Friday, August 10, 2007

Naturalism

This is the apollos.ws site on naturalism, including the Troy Nunley dissertation on the evolutionary argument against naturalism. This is in response to a request by exapologist.

16 comments:

exapologist said...

Thanks, Victor, for the helpful link!

I find Plantinga's argument fascinating. I tend to agree with James Beilby's critique of the argument in his 2003 Philosophia Christi article. If one takes a Lakatosian position about research programs, it seems that Plantinga's criticism doesn't go through. For then the hypothesis of the reliability of our coginitive faculties can be construed as an auxiliary hypothesis with respect to the research program of naturalism, where the plausibility of the latter doesn't essentially depend on a high probability assessment of the former. If naturalism isn't a degenerative research program, then it's rational even if an auxiliary hypothesis undergirding belief in the reliability of our cognitive faculties is low or inscrutable, in which case naturalism isn't self-refuting, contrary to Plantinga's argument.

It's interesting that Beilby came to this conclusion, as he's a conservative Christian who was originally a staunch defender of the argument (he's the guy who edited the collection of papers on the argument entitled "Naturalism Defeated?", and who also defended the argument in the International Journal for the Philosophy of Religion). In the Phil. Christi article, he said (regarding his recantation re: the argument) that his earlier passionate endorsement of Plantinga's argument was due to "the impetuosity of youth"(!).

Anonymous said...

Planting teaches us that demons are causing problems in the world.

He never explains why it is consistent to maintain both that there are demons capable of attacking our cognitive faculties and that we should expect our cognitive faculties to be reliable.

But why should he try to mend the gaping holes in his beliefs? People still buy his books anyway.

Will Kratos said...

Cool resource. Over 30 published papers referenced on that site. Good deal.

Anonymous said...

What is Plantibga's thesis?

Is it 'Increased survival chances do not correlate with increased brain power'?

That is a scientific hypothesis for which Plantinga should produce data.

Brains are very expensive to feed and maintain.

Presumably nature balances off increased survival of more brain power with decreased survival of having to find more food to feed the brain.

But that is a question of the facts of the matter - of little interest to people who see it as their life's work to destroy belief in naturalisn.

Jason Pratt said...

Exap,

I have my own problems with Plantinga's EAAN, and not a few (plus Troy's defense of it); but I'm not sure I could accept the particular one you mention from Beilby (assuming his critique is being reported correctly.) I'm having trouble with someone seriously proposing that a research program can proceed without a preliminary assumption in favor of at least the possible reliability of the cognitive faculties of the agents of the program, including by those agents themselves!

The rationality of a research program depends on the rationality of its agents: it seems to me the program doesn't exist independently of the agents, and can only be an abstract derivative description of actions the agents are doing. (Unless the program is supposed to be construed as a Platonic ideal form or something of that sort, and then I'm unclear why anyone would accept this--except as an attempt at dodging the preliminary necessity of assuming that our cognitive faculties can be reliable. {g} I mean that it wouldn't seem to follow from anything.)

{{He never explains why it is consistent to maintain both that there are demons capable of attacking our cognitive faculties and that we should expect our cognitive faculties to be reliable.}}

I wouldn’t have put it quite that way, but weirdly enough one of my complaints with the EAAN (or with how Plantinga defends it through the concept of defeaters, to be more precise) sort of resembles this. It would take me a while to go into it, though.

{{What is Plantibga's thesis?

Is it 'Increased survival chances do not correlate with increased brain power'?}}

Um, no. And by the way, the rest of us bother to put our names to our posts. That way if we happen to make inept guesses, everyone knows who made the guess. (Or again, if someone has been banned from the site for being flamebait, that way they can rub Victor’s nose and show him his ban was worth nothing. Or whatever. Though on the other hand, if you _were_ banned from the site, and then put up your name, everyone would know you’re not to be paid attention to. So I guess from that perspective being anonymous is a safe tactic; I mean if you’re still hoping to get some attention despite having been banned.)

