Monday, February 05, 2007

Darek Barefoot responds


Lewis's philosophical instincts were good. He didn't develop his thought adequately in this case, but I think he sensed a genuine incoherence problem. I will admit that I have struggled with it mightily.

The key to the argument is that it pertains to naturalistic assumptions that occur epistemologically prior to speculations about the relationship between true beliefs and advantageous ones. Any natural account of reason begins with the assumption of processes that have no inherent sensitivity to truth; specifically, truth cannot be seen as the goal of the great sequence of chemical reactions that constitute evolution. If we stipulate beforehand that we owe our beliefs exclusively to processes that have no teleological dimension, then whether we realize it or not we are left to bootstrap ourselves up to coginitive reliability when it comes to our thoughts about the relationship between natural processes and truth.

Natural selection, specifically, can be sensitive to truth only indirectly. Where beliefs are concerned nature's imperative is their advantageousness in the struggle for survival. The natural perspective must put advantageousness before truth in evaluating human cognition, and therein lies the snag.

To illustrate, suppose that our only way to detect fire was by the smoke that usually accompanies it. Our only fire detector in that case would be a smoke detector. But while a smoke detector might keep us away from fire, it would be nonsensical to try to use a smoke detector alone to discover or verify that smoke usually accompanies fire. In fact, if our sole means of detecting fire were smoke, it is hard to imagine how we could know about fire. To know that smoke comes from fire, we must have some means of detecting fire other than simply by the smoke it produces.

If natural selection is the sole architect of human cognition, the brain is a detector of advantageous behavioral options. Our beliefs are screened only for their contribution to survival. The overlap between advantageous beliefs and true beliefs would be like the overlap between smoke and fire. The only thing we can know about our beliefs, given this account, is that they are advantageous. To know that advantageousness and truth usually go together when it comes to beliefs, we would have to have some means of distinguishing between advantageousness and truth. And that, in turn, would require the ability to detect the two values separately, just as with fire and smoke.

In fact, without the ability to distinguish truth we could not even know that our beliefs were advantageous, since that itself requires truth-detecting ability. The account based on natural selection alone reduces to a purely behavioral theory excluding knowledge of any kind. To make room for our knowledge--of natural selection or anything else--we have to bring in a non-naturalistic source of human reason.

Victor, you proposed the counter-example of supernatural beings who planted thoughts in our minds. But for such an example to be comparable to natural selection, you would have to propose that such beings not only plant thoughts in our minds but that they do so without any conscious regard for the truth or falsehood of the thoughts they are planting. Given those conditions it would indeed be incoherent to speculate that some blind mechanism happens to be aligning the planted thoughts with genuine reason. To know such an alignment could occur we would have to know genuine reason first-hand.

The argument against natural selection as the architect of reason is not a skeptical threat for the same reason that the liar's paradox is not a skeptical threat. If a speaker says to a listener, "I never tell the truth," it is not grounds for either party to conclude that the speaker never tells the truth. It is grounds for both parties to conclude that the speaker has not expresssed himself coherently. The speaker in the present case is naturalism, and to be taken seriously naturalism at a minimum must pass the test of coherence.



At 1/24/2007 06:00:00 AM , Blogger Jason said...


I agree that this can be derived from at least part of Lewis' argument in chp 3; but it doesn't seem to account for where Lewis ends up going in chp 3.

For example, we can ask whether this version of the argument avoids attempting to justify our justification ability. Whether the answer is yes or no, though, the necessity of asking and answering this question is reflected in how Lewis eventually ends out chp 3--yet it is not a topic that seems to be addressed in this explication of Lewis' argument at all. When "_the_ key" to Lewis' argument doesn't involve synching up with his own finale to his argument, it doesn't (prima facie) seem like much of a key, to me. {s} (As useful as it may otherwise be to contemplate and keep in mind.)

On the other hand, if you aren't responding to the debate between Victor and I about what (to put it colorfully) "the most dangerous" part of Lewis' argument is--which is entirely possible given the range of material being collected here on the AfR--then of course I retract this criticism with apologies; apart from the provisional caution about the necessity of checking to ensure we aren't by this argument attempting to justify our justification ability (or a similar attempt of that sort).


At 1/24/2007 10:34:00 AM , Blogger Blue Devil Knight said...

I think Jason is right. Put in a way he might not agree with, if the starting point is "Justify your justification rules" then we end up in a regress from which nobody has ever escaped (though many have tried).

If, on the other hand, the starting point is "Using reason, show how reason (with concern for truth, etc.) could have evolved", then the argument loses its punch (it just isn't that hard to imagine, in broad outline, how truth-sensitive nervous systems could evolve. As Fodor puts it, evolution doesn't directly care about hearts either, but they have evolved just the same).

Ultimately these arguments come down to intuition-pie throwing, and that's not good philosophy (though it is unfortunately one of the most common types of philosophy).

At 1/24/2007 12:33:00 PM , Anonymous Steve Lovell said...


Personally, I've always felt funny about downplaying Lewis's comments on evolution and allowing Flew a small victory on this point, but I haven't been able to see my way to any other position. Your line of thinking therefore has a certain appeal to me from an interpretive point of view, but I'm not sure about the likely success of this line of argument.

In particular, why can't the evolutionary response still be seen as an attempt to "stop doubts arising in the first place" rather than as an attempt to remove doubts that have arisen? There seems to be some sort of assumption that you think (like Derek, and here I see you and him as rather closer than you seem to), that the issue of the cetaris paribus relationship between naturalism and the reliability of our faculties is in some way "epistemologically prior" to any speculations about the relationship between evolution and that reliability.

I'd be interested to hear, by the way, how you think your interpretation of Lewis compares with Plantinga's evolutionary argument against Naturalism and particularly with his "Dreaded Loop" as in "Naturalism Defeated?".

To me, you seem to be going in the same direction as Plantinga here, and I'm just not convinced that the dialectic is right. Again, why can't the naturalist argue that evolution would make reliability probable and say that knowing this prevents the doubts arising in the first place?

At 1/25/2007 07:29:00 AM , Blogger Jason said...


The main point to Lewis’ argument is that there is a more fundamental question that can be asked and answered _before_ going to the evolutionary development question; a more fundamental question that _doesn’t_ involve _starting with_ “justify your justification rules”.

Using reason to show how reason could have evolved in an atheistic reality presupposes that the question of atheism either hasn’t been addressed already or else has been addressed and remains in play. Lewis is going down to the root, though, and is addressing the atheism/not-atheism question as a discovered topical priority. (Even if incompletely by not putting theism to the same kind of conceptual test--does theism also require eventually trying to justify our justification ability?)

It ought to be obvious that _that_ procedure is not intuition-pie throwing (something I agree should be avoided). It’s a careful following out of implications from a proposed position.


If anything, I thought I rather pointed up one of the few things Lewis specifically has to say about evolution in that chapter--an agreement by Lewis that (in my experience) normally is downplayed by parties on either side of the dispute.

In any case, it is Lewis who (apparently) allows Flew the small victory there. This needs to be recognized in a full accounting of his argument. Whatever else Lewis has to say about the efficacy of inherited instinctual reactions being honed by natural selection (and he does have some other very important things to say about it afterward), he is also making an explicit concessionary agreement. The content of that agreement has to make a proportionate difference in the shape of what he’s aiming for subsequently when (immediately afterward) he starts talking about how any expectations produced by such a method cannot be considered valid inferences.

{{In particular, why can't the evolutionary response still be seen as an attempt to "stop doubts arising in the first place" rather than as an attempt to remove doubts that have arisen?}}

My first answer would be that the difference would of course depend on the prior contexts to the response. It looked to me like Flew was attempting to remove doubts already risen by Ernest Gellner (to whom Flew was specifically responding, if I’ve read you correctly.)

My second answer, though, is that I specifically said that even if Flew _is_ attempting to stop those doubts from arising, he may very easily _still_ be shooting himself in the foot along the line ultimately argued by Lewis.

We can use our presumably possibly reliable cognitive faculties to study how our faculties operate and even to draw inferences (whether to liklihoods or to conclusions) about how our cognitive faculties, being discovered to be derivative, have and do come into existence with the properties they have. Despite how it may seem from my answer to BDK above, I even recognize that this can be done (in principle) on the presumption of an atheistic reality (assuming no problems crop up by proposing an atheistic reality per se. Lewis' ultimate point, however, is that inescapable problems do crop up at that prior position.)

We can also use our presumably possibly reliable cognitive faculties to deductively trim off propositions if we discover those propositions necessarily lead us into positions where we have to use our presumably possible reliable cognitive faculties to explain that our cognitive faculties can be possibly reliable. (Which is, I am arguing, what Lewis is ultimately arriving at in MaPS chapter 3, with atheism being discovered to be this kind of proposition, thus being logically trimmed out of the option pool for our acceptance. Not that other versions of the AfR, with some dangerousity {g}, can’t be derived from the material of that chapter, but to me this seems the most dangerous variant, as well as the conclusion Lewis is ultimately arriving at.)

If Flew was presenting an option for deductive removal, on the ground that proposing it leads necessarily only to subsequent fallacies, then he might not be falling foul of Lewis’ argument. (Though so long as he’s talking about evolutionary development then Lewis has outflanked him and gone behind him by at least one topical stage. If atheism is removed from play, there isn’t any point treating evolutionary development in an atheistic context afterward. cf Lewis’ agreement in his opening comments, that there is no point discussing the reports of the resurrection of Jesus in a context of operant theism, if theism has already been removed as an option.)

It seems obvious enough (to me anyway) that Flew is not doing this, though; not so far as has been reported. If he’s trying to stop doubts from arising, he isn’t doing so by removing options discovered to be fatal and proceeding along paths left over and discovered to be clear of that fatality. So, what doubt is he trying to stop from arising, and how is he going about that?

Let me draw a parallel of practical procedure, with close connection to this discussion. Let us suppose (for sake of argument) that I do discover that theism per se, even if it has variants which do _necessarily_ involve calling the possible reliability of my cognitive faculties into question (leaving me unable to defend against this problem without presuming the answer anyway), nevertheless also has one or more variants where this logical problem does not occur. I might discover that, at first glance, the solution would seem to be that “I” am Theos instead of being a creature distinct from Theos. This allows me to proceed; but further along the way I also discover I must be a derivative entity and not Theos Itself. The answer to the puzzle could be delayed a while along the way, because the problem at hand does not involve calling into question something that must necessarily be presumed true in order for me to make any argument. Eventually I may put together enough pieces to suss out a plausible reconciliation of the data (even if not a deductively certain answer at this eventual stage, though any answers reached deductively before then would still need to be respected as such.)

_That_ is the sort of thing Flew (or any similar proponent along his line--or rather the line he used to take, before converting to minimal deism {g}) needs to be doing in order to properly answer the kind of argument Lewis is ultimately deploying in the 3rd chapter of MaPS. (I say ‘the kind of argument’ because what I’ve described is not quite what Lewis does after chp 3; it’s an improved followup.)

First he has to defend against the fatality Lewis is arguing for, regarding atheism itself per se. (He could pretend Lewis’ argument doesn’t exist, of course, but by tautology that isn’t answering the kind of argument Lewis is ultimately deploying. {g}) Even a mutually assured destruction tu quoque (assuming he established it) wouldn’t get him out of his own epistemic hole: he either needs (specifically in answer to Lewis) to show there isn’t a problem where Lewis is saying there’s (ultimately) a problem, or he needs to show that he can get out of the problem without tripping over a fallacy while making the attempt. (Hint: replying that he can move to a subsequent position and make an argument there to keep the prior position from being a threat _in the first place_ is not the right move. The threat has already been made, and part of Lewis’ point is that any attempt at making a subsequent argument in defense of the prior problem necessarily involves presuming the answer to the prior problem. To insist on doing it anyway would be the same as insisting that it’s okay to presume the explanation supposedly being arrived at.)

Then the atheist, having demonstrated that he is not calling the possible reliability of his cognitive faculties into question by proposing atheism (and perhaps showing along the way that theism does necessarily require that?--this would be a fine riposte!), may continue along; so long as he continues to rigorously avoid that fatality and does not try to fudge past it. Eliminating options on his own side of the aisle (if any) that do introduce that fatality subsequently would be an excellent way to keep on his epistemic feet, for instance.

If Flew or any other atheist can do this, more power to them. I can fairly encourage them to try; because by approaching the topics from this direction, I (and we) would be emphasizing a common agreement and charity between us, without building in an initial prejudice for one or another position on which we are divided. Even agnostics can play--for in fact the process begins with positive (not hopeless) agnosticism on the topics eventually being arrived at for discussion. The agnostic is only forbidden to be agnostic about the possible reliability of his own cognitive processes (with analytical testing of the counter-proposition not to be ignored, of course, for sake of completeness); and is only required to keep to the implications of this necessary presumption, whatever they are discovered to be.

Meanwhile, I obviously agree that I believe the relationship between ‘naturalism’ (in order to avoid topical conflation I insist on ‘atheism’) and our cognitive faculties is epistemologically prior to any _inferences_ (much less mere speculations) about the relationship between evolution and that reliability. It would necessarily have to be, in a discussion about metaphysics, which is the discussion we’re engaging in; moreover, I think it can be reasonably concluded that (ironically despite the derivation of the term) physics as a disciplinary study depends upon metaphysical foundations, not vice versa. In any case, so long as an advocate proceeds to appeal to biological evolutionary theory from within an atheistic context, presuming atheism for purposes of argument, that presumption is up for examination and, if possible, removal. And that’s what Lewis is doing (or trying to do anyway.)

That being said, Darek doesn’t look to be presenting his reply above, in a context that arrives at a position logically prior to the effectiveness of evolutionary theory in accounting for our reasoning behavior. He doesn’t seem to be ending up arguing against atheism per se as untenable before (and thus apart from) discussions about evolution and epistemology. He _looks_ (to me) to be arguing that evolutionary theory (necessarily?) fails to account for important qualities of our epistemic claims. Which is not what I am talking about.

(Darek may be doing something along the line of what I am talking about elsewhere, of course. In fact, for all I know he wasn’t actually responding to the debate between me and Victor, in which case the first part of my first comment wouldn’t apply, of course.)

