Hitler was not the first tyrant. Nor was it even the first case of genocide.
Perhaps you are right about God’s purpose (if he existed) but such a god is hardly omnibenevolent. We let our kids learn from their mistakes also, but not to the extent that they broke anything. It also doesn’t cover the problem of natural evil. No matter, I don’t know any god that gives us free rein. The one I grew up with was a hell of a micromanager.
I actually agree with you re: logical conflict. I just think that the Leibnizian defense is only trivially true in that a lot of improbable things ARE possible. The key is that the logical problem of evil asserts incompatibility, which entails an incredible burden. But if I drop the claim of incompatibility, the burden shifts to the theist to show that ALL instance of evil is necessary. After all, one need only find one instance of gratuitous evil and the evidential problem of evil succeeds in showing that the God of classical theism doesn’t exist.
I just realized that put this way, the argument actually reduces to total triviality – ‘assuming there exists a better world than this one, then god can’t be omnimax’ is equivalent to ‘assuming god isn’t omnimax, then god can’t be omnimax’, since only if god isn’t omnimax can there be a better world. (The same doesn’t work the other way around, just so nobody gets ideas: there are countless ways there might not be a better world without god being omnimax – or even existing at all --, first of all the possibility that this might be the only possible world.)
And there’s the rub – how do you ever show that some instance of evil is gratuitous? After all, this precisely amounts to showing that there is a world which is both possible and better than this one. That’s the burden of proof, and it lies squarely with the one propagating the POE attack.
By showing that there is no probable good that could come out of it. A theist could respond that a good beyond my ken could be found, but that’s like saying that a legitimately good moral intention can be found in Hitler’s killing the Jews. Sure, it’s possible, but I don’t think it very likely. I think, the burden is still with the theist to show that every evil that “appears” to be gratuitous to most humans are necessary.
Probable just won’t do. Again, think of bets: there are bets where it’s improbable that you loose, yet on average, you will make a deficit accepting them. This doesn’t work when it’s impossible that you can loose. You’re arguing that there’s a point of improbability where it becomes basically the same thing as impossibility, but that’s just not so.
Yet, even if you showed that there is no possible good to come from a certain evil act, from there it doesn’t follow that the world is necessarily worse off if it occurs: it may be that a possible world in which some evil occurs a completely incidental good thing happens that more than offsets it, without the two being in any way causally linked, or perhaps linked in an indirect way by having a common cause, and thus, this world is still better than one in which said evil doesn’t occurs, at the expense of the good that also doesn’t happen.
Well, if we think BPW works (which I still don’t), and then apply its conclusion to the premises of POE, sure.
But it would be incorrect to characterize POE as including non-omnimaxness as a premise – it does not.
I did actually use the words “logical consistency”, not possibility.
Say if someone were to present me with an argument, I can’t say: “I’ll only accept that argument if you can prove that it is logically consistent”, because that’s impossible. You can demonstrate that you’ve followed the laws of logic, and you can try to find agreement on the premises, but none of that guarantees that there is not an implicit contradiction somewhere.
So the burden of proof goes on trying to show that something is logically inconsistent. This is done by showing a contradiction.
Of course, and I said this. The point is, that the methods of the creationist and the scientist are different, and the creationist methodology is flawed.
The creationist will devise ad hoc excuses to defend a conclusion. This isn’t science.
My point was that BPW does a similar thing – to support the conclusion that this is the best possible world, it throws in an ad hoc “maybe all better worlds are impossible”.
Well to follow this analogy through, our universe is a very curly, very long line indeed. There are infinite straighter, shorter lines that we can conceive of, including lines that only differ minutely from our own, and no apparent reason why god couldn’t do a better job.
Not sure I understand what you’re saying here. The problem of evil posits the contradictory nature of omnimaxness and the existence of evil – that’s one of its premises, right? This amounts to saying that ‘if a better world is possible, god isn’t omnimax’. If you now assume a better world is possible, you render the whole thing circular. Leibniz’ argument doesn’t enter into it at all; I’m not sure what you mean by ‘applying its conclusion to the premises of the POE’.
Well, you can try to pull a tortoise and come up with something like Carroll’s paradox, of course, but in general, if the premises are valid, and the logic is sound, you’re forced to accept the conclusion. Logical manipulations preserve truth values in such a way that if you start out with something true, you end up with something as true. There aren’t any ‘hidden contradictions’ that might suddenly pop out of the woodwork; that’s like saying, even though our algebra gives us 4 as an answer for 2 + 2, it might really be 33.
