Monday, July 23, 2012

A Thomistic Critique of Religious Evidentialism

There is a certain view among Christian philosophers that in order to be justified in one's belief in Christianity one must have studied the best philosophical arguments and come to the conclusion of Christianity through a process of discursive reasoning. Call this "religious evidentialism." The obvious problem is that this would seem to consign anyone who doesn't come to belief in God by studying the arguments into the class of epistemically "unjustified" believers. And that seems to be most believers. I don't think Aquinas would endorse this at all, and I think we can produce a good argument for thinking why this is not the case, on Thomistic grounds:

If this evidentialist approach is right, and most people are unjustified in their belief in Christianity because they didn't study arguments, then at least on a Thomistic view these people are to that extent intellectually failing because their cognitive faculties are failing in some respect and they intentionally act contrary to them. Moreover, to the extent they grow in faith, to the same extent do they grow in being unjustified in their beliefs, and thus grow in intellectual vice. But if Thomism is true, grace perfects nature, and does not destroy or act contrary to it. God doesn't make us do evil things in the process of salvation. He doesn't destroy our nature, but perfects it, i.e. sanctifies us (and on the Thomistic view this is the same as to justify us, in the theological sense). So it can't be true that those faithful who have not studied the arguments are unjustified in their beliefs.

This may provide some warrant for thinking something like Alvin Plantinga's picture of religious epistemology is correct.


21st Century Scholastic said...

Even "unsophisticated" christians believe because of theistic arguments.

Many people trace their belief to the question "where did the cosmos come from?" and refer to the role of God as creator. They believe on the basis of the cosmological argument.
Other wonder at the beauty of various portions of the universe (flowers, stars, etc.) and say: "look at how beautiful this is. How could a God not exist"? They're unconsciously realizing that there is a link between transcendental features of things and their Author - and that is just the fourth way.
And if somebody believes in a Creator and "Director" of nature because of its regularity and prima facie puropsefuless, they're basing their belief on the teleological argument. Etc.

awatkins909 said...

Thanks 21st Century Scholastic. That is an interesting point. If the majority of Christians do in fact reason their way to theism then my argument would not seem to work. A few thoughts however:

Certainly I wouldn't deny many people think this way. But clearly these types of arguments are not sufficient to satisfy the religious evidentialist approach, at least not as they stand anyway. These types of arguments could of course be fleshed out and refined philosophically to be respectable arguments, but as given, these arguments are not strong enough to provide strong epistemic justification for one's belief in God.

It's also not clear to me that such believers are really "arguing" in the strict sense. As an argument this is flimsy and even invalid: (1) This is beautiful. (2) So God exists. It might be more plausible to think that, to take your example, they see the beauty of God's creation (think Psalm 19) and this situation simply produces a properly basic, non-inferentially derived belief in God. I think that this--i.e. not inferring belief via argumentation--is the case with at least very many faithful Christians, if not the majority of them.

Finally, suppose I'm wrong; it still seems highly plausible that at least some faithful and grace-filled Christians do not use arguments, if only a minority of them; so they must be justified in their belief in some other way.

davidus said...

Hi Alfredo, I found your blog some time ago and appreciate reading your thoughts.

'Proper basicality' is something I have found trouble getting clear on,—the question of what exactly counts to distinguish beliefs that are properly basic from those that are not,—perhaps this is something you could clarify in future if you personally hold this view of epistemology.

An alternative picture of justification (one I intuitively hold) might involve Bayesian epistemology, i.e. rationality understood as depending primarily on inferring well/rationally from one's prior beliefs, whatever they are. The case for religious justification would now be determining whether religious beliefs flow well from initial (not necessarily 'basic') beliefs about the world. On this picture, rationality appears dangerously liberal, but I think (1) many initial beliefs (e.g. 'common sense' beliefs) would actually be shared; (2) the principles of thomistic metaphysics, and thus beliefs about God and Christianity, in fact ARE the most rational inferences from (1); (3) even if a believer didn't know her arguments (i.e. couldn't see the connection between prior and posterior beliefs), her beliefs could arguably count as justified as long as (2) is IN FACT justified, and the believer had no additional beliefs that would render inference to Christian faith irrational.

To me this view could share some of the virtues of both evidentialism and proper basicality (a rigorous criteria of justification, without condemning those who have not examined the arguments).


awatkins909 said...

Thanks for reading David.

I should be clear that I am no expert in epistemology, and my post is only really meant to justify a negative thesis, viz. that religious evidentialism is false. If you want I can try to characterize this position more fully if it is not already clear. But as far as my post is concerned I'm not necessarily committed to endorsing any positive thesis such as reformed epistemology.

