Is Plantinga’s Model Of Warranted Theistic Belief

Epistemically Circular?

Richard Brian Davis

Tyndale College


One of the more important lessons to be learned from WCB is that de jure objections to Christian or theistic belief are often based on epistemic principles which fail to satisfy their own criteria. The evidentialist objector, for example, argues that theistic belief isn’t up to snuff because it isn’t self-evident, evident to the senses, or incorrigible; nor does it follow from propositions which are. As Plantinga rightly points out, however, the claim that theistic belief must satisfy these evidentialist criteria in order to be rational, justified, or whatever is itself self-defeating; the claim itself fails to satisfy the very criteria it lays down for epistemic acceptability.

The Circularity Objection

It might be interesting, therefore, to ask whether Plantinga’s models for warranted theistic and Christian belief — the A/C model and its extension — are themselves warranted? More exactly, is Prof. Plantinga warranted in proposing these models? According to one objection — the Circularity Objection — he is not. How does the objection go? Well, for the sake of definiteness, let’s focus our attention on just the Aquinas/Calvin model (A/C, for short). The circularity objector begins by noting that according to the A/C model, theistic belief acquires its warrant by way of the mechanism proposed by the model itself. If this is so, then at first glance it would seem to be the case that

(1) Plantinga is warranted in holding his theistic beliefs only if (2) Plantinga is warranted in believing that his A/C model is correct. But then it looks as though the basis for Plantinga’s theistic belief is his belief that the A/C model is correct — in other words, it appears that (2) is the ground for (1).

To push things further, the critic contends, Plantinga himself admits that (2) is true only if (1) is; that is to say, Plantinga concedes that he is warranted in believing or proposing his model only if theistic belief is warranted for him. This is something of a trivial matter; after all, the A/C model actually includes theistic beliefs. So how could he be warranted in believing the former without first being warranted in believing the latter? And doesn’t this imply that (1) is Plantinga’s ground for (2)? If so, then (1) and (2) ground each other, so that Plantinga’s model ends up acquiring its "warrant" in an epistemically circular (and underhanded) way; in which case he isn’t really warranted in proposing it.

On Plantinga’s Way Out

What shall we say about this objection? Is there a way out for Plantinga here? In the remainder of my paper, I want to briefly examine one such way proposed by Prof. Plantinga, making suggestions here and there as I go. I shall attempt to show that his argument isn’t in fact epistemically circular, but that his way around the objection may in fact concede too much to the objector.

Now it is clear, I think, that Plantinga’s proposal of the A/C model will suffer from epistemic circularity only if he really does affirm (2) — namely, that he is warranted in believing the model. Interestingly enough, however, he never actually makes this claim in the book; nor does he claim that belief in God actually has warrant. What he does say is that his model: (a) is epistemically possible, (b) isn’t philosophically objectionable, if theistic belief is true, and (c) is true (or close to it), if theistic belief is true. Furthermore, as Plantinga correctly points out, he could very well be warranted in asserting the conjunction of (a)–(c) without presupposing that theistic belief had any warrant; but of course this isn’t the case with (2); he couldn’t affirm (2) without assuming that theistic belief had warrant. So the essence of Plantinga’s way out seems to be to let (2) drop out of the picture. Don’t affirm it and the charge of epistemic circularity vanishes.

Here I would like to make a few brief remarks. First, a minor quibble. As Plantinga construes it, a proposition (or model) is epistemically possible if it is "consistent with what we know, where ‘what we know’ is what all (or most) of the participants in the discussion agree on" (p. 169). But wouldn’t the A/C model entail a variety of things that many participants in the discussion would not agree on — for example, that there is such a thing as the sensus divinitatis, a cognitive process, producing in us warranted beliefs about God. This is a specifically theistic belief, and it is far from obvious that it is consistent with, say, the noetic structures of Richard Gale, Michael Martin, or J. J. MacIntosh for that matter. It isn’t clear to me, therefore, that Plantinga is warranted in ascribing epistemic possibility to his model (at least as he appears to be using that concept); in which case it is difficult to see how he could be warranted in believing the conjunction of (a)–(c).

