?

I maybe wrong here, but I always thought that Plantinga was saying something like that it is epistemically possible that there are metaphysical constraints on the worlds God can create. So, "for all we know" it might not be metaphysically possible for God to create a world in which agents always freely choose the good.

I've never understood how Hawthorne & Howard-Snyder's argument threatens that. I guess that means I've probably misunderstood Plantinga!

“There is another world W3 that includes some transworld depraved agents and is such that W3 is better than W2. Just let W3 = [W2 + (E & F)]. Since (E & F) is positive, we know that W3 = [W2 + (E & F)] is better than W2.”

I think that only works if you count F twice. After all, “W2 is a world in which the only free agents are a transworld sanctified agent or two.” That means W2 already has F, which is “the value of having agents exercising their freedom.” Thus, part of the value of W2 is that it is a world with F. It would be misleading, then, to calculate the value of W3 as [W2 + (E & F)], since F is already included in the value of W2. Rather, the value of W3 is simply W2 + E. On this ranking, W2 outweighs W3, since E is a negative.

I’ve probably screwed up here, so let me know :-)

“If (1) God should actualize a world with some transworld depraved agents rather than a world with no free agents at all, then (2) God should actualize a world with some transworld depraved agents in addtion to the transworld sanctified agents rather than a world with only transworld sanctified agents. Since the FWD shows (1) is true, it shows that (2) is true.”

Unless “a world with only transworld sanctified agents” already contains F. If so, then it’s not clear how tossing in a few TWD agents enhances the value of a world that already has F. It looks like it would just add E, a disvalue.

Hey Mike,

I think I understand why Hawthorne and Howard-Snyder thought TWS mattered. I thought in the course of the FWD, Plantinga had claimed that it was not within God's power to create a world with moral good unless it also contained at least one instance of evil and defended the claim that this is possible by defending the further claim that every essence could suffer from TWD. Hawthorne and Howard-Snyder respond by saying that there's more reason to affirm this than there is to affirm the possibility of TWS so Plantinga had not really offered a satisfactory defense. Your point, if I follow (I might not, it's late), is really that H & H-S missed something important. And that is that for any world, W, such that W contains these agents with TWS, there is a world, W', such that we add in a few agents that suffer from TWD and the result is that W' > W. Is the idea then that in light of this point, the FWD is reinstated?

Is this really a version of the "no best" response (i.e., there's no limit to how many times God can add some agents that suffer from TWD to create a world better than a world consisting only of agents that can do no wrong)?

As an aside, to be honest, I’m having trouble linking this issue of additive value with Hawthorne and Howard-Snyder’s original point. If I’m thinking of the same article you are – “Transworld sanctity and Plantinga’s Free Will Defense,” IJPR Vol. 44 No. 1 (August 1998) – their point is to use an ‘epistemic amendment’ argument such that if it’s reasonable to refrain from thinking that TWS is impossible, then it’s reasonable to refrain from thinking that TWD is possible. The former *is* reasonable, and therefore so is the latter. And so we’ve been given no good reason to think that TWD is possible, which unfortunately is the central premise of the FWD. (To this is added a gazillion pieces of advice on how Plantinga could have better constructed the FWD.) Isn’t that how H & H-S deploy the concept of TWS against the FWD? It’s an epistemic argument that has little to do with the concept of additive value, and everything to do with whether we have good reason to accept Plantinga’s key premise.

Now on to your reply. OK. I agree that Plantinga’s assumption in the FWD is exactly as you later state to Clayton: “a world with free agents that sometimes go wrong is better than one with no free agents at all.” And I guess a corollary is that “worlds can be vastly improved by the addition of free agents.” So let’s rank the alternatives:

W-TWS: a world with only transworld sanctified agents, exercising their freedom, and always doing good.
W-TWD: a world with only transworld depraved agents, exercising their freedom, and always ending up doing some bad things.
W-D: a ‘desert’ world, having value of various sorts, but lacking any agents exercising freedom.

Any W-TWD is going to be more valuable than any W-D, since W-D lacks freedom.

In addition, if we keep the number of free agents the same, any W-TWS is going to be more valuable than any W-TWD, because they have the same amount of F, but W-TWS lacks the disvalue of E. (The assumption is that while F+E is a net positive, E has disvalue, so F > F+E.)

Finally, as you note, any W-TWS can have its value enhanced by adding in more free agents. Even if these are TWD-agents, the value of W-TWS will be enhanced, because now it has more freedom, and the corollary above holds. In that case, W-TWS would be converted into (the more valuable) W-TWS/TWD, a world with both sanctified and depraved agents, all free.