{{I'd be interested to hear, by the way, how you think your interpretation of Lewis compares with Plantinga's evolutionary argument against Naturalism and particularly with his "Dreaded Loop" as in "Naturalism Defeated?".}}

Insofar as the Bayesian portion of Plantinga’s “evolutionary argument against Naturalism” goes, it should be obvious enough that I must consider that to be quite different from what Lewis is attempting. I suppose this isn’t what you are referring to, but I thought I’d mention it in passing, since it’s far from being not-discussed by Victor. {g}

Regarding the “Dreaded Loop”, though; let me go back and read it again with that question in mind, and I’ll email a reply (since I expect it will be lengthy {self-critical g}). I’ll cc it to Victor, too, and to anyone else who will drop me a line (here or elsewhere) and requests it. (Darek?) It may take several days, though.

Jason Pratt

At 1/25/2007 07:08:00 PM , Blogger Blue Devil Knight said...

Using reason to show how reason could have evolved in an atheistic reality presupposes that the question of atheism either hasn’t been addressed already or else has been addressed and remains in play. Lewis is going down to the root, though, and is addressing the atheism/not-atheism question as a discovered topical priority.

I have no idea what this means.

At 1/26/2007 06:24:00 AM , Blogger Jason said...

BDK: {{I have no idea what this means.}}

It means Lewis isn't really arguing against evolutionary accounts of human cognition. Which in turn means that critiques of him, whether pro or con, that focus on trying to use him for that purpose, are missing what he's actually after. He's going after atheism, not (primarily) after evolutionary epistemology. He mentions the latter for the sake of providing some examples of principle, but it isn't his main target--which is why he can make a free concession about what amounts to the possible reliability of human cognitive behaviors under what we would now call eliminative materialism (e.g. of the Churchland variety.) This by the way is one extremely obvious difference between his chapter and Plantinga's "Naturalism Defeated"--Plantinga explicitly refuses to make such a concession, and intimates that Patricia Churchland would even agree with him. I think it far more likely that PC would agree with Lewis, when he writes, "A conditioning [of one sort that could be proposed by a modern eliminative materialist, in regard to the development and existence of human cognitive behavior] might serve us as well as reason or in some circumstances [even] better."

After that, any interpretation of MaPS chapter 3 that involves Lewis trying to go after evolutionary epistemology per se, is barking up the wrong tree. He is _not_ arguing for (as Plantinga does) a low likelihood of (R)eason corresponding to truth given the truth of (N)aturalism & (E)volutionary theory. He doesn't end up with a likelihood estimate at all; and the formal violation he claims to occur is not (on his part) anything to do with (E) at all. (He does imagine defenses to his argument involving (E), but doesn't counter the defenses on grounds of (E).)

One potential selling-point to this approach, from the standpoint of the subsequent critic trying to decide whether to accept Lewis' conclusion or not, is that it _doesn't_ involve trying to nix biological evolutionary theory per se.

At 1/26/2007 11:44:00 AM , Blogger Darek Barefoot said...

An aside: Jason, I was responding only to Victor's orignal post on his main blog.

Rather than try to deal with all the comments on my post in detail, I will offer another attempt to clarify my own understanding of what Lewis was driving at:

1) Proponents of naturalism as a rule presuppose a correspondence theory of truth.

2) Naturalism, by attributing reason to natural selection, leaves room only for a utilitarian theory of truth.

Now, to me the utilitarian theory of truth has coherence problems of its own. I leave it to others to defend it if they can. But I think Lewis's point is well taken. The proponent of naturalism cannot have his correspondence theory cake, so to speak, and allow natural selection to devour it and leave nothing behind but a utilitarian theory.

I don't have Lewis's book with me as I write this, but he said that naturalism implies that we believe only what is useful. Realizing this, we will be more humble in future and make no grand claims to great truths. Away with such pretensions. But away, also, with naturalism. As I recall, Lewis doesn't explicitly complete the loop at that point and say, "But we only decided to become humble about the truth of our beliefs because of naturalism, which we are now forced to abandon--at least as far as subscribing to it on a correspondence theory of truth." But this is a fairly clear implication of what he says, in my mind.

At 1/26/2007 02:32:00 PM , Blogger Jason said...


Was that the one that linked to his introductory post on this journal? In either case, Victor was mentioning that I think he's missing accounting for the most important part of Lewis' argument, with the discussion being about the part I was highlighting and the use Lewis is making of it (for better or for worse.)

Otherwise, I don't know what you mean by "Victor's original post on his main blog"--that could be any of several other things. (Not that this would be bad; it would only mean I misunderstood what you were replying to in writing the original text of this thread. In which case, as I said, I retract my crit with apologies. {s!})

{{I will offer another attempt to clarify my own understanding of what Lewis was driving at}}

To which I will reply again that even though this can be derived from parts of Lewis' argument in MaPS chp 3, it doesn't seem to synch up very well with the place he arrived at: a challenge that doesn't hinge on criticising evolutionary theory (or 'natural selection' as he tended to call it, in parlance carrying over from the days before the neo-Darwinian synthesis), though he expected escape attempts to involve appeals to evolutionary processes. Ultimately he's arriving at a formal problem deriving from the basic characteristics of 'naturalism'--which, in how he's using the term in chp 3, pretty consistently means 'atheism'.

{{I don't have Lewis's book with me as I write this, but he said that naturalism implies that we believe only what is useful.}}

That was the second line of defense he expected. The _first_ line of defense he expected was one that did attempt to render what we may call a correspondence theory of truth. It's true that in the case of the second defense, his reply was that in abandoning faith in correspondence claims (so to speak) we would also have to abandon the claim of 'naturalism'. In the case of the first defense, his reply was that the defender must by the nature of the defense be attempting (in his words) a proof that there are such things as proofs.

In one case the imagined example of defense involves appeal to evolutionary theory, but Lewis goes back behind that for his counter-defense. In the second case the imagined example of defense doesn't involve a claim to evolutionary processes at all, and Lewis again makes no counter-defense in criticism of an evolutionary justification per se. (It seems clear enough that his second category of imagined defense would easily fall under his prior fallacy-call of trying to prove we don't have to have such things as proofs, if evolutionary theory _was_ called in to help try to establish the point; but it's telling that he _doesn't_ present such an attempt as an example. He presents something "humbler" instead.)

In all cases then, his ultimate challenge is to 'naturalism', not to b.e.t., nor even strictly to epistemologies derived from b.e.t. He imagined an expected reply from an evolutionist (one I've seen duplicated with some frequency when similarly faced with this challenge!), but his counter-defense wasn't about the 'evolutionary' part. It was only about the _justification_ part. In principle, _any_ justification attempt would have failed to make the defense. It just happens incidentally that the proponents of the position he's going after would likely appeal to b.e.t., so that's the example he gives. _Those_ defenders, however, don't think they're only ending up with a utilitarian theory; and Lewis' counter-defense is _not_ that they're only ending up with a utilitarian theory anyway.

Consequently, then, I conclude: if Lewis' ultimate challenge doesn't necessarily require that the imagined defense against his challenge be a utilitarian defense (neither proposed as such by one set of the imagined defenders nor proposed as being such anyway by him in his counter-riposte), then Lewis was not driving at trying to establish that "naturalism leaves room only for a utilitarian theory of truth".

Granted, this might be derived from earlier portions of his argument anyway--indeed I suppose something like that would have to be derived in order to save his argument's relevancy insofar as someone thinks Lewis drops the ball in the final portion (which is Victor's stated position). Even so, I don't find that it can be what Lewis himself was driving toward.

Jason Pratt

At 1/28/2007 03:44:00 AM , Anonymous steve lovell said...


I agree that you and Plantinga aren't doing the exact same thing. All I meant by saying that you and he seem to be going in the same direction was that his "Dreaded Loop" move has a similar logic to that of your interepretation of Lewis's comments on the relationship between evolution and the reliability of our cognitive faculties.
Plantinga excludes certain replies to his EAAN on the grounds that they depend on evidence to which the interlocutor is not entitled, having already had the reliability of their faculties called into question.

I found the line of argument in your response to me rather difficult to follow, and so I'm probably just going to be asking you to repeat things more slowly ...

As you've discovered in others of our discussions I believe that there are both virtuous and vicious forms of circularity. Anyone seeking to explain the reliability of their own cognitive faculties will be involved an a form of circular reason (although it need not be explicit). You seem to think that in the context of the AfR, the naturalist is committed not only to accepting this form of explanation, but that when it is offered it is offered as an argument to remove doubts over the reliability of those faculties.

I agree that such an argument would be dubious, but I just don't think that's how to interpret the move that Flew makes in responding to Gellner (who is responding to Flew responding to Lewis).

Do this mean that he's doing what you say, in your "hint", should not be done? (i.e. "replying that he can move to a subsequent position and make an argument there to keep the prior position from being a threat _in the first place_")

If by "subsequent" here, you mean something more than mere naturalism, then this is exactly what Flew is doing. But why is this different from finding a version of naturalism which isn't undermined by the AfR? You seem to think this is different because the move is to a position which has more than "metaphysical" content. I just don't see the significance of the metaphysics/physics distinction here.

If you think that metaphysical issues should always be settled before the "physical" issues that relate to them are settled, then I disagree. We don't and shouldn't always think in such a top down fashion.

Is it because the "doubts" are being raised by metaphysical speculations are therefore should be dealt with before we leave metaphysics? That sounds more promising, but I don't believe the premise ... it's from experience that we learn that causes and reasons don't normally overlap.

Epistemology and metaphysics are both tricky topics. I've never particular thought about the epistemology of metaphysics before, so do excuse me if I'm saying things that are straighforwardly ridiculous.


At 1/28/2007 12:36:00 PM , Blogger Jason said...


While I've scanned through "Naturalism Defeated" again to freshen up on it, I haven't completed writing out a comparison between it and what Lewis is doing (or what I find him doing anyway {s}) in MaPS chp 3. Consequently, I'm not yet ready to comment on similarities and differences (if any) between Lewis' climactic summary portion of chp 3 and Plantinga's discussion of the Dreaded Loop move--aside from mentioning that the DreadLoop was originally brought _against_ Plantinga as a criticism _against_ his EAAN. The first question, then, would be whether the DreadLoop nixes Lewis' climactic argument; and if not, does Lewis escape the crit for reasons similar to how Plantinga claims to escape it (insofar as he does claim to)?

I'm not yet prepared to answer that, however. I think it's a fair and interesting enough question, though.

{{As you've discovered in others of our discussions I believe that there are both virtuous and vicious forms of circularity.}}

I was wondering when we would be going to that. {lopsided g}

I believe Lewis' procedure avoids circularity; and I have been very careful to do the same in applying and filling out his argument. Indeed, his whole climactic point rests on calling coup against expected circularity in the defenses.

Similarly, I believe I have a responsibility to my opponents _not_ to engage in circular argumentation: venial, virtuous, or otherwise.

{{Anyone seeking to explain the reliability of their own cognitive faculties will be involved an a form of circular reason}}

I think not. There is a crucial difference between trying to explain how and why our faculties are possibly reliable, and trying to explain _that_ our faculties are possibly reliable. We presume _that_ our faculties are possibly reliable in order to explain _why_ our faculties are possibly reliable (insofar as we can discover this.) This is not a circular argument, virtuous or otherwise.

I think the atheist, in the context of atheism per se, is committed to having to defend _that_ our cognitive faculties are possibly reliable. Any specific explanation attempt given, subsequently, presumes the truth of the notion being threatened. They might as well stop with the sheer presumption that our cognitive faculties are reliable _despite_ being fundamentally non-rational.

_That_ could be an interesting defense; but it will always look weak compared to the fact that theism doesn't involve immediately introducing a challenge to the necessary presumption. If I am going to be required, by holding a position, to sheerly assert an affirmation in the face of a necessary challenge of that affirmation, with no hope of ever being able to do any better, then personally I prefer to go with not-position (so long as it doesn't necessarily result in a similarly inescapable challenge) and work from there. Given a choice between discovering a worldview to believe that requires immediate willed contradiction on my part in order to continue, and one that does not, I go with the latter. Because if I went with the former, I would surrender all claim to make truth claims that were anything other than my own positive assertion.

Of course, I could follow out that position and simply positively assert otherwise again...{g} But then, so much for me arguing to a conclusion as a reconciliation with anyone else. We might as well sheerly assert whatever we happen to feel like, whether we agree with each other or not.

Literally for sake of argument, then, I find I should not accept atheism to be true. Which, in brief, is what I have been saying all along is the functional and practical conclusion to Lewis' chapter 3 argument.

Thanks, btw, for clarifying the Lewis/Flew/Gellner/Flew trail. Flew may be very properly answering Gellner (at least in principle, aside from how effective his actual content is.) Flew is _not_, however, properly answering Lewis, if Lewis is aiming at atheism per se, in the way I find (and am claiming) Lewis is doing.

{{If by "subsequent" here, you mean something more than mere naturalism, then this is exactly what Flew is doing. [JP: we're in agreement, then, on this at least! {g}] But why is this different from finding a version of naturalism which isn't undermined by the AfR?}}

The short answer is that _if_ Flew's answer _is_ (or even can be) appropriate, it will have had nothing to do with happening to find a version of 'naturalism' which isn't undermined by the AfR. It will have had everything to do with denying that (borrowing Lewis' language) we cannot legitimately prove that there are such things as proofs. After that, the form of the subsequent defense is practically incidental, insofar as Lewis' argument goes. We might then proceed to debate whether Flew has in fact found, in this or that evolutionary argument, such a proof that there are such things as legitimate proofs; but having first surrendered a basic logical fallacy, I would not be personally much interested in the subsequent debate. Any and all positions (not flatly self-contradictive in their statements perhaps) can be initially presumed and then be inferred to a conclusion of the truth of that presumption by relying on the truth of that presumption thereby. Big whee. The only question is whether the particular form is valid; and the form of such an 'argument' can _always_ be made valid by reducing to sheer tautology: (1) X is true; (2) therefore X is true. {shrug} After which, we might as well stick with sheerly asserting whatever position we happen to feel like asserting at the moment, thereby saving time and energy for playing Dominions 3. {g}

{{We don't and shouldn't always think in such a top down fashion.}}

This being, itself, a metaphysical and not a "physical" position to take, of course. {g} But my complaint was (as noted above) somewhat more basic than "metaphysics should always be settled first."

