What’s Wrong with the World

The men signed of the cross of Christ go gaily in the dark.


What’s Wrong with the World is dedicated to the defense of what remains of Christendom, the civilization made by the men of the Cross of Christ. Athwart two hostile Powers we stand: the Jihad and Liberalism...read more

Spectaclism versus naturalism

Let’s define spectaclism as the theory that what exists is what my spectacles “tell me” exists, i.e. what I am able to see using spectacles. Naturalism is the theory that what exists is what the natural sciences “tell us” exists, i.e. what we are able to learn via their methods.

In support of their theory, naturalists point to the many predictive and technological successes of the natural sciences. In favor of their own theory, spectaclists could also point to the great predictive and technological accomplishments of spectacle-wearers.

Still, no one believes in spectaclism, and the reason is obvious: That a certain method provides us with reliable and useful information about some domain gives us no reason whatsoever to think that what it tells us exists is all that exists. There are other problems too: What sorts of spectacles are the ones we should rely on to tell us what exists? Bifocals? Sunglasses? What color? And why those, exactly?

Notice that the problem here is a failure to keep in mind that metaphysical questions are prior to epistemological or methodological ones. You have to determine first what exists before you can find out whether spectacles tell you all there is to know about it, exactly which spectacles do the job, etc. Spectaclism, in short, gets things back-asswards.

But here’s the thing: Naturalism is exactly as back-asswards as spectaclism is. It is silly to suggest that what exists is only what natural science tells you exists unless you already know through independent means what exists, can compare the deliverances of all putative natural sciences to it, and determine on that basis that such-and-such putative natural sciences alone capture everything there is.

Of course, naturalists will tell us that there is no alternative to natural science – that common sense perceptual experience, introspection, putative religious experiences, and metaphysical inquiry are trustworthy only to the extent that they are vindicated by the natural sciences. But spectaclists could say something similar: Putative alternative sources of knowledge are to be trusted only to the extent that they can be given a respectable spectaclist foundation. “But that’s ridiculous!” Sure it is. So is naturalism.

“Oh come on, we have independent grounds for holding that more exists than is dreamed of in the spectaclist’s philosophy!” Sure we do, but we also have independent grounds for holding that more exists than is dreamed of in the naturalist’s philosophy – for example, grounds derived from common sense perceptual experience, introspection, putative religious experiences, and metaphysical inquiry. “But those aren’t reliable sources of knowledge!” Oh, you mean the way non-spectaclist sources are not reliable? What’s the difference, exactly? And try not to beg the question this time.

“But, but, but, but… naturalism just can’t be as groundless as that!” Wanna bet?

“But most contemporary academic philosophers are naturalists! How can they all be wrong?” Tsk tsk, come now, everyone knows that arguments from authority went out with the Middle Ages…


Comments (52)

I sense that a discussion about empiricism will arise...

The statement about the priority of metaphysics or epistemology is an interesting one, but I don't actually think I disagree with you in substance, Ed, though I probably wouldn't put it that way. I would be inclined to say, in the same place and to fulfill the same function, that metaphysics should not be constrained by methodology. I use "methodology" there instead of "epistemology" so as to leave open the option of saying that one should not believe in the existence of any entity (metaphysics) without having good reason for believing that entity exists (epistemology), which might seem to be legitimately paraphrased as, "Epistemology is prior to metaphysics." But "good reason" can cover _any rational reason_ at all, including situations where the metaphysical belief is itself foundational--known by direct acquaintance with its truth-maker rather than by inference. What is obviously happening here with naturalism is that metaphysical beliefs are being arbitrarily limited to those beliefs one arrives at by a particular, empirical methodology.

It strikes me, by the way, that there may be a consistency problem here for the naturalist. I'm sure lots of other people have brought this up in the literature, but it just occurred to me: Is not the naturalist's belief in naturalism itself a philosophical belief, hence not handed to him by the methods of science? Therefore it does not meet its own criterion. But that's just an off-the-cuff thought. I'm sure this has all been thrashed out at length elsewhere.


So is Bishop Berkeley's "doctrine" worth a second look (or to a philosophy novice like me, a first look)?

After we came out of the church, we stood talking for some time together of Bishop Berkeley's ingenious sophistry to prove the nonexistence of matter, and that every thing in the universe is merely ideal. I observed, that though we are satisfied his doctrine is not true, it is impossible to refute it. I never shall forget the alacrity with which Johnson answered, striking his foot with mighty force against a large stone, till he rebounded from it -- "I refute it thus."

Boswell: Life

Why would you think that, Jeff? Berkeleyan idealism is by no means the only alternative to naturalism.

What I want to know is if anyone has done any empirical studies on Argument from Authority. It used to be that the moon was made of cheese. Then, a bunch of scientists decided it was rock, and when they got there, so it was. Before that the Earth was flat, etc. In all seriousness though, it's this "naturalistic" idealism (how is it that you claim to understand everything from within a closed system?) and the question you posed in your Rosenberg article (if we're just matter, then why does it matter if you're annoyed that I don't think we're just matter?) that saved me through college. I'm quite happy to see these questions popping up since I'd come to them a while back, and no naturalist has been able to answer them yet.

"It strikes me, by the way, that there may be a consistency problem here for the naturalist. I'm sure lots of other people have brought this up in the literature, but it just occurred to me: Is not the naturalist's belief in naturalism itself a philosophical belief, hence not handed to him by the methods of science? Therefore it does not meet its own criterion. But that's just an off-the-cuff thought. I'm sure this has all been thrashed out at length elsewhere."

Yes, this has been hashed out. Mike Rea, in his World without Design, describes naturalism as a research program that must be assumed rather than justified. And my friend Steve Petersen construes naturalism as a methodology, namely inference to the best explanation, which he admits allows for some kinds of theism (Swinburne's, for example) to count as a kind of naturalism, though I don't think he would allow Thomism to. I'll have to ask him.

Lydia, if your question is what I think it is, the point has definitely been made elsewhere. In short, naturalism is self-refuting because it cannot be proved naturalistically.

