Saturday, October 19, 2013

Is Religious Belief Reasonable?

More specifically: should atheists consider religious belief to be reasonable (for well-informed agents)?  My colleague Kevin Vallier thinks so, aiming "to show that religious beliefs can be held with enough epistemic credence to generate reasons for action that non-believers should take seriously as counting for or against the justification of laws."  I'm not so sure.

Some kind of minimal deism (motivated by the cosmological argument) seems reasonable enough.  But (as I've previously noted) it's a pretty big leap from there to theism, and an even bigger leap to the particular historical and theological claims of religions like Christianity.

Kevin's defense of theism's credibility is rather quick:
It is simply obvious that theism is reasonable to anyone who is acquainted with contemporary philosophy of religion. Nearly all atheists in the literature acknowledge that theistic belief is at least sometimes epistemically justified.

A major worry about appealing to the consensus of philosophers of religion is that this can plausibly be entirely explained by selection effects: it would be odd to specialize in philosophy of religion if one considered it a settled issue.  And if (nearly) the only people who bother to enter the field are those who are antecedently inclined to take religion seriously, their consensus on this point can hardly be taken as evidence that the rest of us ought to do likewise.  So, the crucial question is instead whether further exposure to the philosophy of religion would cause other atheists (who currently judge religious belief to be irrational) to revise their view of the rationality of religious belief.  And I haven't seen any reason to think that's the case.

Moreover, I don't think we need to rely on the indirect evidence of "expert testimony" here, since the first-order arguments are fairly straightforward to assess.  Cosmological and fine-tuning arguments don't support anything stronger than minimal deism.  The ontological argument is obvious sophistry, and the modal version is question-begging.  The design argument was defeated by Darwin; in its place we have the problems of evil and divine silence: the world just doesn't look anything like we'd expect it to if overseen by an omni-max god.  In short, it seems to me, we seem to have every reason to reject theism, and no good reason to accept it.

Things get worse when we bundle in the quirky historical and theological claims of particular world religions.  We know, from the fact that at best all but one of them is false, that holy texts are not in general reliable sources of evidence.  Moreover, the particular contents put forth are often bizarre enough (fitting poorly with everything else we know of the world) to cast yet further doubt on them.  Much of Christian theology (original sin, eternal damnation for honest non-believers, etc.) seems not just bizarre but patently immoral.  It doesn't seem like the kind of thing someone could end up believing as the result of a careful and unbiased assessment of the evidence.

(Am I missing something?  Feel free to comment with a brief explanation, if you think so.)

Of course, even if religious belief is epistemically unwarranted, that doesn't mark it out as warranting any special censure -- we humans are often unreasonable in all sorts of ways, mostly harmless, that aren't worth making a fuss about.  But when religion is used as grounds for promoting a morally misguided social agenda, its lack of justification may become more important to note.


  1. Very briefly, I take it that someone like Kevin might say they have access to evidence which you and I don't and that there is a controversial but not unreasonable epistemology which licenses inferences using that evidence. There are certainly epistemological positions that discount what science tells us and while from my reflective eq those are criticizable, I'm not sure that's the proper standard of reasonableness for what constitutes reasonableness in the political domain.

    1. Hmm, I've previously expressed doubts about "private evidence", and (for the reasons explained in that post) am pretty skeptical of the idea that religious experiences provide significant epistemic support for religious beliefs.

      Perhaps one might reasonably hold that religious belief is (mistaken but) reasonable. (I'm happy to remain neutral on that higher-order question. So I'm not committed to the claim that alternative epistemological positions from mine are unreasonable, as opposed to merely incorrect.) But I think it would take a lot more to show, as Kevin hopes to, that either (i) religious belief really is reasonable, or (ii) that atheists are rationally compelled to believe claim (i).

    2. Isn't there a nice theorem about this something along the lines of:

      If you and I are perfect Bayesians I either have to believe that you are simply lying/misinformed/mislead by your private evidence or our probabilities will both converge to where they would have been had we both had access to the evidence (any disagreement is a disagreement of priors).

