As always, you’re cordially invited to point out the flaw in the argument that 50+ years of debate have failed to find. But just saying so don’t make it so.
The argument doesn’t make your claims though.
Plantinga is at most saying that the existence of evil and an omnimax god are not necessarily incompatible.
This is not the same thing as your claims of (1) That we can take as a premise that there is unavoidable suffering in an omnimax god’s universe let alone (2) That it is irrational to doubt the existence of an omnimax God solely on the observation of apparently unnecessary suffering.
I haven’t looked at your lengthy response to me yet.
The flaw is that an omnipotent being by definition can do anything, therefore there’s no way anything can be “necessary” for it. An omnipotent wants something to be so, and it is.
And decades - millennia in fact - of attempts to evade the issue doesn’t change that. The flaws in all those arguments are blatant, they are just ignored by the people making them because otherwise they’d have to admit that their entire belief system is utter nonsense and this whole debate over Christian dogma is on the level of arguing “Can Superman tear his own head off?”
Just more pretentious, and with the difference that people will actually kill other people over it.
But that doesn’t change the fact that the whole tri-omni concept is full of paradoxes, that people ignore that it’s full of paradoxes, and that we very, very blatantly do not live in a world that would be created by anything approaching such an entity. No amount of sophistry, stubborness and evasion changes any of that.
Primarily, an author is omniscient and omnipotent over their story by virtue of being its causal source: nothing exists in the story except by the author’s knowledge and power. Aquinas and Boethius made similar arguments about God. From this perspective, temporal considerations—whether chapter 1 is fixed before chapter 2—cannot undermine omniscience or omnipotence.
Moreover, it is a mistake to assume that an author produces events in the order characters experience them. Nonlinear narratives aside, authors routinely draft and revise, and it is often unclear whether the first chapter was fixed first, last, or never truly fixed until publication. Even when reading a finished story, an author can alter it mid-chapter, mid-sentence, or even mid-word. This flexibility is even more apparent in the oral tradition: traditional storytelling was often episodic, with the overall plot of an episode predetermined, but with improvisation of details to suit the audience. Today, parents frequently tell bedtime stories that depart from the printed text, and countless authors have done the same. The adaptability of storytelling demonstrates that knowing and controlling a story does not conflict with the ability to modify it.
ETA: The key point is that the “knowledge” required for omniscience applies only to the being’s own domain. In the author analogy, the domain is the story; in theology, it is the universe. Asking whether God is a member of the set of all things leads to an infinite regress or set-theoretic problem. Strictly speaking, this can be avoided by treating God as extra-universal, outside the domain of the set he knows.
~Max
No, that’s still getting things the wrong way round. The argument shows that there are cases where both co-occur, and hence, they must necessarily be compatible.
This is not a claim I’ve made. The argument shows that the set G\cap E is non-empty, and that hence, there is a non-zero average suffering in worlds where a tri-omni God exists. This doesn’t entail that there is suffering in every world in which sich a God exists, and neither is that required.
This is a strict consequence of the FWD. Its conclusion shows that evil and God are consistent, and observing evidence for one proposition consistent with another can never be evidence against that other proposition. The upshot is that it’s possible to have evil in a world with a tri-omni God, so finding evil is perfectly consistent with that. It’s only upon making further assumptions, such as that the amount of evil shouldn’t surpass a certain threshold, that one can hope to find a conflict. Given the conclusion of the FWD, you thus know that there may be evil, but have exactly zero information about how much of it to expect (on average); so any claim that one has exceeded the expectable amount—which is what any claim that one has observed evil that lowers one’s confidence in the existence of a tri-omni God logically amounts to—is one making reference to further information.
Take your time; I’ll wait.
This is not true, of course. An omnipotent being can only do what is possible. The FWD then constructs a case where creating a better world is impossible.
That’s pretending that there isn’t a large secular movement in philosophy that argues against religious claims. But in fact, the reason Plantinga’s argument even exists is as a defense (it’s in the name) against such attacks, most directly the argument of evil as formulated by J. L. Mackie. If things were as you want to pretend they are, such attacks wouldn’t take place. And of course, the FWD has received much discussion from secular circles.