But Bobcat, if it's just a methodology, then nobody gets to say, "The only entities we can believe in are the ones posited by science" or (I remember a statement like this from Rosenberg), "Scientific facts fix all the facts." In other words, inference to the best explanation is a form. It has nothing whatsoever to do with the types of things we conclude using the form. To call IBE "naturalism" is just a category mistake, if "naturalism" is to have anything whatsoever to do with...well...naturalism. As far as "a research program that must be assumed rather than justified," that just sounds to me like a dodge. What does it even mean? Either it's arbitrarily saying, "Nah, nah, I refuse to be held to any standard of evidence for my assumptions" or it seems vacuous.

I actually teach my students that there is no such thing as accuracy in science, since to know whether or not a measure is correct, you have to know the correct value to begin with. Science, correctly done, is an eliminative process. Naturalism, as explained, here, seems to be a constructive process. Someone explain where I am wrong, please. It seems that science and naturalism should part company at some point. Science can study the metaphysical, after all, whereas naturalism cannot.

The Chicken-Who-Has-No-Clue-What-He-Is-Talking-About


I told to go easy -- I'm like the Chicken (who has no clue). I think of metaphysics as equivalent to "Berkeleyan idealism", but I suspect based on your comment that I'm wrong.

Lydia, that's right--Steve doesn't think that one can say a priori that the only entities we can believe in are the ones posited by science (indeed, so far as I know he tries to avoid a priori claims altogether). He thinks that we ought to believe in whatever entities are part of the best explanation of all the stuff there is. I think he calls that naturalism because he's convinced that we can explain things really well without having to posit God, souls, qualia, etc.

As for Rea's point about naturalism as a research program, keep in mind that Rea doesn't accept naturalism, but rather theistic supernaturalism (as he calls it). He thinks that if you're going to be a naturalist, you can't be a realist about science, which is a consequence he thinks many naturalists want to avoid. Basically, though, his point is, if I understand him correctly, one about evidence--namely that what one counts as evidence cannot itself be fixed by evidence, at least when one is making fundamental judgments about what counts as evidence.

It's best not to go by my summary, though. I see you have the book in your university library, so you may want to check it out from there and read chapter 1 yourself.

Jeff, yeah, metaphysics just means (speaking _very_ generally here, other philosophers, don't jump on me) "the study of or conclusions about what exists." Maybe amended or strengthened to "what fundamental sorts of things exist." So a statement like, "Physical entities are the only things that exist" is itself a metaphysical statement. A statement like, "Everything that exists can ultimately be reduced to minds and physical objects, but neither of those can be reduced to the other" is a metaphysical statement.

Idealism is just one metaphysical position--namely, that minds and their features (such as their ideas) are the only things that exist.

It seems so odd, Bobcat, for Steve Peterson to call his methodology "naturalism" if he just means that the evidence he has _does in fact_ (he believes) support a naturalistic conclusion. I mean, I could see his saying that he's a naturalist, but the term "naturalism" would then refer to the position he's come to as the conclusion of his argument, not to the _type_ of argument (e.g., inference to the best explanation, induction, deduction, or whatever).

I've heard statements like the one you describe as Rea's before. I should read the book. I'm very dubious. Evidence is evidence. Perfectly rational agents with identical evidence come to identical conclusions. It's not like we ought to or have to make some sort of a-rational posit about what counts as evidence. Usually I've seen such statements made in terms of "choosing a worldview" or something like that--that you must choose your worldview first, and then that tells you what counts as evidence. Nuh-huh, as far as I'm concerned.

I think he calls that naturalism because he's convinced that we can explain things really well without having to posit God, souls, qualia, etc.

"Posit" is the key word here, Bobcat. Classical philosophy (A-T, Neo-Platonism, etc.) does not "posit" God, the soul, etc. as the "best explanation" of this or that, any more than geometry "posits" the Pythagorean theorem as the "best explanation" of such-and-such geometrical facts; it claims that these things can be metaphysically demonstrated. It isn't a matter of empirical hypothesis formation. But naturalism simply assumes that all good arguments are ultimately a matter of empirical hypothesis formation; that is to say, it just assumes that classical metaphysics is impossible. Even as "just a methodology," then, it is question-begging, since whether the kinds of arguments it will allow are the only good sorts of arguments there are is precisely (part of) what is in question.

And of course, whether we "posit" qualia in particular, or immaterial thoughts, or objective moral properties, or whatever -- that is, whether these things are theoretical as opposed to part of the data which needs to be explained by any good theory -- is itself also a contentious matter, and the usual reason for treating them as "theoretical" is, wouldn't you know it, a "naturalistic" approach to philosophical theorizing.

That's right, kids, yet more question-begging. Somebody, please, make it stop! Make it stop!

Even though I'm a huge fan of IBE (with a Bayesian gloss), I would agree with Ed both a) that it's arbitrary to insist that IBE be the only form of inference used and also b) that such things as thoughts are data to be explained. As a foundationalist, I would insist that IBE cannot get off the ground without beliefs that are certain, and at that point we're talking about plenty of immaterial things already.

Also, deductive reasoning is certainly appropriate in mathematics, not to mention philosophy itself, and I'm inclined to think that induction is an independent reasoning form different from IBE. So no one should be restricting himself to IBE as the only way in town of being justified in believing something.

Well, I've asked Steve to explain himself here, so if he has time maybe he'll show up. It should be noted that I'm not a naturalist, nor do I think IBE is the only way to achieve knowledge of the world. As for thoughts, I suspect that Steve would say that thoughts are data to be explained, because he may believe theory theory.

Hi folks! This being the academic time of year it is, I don't have a ton of time right at the moment, and I can't pretend even to have read all the comments here closely. I would of course eagerly discuss my notion of naturalism, especially after about December 18th ... !

Meanwhile let me point you to the most recent draft of my paper:


I think the view you find there is not subject to the worries you've brought up so far, but of course you might convince me otherwise.


It should be noted that I'm not a naturalist, nor do I think IBE is the only way to achieve knowledge of the world.

I'm hip, Bobcat -- I realize you were just presenting the view without endorsing it.

Steve, thanks for the link, and I will take a look.

Notice that the problem here is a failure to keep in mind that metaphysical questions are prior to epistemological or methodological ones.

