In my 2002 paper I argued that Alvin Plantinga's defence of pure experiential theism (a theism epistemically based on religious experience) against the evidential problem of evil is inappropriately circular.Footnote ¹ In the above article Eric Snider rejects my argument with several claims.Footnote ²

One is that I do not get Plantinga's thought right. According to Snider, Plantinga maintains that someone can be a rational theist even if she has no beliefs about the cause of her belief in God. On this interpretation it is sufficient that her belief be caused by a properly functioning sensus divinitatis. He claims, however, that I attribute to Plantinga a different view in which the rational theist must not only believe she possesses a sensus divinitatis, but that that belief would serve as both the cause and the reason for her belief in God.

I do not think, however, that I have got Plantinga's thought wrong. We can see why by reviewing his views on rationally warranted theism. He maintains that someone can have a rationally warranted belief in God if the belief is caused by a sensus divinitatis, and she has adequately responded to the possible defeaters to theism that she is aware of. Since some theists are simply not aware of any possible defeaters, they thus can have rationally warranted theistic beliefs without having any beliefs about the cause of their belief in God. They may be properly warranted in believing in God simply on the basis of a religious experience which makes it seem like God's existence is certain. This is analogous to the case where someone has a rationally warranted belief about what he sees with his eyes without having any beliefs about his visual perceptual systems; he can rationally believe what he sees so long as his visual systems are properly functioning and he has the relevant percept.

Plantinga, however, wants to defend the rationality of a more informed theist, viz. the modern intellectually sophisticated person who has been exposed to the best available scientific information and philosophic argumentation. This aspiring enlightened theist is aware of several alleged defeaters for theistic belief. In order for her belief in God to have rational warrant she must successfully respond to them, e.g. by providing what I call neutralizers. Indeed, I see much of Plantinga's work as an extended attempt to defend enlightened theism by appealing primarily to religious experiences.

My argument is that Plantinga's particular defence of experience-based enlightened theism fails because it is inappropriately circular. This is because Plantinga responds to the evidential problem of evil by citing a belief that he has a sensus divinitatis. This belief is not a cause of his belief in God; but it does play a critical role in his defence of theism and thus his belief in God is evidentially dependent on it – without this belief, his belief in God could no longer be rationally maintained.

I argue that it is inappropriately circular for Plantinga to rely on the belief that he has a sensus divinitatis since it is itself evidentially dependent on theistic belief. In making this argument, I rely solely on epistemic principles that Plantinga explicitly accepts. This includes the independence constraint on neutralizers which takes it that belief B cannot serve as a neutralizer for belief A when belief B is evidentially dependent on belief A.

Snider does not attempt to show that Plantinga does not accept these principles, or that my conclusion does not follow from them. I thus reject Snider's charge that I am confused about Plantinga's argument. Perhaps Snider thinks that Plantinga should have argued differently, but that is quite another matter.

Putting these differences of Plantinga interpretation aside I acknowledge the point that according to Plantinga a non-enlightened theist could have a rationally warranted belief in God's existence without believing she had a sensus divinitatis. This suggests that an enlightened theist could also have a rationally warranted belief in God without the rational warrant of that belief depending on her belief she has a sensus divinitatis.

This indeed is Snider's position. He defends this view by rejecting the independence constraint on neutralizers. While he acknowledges that the principle may hold in many cases he does not think it holds for core or ‘metaphysical’ beliefs that inform a person's ‘total world and life view’. When a belief in God is core or metaphysical for a theist and she is faced by the evidential problem of evil she may:

Snider makes a descriptive claim here about how a person may feel no obligation to respond to a challenge to her core, metaphysical beliefs. Perhaps this is so for some or even many theists. But I take the spirit of Snider's claim to be normative – that regardless of her feelings the enlightened theist is not rationally bound to provide a neutralizer for challenges to her core, enlightened beliefs such as that provided by the evidential problem of evil. Later, he states this point more explicitly: ‘The believer in God who believes in a basic way is rational, internally and externally, in maintaining her belief in the face of an evidential challenge like Draper's, even if she neither produces nor attempts to produce a defeater to the evidential challenge’ [my italics] (195). Snider's argument here is that one can be ‘rational’ while refusing to respond in any way to challenges to core, metaphysical beliefs.

This strikes me as deeply wrong. Before explaining why, let me concede that there is something to this idea of protecting the ‘core’ of one's beliefs from evidence that shows they are irrational to hold. In particular, one should not jettison cherished, central beliefs that are evidentially challenged when one is not confident in one's reasoning, or in one's evidence. In such a case, however, the proper response is not simply to ignore the challenge, but rather to put off admitting defeat, acknowledging the need for further study.

