Knowledge of God (Blackwell Great Debates in Philosophy). By Alvin Plantinga and Michael Tooley


Pp. ix, 270 , Oxford : Blackwell Publishing , 2008 , $34.95.

Plantinga's Evolutionary Argument Against Naturalism (EAAN), which he's been offering in various forms since Warrant and Proper Function, has drawn considerable attention over the last fifteen years. Some of this attention can be explained by the fact that, like most arguments worth discussing, determining how it ought to be formulated is just as difficult as evaluating its premises. The same explanation can be offered for the ink that's been spilled over the evidential problem of evil. Much of what makes Knowledge of God valuable, then, is that in it Plantinga and Tooley manage to shed a lot of light on how these arguments work, where their weaknesses are, and how they might be patched up in the face of some serious criticisms.

Each author contributes an opening statement, reply, and closing statement. Plantinga goes first. He begins by arguing that if the essential tenets of Christianity are true, then the the odds are good that they're known by Christians. This conclusion falls out of his proper function-based epistemology and the postulation of what Calvin called the Sensus Divinitatis– i.e., ‘a sense of divinity, a faculty, a set of cognitive processes whereby we come to know about God’ (p. 7). Plantinga holds that beliefs are warranted if and only if they are produced by one or more properly-functioning cognitive faculties that are successfully aimed at truth and that are being exercised in the environment for which they were designed. Now, if humans have the Sensus Divinitatis, and this faculty is aimed at truth and is designed to produce belief in the essential tenets of Christianity, then as long as it's functioning properly and operating in the right environment, those beliefs are warranted – indeed, they can be known.

Plantinga then offers three arguments against naturalism. The general argument here is that since naturalism and theism are the only serious research programs around, if naturalism fails, then theism ought to be accepted. Plantinga contends that naturalism (1) cannot account for proper function, and so can't account for properties like healthiness and sickness, (2) entails a form of skepticism, and (3) is incompatible with the view that people have beliefs. Anyone familiar with the last two decades of Plantinga's work has seen these arguments before, although there are various minor differences in their presentation here. One point that emerges more clearly in this version of the EAAN, for example, is that it is really a version of Benacerraf's Dilemma: Just as one might wonder how we can know the truths of mathematics, given that knowledge requires causal interaction and that Platonism is true, so one might wonder about the likelihood that natural selection would produce cognitive structures that correspond to the true propositions, given that natural selection is a causal process and that propositions are causally impotent.

The bulk of Tooley's opening statement is a defense of the evidential version of the problem of evil. At the outset, though, he tries to show that atheism is the default position by arguing that various possibilities other than theism are equally probable a priori, including the existence of an omniscient, omnipotent, and perfectly evil being and the existence of omniscient, omnipotent, and morally indifferent being. If this is right, then the a priori probability of a variety of theism that is religiously satisfying is, at best, one-third. So, before the project of natural theology has even begun, the odds are good that atheism is true.

Tooley's discussion of the problem of evil is carefully nuanced. He explains the pros and cons of at least four different choices that the atheologian must make in constructing the argument, and offers reasons to prefer a concrete as opposed to an abstract, inductive as opposed to a deductive, and deolontological as opposed to an axiological formulation. He also includes a helpful procedure for translating formulations that make use of objective values – like the one he offers – into ones that only make mention of subjective values.

Tooley's version of the problem of evil rests on the following premise: 16) For any action whatever, the logical probability that the total wrongmaking properties of the action outweigh the total rightmaking properties – including ones of which we have no knowledge – given that the action has a wrongmaking property that we know of, and that there are no rightmaking properties that are known to be counterbalancing, is greater than one half (p. 120).

If one assumes that, for any event E, God either directly causes E or permits E, then there is an action that God performs with respect to every event – namely, that of either causing or allowing it to occur. Now consider some horrible event such that permitting it seems to have wrongmaking properties (e.g., causing seemingly-unnecessary suffering, etc.). If it's reasonable to believe that there is no rightmaking property of that event that outweighs its wrongmaking properties, as (16) claims, then it's reasonable to believe that God does not exist at the time of that event, since we can presume that God would not permit the event unless there were some counterbalancing rightmaking property, and yet the event occurred. The fact that we use some principle very much like (16) in ordinary moral reasoning is some evidence for its truth, and if it is true, then Tooley has a powerful argument against theistic belief.