I would say that natural science by definition places epistemology and methodology before metaphysics. Not because scientist don't have any (minimalist?) metaphysical commitments, but the focus is on the questions of how instead of why. Of course there is occasional overlap between the two questions, and given the historical record of those overlapping explanations science has shown the stronger record for accurate predictions.

I just read the paper and liked it a bunch, despite not necessarily accepting it. I'm very curious what ya'll here will think of it.

hey whatever happened to that huge refutation of Kant

Good question, Vernunft, given that I've got so much free time. (Not that I ever said I'd provide a "refutation" of Kant -- just a post about what I think is wrong with him from an Aristotelian-Thomistic-Catholic POV.)

Anyway, I'll get to it when I get to it. Don't hold your breath, though, because it will probably have to wait until next semester, when I teach modern phil again and will thus have reason to think about him anyway.

Heya Bobcat,

My own view is that Petersen's redefining of naturalism ends up making it pretty bland. It's compatible with the existence of God, of immaterial entities, etc, even by his own reckoning. Now, he does equate naturalism with scientism, but he also expands science to include philosophers and philosophical arguments/conclusions - which reminds me of Aquinas treating theology as science (the queen of sciences, if I recall right?) The one line he does draw is against "appeals to mystery", but I think that's very weak (especially given that he seems to have no problem with the idea of positing fundamental entities, as far as his version of naturalism goes).

It seems to me Ed Feser's own views/arguments from TLS can be described as naturalistic given Petersen's naturalism.


I had a quick look at your paper last evening. It's well-written, and you display a good familiarity with the naturalism literature. But I think there are some serious problems with your treatment of IBE.

Here's one way of pressing the trivialization objection. You mention that you think Swinburne is abusing IBE. Suppose you are wrong (as I think you are) and that he is not -- that empirical evidence could, in fact, provide the basis for a legitimate non-deductive argument for the existence of an immortal God, the creator (inter alia) of all physical things, not physical but deigning to manifest His presence sometimes in space and time and thereby to leave the sort of evidence of His existence and actions that we can trace by empirical means. By the standard you seem to be invoking, you would, if the argument were strong enough and robust enough, be forced to admit that such a God exists. However, this is tantamount to selling the naturalist farm. A version of naturalism that is compatible with the existence of an immortal, immaterial creator is not merely toothless: it is naturalism that has lost its dentures and is reduced to gumming on abstracta and muttering about Platonism.

Quite a lot is going to depend, therefore, on that critique of Swinburne's use of IBE -- not on the empirical facts to which Swinburne refers but on the objection that he is abusing the method itself. You sidestep this in the paper with the remark that demonstrating Swinburne's method to be an abuse is not a trivial project. Since Swinburne's formulation is explicitly probabilistic, your options are limited; you might object, as Harman and Foster do, to the explication of IBE in terms of Bayesian probability, or you might object, by parallel to some of Sober's arguments, that the empirical considerations Swinburne uses to justify his probabilities cannot in principle be used where God is concerned. I think that both moves are going to be hard to make, and there isn't much in your bibliography that speaks directly to this problem. For an argument against the former move, see this paper from BJPS 2003; for an argument against the latter, see this one from Philo 2004. For a refutation of Plantinga's criticism of Swinburne, see this paper and this rejoinder to Plantinga's response.

I'm not at all clear that Peterson's form of naturalism is compatible with the existence of _strictly immaterial_ entities. (I read the paper last night.) It _appears_ to me that when he talks about an "explanation" that would (for example) bring God into the realm of what he calls "naturalism," he has in mind the type of example he gives of making astrology part of science by coming up with a _mechanism_ by which the stars really do affect human actions and the events of human life. In that section he mentions the parting of the Red Sea as well. Now, if that's his paradigm case, my impression is that this would end up "allowing" inference to the existence of God only insofar as God's action was explained strictly in terms of a mechanism whereby God affects the world, and that, together with other hints in the paper, leads me to think that it would need to involve turning God (or "God") into a physical being. For example, in one place he criticizes another position for, in his opinion, not being able to give a principled reason for disfavoring a Cartesian dualist who applies for a cognitive science position in a department! He also insists that his position is a good one because it is sufficiently "ideological" and has "teeth" to "exclude" the kinds of things people want naturalism to exclude. The additional statement that God could be the conclusion of a "naturalistic" argument is puzzling in this regard, but I think the whole set (including the marked aversion to Cartesian dualism as a philosophy of mind, which I, of course, took special note of) can be well-explained if we assume that Peterson's notion of explanation would have to involve turning all these entities into physical entities in order to provide the type of mechanistic explanation looked for and implied in the hypothetical regarding astrology.

I had thoughts about Peterson's paper (I think the treatment of mathematics inadequate and the apparently a priori ruling out of a philosophical approach involving "first philosophy" unacceptable), but it's probably unfair to say any of that given that he says he's tied up here at the end of the semester.


I don't see where Petersen is only allowing 'explanations' insofar as it means proposing a mechanism*. He flat out says that Swinburne is using IBE to arrive at God, and that this is fine under his system. He does say that he suspects Swinburne is using IBE "incorrectly", and that it's possible to use IBE "badly". But (and I suspect Petersen is aware of this) people are going to disagree about what is and isn't an incorrect or bad use of IBE. Since his version of naturalism doesn't rule out "theories" (God, soul, etc), only methods (believing in some view "because it is mysterious, or because they hope it is true, or for reasons totally other than explanatory"), that seems to make a tremendous number of theological and philosophical claims IBEs.

* Petersen does quote Alston asking whether what's now considered clearly non-physical - elan vital and psychic forces - should be regarded as physical if physics takes them into account, and he does reply affirmatively. First, the fallout from that re: God strikes me as being the opposite of 'turning God into a physical being'. It's closer to expanding the definition of physical to envelop the traditionally "non-physical" view of God. Second, he then goes on to note that Chalmers' views would fit with his naturalism, even though Chalmers explicitly rejects physicalism - on the grounds that at least what he proposes is "natural", which in turn seems to mean "not an appeal to mystery".