It ought to be obvious that for any scientific procedure, certain metaphysical issues _are_ always accepted as settled first. Despite what BDK was wondering, I am not a logical positivist. Where any such metaphysical issue necessary for science is being called into question, though, science is in no position to resolve the question without (to put it bluntly) cheating.

Jason Pratt

At 1/29/2007 12:07:00 PM , Anonymous steve lovell said...


I worded my comments badly on circularity and justification of belief in the reliability of our faculties. I agree completely with your comments in your last post on this thread.

I'm still bothered, however, that we seem to be talking past one-another and think we may have different conceptions of the dialectic.

If the argument is supposed to show that naturalism is self-defeating, then surely the naturalist can assume anything he likes in addition to naturalism, this doesn't allow him to simultaneously be arguing for naturalism, but then I've never imagined that he was trying to do that in this context. He is only trying to show that naturalism isn't self-defeating, and can pull in any auxilliary assumptions he likes.

Consider as a parallel Plantinga's appeal to actions of demons as an explanation for apparently "natural" evils. Plantinga isn't even committed to the truth of this auxilliary hypothesis, but the very possibility of such agency is sufficient to show (given other things in the context) that the "natural" evils we see are not logically incompatible with the existence of God.

If the argument is not intended to show that naturalism is self-defeating, then of course the probability of the auxilliary hypotheses becomes relevant, particularly the probabilities of evolution (E) on the assumption of naturalism (N) and the probability of the reliability (R) of our faculties on the assumption of both naturalism and evolution. Both have what I call metaphysical and epistemological probabilities (the first being the probability that something would be the case, then second that something is the case), and depending on how the argument is supposed to go either one may be relevant. I must admit to not always being sure which is being used or which is relevant when reading Plantinga and others on these issues.

It seems fairly clear, however, that for your moves to be right you need to be thinking about epistemological probabilities (at least some of the time), and then your claim seems to be that we can only assess P(E|N) and P(R|E&N)
if we are entitled to assume the reliability of our faculties, and the naturalist is required to get decent evaluation of these propositions but is not entitled to assume the reliability of our faculties. Obviously the requirement and the embargo are related, but where do they come from?

Is it because the probabilities, interpreted metaphysically, are low, and as such cast doubt on the reliability of our faculties?

If this is how the argument is supposed to go then it looks very like Plantinga's EAAN.

Unfortunately, the issue of how much the naturalist is allowed to build in to (E & N) is hotly debated. There's also a fair amount of debate over whether there might be a similar problem for theism when it is conjoined with the doctrine of the fall.

I still feel we may well be talking past each other. Am I getting any warmer?


At 1/30/2007 10:45:00 AM , Blogger Jason said...


For the atheist to assume anything he likes in addition to atheism, as auxilliary assumptions to that assumption, certainly means that he is not simultaneously arguing _for_ atheism. It does however mean, by tautology, that he is assuming atheism; thus that he is also assuming atheism is not self-defeating.

To simply assume that Lewis is wrong, and then go on to use atheism as a framework within which to propose consequent beliefs within that framework, is not to answer Lewis' argument; except (tautologically) insofar as 'simply assuming Lewis is wrong' may be considered 'an answer'.

Again: to simply assume that Lewis is wrong, and then go on to use atheism as a framework within which to propose consequent beliefs within that framework, is to assume that atheism is not self-defeating. The atheist is free to sheerly assume this and dismiss Lewis ad hoc; but then, aside from not really answering Lewis, neither can the atheist proceed to try to _thereby_ show that atheism isn't self-defeating. Not without engaging in a circular argumentation; one even you would probably regard as vicious.

Yet again: if the atheist's strategy to argue that atheism isn't self-defeating involves presuming atheism isn't self-defeating, then really he might as well just stop with the presumption and not waste his or anyone else's time with an argument ostensibly designed to arrive at the same position. Unless what he is trying to do is impress the credulous or the unskilled that he is accomplishing something more than sheerly presuming his position (i.e. that atheism isn't self-defeating--in the face of an actual argument saying otherwise, btw, which is _not_ beginning from that position as a presumed truth upon which the argument subsequently depends.)

Or, again: if the 'naturalist' is trying to defend that 'naturalism' is not self-refuting, by appealing to an argument involving evaluations of P(E|N) and/or P(R|E&N), then by your own description he is appealing to an argument that requires presuming a truth (for sake of current argument) that itself must entail what he is trying to defend--that 'naturalism' must not be self-refuting.

He _is_ entitled to assume R, and put _that_ after his vertical bar. Indeed, he necessarily has to! In many circumstances, it would be fine for him to assume N after that vertical bar, in order to evaluate P(E|N) for instance. He is _not_, however, entitled to assess N|N, or N|E&N, or even N|E&R&N, or N|N&anything. And he is _not_ entitled to assess R|anything, either, because (unlike N) R is necessarily going to be _after_ that vertical bar in any case, whether it happens to be mentioned there or not (such as when evaluating P(E|N)).

If you agree completely with what I said in my most recent comments in this thread, then that would seem to include agreeing that we have a responsibility _to our opponents_ _not_ to argue in a circular fashion. Lewis is not doing this. Neither am I (following and expanding on his lead.) The atheist you are describing _is_ however arguing in a circular fashion, trying to get out of (or past) Lewis' argument--which, not-incidentally, is exactly what Lewis expected would happen.

Leaving aside a detailed analysis of the Plantinga example you gave for illustration, I think it is sufficient to point out that it is not (as given) parallel to the atheist defense you described (as given).

Plantinga's auxilliary hypothesis is not a presumption necessary for every argument; he is not trying to encourage (or reassure or head off a doubt) about a belief in that hypothesis; and you haven't yet tried to claim that the position he is aiming for (i.e. that the natural evils we see are not logically incompatible with the existence of God) is something he is presuming for sake of his argument.

Now, if you told me that there was a sceptical argument to the effect that the "natural" evils we see are logically incompatible with the existence of God; and that Plantinga in reply to this was assuming for sake of argument that demons intefere with nature as an auxilliary to also assuming for the sake of his argument that the natural evils we see are logically incompatible with the existence of God; _then_ we would have a parallel with the process you have described the atheist doing as an answer to Lewis.

But then, if Plantinga was proceeding by assuming for the sake of his argument that it is true that the natural evils we see are not logically incompatible with the existence of God--then why is he bothering to make some kind of argument toward the same truth? Not for his own sake--he already presumes it to be true. (Though if he is not prepared to be committed to the truth of his auxilliary hypothesis in order to save his position, I don't know why he would be prepared to be committed to the truth of his primary hypothesis even in order to save his primary hypothesis by arguing for what he is primarily hypothesizing... {shrug!}) For the sake of an opponent? But for an opponent to be 'moved' toward accepting the position being argued _for_, the opponent must first accept for sake of argument the position being argued _for!_

On the other hand, if Plantinga is _not_ beginning by presuming for his argument that the natural evils we see are not logically incompatible with the existence of God, then he might arrive at this by legitimate argument.

(Though, incidentally, if he is not prepared to believe his premises in order to arrive at this, then why should an opponent? Especially when his auxilliary premise is something a sceptic is going to be _at least_ as sceptical about in the first place! If I am asked to assume giant spaghetti monsters exist in order to demonstrate that something I think is logically incoherent about, say, evolutionary theory isn't in fact logically incoherent, well... {wry g})

In which case (not counting the incidental digression), I recommend rephrasing your description of the atheistic defense to Lewis' argument, so that you are not implying the atheist is primarily (or even auxilliaryly! {g}) presuming contra-Lewis in order to build an argument supposedly arriving at the same contra-Lewis position.

{{If the argument is not intended to show that naturalism is self-defeating, then of course the probability of the auxilliary hypotheses becomes relevant}}

This I agree with (in principle), and always have. Though it would be fairer for you to stick with phrasing this in terms of _likelihood_ instead of _probability_. The probability of such things cannot be calculated. The likelihood can be intuitively estimated. Calling it "probability" makes the result seem stronger than it actually is. (I do the same thing sometimes from habit, but I try to watch against doing it unless I really mean probability. It may not make much difference usually, but in _this_ case it makes a real difference.)

However, I do not believe Lewis is ultimately attempting this kind of argument. He wouldn't have climaxed it in the fashion he did, if he had been going that direction.

{{It seems fairly clear, however, that for your moves to be right you need to be thinking about epistemological probabilities (at least some of the time)}}

No. What is happening at the end of Lewis' argument is purely formal, and has nothing in its challenge to do with probabilities or likelihoods at all (epistemological, metaphysical, whatever.) A probabilistic (likelhoodic {g}) defense could be imagined, but it would be thereby barking up the wrong tree. And that's even completely aside from the question of whether the defense is falling into the trap set by Lewis.

The same goes for my own appropriation and expansions of Lewis' AfR. I can present an imagined dialogue between a proponent of what Lewis is (I think) actually doing, and a defender against that argument, where the defender is making probability claims all over the place. But those probability claims make no difference at all, because they still require formally (if tacitly) presuming what the attempts are supposed to be defending, in order for them to even appear to be working for that defense.

{{your claim seems to be that we can only assess P(E|N) and P(R|E&N) if we are entitled to assume the reliability of our faculties, and the naturalist is required to get decent evaluation of these propositions but is not entitled to assume the reliability of our faculties.}}

This, on the other hand, is more-or-less correct as a description of the argument I'm finding in (and using from) Lewis. It's put a little backward as to how it would be arrived at, but it isn't strictly incorrect.

For my argument to be right about this, however, I do not _need_ to be thinking about probabilities or likelihoods.

The embargo, and the mainspring of the argument, comes from |A ('given the truth of atheism'.) And, incidentally, seeing as how the common use of 'naturalism' in the field can entail one or both of two categorically different claims--something both Lewis and Plantinga exhibit--and since I am only discussing one particular claim among those two, could we _please_ for accuracy's sake stick with the claim I am actually talking about in distinction from the claim I am _not_ talking about? It may be promotionally or argumentatively convenient for an atheist to talk about naturalism when he means atheism, but we don't have to do that. (Or, do _you_ have a reason to be jumping over to the category of 'no supernatural entities exist'--a category which need not involve a denial of theism any more than supernaturalism per se necessarily involves a denial of atheism?)

Lewis is saying (in effect): given the truth of atheism, the possible reliability of our cognitive processes is necessarily called into question, and we cannot get out of this without presuming what we need to subsequently defend. The only escape from this is to simply presume the possible reliability of our cognitive processes, which happens to be something we necessarily have to presume anyway in order to make any argument at all.

The atheist is entirely free to presume _that_, and indeed _must be_ presuming that, just like the theist. Neither the atheist nor the theist, however, is free to presume this _and also_ to propose a truth claim that necessarily calls the truth of this presumption into question. (Or, rather, that kind of presumption could perhaps be made, maybe; but nothing at all could be legitimately done with it in subsequent argument to any position.)

Lewis points out that the proposition of atheism puts us in precisely that position of having to question and (if possible) defend the necessary presumption--a presumption that will end up being necessarily presumed anyway in order to make that defense. Therefore, we shouldn't propose atheism for belief. We should propose not-atheism instead. (By "shouldn't" I am here, and previously on this topic, meaning a logical recommendation, not necessarily a moral one; even though I do consider moral recommendations to be also logical ones.)

As part of this, it could and would be pointed out that in our experience we find it right sometimes to attach little-to-no value to beliefs which follow as a result of non-rational behavior. No likelihoods need be appealed to at this point, however; it isn't necessary to do so. I don't have to point out that even if atheism is true rational-from-non-rational behavior must still be rather rare (if it happens at all); and even if I did, the atheist could rightly reply, "Solar fusion is rather rare, too, compared to other behaviors in the universe. Nevertheless, there's the sun!"

So I don't have to mention likelihoods, and I wouldn't really gain anything if I _did_ mention them. All I have to do is ask an atheist to suppose that my beliefs about Christian truth are the result of knee-jerk reaction to stimuli that have pummeled my mind since childhood, and then ask him if he is willing to go be baptized into an honest belief of the same truth claims, on those grounds. Even if he answers, "Well, maybe if I found out you were exceptionally effecient mouth-breathing imbeciles! {g}", he still is appealing to a principle that we ought to question to possible exclusion the propriety of notions derived from non-rational behaviors. But if atheism is true, _all_ our notions (including that atheism is true) derive from non-rational behaviors, whether or not those behaviors somehow become 'rational' later.

Formally, then, we must find ourselves in the quandry Lewis describes, _if_ atheism is proposed to be true.

And that is ultimately the point of his argument.

(Which is only superficially similar to Plantinga's EAAN. As to how it relates to the Dreaded Loop, we'll see. {s})

Jason Pratt

At 1/30/2007 12:30:00 PM , Anonymous steve lovell said...


Sorry to continue to drag this out.

I feel I'm obliged to point out a simple logical theorem:

If a set of propositions {P1, P2, P3, ..., Pn}is logically consistent then, any proper subset of that set is also logically consistent.

Now, if it's true that the set of propositions (P1) to (P4) is logically consistent, then so are (P1) and (P4) taken without (P2) and (P4).

(P1) Naturalism is true
(P2) Evolution is true
(P3) Evolution makes probable the reliability of our faculties
(P4) Our faculties are reliable

Moreover, if the assertion of (P1) to (P4) is not self-defeating, then the assertion of any subset is also not self-defeating.

I simply can't see how you can claim that accepting (P1) necessarily undermines belief in (P4) while also saying that in the presence of (P2) and (P3) this undermining doesn't take place.

This is exactly the logic of Plantinga's point about demonic agency. If the logical problem of evil (or the version of it in play here) is supposed to show that certain evils are logically incompatible with the existence of God, then the mere possibility of demonic-agency being a sufficient explanation for those same evils is sufficient to undermine the argument. (Of course that doesn't mean that all other versions of the argument from evil fail, only the purely logical argument from those particular evils.)

Perhaps we mean different things by "self-defeating"?


At 1/30/2007 12:48:00 PM , Blogger Victor Reppert said...

Jason: Perhaps a numbered-premise list of the "Jason-style" AFR would be helpful here.

At 1/31/2007 06:00:00 AM , Blogger Darek Barefoot said...


