Thursday, June 30, 2005

Richard Carrier Responds

On Wednesday, March 30, 2005, at 12:55 PM, victor reppert wrote:> Richard: I started a blog a couple of months ago. One> series I am developing is going step by step through> your responses to my critique and giving my answers. This is the first installment. http://dangerousidea.blogspot.com/2005/03/dialogue-with-carrier-part-i.html

I finally found the time to read this--I was surprised it is so brief.BTW, I'd like to take a moment to thank you for taking my arguments seriously and patiently and responding rationally and politely--something I miss lately, having had to cope recently with the constant barrage of James Holding's invective now that I have criticized his flagship essay on the origins of Christianity. His treatment of me always calls me to notice and appreciate the kinder, saner voices among Christians who disagree with me. So I really thank you.

Now to the blog entry:
(1) You are correct to ask "Why in the world does he think that I present an argument against the existence of rational inferences" since I did not claim you do in my critique, and never intended to claim so. But in my casual remarks interjected into what you sent me, I did give the mistaken impression (several times) that I was claiming this. That's my fault. What I meant to imply was not that you "present an argument against the existence of rational inferences" but that this is an unforeseen consequence of your AfR. Maybe this will help:If it is the case that rational inferences cannot exist if a particular kind of god (hereafter God) does not exist, then either rational inferences don't exist or God exists. That is what the AfR essentially argues. So let's define the AfR (only for now--we are skipping the nuts and bolts) as: "rational inferences cannot exist if a particular kind of god (hereafter God) does not exist, therefore either rational inferences don't exist or God exists." Let's assume this AfR is true (i.e. its premises and conclusion are true). If this AfR is true then a skeptic can come along and say "Okay, then rational inferences don't exist." How can you refute him? You cannot present a rational argument to the contrary without assuming rational arguments exist (hence begging the question), and you cannot present a rational argument that God exists (and therefore "rational arguments exist") without again assuming God exists (because, per the AfR, only if God exists do rational arguments exist), begging the question again. This is the problem I am talking about. A naturalist does not face this problem because the naturalist does not have to assume God exists, i.e. a naturalist rejects the premise that "rational inferences cannot exist if God does not exist." Now, it a naturalist does face an equivalent problem (as I tried to explain in a Note 6 in my critique online), but even so, it is not the *same* problem, because I am describing a problem that only results from accepting the AfR, and a naturalist does not accept the AfR.But for all that (and I tried to make this clear in the critique itself), I do not believe this particular problem is insurmountable nor do I think you could not surmount it, and hence I do not think this is a fatal objection or even a serious problem for your case. It is merely a problem that will have to be solved some day. And I would be the first to say that naturalism also has its share of problems that arise from its own premises, premises you reject, and therefore I agree naturalism should not leave those problems unsolved, either--I am an equal opportunity critic here, and my own book was an attempt to address problems I think many naturalists wrongly ignore.Hence I started the critique's formulation of this point by noting simply that you don't prove true the premise "If any thesis entails the conclusion that no belief is rationally inferred, then it should be rejected and its denial accepted." That's all. You assert that premise, but do not prove it is true. Nothing more. And as I say, I don't think this premise is false nor do I think you could not prove it true (to a reasonable enough extent), but it is still the case that you merely assume it is true, and therefore your argument is incomplete in that one technical respect (hence I call it a "technical flaw" and not a fatal flaw).But then I argue that this flaw does make things harder for you in a certain way: a naturalist (believes he) can argue from basic observations to the existence of rational arguments, without going all the way to arguing for the existence of something unobserved (God). But if the AfR is true, you have only two options. Either (1) you have to first argue to God and then to the existence of rational arguments, but the only way to argue to God is to presuppose the existence of rational arguments (because God cannot be observed directly, thus there is no other way to argue for his existence but from rational arguments), or (2) you have to argue for the existence of rational arguments, without using rational arguments (and hence presupposing they exist) and without presupposing God exists--and since we can't do that by arguing to God and then to rational arguments, it seems the only means left is to argue from natural facts. But if you can argue from natural facts to the existence of rational arguments, then so can a naturalist, and the AfR is false--unless there is some way to argue from natural facts to the existence of rational arguments that is not available to the naturalist. Again, I suspect there is such an argument. But I think it must be produced before we can get from the AfR to God, rather than from the AfR to worldview agnosticism.Again, there are similar (but not identical) arguments one can advance against naturalism that also invoke the problem of demonstrating that rational arguments exist without presupposing they exist, but the solution to these arguments cannot be the same as the solution required for you (for the simple reason that the argument you need must exclude naturalism).