OK. But while the value of any W-TWS *can* be enhanced by adding TWD-agents, it would be even *more* enhanced by adding TWS-agents. That’s because each TWD-agent adds F+E, whereas each TWS-agent adds F but no E.

So yes, adding in more free agents to W-TWS will always enhance value, even if they are TWD-agents. But does that mean God should go for that addition *if there are better options available to him*? Isn’t that the issue? (Maybe it isn’t.)

Let’s return to the conclusion:

“Conclusion (1):

If (1) God should actualize a world with some transworld depraved agents rather than a world with no free agents at all, then (2) God should actualize a world with some transworld depraved agents in addtion to the transworld sanctified agents rather than a world with only transworld sanctified agents. Since the FWD shows (1) is true, it shows that (2) is true.”

The idea is that if God should actualize W-TWD over W-D, then God should actualize W-TWS/TWD over W-TWS. But what if W-TWS can be enhanced, not by adding more TWD-agents, but by adding more TWS agents? In that case, adding TWS-agents to W-TWS adds more value than adding TWD-agents to W-TWS. In the former case you get more F, but without the disvalue of E, whereas in the latter case you get F+E.

In short, adding TWD-agents adds net value, but adding TWS-agents adds more value per agent added. So why would God add the former to enhance the value of a world, if adding the latter were open to him?

Man, this was wordy. Sorry about that. Feel free to ignore this and just comment on the first paragraph, if you’d like. I suspect we’re not even getting close to the reason why H & H-S think TWS is a problem.

Thanks, Mike, that was helpful. Here's my $.02 on the matter. H & H-S do a nice job calling into question P's claim that it's possible that God cannot create a world containing moral goodness that does not also contain evil. Your response shows that their word is hardly the last word on the FWD, but it still seems that they might be right to say that P can't just say that it's possible that God cannot create a world containing moral goodness without some evil and let that by the final word on the matter. I suppose in response to your point, someone could say (as has been said in response to the 'no best' response) that there are obligations to refrain from adding certain agents knowing that they would perform evil acts even if such an addition makes this better overall, but if this response works, I support it's fair to say that we did not need H and H-S's response at all.

I'm still left worried that there's something odd about the suggestion that God could create a world, W, that contained creatures with creaturely essences that are TWD when there is available a choice to create a different world that is better and contains no creatures whose essences are TWD. Is there a way for you to avoid having to say this?

Hi Mike,

You wrote:

Suppose HH are right and that there is at least one TWS-essence in every world. Now consider the possbility that there are, in addition to a TWS-essences, mostly TWD-essences. On Plantinga's behalf, this seems at least possible

Just to be sure: you mean that this seems epistemically possible, in the sense that we can't rule it out as impossible (although it could turn out that it actually is metaphysically impossible), right? For, at least by my lights, this doesn't seem metaphysically possible; nor does it seem metaphysically impossible. Rather, it appears to me as undecided.

Short version: if we concede (P*), then Plantinga’s (35) is no longer possible, in which case the FWD collapses, doesn’t it? (P*) implies that it *was* within God’s power to create a world containing moral good but no moral evil: just instantiate the TWS essence(s).

Longer version: OK, you concede their (P), in part because you think their epistemic amendment argument is “just intuition trading” and “not at all to the point”. In fact, you concede the stronger (P*), because “nothing interesting follows for FWD”.

But if a TWS essence was open to God to instantiate, doesn’t the FWD collapse? After all, when Plantinga defines the concept of TWD in _God, Freedom, and Evil_, he immediately points out that if we do not *universalize* this thesis – if we instead hold that only some but not all creaturely essences available to God to instantiate have TWD – then it was open to God to just create other people, who weren’t thus afflicted:

“Now we might think that this [defining the concept of TWD] settles the question in favor of the Free Will Defender. But the fact is it doesn’t. For suppose all the people that exist in Kronos, the actual world, suffer from transworld depravity; it doesn’t follow that God could not have created a world containing moral good without creating one containing moral evil. God could have created *other people*. Instead of creating us, i.e., the people that exist in Kronos, he could have created a world containing people, but not containing any of us – or perhaps a world containing some of us along with some others who do not exist in Kronos. And perhaps if He’d done that, He could have created a world containing moral good but no moral evil” (GFE 49).