I read Victor's comment on DI to the effect that he had his doubts about the particular argument we are discussing. I tried to post to that blog, had technical problems and emailed Victor directly whereupon he
entered my post to DI2.

It seems to me that there are two questions in play in the relevant section of Lewis's book. One is, "How can we justify our mental processes once they have been called into question?" The other is, "What kind of explanation of our mental processes is consistent with the reliability which we must assume them to possess?" The first question has no answer and makes no sense. If the naturalist were to try to justify his mental processes on the grounds of natural selection he would be in a vicious circle. Ditto for the theist if he tried to use God's existence the same way.

The second question is coherent and worthwhile. But, Lewis points out, proposing natural selection as the archtect of reason is not a coherent answer. To use his words, "[the naturalist] a history of the evolution of reason which is inconsistent with the claims that he and I both have to make for inference as we actually practise it." Why inconsistent? That is what I was attempting to show somewhere up above.

I don't believe that naturalists generally justify their mental processes by recourse to natural selection. They think they are simply giving an account that is consistent with the assumed reliability of human faculties.

Critical to Lewis's argument in this place is that he has already demonstrated that reason cannot be constituted by blind mechanism. Assuming that his demonstration in that regard has been effective, natural selection cannot ride to the rescue, as it were. Perhaps this is at least generally the direction in which you too are heading.

At 1/31/2007 07:15:00 AM , Blogger Jason said...


That would be a good idea, and I'll work on that for submission presently.


In order to judge the cogency of your P1 to P4, we need to be supplied with the actual _logical relationship_ being supposedly advanced as self-consistent among those propositions.

For example, it is true that the _assertion_ of

(P5) I always lie


(P6) I always tell the truth

is not self-defeating. That is because no logical relationship is being given in the mere assertion of these propositions. If I change premise 6 to

(C0) Therefore I always tell the truth

now there's a logical inconsistency.

If I add any kind of (C0) as following from (P5) and (P6), _now_ there's a logical inconsistency.

The relevant question here is whether (P4) follows as a logically consistent _conclusion_ (not stated simply as another premise or proposition) from premises P1..P3.

So at least rephrase the elements of the set as an argument, or supply a subsequent conclusion to the set, if you wish to test the set's logical consistency per se. (Unless you are simply including a tacit 5th premise, "Set P1..P4 is logically consistent." {g}) For instance, at this time I cannot tell if P4 is supposed to logically follow from P1..P3. If the fourth listed premise is being put into an (hopefully logically consistent) _argument_, though, it really needs to be put back at P1 (or P0!) That premise, or some minimal variant thereof, is necessarily being presumed for any argument involving P1..P3.

I will thus suppose, for instance, that you are not asking me what the problem would be for a naturalist (as a human engaging hereby in a cognitive process) to proceed:

(Necessary P0) Human (including my) cognition is possibly reliable.
(P1) Naturalism (whatever this is taken to mean) is true.
(P2) Evolution (whatever this is taken to mean) is true.
(P3) Evolution makes probable the reliability of our (human) faculties.
(C1) Our (human, including my) cognitive faculties are possibly reliable. (From NecP0, P1, P2, P3)

If you are not asking me what the problem is with this, I need more information about what you _are_ asking me about, in regard to your premise list.

(Whereas, if you _are_ asking me what, in effect, the problem is with this--then we have more basic matters to discuss. {s})

For what it's worth, _apart from Lewis' argument_ (or a similar variant thereof), I would have no problem with the following _argument_ (not just premise list) from an atheist, as being logically proper (consistent and non-circular).

(Necessary P0) Human (or my) cognition is possibly reliable.
(P1) Naturalism (whatever this is taken to mean, including atheism) is true.
(P2) Evolution (whatever this is taken to mean) is true.
(P3) Evolution makes probable the reliability of our (cognitive) faculties (even under naturalism).
(C2, distinct from C1 already listed above) Our (including my) cognitive faculties are probably as well as possibly reliable. (From NecP0, P1, P2, P3)

Notice that in phrasing it this way, I have given the widest scope to what would be entailed in "the reliability of our cognitive faculties." It's the same kind of scope Lewis allows in _his_ argument, where _agreeing_ with Flew about the potential effectiveness of cognitive development via biological evolution in an atheistic reality (insofar as the quote previously referenced from Flew is concerned.) It's also the sort of thing that (apart from Lewis' argument) I have been agreeing all along would be acceptable for the atheist to do.

Jason Pratt

At 1/31/2007 07:39:00 AM , Blogger Jason said...


I agree that atheists do not _generally_ justify their mental processes by recourse to natural selection. They think they are simply giving an account that is consistent with the assumed reliability of human faculties. (Or rather, with the assumed _possible_ reliability of human faculties. There are practical reasons for why I keep including that qualifier in my descriptions.)

Nevertheless, take a good look at the P1..P4 set being reported by Steve up there. That kind of defense (with P4 actually being a conclusion) _is_ common when the atheist is faced with certain challenges following necessarily from the implications of atheism. (Which is why Steve is trying to present it. He knows from experience that the defense is common, and wants to know why it doesn't work, including as a sufficient answer to Lewis.)

It is precisely _those_ implications Lewis is drawing attention to, and building his climactic summary on (as a formal challenge to the propriety of holding _atheism_), when he writes, "[the naturalist] a history of the evolution of reason which is inconsistent with the claims that he and I both have to make for inference as we actually practice it."

I honestly don't know for sure that Lewis has demonstrated (by the time he reaches the climax) that reason cannot be constituted by blind mechanism; he might or might not have, and while his gist is certainly in that direction (at the least he seems to be assuming it), if it came down to brass tacks I might have to acceed that he has not in fact demonstrated this.

It should be noticed, though, that his climactic summary portion _does not require this_ to have been demonstrated for certain yet (much less presumed to be true). If he ends up not relying upon that position, then I expect he wasn't attaching primary weight to trying to establish it.

Jason Pratt

At 1/31/2007 12:43:00 PM , Anonymous steve lovell said...


To be logically consistent or inconsistent a set of propositions do not need to be so related as to form an argument.

Moreover, your (P5) and (P6) while not strictly contradictory are easily brought into contradiction. If we add (P7) we then have.

(P5) I always lie.
(P6) I always tell the truth.
(P7) I have said something.

The would allow us to infer the following two contradictory propositions

(P8) The something I have said is false (from P5 and P7)
(P9) The something I have said is true (from P6 and P7)

And therefore (P5)-(P7) form an inconsistent set of propositions even though they are not themselves structured so as to form an argument.

My (P1) to (P4) weren't intended as an argument but simply as an exposition of the position which you believe is self-defeating.

This set of propositions is self-defeating if and only if it is inconsistent (ie we can use these proposition and other necessary truths to deduce two contradictory propositions) or from these propositions and other necessary truths we can deduce that it is irrational to accept one or more of these propositions.

To say that a position is self-defeating is precisely to claim that it doesn't hold together taken on its own terms, it is to attempt to bring that view down from within, or to provide an immanent critique. But the position you are trying to bring down from within involves all of (P1) to (P4), not just (P1) or (P1) and (P4).

If we cannot deduce a contradiction from (P0) to (P4) and we also cannot deduce from those same premises that it is irrational to accept one or more of (P0) to (P4) then neither can we deduce this from any subset of those propositions.

As an aside, there is something odd about saying that your argument (P0)-(C1) is bad but (P0)-(C2) is okay. I agree, but the oddity is that with the same resources the second attempts to establish more than does the first. Normally, less ambitious arguments are the more likely to succeed. Of course this is only a necessary truth in the sense that less ambitious arguments are more likely to be Sound, and (P0)-(C1) is certainly more likely to be sound than (P0)-(C2). Still, for anyone who doubts the conclusion of these arguments, the second could be persuasive where the first could not (since to doubt the conclusion is to doubt one of the premises). This is why the second is a better argument.

What this aside shows is the importance of the dialectical context of an argument in assessing the merits of that argument. I think that an argument which attempts to show that some position is self-defeating is necessarily an argument in a certain dialectical context: a context which allows the defender of the allegedly self-defeating position to assume rather more than you seem to be allowing. This is why I wonder whether we have different notions of "self-defeat".


At 2/02/2007 08:00:00 PM , Blogger Darek Barefoot said...

To all

I offer the following for consideration:

1. The advantageousness of a belief in the struggle for survival is significantly distinguishable from the truth of the belief.

2. Nature selects biological processes and products solely for their advantageousness, though they may occur originally due to blind chance.

3. The human belief that advantageous beliefs tend to be true is a biological product.

4. Therefore, the belief that advantageous beliefs tend to be true occurs either by blind chance or is selected by nature solely because of its advantageousness.

5. That which is believed either by blind chance or solely because it is advantageous is not justified.

6. Therefore, the belief that advantageous beliefs tend to be true is not justified on 1 - 5 above.

7. Since the belief that advantageous beliefs tend to be true is justified, at least one of 1 - 5 is false.

I should quickly point out that to believe a proposition because it is advantageous is not the same as believing that the proposition is advantageous. It is also not the same thing as believing that the proposition is advantageous and therefore that it must be true. These are critical distinctions. The above sequence presumes that it is possible to believe solely because the belief is advantageous from the standpoint of a reasonable, objective observer. I am not sure that "belief" under such constraints is even possible, hence my earlier observations. But here we will use the word belief broadly. Perhaps ants "believe" that they should build air traps into their nests because, objectively speaking, building nests with air traps promotes the survival of ants. In any case, ants do not subscribe to a proposition concerning air traps, but something motivates them to build air traps and we will label that something a belief.

To me, the most obvious point of doubt for the naturalist is No. 1. Perhaps the argument relies on a specious distinction between the general and the specific. For example, because ants seek out water we might say that ants exhibit "water-seeking behavior." But we cannot make a fundamental distinction between water-seeking behavior and advantageous behavior in general; water-seeking is part of advantageous behavior. Perhaps it is the same with true-belief-seeking behavior.

Attacking No. 1 is not so easy, however. Water-seeking behavior may be distinguishable to some degree from advantageous behavior, but it is not significantly distinguishable as an explanation for ant behavior. That is, water-seeking is a naturalistic explanation for ant behavior only to the extent that water-seeking is advantageous. The same cannot be said for evaluation of beliefs. We cannot hold a belief solely to the extent that the belief is advantageous and still be justified in holding it. For one thing, a belief can be considered justified only to the extent that what the holder perceives to be the reasons for his holding the belief are in fact the reasons for his doing so. None of us perceives himself to hold beliefs solely to the extent that they are advantageous to our survival or the survival of the human species.

At 2/03/2007 02:13:00 AM , Anonymous steve lovell said...


I'd have thought the naturalist ought to deny 3. Beliefs are not the result of biological processes unless those beliefs are "hard wired" or "innate" in some sense.

Rather beliefs are the product of faculties which are themselves the result of biological processes. It's because the faculties are "adaptive" that they get selected.

Of course, for the faculties to be adaptive requires that they produce beliefs which lead to behaviour which is generally adaptive ... but individual beliefs need not be adaptive.

Of course, you might then offer your argument at the "next level up", arguing that the reliability of faculties is distinguishable from their adaptiveness, and that we should not trust the product of faculties which were selected for adaptiveness rather than reliability.
However, this formulation of the argument will need some other modification too, ... what will do the work of your premise 4?

(4) Therefore, the belief that advantageous beliefs tend to be true occurs either by blind chance or is selected by nature solely because of its advantageousness.

When considered as the product of a biological product rather than as biological products themselves, the naturalist has more room in which to put other possible grounds for the belief which are the occasion for the faculties producing the belief, and these extra grounds may be enough to make the belief justified.

Could you have a go at reformulating the argument to avoid this? I think it may be possible, I just don't see how right now.


At 2/05/2007 01:33:00 PM , Blogger Jason said...


The thread ran off the bottom before I got back to it; thus the delay. (Plus I spent Saturday working on finishing a comparison notesheet for Lewis/Plantinga. Or not finishing yet, exactly--I still need to carefully re-read the second half of Plantinga's article to see if he ever provides an escape for a trap he has set for an opponent that looks unavoidable to kill his own position, too.)

{{To be logically consistent or inconsistent a set of propositions do not need to be so related as to form an argument.}}

This is rather like saying that in order to be logically consistent or inconsistent, a set of propositions does not need to be presented in any putatively logical relation. I don't like sort of vaguely presuming logical relationship claims entail between propositions. Either spell out the logical relationships you mean in a set, or else I will treat the premises as being logically unrelated (neither consistently nor inconsistently related.)

{{Moreover, your (P5) and (P6) while not strictly contradictory are easily brought into contradiction.}}

I know. I did it myself, by supplying an ostensible conclusion, creating a logical invalidity: either there is a falsity in one of the premises, or else the conclusion doesn't actually follow from the premises.

As one alternate, I could have combined the premises into a single composite premise; and it would have been logically self-refuting. But only insofar as a logical relationship between them was being tacitly proposed. Where a logical relationship is being proposed among premises, whether stated separately or in composition, that relationship may be judged to be logically consistent or inconsistent.

Or, as another alternate illustration of what _I_ said, I could have set it up the way _you_ set it up. Is it not sufficiently obvious, though, that if you draw inferences from those premises, to mutually exclusive conclusions, thus demonstrating a logical inconsistency, you _have_ structured an argument?? You had to make an argument concerning (P5..P7) in order to demonstrate that the propositions are logically inconsistent (even to functionally claim such a thing, really); even though (P5..P7) taken as themselves are not structured to form an argument.

Very well then--that is another way of demonstrating exactly what I said. So, please provide the logical relationship you are claiming is consistent for (P1..P4). Otherwise, I am going to treat them as logically unrelated propositions. Which means I will not consider them to be logically consistent (nor inconsistent, either, admittedly.)

Whereas, you seem to think that if no logical relationship is proposed concerning them, then they must be logically consistent (since they cannot be logically inconsistent without a logical relationship claim.) No claim of logical relationship at all is made, therefore they are logically consistent in their relationship with each other?? (This is what we would call a self-contradictory proposition, btw. {g} I mean, assuming we are going to think logically about the elements of the proposition.)

When you move on to discussing (P0) through (P4) (instead of (P1) through (P4)), then I will suppose you are including my premise "(P0) Human (including my) cognition is possibly reliable."