Are you even trying to understand my posts? The one of mine I quoted is the critical one, and you repeated stuff about omnipotence which is not relevant. The question is can the god do evil, not about him choosing not to.
Let me spell it out. We have a god who is claimed to be both omnipotent and omnibenevolent, and up to this point it always does good in some sense. No problem. Now, what happens if he does something evil (or which increases suffering)? There are two possibilities. If omnibenevolence is part of his nature, he will be unable to do the evil, since it violated omnibenevolence. Thus, he is not omnipotent, since another god without a claim of omnibenevolence can do this thing.
If omnibenevolence is descriptive, the god can do the evil act, and we can say he was omnibenevolent up to the point he did evil, but is no longer omnibenevolent.
Notice the word “choice” appears nowhere in this question. Saying the god just chooses not to do evil is not an answer.
Skipping the dissertation on novel writing. The novels I wrote had their major sections written out of order, since that made it easier to plant the clues needed for the final section. And of course anyone can modify a story. I read my daughter Nancy Drew books for years, and it was only until she decided to finish one herself when I was traveling that she found the housekeeper’s name was not Hannah Gruesome.
The omniscient / omnipotent paradox is not about our world but about god’s ability. And I can accept that God is outside the universe for the purposes of this argument. Inside or outside, an omniscient god can know everything that happens or is going to happen, just like an author with a really complete outline of the plot before writing. Inside or outside, God can make anything happen, just like a writer can. (Though a writer can fudge things to make logically impossible things happen, like someone picking up a four-sided triangle.)
A writer writing the last chapter first does not necessarily know what is going to be in the first chapter, so the problem is not temporal. The question is if the author has the entire book plotted perfectly, and then he changes something, was he correct in thinking he had the entire book plotted?
I’d say the writer is omnipotent but not omniscient, so no problem. After the book has been completed and published, assuming no changes in later editions, the writer is omniscient but not omnipotent, since he cannot change anything published.
Except maybe in Jasper Fforde’s book world in the Thursday Next series, which makes no sense but he does a great job hiding that from the reader.
No, this is absolutely not a question. The being in question starts out omnipotent, so they absolutely can do evil. If they then choose to only do good, they are also omnibenevolent. If this is possible, omnipotence and omnibenevolence are not in logical tension. Since moreover that clearly is logically possible, there’s no problem with a bi-omni being. Otherwise, you’d have to argue what happens to omnipotence if this scenario is followed:
Comparing god to a fiction writer is a category mistake. Authorial omniscience is historically contingent, fragmented, and unreliable while god’s omniscience is defined as perfect, complete, unchanging, and non-deceptive. Authors play with omniscience, they do not embody it.
I don’t see how this is different from saying that they are not necessarily incompatible. I prefer my description because yours is leading you to make claims that are baseless (e.g. that we should expect non-zero necessary suffering)
This is awful reasoning. The set of “I have a blue car” and “Clowns come from Venus” is non-empty – they are not inconsistent. Hence, given I have a blue car, my expectation should be a non-zero number of Venusian clowns?
We don’t reason from “consistency” in this way, otherwise we could never gain or lose confidence in any claim. My entire life history is consistent with me actually living on the moon and having a hallucination – this doesn’t mean that I don’t have reason to doubt that claim, and every observation of being on earth gives me more reason to doubt.
Every instance of apparently unnecessary and/or random suffering gives me reason to doubt that an all-powerful being tried to make a universe of zero suffering. But sure, God could push me down stairs every day and it would technically be consistent with him being omnibenevolent.
I think you meant omnibenevolent in the last bit.
This is where my thoughts went.
The maxomni god does not choose to be omnipotent - it just is. It does not choose to be omnipresent or omniscient - it just is.
So is omnibenevolence an essential element of what God is? Or is it just the consistent choice to avoid harm?
If it’s an essential part of the being, then is he capable of doing harm?
Ok. But an all-powerful being doesn’t do every possible thing - he chooses what actions he takes, what power he uses.
Similarly, an all-knowing god may know everything, but he chooses how he uses that information. He chooses what to pay attention to for any situation. Right?
So if this god is omnibenevolent by nature, does that just mean he has a predisposition to want to refrain from harm? It’s not a limitation, it’s a description of his desires?
