Thursday, March 10, 2011

Wagering for one's soul and the possibility of Divine Evil

In the previous post, I mentioned that the “God” of the Bible is at least a possible, if not actual being that is very evil. If the description of him is empty, he is a possible being of some possible world (closest to ours with relevant changes) that fits the description. If he is an actual being and “God” of the Bible denotes him, the description may or may not be accurate. The Bible may be wrong about his attributes. He may not be the evil being he is described. In which case it may be that people have mistaken his attributes or it may be that he is actually an omnibenevolent being that has constructed the Bible as a kind of moral test for us. If we resist the Biblical version of him which is on many philosophical accounts is a description of an evil being, we “pass” and if we fall for it by worshiping the Biblical description, we “fail.” What is the consequence of these possibilities?

Consider these four possibilities. Either there is at least one evil being that fits the Biblical bill (henceforth, I will refer to this possible being as "Dog") or there is not. Additionally, either there is at least one omnibenevolent and omnipotent being or there is not (I will call this being just "God" not to be confused, of course, with the Dog which is the Biblical God). There are four total possibilities and they are mutually exclusive: there are at least one Dog that fits the Biblical bill but no God. That both exists. That God exists but Dog does not and that neither exists. In what follows, I will simply ignore the possibility of more than one of each kind of being (it will not affect the argument).

Many philosophers since Augustine have made maltheistic arguments against the Biblical (Abrahamic) God. If we are to interpret the Bible literally, we find that it's inescapable the conclusion, if we are rational, that the being called 'God' there is actually something monstrously evil. His actions in the Bible of causing/inciting/endorsing/ordering genocides, mass murder, "natural" catastrophes, child and wife killings, etc are well known. Even some of our Founding Fathers have remarked how despicable and evil the Abrahamic god is. Jefferson once said (paraphrasing) that a unbiased and reasonable man would have to conclude that the Biblical god is evil or a demon after reading it.

David Lewis in his paper, “Divine Evil” has gone even further in suggesting that the evils perpetrated in these well known examples from the Old and New Testament pale in comparison to the infinite evil of Divine Evil (for all former evils are finite while Divine Evil is infinite). Divine evil is the evil of punishing those to an eternity of suffering (damnation) for facile reasons such as vanity. Yet it is precisely for reasons such as vanity that the Biblical God has given in the Bible (and Koran) for sending people (many of whom presumably lived very virtuous lives) to an eternity of suffering. Making someone suffer infinite torment for an infinite amount of time for facile reasons constitute evil par excellence. It is hard to even imagine what could be more evil. It is infinite Divine Evil. I wholeheartedly agree with this Divine Evil hypothesis, that is, that the being described as the God of the Bible appears to be very evil.

Lewis uses the hypothetical example of a Nazi named "Fritz" to argue his case and who does not act in an evil manner but wholeheartedly supports the Nazi cause of invasion, aggressive wars, and genocide. He admires Hitler very much. Now most of us I think will consider Fritz very evil even if he has not done acts of evil. He is evil simply in virtue of endorsing and advocating those evil acts. He is vicariously evil. That was Lewis's main argument, that people can be vicariously evil in regards to obeying the Biblical commends to worship the evil Biblical being that sends people to torment for eternity for vain reasons much as Fritz can be evil just by admiring and endorsing Hitler. Notice that this vicarious evil need no real object to imbue it with its evil; even if Fritz is a brain-in-the-vat and the Nazi leaders and desired world he endorses is an illusion created by some neuroscientists, he would still seem to be evil in endorsing such evil illusions. Fritz, after all, has no control over whether his world is real or an illusion and if he is guilty of evil in a real world, he is so in a illusory world which is phenomenally identical to that real world. Whether he lives in a real world or an imaginary but phenomenologically identical world is a matter of luck outside of his control and which he presumable is not responsible for.

Assuming that Lewis's argument for the Divine Evil of the Abrahamic God is true, what has this got to do with us? Other than what it has to our moral status there seems to be prudential concerns we must also deal with analogous to Pascal's prudential arguments for believing (or trying to cultivate a belief) in his god.

If the the Abrahamic God is such an infinitely evil being, what prudential arguments analogous to Pascal's wager can we devise to take into account such a conclusion?

In the following table, I will argue in similar decision theoretic terms for what we should believe or disbelieve and maybe be against (that is, cultivate a antipathy or something like it) so that we may avoid eternal damnation and end up with the best chances at getting into heaven.

I have also avoided tricky technical issues with interpretation of the wager and with issues between infinite and finitary decision theory. Alan Hajek employs interesting methods using kinds of ordinal infinite numbers called surreals to remedy the usual problems associated with infinitary decision theory such as problems with mixed strategies (i.e., since run of the mill infinity is reflexive under multiplication of finite probabilities, any mixed strategy will yield infinite utility thereby making a mockery of Pascal's wager), and infinitesimal degrees of beliefs and Roy Sorensen uses more traditional conceptions of (cardinal) infinity to make the wager as fitting as possible to the tasks at hand.

