Skip to content

Voodoo Epistemology

This is the text of comments on Alvin Plantinga’s book, Warranted Christian Belief, that I delivered to a meeting of the Society of Christian Philosophers on Dec. 27, 1999.  

Keith DeRose
But what then is the problem?”   -Plantinga, RBG, p. 75
One of the most natural and most important objections to Alvin Plantinga’s defense of theistic (and specifically Christian) belief is what Plantinga, back when his religious epistemology was young, dubbed the “Great Pumpkin Objection” (GPO). I’ve often taught his original response to GPO in “Reason and Belief in God”(1)(henceforth RBG). It’s always seemed to me that Plantinga misconstrued the objection then. Of course, there’s something problematic about claiming that Plantinga misconstrued the “Great Pumpkin Objection”. “GPO”, after all, is a label Plantinga gave to a particular objection, and he was free to give it to whatever objection he wanted to. What I mean is that the objection, as he responded to it, was made needlessly weak; there’s a better objection — the “real” GPO — in the vicinity, and Plantinga even seemed on the verge of considering it in RBG.

Since I’ve always considered GPO to be a nagging objection, when I received Plantinga’s new book,(2) one of the first things I checked was whether he had updated his treatment of GPO. I was happy to learn that he has. He now treats an objection he calls “Son of Great Pumpkin”, which he attributes to Michael Martin, and this objection seems to be fairly close to what I’ve been thinking of as the “real” problem here.

I consider this “real” objection to be interesting and important. And I think that much of what Plantinga says in response is also interesting and important. But I think Plantinga is even now still making things too easy on himself by construing the objection in some needlessly weak ways, and much of his response takes advantage of this needless weakness. I myself don’t consider this a “killer” objection, even in its best light, but I do think that it’s worth facing squarely.

Plantinga introduced GPO with this paragraph of RBG (the sentence numbering is added):

The Ur-formulation of the Objection: (1) It is tempting to raise the following sort of question. (2) If belief in God can be properly basic, why cannot just any belief be properly basic? (3) Could we not say the same for any bizarre aberration we can think of? (4) What about voodoo or astrology? (5) What about the belief that the Great Pumpkin returns every Halloween? (6) Could I properly take that as basic? (7) Suppose I believe that if I flap my arms with sufficient vigor, I can take off and fly about the room; could I defend myself against the charge of irrationality by claiming this belief is basic? (8) If we say that belief in God is properly basic, will we not be committed to holding that just anything, or nearly anything, can properly be taken as basic, thus throwing wide the gates to irrationalism and superstition? (p. 74)

The root worry behind GPO is that Plantinga’s defensive strategy (claiming that belief in God is properly basic) does, or would if accepted, throw “wide the gates to irrationalism and superstition” (U8), where we’re given samples of “bizarre aberrations” (U3) of “irrationalism” — voodoo, astrology, belief in the Great Pumpkin, and the strange belief that one can fly (U4-6) — to serve as examples of the disreputable lot that we’d be inviting in. We never look carefully at the examples, though, no doubt, they can, and, in the case of voodoo and astrology, actually are held in importantly different ways. I take it the details aren’t important. The objector is protesting: “Look, Plantinga, there are some sets of belief (or possible belief) that even you would want to pass negative epistemic judgment on (where “irrational” is the negative judgment stressed here), right?” And I take it Plantinga — at least back in 1983 — is answering: “Right.” There doesn’t seem to be any hint in 1983 of Plantinga responding: “Well, if you look closely, those beliefs really aren’t so bad.” (At least no hint I can find, and I just read through RBG’s treatment of GPO on the look-out for such a hint.) Which is why, I take it, we never look closely. It’s pointless. The objector and Plantinga are agreed that there are bizarre aberrations of irrationalism, and the question is. . .

Well, there are importantly different questions to be distinguished here, corresponding to two importantly distinct GPOs that are leveled or at least hinted at in the Ur-paragraph. The tough question, asked in U7, is whether, if Plantinga held one of the “bizarre” beliefs, he could “defend [him]self against the charge of irrationality by claiming this belief is properly basic.” Alas, rather than addressing this tough question, Plantinga moved on, in U8, to an easy question: whether, in accepting belief in God as properly basic, Plantinga was thereby committing himself to the “bizarre” beliefs being properly basic as well. And his response was confined to answering the easy question, leaving the hard one behind.