But Snider is not claiming that a theist may hold onto her beliefs while she later seeks to further confirm or disconfirm the challenge posed by the evidential problem of evil. For Snider there is simply no obligation to respond.

Why should one get to protect a core of one's beliefs from rational scrutiny in this way? This is seemingly at odds with the grand philosophic project of subjecting all of one's beliefs to scrutiny with the unsettling possibility that even the most cherished ones may not survive. (Of course, there is also always the hope that the beliefs can survive and be all the more valuable for having survived that process.)

Snider's argument for ‘protecting the core’ rests, so far as I can tell, on this alleged counter-example to the independence constraint on neutralizers:

The suggestion here is that, first, one is obviously entitled to believe that one is holding four jacks if one sees that that is so despite its high objective improbability. I agree that this is so. Second, the suggestion is that this belief (although hardly a core one) runs afoul of the independence constraint and thus the principle must be jettisoned in favour of one that allows one to ignore evidential challenges to one's core beliefs. But I fail to see how believing that I have four jacks violates the independence constraint. It is true that all of one's prior beliefs make the fact that one is holding four jacks highly unlikely. But these beliefs do nothing to undermine my belief (supposing that I have it) that I have a reliable visual faculty. In fact, they do quite the opposite and confirm this belief!

So here is the critical difference between the ‘four jacks’ case and the case of the theist challenged by the evidential problem of evil. The evidential problem of evil challenges a person's belief in theism; moreover, it also challenges her belief that she has a sensus divinitatis since this latter belief is evidentially dependent on her belief in God. This is why it is inappropriately circular to cite a belief in having a sensus divinitatis to defend the challenge that the evidential problem of evil poses to theistic belief.

On the other hand, the fact that any particular dealing of four cards has a low probability shows that it is improbable that I have been dealt four jacks. But it does nothing to show that it is improbable that I have a properly functioning visual system. There is thus nothing inappropriately circular on my relying on the belief that I have a properly functioning visual system to neutralize the challenge provided by the fact that it is improbable that I have been dealt four jacks. That I have a reliable visual faculty is evidentially independent of the belief that I am holding four jacks, and thus there is no violation of the independence constraint in relying on it in this case.

This alleged counter-example to the independence constraint on neutralizers thus fails, and leaves no other reason to embrace Snider's alternative, which overly protects one's cherished beliefs from rational scrutiny.

Perhaps, though, Snider's claim that a theist may rationally ignore the challenge posed by the evidential problem of evil is not about epistemic but rather practical rationality. On a widely held view of practical rationality, one should do whatever one most prefers or wants to do. This could provide support for the view that it would be practically rational to maintain the core of one's beliefs rather than subject them to rational scrutiny since one might prefer to keep these cherished beliefs over being epistemically rational.

For this view to be plausible, though, it is necessary to recognize that one has an epistemic duty to respond to the evidential problem of evil. It's just that on this view that if one feels no obligation to respond to this epistemic duty then one has no practical obligation to do so.

I find the view of practical rationality to be mistaken since I think that epistemic reasons just are basic practical reasons. That said, they can clash with other practical reasons and so we should specifically examine the case of the theist confronted by an evidential challenge to theism.

It is important to note that this is not a case where one's own life or the lives of others would be threatened by following one's epistemic duties. Rather, it is one's cherished beliefs which are at stake, as well as the ways of living closely associated with those beliefs. It is surely relevant, though, that people often survive and even flourish after the loss of such core, cherished beliefs. On the other hand the cost of losing one's philosophic soul (by ignoring evidential challenges to one's beliefs) is so high that I would judge it too great to bear in comparison to potential costs of giving up one's cherished beliefs. But this is taking me away from my point that Snider's proposed alternative to the independence constraint is not an epistemic principle.

The kind of rationality I am interested in – epistemic rationality – is directed towards the truth, and not towards protecting one's beautiful mind from disturbance. This kind of epistemic rationality requires a ‘philosophic disposition’ in which one subjects all one's beliefs to rational scrutiny, especially including the most core or ‘metaphysical’ ones.

Let me finish by expanding on my concluding remarks of the 2002 paper. Qua philosopher, I aim to have epistemically rational beliefs, but I don't aim simply to have such beliefs. This is just too low a bar too easily met. Rather, I aim to have beliefs that are epistemically rational in the light of the best available scientific information and philosophic argumentation. To stand unmoved as one of my most cherished beliefs was evidentially challenged would be to betray the philosopher in me who strives towards the truth, especially in regards to the matters of greatest importance.