Now, if no logical relationship is being proposed among (P0..P4), then of course I agree that we cannot deduce a contradiction or anything else from that set or from any subset of those propositions. Indeed, if no logical relationship is being proposed among them, then I suppose it really makes no difference what order they are related in, either. But neither will I acceed that they are (much less thereby are) logically consistent as propositions.

If a logical relationship is being proposed among them, though, then I want to know what that relationship is. I gave an example of what seemed to me to be the most obvious relationship that might be attempted. If that wasn't the logical relationship you were thinking of, then once again I invite you to provide the relationship yourself. Or, alternately, you may wish to dispute my inclusion of (P0); you are welcome to try, for the sake of exercise, although I don't recommend getting your hopes up about removing it. {g}

So, then: is my problem with (P0..C1) (where P4 is presented as a conclusion from prior premises, as part of a claim/test for logical consistency) that a contradiction arises? Nope. Is it that it is irrational (whether in the Lewisian sense or the Plantingian sense, to borrow a topically relevant distinction) to accept one or more of the premises? Nope. Is it that it is irrational (Lewisian or Plantingian) to accept the fourth element as a conclusion? Nope, my complaint isn't even that. C1 is admittedly a tautological conclusion from P0; with the other elements adding nothing and so being completely trivial--anything at all could have been put in their place, or nothing. It doesn't even matter whether any of the intervening elements are true or not. The intervening premises could have just as easily been:

(P1) "Bablyon 5" is a science-fiction/fantasy television series. (Which, incidentally, is true.)

(P2) Jason Pratt was the creator and chief writer for "Babylon 5". (Which, incidentally, is false.)

(P3) The new contractor for Bittersea Publications is competent at providing cover art for Jason's new novel. (Which, incidentally, I have no idea about yet.)

The nice thing about an argument where the conclusion is equal to its necessary premise, is that anything at all can be stuffed in between without regard to logical coherency. Which also means, by the way, that my critique of such an argument has nothing at all to do with "bringing that view down from within."

(Qualification: had I included a premise between P0 and C1 that involved proposing a mutual exclusion to either of those, then I couldn't have legitimately said that C1 follows from the inclusion of that premise. Otherwise, heck, I could have included as the three intervening premises, "I always lie", "I always tell the truth", and "I have said something"!)

{{As an aside, there is something odd about saying that your argument (P0)-(C1) is bad but (P0)-(C2) is okay. I agree, but the oddity is that with the same resources the second attempts to establish more than does the first.}}

I suppose that's one way to put it. Normally, arguments do involve trying to establish more than just restating the first premise over again as a conclusion. The oddity here is that you think this normal goal is odd! (The first argument is 'bad' because it's a total waste of time and energy, not to say actively misleading. The proponent might as well have stopped with asserting (P0). Dang skippy (C2) should be "more ambitious" than that!)

{{(P0)-(C1) is certainly more likely to be sound than (P0)-(C2).}}

Uh... yeah, I guess that's one way to put it, too. In fact, (P0..C1) as given is _certain_ to be sound. That's because the intervening premises don't make any real difference to the conclusion, which is tautological from P0. This kind of certain-soundness is no credit to the argument, though. On the contrary, _this_ kind of certain-soundness is what kills it as an argument.

{{Still, for anyone who doubts the conclusion of these arguments, the second could be persuasive where the first could not (since to doubt the conclusion is to doubt one of the premises). This is why the second is a better argument.}}

That, and the first argument might as well just have stopped with the mere assertion of P0. The stuff in the middle adds nothing at all to the tautological conclusion. At most, the stuff in the middle could serve to fuddle the slow-witted into being impressed that something was accomplished other than just stating P0 again. Some of us believe we would be sinning to fuddle the slow-witted; for instance, I would call it 'cheating'.

Having said all that, my problem is not that in making a defense (P0..C1) the atheist would be making a self-defeating argument. My problem is that he might as well not be making an argument at all. I _might_ have problems with (P0..C2), when it came down to brass tacks--I would want to see proper establishment for (P2) and (P3) at least, beyond mere assertioning, as well as for (P1) where/if possible. (Though I'd be willing to cut a bit more slack on that one.) But I'm willing to agree the attempt is, in principle, formally legitimate. The first attempt is not. That's why circular reasoning is considered to be a _fallacy_.

(P0..C2), however, is not an answer to Lewis' challenge; which is that if atheism is proposed the proponent has to hold P0 _despite_ the implications of atheism. (P0..C1), on the other hand, gives exactly the kind of defense Lewis was expecting: a circular justification attempt. Which is a fallacy. It doesn't matter what's in between (P0) and (C1)--the attempt is still circular reasoning; either in ignorance of procedure violation (which I suppose is usually), or in an attempt to fuddle the slow-witted (as I put it above.)

When the proponent has to necessarily presume the position he is trying to defend by argument, then in _that_ sense the proponent is defeating himself.

This might be a good time to point out that in all the discussion above _I_ have never been claiming that atheism is self-defeating (per se). (Plantinga claims Naturalism is defeated, by which he means atheism contextually everywhere except in his opening definition of naturalism where he means something categorically different than atheism {cough}{g}--though he is no worse than Lewis on that score, to be fair--but Plantinga is talking about a very specific and technical kind of 'defeat'.) If I had said that (or, anywhere I missed despite making a term search, did say it), I would have meant it in the somewhat untechnical and prosaic, but practical, sense I stated above.

I _did_ reply to attempts on your part to claim that all an atheist has to do, in regard to Lewis, is show that some kind of atheistic epistemological argument _isn't_ self-defeating, though; where, by the way, I even noted beforehand that the atheist couldn't do this "without engaging in a circular argumentation; one even you would probably regard as vicious." (Except, evidently, you didn't consider that circular argumentation to be especially problematic really. Yeesh...{sigh}{s}) But this line of reply on my part is _not_ the same as me claiming that atheism is (in some technical way, Plantingian or otherwise) self-defeating.


At 2/05/2007 01:36:00 PM , Blogger Jason said...


{{1. The advantageousness of a belief in the struggle for survival is significantly distinguishable from the truth of the belief.}}

Granted; but is there anyone here on any side of the aisle who seriously doubts that correspondence to truth is more of an aide to survival-to-replication, generally speaking, than a lack of such correspondence? Even on an eliminative materialistic view, truth as such cannot be expected to "take the hindmost" generally. If my reactions concerning an opportunity to spread my seed around don't synch up with actual facts very well, I'm likely to try seeding a hole in a wall, or another male, or a female who is too old or too young or too sick or too crippled (or not the right species!) to effectively bear children, or a female who is not in the mood to receive my attentions for various other reasons (not necessarily 'reasons' of 'her own', though) and so who may attempt and succeed to escape, or a female who is already receiving the attentions of another male who is capable of ending my access to her in various ways (assuming the female cannot), etc. etc. And that's on a purely unconscious, reactive level.

All things considered, and with respect to (for instance) Plantinga's otherwise correct technical observations, a lack of correspondence in cognitive beliefs (assuming those even exist in any form--do ants have 'beliefs'?) to actual facts is more of an evolutionary hinderance, and the presence of such a correpsondence is more of an evolutionary benefit, than otherwise; occasional exceptions notwithstanding. The Churchland Dictum is fun to pick at, but it's more color than substance; and I expect the Churchlands themselves would actually agree with this: what the dictum really means is that _where there is a conflict_ between impulses, the four Fs will tend to take instinctual priority at the given moment of behavior over an impulse to 'seek truth'. When a screech screeches, the first instinct is usually to jump away from the screech, not to look for what is causing the screech. (A principle I feature in a deer hunt in the opening scene of my novel, btw.)

Consequently, then, I would have to fairly insist on a related premise 1.5 (or 2.5 maybe) in there somewhere. Which is going to make some kind of addendum difference to item 4, though I'm not entirely sure how to phrase it. (e.g., "...and, from 1.5, this advantageousness in believing that advantageous beliefs tend to be true, has some serious probability of being in correspondence to actual facts"? Something like that.)

This might make a real difference in item 5, too. Not sure the difference would necessarily alter the conclusion at 6, but... {shrug}

Keep in mind this is coming from someone who has technical problems with the biological theory per se, and who thinks that in even the best possible circumstances non-rational behaviors cannot be transformed into actually rational behaviors--but who affirms the existence (and some special characteristics) of rational behaviors anyway.

In any case, I would be extremely hesitant to consider 'beliefs' to actually exist apart from intentional action; and once conscious action is on the table, then we already have something in prima facie dispute with atheism: intentive actions derive their existence and character from-and-only-from automatic non-intentional behaviors.


At 2/06/2007 06:49:00 AM , Blogger Darek Barefoot said...


Put the following statement, without further comment, to assorted proponents of atheistic naturalism:

"The human faculty of reason is a product of evolutionary biology."

I will be shocked if you encounter many who will disagree. You might even strengthen the statement with the word "entirely"--"entirely the product of evolutionary biology"--and I doubt the results will change. In the case of those who agree, follow up with the question, "If the faculty of reason is the product of evolutionary biology, isn't it necessarily the case that beliefs generated by that faculty are also products of evolutionary biology?" I think that most will say yes.

I don't see how we can say that a product of a biological process is not a biological product itself. To illustrate, the product of a human process is a human product--or human artifact or human effect, since any of these terms will do. The effect of a human effect is itself a human effect.

One of the crucial points of naturalism is precisely that man in no sense transcends biology. A skyscraper is no less a biological product than is a termite hill in Kenya. The migration of humans to North America was no less a biological phenomenon than was the migration of finches to the Galapagos archipelago. All human behavior is neurologically based, and all neurological structures and processes are shaped exclusively by the two forces of blind chance and natural selection. The explanation of the origin and function of any human organ--the thyroid, the liver, the femur, the caratid artery--ultimately is exhausted by chance and selection for fitness. There is no exception for the brain/brain functions whatsoever--or else we are outside naturalism.

Of course, naturalists always try to carve out an exception for their own thoughts. They must place themselves somehow above or outside of the relentless gear-grinding of nature to buy justification for their speculations about nature. That was Lewis's point. The only resources nature provides them with are biological structures, biological functions, biological interactions with the environment, and chance.

If the naturalist wants to say that the "faculty" of reason, while conditioned by biology, somehow transcends biological products/processes--or biological effects, if you prefer--I am happy to agree. But when we shake hands on this conclusion we will find that we are standing outside the borders of naturalism.

"We are the evolved products of natural selection, which operates without intention, foresight or purpose. Nothing about us escapes being included in the physical universe, or escapes being shaped by the various processes - physical, biological, psychological, and social - that science describes...
"Human beings act the way they do because of the various influences that shape them, whether these be biological or social, genetic or environmental. We do not have the capacity to act outside the causal connections that link us in every respect to the rest of the world." --from "Tenets of Naturalism" (

Yes, on both sides of the aisle we agree that true beliefs are generally more adaptive. The question is whether the naturalistic account is consistent with the justified nature of this common belief. I still believe my sequence shows that it is not.

At 2/07/2007 04:45:00 AM , Anonymous Steve Lovell said...


I we insist on calling beliefs "biological products", then you can no longer say that all biological products occur because they are adaptive or through blind chance.

We agree that the naturalist thinks our faculties are the biological products and are either the product of blind chance or are adaptive.

But suppose the faculties are adaptive, they will be adaptive because the enable behaviour that is responsive to the environment, and because the behaviour is generally adaptive. But then the environmental situation will have a key explanatary role when in comes to specific pieces of behaviour. The same goes for beliefs, the evolutionary account of our faculties allows that the beliefs they produce will vary depending on the environment the agent finds herself in, which then gives the environment an explanatary role. Your (2) therefore implies that beliefs are either not biological products or may not be their solely because they are adaptive or the result of blind chance ... rather those beliefs are held because the person has faculties which are adaptive and because the environment is so-and-thus ... and the environment being so-and-thus may infact be grounds for the belief as well as the cause of the belief.


The behaviour itself cannot be

At 2/07/2007 05:36:00 AM , Anonymous Steve Lovell said...


From your initial presentation of the argument:

If Lewis' demonstration is properly accurate, ... then we will have deduced, not strictly
that atheism is false, but that so long as we take our own reasoning
seriously we ought to reject atheism as false--and accept not-atheism as

If this isn't an argument that atheism (or in my view Naturalism) is self-defeating, then I don't know what is. You also quote with approval the title of J.R. Lucas's paper and the previous title(s) for Miracles Chapter 3.

I really don't know where you're going with your comments on whether certain propositions are or are not logically related. Propositions stand in logical relationships (or don't) regardless of whether they are asserted to stand in those relationships and regardless of whether they are structured to form an argument. Moreover, if an argument is invalid, that hardly makes the set of propositions which constitute the argument inconsistent. Your (1) and (2) do not become any more inconsistent by adding "therefore" to the beginning of (2).

I agree that to demonstrate an inconsistency between propositions I did use those propositions as premises and then derived a contradition. But that was to demonstrate an inconsistency ... and the demonstration of a fact is different from the fact itself.

Now naturalists hold that the four propositions I offered are consistent.

(P1) Naturalism is true
(P2) Evolution is true
(P3) Evolution makes probable the reliability of our faculties
(P4) Our faculties are reliable

Here the role of (P2) and (P3) is simply to make the conistency of (P1) and (P4) seem more intuitively obvious.

I agree that I haven't demonstrated a consistency between the propositions. But then why do I have to? The onus is on the defender of the argument against naturalism/atheism to demonstrate that the four propositions I offered are inconsistent ... remember it is you that is trying to pursuade the naturalist here and not the other way about.

I may still be wildly misinterpreting you here, I just don't have handle on how your argument is supposed to work.

Whether I've understood or misunderstood, I'm still hoping to see a numbered proposition version of the Jason-style AfR.


At 2/07/2007 10:09:00 AM , Blogger Jason said...