To see the difference between the two questions, consider Plantinga’s answer, which I claim answers only the easy question. We pick up his answer at the end:

So, the Reformed epistemologist can properly hold that belief in the Great Pumpkin is not properly basic, even though he holds that belief in God is properly basic and even if he has no full-fledged criterion of proper basicality. Of course he is committed to supposing that there is a relevant difference between belief in God and belief in the Great Pumpkin if he holds that the former but not the latter is properly basic. But that should prove no great embarrassment; there are plenty of candidates. These candidates are to be found in the neighborhood of the conditions that justify and ground belief in God — conditions I shall discuss in the next section. Thus, for example, the Reformed epistemologist may concur with Calvin in holding that God has implanted in us a natural tendency to see his hand in the world around us; the same cannot be said for the Great Pumpkin, there being no Great Pumpkin and no natural tendency to accept beliefs about the Great Pumpkin. (RBG, p. 78)

Of course, believers in the Great Pumpkin, if there were any, would object to that last part. (Well, of course, they’d object to a lot of the treatment they receive in this discussion!) Is it fair for Plantinga to simply declare: “there being no Great Pumpkin and no natural tendency to accept beliefs about the Great Pumpkin”? Well, that is the proper response if he’s answering the easy question. Nothing we’ve seen has any tendency to commit Plantinga to accepting the Pumkinites’ belief as properly basic.

But Plantinga’s above response naturally triggers the question: “Well, won’t the Pupkinites be able to just as well say that belief in the Great Pumpkin is properly basic, but belief in the God of traditional theism is not, there being no God of traditional theism, and no natural tendency to believe in Him?” This is moving to the hard question, which does not concern whether Plantinga is committed to the proper basicality of Pumpkinism. The asker of the hard question sees that neither Plantinga nor Reformed Pumpkinites are committed to the proper basicality of the other’s beliefs. It concerns rather whether the Pumpkinites can use Plantinga’s defense as well as Plantinga can.

Plantinga formulates Martin’s “Son of Great Pumpkin” objection as follows:

1. If Reformed epistemologists can legitimately claim that belief in God is rationally acceptable in the basic way, then for any other belief accepted in some community, the epistemologists of that community could legitimately claim that it was properly basic, no matter how bizarre the belief.

2. The consequent of 1 is false.

So, 3. The Reformed epistemologist can’t legitimately claim that belief in God is rationally acceptable in the basic way. (forthcoming, p. 28)

Plantinga kicks off his reponse as follows:

Is this a good argument? One initial problem is that the argument is pretty loosely stated; Martin doesn’t tell us what he means by ‘rational’, and he doesn’t tell us what he means by ‘legitimately’.  (p. 346)

Plantinga then proposes various things that Martin could mean by those two terms, concluding that in no case do you get a good anti-Plantinga argument; he concludes that Martin’s argument

is multiply ambiguous, inheriting the multiple ambiguity of ‘legitimately’ and ‘rationality’. Most of the disambiguations, however, show no promise at all. The last disambiguation, where both ‘legitimately’ and ‘rationality’ are both understood as referring to warrant, is at least interesting; the argument so construed, however, suffers from the annoying defect of having a false premise. Son of Great Pumpkin does no better than Great Pumpkin. (pp. 349)

I can’t here go into detail about all the “disambiguations” Plantinga considers. But there’s plenty of room to worry that much of Plantinga’s treatment of Martin is unfair — particularly his treatment of Martin’s use of “legitimately.” None of Plantinga’s proposed “disambiguations” come close to getting it right, in my opinion. (My best guess as to what’s meant by “legitimately” here is (roughly): with the backing of a successful defensive argument.) One might as well pick on Plantinga’s use of “does no better than” in the above quotation, proposing fairly technical and unpromising possible specifications of it from within one’s own theory of arguments:

Plantinga argues that Son of Great Pumpkin “does no better than” does Great Pumpkin. But that’s very loosely stated and highly ambiguous. Plantinga never tells us what he means by “does no better than.” As we’ve seen, I’ve identified 38 different things that can be meant by saying that one argument “does better” than another. So, we’ll have to disambiguate Plantinga’s claim, and see if his argument has much promise on any of them….

If none of the unpromising 38 proposals seemed accurate reflections of what he meant by “does no better than”, Plantinga would be within his rights to just reject them all. That none of the 38 identified “disambiguations” makes the argument come out “good” (whatever that might mean!) stings only if those are the 38 correct disambiguations. I don’t think any of Plantinga’s proposals accurately reflect Martin’s meaning. But rather than getting bogged down in issues about what exactly Martin meant, and whether Plantinga is construing him correctly, let me just cut right to the part where I present GPO as I think it’s best formulated.