How’s this ad hoc in any way? This is a possibility from the very beginning, if there are any worlds at all that are impossible, i.e. if god is bound by logic, for example. From this, it directly follows that any given world may be impossible; it’s not added in or tacked on as a defence of the argument, it is the argument.
Yes, but that’s just because we do not know the geometry.
No, I’m not arguing this. What I’m arguing is that an instance of gratuitous evil is enough to show that classical theism is false. Every instance of evil that appears gratuitous makes it likely that there is indeed gratuitous evil. The theist who wishes to deny this MUST show that ALL instances of evil is necessary.
In other words, you’re saying that the evil act was necessary for the good thing to happen - which I do not discount. What I’m merely saying is that the theist would have to argue that ALL appearances of evil ARE necessary, in order to preserve classical theism. I find that burden to be much, much greater than my showing that SOME of the evil I see are gratuitous.
No. The theist (or anybody else arguing in favour of Leibniz principle) does not need to deny that any instance of apparently gratuitous evil makes the existence of gratuitous evil more likely; I’ve agreed with that several times within this thread. But that’s not the point. The thing is, again, possibility is sufficient. And thus, anybody challenging the argument must eliminate this possibility, and hence, show that (at least) one instance of evil is gratuitous; otherwise, it remains possible that none is. That’s just the misunderstanding I’ve tried unsuccessfully to argue out about the past two pages, and it is exactly where the analogy to the whole bet thing lies.
That’s not what I’m saying. I’m saying that it might be that there are two possible worlds which differ in two respects: some horrible act of evil occurring in world A, not in world B; and something incredibly good, happening similarly in world A, not in world B. In world B, then, the evil thing didn’t happen; but depending on the good thing, it might still be worse in total than world A. Say, the evil thing is somebody kicking a defenseless dog in America, and the good thing is somebody saving the life of a pregnant woman who is going to give birth to the next Beethoven in China, at exactly the same instant, such that no causal influence can travel between the two. The two events are therefore not connected, but it still might be that both worlds are possible, and hence, that world B is worse off in total, even though the evil has no redeeming consequences whatsoever.
Again, he must merely argue that it’s possible that all instances of evil are necessary.
Well, even to show any single instance of evil gratuitous, you’d have to exhibit a world differing from ours in only that instance of evil, show it to be possible, and be able to predict its entire future. Failing to meet these requirements means that it is possible that the evil was in fact necessary.
I don’t get how this reasoning is circular. It makes an important point about the nature of reality and god.
Of course the premises entail the conclusion…that’s the point.
There is no shortage of people that believe in an omnimax god and that our universe is not the best possible. The POE points out the contradictory nature of that position.
I shouldn’t have used the word “argument”. I’m really thinking of scenarios.
For which, logical consistency is generally assumed.
If I’m up on a murder charge, saying that it has not been demonstrated that the evidence against me is logically consistent would not be an effective defence. Logical consistency is assumed, it’s up to me to show that there is an inconsistency.
This example may seem contrived but it is what we do frequently every day: we play out scenarios and logical consistency is the default assumption.
No, BPW is simply this: “Perhaps we’re in the best possible reality”.
This is vulnerable to the obvious retort that one can trivially think of realities that are better than ours and are, on the face of it, possible (the default position).
To this, BPW has to add the ad hoc defence: “Well, maybe they are impossible for some reason that we are not aware”.
It’s really no different from “God works in mysterious ways”; immune from formal logical arguments, but practically and interllectually useless.
For our universe to be the simplest line of goodness, it must be one helluva twisted topology.
And at the risk of taking the analogy too far, where has this complex geometry come from, and why is god limited by it?
I appreciate where you’re coming from on this, and I already conceded that the argument may work from this POV. Evilness does not automatically entail non-omnimaxness.
What you are not listening to, is that from a practical POV the problem remains. Because while evilness and omnimaxness may coexist, every instance of evil makes us less confident that an omnimax god may exist.
Why? Because it requires another infinite set of contradictions to explain away all the scenarios which would have been better.