One may also employ very large finite utilities to stand for heaven (and inversely, large negative utilities for hell) but I prefer the infinite strategies as it is closer to our religious and philosophical notion of the utility of heaven and hell but other problems with the nature of non standard uses of infinite numbers inevitably rear their heads making trouble for Pascal's interpretation of the wager as Hajek shows. Since juggling infinite utilities seems to cause technical trouble (tractable or intractable is a matter of dispute) so I will try to avoid such technical issues with the ambiguous utilities of -H (which is hell and may or may not be infinite negative utility) and +H will mean heaven (which may or may not be infinitely positive utility). I think the wager can still be analyzed rationally using just the notion of dominance from decision theory.

-H and +H, even if finite, will likely be under any plausible interpretation, very, very large utilities in opposite directions. One is to be avoided at nearly all costs and the other, desired at equally far reaching costs. I will use “Indeterminate” to be somewhere between these utilities as it is indeterminate what our reward/punishment will be under some such scenario. However, indeterminate leaning towards some direction (such as -H or +H) as it may be will lean towards that direction. So we have the current utility spectrum from least desirable to most (for us).

(-H) is less than (Indeterminate) is less than (+H)

I will use the palindrome "Dog" to denote the Biblical god and assuming that Lewis's argument for its evil is true (and in fact, it may be a sort of analytical truth that a being displaying those qualities are evil as I think Lewis would agree). I will also use "God" to denote a possibly existing being who really is omnibenevolent, omnipotent, omniscient, etc; this is the being with the properties traditionally thought to belong to God's. Thus my terms are:

Dog=Abrahamic/Biblical god=Devil

God=Omnibenevolent, creator...etc

I have assumed (much as Pascal and Hajek's versions of the wager) that your subjective probability for all these possibilities are positive (or all >0). The first three columns of this table denotes the possible situation when God is all powerful with ability to send people to heaven/hell but Dog does not have such powers, and this would correspond with many of our understanding of God and the devil. Dog may claim to hold such powers to be able to send people to heaven/hell as he does in the Bible (if the Divine Evil conclusion is true as we are assuming). The fourth and fifth columns represents possible scenarios where both God and Dog have powers to send us to -H or +H. We have the following 2 by 6 decision matrix:

God & Dog exists

Dog only

God only

God & Dog exists (both equally powerful)

Dog only (Dog has power to send people to -H)

Neither exists

Wager for Dog




Indeterminate (but leans towards -H)

Indeterminate (lean towards -H)


Against Dog







The columns represent these exhaustive possibilities. Either both God and Dog exists, or Dog exists but God does not or God does and Dog does not or neither exists (last column).

When God and Dog exists but Dog lacks powers to send us to -H (first column), it is prudent to wager against Dog. When Dog exists without God but Dog lacks the requisite powers to send us to -H (merely bluffing to, e.g.) as is the case in the second column, the results are indeterminate much as what happens when we die in an atheist's world. When both God and Dog exists and both are equally powerful, it is indeterminate who will get our souls but I am leaning towards -H for those who wager for Dog as both God may want to punish us for our vicarious evil in wagering for Dog and Dog may want to punish us just because he is Dog. He may repay our good turn for our wagering for him with an evil turn. He is evil after all! In that case, both God and Dog will agree on our fate. But he may also wish to reward us for our wagering for him but God may have other ideas as he may wish to punish us. Anyway, I am leaning towards Against Dog as the rational choice in this scenario even though I have labeled it indeterminate what our fate is if we so choose for that reason.

In the fifth column, only Dog exists and he has the power to send us to -H. If we wager for him, he may show us some good will and not send us to -H. But he may also lie (again, he is evil after all!) and decide to send us there anyway against his promises in the Bible to send all faithful to +H. If we wager against him he will certainly send us there.

The last column is where neither God nor Dog exists and in which case we just die and what happens to us is indeterminate or likely we just get a utility of nothing.

Notice the asymmetry. If we ignore the fifth column where Dog has the power to send people to -H and there is no God to check his powers, we have a decision matrix that is dominant for “Against Dog”. That is, if we ignore the possibility that Dog is the only power in the heavens without God to check his evil ways, we would be irrational not to wager against Dog. Even if that columns is true, we may only have little prudential reason to choose to wager for Dog as he might deceive us and send us to -H anyway. It makes sense that the devil pays a good turn with a bad one so given this there is little reason we should decide to wager for him even if this column is in fact the case. This (especially combined with the moral reasons) suggests we should wager against Dog.