JRP

Unknown said...

If we simply take naturalism as a reasearch programme and assume that our cognitive faculties are reliable, should not naturalism be expected to be able to account for that reliability to be a fruitful programme?

Oh, and all of us are interested in 'the facts of the matter'. Or at least I'd expect some evidence to be produced if you're going to start accusing ppl of acting otherwise.

exapologist said...

Hi jason and mattghg,

If evolutionary theory (sorry, in my last comment I said "naturalism") is a productive research program, then it's rational to accept it even if some of its auxillary hypotheses have a probability that are low or inscrutable. Well, one auxillary hypothesis of evolutionary theory is that the beliefs of prehistoric hominids are causally connected with behavior and are adaptive, and this, in turn, is supposed to explain the reliability of our cognitive faculties. And while (as Plantinga points out), the probability of this may be low or inscrutable, it may yet be rational to accept that our cognitive faculties are reliable so long as evolutionary theory remains a productive research program.

Now I know it's tempting to come back with Plantinga's point that we have an all-purpose preemptive defeater for claims about the reliability of our cognitive faculties, since any such belief will be generated from a faculty whose reliabiliity has been called into question. But the problem is that Plantinga's preemptive defeater doesn't get going until he *first* gives us good reason to distrust our cognitive faculties on the conjunction of naturalism and evolutionary theory. But again, given a Lakatosian philosophy of science, a preemptive defeater of this sort requires that Plantinga do something much, much stronger -- *he has to show that evolutionary theory is a degenerative research program*. But he's come nohwere near doing *that*.

Jason Pratt said...

Mattghg: {{If we simply take [b.e.t., per Exap's topica correction] as a reasearch programme and assume that our cognitive faculties are reliable, should not naturalism be expected to be able to account for that reliability to be a fruitful programme?}}

Depends on what we mean by 'account for'; and also on the extent to which the assumption of adequately reliable cognitive faculties has to be done in _any_ case. If the intention is to _justify_ our justification abilities, then something has gone wrong with the program at a fundamental level somewhere. (Which incidentally is one of my main gripes with Reformed-presuppositionalist arguments _to_ reason.) Ditto, or even more-than-ditto, if the result of the program arrives at a denigration of our justification abilities (which then have to be justified in order to keep the denigration from being self-refuting or at least universally solvent.)

An explanatory description of how reasoning capability contributes to competition and survival in an evolutionary fashion, is fine as far as I can tell; and I'm still okay with the project insofar as it involves discussing how genetic contributors to the effectiveness of our reasoning capabilities are affected through multiple generations of breeding, via natural selection (or artificial selection as the case may sometimes be!--for once we start talking about intentional actions of humans, we’re in the artificial category now. A distinction of very considerable importance, by the way.)

When the b.e.t is being marshalled to help underwrite claims of naturalism, though, then I’m going to start talking about the philosophical topic of naturalism and/or atheism, at which point I’m going to begin bringing up an analysis of fundamental claims being made by proponents. Now we’re doing metaphysics. And if the metaphysical math doesn’t add up in principle, then that makes a difference in how we ought to be interpreting the subsidary data.

For what it’s worth, I would probably go past “reliability of our cognitive faculties” sooner or later; for much the same reason that Lewis in MaPS is quite content to allow that if naturalism (actually atheism per his contexts) is true then our cognitive faculties might function as reliably as reason _or in some circumstances even better_. But he distinguishes between reason and instinct, and while he isn’t entirely clear about the character of that distinction, it’s utterly important to the thrust of his argument vs. atheism in MaPS.

I will note in passing that this clear allowance on Lewis’ part, which could have just as easily been written by one of the Churchlands, is a key piece of evidence for distinguishing the major differences between Lewis’ Argument from Reason and Plantinga’s EAAN. (I have a long paper on this around the office here somewhere, with comments from Steve Lovell I’ve never gotten around to reading but meant to, now that I think of it...)