We’re arguing past each other. I agree that the Leibnizian defense weakens the Logical Problem of Evil in the same way that the Free Will defense does. You are, however, trying to use the Leibnizian defense in the service of rebutting the Evidential Problem of Evil which I am alluding to. It simply doesn’t work. You’re saying that mere possibility is sufficient to resolve the Logical Problem of Evil. I agree. This is because, the proponent of the LPOE is arguing from incompatibility. The point I was trying to make is that by dropping claims of incompatibility and transforming the argument into an evidential problem of evil (which argues that theism is merely unlikely), the burden shifts to the theist to show that EVERY EVIL is necessary (I’m using the term loosely here…I take necessary to mean that the existence of the purported evil and a greater good is somehow connected in best-possible-world sort of way).
Again, it’s possible in the same way that it’s possible for Hitler to be really, really good. So what? Is it likely? (again, I’m not arguing for incompatilibity here, so arguing about the unlikelihood of the scenario you’re describing IS relevant)
‘Assuming god isn’t omnimax; then, god isn’t omnimax’ makes no point at all, it states merely a tautology.
Right, but if the premises entail the conclusion by assuming it, the argument is vacuous.
That at least I can agree with.
Ah, I see what you mean. Yes, in general, when I say ‘I played football yesterday’, nobody would doubt this on the grounds that it might be logically inconsistent – i.e. that there might not be a possible world in which I did, in fact, play football yesterday (though those who know me well would perhaps come close to such a thought ;)).
The key factor here, though, is that my saying ‘I played football yesterday’ and the assumption that this is logically impossible do not stand on equal grounds. The assumptions that a given world may be possible, or that it may not, however, do.
Which only works if it is true that we may be in the best possible reality, which is equivalent to every other reality being either worse or impossible – ‘every other reality may be either worse or impossible’ and ‘perhaps we live in the best possible reality’ say exactly the same thing, so if you grant me that I showed one, you can’t accuse me of having tacked on the other.
I still have problems with this ‘trivial’ imagining of other realities. I could claim I can easily imagine square circles, but have I done so, actually? Similarly, when you claim that you can imagine whole realities, what am I to make of this claim?
Even granting your criticisms, you yourself don’t seem to believe it entirely useless if you accept that the argument shows that the co-existence of evil and an omnimax god isn’t in principle impossible; I’d say that this is quite a feat regardless, actually.
It’s given by the constraints of possibility, which I suppose come from logic, essentially. As to where logic comes from – well, I have some thoughts on how logic may ‘assemble itself’, in a sense, but this is really beyond the scope of this discussion.
I don’t fail to appreciate this (or at least, I don’t think I do), but the implication from your posts is that it’s merely a quantitative difference, that given the improbability of god’s omnimaxness, one might as well regard it as impossible, and carry on regardless. This, I think, is wrong; if you’re willing to continue this line of thought at all, the ‘evidential problem of evil’, as Anduril dubbed it, should at least be recognized to be of a completely different metaphysical import as the original problem of evil.
I don’t see where you’re seeing this supposedly infinite set of contradictions. There is no ‘new’ contradiction necessary for every possible world, or else, you would have to burden every argument likewise – since it is possible that there are infinitely many refutations (say, to its premises, or to the logic applied), every single one of those must be assumed to be unsound by the maker of the argument, hence, the argument assumes infinitely many ad hoc hypotheses, hence, the argument is not a good one and can be disregarded.
But how do you establish likelihood in this setting? You have no grounds against which to gauge it – that’s where Leibniz’ argument is relevant for the evidential POE as well. There’s no way of telling which worlds are or aren’t possible, and hence, especially is there no way of telling which worlds are or aren’t likely.
One could equally well argue towards the other direction: I could claim to be able to imagine infinitely many more ways for the world to be more evil than for it to be better. Hence, if we assume that some evil is always present, that this world is best would become more likely with each new way for it to be more evil I could think up; in other words, the likelihoods we assign are exclusively functions of our imagination. How can we pretend those to make arguments about the nature of reality?
Well, put that way it does. Can you show me how POE does that?
Why do you suppose that?
When proposing future scenarios, or even hypotheticals, logical consistency is assumed, just like with the examples we’ve given. I don’t see why BPW should be treated differently.
Well left like that, there isn’t a problem.
However, since there is no reason to suppose all other alternatives are impossible, you have repeatedly made the suggestion that maybe there is some reason why all the others are impossible, of which we are unaware.
Well, maybe there is. And maybe god has mysterious reasons.
To believe it? In the case of the universe devoid of life, for example, such a universe can be defined using rigorous mathematics.
I don’t know what else can be done.
Quite. And this thread has revealed that to me at least. The way I’d put it, is that the POE makes a formal logical claim, and it should not. It should be written more empirically.
