Sunday, September 09, 2007

A Epistemological Dilemma For Evolutionary Naturalists

1. A good case can be made that externalist theories of justification (or warrant) are undercut by evolutionary naturalism in that with the conjunction of the two we have a defeater for our belief in the reliability of our cognitive faculties (i.e., that part of our noetic structure responsible for producing true beliefs).

2. A good case can be made that internalist theories of justification (or warrant) imply (or are assumed by) substance dualism in that these theories seem to be motivated by first-person perspectives of the knowing and experiencing subjects, they seem to assume a same-self as the one having access to and being aware of the justifying reasons for belief, and they seem to imply some other features which are problematic for evolutionary naturalism: At least on some readings, libertarian free-will is assumed due to the ought-implies-can principle implicit in deontologism, and most analysis of internalist theories of justification and warrant attach justification (or warrant) to beliefs which, on some naturalist theories, are queer entities that need to be eliminated, and also requiring a normativity condition at odds with descriptive accounts of the world evolutionary naturalists must give, and lastly doxastic voluntarism seems to be implied by internalists - this seems to be at odds with the pictire of the world that naturalists give us. Indeed, this is why many naturalists hold to an externalist theory of knowledge, calling it "naturalized epistemology." It should be granted that subtance dualism is incompatible with evolutionary naturalism.

3. Externalist and internalist models seem to cover the field of possible ways justification or warrant is conferred on beliefs (or C-fibers firing!).

4. Therefore, a good case can be made that evolutionary naturalism undercuts the only possible ways justification or warrant can be conferred on beliefs.

Premise 1 is supported by arguments like Plantinga's EAAN. Premise 2 is supported by arguments like Reppert's AFR.


  1. "1. A good case can be made that externalist theories of justification (or warrant) are undercut by evolutionary naturalism in that with the conjunction of the two we have a defeater for our belief in the reliability of our cognitive faculties (i.e., that part of our noetic structure responsible for producing true beliefs)."

    Hmm. I wonder about that.

    I once heard Greg Koukl give an interesting answer to someone who called in to STR and asked how he knows that his senses are reliable. He said he can make a good case for the basic reliability of the senses--by pointing out that he's alive. Every day he gets into a big metal box and zips around with other people in big metal boxes zipping around him--and yet he survives. (Of course, we could still be living in the Matrix or be the dream of a butterfly.) There can still be flaws in our sensory perceptions, but they are reliable.

    That's basically a natural selection type of argument, and it does seem valid to me. Why can't a naturalist appeal to the same principle for the basic reliability of our cognitive functions? There may still be flaws in our ways of thinking and knowing, but they're valid to the extent that our survival depends on them.


  2. And so are you unfamiliar with Plantinga's EAAN? Or, do you find problems with that?

  3. Basically, the defeater obtains against the *reflective naturalist,* and so Koukl wouldn't have the defeater. Furthermore, Koukl's argument, offered by the naturalist, assumes that adaptivity = truth.

    So, everytime I saw a man-eating tiger, I could form the belief that a marathin in the other direction was about to start, this would get my body parts in the right place to survive, and so my beliefs would keep me alive, but this wouldn't mean that they were true. This argument is an alethic rationality defeater.

  4. P.S. Tim, here's a short statement of EAAN

  5. P.P.S. and, I'd also be surprised in Koukl disagreed with EAAN. Moreland uses it in his Philosophical Foundations for a Christian Worldview, and I know Koukl uses that book and has Moreland as a prof. Doesn't necessarily imply that he endorces EAAN, but I find it highly likely.

  6. Ah, wonderful! Thank you for the link. I'll contemplate it more and get back to you.

    No, I did not have prior familarity with Plantinga's EAAN specically--I do believe I've seen many of the components of the argument, but not in that rigorous form.

    You could rightfully ask why I commented on your 1.) without reading EAAN when you'd referred to it to support 1.). Basically, I commented because I've been thinking about it for a while, and I know naturalists tend to think along those lines--so I thought it would be profitable to deal with it specifically & explicitly. (I probably should have started with "I'm saying this without familiarizing myself with EAAN, sorry if it completely addresses what I'm about to say.")

    Anyway, like said, I'll get back to you. I do think I'll have something to say.

    Oh, and I agree with you about Koukl--I would be surprised if he didn't endorse EAAN.

  7. Tim,

    Koukl's answer to the caller begs the question (at least, it does as you've paraphrased it).

    His claim that "every day he gets into a big metal box and zips around with other people in big metal boxes zipping around him--and yet he survives" takes for granted the basic reliability of his senses. You can't appeal to empirical knowledge as a premise in an argument for the reliability of your senses. That's why it's such a challenging philosophical problem. :)

    Your parenthetical comment about the Matrix indicates that you already recognise this. For if we're living in the Matrix then our sensory experience isn't reliable.