Exap: {{If evolutionary theory (sorry, in my last comment I said "naturalism") is a productive research program, then it's rational to accept it even if some of its auxillary hypotheses have a probability that are low or inscrutable.}}

That’s admittedly true; though that isn’t the same as accepting those auxillary hypotheses! Also, as noted above, this means nothing if one of those ostensibly auxillary hypotheses is actually a necessary presumption for doing the work on b.e.t. at all. At _that_ point, a conflict has arisen, even if the ostensibly _auxillary_ hypothesis ended up having a reasonably fair liklihood under b.e.t. (And if its liklihood is being graded in light of b.e.t., then we’re probably talking about a conclusion somewhere, not just about a hypothesis.)

Insofar as the evolutionary program being a degenerative research program goes, if one result of it is that a faculty which is necessary to underwrite the program is called into distrust as a result _of_ the program, then that looks sufficiently degenerative to me. But, since we’re only talking probability, I can see that if one tries to point to successes elsewhere--and the successes needn’t even be in the b.e.t. program, I suppose--even though that can only be done by appealing to the faculty put into probabilitistic distrust _by_ the program, that could be considered evidence that the probability is met after all. Improbability isn’t impossibility.

What would be tricky there, would be to make the appeal without at the same time trying to re-establish the _trust_ involved. If the reliability of our cognitive faculties has been called into doubt, the trust cannot be formally re-established by a method that requires us to trust the possible reliability of our cognitive faculties!

But then, the problem could perhaps be avoided from the outset by treating reason as a given reliability (adequately if not perfectly so), instead of as an auxillary hypothesis. That might be a slick way for an opponent of Plantinga’s EAAN to use Lewis’ own AfR procedure to trump the EAAN attempt.

‘course, using _that_ idea might be pretty dangerous... {g}

JRP

exapologist said...

Hi Jason,

On a Lakatosian philosophy of science, a belief entailed or suggested by a research program can accrue justification "globally -- via the program as a whole, and not necessarily just "locally" -- via the probability of an auxilary hypotehesis alone. One reason why Lakatos thought this was because of many episodes in the history of science, where theories that turned out to be successful had local claims and auxiliary hypotheses that were low or inscrutable for a long period of time -- the evidence often seemed to *disconfirm* the theories. So, for example, the heliocentric research program had many such problems in their Keplerian and Copernican incarnations. But the *global* justification of the research program that accrued to it in virtue of its broad explanatory scope and power, its suggesting of many new and fruitful lines of research, its generation of novel predictions, etc. -- made it reasonable to pursue it, despite being plagued by many apparently disconfirming observations. And -- lo and behold! -- it turned out that the theory was *true*! So Lakatos is trying to avoid a naive kind of Popperian falsificationism that would make it virtually impossible for a research program to get going.

In light of this picture, the application goes something like this: there are worries about recent and past incarnations of auxiliary hypotheses of evolutionary theory that relate to beliefs about the reliability of our cognitive faculties. However, on a Lakatosian philosophy of science, that's not enough to defeat the claim to cognitive faculty reliability, as the justification for this claim doesn't derive directly from local auxiliary hypotheses, but rather from the global justification of the research program of evolutionary theory taken as a whole. So for a Lakatosian like me who also accepts the conjunction of naturalism and evolution, I don't yet have sufficient reason to doubt my cognitive faculties, in which case Plantinga's all-purpose preemptive defeater can't get going. He needs to get me to a point where I doubt the reliability of my cognitive faculties for his all-purpose Reliability defeater to kick in, and he hasn't come anywhere near that with his criticisms of local auxilary hypotheses. So long as evolutionary theory remains a productive research program, I'm entitled to think that *some story or other* must be true that accounts for cognitive reliability, even if current incarnations of the relevant auxiliary hypotheses are inadequate.

Jason Pratt said...

