The definition of vegetarianism at best entails an intention of not buying/handling meat, and even that only in a ceteris paribus way absent defeaters like a household member on a paleo diet or a carnivorous pet. It entails, on its own, nothing about the ability to make that intention manifest, and thus, nothing about what observations are to be expected. For this, background knowledge about the actual world, such as that people are typically in control of what they buy and do, is necessary. Only upon asserting some thesis like, ‘vegetarians are typically able to avoid buying meat’ or some such does it become surprising if a vegetarian does, in fact, buy meat.
Likewise with criminal cases: only once you know what observations are to be expected can you tell whether an observation, such as fingerprints at the scene or the murder weapon, are surprising given a hypothesis of innocence, and thus, is in tension with that hypothesis.
Can you rephrase this to be less… ambiguous? I’m not asking you to prove tri-omni God would not allow gratuitous suffering. I’m asking you to admit it.
The logical problem of evil asserts it is logically impossible for tri-omni God allow any suffering:
No tri-omni Gods are beings who would allow suffering.
It is resolved by first proving the counter-proposition:
No tri-omni Gods are beings who would allow gratuitous suffering. (residuum)
All tri-omni Gods are beings who would allow necessary suffering.
All beings who would allow necessary suffering are beings who would allow suffering.
All tri-omni Gods are beings who would allow necessary suffering. (2)
All beings who would allow necessary suffering are beings who would allow suffering. (3)
Therefore, all tri-omni Gods are beings who would allow suffering.
These are categorical propositions and do not imply any category is populated; “would” does not imply actual action; you can “apply” a proposition with modus ponens, &etc. From here, theodicies can work without implying a logical contradiction between observed suffering and the nature of tri-omni God. This strong logical claim was for the most part abandoned after Plantinga’s free will defense in the 1970s. And you agreed the logical claim is defeated by the mere possibility that no suffering is gratuitous. With that in mind…
Max Presents: The Evidentiary Problem of Evil
Assumptions
No tri-omni Gods are beings who would allow gratuitous suffering. (arguendo)
All beings who would allow persistent, seemingly gratuitous suffering yet refuse to provide an intelligible explanation are beings who would allow gratuitous suffering. (inductive)
All beings with omnipotence over the observed world are beings who would allow persistent, seemingly gratuitous suffering yet refuse to provide an intelligible explanation. (arguendo)
Argument
All beings who would allow persistent, seemingly gratuitous suffering yet refuse to provide an intelligible explanation are beings who would allow gratuitous suffering. (inductive) (2)
All beings with omnipotence over the observed world are beings who would allow persistent, seemingly gratuitous suffering yet refuse to provide an intelligible explanation. (3)
Therefore, all beings with omnipotence over the observed world are beings who would allow gratuitous suffering. (inductively supported)
No tri-omni Gods are beings who would allow gratuitous suffering. (1)
All beings with omnipotence over the observed world are beings who would allow gratuitous suffering. (inductively supported) (c)
Therefore, no tri-omni Gods are beings with omnipotence over the observed world. (inductively supported)
For both (2) and (3), the qualifiers “persistent” and “seemingly” stand in for evidence, refusal to provide an explanation is something of a moral claim, and the intelligibility of an explanation is thrown in for good measure. The particulars are not important; the common term is intentionally burdened with controversial assumptions that I think you would agree to. It can be reformed as needed, but it can’t be omitted or made airtight.
The evidence would come in to support the inductive proposition, (2), which is of course the weak point. This is what skeptical theism attacks, specifically by emphasizing the possibility that no observed suffering is gratuitous. Full theodicies exploit this vulnerability further by relying on religious grounds that are generally incompatible with deism (Hume) or naturalism (Rowe). But for those who agree with the religious arguments, the evidentiary problem of evil is solved, and arguably the problem of evil is primarily aimed at those traditional theists to begin with.
Your argument is not the standard evidentiary problem of evil. As best as I can tell, it goes something like this:
Max Attempts: Mijin's Evidentiary Problem of Evil
Assumptions
No tri-omni Gods are beings whose acts tend to reduce our confidence in their benevolence. (inductive)
All beings with omnipotence over the observed world are beings who tend to allow lots of suffering. (arguendo)
All beings who tend to allow lots of suffering are beings whose acts tend to reduce our confidence in their benevolence. (inductive)
Argument
All beings with omnipotence over the observed world are beings who tend to allow lots of suffering. (2)
All beings who tend to allow lots of suffering are beings whose acts tend to reduce our confidence in their benevolence. (inductive) (3)
Therefore, all beings with omnipotence over the observed world are beings whose acts tend to reduce our confidence in their benevolence. (inductively supported)
No tri-omni Gods are beings whose acts tend to reduce our confidence in their benevolence. (1)
All beings with omnipotence over the observed world are beings whose acts tend to reduce our confidence in their benevolence. (inductively supported) (c)
Therefore, no tri-omni Gods are beings with omnipotence over the observed world. (inductively supported)
Note these conclusions are supported by inductive/probabilistic logic, not ontologically certain deductions.