{{If this isn't an argument that atheism (or in my view Naturalism) is self-defeating, then I don't know what is.}}

Clearly what I wrote is not the same as saying that a proposition is inherently self-contradictory. (Something along that line might be proposed, perhaps, but I'm having trouble coming up with it.) Nor does the argument depend on ascertaining that contradictions are occuring between separate premises. (This line of attack could, I think, be developed in particular cases, but the ultimate challenge of the Lewisian AfR doesn't depend on charging particularly self-contradictory premises.) So much for two technical definitions of self-defeating. The notion of defeasibility borrowed and applied by Plantinga, to take another definition, might be said to apply, but only incidentally: Lewis' ultimate point isn't to find some reason to doubt (R) given (N). (Notably, Plantinga recognizes that for such a defeater a sheerly imagined posit would work just as well, which is why he is willing to call in the Cartesian demon against the naturalist. Rather ironic, given that Plantinga is rightly trying to avoid the faults he recognizes in classical foundationalism as exemplified by Descartes. But I will discuss this in the emailed paper presently. Suffice to say for now that Lewis' argument does not attempt such a thing, and would be leveled against the Cartesian demon just as much as against atheism.)

What's left over is the "somewhat untechnical and prosaic but practical sense" of self-defeat that I agreed to: e.g. (though not limited to) "When the proponent has to necessarily presume the position he is trying to defend by argument, then in _that_ sense the proponent is defeating himself." I could have given other examples along a similar line, such as: if the proponent chooses to call his own existence as a rational entity into question, in order to escape the implications of the Lewisian AfR (which I've seen sceptical proponents do, btw--you and I both know one sceptic by name who frequently comments on this forum, who has routinely done this, for instance), then that proponent is only defeating himself.

Yes, I positively referenced Lewis' original title to his chapter, and also the chosen title of the Lucas rematch with Anscombe after Lewis' death. This was presented as evidence (among other things) in favor of Lewis' intention to deductively remove atheism as a viable option for belief--an intention which anyone would, I think, agree is rather more dangerous than an intent at inductively challenging atheism. (Which was the original topic of the discussion: what is the most _dangerous_ aspect of Lewis' AfR in MaPS chp 3?)

I'm making this point because you have been fairly constantly claiming that an opponent defending against Lewis only has to defend that a set of the defender's own propositions is not "self-defeating", in one of a couple of technical senses (to which I myself am even willing to add Plantinga's technical sense--which you yourself referenced for topicality.) So, I am replying: the argument I've been trying to refer to from Lewis, does not involve _that_ kind of challenge in the first place.

Aside from which, I agree that a bunch of logically unrelated premises cannot be said to be self-defeating--nor logically consistent, either--and I presented a logical relationship of those premises which I find to be rather common as defense attempts in the field, which I can quite readily agree is also not self-defeating in any of the three technical senses at hand: no circular argument with irrelevant mediant premises is self-defeating. But neither is it an effective reply to Lewis' challenge (on the contrary, he expects it.) Again, I gave a slightly tweaked variant of that defense-attempt argument which I allowed would _in other regards_ be logically proper to try, and which formally (as stated) would not be self-defeating--but which again doesn't really address Lewis' challenge, if given as a reply to it.

I am agreeing, then, that in any of the three basic defense attempts discussed here, none of them are technically self-defeating. I am simply pointing out that what I have been extracting and applying from Lewis' argument, as Lewis himself ultimately arrived at, is not a claim of self-defeat (in any of three discussed senses anyway.)

Now, it may perhaps be possible to set up some variant of the Lewisian AfR which shoots for a technical self-defeat charge--Lewis clearly thought his first attempt did something of this sort. As I noted a minute ago, I can think of a couple of ways of trying it against _particular attempts_ being made by proponents of atheism (in regard to human epistemology). But I do not see yet that such attempts are necessary, and it seems to me they would only have limited application, to specific claim-attempts. Lewis' actual ultimate argument, on the other hand, covers the gamut in principle.

{{[I assert that] Propositions stand in logical relationships (or don't) regardless of whether they are asserted to stand in those relationships and regardless of whether they are structured to form an [arrangement of logical relationship, i.e. an] argument.}}

Ooooookay! {g} I honestly don't know what to say to this--other than it's nonsense, and I would try to avoid taking that position myself, even for convenience's sake.

{{Moreover, if an argument is invalid, that hardly makes the set of propositions which constitute the argument inconsistent.}}

Taken purely as propositions, no: I myself pointed that out, first thing. Taken as a particular invalid argument, yes--inconsistent in that way. On the other hand, if a valid argument is made from them, then they are logically consistent in that fashion; even if the argument is otherwise worthless, thus not to be appealed to for any reason (such as in a circular argument. Circular arguments are worthless, but logically consistent.)

But taken purely as propositions, neither can they be considered logically consistent, any more than they can be considered logically inconsistent. No logical relationship, no logical consistency or inconsistency. Logical relationship, though, presupposes that which needs expression by argument, not by just stating the propositions as premises. Merely hinted arguments can be useful for sneaking things under an opponent's radar, but for purposes of actual evaluation of ostensible logical relationships, they have to be brought into the open.

{{Now naturalists hold that the four propositions I offered are [logically] consistent.}}

I agreed, and provided an argument to demonstrate what that logical consistency turns out to be: a worthless (even though logically consistent) circular argument (exactly as Lewis predicted). If you think they're after some kind of other logical consistency, you're welcome to try demonstrating it. If you (or they) are not going to try demonstrating the logical consistency but simply fall back onto asserting it exists for a given set of propositions, then I have no reason to treat that assertion as any actual logical consistency: it has no practical effect. (And that's aside from the nonsensical claim you made earlier.) Demonstrate it, or don't. No demonstration, no effect. I don't take bluff as an answer to the challenge of an argument.

{{Here the role of (P2) and (P3) is simply to make the conistency of (P1) and (P4) seem more intuitively obvious.}}

If it comes to that, I myself have essentially said that the role of (P1..P3) is to make the consistency of (P0) and (P4) _seem_ more intuitively obvious. Or rather, to make the entirely obvious logical consistency of (P0) and (P4) seem to be some _other_ kind of logical consistency: one which isn't merely circular (thus worthless).

A lot of that illusion's success hangs on ignoring (P0), while nevertheless necessarily presupposing it.

{{I agree that I haven't demonstrated a consistency between the propositions. But then why do I have to?}}

Because anyone can just claim a logical consistency is there without demonstrating it, whether the consistency is there or not. You claim our opponents have a logical consistency in those propositions, that answers Lewis' argument; I say fine, show it to me. You flatly refuse to do so.

Since I went on to actually demonstrate the logical consistency, I find it terribly amusing that you're trying to find ways to keep from having to demonstrate the logical consistency! Ah, so all a defender against Lewis' argument has to do is simply assert that a group of premises has logical consistency, and it's all up for Lewis, eh? The onus is on the Lewisicist proponent to show that the propositions _aren't_ consistent, instead of on the defender against Lewis' argument to show that they _are_.

But it's too late: I already did it (since you were refusing to do so)! We're past that part now; the genie is out of the bottle, I uncorked the puppy and shook the logical consistency of those propositions out onto the table for you. There he is: a worthless circular argument. Just as Lewis predicted, not incidentally. Why is this not a yay-team moment!?

{{Whether I've understood or misunderstood, I'm still hoping to see a numbered proposition version of the Jason-style AfR.}}

It'll show up presently. Aside from working on this thread, and finishing up the EAAN/Lewisian AfR comparison/contrast paper, I'm going to take the opportunity to revise a chapter of an unpublished book while I'm at it. You may safely expect, however, that when I do present it, I won't be just giving a bunch of propositions and hinting that there is some kind of logical relationship (much moreso consistency) among them, while insisting that the onus is on the opponent to demonstrate there is no such thing there.


At 2/07/2007 12:02:00 PM , Anonymous steve lovell said...


This is ridiculous. What is your argument?

Firstly, a valid argument from certain premises to a conclusion does nothing to demonstrate that the premises are consistent either with that conclusion or among themselves.

For example, the following argument is deductively valid ...

(1) X
(2) not-X
(3) Therefore not-X

Now premise (1) is unnecessary, and the argument is circular, but the argument is nevertheless valid. But the premises are still inconsistent.

To show that a set of propositions are consistent is to show that they do not entail a contradiction. There are various ways this could be done ... but again, I still don't think that the onus is on the naturalist to do it. Rather, you are arguing that naturalism is in some way (technical/non-technical or otherwise) self-defeating, therefore the onus is on you to show that this is the case, and if showing that this is the case requires showing that there is some inconsistency in the naturalists position (or showing that there is an inconistency in asserting the naturalists position, then you have to show this, it's not down to the naturalist to show that this isn't the case.

So far, the only reason I can see that you've offered for thinking that any sort of "onus" has been placed on the naturalist is your assertion that naturalism provides
some prima-facie reason to doubt the reliability of our faculties, and that this doubt creeps in at a metaphysical level. But I haven't seen anything resembling an argument for this claim, though given the extent of misunderstandings between us, I may well have missed it or misread it.

On "self-defeat", I take your argument to have the following structure ...

(SD1) If naturalism is true then we should doubt the reliability of our faculties.
(SD2) If we should doubt the reliability of our faculties, then we should doubt the beliefs those faculties produce.
(SD3) If we should doubt the beliefs our faculties produce then we should doubt our belief in naturalism.
(SD4) Therefore, if naturalism is true then we should doubt our belief in naturalism. [From SD1,SD2 and SD3]
(SD5) Therefore, either naturalism is false or we should doubt our belief in naturalism. [From SD4]

You do seem to be arguing this way, and its a perfectly valid argument. It is, moreover, an argument that naturalism is self-defeating (hence the "SD"s).

What I want to see is the support for (SD1).


At 2/07/2007 03:49:00 PM , Blogger Jason said...


While I am still highly concerned about your attempts at using propositions (such as when you tried to demonstrate how a valid conclusion following from its premises doesn't necessarily have to involve logical consistency between that conclusion and its premises--which attempt involved using a conclusion which simply ignored one of its premises in order to reach a valid conclusion from its remaining single premise! {sigh}), I think I understand now where the mis-expectation is.

I'm starting to suspect you think the situation is this:

An atheist argues to some conclusion either equal to or similar to P4 in virtue of P1..P3. (To which, btw, I insist that an actual _ARGUMENT_ be given; not just some propositions ladled out with vague handwaving about them being logically consistent.) Lewis (and/or I) detect a contradiction in the argument. We point out the contradiction, maybe adding that this contradiction is going to always follow from holding to P1. Thus we declare the atheistic argument to be logically inconsistent, and maybe permanently inconsistent so long as P1 (involving the truth of atheism) is included.

However: _this is not the Lewisian procedure_. Nor mine. I'm pretty sure I've said up there somewhere already (though it's a long thread and I've been trying to compose the Plantinga/Lewis comparison, too {sigh} so I may be misremembering, admittedly), that the Lewisian AfR does not require the certainty of some kind of self-contradiction in an atheistic claim. (In that regard, if nothing else, the Lewisian challenge resembles a Plantingian defeater attempt.)

This would explain why my various _agreements_ about various ways of using P1..P4 (with or without P0) _not_ being self-defeating and/or _not_ being logically inconsistent, are throwing you off so badly. Were you expecting me to be trying to nail down a logical inconsistency in an atheistic argument instead?

I'm not. Neither was Lewis (ultimately). His argument (or so I have found, so far) was actually _MORE_ dangerous than that. {s}

{{So far, the only reason I can see that you've offered for thinking that any sort of "onus" has been placed on the naturalist is your assertion that naturalism provides some prima-facie reason to doubt the reliability of our faculties, and that this doubt creeps in at a metaphysical level.}}

Incidentally, I _have_ in fact given arguments for that claim, though not formally enumerated as such, way back up in all that material somewhere, going back to the big post. But at the moment that's beside the point.

Yes, I'm saying (and saying that Lewis is saying, and I _have_ been continuously saying this with monotonous regularity) an "onus" has been placed on the atheist by the implications of atheism. The "onus" is that in light of those implications, they are going to end up having to try to establish by argument (not by vague handwaving) that a faculty they already must necessarily presume to be possibly reliable for any argument at all (including whatever they come up with for defense), can be possibly reliable. (Or as a minor variant, they end up having to establish by argument that a faculty either exists or doesn't have to exist, which they are already necessarily presuming exists-or-doesn't-have-to-exist for the sake of whatever argument they're going to make for the defense.)

No actual logical inconsistency in whatever argument the atheist is coming up with, has to be ascertained. (That would be okay to do, too, but it isn't necessary for the Lewisian AfR.) I have said this more than once already in previous comments (and/or the prior big post.) In fact, the atheistic defense is _expected_ to be logically _consistent_ (though if it isn't, that's okay, too. {g}) Just also worthless for establishing, as an argument, what needs affirming _despite_ the implications of atheism.

If I say until I'm blue in the face that Lewis is setting up a situation where the atheist (specifically thanks to his atheism) lands in a circular argumentation effort no matter which way he jumps or how far--how in all the names of God is _that_ supposed to be the same as saying that Lewis is aiming at establishing a logical inconsistency in the atheist's argument??

{{(SD1) If naturalism is true then we should doubt the reliability of our faculties.}}

Close enough for the moment. (This is clearly supposed to be a conclusion in itself, but conclusions can be treated as premises subsequently. For immediate discussion it can be treated as established already. But you're absolutely right to ask for support for it. This, incidentally, is what most of MaPS chapter 3 is working to establish.)

{{(SD2) If we should doubt the reliability of our faculties, then we should doubt the beliefs those faculties produce.}}

That's one way to go, and not necessarily a bad way to try; but no, it's not the way I would take (following Lewis).

It would be more like SD2alt1: if the reliability of our faculties is in doubt, then we had better try to establish that reliability somehow, if we expect anyone to take us seriously.

After that, the divergence from your SD trail becomes more pronounced, I think. Thus, SD3 would be something more like: but SD0, which we ourselves have to be presuming in order to make any argument, including in order save ourselves from SD1, is that our faculties are reliable. (I would be more specific than that in regard to the reliability, in various ways, which is why I have kept talking in terms of "our faculties are possibly reliable" and "can be possibly reliable", etc.)

It _is_ true, of course, that (on the SD3 you gave) if we should doubt the beliefs our faculties produce then we should doubt our belief in naturalism (with more-or-less dangerous variants of that statement {g}); but that isn't really necessary for the Lewisian AfR, though he can mention it in passing.

