Georges Rey, Philosophy, Univ of Maryland, College Park 20742; georey@carnap.umd.edu



(long version) Thurs 16 Nov 2000



to appear in Acta Analytica



Meta-atheism:

Reasons to Think that Few People Actually Believe in God (1)



I'm not a professional philosopher of religion and have no special knowledge of theology. However, I regularly teach an introductory course in philosophy in which I discuss the standard arguments for the existence of God. The exercise has produced in me a certain incredulity: I have come increasingly to wonder how such extremely smart people, like Aquinas or Descartes, could advance such patently bad arguments, as I think most philosophers (even those who claim to "believe") would take those arguments to be. At any rate, I find it hard to believe that anyone really buys the "ontological argument," or any of Aquinas' "five ways." Existence may or may not be a predicate, and there may or may not be unmoved movers, uncaused causers, and undesigned teleological systems, but these arguments don't remotely establish their intended conclusion, the existence of anything like the traditional (omniscient, omnipotent, omnibenevolent) Christian God. And "religious experiences," no matter how ecstatic or profound, could obviously be explained by any number of other far more modest hypotheses. (2) I began to wonder whether the arguments were ever really intended seriously; and this led me to wonder whether anyone actually believed their conclusion. That is, I began to wonder whether anyone really did believe in God.

Well, obviously lots of people claim to, and seem to live and sometimes die for such beliefs. It's certainly risky for me to second-guess them on that score just because of some bad arguments --after all, don't people know what they themselves believe, and ordinarily believe what they sincerely avow?

Maybe not. People seem to be susceptible to all manner of confusion and often deeply motivated distortions of their own psychological lives. (3) For starters, there are the formidable difficulties of expressing oneself clearly in language, of saying -and thinking!- exactly "what one means." And, related to that, there is the familiar phenomenon of adjusting what one says --and thinks-- in the light of the demands and expectations of one's audience (think of the "different ways you express yourself" on a topic to a friend, a foe, an interviewer, a student, a judge in court: which of these is one's "real" belief?). But there's also the phenomenon of self-deception: people often claim to believe things that they merely want or are committed to believing, even though "at some level" they know the belief is false. There are the standard examples of someone ignoring the symptoms they have of some dread disease, or the obvious evidence of the infidelities of their spouse; or doting parents exaggerating, even to themselves, the talents of their child. In most of these cases it is because we have reason to suppose that the people involved are otherwise quite intelligent enough to draw the conclusions that they surprisingly resist that we suppose there must be something else at work.

This is my hunch about what passes as "religious belief" (although I expect the issues about expression and intended audience may also play a role). And so I find myself taking seriously the following hypothesis, which (for lack of a better name) I call meta-atheism:



Despite appearances, not many people -particularly, not many adults who've been exposed to standard Western science- seriously believe in God; most of those who sincerely claim to do so are self-deceived.



Notice that, strictly speaking, meta-atheism doesn't entail atheism: it's a view not about God and whether He exists, but about whether people actually believe that He does (hence the prefix `meta-' which, in philosophy, has come to mean, roughly, `at a second-order level about...'). Even someone who did take themselves to be a serious theist might find this thesis interesting, if only for the light it sheds upon the difficulty (sometimes noted by the devout) of actually believing. But, of course, my own interest in the view is in fact motivated by what seems to me the overwhelming obviousness of atheism. I'm afraid that I really don't think the question of the existence of God is much more "open" than the question of the existence of elves or ghosts.

I should say right away roughly what I shall mean by `God.' I'm most familiar with Christian conceptions, and in the short space here will focus upon them, although I expect much of what I say could be applied to others. What seems to me essential to most conceptions, and is what bothers atheists, is that God is a psychological being, i.e., a being capable of some or other mental state, such as knowing, caring, loving, disapproving. What the theist usually asserts that the atheist denies is that there is some psychological being who is not subject to the known laws of the physical universe, knows about our lives, cares about the good, and either created the physical world or can intervene in it. And, at least in Christianity, He is in charge of a person's whereabouts in an afterlife. If you think of God as something other than a psychological being of this sort, or that talk of God is simply a metaphorical or "symbolic" way of talking about love, the possibility of goodness, or the Big Bang, then much of what I say may not apply.