  8. James,

    Actually, my parenthetical remark was part of the paraphrase. Koukl didn't beg that aspect of the question--he discussed it with the caller.

    His answer was, unless you are willing to entertain the idea of a wholesale fantasy world, you have good reason to trust the basic reliability of the senses.

    His specific response to the Matrix objection was to say, sure, it's possible. Just because something is possible doesn't mean that it's reasonable for me to believe that it is so.

    And I think that's a satisfying enough response. Not "proof", but satisfying for me. I can't disprove things like Last Thursdayism, but I don't worry about them.

  9. Tim,

    Thanks for the clarification.

    I guess I'm now left somewhat confused as to what Koukl meant by "make a good case". I assumed he meant an anti-skeptical argument. It now sounds as though he was making a pragmatic argument. That's fine in certain contexts, but a pragmatic argument would be no use to a naturalist who wants to defeat EAAN because pragmatic justification and epistemic justification are quite different beasts.

    A pragmatic justification says "we should believe P because that would have good consequence C" whereas an epistemic justification says "we should believe P because it's likely to be true for reason R".

  10. The call wasn't on EAAN; the caller asked how Greg knows that his senses are reliable, and his answer was restricted to that topic. (Koukl is a Christian apologist, BTW, in case I gave the wrong impression.) I made the application of similar lines of reasoning to the Paul's point #1 up above.

  11. Tim,

    "how Greg knows that his senses are reliable"

    And I think you make James' point. Greg said he knew that P, based on P.

    I think many philosophers, Wlliam Alston for example, have demonstrated that there is no non-circular argument for how you know that your senses are reliable.

    But, knowing things by our sense are a paradigm case of knowledge.

    So, now what?

    I'd say, following others, we take it as basic. That is, you don't need propositional evidence (e.g., evidence about the survival rate of huamns in moving tin cans!) in its favor.

    So, you may ask, why can't the naturalist take the reliability of his senses as basic and thus respond to EAAN that way.

    Well, much of this depends upon prior argumentation about the probablity of the reliablility of your congative faculties. No time to get into that - but I did offer a link above. Suffice it to say, if the reflective naturalist grants that P (R/NE) is low or inscrutable, this defeater-deflector doesn't work for him.

    So, giys like Bergmann have taken this route in trying to answer EAAN. Since I have made a few comments about that in other places, I'll just re-post them. Remember, they are in the context of responding to a particular person, and Bergmann in general:

    Relying on Bergmann's idea of R being properly basic, and so it's a defeater-deflector for EAAN, seems to miss the point of EAAN. Of course proper function would demand continued belief in R, but this is not because this portion of your cognitive faculties are aimed at truth, but, rather, at the avoidance of cognitive disaster. A person S may be in a situation - say, lost in a snow storm on top of a mountain - and S may see a ridge that S thinks could be leaped to. Based on perception, this belief is basic to S. But, S would not have thought this if S were not in this survival situation. So S maintains this belief that the chasm is able to be jumped. Proper function requires this belief to be maintained. The optimistic overrider has kicked in. But the faculties governing this have some other virtue in mind - survival rather than true belief. In normal, reflective situations, S would not form said belief.

    Or, suppose S ingests agent XX, a hallucinogenic drug, producing hallucinations in 90% of those who take XX. Proper function would require assuming R so as to avoid cognitive disaster. So, S has powerful inclinations to continue on in belief in R, even though S has come to believe that P(R / XX) is low or inscrutable, and S may take it in a basic way, but of course these powerful inclinations don't count as evidence for R. S would have this inclination whether she was in or out of the lucky 10%.

    One can only use Bergmann's position that R has non-propositional evidence for it, raising the probability thesis, in cases of an unreflective naturalist. But, can the reflective naturalist have this? Well, if so, then R has warrant for him only if R is produced by proper functioning cognitive faculties successfully aimed at truth. But his beliefs are only successfully aimed at truth if they are reliable, and this is precisely what's in question. And, just to say that he has a strong inclination to believe R, doesn’t get him where he wants (see (6) above). And, if he assumes N & E, he has no reason to assume that his beliefs are aimed at truth. So non-propositional evidence doesn't help S.

    If all it took was that a belief B could not be defeated D because B had non-propositional evidence, then it would appears that basic beliefs could not be defeated. Say that you go inside a widget producing factory. You see the widgets, and they appear red to you. The belief is formed by your senses, which proper function demands you believe are reliable, and hence the belief is basic. But, the shop manager, and close friend, tells you that the widgets have a red light illuminating them so as to detect otherwise unnoticeable defects in the widgets. He says that there are actually very few red widgets coming down the assembly line. It would appear that your basic belief was defeated that what you were seeing was a red widget because the probability that you were seeing a red widget was low.