Fwiw, I have no particular care to dispute that the SD argument you presented is that naturalism is self-defeating in one or another way. I _do_ care to dispute that the AfR I'm promoting involves claiming that naturalism (or atheism rather) is self-defeating in some various technical ways; for instance that proposing atheism as part of an argument necessarily renders that argument logically inconsistent. That isn't what the Lewisian AfR is ultimately after, consequently defenses designed to escape that kind of claim of self-defeat do not apply.


At 2/07/2007 10:02:00 PM , Blogger Darek Barefoot said...


"Biological" explanationsl always presuppose the nature of the environment of the organism. It is hardly inaccurate to say that there is a biological explanation of why the lungs have the shape and structure they do, even though this explanation makes no sense apart from the recognition of a certain gaseous composition of the atmospere the lungs must breathe. To say that because atmospheric composition must be taken into account the explanation of the lung cannot be fully biological is, I think, to draw a false distinction.

The changing of environments does not obviate the completeness of explanation from chance and selection. If the climate in the arctic becomes warm, polar bear fur may no longer be adaptive. But the explanation of why polar bear fur has the structure that it does can (on evolutionary assumptions, at least) be fully explained as the result of chance genetic variations and selection for the environment in which the fur emerged.

Given naturalism, the brain is an organism for producing, among other things, beliefs; in principle this is just like the liver producing bile. And the only criterion for selection of beliefs or bile--as for the processes involved--is contribution to adaptive fitness to the environment in which they emerge. But our motivation for believing that bile and beliefs can be adaptive cannot be solely that we gain adaptive advantage by so believing. At least, it cannot be our criterion if that belief is to be justified.

Look at the following:

1. Believing what is true usually makes Pollyanna happy.
2. Pollyanna's beliefs may be justified.

These statements are consistent. But now look:

1a. Pollyanna believes what she does only because doing so makes her happy.
2a. Pollyanna believes that she is usually made happy by believing what is true.
3a. Pollyanna's beliefs may be justified.

These statements are incompatible. No. 2a cannot reconcile the conflict between 1a and 3a, although at first reading it may seem that it does. 1a tyrannizes over 2a. Now substitute human beings for Pollyanna and "confers adaptive advantage" for "makes her happy."

Also, remember that nature can only provide accidental processes/products/state changes for selection to choose among. If it cannot be accidentally be produced, nature cannot select it on a purely naturalistic account. If an accidental change takes place in a hormone-producing gland and a hormone is conseqently produced with a different chemical structure, the new hormone--whether advantageous or not--is necessarily a chance product. This may work well enough for physical products and processes (ID proponents would differ, of course) but not for rational belief-forming processes.

If thought B follows thought A accidentally, then even if B could be inferred on the basis of A it has not in fact been so inferred. It is simply not in the nature of an accidental sequence of thought that it can also be a rational sequence. Yet that is absolutely the only way the sequence of B followed by A can be explained in naturalistic evolutionary terms. If a brain process is sightly altered by a spin of the genetic-chemical roulette wheel, the states generated as a result are themselves accidental (or, incidental). If we stipulate consciousness, the evolutionary process can get us as far as advantageous hunches--impulsive beliefs generated by chemical surges--but not sequences of thought in which the succession of A by B occurs because of a logical relationship between them.

At 2/08/2007 05:25:00 AM , Anonymous Steve Lovell said...


I agree that the environment is always important in the explanation of biological products, but it plays a different role in the explanation of "adaptions" and of "behaviours".

This is because behaviours are not "coded in the DNA" as it were. We don't inherit all our behavious or beliefs from our parents (at least not in the same way as we inherit our eye colour or other physical traits).

I also agree that your three propositions 1a,2a,3a are inconsistent. However, they would not be inconsistent if we were to remove the "only" from 1a. Here you're running into issues about whether two different sorts of explanation are mutually exclusive or not. This is exactly the line of criticism of Lewis used by Anscombe and Flew. They both contend that "causal explanations" and "reason explanations" are not incompatible. Of course as Lewis points out if naturalism is true and our beliefs are ever justified, then the relationship between the causes of our beliefs and the reasons for them cannot be accidental. My point is that a persons environment being a particular way as they form a belief may both cause the belief and provide reasons for the belief. You need to show either that the two kinds of explanation are mutually exclusive, or that evolutionary history cannot have so wired our brains as to make the causal and reason-based explanations reliably obtain together.


I think we're finally getting somewhere. What I want to see spelled out is the part of your argument that places the onus on the naturalist to justify his confidence in his faculties.

Remember, you need to do this in such a way as the prevent the naturalist from appealing to evolution as a way of resisting the imposition of the (impossible)burden of proof rather than as a way of removing that burden.


At 2/08/2007 06:07:00 AM , Blogger Jason said...


{{Remember, you need to do this in such a way as the prevent the naturalist from appealing to evolution as a way of resisting the imposition of the (impossible)burden of proof rather than as a way of removing that burden.}}

On the contrary, the Lewisian AfR works precisely from the atheist attempting to appeal to anything subsequently to the challenge. Why would I try to stop the atheist from appealing to evolution as a way of resisting the challenge? That's precisely the sort of thing the Lewis is expecting and counting on! If the atheist _doesn't_ try to make an appeal, the circularity of his appeal won't be exposed--but then the challenge remains where it is, too. It's a losing situation any way the atheist jumps (or doesn't jump). The checkmate character of the challenge is probably why Lewis either never considered or never bothered to run it in regard to theism, too.

This is an example of why I keep saying the Lewisian AfR is commonly misunderstood in the field. It isn't about trying to prevent the atheist from doing something. It's about discovering that making no move is fatal either to a necessary presumption or else to atheism; and _also_ discovering that making any move involves trying to save the necessary presumption from the implications of atheism, which save must itself necessarily involve a merely circular (thus worthless) argument.

When the king is zorched whether he stays in one place or tries to move anywhere, the game is over for him. (Which is not really a good analogy, since in chess if the black king is checkmated it isn't necessary to test whether the same arrangement checkmates the white king, too. {s} Well, usually anyway. Fun chess problem: come up with a checkmate for black king that risks checkmating white king, too, at least to the point of having to test it. For analogy's sake, don't use the white king in the checkmate attempt.)

Anyway, I point this out (yet again) because it's important to understand the general way the argument (ostensibly) works, in order not to misjudge it. The success or failure of the argument has to be judged on its own terms, and trying to prevent the opponent from making a move is _not_ one of its aims.


At 2/08/2007 06:24:00 AM , Anonymous Steve Lovell said...


I'm very sorry but you don't seem to get this point.

The part of the arugment I want to see is that part that supports something along the lines of (SD1).

Now suppose you have an argument for SD1 (I still haven't seen it btw). What the evolutionary move is an attempt to do is to show that the argument for SD1 is a bad one. It is not an attempt to accept the argument for SD1 but say that we have other evidence which outweighs your argument.

This is what you have to prevent the naturalist/atheist doing. But to see whether you can prevent the naturalist/atheist doing that we have to see the argument for SD1.

I know you said you gave that argument already, and there are some passages that look vaguely like arguments for that conclusion, but I haven't seen anything which would prevent the evolutionary move being a non-circular attempt to resist the argument for SD1.


At 2/09/2007 06:16:00 AM , Blogger Jason said...


{{The part of the arugment I want to see is that part that supports something along the lines of (SD1).}}

I know. I’m working on that. I specifically said you were absolutely right to ask for that.

If you’re expecting this to involve simply trying to keep the atheist from moving onto a subsequent explanation, though, then you’re going to be disappointed and/or confused.

I have numerous other things I’m trying to do (helping manage a manufacturing facility for instance); and I’m trying to figure out how to do the enumeration in such a way that it doesn’t require me having to go into 330 pages worth of prior discussion and illustration. (I’m not kidding about that number.) As it is, I’m worried that in trying to broaden out P0 so as to avoid having to discuss a bunch of other things, I’m leaving out an important element (one Darek is getting into the account, btw. {s})

Meanwhile, I’m _also_ trying to forestall potential mis-objections, so that when I eventually do X I don’t have to spend time trying to reply to objections which expect I am really doing Y or Z or F or something else other than X.

And you’re still expecting me to be doing Y (or something), not X. Otherwise you would have noticed that SD2alt1 anticipates exactly the move you’re expecting the atheist to make, in applying to evolutionary theory, as an answer to SD1. (If the atheist wants to answer SD1 while also trying to avoid ‘answering’ SD1, that’s his problem, not mine.)

Also keep in mind that in commenting on SD1 I said “close enough for the moment”. That was because I could see that it may turn out to be somewhat different when it arrives.

{{I haven't seen anything which would prevent the evolutionary move being a non-circular attempt to resist the argument for SD1.}}

If the evolutionary move to resist the argument for SD1 (whatever that turns out to be) involves presuming its conclusion, then it’s a circular argument attempting to resist the argument for SD1. We discussed this a long time ago.

Again, if an answer to resist SD1 looks essentially like this:

(Necessary P0) Human (including my) cognition is possibly reliable.

(P1) Naturalism (whatever this is taken to mean) is true.

(P2) Evolution (whatever this is taken to mean) is true.

(P3) Evolution makes probable the reliability of our (human) faculties.

(C1) Our (human, including my) cognitive faculties are possibly reliable. (From NecP0, P1, P2, P3)

--then the attempt to resist is not noncircular. (For that matter, it could be given without being specifically an attempt at resisting an SD1; but it would still be circular.)

I will try to work more on the argument Saturday. (I spent all yesterday afternoon and last night working out one enumeration and then deciding I had done it wrongly. I’m not delaying, I’m just _busy_, and it takes time to do and it’s difficult and I want to get it right.) Honestly though I’m going to be struggling not to just collapse today at work from mental exhaustion, so I may have to delay getting back to this until Monday. Maybe longer. (I have a chronic illness I expect to be strongly kicking in again soon, with early signs already it’s going to happen right on schedule. Exhaustion makes it worse. God knows I would like to get through the next few weeks without begging Him to just kill me and get it over with again...)


PS: great, today's anti-spam word verification exercise is "eyejuse". {wry s}

At 2/09/2007 10:54:00 AM , Anonymous steve lovell said...


Suppose that the argument you are giving has this form:

(1) X
(2) If X then Something SD1-ish
(3) Something SD1-ish

Now, until you have established (3) you have not placed any burden of proof on the naturalist/atheist to demonstrate the reliablity of our faculties.

As such, in the process of objecting to the argument (1) to (3) the naturalist cannot be accused of circularity in appealing to evolution.

Of course, depending on what the argument (1) to (3) actually is, the Flew-ian evolutionary reply may or may not succeed in preventing the inference to (3).


At 2/10/2007 11:25:00 AM , Blogger Darek Barefoot said...


Well, I agree to an extent. Natural selection and chance do not exclude reason unless they are invoked as its sole sources.

And I should have acknowledged more emphatically that it cannot be said of every biological event unequivocally that it occurs because it is advantageous. As I said before, a biological feature or behavior that is adaptive in one enironment may not be so in another. But the origin of the behavior or function may always be referred either to its adaptive value in the environment in which it emerged or to blind chance.

A belief that is adaptive in one environment may indeed not be in another. But I do not see how a neurological process that yields unjustified beliefs can, with a mere change of physical surroundings, be transformed into a process that yields justified ones. If this is your contention you need to demonstrate it rather than simply assert it.

I will attempt yet another analysis. Take a belief that is advantageous. Is the adaptive advantage of the belief either enhanced or dimished by the fact that the belief is held justifiably? I don't see how it affects the advantageousness of the belief in the slightest. If you believe that you should not be on a bald hill in a thunderstorm, and if the belief motivates you to get off the hill, you enjoy the full advantageousness of the belief whether you hold it justifiably or not. Therefore we cannot say that the rationality of a belief is an aspect of its advantageousness. And If a belief need not be held rationally or justifiably in order to be advantageous, then how can nature select for that property in beliefs? To summarize:

1. Nature selects beliefs only for those qualities that contribute to advantageousness.

2. The advantageousness of beliefs is unaffected by whether or not they are held justifiably.

3. Therefore nature does not select beliefs because they are held justifiably.

This does not mean that nature chases out rationality. But it amounts to a big problem for anyone who answers the question, "How could evolution give us the power of reason?" by saying, "Obviously, it can do so because reason promotes survival." Victor was hit by that shallow answer on an evolution blog just the other day.

But wait. A possible escape route targets belief-forming processes as follows:

1a. Neural processes that are capable of producing justified beliefs are more likely to produce advantageous beliefs than are processes lacking that capability.

2a. Nature selects for neural processes that are most likely to produce advantageous beliefs.

3a. Therefore, nature selects neural processes capable of producing justified beliefs.

This attempt, however, can only work if 1 - 3 is not airtight, because of that old problem of recursion. With nature as the only resource, 1a, on which the escape route depends, is blocked as follows:

1b. Nature selects beliefs only for those qualities that contribute to advantageousness (1, above).

2b. The advantageousness of the belief, "rational belief-forming processes are more likely than others to produce advantageous beliefs," is unaffected by whether or not that belief is held justifiably.

3b. Therefore nature does not select the belief, "rational belief-forming processes are more likely than others to produce advantageous beliefs," because it is held justifiably.

Hold on!, the naturalist says. Nature may not select that belief, but we are free to see that it's reasonable. Yes, we are free to hold it justifiably--outside of nature. Natural selection cannot favor the belief because it is rationally held, yet something in us does so favor it. Our third door is chance, but if we believe it by chance it likewise is not justified.

This reasoning gives a degree of support to the conceptual arguments about the possibility of p-zombies from Chalmers, et al. My favorite example is social insects. Social insects organize, plan for the future, cooperate and exhibit fairly sophisticated problem-solving skills through what I think we must label computational means--apart from justified beliefs. Certainly the behavior of ants and bees is less sophisticated than that of human beings, but then the computational resources of an ant's central nervous system are quite a bit smaller than those of a human being.

At 2/10/2007 11:52:00 AM , Anonymous steve lovell said...


Why isn't the argument 1b to 3b just as vulnerable to the challenge of 1a to 3a as the argument 1 to 3?

After all, 1b is all about the adaptiveness of a single belief, and as 1b to 3b seems to show, the naturalist/evolutionist isn't commited to all beliefs occuring either by chance or being adaptive.