I want to emphasize that I am not intending what I say here to be critical of any religious practices (meditating, attending church) or the serious good that religions sometimes (although not always) do. I am concerned only with the content of what many religious people say they believe. This is because the more seriously and carefully I think about what they actually say, the more utterly bizarre and ludicrous I find it. It seems, quite frankly, mad. At any rate, beliefs that there are invisible psychological agents, with larger than life powers, with whom one is some special "super-natural" communication, who love, scold, disapprove, command, forgive -think about it: these are the sorts of beliefs that, in any other, non-religious context, are associated with patently schizophrenic delusions!

Now, of course, I don't think that most religious people are in fact schizophrenic. Nor do I think all religious people are being insincere. Rather, the meta-atheism I want to defend is the view that many people who sincerely claim to believe in God are self-deceived, which, as some of the other cases I've already mentioned show, can be entirely "normal," and even morally benign (nothing like a little self-deception to keep an otherwise querulous family together!) My view is, of course, a kind of extension of the familiar observation that most religious stories involve patent wishful thinking and rehearsal of childhood dramas with one's parents. But I would also want to include other influences, for example, loyalty to one's family or other social groups, powerful commitments and identifications, or simple resistence to changing significant public stances and avowals.

A verbal issue: along lines of my (1988), I'm inclined to describe the results of self-deception as not genuine belief, but, rather, as things people sincerely "think" or "avow" whether or not they actually believe them. But nothing turns on this. I can well imagine someone thinking of self-deceptive beliefs as genuine beliefs, and as simply manifesting ways in which people's beliefs can be bizarrely irrational and compartmentalized. What concerns me is not the label, but the psychological structure: all I want to claim is that for most contemporary adults in our culture, there is some level at which they know very well the religious stories are false, even if they manage to get themselves to "believe", avow, defend and even die for them on the surface.

I don't pretend for a moment to be able to establish meta-atheism conclusively. I certainly recognize that there's a lot to be said that would appear to argue against it. Much depends upon having a much clearer understanding of such really quite complex states and processes of "belief", "avowal" and "self-deception" than I think anyone yet has, an understanding that depends upon a much subtler empirical psychology than is presently available (cf. my (1997:32-4,187-9)). I fully expect that the full story in the area will allow for a wide variety of different sorts of "belief." All that I really hope to do here is to put my hypothesis in the running, calling attention to a number of striking peculiarities of religious thinking that I think it may help explain. Indeed, it's really these peculiarities that interest me most. There are (at my last count) nine:



(1) Tolerance of Otherwise Delusional Claims: I don't think you need to be an atheist to have the reaction I've mentioned to the content of religious claims. I submit that, were the claims about a supernatural entity who loves, commands, scolds, forgives, etc., to be encountered in a fashion removed from the rich, "respectable" aesthetic and cultural traditions in which they are standardly presented, they would be widely regarded as psychotic. Think, for instance, of how most normal, even religious people take the claims of various cults, in which some charismatic figure convinces people that he is the voice of God and convinces them to renounce their present lives and follow him in various peculiar practices (think of the Koresh cult in Texas, or the recent "Heaven's Gate" cult surrounding the Hale-Bopp comet) and then remember that many religions -notably Christianity-- were themselves once just such "cults" (see Pagels 1989). Or, to pass on to related doctrines, think about what you would make of someone -again in any other context- who said they could really change wine into blood, or bread into flesh, or who thought that some kind of justice or other good would be realized by having a perfectly innocent person die for the sins of everyone else (imagine a judge in a local court deciding that, because he so loves the guilty defendant, he will conceive and then sacrifice a son to atone for the crimes!). In any non-religious context, such proposals would, I submit, be regarded as sheer lunacy.

A caveat: (4) Many people who should know better are prey to such (as we call them) "superstitions": knocking on wood after boasting, wearing the socks one wore in hitting the home run, worrying about the next air flight because one has flown so many times so far without accident. At least for many of us, what's peculiar about these "beliefs" is that they persist at some level despite our seriously disavowing them -and despite their failure to be integrated into the rest of our thought (they display precisely the same sort of detail resistence I noted about religious claims). They might be called "ossified beliefs": thin, isolated beliefs that have become rigid and aren't removed by rational reflection. Other examples might be the "neurotic" beliefs Freudians ascribe to us on the basis of irrational behavior -e.g. regarding murderous fathers and castrating mothers.

I wouldn't be surprised if some religious beliefs are also of this ossified sort (when theists suspect atheists are themselves self-deceptive about their atheism, it may well be such ossified beliefs they have in mind). Of course, many religions are at pains to distinguish belief in God from "mere superstition." Religious belief becomes self-deceptive, on my view, when the belief is not merely noted, in the detached way that one notes one's superstitions, but wholly endorsed, regarding it as somehow something more than mere superstition. It's this further attempt to integrate religious beliefs into, as it were, serious belief that strikes me, for the reasons I have given, as going against what most people with a high school education know very well to be true.