That really seems to mean the bar for IBE is really low - 'don't make a raw appeal to mystery'. If that's the case, then I think it's clear that swarms of philosophical and theological ideas can jump that bar and be "naturalistic" without modifying those ideas. And oddly enough, I can think of some materialist naturalists who would no longer be naturalists given this definition (some of the New Mysterians, like Colin McGinn.)

if physics takes them into account

I take this to mean _gives a physical theory of them_. Which would make the elan vital and psychic forces physical forces, explicable in physical terms. I mean, physics qua physics certainly isn't going to take God into account. Angels, possibly could be, if they have different kinds of bodies rather than being strictly disembodied. (C.S. Lewis imagines something like this with his eldila in his space trilogy. The eldila could in theory be taken into account by a sufficiently advanced physics.) But God is strictly non-physical, except in the incarnation, and God the Father and the eternal Son will never be taken into account by physics in the sense that seems to be imagined for the elan vital.

I suspect, too, that the explication of Swinburne's supposedly "incorrect" use of IBE would involve his not demanding or giving this kind of physics-based explanation.

That, however, is conjecture. But the notion of an "incorrect" use of IBE covers a multitude of prejudices, prejudices that definitely seem to me indicated by the business about the Cartesian mind. If you're just going to _define_ that as "mystery," well, there you go. I would say myself that I use IBE to conclude that the people around me have minds *all the time*. My own mind I know by direct acquaintance and hence not by inference. :-)

I suppose while I'm at it I might as well add this: Peterson's paper does definitely limit knowledge-gaining to the conclusions of IBE. This can't be right, as IBE has to have starting points which will not themselves be the conclusions of an IBE. Classical foundationalism is true, so... :-) I realize the paper couldn't deal with every possible objection, but I think it should at least _acknowledge_ that a foundationalist, and still more a classical foundationalist, will not agree with the epistemological stipulation that IBE is the only type of knowledge-gaining enterprise. (I don't have the paper up to get the exact wording, but it was something to that effect.)

"I realize the paper couldn't deal with every possible objection, but I think it should at least _acknowledge_ that a foundationalist, and still more a classical foundationalist, will not agree with the epistemological stipulation that IBE is the only type of knowledge-gaining enterprise."

From what I recall, and Steve can correct me if I'm wrong, this paper was intended just to be programmatic. Is your point, then, that even in a programmatic overview of his project, he should mention this disagreement with foundationalists?

I think, but not for the record, that Steve is an internalist coherentist, so I think you like at least half of his approach to epistemology.


Even given what "physics" implies, look at Petersen's talk of Chalmers. He seems entirely willing to accept 'immaterial realities as fundamental properties' as compatible with his version of naturalism. Given that, and given his including philosophy as science, I really don't see how he'd rule out, say.. most/all of the arguments/conclusions Ed gives in TLS as being compatible with (his) naturalism. Sure, Ed is absolutely not developing an 'inference' or giving a probablistic argument for God or anything (far from it), but he's also not 'appealing to mystery'.

As for physics meaning "gives a physical theory of them", a lot of that hinges on what is physical. And since quantum physics' arrival on that scene, that seems to be a discussion that has yet to end. Again, Petersen seems to recognize the problem of defining physical given his footnote to Stoljar - which again says to me that Petersen is talking about physics (well, physicists) expanding the definition of physical to include what we used to regard as immaterial.

Actually, that's a question I'd direct to Ed Feser himself, if he read Stevensen's paper. Ed, do you think Stevensen's definition of naturalism allows Aristotileanism and some/most/all of your TLS arguments to be categorized as 'naturalistic'? Inquiring minds want to know!

I tend to think it would be good, Bobcat, even in a programmatic paper, to state that one is *ruling out foundationalism ab initio as unscientific*. Remember, he says that "scientism=explanationism."

The thing is, Joseph, that Peterson expressly says that he wants his position to have "teeth" and to be "ideological." Them's pretty strong words. He criticizes other positions as being merely a sort of "hurray for science" and _in particular_ criticizes more pluralistic forms of naturalism for not being able to rule out (his examples) allowing ID to be taught in schools and being indifferent between a physicalist and a Cartesian applying for a cog. sci. specialty job.

Now, look: The ID folks insist, just as much as does Richard Swinburne, that they are using IBE to arrive at their conclusions. I admit freely that there are some confusions there, because frankly, Dembski's filter ain't IBE, even though Dembski doesn't realize that. It's a point I have made both to Dembski and to Steve Meyer. However, the Dembskian filter is but one way to construe ID, and IBE is very definitely on the table as a way to construe it. How in the world can he allow that it's at least in principle possible that Swinburne's arguments for the existence of God could be "naturalistic" but rule out in principle the possibility that ID could be scientific, taught in schools, etc.? Wouldn't the two be very much on a par? And why is it not possible that a Cartesian mind could be a good explanation?

I don't claim to understand how his position avoids falling onto the very horn of the dilemma that you, Joseph, seem to take to be an _advantage_--namely, that it's "bland" (your word). But that's exactly what Peterson is trying to avoid.

My _guess_ is that in the end the non-bland aspects of Peterson's naturalism are lurking in the details. For example, he could use his "unificationism" to try to rule out the invocation of non-material entities on the grounds that then we aren't "unifying" things as much as if we hold out indefinitely for explaining everything in purely material terms. Or he could, as I've suggested, argue that Swinburne's (and the ID theorists') arguments don't fulfill his criteria of explanationism, because they don't provide a mechanism.

But in this case, of course, the specter of sheer dogmatism raises its head once again.

My own opinion is that there is no way to avoid sheer dogmatism on the one hand and a _purely_ contingent (and rather bland) naturalism on the other hand, where the latter is merely something like, "I could come to be something other than a naturalist if I were convinced that God exists [or whatever], but I'm not convinced yet." But I'm quite sure this latter isn't at all what Peterson is aiming for.

Earlier, Lydia wrote:

"Evidence is evidence. Perfectly rational agents with identical evidence come to identical conclusions. It's not like we ought to or have to make some sort of a-rational posit about what counts as evidence. Usually I've seen such statements made in terms of "choosing a worldview" or something like that--that you must choose your worldview first, and then that tells you what counts as evidence. Nuh-huh, as far as I'm concerned."

I meant to ask at the time: do you know of any papers/books, perhaps of yours or Tim's that address this topic?