Like other philosophers, Plantinga often evaluates arguments by means of various simple terms of appraisal. A positive argument, or an objection, or a response to an objection is said to “succeed” or to “not succeed” or “fail”. One of Plantinga’s favorite terms of appraisal as applied to arguments seems to be “cogent”. And, as we’ve seen above, Plantinga uses comparative constructions, saying that one argument is “better than”, or, as above, “no better than” another. As Plantinga himself realizes,(3) it’s no easy task to say what it is for an argument to be successful. Following a lead of Plantinga’s, some of the difficulties can be brought out by means of examples like:

Argument A
1. Either God exists or 2+2=5
2. It’s not the case that 2+2=5
So, C. God exists

Argument B
1. Either God doesn’t exist or 2+2=5
2. It’s not the case that 2+2=5
So, C. God doesn’t exist

Though there may turn out to be importantly different things that can be meant by calling an argument “successful”, I take it that both of these arguments, in just about any way we should be interested in, are unsuccessful. The philosopher who claimed to have a successful argument for God’s existence would be wrong, if it’s Argument A she’s referring to. Ditto for the philosopher who would claim that she has a successful argument for God’s non-existence, if it’s B that she has in her pocket. Yet, on the assumption that one of the pair “God exists” and “God does not exist” is true, one of the above arguments is sound: They’re both valid, they each have a true second premise (the same second premise, as it turns out), and the one with a true conclusion also has a true first premise. Soundness is not sufficient for successfulness, on any proper use of “successful”, as applied to arguments. (Nor is it necessary, in my view.) About the argument of the above that is sound, we may ask: What’s wrong with this sound argument that renders it unsuccessful? I trust everyone here has some ideas about how to begin to answer this question. I suspect that few, if any of us, can fully answer the question of what success in an argument consists in. I trust that those of us who cannot fully answer that question will continue to evaluate arguments as successful/unsuccessful even before we have a complete account in hand. At any rate, Plantinga seems to be willing to let such evaluations pass without complaining that they’re problematically unclear.

The hard question compares Plantinga’s defensive strategy, as an argument, in terms of that completely legitimate but difficult-to-analyze property of success, with an imagined use of that same strategy by a Pumpkinite. The question is whether Plantinga’s use of the strategy is any more successful, or cogent, perhaps, than is the Pumpkinite’s. To the objector, it seems that, and there seems to be nothing to block the conclusion that, to use Plantinga’s own phrase, Plantinga’s defense is “no better than” the Pumpkinites’ defense. So, Plantinga is getting very close to the real question/objection, when he asks himself in the current book what we might call:

The Pointed Question: Now couldn’t this be argued with equal cogency with respect to any set of beliefs, no matter how weird? (p. 350)

But this is still, in one important way, making things too easy on himself. A major part of Plantinga’s response to the Pointed Question consists in citing examples of beliefs that could not be defended by means of his strategy. Of course, this allows Plantinga to answer the Pointed Question, as formulated above, in the negative, but it won’t do as a response to the real problem. For the objector need not claim that just any set of beliefs can be defended Plantinga-style. For our “root worry” that accepting Plantinga’s defense would be to throw “wide the gates to irrationalism”, it’s not necessary that all irrationalisms would be invited into the realm of the successfully defensible (the gates needn’t be open that wide before we’re rightly worried that they’re open too wide); it’s enough that there be some“bizarre” (to use the term of 1983) or “weird” (the term in the Pointed Question; I take it these are supposed to pick out the same category) “aberrations” of “irrationalism” which would be let in.

Plantinga at least partially recognizes this. After asking himself the Pointed Question, he answers it as follows:

The Pointed Answer: Certainly not. Many propositions are not such that if they are true, then very likely they have warrant: the proposition No beliefs have warrant comes to mind. (p. 350)

Some background is needed here. Plantinga’s defensive strategy now goes beyond claiming that the beliefs in question are properly basic and so not in need of evidential support (though I think the change I’m about to mention was already implicit to some degree back in Plantinga’s writings in the early 80’s). The defensive strategy now crucially includes the claim that the beliefs Plantinga seeks to defend are such that if they are true, then they are likely warranted. In his Pointed Answer, Plantinga is pointing out that not just any set of beliefs will have this feature. But Plantinga then realizes that the objector needn’t be claiming that just any set of beliefs can be defended Plantinga-style. So Plantinga has the objector regroup; immediately after the above Pointed Answer, Plantinga writes:

But, you say, isn’t this just a bit of logical legerdemain; are there any systems of belief seriously analogous to Christian belief for which these claims cannot be made? (p. 350)

Here Plantinga is considering the somewhat more subtle:

Pointed Question 2: Now couldn’t this be argued with equal cogency with respect to any set of beliefs seriously analogous to Christian beliefs, no matter how weird?