(2) Reliance on Texts: Many of these otherwise outlandish religious claims derive an air of legitimacy, of course, from their reliance on a specific set of usually archaic texts, whose claims are presented dogmatically (indeed, the primary meaning of `dogma' has precisely to do with religious proclamations). The texts standardly serves as the sole basis for various claims that are regarded as essentially uncontestable -certainly not contested on the basis of any non-textual evidence. As many have noted (e.g Wittgenstein 1966), they are not presented as hypotheses, to be either confirmed or disconfirmed by further research. They are usually adopted or renounced not on the basis of serious evidence, but as a matter of "faith" or "conversion" (see also #(9) below).

By contrast, there are no such texts or creeds for common sense or science. In general, we know very well that truths about the world are largely not revealed by the contents of some text in the way that religious claims often are, and so, even though people of course rely on the texts for the transmission of information, the texts are every few years being challenged, revised, updated and eventually replaced. Contrary to popular ideas, there is no set of claims, much less any single text, to which all scientists need subscribe to be good scientists: indeed, as recent science shows, there is no claim so sacrosanct that some good scientist (or scientifically minded philosopher) might not challenge it. To use the nice metaphor of "Neurath's boat" emphasized by Quine, in both science and commonsense we are like mariners on the open sea who have to repair their boat while remaining afloat in it, standing now on one plank to repair a second, only to stand on the second to repair the first.



(3) Detail Resistence: This continual mutual revision and adjustment of ordinary beliefs is related to the multifarious ways in which they are interconnected, any one of them having logical or evidential relations to indefinite numbers of the others. For example, beliefs about whether O.J. Simpson murdered Nicole are connected to beliefs about cars, freeways, airports, police, DNA -which in turn connects them to beliefs about cities, governments, history and cosmology. And one expects there to be in this way indefinite numbers of details that could be filled out in regard to these connections. If doubts are raised about the details, they can rebound to any one of the connected beliefs: thus, evidence against a particular theory of DNA would have given jurors reason to doubt that OJ was at the scene of the crime. And if someone were to suggest that there was some third party that murdered Nicole, then one would expect there to be further details -e.g., further fingerprints, DNA- that would serve as crucial evidence. If there were no such details, one would be reasonably sceptical: absence of evidence, especially when you look for it, becomes evidence of absence, and most people know this, which is one good reason they have for not believing in fairies or little green men controlling us from Mars.

By contrast, literally understood, religious claims are oddly detail-resistant. Perhaps the most dramatic cases are the claims about creation. Whereas scientists regularly ask about the details of the "Big Bang" --there is an entire book, for example, about what happened in the first three minutes (see Weinberg 1977)-- it seems perfectly silly to inquire into similar details of just how God did it. Just how did his saying, "Let there be light," actually bring about light? How did He "say" anything at all (does He have a tongue)? Or, if He merely "designed" the world or the species in it, how did He do this (are there blue-prints of the individual particles/ animals)? Was it just the quarks, the DNA, or the whole body? Or just some general directives that were executed by some angelic contractors? At what specific point does He --could He possibly-- intervene in the natural course of events without causing utter havoc? Does anyone really think there is some set of truths answering these questions? Perhaps; but it is striking how there is nothing like the systematic research on them, in anything like the way that there is massive, on-going systematic research into the indefinitely subtle details of biology and cosmology. As Kitcher (1982:ch 5) points out, even so-called "Creation Science" is concerned only with resisting evolutionary biology, not with seriously investigating any of the massive details that would be required for the Creation story actually to be confirmed (imagine there being careful investigation of radio-isotopes, sedimentary layers and the fossil record to establish precisely how, when and where God created atoms and compounds, as well as the full array of biological species.)

Of course, theologians do discuss details. As I mentioned, I'm not a scholar of theology, however, I'm willing to wager that few of the details they discuss are of the evidential sort that we ordinarily expect of ordinary claims about the world, i.e. claims that link the theological to crucial data that would be better explained by the theological than by any competing hypothesis. Elaborations of the theological stories without this property -mere stories about "angels on the head of a pin"- do not constitute such details: they are merely like the elaboration of a fictional story. If there really are serious attempts to narrow down the details of God's activities by, e.g., reference to the fossil record, or systematic studies of the effects of prayer, then I stand corrected. But I'd also wager that most "believers" would find such efforts silly, perhaps even "sacreligious."