I think I get where you're coming from now, and I guess I should be more clear.

I know Petersen states that he wants his naturalism to have "bite", to be ideological, etc. I think he fails badly. Naturalism, as he defines it, is wide open to all kinds of claims, ideas, and conclusions that naturalism is typically supposed to rule out. His main example of what his naturalism has "bite" against is pretty meager - people who believe in things precisely because they are mysterious. That's not ID, cartesian dualism, hylomorphic dualism, classical or other types of theism, idealism, etc. In fact, I think it covers a very small subset even of the one group Petersen points at (non-philosophers, which I take to be the 'average man'.)

If the 'bite' is going to come from anywhere, it's not going to be from Petersen's stated general ground rules for his "naturalism", but from arguments about what is or isn't the IBE itself. Hence his suspicions that Swinburne's IBE is flawed in some way, or "bad". And I'm guessing that would be the route of attack he'd pick for other topics as well, ID included: 'Okay, ID is an IBE, but it's a bad IBE, or an incorrect IBE.' But as I pointed out, people are going to disagree about what is or isn't a correct/good or incorrect/bad IBE. You're probably right that dogmatism could be smuggled into this "naturalism" framework one way or the other - but in that case, the "bite" is coming from the dogmatism and the smuggling. The "naturalism" is still as bland as ever.

I'd also point out that this strategy is suicidal for someone trying to be anti-ID. Right now the best argument against ID (and even it has some problems) is that ID simply is not science: Science has a certain, very narrow set of commitments and can't rule (positively or negatively) on the sort of questions ID brings up. Petersen's definition makes ID (and, I still maintain, Aristotileanism and idealism and vastly more) science - that best argument is gone. In its place would have to be an argument that while ID (among other things) is science, it's bad science. To say that that's going to be a hard case to make, given how re-drawn the lines of "science" are under this naturalism, is an understatement.

And I'm not really saying Petersen's view of naturalism has "advantages", much less endorsing it. In fact, I think it doesn't accomplish what he says he'd like it to, and its only hope of having some "bite" will either rely on dogmatism (as you state) or will be severely diluted by disagreements on what the standards should be for 'best', 'correct', or 'good' inferences. All while opening the gates to the enemy, so to speak.

Hi folks! I checked back at this site and, predictably, can't resist discussion of my own work. It's so much more fun than grading and administrativa!

Thanks very much for your thoughts and comments. Let's see how much I manage to address here after a regrettably cursory reading.

First, it's important to remember that my naturalism is methodological. As I see it, this already means that I don't get to rule out any entities or theories as non-naturalistic "for free", as I would put it - not ID, not even God. Ontological naturalism of the type the original post addresses has serious problems, I think, in Hempel's dilemma. For such reasons I think to assert baldly that "God is not a naturalistic entity" is empty rhetoric. The only way I can make sense of such a claim, on my view, is "God would not be endorsed by ideal and thorough application of IBE." That's still empty rhetoric until cashed out with reasons to show that ideal IBE would not endorse such an entity. Of course I think ideal IBE would not endorse God, but I think work needs to be done to show it. The same goes with ID (with less work required, probably). Again, the same is true of any methodological naturalism; pick any entity e and some knowledge-gaining method M, and I bet you could abuse M badly enough to endorse e.

  • Incidentally I'm not sure the best route against ID is to call it "not science" - any attempt to rule it out as a science is likely to backfire, since whatever you say is science, they just have to do that badly enough. Often I think ID is not science even on my very inclusive picture. But to the extent it does attempt IBE, I'm inclined to think it's simply rotten science. (The problem in the political arena is that "bad science" isn't enough to make ID constitutionally illicit. Luckily, ID is also pretty baldly religiously motivated. If a group of people wanted to insist on giving alchemy equal time with chemistry, I'm not sure what the courts could say!)

You may say that this removes all teeth from my view. Well one way to look at my main goal here is to rally those on board with the methodology - what I think of as "naturalistic" groundrules for gaining knowledge. In the realm of philosophy of religion this means ruling out knowledge via faith, revelation, etc. (Unless revelation is treated simply as testimony from an authority to which one has independently reasoned via IBE - if so then revelation itself is thrown into the IBE hopper as a theoretical posit up for grabs.) If the theist's claim is that - for example - knowledge of the holy trinity is gained by IBE, then I am interested to address this issue on IBE grounds. If the claim is that knowledge of the holy trinity is gained by some other method, then I can't reason with such a position; all I can do is make the normative move of throwing them definitively out of the naturalism camp - even my rather inclusive one! Perhaps I underestimate how many people (philosophers even) came to all the tenets of their religion based on pure IBE? If so, great; if we're agreed on those groundrules, we can do the IBE disputes themselves.

  • Incidentally I regret my way-too-quick potshots in my draft against van Inwagen, Chisholm, and Moore. I think the best case there, though, is against van Inwagen - he has a gorgeous IBE against the existence of free will, and then concludes that we must have it somehow, but who-knows-how. I can't imagine a more explicit appeal to mystery. And from what I know of McGinn's work, I think it's different - he simply says "qualia are a mystery, and maybe essentially so." That's very different from taking a position and defending it by saying it's a mystery how it could be true. (Though McGinn does throw in the towel a little soon for my tastes.)

And yes, of course a great deal of the teeth will come from working out the details of what IBE is. This is of course something I'm working on. (Fancy Turing machine model of unificationism in progress!) I am not against positing new fundamental entites or properties if they make a great deal of sense of things. I am not clear what it means to call such entities "physical" or not, though I think Chalmers' notion of the physical as what's subject to "structural / functional" explanation isn't bad. I am against positing entities if they seem ad hoc or at least as mysterious as the things being explained. Of course all depends on the details of what counts as ad hoc and mysterious. I am inclined to think that something like mechanism is required, unless (as in Chalmers' qualia) it's not the kind of thing with causal powers where mechanisms would be expected. But of course even the claim that "causal stories should involve mechanisms" is a defeasible claim I've reached via IBE, and I'd be willing to reconsider that too if IBE warrants I do so.