      Now surely we aren't perfect Bayesians but the general logic stands. Either I think you are lying (which does nothing to help make your belief reasonable), being affected by a reason altering state similar to certain kinds of drug induced beliefs which should be given no credit (the feeling of religious rapture may be overpowering but that doesn't mean I should accept it as evidence) which makes me even less inclined to call your belief reasonable or I should find your acceptance of said evidence to itself be compelling evidence for your theistic belief.

    3. The theorem is more strict than that. If you are both rational, have common knowledge (this is true if it's given that neither of you is lying, you both state your degrees of belief, repeatedly, and the fact that neither of you is a liar or irrational is common knowledge), and you have common priors, your beliefs will converge. Equivalently, if you have common knowledge then your evidence will converge. It's possible we'll disagree based on one of us thinking a god is more likely a priori, but no amount of private evidence or misinformation will lead to our beliefs differring.

    4. Private evidence only exists insomuch as I think you might be lying. If you have some feeling come over you, and you tell me about this, we both know you had that feeling come over you. There is no reason why we wouldn't both consider it to be equally powerful evidence.

  2. Yeah, I also think, at best, religious belief is mistaken but reasonable and even that a view on which evidence can be private is mistaken but reasonable. It's hard to say whether Kevin needs to say that atheists are compelled to recognize the reasonableness of his beliefs in one sense. Think about what a quasi realist might say about anti realism. There is a sense in which they deny the view is reasonable and a sense in which they don't. Or think about a Bayesian with crazy priors over which epistemic principles are true about what counts as evidence. From my perspective those are bad priors but that's because I've already been updating on my own. Anyway, that's the best case I can make on behalf of Kevin. I've read some of the religious epistemology literature and take it to be as reasonable as various other philosophical views that I'm not willing to dismiss as unreasonable even though I think they are mistaken, like Utilitarianism (I kid).

  3. How about the straightforward argument that there are a great many mainstream religious beliefs whose central tenants are contradictory. No matter what both the atheist and the theist must assign an average probability of no more than 1/n (where n is the number of such mainstream religious beliefs) to such beliefs. Thus, for at least some mainstream religious beliefs, the atheist MUST consider them to be unreasonable. It would strain credibility to claim there is some general argument for the atheist to find *some* mainstream religious beliefs reasonable when we know he must find many (even the majority) of mainstream religious beliefs unreasonable. Indeed, the existence of such an argument, coupled with the probabilistic bound given before would seem to guarantee the existence of an argument that some particular religion has reasonably high probability (substantially greater than say the 1/20 bound on the average probability of a mainstream religion being true).

    But why does this argument seem to have some pull if it is so wrong. I suggest it is a subtle confusion about the meaning of the word reasonable. On the one hand one might use reasonable to simply mean 'Right thinking people who start with sufficiently different prior probabilities might reach that conclusion without adopting any beliefs that render them dysfunctional or unfit to live in modern society.' (replace prior probability with whatever notion of foundational belief you prefer). In this sense sure, aethists are required to believe that religious belief is reasonable, but only reasonable in the same sense that belief in UFO abductions, the illuminati etc.. et.. are reasonable since appropriate priors surely also justify those conclusions and the believers are just as capable of functioning in human society.

    A further confusion exists between the notion of 'many people find persuasive/reasonable' with 'is reasonable to believe.' The wrong answer to the Monty Hall problem is granted credence by a vast number of experts and non-experts alike when first presented but that in no way makes that wrong answer reasonable. No matter how many people are likely to fall victim to a fallacy doesn't make it any less of a fallacy.

    1. In other words move along there is nothing to see here:

      Religious belief is reasonable in exactly the unsurprising ways that everyone agrees it is reasonable, it's the sort of thing that almost all of us can be convinced into believing and many of us have, it doesn't require violating any laws of probability or adopting behavior inappropriate to modern society. Also unsurprisingly, calling someone's religious belief unreasonable is particularly likely to make them resistant to persuasion in ways that calling other philosophical beliefs unreasonable will not.

      Is there anything else here?

    2. Hi Peter, I don't think your "straightforward argument" works, because one might hold the "permissive" epistemic view that a wide range of credences can be reasonable (even given fixed evidence). That is, one could give low credence to a view whilst recognizing that others might reasonably give it much higher credence. So that leaves open the possibility that each religious view is such that it can reasonably be given high credence (and the others low credence).