    Or, say that S comes to believe that she is a brain-in-the-vat (BIV). Further, so has no reason to believe that the Alpha Centaurian super scientists care whether she has mostly true beliefs or not. So, she believes P (R/ BIV) as low or inscrutable. Doesn't she have a defeater for R, no matter how strongly she continues to assume R on an every day basis. After all, the Alpha Centaurians needs to see her in every day life and so have constructed her to continue to believe R, and act as if R, because to not act that way would lead her into cognitive despair. But, when she reflects on her thoughts, she quires a defeater for R, and so gives it up. Same with the naturalist.

    Lastly, S is irrational because proper function for internal rationality that S give up one of S's beliefs: R v N&E. If S came to believe that she had ingested agent XX - which caused hallucinations in 90% of those who took it, and S also held to R, proper function would require S to drop R. Or, say that S believed his head made of glass, and believed that he could play football without a helmet, then he would be internally irrational to not give up one of the beliefs (if, following Alston, one wishes to say that S should drop N&E rather than R, I still think that gets you to ~R, but just takes the long way. For without a story on the purpose of your cognitive faculties, how they got here, or anything, S should remain agnostic A about R, and so P(R/ A) would be inscrutable, and the defeater for R has not left. Though she still might act as if R, upon reflection, rationality demands R to be dropped. We would call someone who believed both of the above propositions, irrational. If S is in a similar situation with P(R / N&E), then S is irrational for holding to R in the Humean and alethic-rationality way.

  12. Paul,

    Thanks again for the link. It was a pleasure to read. However, I'm not persuaded that EAAN is a compelling defeater for naturalism.

    Premise (1) is "P(R | N&E) is low", or the probability is low that our cognitive faculties are reliable given that they have come to be through naturalistic evolution. That's where I see the problem; I agree with (2), (3), and (4). So I'm looking specifically at the four paragraphs where he supports (1).

    The problem, as I see it, is that he does not argue primarily in terms of the adaptivity of cognitive faculties or reasoning processes. Instead, he talks about the adaptivity of particular beliefs. When he goes to connect this to cognitive faculties, I think he uses a mistaken assumption. I find his argument compelling against a naturalist's confidence in the truth of instinctual or intuitive beliefs. It is less compelling against reasoning processes.[1]

    The first two paragraphs deal with neurophysiological (NP) properties--having to do with the neurons & such that relate to our beliefs. I don't have any comments here. It's mostly straightforward, and I'm not familiar enough with the bit about non-reductive materialism to be sure how that detail might affects things. (I'm a physics nerd, not a trained philosopher.)

    Go to the fourth paragraph. I agree that the adaptivity of a belief does not depend on the truth of the belief. Your tiger example works. (And in fact, this is precisely the type of thing some naturalists will argue about human morality. That is, we may have a moral intuition, but it is simply an adaptive instinct that is not grounded in objective normative principles.) However, for evolutionary purposes, we are not interested in the adaptivity of beliefs (AB) so much as in the adaptivity of the reasoning processes (AR) that produce them. AB will act to establish (easily erroneous) instinctive beliefs in our minds; AR will act to establish our cognitive faculties. The two are going to be related, of course, but we do hae to get to AR.

    I agree with this sentence:
    Well, since the truth of B doesn't make a difference to the adaptivity of B, B could indeed be true, but is equally likely to be false; we'd have to estimate the probability that it is true as about the same as the probability that it is false.

    But then he tries to connect this to the reliability of cognitive faculties (going to AR) by looking at the truth or falseness of multiple beliefs:
    But that means that it is improbable that the believer in question has reliable cognitive faculties, i.e., faculties that produce a sufficient preponderance of true over false beliefs. For example, if so, if the believer in question has 1000 independent beliefs, each as likely to be false as true, the probability that, say, 3/4 of them are true (and this would be a modest requirement for reliability) will be very low—less than 10−58. So P(B | N&E&NRM) specified to these creatures will be low.

    It seems to me that he made a basic mistake. He said, "...if the believer in question has 1000 independent beliefs...[bold added]". The beliefs are not independent. We're not talking about some set of instinctive beliefs--those could be independent. We're talking about a set of beliefs all produced by the same reasoning processes. The relevant question is not how likely B is to be true. The relevant question is, how likely is it that unreliable cognitive faculties would consistently produce beliefs that are adaptive but not true?