(2) You do try to say now "that belief in rational inferences is properly basic." That, indeed, would constitute an argument against the PS objection--whether it is successful or not, it would at least acknowledge the problem and attempt to solve it. My point was that you did not even do this in the book. But I have no objection to your trying to advance arguments now that close the gap I pointed out, including this argument, that "belief in rational inferences is properly basic." I have certain objections to that argument, but that does go beyond your book and my critique of it. Nevertheless, for your benefit I'll digress on it now, now that you are advancing it as an argument:To be properly basic requires that something be undeniable. Because if it could be false, then it could be as false as true, and therefore to assert that it is true, or even more probably true than false, you must appeal to some evidence--but by definition, if you need to appeal to some evidence before asserting X is true, the fact that X is true cannot be properly basic.That is why (as I explain in my critique of Rea: http://www.infidels.org/library/modern/richard_carrier/rea.shtml) the only candidates I can find for properly basic beliefs are beliefs that certain experiences exist (that I am seeing the color red now is a properly basic belief--or for really extreme skeptics, that there is a seeing of the color red now is a properly basic belief--but that I am seeing a red tomato is not). Everything else is derived from (argued from) this evidence and therefore not basic.The problem for you is this: you can certainly say that "there is an experience now that X is a rational inference" is a properly basic belief, but that X *is* a rational inference is still not basic, because an experience of X being Y does not entail that X is Y. For example, you can say there is the image of a red tomato in your visual field, and that would be a properly basic belief--but as you know, whether there really is a red tomato there, apart from the image of one in your visual field, is another question altogether, and therefore we are beyond properly basic beliefs now. Hence, that rational inferences *exist* cannot be a properly basic belief. That they *seem* to exist can be a properly basic belief, but that's all.Ultimately, I think "rational inferences exist" is a theory about experiences (since it could be false) and thus must be constructed from experiences. How I (as a naturalist) would do that without using rational inferences is a separate problem (and a problem I acknowledge), but however I accomplish this, it will probably be different from how you would do it, because whatever strategy I use, you probably could not use it (for my strategy would be consistent with naturalism, and if the existence of rational inferences is consistent with naturalism, then the AfR is false).In other words, I assume that, without presupposing God is true (or even that naturalism is true), we can produce a justified belief that rational inferences exist--but this will only be from evidence, never a basic belief. And I think the best evidence I have that rational inferences exist is a cumulative case from basic experiences (colors, thoughts, etc.) to physical systems (and the distinctions between truth-finding and truth-missing machines) to aggregate observed facts (the causes and consequences of these different machines in operation), which (if all this is correct) entails that the AfR is false. So you could not use that strategy. But maybe you can find a different strategy. I have never argued that you can't. But nevertheless, you will have to--and that is my point.(3) To draw the point further regarding both (1) and (2) above, when you say "Can't I just point out the disastrous epistemic consequences of not believing in rational inferences?" you are missing the point of the PS objection. A PS proponent would answer: "Yes, the epistemic consequences are disastrous, but reality does not conform to the way we want things to be, and for all we know, the way things are may indeed be epistemically disastrous and there is nothing you can do about that."Let me remind you that, as I say in my critique, I reject PS, so I am not advocating PS here, but that is what a PS proponent would argue. Thus, a PS proponent would say (co-opting your own words) "maybe there are no mathematical inferences and no one ever does science as you understand it and though there may be all sorts of overpaid people who think they are scientists, it may be that there is really no such thing." Moreover, a PS proponent would argue that, if the AfR is true (in the sense defined above in (1)), then we have evidence that all this *is* the case, though because of that very fact, we cannot know whether it is the case because if there are no rational inferences, then we cannot infer from this evidence what is the case, and therefore agnosticism in the form of radical skepticism is what we are left with. You would respond, "But if God exists, we are not in that predicament." To which the PS proponent would respond, "Well how do you know God exists then?" And that is where the problem begins: how do you answer?