{{However, on a Lakatosian philosophy of science, that's not enough to defeat the claim to cognitive faculty reliability, as the justification for this claim doesn't derive directly from local auxiliary hypotheses, but rather from the global justification of the research program of evolutionary theory taken as a whole. So for a Lakatosian like me who also accepts the conjunction of naturalism and evolution, I don't yet have sufficient reason to doubt my cognitive faculties, in which case Plantinga's all-purpose preemptive defeater can't get going.}}

This is very similar to my own proposed defense against the EAAN, in my previous comment; except that I think I avoided the circular justification involved in appealing by a faculty whose reliability has been suggested to be in doubt, to other results derived from the research program by that faculty. {g}

Leaving that aside--though I don't believe it should be left aside {s}--I think you're actually continuing through the EAAN paradigm insofar as it relies (in any legitimate fashion, which I have some doubts about) on a Bayesian evaluation. A defeater, by Plantinga's own description, is not supposed to be conflated with deductive disconfirmation. At best it can only count as x-amount-of-intuitive-weight against N&E (conjoined per the EAAN). If an individual, such as yourself, then adduces other weights, such as by appeal to results of the program in other regards (again setting aside the special formal problem I think is involved in doing this, in this particular case), then that individual may still decide things weigh out in favor of N&E. I don't see that the EAAN can really close off against that result without appealing to the formal special-case factors; in effect borrowing strength from a different argument altogether (such as Lewis' AfR--which as noted above is not directed against evolution per se anyway.)

I wish you had addressed the topic of auxillary hypothesis vs. preliminary assumption, though.

JRP

exapologist said...

Hi Jason,

I don't think I'm engaged in any sort of circularity regarding justification. On my view, belief in the reliability of my cognitive faculties need not be justified before I can trust them. I hold to a kind of "internalism about justification; externalism about knowledge" epistemology (though not quite of the sort Plantinga holds to). With respect to internal justification, I hold to Michael Williams' "default and challenge" model, according to which beliefs issuing from fundamental sources of justification (perception, memory, introspection, rational insight, reasoning -- and (on my view, even testimony is a basic source (see Coady, Graham, et al for a defense of the latter) are "innocent until proven guily" under normal conditions. If so, then I'm entitled to trust them, under normal conditions, until given principled grounds for thinking otherwise, and so I need not justify their reliability before I trust them. As William Alston has nicely argued in a number of places, any such attempt involves "epistemic circularity" (relying on the very faculties whose trustworthiness you're trying to justify). In any case, given my Lakatosian stance, Plantinga's EAAN isn't sufficiently strong to undermine this prima facie justification, for reasons mentioned previously. (regarding Plantinga on defeaters: I'm not sure if you're right. Plantinga follows Pollock in the basic distinction between undercutting and rebutting defeaters, where a rebutting defeater is any consideration sufficient to make it reasonable to believe that not-P, and an undercutting defeater is any consideration that cancels your justification for believing that P. Now he does a wonderful job of giving a more sophisticated typology of defeaters than Pollock in WCB and EAAN (alethic vs. rationality defeaters, etc.) but for some reason I don't think he precludes deductive arguments from playing the role of defeaters, no? It's been several years since I've read that literature, though, so you may be right). When it comes to knowledge (as opposed to justification), I hold to an externalist constraint that, in addition to being internally justified, my belief must be produced by a reliable process. If so, and a given belief of mine is so produced, then my belief constitutes knowledge even if I don't know that the process is in fact reliable -- it's sufficient that it *is* reliable.

Back to Plantinga and EAAN: the neo-Darwinian research programme (construed along Lakatosian lines) functions, in this context at least, as a way of showing that Plantinga's worries about the reliability of our cognitive faculties on the conjunction of N&E aren't sufficient to call them into doubt to the point that they lack prima facie justification. Now you mention the auxilary hypothesis regarding the evolutionary basis for such reliability in relation to the justifcatory basis for my trusting my cognitive faculties, but that's not the role such the hypothesis plays for me. As I mentioned above, such beliefs need no such justification before I'm internally justified in trusting them.

