Formel debat på FRDB: The logical problem of evil

Jeg deltager i en formel debat på FRDB. URL:

Her er min modstanders første indlæg og mit svar:

Wiploc’s first post

The Logical Problem of Evil: If god exists, why is there evil? Doesn’t the existence of evil prove that god does not exist? The answer is yes. If evil exists, the tri-omni god does not exist.

Tri-Omni: A tri-omni god is omnipotent (all powerful), omniscient (all knowing), and omnibenevolent (all good-wishing).

Omnipotence: A truly omnipotent god could do anything, including contralogical things like making square circles. But you can’t draw logical conclusions about what would happen if logic didn’t work. So this discussion is about a punk-omnipotent god: one who can do anything except violate logic. And, from here on, that’s what I shall mean by “omnipotent.”

The absence of evil is not a square circle. An omnipotent god could prevent evil if he wanted to.

Omnibenevolence: An omnibenevolent god would choose to eliminate evil if he could. If there were a god both omnipotent and omnibenevolent, he both could and would eliminate evil. There would be no evil.

Therefore, if evil exists, it follows that the omnipotent omnibenevolent god does not exist.

Omniscience: We throw in omniscience so nobody can say, “Well, what if god was that powerful, and was that benevolent, but was too stupid or ignorant to realize that evil exists? (What, for instance, if Jehovah assumed circumcision would cure all our problems?)

An omniscient god knows everything. He knows the future; knows counter-factuals (what if Hitler had died as a child?); knew at the time of creation (per Plantinga) every choice that would ever be made in every possible and impossible world. (Which means that, if he chose to create this world, he did so knowing that it would contain evil.)

Tri-omni gods, then, are omniscient, omnibenevolent, andomnipotent: they know all about evil, want to prevent evil, and are able to do so. If such a god existed, there would be no evil. If evil exists, there is no such god. Evil and the tri-omni god cannot logically coexist.

Benevolence is the desire for good, the desire that there not be evil. In the case of omnibenevolence, this desire is total, infinite, unalloyed, unconflicted, on the front burner.

Good and evil: I hardly care how these are defined. Given the definitions already established above, it is clear that the tri-omni god cannot exist if evil exists. But, nonetheless, here’s my understanding: good is anything that causes happiness; evil is anything that causes unhappiness. Stubbing your toe is both good and evil if it makes one person happy and another person unhappy.

Evil is distinguishable from sin. Sin is doubting or disobeying god. In the Adam and Eve story, evil is the punishment for sin. Sin that causes unhappiness is moral evil.

Those are just the definitions that I know. I’m not wedded to them. If Deleet, says, “What if god wanted to give people free will more than he wanted people to be happy?” that will mean that, by definition, god is not omnibenevolent. But, if he wants, we can change the definition. We can say that free will and happiness are both good. Then, god can have both of those desires and still be omnibenevolent. Of course, if those desires conflict, then god isn’t omnipotent—an omnipotent god could have both free will and happiness.

It hardly matters how we define good and evil, so long as they are opposites. If good were blue, and evil is other colors, a benevolent god would want things to be blue. An omnibenevolent god would want everything to be blue. The existence of other colors would prove that one or more of the following statements is true:

1. God isn’t omnibenevolent (he doesn’t want everything to be blue).

2. God isn’t omnipotent (he isn’t able to make everything blue).

3. God isn’t omniscient (he isn’t smart/knowing/wise enough to exploit his omnipotence so as to achieve omni-blueness).

So the LPoE (logical problem of evil) is bulletproof almost regardless of how good and evil are defined.

Five relevant responses to the LPoE: There are many irrelevant responses to the LPoE (logical problem of evil), like, “Atheists suck!” or like, “If god doesn’t exist, then morality doesn’t exist either.” But there are only five relevant responses:

1. God isn’t really omnipotent.

2. God isn’t really omniscient.

3. God isn’t really omnibenevolent.

4. Evil doesn’t really exist.

5. Belief in tri-omni gods is contra-logical.

Each of these “defenses” amounts to a concession that the LPoE is correct. The art, then, of “defending” against the LPoE consists entirely of not realizing what you have conceded. Or at least of not letting your audience realize it.

Thus, Plantinga says god is omnibenevolent, but follows that with the deadpan question, “Why would you believe that a good god would be good to you?” (Yes, I’m paraphrasing.) Thus, he claims god is “omnibenevolent,” but then he un-defines “benevolence” so his claim becomes meaningless. Thus, he concedes #3 without recognizing the concession.

All “defenses” against the LPoE are based on this sort of equivocation.

Conclusion: If a god were omnipotent, he could eliminate evil. If a god were omnibenevolent, he would want to eliminate evil. If he were omniscient, he would be smart enough to eliminate evil. If he were all three, there would be no evil.

Therefore: if evil exists, there is no tri-omni god.

Deleet’s first post


I would like to thank the people who make this debate possible and most of all my opponent, Wiploc, for debating me. I have more objections than the ones in this post but the limited amount of space restricts me from posting them until later. My position is that the LPoE is unsound and the EPoE is sound.

Problems, evil problems!

I would like to distinguish between the two different kinds of evil problems. The first one, which is the one this debate is about, is called the logical one. This is because that the problem, which is actually an argument, states that evil and a tri-onni god are impossible together; they cannot both exist at the same time. The logical problem of evil (LPoE) is committed to claiming that the set of propositions hereunder are inconsistent:

Set 1

1. God is omnipotent

2. God is omnibenevolent

3. God is omniscient.

4. Evil exists.

5. God exists.

The argument then goes that (1) through (4) are true and therefore (5) is false. In other words the LPoE is committed to this implication:

A. If (God is omnipotent and god is omnibenevolent and god is omniscient and god exists) then evil is nonexistent. [(1∧2∧3∧4)→ ¬ 5]

The evidential problem of evil is not committed to (A) but only to making the claim that (roughly) (1) through (4) implies that (5) is very probably false.