Plantinga’s answer is to argue (after a long discussion) that “it is possible that *every* creaturely essence… suffers from transworld depravity” (53, my emphasis). If so, then “it’s possible that God could not have created a world containing moral good but no moral evil” (53). Plantinga concedes that, given this possibility, God “might have been able to create worlds in which moral evil is very considerably outweighed by moral good” (53). But he regards that as irrelevant as a way of rebutting Mackie. What matters is that the possibility of ‘global TWD’ implies that “it was not within His power to create worlds containing moral good but no moral evil – and this despite the fact that He is omnipotent” (53).

So if Plantinga is going to rebut Mackie, and prove that God and evil are logically consistent, then he needs the thesis that it’s possible that every essence has TWD. (P*) contradicts this. If (P*) were the case, then it was open to God to create a world which contained moral good but no moral evil. If so, then we don’t have a proof that God and evil are logically consistent, because we no longer have a model within which (1) and (3) are both true. (In short, we can’t deduce (3) from (1), (35), and (36) taken together, if (35) is not so much as possible.)

I think you’re right that adding TWD agents involves adding net value to a world, such that the moral evil might get outweighed by either moral good or the value of freedom itself. But that, by itself, doesn’t show that God and evil are logically compatible. In addition, we need (35). Or are you saying we don’t need Plantinga’s (35) in order for Plantinga’s FWD to work? Perhaps (for instance) it’s just too strong of a claim, and Plantinga shouldn’t have hinged his case on it?

Hi Mike,

I'm not so sure. The argument that this is epistemically possible but metaphysically impossible would have to appeal, I think, to some necessary a posteriori proposition with which it is inconsistent. So, for instance, it would have to be inconsistent, say, with water being H2O. But how could any necessary a posteriori proposition present a metaphysical problem here? God determines in each world what is in the lakes and streams, and the same goes for every a posteriori necessary proposition. The contingent facts upon which they depend are things God can determine. My guess is that you're actually worried that the situation Plantinga describes is not epistemically possible.

I'm not quite going so far as to say that the world under consideration is epistemically possible but metaphysically impossible; I'm making the weaker claim that it's epistemically possible, but I can't tell whether it's metaphysically possible or not.

I'm thinking of Yablo's (1993) criterion of modal appearance as a necessary condition for the PFJ of a modal claim. (Leaving aside Yablo's two-part account of modal verification, at least for now) In order for the claim that P is possible to be PFJ for me, it must be accompanied by a sense of conviction that P is possible. But I lack such conviction about Plantinga's claim; so, at least with respect doxastic force, Plantinga's claim doesn't appear to me as possible. Of course, I also lack conviction that Plantinga's claim is impossible, as that claim lacks doxastic force for me as well. Therefore, Plantinga's claim is undecidable for me, as things currently stand.

As I said, I don't want to commit to the impossibility of Plantinga's claim, but about the point re: a posteriori necessities: couldn't there be a relevant a posteriori necessity claim here? For example, suppose, as is not terribly implausible, that it's metaphysically necessary that all agents have compatibilist freedom at best...

Actually, I'm not primarily concerned about whether it's metaphysically necessary that agents have compatibilist freedom (at best). The primary problem I mentioned was the absence of doxastic force. I don't (yet) see a reason why a posteriori necessities are the only sorts of doxastic force-stoppers.

In any case, I'm not sure I quite have your account of modal epistemology straight. Is it basically Chalmers' account -- primary intensions (backed by objectual imaginings that pass Yablo's two-clause modal verification test) as PFJ of metaphysical possibility?

Do the contributors to the conversation so far have a judgment on whether a solution to the problem of heavenly freedom would or would not have provide an adequate defence of the FWD here at this point?

Mike,

Isn't the problem of heavenly freedom essentially Mackie's objection? Do you take Hawthorne's objeciton to be essentially different than Mackie's?

If part of the solution to the problem of heavenly freedom were to explain why God can't create TWS agents in any possible world, that would seem to be of help here.

Mike,

I think that is on the right track. I take Mackie (as well as Oppy) to argue that if agents can be free and morally impeccable in heaven, then they can be so on earth. If God could have created a segment of a world where moral impeccability and freedom are compatible, then he could have created a whole world where that is so. If this is so, the FWD fails.

To be sure lots of theists disbelieve many traditional doctrines, but Mackie's objection touches those who adhere to a traditional perspective on heaven. Platninga seems to be one of them in thinking that agents in heaven are free and morally impeccable.