Incidentally, as for what counts as data vs theoretical posit, I'm tentatively inclined to take a fairly dialectical position on that. That is, it's a datum if taken as an agreed-upon point for the purposes of reasoning; it's a theoretical claim if it's in dispute. For example, I would happily grant the intuition that we have minds, and even that we have beliefs, as data. But a) that's quite different from the theoretical posit, relative to a discussion among us, that the mind is realized in an immaterial thing, and b) it is possible that in the course of seeking best explanations we throw out some intuitions as bad data. This is what I take the Churchlands to be doing with the intuition that we have beliefs, for example - they explain it away. This is to my mind no different from a physicist explaining away bad data from a smudge on a lens, or even just because the data are too implausible given the rest of theory (as physicists routinely do). (I don't necessarily agree with the Churchlands - I'm just giving an example.) Thus in a substantive (rather than methodological) dispute with a foundationalist, I am unlikely to balk unless their "foundations" are not agreed-upon relative to our debate. If the foundation is something like "God exists", and this controversial point is immovable rather than up for IBE grabs, then that person is anti-naturalist in my book.

  • Another side note: according to my understanding of Bas van Fraassen, he and his scientific anti-naturalist ilk count as something like "spectaclists". It's not as obviously a ludicrous view as the original post makes it out to be (though I don't agree with it myself, as you can tell).

Thanks again for your thoughts and comments. I appreciate the careful attention you've given my work!


Hello all, my comments on Steve's paper ended up being extensive enough that I turned them into a separate blog post, which you can find over at my own blog:



I don't think it's true that the same goes for 'any methodological naturalism' re: ID, and the case doesn't seem as simple as you make it. Me, I don't think ID or no-ID is science anyway, but there are people who make ID claims that are disconnected from any typical religious motivation. (Nick Bostrom and David Chalmers come to mind immediately.) And the religious motivation card, even if valid, can be reversed - there's certainly some people who don't like ID because of their own religious commitments.

What I really have to ask in the end is: Why would you call what you're crafting here "naturalism" anyway? It seems so divorced from what most naturalists tend to be going on about, and in principle the field is wide open for IBEs to arrive at conclusions of God, cartesian dualism, idealism, and who knows what else. Now, I know you have a hearty faith (ironically enough!) that the 'proper' IBEs will ultimately point away from God, etc. But not only is that not at all clear, but conceivably (and to me, it seems very likely) the IBE may either point to God, to souls, etc, or we may end up with multiple "IBEs" about the same question with no clear way to totally resolve which is really the 'best' IBE.

It actually seems analogous to people talking about the importance of (say) reason, logic, etc in coming to conclusions. It's great. Except for all those differing conclusions so many come to on a broad variety of topics.

In defense of Petersen's view as qualifying as a naturalism, perhaps the following analogy will help:

Let's define a libertarian as anyone who holds that everyone owns herself, and no one is allowed to intrude on her self-ownership. Now, G. A. Cohen and Robert Nozick both start with the presupposition of self-ownership. From this presupposition, though, Cohen reaches Marxism, and Nozick reaches the night-watchmen state. It would be hard to imagine people who endorse policies that are more opposed to each other. And yet they have the same starting point. Arguably, they are both libertarians, and they both differ from those who deny the self-ownership premise. Now, we call Nozick a libertarian and Cohen a Marxist. But it's not clear that we have to; we could call Nozick a right-libertarian and Cohen a left-libertarian.

Perhaps we could make the same move with Petersen's kind of naturalism: it's a starting point that allows for a bewildering array of different philosophical "policies", but its starting point still excludes other views that are popular.

Yeah, but the thing is, Bobcat, that in that example the "libertarian" label seems to become trivial -- as many Nozick-style libertarians have complained vis-a-vis the use of the "left-libertarian" label for people like Cohen, Steiner, Otsuka, et al. (And I've often argued that "self-ownership" turns out, on analysis, to be far less determinate and interesting than libertarians suppose, in part for the sorts of reasons underlying what you're saying.) So the parallel seems apt, but in a way opposite to what you intend: In both the "libertarianism" and "naturalism" cases what seemed to be an interesting and important position ends up being indeterminate and trivial.

"So the parallel seems apt, but in a way opposite to what you intend: In both the 'libertarianism' and 'naturalism' cases what seemed to be an interesting and important position ends up being indeterminate and trivial."

Perhaps, but the fact that you, yourself, have written persuasively against the doctrine of self-ownership suggests that it's a position that one can reject. The same would go for extreme IBEism of Petersen's sort. If true, wouldn't that change Petersen's position into something more than trivial or indeterminate?

But what I've argued against are certain specific versions of the self-ownership thesis, while allowing that there is some, not-that-interesting sense in which we own ourselves. Same, mutatis mutandis, for naturalism.

I still don't understand why Petersen has that passage in the paper about not disfavoring the Cartesian in hiring if his view is supposed to be so toothless and so readily compatible with Cartesianism. And what about the stuff at the end about "first philosophy"? If things like sense data, the existence of the self, and such are allowed to be foundations, where's the quarrel with first philosophy anyway? And if foundations of that sort are allowed, then shouldn't the statement about IBE as the only way to knowledge be qualified?

But perhaps it's just that we shouldn't underestimate the expected "teeth" in the "fancy Turing machine model of unificationism."

I would think (haven't had time to read Ed's post yet) that the Petersen position would be incompatible with Ed's A-T view because of the enormous importance of deductive reasoning in the A-T view.

For that matter, would not the Petersen version of naturalism also rule out even W.L. Craig's cosmological argument, since it is deductive in form?

Well, that article is kind of fun. A physicist posts that there are 4 fundamental forces, gravity, electro-magnetism, strong and weak forces. Never mind the fact that these are mere mysterious names for he-knows-not-what. Or rather, do pay attention to that. By these 4 mysterious forces, he explains many ordinary occurrences. The classical philosophical theist, on the other hand, posits forms, matter, and final causes, and by it he explains much that constitutes ordinary every day experience (I won't say he "posits" agents, because that is an accepted data point). It would be very difficult, under Petersen's construal, to show that what Aristotle did in his Physics (and following on in his On the Soul ) was not science at the time. And if science at that time, then it would be possible for it to remain science now. If it were the case that some other later explanation BOTH contradicted Aristotle's conclusions AND came up with a more satisfactory (more unifying) explanation of the very same phenomena, then we might be justified in calling Aristotle's effort bad, or defective, science, but (1) that would not cast it out of the field of science, any more than Newton's work is treated as unscientific now that it has been superceded with relativity and quantum theory; and (2) it simply isn't true that any later theory more successfully handles the very same phenomena that Aristotle's science explains, or even that the later work of scientists categorically contradicts the main theses of Aristotle's. Hence, according to Petersen's standard, there is every reason to claim that classical philosophy is good science. And included in that is natural theology.