The question of this debate is then: is (A) true? Before I shall attempt to answer that however I will discuss the three attributes attributed to God.


This is, I think, the least controversial attribute. By that I mean that it is the least controversial attribute in terms of what it means. Wiploc thinks that omniscience means that one “knows everything”. I prima facie agree with this. Wiploc though believes that this implies that one knows what choices people will make before they make them. This may be impossible so I’ll just reword omniscience to: one knows everything that is possible to know. I hope that Wiploc will accept this.


Wiploc distinguishes between true omnipotent and punk omnipotence. The first includes the ability to “defy logic” and the second does not. The second is characterized (roughly) like this: one can do everything that is logically possible.

I used to be in favor of a such characterization of punk omnipotence but I have changed my mind. There is nothing logically impossible about traveling north of the north pole yet God presumably cannot do this. This is because logical possibility is not the only condition an action have to meet in order to be realizable. There is also what I call environmental possibility; being in position to do so.[i][ii]

Otherwise I accept Wiploc’s use of punk omnipotence instead of true omnipotence.


This one, I think, is the most controversial attribute. This is because it is not clear what a being who is omnibenevolent wills. The prefix ‘omni-‘ simply means ‘all-‘. So we have to look up ‘benevolence’:

1. Desire to do good to others; goodwill; charitableness: to be filled with benevolence toward one’s fellow creatures.

2. An act of kindness; a charitable gift.

3. An inclination to perform kind, charitable acts.

4. A kindly act.

5. A gift given out of generosity.

6. Disposition to do good.

7. An inclination to do kind or charitable acts.

8. An act intending or showing kindness and good will.

Etc. [iii]

As we can see per the above it is not clear that ‘benevolence’ is the desire to remove evil. Perhaps something else is meant by an omnibenevolent person. Note that ‘do good’ can maybe be interpreted as ‘remove evil’ but it’s prima facie not very clear and needs a further and maybe implausible analysis.

Wiploc thinks that benevolence implies the will to remove evil, but this is not what the dictionary says. I’ll go with the dictionary until Wiploc can argue otherwise. Without ‘benevolence’ implying the will to remove evil, then (A) is false and thus the LPoE unsound.

A possible morally sufficient reason to not eliminate evil

The strongest objection to the LPoE (but not the EPoE) is to claim that there is a possible morally sufficient reason as to why God has not removed evil. If there is such a reason, then even an omnibenevolent being such as God does not want to remove evil if he could:

B. There is a possible morally sufficient reason for god not to remove evil.

If (B) is true, then (A) is false. So the LPoE depends on the falseness of (B) to work viz. if (B) is true, then (A) is false and the LPoE is unsound.

So the question now is: is (B) true?


How do we determine whether something is possible? First off we must ask: what kind of possibility are we talking about? We’re talking about logical possibility viz. the non-contradictory nature of something. The most widespread answer to the question above is conceivability. It goes roughly like this: If I can conceive of p being true without discovering a contradiction, then p is logically possible.

However I don’t think this is a good method because it gives wrong answers; if we later discover that something is logically impossible, then it was not possible before even though we could conceive of it without contradiction. Second I don’t think there is any reason to adopt this method.

I propose that we simply assume logical possibility instead of “making up an excuse” to believe that p is possible.

Given either of the “methods” above then (B) is true. This is because that I can conceive of there being a morally sufficient reason and so can others.[iv]

And so if (B) is true, then (A) is false and the LPoE is unsound.

i Such a characterization is dicussed in Omnipotence, Stanford Encyclopedia of Philosophy. URL:

ii See my earlier paper on the subject of contradictory divine attributes The incompatibility of omnipotence and omniscience. URL:

iii I found these at in a search for ‘benevolence’. I have omitted some irrelevant entries. URL:

iv See for instance the ones in this article. Also see Platinga’s book mentioned in the article. URL:

This Post Has One Comment

  1. A Theist

    Your friend’s argument seems to be:

    1. Necessarily, if it is both the case that an omni-God exists and evil is not a necessary existent, then evil would not exist in the actual world.

    2. Evil exists in the actual world.

    3. Thus, it is not both that case that an omni-God exists and evil is not a necessary existent.

    4. Thus, if it is possible that evil does not exist, then an omni-God does not exist.

    5. Possibly, evil does not exist.

    6. Thus, omni-God does not exist.

    But your friend has not given any reason to think that his understanding of omnibenevolence, which aims to give some of the grounds for the truth of premise (1), is true. Your friend’s argument seems to operate on the assumption that an omni-God would have actualized a possible world without evil over the actual world. But a possible world without evil is consistent with a possible world wherein nothing exists save God. If such a world is possible, then your friend would be committed to the position that God would have actualized that possible world over the actual world. That’s tantamount to saying that a world without existents other than God (and thus a world without human existence) is better than the actual world. Thus, if God had to actualize either the actual world or a world with just Him as an existent, then He’d have to actualize the world wherein He is the only existent. But that does not seem tenable.

    Your friend should also hesitate to state that evil and goodness can be defined however we wish. Theists might understand evil through the divine command theory and thus premise (2) implies an omni-God’s existence. So from premise (4) and (5) it would follow that an omni-God does not exist, and from premise (2) it would follow that an omni-God exists, and hence a contradiction is deduced. Contradictions entail that one or more of the premises are false, and unless your friend wants to argue that all of the above premises are essential to omni-theism, then his argument fails because one of his premises must be false.

Leave a Reply