Some of this resistence to detail could, of course, be attributed to intellectual sloth. But not all of it. After all, if the religious stories really were true, an incredible lot would depend upon getting the details right (if you believe the wrong story, you could risk winding up in hell forever!). However, when I ask "believers" these kinds of questions of detail, I am invariably met with incredulity that I even think they're relevant.

Indeed, I find there are three standard reactions: either the claims are not to be understood literally (in which case, fine: they are not literally believed); or they appeal to "mystery" (to which I will return shortly); but more often they simply giggle or make some other indication that I can't possibly be asking these questions seriously. The questions are regarded as somehow inappropriate. I have never encountered the kind of response that would be elicited by questions about how , e.g., O.J. got to the airport in time, or about how just how big the Bang was. To these latter questions, people will, of course, usually find the question relevant, and maybe even interesting. They might not know the answer, and perhaps not particularly care to find out: but they appreciate its pertinence and assume there is some intelligible way of finding out -and that, if there's not, or the answer came out wrong, then that would be a reason to doubt the purported event actually occurred.



(4) Similarity to Fiction

Indeed, this resistence to detail is strikingly similar to the same resistence one encounters in dealing with fiction. It seems as silly to ask the kind of detailed questions about God as it does for someone to ask for details about fictional characters, e.g.: What did Hamlet have for breakfast? Just how did Dorothy and Toto make it over the rainbow? These questions are obviously silly and have no real answers -the text pretty much exhausts what can be said about the characters. So, in keeping with the reliance on texts and appeals to non-literality that we've already noted, perhaps religious claims are simply understood to be fiction from the start.



(5) Merely Symbolic Status of the Stories: Indeed, notice that much of the power of religious claims doesn't really consist in their literal truth. Take for example the tremendously moving story of the betrayal and crucifixion of Jesus, and ask yourself whether, were we actually trying to achieve justice in the world, his "dying for our sins" would be even remotely appropriate. In the first place, as I mentioned, the idea of an innocent person being sacrificed to expiate someone else's sins is really pretty wild.

But, secondly, supposing that this kind of proxy atonement did make sense, the question should certainly arise in the specific case of Jesus whether He the actually did suffer enough! I don't mean to say that His betrayal and crucifixion weren't pretty awful; but can they really "balance" all the horror of Ghenghis Kahn, WWII, the Gulag, or what death squads routinely do to their victims in Latin America? These are crucifixions multiplied tens of millions fold. -But, of course, all of this is less relevant if we are to take the passion story as merely symbolic fiction, i.e. not as actually a rectifying of wrongs. Mere symbols, after all, needn't share the magnitudes of what they symbolize.



(6) Appeals to "Mystery": Confronted with many of the above oddities, many theists claim God is a "mystery" --indeed, I once heard a famous convert, Malcolm Muggeridge, claim "mystery" as his reason for believing! But ignorance (=mystery!) is standardly a reason to not believe something. Imagine the police arresting you merely because it's a "mystery" how you could have murdered Smith! Just so: if it's really a complete mystery how God designed or created the world, then obviously that's a reason to suspect it's simply not true that He did -and, my point is that this is sufficiently obvious that everyone knows it, and simply pretends that religion affords some very odd exception.

It is often claimed that believers are willing to tolerate the mysteries surrounding God because they have an additional belief, viz., that they also can't know about God's ways. Now, first of all, this is disingenuous. In the first place, there are all the things claimed to be known about God and his ways in the Bible. Moreover, many people believe that He's responsible for when people live and die, suppose Him to be responsive to petitionary prayer. But, of course, these then are precisely the points at which the God hypothesis is vulnerable to obvious disconfirmation: too much happens that's hard to believe is the result of a benign omnipotent being; too little happens that's an answer to prayer.

Of course, people do tolerate plenty of mysteries about how the world works. Most people have only the dimmest idea about how things live and grow or how it that thoughts are related to action. But in these cases the evidence for the postulated processes is of course overwhelming and --apart from some highly theoretic (and unsuccessful) suggestions of behaviorists- completely uncontroversial: ordinary people haven't the slightest reason to doubt that things grow, or that thought causes action despite the mystery about how it occurs. By contrast, anyone aware of the basic ideas of contemporary science and the conspicuous lack of evidence of a God have plenty of reason to doubt the existence His existence. In such a case, mystery can be no refuge.