Now, from my understanding, a belief is properly basic if and only if it is justified and it does not depend for its justification on other beliefs. So, for instance, maybe my belief in the external world is properly basic, and I am justified in believing in it but my being justified in this belief is not due to its being inferred from some other belief.

The Bayesian type position seems to me to be closer to the evidentialist approach, and I would argue it must ultimately reduce to it, at least as you've characterized it. Let me explain.

First, we have to be careful here in our use of terms like 'inferred', 'flow', and 'justified'. Given your characterization of the Bayesian approach as involving inferring well from prior beliefs, an individual believer's epistemic justification would not in fact reduce to determining whether religious beliefs objectively flow well from such initial beliefs, at least on certain interpretations of the notion of '; for if by 'q flows from A's belief that p' you mean something like 'q is entailed by p' then one can believe something which entails another belief but still not be justified in believing the latter. For instance, suppose God exists, and moreover exists necessarily; then the proposition that the sky is blue objectively entails the proposition that God exists (for anything entails a necessarily true proposition). But clearly I'm not justified in believing in God because I believe in the sky being blue.

You might instead mean by 'q flows from A's belief that p' something like 'there is a sound argument from p to q (which A need not necessarily know about)'. However, then it is still not obvious or clear that just because q flows from p that I'm justified in believing that q. For instance, most people know enough mathematics such that from the basic propositions of arithmetic they know there is a sound argument to the proposition that (all the digits of pi) + (all the digits of pi squared) = some huge number n. But surely they aren't justified in simply believing this out of the blue, even if they happened to get the right answer for n.

Also, we must keep in mind that it is an individual's beliefs which are justified, not propositions. Propositions are not justified or unjustified, but rather they are true or false. The proposition that all potency exists mixed with some actuality is not justified, but rather true. My belief in this proposition on the other hand is what is justified or not.

It seems then that the only way to really understand 'q flows well from A's belief that p' is as 'q can be (soundly) argued for by A from p'. And this seems to reduce to the evidentialist approach.

Also, I'm curious, what do you think about my argument in the post?

davidus said...

Thanks for reading my (as I now see, rather sloppy) post with charity, Alfredo. :-) I understand the post doesn't actually concern itself with a positive thesis about epistemology. I also agree with your other observations, and I see that when I talked, e.g. about belief about a prop being 'in fact' justified (vs. some individual's faith in that same belief), it sounded pretty dodgy. Inadequate as my expression of my view may be, I do still want to add something to save the 'Bayesian' position I have in mind from collapsing into evidentialism as you describe it. What I mean by a statement in the vein of, 'q IN FACT flows well from A's belief that p', is that A's belief in q is entailed by her prior beliefs, IF she in fact has perfect epistemic access to all her beliefs. (I.e. if she is a perfect rational agent). But nobody has such ideal rationality/epistemic access. My thought is that individuals who believe in God seemingly without 'arguments' are those who in fact have the right overall set of beliefs to constitute justified true belief, but simply don't have the epistemic access to said beliefs. This would be because natural (initial) belief-formation of would accord with the nature of reality, thus providing the necessary set of beliefs.

Do you think this holds water?

And, forgive me for not commenting on your actual argument—for I think I agree completely with it. It seems correct to me that evidentialism as you describe it could not allow for the increase in faith (without reasons) to count as meritorious in any way, and that given the Thomistic view of grace and nature, an increase in faith in any context would have to involve justified belief in God. I definitely suppose that Thomas would have held belief in God to be naturally justified in some way or other (as I imagine you would).

awatkins909 said...

Thank you for commenting!

My worry here would be that this analysis would still be susceptible to counter-example in the manner of my fifth full paragraph in my last comment. Let me try to explain a little more. So, what we have then is this definition of 'flow':

(FLO)'q in fact flows well from A's belief that p' just in case if A had perfect knowledge of her prior beliefs and all the sound arguments based on them, then she would infer q from p.

Do you think that is an accurate representation of what you're saying, or at least equivalent? I'm using reference to arguments here rather than 'entailment' since the latter seems much more easily open to counter-example, such as that in my fourth full paragraph. Would you then understand epistemic justification as follows?

(JUS) 'A is justified in believing q' if q in fact flows well from one of A's prior beliefs that p.

Let me know if you think these definitions accurately represent what you're saying. Now, if they do, then it could still be the case that A just happens to get lucky and guess the right answer to the equation '(all the digits of pi) + (all the digits of pi squared)' and be justified in her belief in this. For if A were perfectly rational in the sense we're working with, she would see that from her very basic arithmetical beliefs one could soundly argue to the right answer of this. But if one can simply guess what's true and be justified in this belief, this doesn't capture the sense of justification we're working with.