Re: your question about the difference between various sorts of propositions in a Lakatosian research programme: According to Lakatos, there are at least two basic sorts of claims in a research program:

(i) the "hard core"", which contains the central theoretical and methodological assumptions. These propositions are not to be given up in light of new evidence without thereby abandoning the research program itself. They also function to determine whether a new theory is a part of the same research programme, as opposed to a theory that merely resembles it in various ways. Lakatos came up with the "hard core" classification to avoid Popper's naive falsificationism and Kuhn's challenge that each new broad theory is utterly unrelated to previous theories (in which case there would be no such thing as scientific *progress*).

(ii) auxiliary hypotheses are sets of propositions that are naturally suggested by the hard core, but aren't entailed by it. As such, they can often be exanded, radically modified, and in some cases even discarded in light of new evidence without thereby discarding the research program. Such hypotheses thus function as a "protective belt" around the hard core, shielding it from"defeaters" by apparent counterevidence that a naive Popperian would take to refute the research program. According to Lakatos, a research program is *productive* if it's in a state of producing auxiliary hypotheses that make novel predictions, suggest new and promising lines of inquiry, etc. And a research program is *degenerative* if it's auxilary hypotheses make few if any novel predicitions, fail to suggest new lines of inquiry, and/or the hypotheses have no other function beyond merely protecting the hard core from refutation.

exapologist said...

Whoops!

"...he does a wonderful job of giving a more sophisticated typology of defeaters than Pollock in WCB and EAAN (alethic vs. rationality defeaters, etc.)"

It should read: "he does a wonderful job...in WCB and (his portions of *Naturalism Defeated*..."

Sorry for the general messiness of the post as well, but I'm racing to meet some deadlines, and I only had a minute. I expect tol be able to comment again in about a week or so.

Jason Pratt said...

Hi, Ex! Thanks for the post-in-a-hurry. {g}

{{On my view, belief in the reliability of my cognitive faculties need not be justified before I can trust them.}}

Certainly no disagreement from me on that; and yet this isn’t the defense you were giving (or seemed to be giving) against the EAAN, if P(R:E&N) turns out to be low-or-inscrutable. You were appealing, as I noted, to an effectively Bayesian counterweighting evaluation; which as I noted would be a proper move to make _if_ (R) was an auxillary hypothesis instead of a preliminary assumption. If (R) is a preliminary assumption, then there might be a different defense against the EAAN per se, but it wouldn’t be an appeal to counterweighting evidence for purposes of justifying the rationality of the program against a threat of the program denigrating rationality-at-all.

{{for some reason I don't think [Plantinga] precludes deductive arguments from playing the role of defeaters, no?}}

He doesn’t preclude them from playing the role of defeaters, but his EAAN is avowedly not a deductive argument. (Except insofar as he seems to be trying to get a deductive argument out of it perhaps. {g})

{{I hold to an externalist constraint that, in addition to being internally justified, my belief must be produced by a reliable process. If so, and a given belief of mine is so produced, then my belief constitutes knowledge even if I don't know that the process is in fact reliable -- it's sufficient that it *is* reliable.}}

Admittedly, if a belief of yours is produced by a reliable process, then your belief constitutes knowledge even if you yourself don’t know that the process is in fact reliable -- it’s sufficient that the process is (your own ignorance notwithstanding) reliable. Otherwise we’d all have to be epistemic philosophers before we knew anything. {g}

This is not very helpful though when the reliability of the process, or even the probability of its reliability, has been judged to be low or inscrutable! An appeal will have to be made at that point in defense against the judgment. The “If reliable” has already been answered: probability of reliability given the process is either low or inscrutable.

