The Problem of the Criterion

There is an upcoming conference on the "Problem of the Criterion." The conference organizers introduce the dilemma as follows:

To know whether things really are as they seem to be, we must have a procedure for distinguishing appearances that are true from appearances that are false. But to know whether our procedure is a good procedure, we have to know whether it really succeeds in distinguishing appearances that are true from appearances that are false. And we cannot know whether it does really succeed unless we already know which appearances are true and which ones are false. And so we are caught in a circle.

They also observe that only one book-length treatment has been given to this problem. I wonder if certain arguments which do receive much attention could count as special cases of the Problem of the Criterion. I'm thinking particularly of Alvin Plantinga's "Evolutionary Argument Against Naturalism" (stated succinctly here). Plantinga's argument seeks to show that the conjunction of standard evolutionary theory and naturalism provides a defeater for the proposition that our cognitive faculties are reliable. On the conjunction, the probability that our faculties are reliable is said to be either low or inscrutable.

I wonder if this argument could be construed as developing a special case of the Problem of the Criterion in the following way: the conjunction of evolutionary theory and naturalism precludes the possibility that we could distinguish appearance from reality or identify methods which could do so. This argument (if successful) would provide an additional reason beyond circularity for thinking the Problem of the Criterion is unsolvable (given evolutionary theory and naturalism).

There would then arise a Plantinga-style solution for the Problem of the Criterion, which would be any non-naturalistic view that when conjoined with evolutionary theory makes probable the reliability of our cognitive faculties. On Plantinga's view, theism provides good reasons to think our cognitive faculties are reliable. If we allow our cognitive faculties to stand in as a ridiculously general "procedure," as formulated above, then perhaps we've solved this particular epistemological chicken-egg problem.

But it seems to me that Plantinga's theory is in danger of falling victim to the basic thrust of the Problem. If we take God's existence to be the truth distinguished from reality, and our cognitive faculties as the reliable procedure, then how can Plantinga escape the accusation of circularity? Even taking for granted all of Plantinga's foundational claims, crucially the basicality of Christian theism, it seems that in the most general way our cognitive faculties are part of what produces Plantinga's theism. I suspect that Plantinga would say there is no circularity here, because justification in the basic way is not dependent upon arguments or other propositions, and therefore not dependent on the proposition that our faculties are reliable.

So is there a useful application of or solution to the Problem in Plantinga-esque arguments? Is the "evolutionary argument against naturalism" a special case of the Problem? Second, does the basicality of theism, conjoined with our reasons for thinking we have reliable cognitive faculties, constitute a circle?

Do we need defeater-defeaters?

In the previous post, Ben wrote about a good principle of epistemic humility, that although we must believe each member of the set of our beliefs, we should nevertheless believe the set to be, as a conjunction of all contained propositions, false. In other words, Ben can believe p, r, and q, but nevertheless believe -(p,r,q). This funny state of affairs is merely a moderately rigorous way of expressing the fact that we generally think we hold some false beliefs; we just don't know what those beliefs are.

Like the sultry muse he is, Ben reminded me of some thoughts I've had which are related to this subject. In epistemology the concepts of defeaters and defeasibility loom large. Defeasible beliefs (or propositions) are those that can be defeated. A defeater for some belief or proposition is generally thought to be some other belief, proposition, state, or whatever, which renders the original belief or proposition refuted, or at least counts towards its refutation. I'm not sure there are rigid edges here, and I'm not about to read this to find out, for fear of discovering my own thoughts to be obsolete.

Especially in apologetic contexts (say, when defending moral realism, theism, or utilitarianism), there seems to be an epistemic responsibility to conduct oneself according to at least the following guidelines (not comprehensive):

(1) One should address defeaters as one is made aware of them.

(2) If one cannot offer defeater-defeaters, the target belief is plausibly defeated. Abandon it.

(3) Defeaters of the objective, propositional sort are especially defeating - respond with extra-seriousness to them.