Incidentally, as for what counts as data vs theoretical posit, I'm tentatively inclined to take a fairly dialectical position on that. That is, it's a datum if taken as an agreed-upon point for the purposes of reasoning;

That could raise some worthwhile discussions. Let's take the question of whether the following is data: the claims of some thousands of peasants and others who saw the sun dance around in the sky and apparently dash near the ground, and also the widely reported claims of the totally soggy ground being dry right afterwords - all happening in Portugal in 1917. In my view, this would be a "datum" - or rather several data. There could be lots of theories as to whether people "experienced" these events with their eyes and the vision centers in their brains, or rather "experienced" them rather on account of other faculties in the brain that made them "see" things in a way that was independent of what a normal eye would report. But most naturalists would prefer to discount them as data altogether, as being _outliers_ that don't really need to be explained in detail. In fact, most naturalists prefer to think of ALL miracle accounts as either outright mistaken sensory impressions, or defective interior experiences that are inherently unworthy of an IBE account except as _explaining away_ the event.

Another datum occurs to me: the many, many death-back-to-life reports by people who stopped breathing for a time and were pulled back from that condition. So many of them report the experience "moving toward a bright light and a great peace", and enough of these accounts are clearly not influenced by the person hearing prior accounts, that I would consider this a significant datum not readily admitting of being _explained away_.

But I would go much farther than that, and suggest there are numerous items of qualifying _data_ that are ostensibly, at least superficially to be called "supernatural", and that the existing best explanation is precisely that they are supernatural, and will remain so until someone comes up with a better IBE that positively excludes spiritual separated substances as being involved: such as, for example, the repeated reports by Allied pilots of seeing a Franciscan monk, up in the sky over Italy, preventing them from from bombing a certain location - reports by men who had NEVER seen a picture of Padre Pio, and all gave recognizable accounts of him after their flights. The fact is that there are many, many such miracles that are attested to by people who had no a priori basis for being hoodwinked or brain-scrambled into the particular manifestation they experienced. Excluding these as _data_ is just contrary to sound science.

But when you accept these in principle, you open the door to something much more difficult for naturalism to deal with, namely the interior experiences people have of God's presence to them, and other spiritual experiences. Certainly the fact that this is interior means that there is trouble establishing that person A is talking about the same thing that person B is talking about, and if you cannot establish that you have trouble with whether this is an "accepted" or agreed-upon point for reasoning. But this a relatively modest hurdle, not a fundamental abyss. You can get over it with some work. You cannot a priori say that this experience is not part the data to be accounted for. And if you want to try to account for the experiences of a St. Teresa of Avila or similar mystic in a way that satisfies the person who experiences the event, with an IBE account that does not include spiritual and supernatural realities, well, I have never heard of one. So, at least for now, I would suggest that the kind of account that is the BEST account right now is one that includes a supernatural being, and therefore at this time IBE must be considered consistent with theism.

Well, this whole discussion has been interesting, but one question I want to know is: assuming you reject Petersen's account of naturalism, how do you specify naturalism? Is naturalism the view simply that there is no God? No God and no souls? No God, no souls, no free will?

Regardless, how do we, as supernaturalists, define supernaturalism? What is naturalism such that our view is super in comparison to it? Or do we not care about what supernaturalism is?

I suppose this last view would be my own: I believe in God, in free will, in the self, in a moral right and wrong, etc., and I don't much care how you classify me.

Well, this whole discussion has been interesting, but one question I want to know is: assuming you reject Petersen's account of naturalism, how do you specify naturalism?

The assumption that mysteries in this world can be solved through measurement and testable theories.

I would take it that "naturalism" entails physicalism, at a minimum, as a metaphysical position. That's certainly how I have always encountered the use of the term. Further than that, there are various projects to "naturalize" this or that area of philosophy or endeavor, where that usually means getting rid of things like irreducible beliefs or experiences and replacing them with physical descriptions, sometimes with some sort of riff on Darwinism.

It would be possible and might be relevant in some contexts to define the term "natural" (without the -ism) as referring to any event or entity that follows created regularities. In this use of the term, my moving my arm is natural, even though it involves, on my view, the interaction of an irreducible mind with the physical world. But it follows created regularities in the sense that it is natural for me to be able to lift my arm by willing it but not natural for me to be able to lift my computer by willing it. God might, of course, create a psi race whose mind-body regularities were different.

But this latter sense of the term "natural" has little to do with "naturalism," both because I don't know of anyone who would say, "I'm a Cartesian dualist and a naturalist" and because the -ism seems to involve an extra exclusionary notion--or at least, as in Petersen's case, a motivation to exclude--concerning things that are not natural. So as far as I can tell one would have to disbelieve at least in those events that are strictly speaking supernatural (e.g., miracles) in order to be a naturalist on any construal of the natural order.

Let us grant Robert Pennock's argument that it is acceptable to search the literature for references to supernatural phenomena in order to rule such out on a descriptive basis in order to support the idea that methodological naturalism is a ground rule of science.

Now let's turn to the question of whether invisible ninjas might be producing phenomena that we see in the laboratory. Invisible ninjas are not invisible due to supernatural means; let us suppose that they only natural means. Hence, invisible ninjas are natural agents. We search the literature for a reference to invisible ninjas and find none. Hence, we conclude that methodological supernaturalism is a ground rule of science, using Pennock's reasoning.

I also have a problem with scientistic (from "scientism") rhetoric leveraging science which has produced useful technology to support "science" which has produced no useful technology. Didn't Cartwright argue the point about the connection between theory and technology? And didn't Hacking have something useful to say about the state of our knowledge when intervention is possible (say, with respect to Maxwell's Equations) versus its state when intervention is impossible (say, with respect to the Common Descent proposition)?