To make the point more vivid let me give an illustration. If we were to ask A why she believes whatever number she gives is the right answer she would say, "No reason, just a guess." But surely she's not justified here, despite the fact that if she were perfectly knowledgeable she could argue from her prior beliefs to the right answer.

So my worry here is that this leaves us with the same problem as before.

awatkins909 said...

BTW, just curious, about how long ago do you think it was that you found this blog?

davidus said...

Pt 1

Thanks again for the good reply! I fear I still haven't quite presented my view with clarity (specifically, I am doing a real bad job of describing the Bayesian picture. As you have correctly noted, my use of 'entailment' is indeed unsuitable, as is my description of 'flow').

I'll give a shot at re-describing my view of the Bayesian picture. What actually happens in the Bayesian picture, as I understand it, is that a person's credences (*degrees* of belief) are formed through a process of conditionalising on prior credences given *evidence*, i.e. experiences. E.g. person A has an experience *z* of a tiger for the first time, comes to update his prior credences given this evidence such that his prior credence in the existence of tigers, q1, (a fairly low credence) is replaced by a posterior credence, q2 (fairly high). Where do actual beliefs come in? in addition to these credences, A might have full-fledged prior *beliefs* about what kinds of credences would be sufficient for belief that [q] (that tigers exist). The prior beliefs might be belief in the following counterfactual [x]: "if b(q2 & y), then b(q)", where b is the property of being believed by A, and y is the appropriate set of all prior beliefs that A has. So given the experience *z*, A conditionalises on (q1 / z) = q2, and from this infers from x (given y), that he is to change belief from ~q to q.
At this point I am not sure how to characterise what the process of coming to believe q is, but I guess it needs more characterisation than just an inference from some p. Again, to refer to the above scenario, p constitutes a combination of belief in x, belief in q2, and all other beliefs y. Now, if A were to be an ideal rational person, she would have (1) a complete set of prior beliefs of kind [x] about how she would conditionalise on every sort of given evidence to get posterior beliefs, and (2) perfect 'memory' of all these beliefs in the sense that she would see immediately when, where, how, and why she would hold all of her posterior beliefs. Now if (1) and (2) are satisfied, then I'd say A has 'perfect epistemic access', i.e. she would respond in a completely coherent way to any form of evidence of kind *z* which, combined with her beliefs x, would lead her to belief in God.

Now Bayesianism isn't perfect because it leaves unanswered the question: where do the INITIAL prior beliefs come from? My assumption is just that a minimal set of prior beliefs (maybe basic beliefs of some sort about immediate objective reality) are formed. This is the 'hole' in Bayesianism I think. But it can be patched.

Now, for such a person A who satisfies both (1) and (2), any given belief q that A comes to believe will count as a belief that 'in fact flows well' from p.

davidus said...

Pt 2

I'll try to rephrase (FLO):

(FLO*) 'q in fact flows well from A's belief that p' just in case if A had 'perfect epistemic access' (i.e. 1 and 2), she would infer q from p.

I agree with your characterisation of (JUS), but I think my use of 'perfect epistemic access' is deviant enough to make the mathematical counterexample not be a worry. If A hadn't computed the complex equation before, she wouldn't believe she knew the answer. Even if she had done it before, but didn't believe that when confronted with the experience of being shown the complex equation again, she would be able to do the answer without calculation, then her guess wouldn't be in accord with p and hence would be unjustified. To clarify: my use of epistemic access just means, knowing how one will react to evidence given perfect memory of one's beliefs.

Now in the absence of (1) and (2), A could fulfil all the criteria for (FLO*) and (JUS) while failing to be able to articulate her reasons. For example if she doesn't 'remember' all her prior beliefs y, she won't be able to actually articulate why q2 leads to q.

RE: God. For any A that believes in God, the case can be made that this is because she firstly has the right kinds of prior beliefs about the world p, (which includes exposure to the right kinds of evidence). This would be enough for her belief in God to be an inference of q from p, although she might not be able to articulate this because she doesn't have perfect epistemic access.

I haven't formulated any of these thoughts on paper before, so if I am still being incoherent/unclear, I'll leave these thoughts for another time and save you the trouble. :-)

—I found this blog about three or four weeks ago while doing some google trawling on thomism and modality, I think. Props again to the blog—nice clear posts. It's great to find fellow trad Catholic phil students around. Not many from where I am from! :-)