The defense however cannot be made on grounds reducing back to the process, since the process was already judged to be unreliable. (There are other potential ways to defend, but not by that method--not without circularity of appeal.)

{{Now you mention the auxilary hypothesis regarding the evolutionary basis for such reliability in relation to the justifcatory basis for my trusting my cognitive faculties, but that's not the role such the hypothesis plays for me.}}

Would you say the role it plays for you corresponds to the first or to the second basic sort of claims in a Lakatosian research program? (It would seem to be the first, but then I’m confused why you would bother to recognize it as the second. The two groups do not seem to be capable of overlapping, by definition.)

JRP

exapologist said...

I'm back. Sorry for the delay, Jason.

I think we're somehow not on the same page. My discussion about internal and external epistemic conditions pertain to prima facie justification regarding faculty reliability; appeal to a Lakatosian construal of the evolutionary research program pertains to ultima facie justification regarding faculty reliability. The idea is that I'm initially justified in trusting my faculties (and perhaps I have knowledge here if it meets the externalist condition), and that when Plantinga's worries are considered, they don't undercut that prima facie justification. For the reliability claim embedded within the relevant auxiliary hypothesis accrues justification from the research program taken as a whole, and not merely from the local justification of that auxiliary hypothesis. So there's no circularity here that I can see.

Best,

EA

Jason Pratt said...

Hi, Ex! Glad you're back safe.

I'm not entirely sure you answered my question. I say 'not entirely sure', because it _looks_ as though you answered my question by saying (in effect, though not in so many words) that the hypothesis functions for you both as a category (i) _and_ as a category (ii) Lakatosian claim (and importantly so either way); but then you go on to say that you don't see any circularity in this, so I'm left wondering whether I understood you correctly.

Have you been talking about two categorically different hypotheses the whole time, perhaps (one you're putting in category (i), and one you're putting in category (ii)), and I just haven't figured that out?

JRP

PS: btw Victor, would this be a good thread to ditto over to DangIdea2? It seems to have more than a little to do with _some_ version (or versions) of the AfR.

exapologist said...

Hi Jason,

Thanks. I'm still behind on one deadline, but I'm learning to come to terms with the fact that I'll be late on it no matter how hard I try. :-)

But on to your comments: Actually, I'm not sure if I've held a consistent position throughout this thread, as I'm untangling my thoughts about Beilby's article as I talk about it with you here.

But my gisty way of putting my Beilby-style response to EAAN is that I start of with prima facie justified (henceforth 'PFJ') belief in the reliability of my faculties. Then Plantinga comes along and tries to defeat my PFJ by saying, "Hey there, you're an evolutionist who's also a naturalist, right? Well guess what? It turns out that current hypotheses in that vein that touch on our cognitive faculties provide no explanatory account that would give us reason to suggest that they evolved so as to be truth-aimed. If so, then you have no good reason to trust your faculties." And then I come back at Plantinga and say, "Wait a minute. You haven't yet given me sufficient reason to doubt the reliability of my faculties. All you've done is poke holes in auxiliary hypotheses pertaining to the aetiology of my faculties. But that falsely assumes that evolutionary accounts of faculty reliability accrue justification locally, directly from the relevant current auxiliary hypotheses. Rather, the justification comes from the global justification enjoyed by the evolutionary research program as a whole. It may well be true that current auxilary accounts have massive problems with them. But so long as the evolutionary research program is productive and not degenerative, we're justified in thinking that *some evolutionary account or other (whether currently correctly formulated or not)* must be the correct explanation of faculty reliability.:

It's important to point out that the aetological accounts of faculty reliability aren't contributors of the PFJ of faculty reliability. Plantinga would agree. He thinks our ordinary perceptual, memorial, introspective, etc. beliefs are properly basic, prior to his aetiological account of facultiy reliability in terms of God. Similarly, at least in my view, naturalistic aetiological accounts of faculty reliability aren't contributors to the PFJ of faculty reliability.

Best,

EA