Friday, August 22, 2008

A quick note on Plantinga's religious epistemology

Inspired a bit by some discussions taking place over at Prosblogion, I thought I'd briefly discuss a problem I have with Plantinga's religious epistemology. Some of you might know that there's about nothing that P and I agree on, so I thought I'd start by stating that I'm actually far more sympathetic to his claim we cannot properly determine whether belief in God is warranted, warranted in the way properly basic beliefs are, or rational unless we can offer reasons for believing God doesn't exist. That's because I'm quite willing to accept both a sort of externalism on which one can know non-inferentially that something is so relying on a source one has no (internal, if you like) reason to take to be a reliable source and thus far nothing we've said rules out this: that on the hypothesis that theism is true, it would be quite likely that God designed us in such a way that we could come to have warranted beliefs about God (either basically, by means of something like sensus divinitatus, by relying on testimony, the occasional sky writing, etc...). Having granted that much, it seems Plantinga can then go one step further. Atheists and agnostics might be rational in their beliefs, but that we can rationally accept arguments from evil gives us no reason to be unfriendly atheists. For the rationality of religious belief, you see, depends on the totality of an individual's evidence or grounds. It may be that the theist has grounds the atheists and agnostics don't.

This clever move I'll refer to as Plantinga's Pincer movement. I won't call it this for any good reason except that I needed a cool sounding word that began with 'p', and that's what came to mind. Here's my worry. I haven't seen any careful argument for the following claim:
(*) On the hypothesis that theism is true, it is quite likely that some of us would be created in such a way that we could have warranted beliefs about God.

I'd agree that if the conditional probability of our having warranted beliefs on the hypothesis of theism were high, Plantinga's Pincer movement would be quite impressive. Not as impressive as Hannibal at the Battle of Cannae, mind you, but impressive. However, I don't see that Plantinga has given good reason for thinking that this probability is high. I think the probability is either low or inscrutable. His argument that it would be high in WCB is, basically, that God loves us and created us in God's image. The problem is that I see no inconsistency in saying that God loves us, created us in God's image, but equipped us with little by way of brains. At least, that strikes me as no more inconsistent than saying that God loves us, created us in God's image, but some of us are like Dahmer, some of us are swept away to sea during a tsunami, some of us are left to die slowly in burning buildings, to fend for ourselves when cornered by machete wielding mobs, among the millions drowned when God called for a 'Do Over' and sent the floods, etc... Evil can do double duty. It seems to constitute evidence against God's existence. It also seems to constitute evidence that (*) is false or, at the very least, not something we have reason to accept. And, by Plantinga's rules, it seems that if the conditional probability that you are reliable on some matter is either low or inscrutable given your evidence and grounds, you'd be irrational to continue to hold beliefs on the subject. If it's (*) that's keeping the epistemic wolves at bay for the mildly reflective theist, I don't think Plantinga's Pincer movement is quite as useful as some might have thought. But, it's better than maybe you thought I'd think it was.

One footnote. I've seen an occasional reference to Plantinga's Two Dozen or So arguments for theism (here) and I did a quick poll in CO. 50% of respondents said that much of this is either tongue in cheek or just his playing devil's advocate and 50% said he was being dead serious. Does anyone know?


Timmo said...

There's a more detailed scriptural case for what Plantinga calls the Aquinas-Calvin model, i.e. taking theistic beliefs to be produced in the believer by a certain cognitive faculty, a sensus divinitatis, than you suggest in your post. You might check out Romans 1:18-32. God is supposed to have made it "plain" to us that He exists and that we are "without excuse." Even the wicked retain knowledge of God.

The burden of St.Paul's testimony here seems to be that non-inferential knowledge is available to all; God has created us so that His hand can be clearly seen in the world. It is our moral corruption spilling over into epistemic corruption -- the so-called noetic effects of sin -- which is responsible for the obfuscation of God's reality.

Then, it might be better to formulate Plantinga's "pincer" this way:

(**) On the hypothesis that Christian theism is true, it is quite likely that some of us would be created in such a way that we could have warranted beliefs about God.

This latter claim (**) is one that Plantinga can defend with an appeal to Scripture. Whether (**) is correct or not, I am not sure, but there is some non-trivial exegesis to be done here!

Mike Almeida said...

(*) On the hypothesis that theism is true, it is quite likely that some of us would be created in such a way that we could have warranted beliefs about God.