Moreover, what's peculiar about the belief about our supposed inability to know about God's ways is that the inability is so arbitrarily and inexplicably strong: why should there be no normal evidence of his existence? Why shouldn't it establishable in the same way as the existence of bacteria or the Big Bang? In any case, it's not as though the religious try to do what they might do in these other cases, namely, think of clever, indirect ways of finding out. No, the "mystery" is supposed to be "deeper" and far more impenetrable than that. I can't imagine what sustains such conviction -mind you, not about God, but about the knowability of God's ways- except perhaps an unconscious realization that there of course couldn't ever be serious evidence for something that doesn't actually exist.



(7) Betrayal by Reactions and Behavior: People's reactions and behavior --for example, grief, mourning at a friend's death-- do not seem seriously affected by the claimed prospects of a Hereafter. Contrast the reactions in two situations of a loving, "believing" couple who are each seriously ill: in the first, the wife has to be sent off to a luxurious resort for care for two years before the husband can come and join her for an indefinite time thereafter; in the second, the wife is about to die, and the husband has been told he will follow in two years. If, in the second case, there really were the genuine belief in a heavenly afterlife that (let us suppose) they both avow, why shouldn't the husband feel as glad as in the first case -indeed, even gladder, given the prospect of eternal bliss!? However, I bet he'd grieve and mourn "the loss" like anyone else. Note in this connection how even the most religious observances with respect to death are deeply lugubrious, and imagine the absurdity of performing a requiem mass on behalf of someone you won't see for a few years because she has gone to a luxurious resort!

Or consider petitionary prayer: why aren't people who believe in it disposed to have the National Institute of Health do a controlled study (say, of the different sorts of prayer) as they would were they interested in the claim whether soy beans cure cancer (5) And, in any case, why do none of them expect prayer to cure wooden legs? Or bring back Aunt Martha after twenty-five years? I suggest that there are obvious limits to people's self-deception, and they know full well that God couldn't really intervene in that preposterous a way.



(8) Patent Sophistry: Purported theists do not take the familiar arguments of atheists lying down. There is a small industry of efforts in certain philosophical circles to show that religious belief is no worse off than our beliefs about the existence of the external world or other minds. Most conspicuously, Alvin Plantinga (2000) tries to assimilate religious belief to the kinds of "basic beliefs" that "foundationalist" epistemologists claim are required for us to have knowledge of anything, e.g. beliefs about the reliability of memory, sense perception, basic arithmetic.

I think it's a serious mistake to think that discussions about theism really turn on difficult philosophical issues regarding the ultimate justification of belief. G.E. Moore observed that it's a requirement on any credible theory of knowledge that it not deny that normal people know they have two hands. As a corollary of that observation, it should also not tolerate the delusions of schizophrenics. Any defense of religious knowledge would need to present a theory that meets both of these demands, and, for some of the reasons I've sketched, it is difficult to see how it could do so, whether or not it's "foundational." In particular, the question is not whether there are or aren't "basic" or foundational beliefs, but why on earth anyone should think that belief in the existence of anything with the extravagant implications of God should figure among them; or, even if it does, why the utter failure of any of these implications to be independently confirmed wouldn't be an overwhelming reason to scotch the belief, basic or otherwise. Beliefs acquired by unassisted vision, be they ever so basic, are soon undermined by noticing you're not seeing things smack in front of you -or "seeing" things that are patently not there. In any case, there obviously needs to be independent evidence for the reliability of any perceptual or cognitive process if its testimony on a controversial topic is to be believed, and it's just such evidence that seems patently unavailable in the case of theistic claims. And most everyone knows this.



(9) Appeals to "faith": Indeed, most religious people readily recognize the failure of evidence but then go on to claim that religious beliefs are matters of "faith," not evidence (in an extreme case, like that of Tertullian or Kirkegaard, claiming to believe precisely "because it is absurd"). But try thinking something of the form:



I believe that p, however I don't have adequate evidence or reasons for believing it.



where you substitute for `p' some non-religious claim, e.g. "It's raining now in Vienna," "2+2=37," "the number of stars is even," "Columbus sailed in 1962." Imagine how baffling it would be to if someone claimed merely to "have faith" about these things! As Adler (1999) points out, there seems to be something "impossible," even "conceptually incoherent" about it, a little like the incoherence of thinking you know something, but being nevertheless convinced it isn't true.