At 2/12/2007 06:35:00 AM , Blogger Darek Barefoot said...


Good point. To the extent that 1a - 3a is ambiguous about the options on the table for belief-forming processes, the status of the argument is ambiguous.

Perhaps it would help if I were more specific about the mechanics of this question. We know that chemicals in the brain can compel belief. Say that a man takes his girlfriend to a nightclub where she trades jokes with a handsome stranger. What the man in a sober state would take to be an innocent bit of conversation he may, after downing several shots of liquor, believe to be a shameless flirtation. Likewise, chemical surges in the brain of a schizophrenic may compel him to believe that Martians have planted a computer chip in his brain or that any stranger who looks at him intends to kill him.

Imagine that under various external stimuli, the brains of thousands of human progenitors underwent chemical surges prompting beliefs of all kinds. Among these compulsions there were a few that, by chance, corresponded to what we would consider rational responses. Say one such stimulus was lightning striking and killing a member of a group of hominids on the top of a treeless hill during a thunderstorm. Suppose another member of the group observed this event and happened to experience an advantageous chemical surge compelling him to believe that he would be in danger under similar circumstances. His belief would be justified to us, that is, we believe for good reasons that one should fear being on treeless hills during electrical storms. But, given the chemical compulsion behind the hominid's belief, would he himself have been justified in holding it? Remember, the compulsion arose by chance and was preserved by a mechanism--natural selection--that is insensitive to whether a belief is held justifiably or not. Suppose, however, that through gradual accumulation of similar advantageous variations in the triggers for neurochemical secretion, each small instance of which arose by chance, systematic alignment of compulsive belief with rationality resulted. Doesn't this give us a plausible explanation of human reason?

Lewis thought that this account was incoherent. A chemically compelled belief that by chance happens to be true is not justified. The mere fact that this chance belief besides being true happens to be advantageous is not enough to move it from the "unjustified" to the "justified" category. The presupposition of justification for at least some of our beliefs is therefore enough to falsify the account. Does naturalism have the resources to give an explanation that is essentially different than the one above? I think that the best hope for the naturalist is to emphasize the systematic nature of belief-producing processes. But this emphasis only disguises the fact that chance variations in the physical structures governing chemical secretion must nevertheless be the source of all the variation in resultant processes and therefore of their belief products as well.

At 2/12/2007 11:42:00 AM , Blogger Jason said...


hurting badly at the moment, so brief reply.

{{As such, in the process of objecting to the argument (1) to (3) the naturalist cannot be accused of circularity in appealing to evolution.}}

The naturalist can be accused of circularity in appealing to evolution if the appeal is circular (whether or not the appeal happens to be made in objection to argument 1..3.) If the appeal is not circular, then as I specifically gave an example of myself, no accusation of circularity of course. (Though as also exemplified in the example I myself gave for such a thing, way back up in the comments somewhere, neither was that particular appeal actually an answer to Lewis' challenge.)

I also mentioned there were (in principle) ways of objecting to (in essence) 1..3 that need not involve the risk of circularity by appeal to subsequent justification efforts; such as challenges about the content of 1..3 in the first place.

back under my rock for a while


PS: I'll read and address the comments you sent on my Lewis/Plantinga paper, too, later, when I'm not trying to remind myself to breathe. Nominally I plan to work on the enumeration argument first, though.

PPS: Darek writes, "His belief would be justified to us, that is, we believe for good reasons that one should fear being on treeless hills during electrical storms. But, given the chemical compulsion behind the hominid's belief, would he himself have been justified in holding it? "

That's an excellent example of avoiding the externalist fallacy.

At 2/12/2007 12:17:00 PM , Anonymous steve lovell said...


I like this. Essentially, what's going on is that on this account natural selection is bringing our beliefs into correspondance with reason, but without our beliefs ever actually being influenced by reason. This, to my mind seems exactly the right view of how things must finally be on the naturalists view. It's a nice way of fleshing out Lewis's comment that "it is not conceivable that any improvement of responses could ever turn them into acts of insight".

However, the naturalist now has one final trick up his sleeve: psychologism. On this view reason just is the way that the brains of a (priveleged) class of people are wired to respond when in the right circumstances.

This is the line of response favoured by John Aach in his paper "Pyschologism Reconsidered: A Re-evaluation of the Arguments of Frege and Husserl" Synthese 1990 (85): 315-338.

Personally, I've tended to think that this is the naturalists best move in response to the AfR, and that otherwise the coincidence between belief and reason remains ultimately just that: a coincidence. One can tweak the definitions of "causation" to attempt to show that the two sides of this coincidence are causally related (this appears to be what Kim is attempting), but I've never found this very pursuasive.

The psychologism move avoids this by collapsing the two sides of the relationship to a single metaphysical level. I think Lewis is aware of this move (he called it behaviourism - see chapter 13 of SbJ), but found it literally incredible. So do I. I ought to read up on the arguments against it, but I've never found the time (and tend to find Frege and Husserl and their commentators rather difficult).

Obviously, it's a form of relativism and has many of the usual problems associated with that. Moreover, I can't see how you'd get any normative force out of reason on this account. (Lewis's material on "Maps" in Pilgrim's Regress is relevant here.)

Thoughts anyone?


At 2/12/2007 09:22:00 PM , Blogger Darek Barefoot said...


We come out at pretty much the same place. I haven't read the material on psychologism, but indeed it seems to abandon objective norms of reason . . . which doesn't make it a very good response.

Alignment versus identity is indeed the issue. Syntax versus semantics. Numbers do not flow through the circuits of a calculator, rather, representations of numbers do. But numbers, not just representations, somehow flow through my mind--else I could not grasp the difference between numbers and their representations.

A final point: Lewis seemed to think that logical necessity makes the issue stand out better. I believe he was right. I can "see" in my mind that 2 + 2 must equal 4 and that no other answer is possible. I cannot identify this as mechanical computation because a calculator can be wired to say that 2 + 2 = 5 or any other number for that matter.

Or, approach it from the standpoint of chemically-generated beliefs. Presumably my brain could be chemically "wired" so that I could "see" that 2 + 2 = 5 in just the same way that I now "see" that 2 + 2 = 4, except that I wouldn't survive very well if I did so. But this is just not conceivable. No one could ever be chemically induced to "see" that 2 + 2 can only equal 5 the same way we see that it can only equal 4.

At 2/13/2007 08:49:00 AM , Anonymous steve lovell said...

Darek and All,

First, I'd like to apologise for having refered to you as Derek in all my comments thus far. It wasn't deliberate.

After making my last comment it occured to me that I'd been a bit hasty in saying that the naturalist has "one final trick up his sleeve". There is at least one other way the naturalist can wriggle out of this one: externalism.

According to Externalists what turns true belief into knowledge need not be something within the cognitive perspective of the agent in question. Reliabilists, for example would just require that the belief be formed by a process that generally results in true beliefs. This reliability would certainly be consistent with the naturalist account of belief we've (mostly you've) stetched.

Vic has some material on the blog somewhere about internalism and externalism, but it got rather technical and I bailed, not sure how relevant it would have been to this concern either; I'll have to look it out again, I think it's among things that Vic redated.

I'm guessing many naturalists would find this more palatable than psychologism, but I still tend to think that psychologism is the stronger response.

Out of interest, I'd be interested in hearing people's views on the psychologism and externalism moves. Also, I'd be interested in arguments for/against psychologism generally. One reflection of mine was that reason ought to have some internal relation to truth, but this would seem to mean that adherents of psychologism (psychologicists?) would have to hold to some sort of coherence theory of truth. Moreover, since coherence is a logical term the coherence in question could itself be psychologised, making the view doubly relativised.


At 2/13/2007 11:52:00 AM , Anonymous steve lovell said...

Just dug out the previous dialogue on internalism and externalism, and it's actually rather sparse, the posts I was thinking of quickly went rather "off topic" as far as this is concerned. Essentially what we have on the blog so far is the assertion that externalism is misguided, but we also have a link to Depoe pointing out that Plantinga is an externalist. Indeed, it occurs to me that a variant of Plantinga's proper function based on evolution rather than God would be an interesting epistemology to pit against the AfR.

Where do we start in thinking about this?


At 2/13/2007 12:46:00 PM , Blogger Deuce said...

Hi darek, steve,

Been watching your comments with interest, and thought I'd add a couple things.

You've been starting with the assumption that true beliefs will generally have survival value, but are pointing out that this doesn't account for rationally held beliefs, but at best for an incidental correlation between some beliefs and the truth.

However, I don't think that even that assumption can be taken as a given, except for the sake of argument. What, for instance, if the belief that a rabid lion is in front of you, and that you need to get away, causes you to run up and pet the kitty? Or heck, what if that same belief causes you to do something totally unrelated, like wash your hair?

The naturalist could, at this point, simply declare that, say, a brain state that causes you to run away simply is identical to the belief that there is a lion in front of you and that you need to get away, and so forth. However, this is radical behaviorism, which of course has a whole host of problems, and I think this answer is about as nutty as psychologism. And barring identity, it must be accepted that the correlation between between a belief and the action that results is contingent.

So there's actually two questions at play here. On the one hand, how to account for beliefs being correlated with correct corresponding actions, and on the other hand, how to account for them being correlated with the truth. Accounting for each requires that the other be taken as a given. And then, even if the naturalist takes both of these as givens, there's the additional problem, pointed out by you two, that they've only accounted for a general correlation, not rationality.

At 2/13/2007 09:57:00 PM , Blogger Darek Barefoot said...

Steve and Jason

An overdue disclaimer: I have no formal training in philosophy or logic, and my informal training--if it deserves to be called that--is minimal. Nor do I have much spare time to supplement it now. One of the reasons that I have not commented on Jason's posts is that I'm not sure I followed his argument sufficiently well. I am at this point barely acquainted with the externalism-internalism debate and so can offer only impromptu thoughts.

Every philosophical argument is accompanied by an implied invitation and challenge: "If you think carefully about what I am saying you will conclude that it is reasonable, and since it is reasonable it is likely to be true." My impression is that externalism issues this invitation and repudiates it at the same time: "If you think carefully about what I am saying you will conclude that what makes it reasonable is not what someone concludes about it after thinking carefully, but the way the world actually is." It just doesn't sound coherent to me. But, perhaps externalism simply places justification at a fair distance from reason proper, in which case I'm not sure what the fuss, or the point, really is.

Recently I came up with a proposition that I will call, for lack of a better term, a corollary of reason. It is this: A belief may be considered rationally held only to the extent that what the holder consciously perceives to be the reasons for his holding the belief are in fact the reasons for his doing so.

Till this point I have not been able to shake this corollary, for what that's worth. It doesn't mean that rationally held beliefs are the only ones worth having, and it allows for degres of rationality of beliefs as well. However, it does raise a conceptual barrier against assaults on introspective logical analysis. It says that there is an absolute limit beyond which we cannot coherently accept that the deliberations of logic are at some deeper level unrecognizable. It implies that we cannot be so utterly confused about the difference between reasons and causes that non-rational mechanisms can--surprise!--be the hidden drivers of rational thought. In the present case, it may point to the characteristics of the deliberative process itself, rather than external circumstances, as the essential gauge of rationality. But, again, I cannot pretend to have analyzed the issues in great depth.

At 2/14/2007 12:49:00 PM , Anonymous steve lovell said...

Deuce and All,

I quite agree. We've been allowing that the naturalist has an account of the content of beliefs which links it in some way with the behaviour which that belief (in combination with others and some desires or goals) would make rational. Plantinga discusses these issues in his presentation of the Evolutionary Argument Against Naturalism. Personally, I've never been able to work up much concern for the naturalist on this point. All my professors at university were functionalists (not quite the same as behaviourism, but very close), and as you point out, if functionalism can be motivated it seems to avoid the problem at issue.

For those interested in the Externalists rejoinder, I've just added a comment to Victors redated posting on Depoe's discussion of internalism, externalism and the AfR. In essence, I explain there why I don't think externalism meets the challenge that we've been putting to naturalism here, but I'm still not sure what to say about psychologism.


At 2/15/2007 11:28:00 AM , Blogger Edward T. Babinski said...

This comment has been removed by the author.

At 2/15/2007 11:38:00 AM , Blogger Edward T. Babinski said...

Perhaps Vic hasn't gotten to the point of fully acknowledging that there are philosophers who argue that "theism" employs arguments that are equally as "circular" as the ones he finds so unsatisfactory in "naturalism" (especially whenever arguments about huge abstractions are analyzed in a strictly philosophical fashion).

So maybe philosophy itself has its limits (whether they be "natural" philosophical limits or "supernatural" philosophical limits is yet another matter for theistic and nontheistic philosophers to argue over ad infinitum).

I do know however that philosophy is such a flexible mode of "explaining big abstract questions" that even Christian philosophers are on multiple sides of the brain-mind question:

If the link above fails to show, I have added a space in the second version of the link below, i.e., a space between .us/ and /religion

Just delete that space and you can find the article that itself contains lots of links to articles on the topic of C. S. Lewis's "dangerous" [sic] idea.

At 2/16/2007 02:55:00 PM , Blogger Jason said...

Perhaps Ed hasn't gotten to the point of fully acknowledging that Victor has worked and studied for decades in a more-or-less Reformed theology venue, and allows theologians who actually promote circular reasoning to duke it out on his web journal with other theologians who don't, when he isn't calling coup on various circular theism tactics himself; thus demonstrating (as if his repost of JohnnyDee's question wasn't blatantly obvious enough) that Victor and the rest of us here are very well aware of circular-argumentation claims (pro and con) in regard to theistic apologetics, thank you very much.

I mention this, not so much for Ed's benefit, since he has had just as much and in some cases more opportunity than most site visitors have had, to know that Victor is aware of theistic circular argumentation attempts and claims thereof, and yet thought it pertinent, or perhaps just conveniently impertinent, to write what he did anyway; but for site visitors who may not have had Ed's opportunities and so who may be misled into thinking wrongly about Victor by Ed's incompetency. Whatever disagreements I may have with Victor, I don't like people unfairly representing him. (Plus I'm easily annoyed in my current condition. If I'm going to puke I might as well puke on fools who volunteer for nothing better...)



Post a Comment

Subscribe to Post Comments [Atom]

<< Home