In Plantinga's first reply he attempts to undercut Tooley's case for atheism as the default position by arguing that we don't see that various alternatives to theisms are equiprobable; rather, we fail to see that the various alternatives aren't equiprobable, and we easily confuse this for having a much stronger insight. (Oddly, Tooley doesn't use the same strategy to criticize Plantinga's argument that wholly material entities can't have beliefs, even though Plantinga's modal intuitions aren't obviously on firmer ground than Tooley's probabilistic ones.) He goes on to contend that we don't have any reason to affirm or to deny (16), the crucial premise in Tooley's version of the evidential argument from evil. Plantinga thinks that we don't have enough moral knowledge to tell one way or the other about the distribution of rightmaking and wrongmaking properties. He also argues that even if theism is improbable, belief in theism might still be justified; this, of course, is due to the fact that belief in theism might be produced by a cognitive faculty that is functioning properly, and so on. Plantinga concludes by arguing that evil is not a defeater for belief in God.

Tooley's first reply begins with a few considerations that tell against Plantinga's externalism. He then offers a causal theory of content that he takes to be compatible with materialism; this serves the dual purpose of providing a response to Plantinga's argument that material things can't belief anything and of laying the foundation for his criticism of the EAAN. Tooley points out that the EAAN depends on a premise much like the following:

(7) If Darwinian evolution is true, then even if it is true both that neural states of type N are, in an organism H, reliable indicators of the presence of an instance of property P, and also that states of type N have content C, there is no reason why content C need be related to property P (p. 208).

But if a causal theory of content is true, then (7) is probably false; if some type of neural state in an organism is a reliable indicator of the presence of an instance of some property, then that type of neural state does have the relevant content. And if Darwinian evolution is true, then we have good reason to think that that there are types of neural states that are reliable indicators of the presence of an instance of various properties, since it's hard to see how any creature could survive whose neural states were unreliable in this regard. So, Plantinga's EAAN stands or falls in part on at least one of his other arguments against naturalism.

In his closing comments, Plantinga argues that Tooley's reductive theory of content is unable to explain how we could represent specific objects, properties, and relations in the way that we do; if this is right, then the EAAN emerges unscathed. In his, Tooley argues that Plantinga hasn't captured the relevant notion of epistemic responsibility that is at issue when discussing belief in God, that we have no reason to believe that there is a reliable belief-forming faculty devoted to religious beliefs, and that we can make inferences about God's character based on the existence of evil, thereby acquiring a defeater for theistic belief.

Knowledge of God is an excellent volume. The debate between Plantinga and Tooley is rigorous and wide-ranging, covering far more than I've had room to indicate here. The insights provided by their dialectic make this a much more helpful than a pair of monographs would be. For all that, though, Knowledge of God is not all that it could have been. One editorial quibble is that by the time Tooley has the opportunity to reply to Plantinga's opening statement, 115 pages have gone by; the result is that one has to flip back and forth in order to recall the arguments that are being criticized, an irritation that could have been avoided had the book been split into two parts, the first devoted to Plantinga's opening statement, the second devoted to Tooley's. A more substantive complaint is that it's not at all clear what the debate is about. Are they trying to determine whether someone could know that God exists? Whether an atheist or agnostic should know that God exists? Whether theists shouldn't claim to know that God exists? Or is the debate really about warrant, or justification, or epistemic responsibility? Neither the title nor the text settles these questions. Now, I am not suggesting that Plantinga and Tooley aren't sensitive to these issues; they make plenty of remarks that indicate their views. The problem is that they make little effort to come to a consensus, or at least to be explicit about how they differ (although Tooley goes some distance toward the latter in his closing comments). The upshot is that while Plantinga and Tooley are in top form when discussing the EAAN and the evidential problem of evil, they often seem to be talking past when it comes to the significance of these arguments. Some of this vagueness is probably inherent, but it might have been mitigated in an introduction, co-authored or otherwise, which the book lacks.

Of course, none of this should keep anyone from working through Knowledge of God, which is probably required reading for philosophers of religion. Any minor flaws that the book has are certainly outweighed by the fact that it is, without question, an extremely valuable contribution to the literature.