The End of Skeptical Theism? (Part 1) - Rowe's Evidential Argument

This post is part of my series The End of Skeptical Theism?. For an index and introduction, see here.

The primary reference point for all discussions of skeptical theism is the evidential problem of evil. I've discussed this before, but I want to do so in more detail now because it is essential to understand the strengths and weaknesses of the argument before assessing skeptical theism. After all, skeptical theism was originally introduced to cope with the evidential problem of evil.

Three things will be discussed in the remainder of this post. First, I will look at two versions of the evidential problem, both from William Rowe. Second, I will point out the great strength of Rowe's argument, namely that it sets the bar exceptionally high for any potential theodicy. And third, I will point out the aspect of the argument that is challenged by skeptical theists.

Although this post is reasonably detailed, it is obviously not an exhaustive treatment of the topic. For that, I would highly recommend this article by Nick Trakakis.

A Dead Fawn, E1 of Rowe's 1988 argument

1. Two Versions of Rowe's Argument
Probably the most famous presentation of the evidential problem of evil comes from William Rowe's 1979 article. Rowe focuses on evil in the form of intense human and animal suffering, and then gives us the following argument:

• (1) There exist instances of intense suffering which an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse. (Call this the "Factual" or "Evidential" premise)

• (2) An omniscient, wholly good being would prevent the occurrence of any intense suffering it could, unless it could not do so without thereby losing some greater good or permitting some evil equally bad or worse (Call this the "Theological" premise).

• (3) Therefore, (probably) there does not exist an omnipotent, omniscient, wholly good being.

The argument is evidential, as opposed to logical, in that it does not pin all its hopes on the logical incompatibility of God and any amount of evil (however slight). Instead, its hopes are pinned on the volume, intensity and apparent pointlessness of the evil we observe.

In 1988, Rowe presented a slightly different version of the argument which went as follows:

• (P) No good we know of justifies an omnipotent, omniscient, perfectly good being in permitting E1 and E2.

• (Q) No good at all justifies an omnipotent, omniscient, perfectly good being in permitting E1 and E2.

• (not-G) There (probably) is no omnipotent, omniscient, perfectly good being.

In this case E1 stands for the suffering and death of a fawn trapped in a forest fire; and E2 is an actual case involving the rape and murder of a five-year-old girl. Rowe focuses on these because they are striking instances of both moral and natural evil.

I think a truly effective presentation of the argument will have many examples to draw upon (E1....En) and won't get bogged down in just two instances of evil. Hasker emphasises this point in his criticism of skeptical theism, which we'll get to in due course.

Anyway, the two versions of the argument are mapped out below.

As you can see they are slightly different in terms of their logical structure. These differences are important when trying to work out the strengths and weaknesses of the argument.


2. Setting the Bar High
The great strength of the argument is obvious when you consider the theological premise in the 1979 version. Effectively, this premise states that suffering (or evil more generally) is only permissible if outweighed or defeated by some greater good. But given the type of being God is supposed to be, this can only be the case when the following condition is met:

• The existence of the original evil was logically necessary for the greater good to obtain. The "logically necessary"-condition is warranted because God is an omnipotent being who has the power to do everything that it is logically possible to do.

This would seem to rule out any argument suggesting that the evil is defeated by the mere fact that it caused or brought about some good consequences. This is because presumably, given his omnipotence and the fact that he can exert his power in a basic manner, God could (logically) have achieved those consequences without the intervening evil.

It is only right to point out that this has an obvious corollary: a good state of affairs is not necessarily defeated just because it has evil consequences. Those consequences might have been logically necessary for bringing about the good.

As can be seen, the theological premise sets the bar exceptionally high for defeating the evidential argument. The theist has four options when responding.

1. They can try to identify the logically necessary greater good that God is trying to bring about. This involves the construction of a theodicy. 
2. They can concede the force of the argument but argue that since its conclusion is probabilistic (it is "evidential" after all) there are other considerations that tip the balance in favour of the existence of God. 
3. They could argue that their understanding of God does not commit them to accepting the tri-omni properties suggested by Rowe. 
4. They can take the skeptical theist route. This route switches the focus from the theological premise to the factual or evidential premise, and to the questionable inference that seems to underlie Rowe's argument.

3. The Questionable Inference

The questionable inference is clearly identified in the diagram depicting the 1988 version of Rowe's argument. It is also implicitly propping-up premise 1 of the 1979 version. The problem is that Rowe jumps from observed instances of intense suffering, for which we cannot locate a logically necessary reason, to the fact that no such reason exists. 

This inference -- from potentially gratuitous evil to actually gratuitous evil -- can be supported by something I am going to call "the principle of warranted induction". According to this principle, we are warranted (within our epistemic rights) when we make generalisations from our own understanding and experiences. This principle is obviously not absolute since all inductive generalisations are open to defeat. Nonetheless, accepting it as a defeasible rule seems reasonable and allows us to make the following argument in support of P1 from the 1979 argument:

• (i) There is an enormous variety and profusion of intense human and animal suffering in the world.

• (ii) Much of this suffering seems logically unrelated to any outweighing greater good. In other words, we can't see why it is logically necessary for some greater good.

• (iii) If it seems to be case, it probably is the case (Principle of warranted induction).

• (iv) There exist instances of intense suffering which an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse.

(iv) is identical to P1 from 1979.

The essence of skeptical theism is the rejection of this argument. Its proponents maintain that we are not warranted in making the generalisation from what appears to be the case to what actually is the case. 

Ideally, they want this skepticism to only affect our ability to make moral judgments about God. They do not want it to expand into their day-to-day moral judgments, their understanding of the natural world, their reading of scripture and their knowledge of their own religious experiences. Critics argue that their skepticism cannot be confined in this manner.

We'll see why the critics think this in the remainder of the series.