Not really, since as I’ve been arguing, the context in which we’d use the evidential POE and the “normal” POE are the same.
Indeed, I’d always understood POE to be making more of an empirical argument than a formal logical one. After all, I can think of other ways one can bite the bullet. e.g. What is evilness? Perhaps our notions of evilness are completely out of kilter with god’s.
Your response here doesn’t work.
When I write an argument, I make no reference to refutations. I don’t say “Assuming there are no refutations…”. They may well exist, and my argument may well be unsound. But the burden of proof is on someone claiming that there is a refutation.
OTOH, BPW, at least as you’ve been describing it, explicitly supposes that there are reasons that better realities are impossible. Hence it carries a burden of proof vs the evidential POE.
I totally agree that it would be hard to establish likelihood in the scenario we are discussing. However, I think it’s obvious that the likelihood that one of the evils we are seeing is gratuitous and avoidable is much higher than the likelihood that all of them are necessary in some way. So the fact that you can imagine infinitely many more ways for the world to be more evil than for it to be better doesn’t change anything. I need only one instance of evil to be gratuitous - the theist needs all instances of evil to be necessary in some way.
I’ve been away for a while, so I shall be short and sweet. I DO have the logical chops to assess the argument on its logical basis, and it is severly lacking.
To illustrate, here is an argument of similar form to the one in the OP:
- It may or may not be the case that Obama wears boxer shorts.
- I wear boxer shorts - just like Obama!
- Conclusion: I may or may not be Obama. Who might or might not wear boxer shorts.
As arguments go, it’s not one. There’s simply no “there” there - aside from fallacious rhetorical handwaving, and/or a gross misunderstanding of what the term “may” implies in a logical argument.
Here’e the thing about “X may be the case”: it must logically mean one of the following:
- “X is the case or X is not the case.” (‘may or may not’ sense.)
- “It is not the case that X is not the case.” (‘established as possible’ sense.)
2, we’ll note, is logically identical to “X is true”. Which is obviously not the same thing as 1; 1 is axiomatically true from a logical sense but basically deviod of information, and 2 is informative but needs to be either explicitly assumed or explicitly proven.
When “may” is introduced into the argument in question, it is clearly introduced in the ‘1’ sense, in statement 8. In the remainder of the argument, though, it’s used in sense 2. Sense 2 derives from the ‘permissive’ use of the english word “may”: “You may have the cake” is typically assumed to basically mean “yes”. This is the sense the argument is leaning on; it’s possible we’re in the best possible world, they say, as though that actually meant something. And in english it does, but in logic it doesn’t, and equivocating the two situations is where this wad of fallacy gets what little argumentive strength it claims to have.
The original POE doesn’t. The evidential one does: you put it as ‘assuming there is a better possible world, then god isn’t omnimax’. ‘Assuming there is a better possible world’ is precisely equivalent to ‘assuming god isn’t omnimax’, since only then is there such a better possible world.
Because the logical consistency of the proposed world is exactly the thing under debate, not whether or not I played football. It’s precisely Hume’s problem of induction: Is it possible that the sun will rise tomorrow? Well, yeah, sure, but it may not – we have no logical grounds to conclude that it will. Is it possible that some given world is consistent? Well, again, sure; but we have no logical grounds concluding that it is.
Well, yes, I’ve played some games with the hypotheticals, most of which were deliberately over the top, just to show that a possible reason can always be found. And some actions of an omnimax god may seem mysterious to us, just as some actions of a mother may seem mysterious to a child; the kid might think it’s evil that he doesn’t get any more sweets, but the mother is simply acting in his best interest, from a point of view the child doesn’t possess. Such mysteriousness isn’t troublesome. Only when this mysteriousness is offered as an answer to, for instance, an inconsistency in the teaching attributed to god, then it is insufficient.
That’s not really very easy, either. Sure, you could take a vacuum solution of Einstein’s equations to model a consistent, empty universe; but that’s only consistent if you don’t talk about its beginning, for instance. And as for such things like ‘the empty set’, I don’t even know what it would mean for that to exist.
Yes, that’s precisely what’s troubling me. It’s again like saying ‘as long as it’s really improbable for me to loose, it’s always wise to bet every amount’.
That’s exactly what I mean – everybody who brings up arguments like ‘perhaps such-and-such a world is possible and less evil’ is essentially saying ‘perhaps there is an argument that can refute yours’. And that may well be, but as you say – the burden of proof is on those bringing up the alleged refutation.