On the other hand, issues of faith do arise precisely in those cases in which a person is asked to manifest their loyalty to a person or cause despite the evidence that might otherwise undermine it: thus, a father has faith in his son's honesty despite what the police say, or someone remains "true" to a political cause in the face of evidence of corruption. But, of course, such cases are precisely ones that lay the ground for the kind of self-deception that I have been arguing is characteristic of contemporary religious claims.



References

Adler, J. (1990), "Conservatism and Tacit Confirmation," Mind XCIX, pp559-70

Adler, J. (1999), "The Ethics of Belief: Off the Wrong Track," Midwest Studies 23:

New Directions in Philosophy, ed. P. French and H. Wettstein, pp267-85

Bach, K. (1981), "An Analysis of Self-deception," Philosophy and Phenomeno

logical Research, 41 (March), pp. 351-370.

Kitcher, P. (1982) Abusing Science: the Case Against Creationism, Cambridge: MIT

Martin, R. and Kolak, D. (2001), Wisdom Without Answers: a Brief Introduction to Philosophy,

Belmont CA: Wadsworth

McLaughlin, B. and Rorty, A. (1988), Perspectives on Self-Deception, Berkeley: Univ of Calif

Nisbett, R. and Wilson, T. (1977), "On Telling More Than We Can Know,"

Psychological Review, 84#3:231-59

Pagels, E. (1969), Adam, Eve, and the Serpent, New York: Vintage

Plantinga, A. (2000). Warranted Christian Belief. New York: Oxford University Press.

Rey, G. (1988), "Towards a Computational Account of Akrasia and Self-Deception,"

in McLaughlin and A. Rorty (1988), pp264-96

Rey, G. (1997), Contemporary Philosophy of Mind: a Contentiously Classical

Approach, Oxford: Blackwell

Sober, E. (2001), Core Questions in Philosophy, 3rd ed, New York: Macmillan

Weinberg, S. (1977), The First Three Minutes: a Modern View of the Origin of the

Universe, New York: Basic Books

Weinberg, S. (1999), "Is the Universe Designed?", New York Review of Books, XLVI, 16,

(Oct 21), pp46-8

Wittgenstein, L. (1966), Lectures and Conversations on Religious Belief. Berkeley:

University of California

1. . This is a long verison of a short piece on the same topic that will appear in Martin and Kolak (2001), with a commentary by Christopher Bernard and my replies to that commentary (most of which I've integrated here). The impetus to the central idea arose from some stray remarks some years ago of two of my teachers, Rogers Albritton and Hilary Putnam, although I doubt either of them would recognize (much less endorse) what I have made of them. I'm grateful to Bernard and to Michael Slote for discussions and comments, and to Profs Martin and Kolak for the encouragement to write up my views, and their permission to reproduce the material elsewhere.

2. . For perfectly adequate refutations of the standard arguments, see most any introductory text, e.g. Sober (2001:pt II). The simplest thing to notice about most of the arguments is that they don't establish the existence of a psychological being of any sort (why should the uncaused causer have a mind?), much less of a unique one with the hyperbolic properties in question. And, of course, "religious experiences" by themselves can't establish much of anything, any more than dreams of ghosts do: what would need to be shown is that God --or ghosts-- would be the best explanation of those experiences, but this no one has even seriously begun to do. The only argument that comes close to being plausible is the argument from biological design, but it can't be taken seriously since Darwin (for an excellent recent discussion of "global design" see Weinberg 1999). --Of course, I haven't kept up with all the very latest versions of the traditional arguments; but nor have I followed all the latest proposals for a perpetual motion machine (which, on many versions of theism, God seems to be!).

3. . For discussion of this point see my contribution to McLaughlin and Rorty (1988), which enlarges a discussion well under way in Stich (1983)and Bach (1981). Some of that discussion was influenced by the classic paper of Nisbett and Wilson (1977) in which they adduce considerable evidence that much purported "introspection" of one's own psychological states is often merely the imposition upon oneself of the same kind of (often mistaken) popular psychology that we use to understand others.

4. . Due to an objection of Chris Bernard, who argued that people's beliefs can't be inferred from their behavior, since even people who don't (seem to) believe in ghosts prefer not to sleep in graveyards.

5. . Note that they could do this without disrespectfully praying with such a test in mind: they could simply do demographic studies of the incidence of cures with different religious sects.

I put aside for the nonce the extreme peculiarities of "belief" in petitionary prayer at all --which suggests that people believe that an omnipotent, omnibenevolent God would just as soon, e.g., let a young child die slowly from an excrutiating cancer except that He's heard your prayers!?