It feels like you two are arguing about how you agree with each other, because you can’t agree on the words to use.
@Voyager agrees that the FWD provides room for there to be minimal evil.
@Half_Man_Half_Wit agrees that measuring the quantity of evil does allow one to question the omnibenevolence of a proposed god.
The argument is over the context of evaluating the data points of harm.
@Voyager says the quantity of harm seems excessive. The accumulation of experience allows one to doubt the premise.
@Half_Man_Half_Wit says the FWD doesn’t limit the amount of harm, so there has to be some additional critera or data to set the expected threshold allowed.
It seems like the accumulation of experience meets the requirement of additional criteria or data. Expectation is set by experience. Experience is not limited to just god behavior. Measuring human behavior gives context to compare against.
@Irishman Firstly note you’ve misattributed some of my words to @Voyager.
But secondly, when @Half_Man_Half_Wit first brought up Plantinga I immediately agreed that it showed that, technically speaking, the existence of evil and an omnimax god are not a logical impossibility. So it is not me that is dug in here and reluctant to find agreement.
All I had said was that I disagreed with the further claim that it is not logical to doubt an omnimax god on the basis of the experience of suffering. And I have found the “many worlds” argument for that completely flawed. The fact that it can be used to argue literally any claim and handwave any evidence is a clue that it’s completely bankrupt.
(And in the last few posts several other points have been made that I disagree with, but for this post I won’t go into those)
It’s the difference between ‘it will not necessarily rain tomorrow’ and ‘it will necessarily not rain tomorrow’: the first one allows for circumstances in which there is rain tomorrow, i.e. in which evil and God are incompatible, while the second excludes this. But it’s nonsense to think that propositions could be compatible or incompatible based on additional circumstances; if there exist circumstances in which they simultaneously hold, then they are compatible, period.
On average. That’s just a different way of saying the same thing: if the intersection of God and evil is non-empty, then there’s a certain probability with which you will experience evil despite God’s existence, and hence, a non-zero expected amount of evil. Like if there’s one in ten worlds that has x evil, then on average you should expect to find x/10 evil.
It’s not clear to me that this is the case. I’d expect the set of worlds in which clowns come from Venus to be empty, since it’s not possible for there to be life on Venus.
But also, again, this isn’t something I’ve made up just for your benefit—it’s bog-standard Bayesian reasoning. There is literally nothing controversial here.
This is just not true. If you have some beliefs on the basis of which you can assign expected probabilities to the truth of a hypothesis, and your observations given that hypothesis, then there’s a single rule according to which one can rationally update one’s belief in that hypothesis. You can of course do things differently, but then, you’re not acting rationally.
Only if you have some reason to believe that this suffering is more than you should expect, given the existence of a tri-omni God. However, you have no such information.
Go back to the illustration. If you observe evil, you know only one thing, and that’s that you are somewhere in the set E. Any further specification of your ‘location’ within the set of possible worlds necessarily needs further information. If you are saying that ‘excess’ evil gives you reason to believe that you are not in the overlap E \cap G, then you are saying that you have additional knowledge about your position within the set E. These are simply logically fully equivalent statements.
Perhaps this old explainer I’ve written on the topic of Bayesian reasoning can help you get a feeling for the type of argument being made:
The Demon And The Reverend: How Doubt Unites Us
The example of the cookie jars given there can be readily recast into an argument about which possible world one finds oneself in. Each cookie jar is a possible world; some have more chocolate cookies (i.e. evil, although I balk at this use of metaphor somewhat), some have less. The crucial point is that you need some estimation of how many chocolate cookies to expect in order to get started—in order to know how unusual it is to draw a chocolate cookie. But you don’t have this; you only know that there are at least some cookie jars with chocolate cookies in them. The possible worlds is just a different visualisation of this sort of argument, but this should be enough to show that just flat-out rejecting it is simply not a valid response.
Anyhow, I’ll give you the opportunity to reply to the other points I raised that you had deferred earlier.
Only if one has reason to believe that there is a certain threshold of evil that would not be exceeded if there were a tri-omni God, which would have to be argued for. But @Mijin claims that no such further argument is needed, that one can take every instance of evil as evidence against the proposition that there is a tri-omni God. This is a substantive disagreement.