Hey Clayton,

It's not so difficult to make (*) plausible, if you're allowed to make lots of theological assumptions. This is effectively what happens when Plantinga gives a special epistemological role to the Holy Spirit in WCB, chp. 8. So, if the argument is that, "if this sort of trinitarian God exists and one person of the trinity plays this epistemological role in causing you to believe the great truths of Scripture, then the belief in God will likely be warranted", then we have a pretty good reason to believe (*). But now the atheist is going to balk, very naturally, at this theologically full blooded conception of God.
But even the A/C model in chp. 6 is sufficiently richly theological to make (*) plausible. The atheist is taking on too much in deciding to show that the A/C model (extended or not) is false. After all, Plantinga does not show that it is true. He simply offers it as something epistemically possible--it is effectively another defense. I think the right atheological response is to offer an at least equally plausible story in which God does not figure at all.

D.C. Cramer said...

RE: 'two dozen or so arguments for theism', see the new Cambridge book, Alvin Plantinga, ed., Deane-Peter Baker, which includes an appendix with 'two dozen...', including a preface on them by Plantinga. Might clear things up a bit. (Then again, might not!)

Mike Almeida said...

This latter claim (**) is one that Plantinga can defend with an appeal to Scripture

This is true, as I also suggest (I didn't see Tim's post), but it does leave one wondering what it is to 'defend a claim'. In fact, this route promises much more than it delivers. Some theists try to manage the problem of evil in the same way. Instead of trying to fend off the problem with austere theism (AT), they assume adorned theism (ADT). Adorned theism does appear to manage the problem better--indeed on traditional Christian theism, you should actually expect to see lots of apparently unjustified evil!
But what is actually happening is that you're trading off one problem for another. When you assume ADT, your prior probability for theism sinks dramatically. It is much less likely that some adorned theistic view is true than that some austere view is true. So, though the existence of apparently unjustified evil does not lower (much) the probability of ADT, its' priors are already so low, that you might be better off managing the problem of evil with the AT.
The same will be true with the extended the A/C model. The extended model will give us good reason to believe (**), but the priors on the extended model will be much lower than the priors on the limited model. Basically the more you build into the A/C model to preclude coutnerexample, the lower the prior probability. So what looks a lot like cheating--building into your view protection against counterexamples--actually costs you lots on the front end. Wykstra refreshingly acknowledges this, calling it the problem of "top heavy theories".

Robert said...

I like your tie-in to the problem of evil point, but I'm not sure I get the overall gist of your argument. Why does anyone need to have confidence in (*)? Doesn't it just have to be true for Plantinga to get all he needs? I guess I might not be getting the "pincer."

Mike Almeida said...

(**) On the hypothesis that traditional Christian theism is true C, it is quite likely that some of us would be created in such a way that we could have warranted beliefs about God B.

Robert, I'm not sure I'm following your question. My post was just cautionary. I'm quick to agree that,

1. Pr(B /C) is pretty high.

But I urge that we shouldn't be so impressed with this fact. It is also true that,

2. Pr(E /C) is pretty high.

That is, the probability that we observe just the sorts of evils we do observe is pretty high on a traditional Christian view.

But so what? What we are conditionalizing on is C, and C does not have a high prior probability. Or, many atheists will so argue. I'm sort of sympathetic the the atheistic shrug here. What's the big deal if we let some really improbable assumptions do all the heavy epistemologial work in explaining the evil we observe or in explaining how we could have warrant in believing that God exists? This would be a big deal if there were (contrary to fact) some atheistic argument to the effect that it is IMPOSSIBLE that God should exist along with our having warrant in believing he exists. The argument would then show that, well, it is possible. To be fair, it shows that it is epistemically possible.

Robert said...

Hi Mike,

I think there's been a mix-up--my fault. I intended my post as a question for Clayton's initial post, and as a sort of defense of plantinga.

Clayton said...

Hey Timmo,

But I don't see how your (**) gets around the problem. I'm not denying that someone given some evidence might be justified in theistic beliefs, but how someone who has the sort of evidence we do about evil in the world could be.

Same thing, but I need to think longer about what you and Timmo are saying. I'm a bit tired.

I think you might be right that on some epistemological views, it's the truth of (*) that matters. I'm assuming something P seems to be assuming in his argument that evolutionary naturalism is self-defeating.