    What kind of mind would be able consistently to produce false-but-adaptive beliefs in a wide variety of situations? That probability seems much harder to evaluate than P(B | N&E&NRM). I'm not sure how to come up with a good estimate, but my intuition tells me it's low. That is, my intuition still tells me that any cognitive faculties that consistently produce adaptive beliefs in a variety of situations is going to be producing predominately true beliefs.

    I realize the link you gave me is just a short statement of EAAN; I'd be curious to know if any of Plantinga's other developments of EAAN deal with this.

    [1] I suppose you can't completely decouple our reasoning processes from intuitive assumptions. I haven't fully thought through that aspect. For the moment, I think we might be able to get by if we distinguish between intuitive beliefs that "stand alone" (e.g. a moral intuition) and intuitive beliefs that form the backbone of all reasoning processes (e.g. non-contradiction). In other words, when I talk about cognitive faculties, include the latter but not the former.

  13. Paul,

    P.S. I wrote the above before I'd read your last comment.

  14. Tim,

    Yes, Plantinga deals precisely with those points (briefly) in Warrant and Proper Function...more fully in Warranted Christian Belief....more fully than that in Naturalism Defeated (e.g., in his positive case and in his reply to others, like, say, Fodor)....and also Troy Nunely deals with those objections, and a host of others in his doctoral dissertation on EAAN.

    Here's a link to that:

    and here's a link to Plantinga's most developed online version of the argument:

    FWIW, anti-EAAN guys like Talbott and Bergmann grant (1).

    Anyhow, look at the way naturalists like Dawkins and Hitchens argue. They say that evolutionary processes produced theistic-minded creatures and, at one point in time, this had survival value...just not anymore. So, as a Christian, entire sets of my beliefs are false. I believe all things are to be done to the glory of God. Therefore entire sets of beliefs are false. And, it is the naturalist who as made this argument! (Btw, no one is arguing that every single belief *is* false. In fact, take the tiger example. We still believe that there are teeth, fur, orangish tinge, about 16 ft of length, etc. So, true and false beliefs are mixed together. But, if you add in "marathon-signaler" to the belief about the tiger you saw, then the belief is false, even though parts of the belief were not.

    Or, what if I believed I was inside a giant video game? I got points for dodging bad guys viz. cars, trees, lions, for getting promoted at work, etc. Entire sets of my beliefs would be false but adaptive (e.g., my "boss" is a computer program designed by game designers from Alpha Centauri).

    Or, take Plantinga: "Suppose (as with several naturalists, Ruse, Wilson) that belief in God, while false, is adaptive. So, suppose a tribe of cognitively gifted creatures believe that everything (except God himself) has been created by God; they therefore think that everything is a creature, something created by God. Suppose further that their only way of referring to the various things in the environment is by way of definite descriptions as 'the tree creature before me' or 'the tiger creature approaching me.' Suppose further that all their beliefs are properly expresses by singular sentences whose subjects are definite descriptions expressing properties that entail the property of creaturehood -- such sentences as 'The tiger creature approaching me is dangerous' or 'The tree creature before me is full of apple creatures.' Suppose, finally, that their definite descriptions work the way Bertrand Russell thought definite descriptions work: 'The tallest man in Boston is wise.' Then from the naturalist perspective, all their beliefs are false. Yet these can still be adaptive: al they have to do is ascribe the right properties to the right creatures.

    There are plenty of other ways in which systems of mainly false beliefs can be adaptive. These tribespeople, for example, might also think that in any situation calling for action, there is just one action which has the nonnatural action property of being right; all others have the nonnatural property of being wrong. Given that their isn't any nonnatural property of being right, their judgments about actions will be for the most part mistaken -- though adaptive, if they ascribe nonnatural rightness to fitness-promoting actions and take the actions they think are nonnaturally right." (ND, 260-261)

    He goes on to cite arguments like those of Michael Rea's who argue that naturalism implies an ontology where there aren't really any objects - although there is a gunk or a goo that may, they say, be propertied at different times and different places. What if Rea is correct? Seems that all sets of beliefs which include "objects" are, on naturalist terms, false.

    We can offer thousands of these. So, what is the probability of R? Seems inscrutable. Why is it that out of the thousands and thousands of ways sets of beliefs could be false but adaptive, we have the one that is adaptive and true? How do we figure that out?

    That should be enough to get the ball started (especially with those links). Hope the resources help.

  15. Tim,

    I should clarify. The purpose of my original comment was not to criticise Koukl, but rather to point out that his answer to the caller (whatever its merits in other contexts) would not help a naturalist to defeat EAAN.

    I realise that Koukl was not addressing EAAN. I was commenting on your application of his answer to EAAN.

    FWIW, I'm familiar with Koukl's ministry and I think very highly of him.

  16. Thanks for those resources, Paul. I'll continue to look into it.

    James: Ah, OK.