It doesn’t anymore than any other argument explicitly supposes that there are reasons every counterargument fails.
Those two sentences seem to assert contradictory stances…?
Well, no. He merely needs for it to be possible that all instances of evil are necessary. Really, there is a difference there, and an important one.
This is the same old confusion – this establishes x as necessary, not as possible; it’s not a meaning of may, it’s a meaning of must. ‘Not possibly not x’ means ‘necessarily x’. The argument arrives at ‘possibly x’ which is equivalent to ‘not necessarily not x’.
Exactly. How could it then be equivalent to ‘x may be the case’, or ‘x is possible’?
Well, I’m at best casually familiar with Plantinga’s argument, which seems more of a refinement of the Leibnizian one, essentially asserting that while it may not be impossible to create a world free of evil, it still may be impossible to create a world that both incorporates free choice and no evil. This has been lightly touched on in the thread. I’ve never read Lewis’ Problem of Pain, so I’m not sure what new elements he brings to the table; but if you think the discussion benefits by bringing those ideas into it, then sure, have at it.
Personally, I think every rebuttal that explicitly depends on free will has to wrestle with problems perhaps greater than those it intends to solve. As I see it, nobody has yet managed as much as defining ‘free will’ in a sensible way, much less showing that it exists at all, but that’s really a topic for another thread.
Also note that the logic of a free will defense is essentially equal to the Leibnizian possible-worlds approach, if one just shifts the interpretation of the modalities: ‘possible’ becomes ‘(morally) permissible’ and ‘necessary’ becomes ‘(morally) obligatory’. The problem of evil, ‘if god is omnimax, it is impossible for evil to exist’ then becomes ‘if god is omnimax, it is morally impermissible for him to allow evil to exist’, and, in arguing against this dichotomy, from ‘show that it is possible for an omnimax god to co-exist with a world in which evil exists’ we then get ‘show that it is morally permissible for an omnimax god to let evil exist’. We then need to find a morally sufficient reason for god’s tolerance of evil, so that he may be overall good despite the obvious existence of evil – in the same way a mother may be overall good despite wilfully letting harm come to her child – in the form of a painful vaccination, for instance. Instead of the best of all possible worlds, the freest of all possible worlds is then postulated as giving such a sufficient reason, with the distinction that this world may include evil due to its inhabitants’ freedom to do bad things.
This ‘shift of interpretation’ of the modalities, however, entails a shift towards deontic logic, which is notoriously paradox-laden and thus might not be as sure a footing as classical modal logic.
Sorry for the third post in a row, but I keep coming back to this. There’s one criticism of free will defenses that pops always up, as Voyager has noted, the problem of natural evil – how can free will justify suffering caused by natural forces? Now, typically, this is done away with a reference to original sin, that somehow the fall of man (this being brought on by free choice) is what caused us to live in a world where natural disasters exist. This one can either buy or not, but, even given this – how about natural evil with respect to animals? Typically, we think of causing (needless) harm to an animal as an instance of evil. So what’s the justification for letting animals die horribly due to disease, flood, fire, etc.? (Now that I think about it, did Voyager bring this up, too? I seem to remember something along those lines…) I haven’t seen this discussed anywhere; you wouldn’t happen to know whether or not this has received any attention?
Also, I think I might have to eat crow and retract – well, relativize 
– some of what I said re: the evidential POE. I think I misunderstood that it applies to a slightly different framework than the logical POE, thinking both to be intended to work on the same grounds. However, the logical POE works in a situation where a pre-established faith is challenged and purportedly shown inconsistent, through the alleged inconsistency between the existence of evil and the existence of an omnimax god. Under these circumstances, I don’t believe it works, in accordance with the arguments in this thread; and similarly, when applied in this framework, the evidential POE holds no sway, either, being essentially a weaker version of the logical one.
However, things are different if one looks at a situation where the playing field is level, so to speak – when the discussion is concerned with whether or not god exists, and both possibilities are on the table. In this situation, you could assign a likelihood to god’s existence (resp. his non-existence – let’s just for the sake of discussion assume both to be 0.5), and also, in this situation, evidence of the existence of evil influences this likelihood, i.e. makes the hypothesis that god doesn’t exist more reasonable. It is then an argument capable of influencing the fence-sitter, but not one to show the believer wrong, which I thought it was intended to be (and which the logical POE is intended to be). So, sorry for being somewhat obstinate, if this was the intended framework for the argument.