Mirror of Justice

Michael Perry called Professor Eberle’s comments to my attention and kindly invited me to respond.

I shall skip Professor Eberle’s first paragraph, which simply plays rhetorical tricks with my comments and is irrelevant to the issues of substance.

Professor Eberle would like to minimize the significance of Wieseltier’s mistakes by describing them as just a matter of not “hav[ing] all of his philosophical ps and qs together,” and as “minor interpretive points,” and as showing that he “isn’t a professional philosopher.” I take it Professor Eberle is conceding, delicately, that Wiesletier is, indeed, ignorant about the subject on which he is writing: to wit, why one might think scientific explanations of psychological and social phenomena might be worth pursuing; how Dennett’s project relates to Hume’s; what Hume’s theism does or doesn’t amount to; and the relevance of the etiology of belief to the epistemic status of that belief. (This last point Professor Eberle describes, correctly, as “a pretty standard point in contemporary epistemology,” which, of course, confirms my point about Wieseltier’s embarrassing ignorance of the subject on which he opines.) Professor Eberle can not really be serious that to be mistaken on all these points is somehow trivial; they are central to the review and Wieseltier’s criticism of Dennett.

Professor Eberle says correctly that Dennett’s speculative story about the origin of religious belief “can be used to discredit claims to religious knowledge only if the story Dennett tells about the history of religious belief…is both true and provides reason to believe that religious belief is formed unreliably.” As I explicitly noted in my comments on the review, Wieseltier’s one fair point is about the speculative status of the story Dennett is telling in the book. Professor Eberle then says, wrongly, that, “If Dennett’s narrative isn’t actually true, then Leiter’s repeated reference to the epistemic relevance of the etiology of belief is itself irrelevant.” This non-sequitur surprises me, for I did not think the basic point here was a hard one. The etiology of belief can be relevant to its epistemic warrant (which Professor Eberle concedes, and which Wieseltier denies, because he is ignorant); Dennett offers a speculative story about the etiology of religious belief, involving an extrapolation from paradigmatic explanations in evolutionary biology; if the speculative story is true (or well-confirmed, as I put it), then it raises questions about the epistemic warrant for religious beliefs. That the argument has the form of a conditional, whose antecedent is undecided, does not make the point about etiology irrelevant—and especially not when Wieseltier denies the purely epistemic point!

The Plantinga argument for the self-refuting nature of evolutionary naturalism is the subject of a large secondary literature, but it is not relevant to the “game” I was playing in my comments on Wieseltier. If Dennett’s (or anyone’s) argument were that any cognitive capacities that have an evolutionary explanation are epistemically suspect, then the Plantinga argument would be clearly relevant. But that is not Dennett’s (or anyone’s) argument.

I take it Dennett’s (and everyone else’s) argument has the same structure as Hume’s original one: where beliefs lack a rational warrant, we need some explanation for why the beliefs are held nonetheless. Dennett favors an evolutionary one; Hume favored a different one; Freud had another; Marx another; Nietzsche another; and so on. I take it, then, that what Professor Eberle wants to claim is this: since the naturalistic explanations in the offing do not involve reliable methods for belief formation, if they are the correct explanations, then they undermine the epistemic warrant (if not the truth) of the beliefs in question. But then, so the argument goes, doesn’t the same consideration count against the reliability of our cognitive faculties that are the product of these same unreliable mechanisms?

Notice, to start, that an unreliable mechanism [e.g., natural selection] for forming beliefs is not, a fortiori, a mechanism that necessarily produces unreliable belief-forming cognitive mechanisms. That requires a separate argument. Plantinga has one, about which more in a moment. Wieseltier runs these points together in the review; Eberle appears to do the same, but perhaps this was simply due to being terse.

Plantinga’s is an interesting, if not successful, argument, but it is still, I think, irrelevant to the review of Dennett. (Professof Eberle is probably right that Wieseltier is recycling the Plantinga argument, which gets a lot of play in pious circles; alas, the responses to Plantinga don’t seem to get the same kind of play.) The structure of Dennet’s argument, again, runs this way: (1) religious beliefs (e.g., in God or gods, in an immaterial soul, in eternal life) are either false or lack rational warrant; (2) many people hold religious beliefs, which requires some explanation; (3) the explanations in the offing may loosen religion’s grip on people since, if true (or found plausible), they will persuade people that they do not hold their religious beliefs for good reasons. None of these points requires a general defense of the reliability of our rational and cognitive faculties, the issue raised by the Plantinga argument. If they aren’t reliable, then all arguments fail to get started in philosophy, not just this one. I’m of the view that no one has a good response to the global skeptic, and that’s what is being asked for here. If that’s a real philosophical problem, it’s a problem for Plantinga as well as Dennett.

Even if we can’t respond to the global skeptic, we can respond to Plantinga’s case for that skepticism. Plantinga’s precise argument turns on some technical details about probabilities, that don’t admit of simple explanation (but which I would have thought were dealt with decisively a few years ago in a paper by Branden Fitelson and Elliott Sober on “Plantinga’s Probability Arguments Against Evolutionary Naturalism” [which I think appeared in Philosophy of Science, not sure about that]). Crudely, Plantinga thinks the probability of our cognitive mechanisms being reliable given the truth of evolutionary naturalism is low; but since we have reason to think our cognitive mechanisms are reliable, we have reason to doubt evolutionary naturalism and, moreover, reason to accept theism (which makes more probable the reliability of our cognitive mechanisms).

This argument involves a lot of quite dubious moves and assumptions: e.g., (i) that the prior probability of evolutionary naturalism being true is about the same as the prior probability of theism being true; (ii) that the reliability of our cognitive mechanisms is to be assessed in toto (but when we parse them into, e.g., perception, mathematical knowledge, decision-making under uncertainty, memory, and so on, it turns out that their reliability varies wildly, and in ways that evolutionary naturalism fits rather neatly (and which theism has a lot of trouble explaining)); (iii) that beliefs and actions are not causally connected in ways that would suffice for natural selection operating on actions to nonetheless favor reliable belief-producing mechanisms; and (iv) that the only ground an evolutionary naturalist has for thinking our cognitive mechanisms are reliable is evolutionary naturalism itself. None of these moves or assumptions are, I think, defensible, and so Plantinga’s argument fails.

I will go out on a limb and venture that Wieseltier, who mimics in summary form Plantinga’s argument, neither understands these issues nor cares about them. I think philosophers should care about them, and I also would have thought that, by this point in time, Plantinga’s clever arguments had been responded to effectively, such that something new is needed by way of counter-argument for any philosophers, not just naturalists, to be worried about them. At best, the Plantinga argument recommends fallibilism, that is, a willingness to admit that one is wrong. But that just aligns Plantinga with the arch-naturalist Quine. Fallibism is a good epistemic posture, at least with respect to hard questions like these. But it hasn’t much to do with Dennett’s attempt to give a naturalistic account of religious belief, or with Wieseltier’s confused comments on that attempt.
_______________
mp