      The better argument along these lines, I think, is the one I gave in the OP, about how it at least shows that holy texts are not, in general, reliable sources of evidence.

    3. Hence the second half of my post (and my post above).

      Yes, you are right that they can all be reasonable if we expand the definition of reasonable to be something like "Believable by people with appropriate priors." I specifically grant that interpretation but note that it isn't the intent of the argument because religion comes out as no more reasonable than belief in UFOs.

      The fact that we all have different evidence is a red-herring. I either believe your reports about the evidence you have, in which case I take that evidence into account myself and end up giving your particular religious belief high credence, or I don't accept your reports about your evidence in which case those reports surely can't be reasons for me to think your view is reasonable. In other words it does all come down to priors.

      I imagine the case you are thinking of is what if I know such and such a person has always gotten whatever they prayed for. However, this doesn't make their belief any more reasonable since they know this isn't true for others so they are left evaluating the same information I am "Does the fact that a person who acts way X, Y and Z has always had their prayers answered make it likely that religious belief is true," and this is evaluated independently of who we both are.

  4. I suspect there is also the question of whether religious belief is reasonable in the sense of whether it should be subject to ridicule and derision the way we do with belief in a flat earth or the like. Here, for purely practical reasons, one would say religious belief is reasonable in the general cultural context since it would be counterproductive to ridicule a majority view (unlike the deterrant effect it can have against a minority view) is often indoctrinated at a very young age and is particularly resistant to change etc..

    Again, yes there are meanings of reasonable on which religious belief should be accepted as reasonable by the atheist but once unpacked these notions of reasonable turn out to be totally unsurprising.

    Since you are the one bringing up the subject the burden is on you to identify a sense of the word reasonable on which the idea that atheists should accept religious beliefs as reasonable is, at the very least, not immediately absurd or ruled out by straightforward probabilistic arguments yet still retains some interesting consequences.

    Absent an identification of such a sense of reasonable, e.g., if you simply say I mean reasonable the way we all mean reasonable, the best I can do is shrug and say I don't see the claim/issue here.

  5. I think you're spot-on with these observations, which are worth re-hashing every now and again to remind everyone of just how enormous the gulf is between deism and any particular religious system. I wonder, though, what you think of the Pascal/James line, which is (broadly speaking) pragmatic. Pascal's argument is probably no good, but James' has always seemed less bad to me.

    1. This comment has been removed by the author.

    2. Yes, I basically agree with Dave here.


  6. Clayton Littlejohn is known for denying one can have a justified false belief. Yet even he in his strict forbidding still admits a false belief can be "excusable" and even "reasonable."

    1. Of course, I'm not suggesting that religious belief is unreasonable because false. It's unreasonable because it's obviously false. (Some, but not all, false beliefs can be reasonable -- namely, the non-obviously false ones.)

    2. You'll have to give us more than just an assertion of falsehood. How is religious belief obviously false? I could say the same about Atheism or Physicalism, but I would have to provide more "butter" to my statement.

  7. "Things get worse when we bundle in the quirky historical and theological claims of particular world religions. We know, from the fact that at best all but one of them is false, that holy texts are not in general reliable sources of evidence". What makes you think that? What reason is there to think that only one religion can be true? The claims of one religion, say, Vaishnavism, do not contradict anything which Sikhism essentially holds to be true. Another thing, we can use history to prove that the events narrated in the various religious texts are historical, which would make them reliable sources of evidence.

  8. Thanks for the interesting post!

    Consider the following: A person S has enough evidence to satisfy the justification condition for knowledge with respect to a basic belief B iff among all the propositions that S understands, the propositional content of B best explains why S's non-propositional experiences occur in a particular way and at a particular time (rather than in other ways and at other times). S's having adequate evidence for the truth of her belief B does not entail her being able to articulate her epistemic reasons to anyone else. And even if S describes her experiences (i.e. the ones that justify her belief B) to other people, such testimony does not automatically provide those other people with enough evidence to justify their (potential) belief in the propositional content of B.

    1. Kind of like Nozick's theory of EJ.

    2. Hi Ananta, I thought Nozick held to an externalist account of warrant? S knows P iff P is true, S believes P, and S's belief that P "tracks the truth" of P.