Plantinga responds as follows. He registers his suspicion that his strategy could be successfully used to defend some systems of belief other than Christianity; the ones he lists are “Judaism, Islam, some forms of Hinduism, some forms of Buddhism, some forms of American Indian religion” (p. 350).  (One begins to wonder why Christianity, Judaism, and Islam, as opposed to Hinduism and Buddhism, are such that they don’t need the “some forms of” qualification. Could it be that all forms of the former pass muster in this regard?) But he also finds sets of beliefs that he thinks could not be defended in his manner; he writes that his defense won’t work for “voodooism, or the belief that the earth is flat, or Humean skepticism, or philosophical naturalism.” All of these, I take it, are supposed to be “seriously analogous” to Christian belief. (My guess is that the first two, and possibly the third, but not the fourth, are also thought of as examples of “weird” / “bizarre” beliefs.) So, Pointed Question 2, like its predecessor, rightly receives a negative answer.

But while Plantinga is right to sense that the objector needn’t claim that all sets of beliefs are Plantinga-defensible, I think he doesn’t consider the objector’s best option. As I said above, the real worry here is that there are some bizarre/weird aberrations of irrationality that would be Plantinga-defensible — i.e., would be sets of beliefs with respect to which Plantinga’s defensive strategy would work as well as it does in its actual use. Following Plantinga’s own wording (some from the early 80’s, some from now) as closely as possible, this yields:

The Real Pointed Question: Now couldn’t this be argued with equal cogency with respect to some weird/bizzare aberrations of irrationalism?,

where “this” refers to Plantinga’s defensive strategy.

Well, that’s the real pointed question, I think — or at least close to it. My main goal here has been to get it out, so it can be faced squarely. To turn it into an anti-Plantinga argument, we’d end up with something like:

1. There are some possible wildly bizarre/weird aberrations of irrationalism that are Plantinga-defensible (i.e., are such that Plantinga’s defensive strategy against the charge of irrationality would be as successful in defense of them as it is in Plantinga’s hands in defense of Christian belief).

2. Plantinga’s strategy could not be used to successfully defend the wildly bizarre/weird aberrations against the charge of irrationality.

So, 3. Plantinga’s defensive strategy does not provide a successful defense of Christian belief against the charge of irrationality.

As I said at the outset, I don’t take this to be a “killer” objection. But it is worrisome.

Here, I’ve been focusing on aspects of Plantinga’s response that seem to me to take advantage of ways in which the objection/question was being construed in needlessly weak ways. But don’t be misled by this. There are aspects of his response that I think speak to the real problem.

I won’t say much about Plantinga’s answer. But let me close by pointing out one way in which Plantinga’s answer to this type of problem seems to me to have undergone a startling reversal since the early 80’s. In RBG, it’s the charge of irrationality (rather than a lack of warrant) that’s the focus of the objection, and the objector seems to assume that Plantinga will join him in condemning as irrational the “bizarre aberrations”. And as I noted above, Plantinga seems to verify this assumption. To repeat myself: The objector is protesting: “Look, Plantinga, there are some sets of belief (or possible belief) that even you would want to pass negative epistemic judgment on (where “irrational” is the negative judgment stressed here), right?” And I take it Plantinga — at least back in 1983 — is answering: “Right.” There doesn’t seem to be any hint in 1983 of Plantinga responding: “Well, if you look closely, those beliefs really aren’t so bad.” (At least no hint I can find, and I just read through all of RBG on the look-out for such a hint.) Which is why, I take it, we never look closely. It’s pointless. The objector and Plantinga are agreed that there are bizarre aberrations of irrationalism, and the question is. . .