This argument has two weak points in its two inductive propositions, (1) and especially (3). The first proposition, (1), is vulnerable to skeptical theism (as in my formulation). The third proposition, (3), is very problematic. Surgeons and mothers prove that benevolence is measured by instrumentality, not an absence of suffering, because we can understand the reasons. In the words of Bayes, we use priors to give weight to mitigating factors when we evaluate the evidence. To call this distinction ‘special pleading’ is to mistake a hospital for a torture chamber simply because both contain scalpels.
There’s nothing ambiguous about it and nothing “missing”. The definition of omnibenevolence – of being maximally good and loving – entails not causing needless suffering.
No-one would be questioning this if I were talking about doubting that a serial killer who tortures and rapes his victims before killing them was omnibenevolent.
As I say, it’s special pleading for god as it always is.
BTW please mention “emphasis added” when adding emphasis to quotes.
Wrong immediately. I wouldn’t make that claim. This is sneaking in the logical problem of evil again.
Under the evidential problem, I simply have good grounds to doubt. It remains possible, hypothetically, that an omnimax god exists. The same as it’s hypothetically possible that Freddy Krueger is perfectly loving, though I have good grounds to doubt it.
So (1) doesn’t follow.
Exactly. So only instances of needless or gratuitous suffering yield evidence against tri-omni God. Consequently, pointing at evil, or suffering, in general does not constitute such evidence. Rather, it has to be argued that some instance of suffering is actually needless. Which is all anybody ever said.
No it isn’t. It’s all you’ve said, but I was replying to @Max_S, who has now repeatedly questioned the claim that omnibenevolence is contrary to gratuitous suffering. Which I don’t even consider a claim here, it’s just the generally-agreed definition.
Fine by me, but just to be clear: you’re now agreeing that only needless suffering is in conflict with a tri-omni God, and no longer hold that any suffering is, yes?
By “fine by me” are you admitting to getting the wrong end of the stick there about who said what? Are you actually capable of admitting to being wrong about anything?
Now, in terms of the point you’re putting to me: 700 posts ago I agreed with the position that the logical problem of evil is solved by postulating that God might have mysterious ways or whatever such that it’s actually necessary suffering and the best possible world. So, no, it’s not in conflict logically.
However, my point remains that this doesn’t defeat the evidential problem: that apparently needless suffering gives us grounds for doubting that god is omnibenevolent.
It’s exactly the same as saying that seeing Freddy Krueger go into people dreams and savagely kill them, gives me grounds for doubting he is omnibenevolent. Even though it doesn’t prove it in an absolute sense, because as far as we know it could be essential suffering.
I didn’t understand that @Max_S means for omnibenevolence to include the possibility of admitting gratuitous suffering, but if he does, then I’ve misunderstood him. That’s no skin of my back.
(My bolding)
This is, however, not how you’ve framed your version of the evidential problem in the past, where you were adamant that any suffering at all constituted evidence against the existence of tri-omni God. The above statement is contrary to this: because then, there is suffering (or evil) that is not grounds for doubting omnibenevolent God, which, again, is the sum total of everything I’ve ever argued for. On this formulation, for evil to count against God, you have to give some reason why it seems gratuitous; you can’t just point at any evil act and claim it yields evidence against God.
And I remain so. Because all suffering is apparently unnecessary until we learn otherwise.
Suggesting otherwise means positing additional factors, entities etc.
Or, to be strict, the default position should be we don’t know if mitigating factors exist. But, we aren’t just classifying a singular instance of suffering. We’re looking at a vast amount of suffering, of many different flavors. Given that reality, the position that all instances have a (as yet unknown) purpose should be taken as less and less plausible.
Meanwhile the evidential problem of evil does not need to make the claim that no instance of suffering has a purpose.
I sincerely apologize for omitting the “emphasis added.”
I agree that the “definition of omnibenevolence … entails not causing needless suffering” (where “needless” is relative to the knowledge of the omnibenevolent actor). I do not ask you to prove that statement. I ask you to admit that statement is true because I use it as the first assumption for my evidentiary problem of evil:
No tri-omni Gods are beings who would allow gratuitous suffering. (arguendo)
You offered no comment on my take of the evidentiary problem of evil, so I don’t know your opinion of it.
It appears that you disagree with my understanding of your position. That is your right, but would you explain how you “simply have good grounds to doubt” after rejecting the two arguments I presented, above? How do you reach the conclusion that no tri-omni Gods are beings with omnipotence over the observed world? What do you have to assume to get there? Are those assumptions reasonable in the context of the evidentiary problem of evil? I am clearly struggling to pin down your position to a comprehensible form.