      The internalist-foundationalist account of justification that I briefly summarized above comes from Paul Moser's Knowledge and Evidence (1989). This is relevant to Richard's post because he doubts that individuals can have justified beliefs based on private evidence. If individuals can have justified beliefs based on private evidence, however, then how can he be justified in believing that NO well-informed agents have justified religious (or theistic) beliefs? In order to know or be justified in believing that, he would need to be justified in believing that no well-informed agents have adequate private evidence for the existence of God. But he does not have access to the private evidence base of ALL well-informed agents!

  9. [Note: Richard just gave me permission to post this extra-long comment in 2 pieces. Since I wrote this, I see there are some similar points made by Ananta Vasudev.]

    I am a Christian. In order for your claim to be true, you need to be able to show that I am unreasonable or else not well-informed. (You don't necessarily need to convince me, but you need to convince yourself that you ought to have convinced me!)

    1. Regarding your "first order" philosophical arguments for theism, I mostly agree, except:

    (a) saying that the Design Argument was defeated by Darwin presupposes that there is only one design argument, based on the origin and adaptation of species. But there are others (for example, the fine-tuning argument you mention belongs to the Design category), and I think the one refuted
    by Darwin wasn't even the most important one historically.

    (b) I agree that the Argument from Evil has significant force, but I don't think it clinches the case. One has to show that God could have no good reason for allowing evil. This is hard to show, and at least some defences seem vaguely plausible (e.g. the idea that suffering builds character is in accordance with common sense).

    (c) Although you point out that the philosophical arguments here only support minimal deism, they raise the probability of stronger religious claims as well. Also note that they support monotheism over polytheism. This already acts as a filter on the large number of religious claims out there.

  10. 2. Regarding the "quirky historical and theological claims of particular world religions", note that it is precisely these quirky historical claims which allow for empirical testing through historical research. In other words, you need to engage in an actual historical argument, in order to assess whether the various claims of divine relevation (e.g. eyewitness accounts of ancient or modern miracles, fulfilled prophecies, visions etc.) are or are not plausible. You don't even begin to analyze this in your post, but in my view this is exactly where the solidly convincing evidence is located.

    (a) "We know, from the fact that at best all but one of them is false, that holy texts are not in general reliable sources of evidence." This is way too superficial. First of all, it does not actually follow from one religion being true that the rest are all false. For example, if Christianity is true, than Judaism is also true. Christianity is also largely compatible with the claims of monotheistic philosophers such as Plato. We are opposed to polytheism and idolatry, and we think that there have been some false prophets, but that does not mean that everything in a non-Christian religious book is 100% false.

    (b) Furthermore, the category "holy text" lumps together lots of different eclectic material (ethics, philosophy, etc.) not all of which are even obviously claims to divine revelation (for example, the post by Kevin Vallier inaccurately includes Confucianism in that category). If we restrict to claims of explicit divine revelation taking place in historical times, the field of religions is narrowed down considerably, and if we eliminate obvious impostures it is even narrower.

    (c) Also, it happens all the time (in science, for example) that of several competing views, one is true and the rest are false. The mere fact that there are competing claims is weak sauce. If you want to convincingly support your broad thesis, you need to actually sift the relevant historical evidence. I could go into more detail here about how *I* sift the evidence, but you asked for responses to be brief, so for now I will just say that your argument here is barely an argument at all.

    3. I admit that Christianity is weird, but then again so are some other things which are supported by empirical evidence (e.g. quantum mechanics).

    I deny that Christianity is patently immoral. For example, I don't think that "eternal damnation for honest non-believers" is taught anywhere in the New Testament, but rather the opposite (see John 3:19-20, 15:22-24, Matt. 12:32, Acts 17:30, Romans 2:9-16). Similarly, I don't believe in a version of "original sin" which would imply that we are literally guilty for something our ancestors did. On the other hand, it is common sense to say that the choices our ancestors made have a moral influence on us. And it is also pretty clear that nearly all of us fall very short of what ethics says we ought to do.

    In order to argue that Christianity is intrinsically absurd, you need to argue against whichever version of Christian theology is the most reasonable. That may require looking beyond the versions which have been presented to you, to figure out what a more reasonable Christian *would* have said.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.