Hi MIke,

I'm taking doxastic force to be (at least) a constituent of whether something seems or appears to one to be the case; I'm taking the way things seem as PFJ for how things are; and I'm taking the sort of modal evidence at issue to be modal seemings. So, I'm taking the absence of doxastic force as an indicator that the modal claim in question isn't PFJ.

Mike: "It is perfectly consistent with this assumption that TWD-agents also do at most one thing wrong. It is also perfectly consistent with this assumption that every TWD-agent performs a wrong action before they die. If that is true, then heaven could be filled with TWD-agents who never go wrong (i.e. never go wrong in heaven)."

Paul: What about an embryo (or whenever you think the fetus is a person) who dies in the womb? There was no "wrong action performed before they died" and yet in heaven (assuming those human persons who die in the womb go to heaven) they also never perform a wrong action. Yet they are free in heaven, and never go wrong, and never went wrong.

No, I do want to take seemings or appearances as PFJ in modal matters; it's just that Plantinga's TWD claim doesn't seem or appear to me as possible or as impossible; it's undecided for me.

Think of Van Cleve's weak/strong conceivability distinction. Weak conceivability is the inability to see that P is impossible, and strong conceivability is the ability to see that P is possible. Since only strong conceivability involves the appearance of possibility, only it is PFJ for possibility. And the problem, it seems to me, is that Plantinga's TWD claim is only weakly conceivable.

I'd have to think about the analogy for a while, but off the cuff, the account strikes me as too permissive. Suppose I try, several times, to falsify the claim that, possibly, there is a run of five 7s in the decimal expansion of pi, each time failing. Do I now have decent evidence that, possibly, there is no such run (a point from Van Inwagen)? And of course there's always the Goldbach's Conjecture case...

For a non-math case, how about the following: I try, several times (several centuries?), to falsify the theory that, possibly, the natural world exists and God does not. No luck so far. It looks as though I have a good reason to deny that there is -- and isnt! -- an Anselmian being on your account, no?

I'm not sure how the contradictory claims get justified.

I try, several times, to falsify the theory that maximal excellence is exemplified in at least one possible world. On your account, I take it, I'm now justified in believing that maximal excellence is exemplified in at least one possible world, in which case (assuming I accept Axiom S5, and I do) I'm justified in believing it's exemplified in all possible worlds, including the actual world.

I try, several times, to falsify the theory that there is a possible world at which there are just the fundamental particles (or strings, or...), and all else logically supervenes on that. On your account, I take it, I'm now justified in believing that there is such a possible world.

Something has to give. It seems to me that it's your account, no?

I'm not so sure. To be charitable to Van Cleve, I think his point is that merely failing to see that P is impossible is not, all by itself, evidence of P's possibility (How could mere ignorance increase the positive epistemic status of a proposition?). I'm not sure what he would would say about weak possibility if it were modified so as to account for your qualification re: (many) failed attempts at showing P's impossibility.

Mike,

Sorry for the delay; I took Saturday and Sunday off.

OK, your reply seems to answer all the questions I had. Essentially, you are replacing Plantinga’s (31) “Every essence suffers from transworld depravity” with a weaker thesis, namely, (P*). At first I thought that involved the abandonment of what Plantinga regarded as crucial to the FWD. But I was mistaken. On pp. 189-90 of NN Plantinga says that “the conjunction of (31) with (32) is not the only proposition that can play the role of r in the Free Will Defense,” and he goes on to specify a (33) that is weaker than (31), but still does the trick. It’s not the same as the TWS possibility/necessity you’ve been raising, but it’s in the same spirit. So there’s flexibility here that I was missing. Thanks for taking the time to make this clear for me.

However, it seems to me that “FWD implicitly assumes that God actualizes the best world he can” is a bit strong. I didn’t find that implicitly in NN 164ff. Yes, a world containing free creatures “is more valuable, all else being equal, than a world containing no free creatures” (166). But “perhaps for any world you pick, there is a better” (168). In that case, the FWD would be implicitly assuming something that God cannot do. Alternatively, perhaps God only actualizes a best *kind* of world, such as a world with significant freedom. That would explain the preference for a free-agent world over a no-agent world.

Yes, you're right; feasibility does the trick. The feasible worlds could be ranked in value, and it could just be a brute fact that, given the relevant counterfactuals, there is a best feasible world. Any better worlds might be possible, but not feasible.

In the recent January 2009 PPR, Plantinga responds to Hawthorne and Howard-Snyder, starting at p. 183.