Not at all. As you now accept, all suffering comes in two classes: necessary and gratuitous. Only the latter kind yields valid evidence against tri-omni God. Beyond that, we have no knowledge of when a particular instance of suffering should fall in either class. If you thus just point to a particular instance of suffering without giving any reason to think that it is gratuitous, you haven’t given any reason to consider it valid evidence.
Again this isn’t any different from evidence in more mundane cases. Fingerprints on the scene are only evidence against the innocence of the person whose fingerprints they are if we shouldn’t expect them to be present. If it’s just fingerprints of a random person, we have of course ample background knowledge to judge it unlikely that they should’ve been at the scene, since the vast majority of people will never have been in any given room; but if that background knowledge changes, or fails to be apply, then so does the relevance of the evidence. But it is only in light of that background knowledge that the relevance of the evidence is decided. Pointing to observations and claiming them to be evidence without specifying that background is then to implicitly posit that some such background exists—which however nobody has any reason to grant you without further argument.
The question isn’t about any ‘mitigating factors’. The point is simply that we don’t know without further information or argument whether any given instance of suffering is necessary. Consequently, if you want to paint any such instance as unnecessary, you have to make a case for that. Again, that’s how the evidential PoE always proceeds: by arguing that some instance of evil is probably gratuitous.
That simply doesn’t follow without further assumptions. More fingerprints on the scene by the butler don’t make them any more likely to be evidence of guilt.
But it does need to make the claim that some particular instance of suffering (probably) is gratuitous. Otherwise you simply haven’t given any reason to consider it evidence against the possibility of a tri-omni God.
In fact, if what you’re saying here were true, then there would be cases where the rational evaluation of the available evidence would systematically lead one away from the truth, which would obviously be disastrous for evidential reasoning. For consider the case where we actually live in the best possible world, and God actually exists: according to your position, we would still be justified in considering any evil to lower our credence in the existence of God. Consequently, upon gathering more and more evidence, our beliefs would rationally diverge further and further away from the truth. Thus, evidential reasoning would fail to update our beliefs in the direction of the true hypothesis. In other words, if that were how empirical reasoning worked in this case, then we ought not trust it, meaning the evidential PoE pulls the epistemic rug out from under itself.
Luckily, of course, this is just due to the erroneous assumption that all evil is evidence against God’s existence. Correcting this assumption, evidential reasoning will lead us towards ground truth again. In fact, in a sufficiently formalized context, this is provable: Schwartz’s theorem tells us that the posterior probability concentrates on the ground truth with increasing number of observations.
Ah I see. I thought you were suggesting that this was part of my argument, and it isn’t. Anyway, glad we’re on the same page then as to definition of omnibenevolent.
Well multiple arguments were presented. But I’ll take a look at that one:
In terms of these premises, I agree with (1) as just mentioned.
(2) immediately seems unnecessarily weaseling and equivocal. Where did the need for persistence, or an “intelligible explanation” come from? And no, the set of suffering may be greater than the set of gratuitous suffering. So I would disagree that a being allowing one necessarily means allowing the other.
(3) Is ok, as long as you add omniscience. An omnipotent being hypothetically might not know that I am suffering.
The middle term–beings who would allow persistent, seemingly gratuitous suffering yet refuse to provide an intelligible explanation–is an attempt to anticipate skeptical theism. The combination of seemingly gratuitous suffering and refusal to provide an intelligible explanation mean we have no known mitigating factors. Persistent means this is a pattern of behavior, not a one-off incident. If we merely assert “all beings with omnipotence over the observed worlds are beings who would allow seemingly gratuitous suffering,” that leaves a much stronger objection to the inductive leap that all such beings likely allow gratuitous suffering.
You note that the “set of suffering may be greater than the set of gratuitous suffering.” That is true but it does not bear against the assertions. Assertion (2) is a generalization. It comes from an inductive argument (and is marked as such): if it looks like a duck, swims like a duck, and quacks like a duck, then it probably is a duck; here, if a being would allow persistent, seemingly gratuitous suffering without an intelligible explanation, the assertion says it is reasonable to assume the being would allow gratuitous suffering. This means the proposition is not one of certainty, nor are any conclusions drawn from it, but that’s to be expected given the nature of the main argument.
I think omnipotence implies omniscience, but I’ll admit (3) could be reformed to add omniscience.
I don’t think it’s productive to try to “head off” particular arguments when outlining the the problem of evil. It reads a lot like strawmanning the ePOE, though I don’t think you intended to.
It’s really not that complicated. Seeing apparently unnecessary suffering gives us reason to think there’s unnecessary suffering.