Finally, I note that Brian Leiter is engaging in demagoguery against poor Thomas Nagel over on his blog. He seems to be trying to form a mob of ideologues posing as philosophers to throw dogmatic, rhetorical stones at Nagel. Those of us who think that Mill had something useful to say should oppose Leiter vigorously, it seems to me.

Not to stray OT, but the problem with getting people to oppose Leiter vigorously is as follows:

--Those whom he can't hurt and who despise him think he is beneath contempt and hence try to waste as little of their time on him as possible.

--Those whom he can hurt, whether they despise him or not, or those who _think_ he can hurt them, are afraid to oppose him vigorously.

But it's telling that he's starting an anti-Nagel campaign. Not surprising, but telling.


you've said in a number of previous posts (both here and on your personal blog that metaphysical questions are prior to epistemological or methodological ones - this may seem a dumb question but how do we build from a metaphysical position say realism to an epistimalogical position ?

Hi again folks!

Thanks again for your thoughts. I've posted a response over at Edward Feser's blog to the general theme that my naturalism has no teeth, but a couple more thoughts here.

First, it is easy to say that you would define naturalism as "physicalism" or "testability" until you look at the complications of what counts as "physical" or "testable". Ontological physicalism is just as susceptible to Hempel's dilemma as ontological naturalism. And it is not at all obvious that naturalism entails physicalism, and that Chalmers can't be (as he claims to be) an anti-physicalist naturalist. (You might not agree with Chalmers - I'm inclined not to about qualia - but in my experience it would a mistake to assume that guy is obviously wrong about anything!) Meanwhile the notion of what's "testable" is enormously complicated post Quine-Duhem. In fact I think the best notion we can recover of "testable" is exactly IBE.

Second, a sidenote on deduction: of course I am not opposed to deduction as an inference method. But deduction - say from premises P_1 ... P_n to conclusion Q - is just noticing that a set {P_1 ... P_n, ~Q} is inconsistent, plus an expression of a preference for which of those to throw out. This kind of point is familiar to those who have heard the old philosophers' joke that "one person's modus ponens is another's modus tollens." On what grounds do we throw out the one claim whose denial becomes the "conclusion"? For that matter, on what grounds do we determine what counts as "inconsistent", given various specifications of logic? Guess what my answer is to those questions!

Okay, ironically enough, time to go decorate a Christmas tree. (Hey, I'm a cultural Christian anyway! It's part of my ethnic heritage, and I celebrate it the way the non-superstitious probably celebrate Halloween.)

Thanks again for taking the time to discuss my work so carefully - it's been really helpful for me.


Hi, Steve,

If you get a chance after decorating the Christmas tree, I'd love a more specific answer about foundationalism. Let me put it starkly: If IBE is the only route to knowledge, then doesn't that commit you to saying that nothing can be known non-inferentially? All of your discussion of dialectical foundationalism doesn't really answer that question. It's sort of a pragmatic answer, not an epistemological one.

I would take it that someone who uses a deductive form of argument for a metaphysical conclusion either

a) thinks that his premises are non-inferentially known--say, are self-evident, known a priori, analytically true, or something like that


b) thinks that his premises are so overwhelmingly supported by empirical evidence, so minimal, so widely agreed-upon, that his interlocutor, who presumably doesn't like his conclusion, will not pick his empirical premise(s) to attack.

And perhaps a combination of these. For example, you might start with a premise concerning the existence of the physical world right now, add in the PSR (knowable a priori), stir, and end up with some sort of cosmological argument.

So if you have no problem with deduction, and if you have no problem with a priori knowledge (but you _must_ have a problem with a priori knowledge, right?), then you're going to have trouble keeping all the interesting arguments in the IBE box.

Hullo Lydia, done with the tree!

As for deduction, I of course don't hold to an analytic / synthetic distinction, so I hear your (a) and (b) options both as saying the deductor holds the premises closer to the center of her web of beliefs than the denial of the conclusion. On what grounds? IBE, of course. The premises may be so close to the middle that they look like the options in (a). (And notice the denial of the conclusion may be only slightly farther out, as in so many philosophical puzzles.) I think if someone counterfactually held any one claim indefeasibly, though, despite data that would make it a poor factor of the whole explanatory scheme, then that someone would be (epistemically not pragmatically) a non-naturalistic foundationalist in my book.

As for whether something can be known non-inferentially, I guess a lot depends on your notion of inference. I'm inclined to think that even animals have knowledge, and that even the "tuna - kitchen - now" type representations my cats have are inferences to the best explanation (like the Peirce quotation in my footnote 19), just fairly rudimentary ones. Also, as Bobcat suggested, I'm a coherentist about internal justification. (I'm a "foundherentist" really, if you press me - and actually I'm strictly an externalist about justification generally, since though I'm interested in internal norms I think there are external ones too.) Anyway I think a belief (or mental representation of whatever sort) had better sit well in the IBE sense with the rest of the beliefs (or representations) to count as knowledge. I do not think such IBE reasoning always needs to be conscious, however; perhaps that's needed for what Sosa would call "reflective" knowledge.

Once the school week starts up again tomorrow I'll be unlikely to be able to come out and play much, though of course I'm interested to keep discussing these points. I've always found it very instructive to "reach across the aisle" where I can and see what a dialog looks like across very different views. It's all too rare these days, don't you think?


Funny how my previous post that I'm sure that I posted at Ed's site regarding Steve Petersen's paper ended up here on WWWTW.

Sorry about the OT comment, Lydia. My defense is that it was minimal and that I thought it important.

Post a comment

Bold Italic Underline Quote

Note: In order to limit duplicate comments, please submit a comment only once. A comment may take a few minutes to appear beneath the article.

Although this site does not actively hold comments for moderation, some comments are automatically held by the blog system. For best results, limit the number of links (including links in your signature line to your own website) to under 3 per comment as all comments with a large number of links will be automatically held. If your comment is held for any reason, please be patient and an author or administrator will approve it. Do not resubmit the same comment as subsequent submissions of the same comment will be held as well.