But that was then. This is now. Plantinga won’t simply answer a question about rationality. Instead, as response to Martin shows, he’ll “disambiguate” it into different things that can be meant by “rational”. But — and here comes the startling reversal — in most of the senses of “rational” that Plantinga recognizes, he now claims that Pumpkinites and voodoists are rational in their beliefs, sometimes obviously rational:

Martin doesn’t tell us what he means by ‘rational’….Perhaps the best candidates would be rationality as justification (deontological justification), internal rationality, and rationality in the sense of warrant. We needn’t linger long over rationality as justification: obviously the voodooists could be within their intellectual rights in thinking what they do think (if only by virtue of cognitive malfunction); hence they could be justified… (p. 346)

Well, suppose we specify the argument to internal rationality; take ‘rationally acceptable’ to mean ‘internally rational’. Then again the answer is pretty easy. A belief is internally rational if it is produced by faculties functioning properly ‘downstream from experience’ — if, given your experience (including doxastic experience) at the time in question, it is compatible with proper function that you accept the belief in question. That could certainly be so for voodooists.  (p. 346)

So, on two of three things Plantinga thinks can be meant by calling a belief “rational” — or at least on two of the three “best candidates” — he thinks the bizarre/weird beliefs are rational. (Or at least can be rational. But then that misses the point of the objection, which is that they can be irrational, but Plantinga-defensible.(4)) What’s more, the single sense of “rational” (at least among the best candidates) he can find on which he thinks the bizarre beliefs are not rational, turns out, I think, not to be any good sense of “rational” at all. “Rationality in the sense of warrant”, that is, is no sense of “rationality” at all, according to me. My “BIV twin” — a poor soul who’s had all the same experiences as I’ve had, and who has formed all the same beliefs I have, but who’s external world beliefs are mostly false because he’s a brain-in-a-vat — doesn’t seem to me, in any good sense of “rational”, to be more irrational than me, though many of his beliefs are unwarranted (in Plantinga’s stipulated sense) to a very high degree, where my corresponding beliefs are highly warranted (in that same stipulated sense).

At any rate, the important point here is that Plantinga now takes these weird/bizarre beliefs to be rational, in at least many of the senses of “rational” he recognizes. I see no hint of that in RBG. But I may well be getting something wrong here. For if Plantinga really has so dramatically shifted his position on a matter so important to the GPO, then his response to Martin is puzzling, to say the least. Martin’s objection was aimed at RBG. If Plantinga really has so drastically changed his position, his response should have been along the lines of:

Well, I’ve changed my mind about that. Back then, I agreed with the objector that the Pumpkinites, voodooists, etc., were irrational. I now think I was wrong about that, and that if you look closely, on most of the best senses of “rational”, those absolutely bizarre, wild, weird beliefs are perfectly rational.  In fact, obviously so.  I don’t know what I could have been thinking back then…

But he doesn’t say anything like that. Rather, he writes as if he’s defending the same position he held in RBG. Perhaps Prof. Plantinga can clear this up for me.

At any rate, I greatly enjoyed his book, and recommend it to anyone seriously interested in religious epistemology. I hope that I have here helped to clarify what the nagging Great Pumpkin Objection does — or at least — can amount to.


1. “Reason and Belief in God,” in A. Plantinga and N. Wolterstorff, ed., Faith and Rationality (University of Notre Dame Press, 1983), pp. 16-93. 

2. Warranted Christian Belief, forthcoming, Oxford UP. Page references are to the manuscript copy I received, Chapter 10.
Update 7/29/2005: My page references to WCB are now to the actual, published book, which came out in 2000.

3. See God, Freedom, and Evil (Grand Rapids: Eerdmans, 1974), p. 112. 

4. One possible way to read all this so that Plantinga hasn’t drastically changed his mind is to fasten on the “could”s in the above: Maybe all Plantinga is saying now is that there are some possible forms of these “weird” beliefs that would be rational, while he still agrees with his former self that they’d be irrational as one would typically imagine them. But this would completely miss the point. From the Ur-beginning, the GP Objector’s use of Pumpkinism, Voodooism, etc., has been for them to serve as examples of bizarre aberrations of irrationality, so it’s only the most bizarre and most clearly irrational forms of these beliefs that are relevant. If what Plantinga means here by saying that voodooists “could” be rational is that there are some possible forms of voodooism — perhaps less outlandish than non-voodooists typically imagine voodooism as being — that are or could be rational in the relevant sense, he doesn’t begin to address the GPO. The GP Objector never thought that all possible forms of voodooism would have to be irrational.

Indeed, with regard to voodooism, which has many actual adherents and practitioners, I’ve been taking it that the details of how it’s actually practiced are irrelevant to the current discussion, and what’s relevant is the (bizarre and outlandish) voodooism of the typical non-voodooist’s imagination.

Skip to toolbar