Freddy Krueger committing violent sadistic acts apparently for the fun of it gives us reason to suspect he isn’t omnibenevolent. It doesn’t prove anything – few empirical observations do – but it puts weight on the side of the scales that is “Freddy Krueger is not perfectly loving” based only on what we currently know.
Agreed? This isn’t the end of the argument, but I want to see how you answer this first.
I’m not personally familiar with Freddy Krueger, but I think I can agree with that. However, as a categorical proposition, it is unreasonable to say people who allow apparently unnecessary suffering are (probably) people who allow actually unnecessary suffering.
I seem to have lost the grace of your reply again, but I think this represents genuine progress. You’ve retreated from the claim that ‘all evil is grounds to doubt (tri-omni) God’s existence’ to one that ‘only some kinds of evil are grounds to doubt (tri-omni) God’s existence, but we can pretend like all evil is of that kind’. I think it’s much easier to see why that can’t be right, either, so maybe this won’t take another 700 posts.
First, note that if you were to posit this kind of evidential PoE to a theist, they can just outright reject it: since it will necessarily yield the wrong conclusion in the case of God’s existence (as all evidence can only lead to doubting God’s existence, regardless of the actual truth of the matter), it’s just not a reliable guide to truth, even probabilistically. So the argument doesn’t have any persuasive force.
Second, it’s also completely superfluous: if you think this kind of reasoning is sound, then you can conclude God’s nonexistence even in the absence of any evidence of evil, since you know both to expect evil, and that this evil could only adjust your credence towards God’s nonexistence, so you can go ahead and do so without even having witnessed any evil. This sort of ‘evidential’ argument becomes independent of any evidence, because any evidence at all only supports its intended conclusion.
Third, the reason you give that we can treat all evil as ‘apparently unnecessary’ is unsound. You seem to (somewhat misguidedly, but then almost nobody gets this right) appeal to a kind of Occam’s razor-like claim that we shouldn’t posit grounds to make some evil necessary in the absence of knowledge that such grounds exist. But of course, we do already know that such grounds exist, because we already know that necessary evil exists! So any evil we witness may, in principle, fall into either the categories of necessary or unnecessary, and if we point to the former as grounds to doubt (tri-omni) God’s existence, we’re just wrong.
So the only way to make an evidential argument that is persuasive and nontrivial is to give a reason why some evil is probably unnecessary—which (just to give you a chance to ignore this point one more time! )—is why every actual formulation of the evidential PoE starts out by doing exactly that.
We had used necessary/gratuitous as descriptors of cosmic instrumentality. Mijin has not admitted necessary evil/suffering actually exists, only that it may exist. Nor has he admitted to “expecting” any evil/suffering given a tri-omni God. It could very well be that Mijin includes all observed suffering when he says, “apparently unnecessary suffering.”
What is possible ought to be expected: if it’s possible it will rain, it’s reasonable to bring an umbrella. (At least, it’s unreasonable to be surprised upon the possible occurring.) Besides, if there’s evil possible in a world with a tri-omni God—if there is a world W such that there is evil E and that world meets God’s criteria to instantiate it C (such as allowing for nontrivial free will)—then there’s necessary evil in a world with a tri-omni God, as there could not be a world W’ with evil E’ < E that similarly meets criteria C, otherwise, W would not have been possible for God to instantiate since allowing evil greater than E’ would clearly be unnecessary. (I don’t know @Mijin’s stance with respect to these points, since so far, he’s declined to address them, but that doesn’t mean I have to drop them.)
Besides (and for a fourth reason @Mijin’s stance seems rather unreasonable), there is also lots of apparently necessary suffering in the world. Take a baby crying upon receiving a vaccine, or just any sort of pain that leads one to avoid greater harm, and so on. Sure, one might claim that there are other ways for God to achieve the same ends, but that would at least need an argument to be convincing: it’s not obviously true. To appeal to @Mijin’s favorite hypothetical in reverse, one could imagine a world in which all suffering were exclusively of that sort, and yet, it would still be reasonable to conclude God (probably) doesn’t exist, since one can just take all suffering to be unnecessary without giving a further reason.
An evidential argument that can only ever reach a single conclusion simply isn’t an evidential argument.
Ah but the petard I had prepared is that we don’t know that Freddy Krueger isn’t god. And apparently we are not to draw this inference about gods, so how does this work exactly?
It depends what you mean by categorical proposition here. I am not making an absolute claim. No-one is saying that observing X proves that X exists. Only that observing X gives us reason to suspect that X exists.
Yes this is correct. On the last point, yes all suffering is apparently unnecessary until we know otherwise. (and of course I mean this in the cosmic sense, very obviously physical and emotional suffering evolved to serve a function)
)—is why every actual formulation of the evidential PoE starts out by doing exactly that.