I think all of the above can be challenged. I'll just flirt with each of them briefly and see if anyone comments. I'll notate the criticisms respectively with letters (1a, 1b... etc.).

(1a) It could be that one is made aware of more defeaters than one could possibly handle, given multiple limitations (time, money, psychological tolerance, competing obligations...). Here we might say that the epistemic ought is blocked by a multiply realizable can't. So perhaps (1) can be given side constraints, but it seems like there are a burdensome number of constraints in this direction. Maybe the gloss "within one's abilities" will suffice. But then this is a rather weak sort of obligation! I have all sorts of more important things to do than, say, defeat defeaters.

(1b) It seems like not just any defeater can obligate us. Were I independently wealthy I could spend all my time coming up with defeaters, and then go around harassing responsible intellectuals with them ("Sorry guys, can't go to the movies, have to study tonight; that reclusive philosophy jackass just took advantage of my defeasibility again."). Something like this is actually occurs with Christian evangelists, political activists, and new atheists. So defeaters have to be compelling in various ways. But what makes them compelling enough to obligate us to counter-defeat them?

(1c) The whole obligation itself has a suspect aspect to it, which is that we know (think of Ben's post) that some of our beliefs are actually false. So any defeater of any of our beliefs could be a successful defeater of a false belief. Furthermore, given the likely falsehood of several of our beliefs, isn't it incumbent upon us to figure out which beliefs those are? But I think there is no such obligation. But if there is no such obligation here, why not? And whatever this reason is, wouldn't it mitigate our obligation to address defeaters?

(2a) Despite my many words so far, this is the objection most interesting to me. I almost think this should constitute the whole post. It will be a very common assumption that if a defeater is offered to me and then I do my best to address it, I am obligated to abandon my belief if no defeater-defeater is forthcoming. I've been starting to think that we're just kidding ourselves here. Although it seems unsavory to, for example, simultaneously find the problem of evil convincing but maintain Christian theism, I think we do a version of this all the time. I suspect that we are aware of any number of propositions contrary to our own such that we have no answer to them and they seem true to us. I think we reconcile things by supposing something like the epistemic humility discussed in Ben's post and above - surely we're missing something, even though we know not what. This is one reason why, for example, I maintain some form of moral realism. I can't see any promising way to, for example, defeat defeaters along the lines of "Our natural explanations of the universe are complete without spooky morals." But (along with other reasons for accepting moral realism) I just hold out for some relevant belief of mine other than belief in moral realism being false. Or, better for me, some proposition that makes the otherwise convincing defeater true might be false. So what do you think? Is it possible to be rational and simultaneously believe p and some defeater for p? Before anyone has a stroke, notice that a defeater for p is not the same as -p. Its relation to -p can be any number of things, as testified in the Internet Encyclopedia of Philosophy article, linked above. But I'm trying to push this point for defeaters that are as close to -p as possible. I'm thinking of cases where we end up saying to ourselves, "If R, then -P. R. I believe P, so I'll just wait around here for a bit."

(3a) I wonder if this is really true, since there's something not straightforwardly logical/propositional about why propositional defeaters make it especially incumbent upon rational persons to abandon their defeated beliefs. If we take the abandonment of a defeated belief to be in some way volitional, I think we can understand my claim here. Certain complex beliefs (e.g. belief that the death penalty is justified) are volitional in the sense that we are deeply and consciously involved in seeking out evidence, weighing evidence, acting on our beliefs (e.g. with voting, arguments, maybe pulling levers, choosing cases to litigate, etc.). So it is that I think someone can, "despite the evidence," continue to think the death penalty is legitimate or illegitimate. So anyway, propositional or objective defeaters for these complex beliefs are supposed to inspire within us a certain tendency to abandon the defeated belief. But this is also what, say, a basically acquired belief is supposed to do. For example, someone is made to pull the lever himself, can't do it, and walks away feeling the force of a defeater for his pro-death penalty belief. There may be nothing to my objection here, since even though the goal of the two defeaters are the same, maybe we still should take the propositional, evidential, objective one more seriously. So focus on my (2a).