Sure. If one has been to a million different homes, one has built up expectations about how good hosts behave, and can compare those against any new host. If one could analogously sample from the distribution of possible worlds, one could gain information about the amount of typical evil one should expect within a world in which a tri-omni God exists, and could compare the actual evil in this world against it. However, this hits some obvious difficulties. What one can do—and what usually is done, e.g. in Rowe’s argument—is to explore this space of possible worlds virtually, by making plausible assumptions and arguments regarding the expected amount of evil. But this is an absolutely necessary step before one can point to any given amount of evil as being excessive!
No, it just claims it does, but its claim makes no sense and directly contradicts reality. It uses a nonsensical concept to defend a claim that is blatantly false.
Causation and time are irrelevant for an omnipotent, anyway. They can make things happen without an in-universe cause, and don’t need to create a universe that’s consistent over time. Cause and effect only hold in a universe under an omnipotent so long as the omnipotent chooses for it to be so; an omnipotent could trivially make the universe of a microsecond ago entirely different than the universe of now, the universe of a second in the future nonexistent, and for yet another configuration of the universe to exist in the second after that.
Oops, I’m so sorry. I confused myself.
Didn’t respond to my point at all, did you? You are just asserting that a bi-omni being is possible, and basically defending that assertion by ruling out all cases which can cause conflict between those omnis.
That they are both possible is a hypothesis, which I think I have falsified many times over, and excluding the situation where it is falsified as you do does not solve the problem at all. Falsifying the hypothesis does not falsify the possibility of omnipotence, nor the possibility of omnibenevolence, but just the possibility that they exist in the same being.
Your continuing to use the concept of choice shows how bankrupt your argument is. All possible choices have to be considered, not just a useful subset.
Well, by choosing bad I meant increasing the amount of suffering beyond the minimum, but close enough. And your statement of my position later in the post is correct.
Ah yes, proof by vigorous assertion. Well if that’s not convincing!
I have responded to your point over and over, to no avail. In contrast, I have given an explicit construction that shows the consistency of omnipotence and omnibenevolence by showing how a being that is omnipotent can be omnibenevolent, thus showing the two concepts to be consistent. You have ignored this three times now, so you shouldn’t really get all in a huff about whether I properly reply to your arguments.
And again, I’ve detailed an explicit situation where omnipotence and omnibenevolence co-occur, so no, you haven’t falsified anything. You’re still just affirming a contradictory position where on the one hand, choosing one alternative over the other doesn’t take any ability away, and on the other hand, somehow always choosing good means you can’t so evil.
An omnipotent being chooses to always do good. It’s thus omnibenevolent. That’s it. No issue at all.
No, supporting a claim by pointing at the real world. Sneering and insults don’t change that, nor does a repeated failure to address the point. The world is, objectively, nowhere near the best possible world. Pretending otherwise because it means the tri-omni god cannot possibly be true won’t change that.
Except the argument doesn’t really entail any empirical claims. Rather, it is about what one can can conclude from a certain empirical datum (the existence of evil). So you’re already taking aim at the wrong target, which you then proceed to miss.
First of all, you can’t possibly know that. It could be, for instance, that this is the only possible world—the only self-consistent set of natural laws and initial conditions. Without additional assumptions, we can say very little about what worlds are possible, and what we can say then depends on the plausibility of the assumptions we make. The idea that we could just without further ado point at this world and claim anything about where it is located in the set of possible worlds is just nonsense.
But more importantly, and it’s really a mystery to me how this isn’t clear at this point, the FWD doesn’t depend on this being the best possible world. Even if you could definitely prove that this isn’t the best possible world, this wouldn’t impact the logic of the argument! All it shows is that it is possible that there is non-zero evil in the best possible world, hence, that God and evil are not incompatible, and that observing evil this isn’t grounds to conclude the non-existence of God.
There are three possible reactions when faced with an argument opposing one’s predilections. One can grudgingly accept it; one can seek to fight it’s logic, by showing the argumentation fallacious or casting doubt on its premises; or one can just blithely ignore it. The first two are rational reactions; you chose the third. The thing you’re pointing to is not reality, but merely your own unreflected dogmatism.