Secondly, I’m wondering whether Plantinga isn’t in a position to affirm (2) and avoid the charge of circularity. For example, I assume that Plantinga believes that he is warranted in accepting (a)–(c). If this is the case, then he (probably) also believes that he is warranted in accepting

  1. If theistic belief is true, then the A/C model is true (or close to it).
(Here we might note in passing that Prof. Plantinga does think his A/C model is true: "I believe," he says, "that the models I shall present are not only possible and beyond philosophical challenge but also true " (p. 169).) Now if Plantinga is warranted in believing (3) and, further, if his theistic belief is warranted, why wouldn’t it follow that he was also warranted in believing his A/C model? Well, according to Plantinga, "It is not the case … that if [theistic] belief has warrant for me, then the model must also have warrant for me" (p. 352).

But why not? This would be the case, Plantinga says, provided that he argued for the truth of theistic belief employing an argument, one premise of which was the A/C model itself. But (i) he never argues in this way and (ii) it would be viciously circular to do so. This seems right, but couldn’t there be an argument or arguments for theism that didn’t contain the A/C model as a premise? I don’t see why not. In fact, it’s hard to think of a theistic argument that does make use of such a premise.

Now this proposal poses a difficulty. If the A/C model gets it warrant by way of warrant transfer from (3)’s antecedent, then in order for Plantinga to show that the A/C model is warranted, it might be thought that he would first have to show that theism is true. But Plantinga says that he doesn’t "know how to do something one could sensibly call ‘showing’ that … [theism ] is true" (p. 170). The assumption here seems to be that if theistic belief acquires its warranted by way of argument (as opposed to the way outlined in Plantinga’s model), then that argument would have to show or demonstrate or prove the truth of theism. And Plantinga doesn’t think this can be done.

But why not? What would it take to show or demonstrate the truth of some proposition or another? Well, Plantinga doesn’t say exactly. But in his book God, Freedom, and Evil he says that a successful piece of natural theology would prove God’s existence; however, in order to do that a theistic argument must not only be sound; its premises must be such that they are accepted by everyone who understands them or at least nearly every rational person. Putting these various ideas together, then, we get:

(4) A person S is warranted in accepting a belief p by way of an argument A only if A proves its conclusion p — that is, only if A is sound and all of A’s premises are such that if S understands them, then S accepts them. Now (4) doesn’t seem to be self-evident, evident to the senses, or incorrigible; nor does it seem to be properly basic with respect to warrant. So why should we think it’s true? Can we perhaps give an argument in its favor? Well, any successful argument for (4) would have to prove that (4) is true. But then the premises of that argument would have to be such that anyone who understood them also accepted them. It seems to me rather doubtful that there are any such premises, in which case whoever accepts (4) on the basis of argument probably isn’t warranted in doing so. Moreover, if (4) is true, probably no philosopher has ever been warranted in accepting any interesting conclusion on the basis of philosophical argument. And surely this is going too far.

The problem here is with the standard (4) lays down; it isn’t at all realistic. So suppose we apply a more realistic standard for what counts as a proof. What then? Well, if we do that, then Plantinga is happy to grant that there are a variety of good (that is, successful) theistic arguments. With respect to his own version of the ontological argument, for example, Plantinga says that it "provides as good grounds for the existence of God as does any serious philosophical argument for any important philosophical conclusion" And in a number of places in the book, Plantinga says that there are at least two dozen or so good theistic arguments; presumably this same compliment would also have to be paid to them But then given this more relaxed standard, it seems that Plantinga would be perfectly justified in claiming that these arguments constitute proofs, and so warrant his belief in theism. Indeed, in another place Plantinga admits that the arguments of natural theology could serve to significantly increase the warrant of one’s belief in God (taken as basic) even to the point of nudging that belief "over the boundary separating knowledge from mere true belief."

The upshot of all this is that if theistic arguments are warrant conferring, then provided that Plantinga is warranted in believing (3), he can transfer warrant from antecedent to consequent and hold that his A/C model is warranted. It doesn’t follow, however, that this procedure is epistemically circular. For although warrant accrues to the model by way of argument, the source of this warrant is the theistic proofs and not the A/C model. At any rate, it is not even clear that Plantinga needs these proofs to complete the warrant transfer. The proofs are only needed for showing that theism is warranted and true. Being warranted and showing that one is warranted, however, are two different things. Plantinga could base his claim that he is warranted in proposing his model on the fact that he is warranted in believing (3) and, further, that his theistic belief is warranted (even if this isn’t shown to be the case). As long as the conditions for warranted belief are satisfied, Plantinga’s belief in God will have warrant. Is this epistemically circular? Not at all. For if the A/C model works the way Plantinga says it does, one could be warranted in believing in God without having to presuppose or assume the model at all. No doubt this is the way things work with most theists.