(4) "Carrier seems to subscribe to a very strong form of epistemological foundationalism" -- I really hate terminology like this, because it implies baggage, and it is not the case that every foundationalist shares the same baggage. To pin me as holding to "a very strong form of epistemological foundationalism" risks leading people (and yourself) to infer things about my views that follow from the textbook definitions of "very strong forms of epistemological foundationalism" but are not in fact true about what I actually hold. Best to avoid terminology and address the facts (or inquire as to the facts) regarding what I actually do hold (my critique of Rea will help, but my book gives a far longer explanation of my epistemology).Case in point: you say for my "very strong form of epistemological foundationalism" that "there is a fact of the matter as to whether some claim or other rests upon a foundation, and this does not vary from person to person." I do not hold that. Rather, I hold that "there is a fact of the matter as to whether some claim or other CAN BE RESTED upon a foundation, and this COULD vary from person to person," since (a) you and I can have completely different evidence that X is true, owing to our having access to different information, but each of our sets of evidence can be sufficient to believe X is true; and (b) once we have established the trustworthiness of a method or procedure foundationally, we do not have to recreate this proof every time, and so we can be justified in believing things without direct foundations so long as we are justified in inferring that the foundations exist.The garden variety example of the latter is trusting what our friends say: we don't have to check their claim that X, though we do have to have evidence that they have been and thus are now very likely to be trustworthy about X--thus, once we have established their trustworthiness, etc., we don't need the foundation for X itself--until something comes up that challenges our inferences here. Thus, in this example, I believe on good evidence that there is a foundation for believing X, simply because I believe on good evidence that my friend would not claim X unless there was a foundation for believing X.This example oversimplifies things, since the reality is that we are engaging in this kind of indirect inference all the time, in many more ways than in regard to human testimony, and my friend's claim that X is true is itself constructed from a matrix of such inferences, and so on. But the simplified example captures the basic structure of what is going on--and that it is tenuous is precisely why we often find ourselves in error. But we find ourselves correct often enough to trust the inferences involved, even though we know they must fail us occasionally. Certainly, if there were a better way available to us (if we could literally be everywhere at every time, if we had the ability to directly sense everything that exists, and so on) we would use it. But we can only use the best tools available to us.You then say "If something is properly basic, or properly grounded, for one person, it must be properly basic or properly grounded for everyone." I am not sure what you mean by this. It does not sound like anything I believe, but I can't say for sure, since I am not sure what you are saying exactly. But as noted above, it is not my view that (as you then say) "there is a set of propositions that are properly basic, and other beliefs are justified just in case they are supported by those beliefs." Rather, it is my view that "there is a set of beliefs that are properly basic, and other beliefs are justified just in case they CAN BE supported by those beliefs" or in a different direction "what makes a belief true is just that it can be justified in reference to some set of beliefs that are properly basic" (though different sets can often be sufficient, so one does not always need the same set). In effect, that last statement is a simple foundationalist definition of "truth."[And of course we are both leaving out the difference between all beliefs as such (all the beliefs that can be), and beliefs an individual actually holds, which obviously is a much smaller set than the beliefs that could be had, which is why we can generally only know something by going out and looking for it, in order to increase the number of our relevant beliefs. But I note this only in case it wasn't obvious.]Again, when you say "this is the thrust of Descartes' project" you seem to have my epistemology all wrong--for I reject Cartesian epistemology. I am very much with his enemies, the empiricists. Descartes wanted to build all knowledge from internal contemplation. In contrast, I think basic beliefs can only be acquired by making empirical observations, and knowledge can only increase by increasing one's store of basic beliefs (even to the extent that knowledge can be increased by understanding better the beliefs we already have, this very activity generates some new basic beliefs in regard to internal sensations). Likewise, Descartes was insensitive to the realities of human practical limitations (such as the complexities of dealing with testimony, of both other people and our own senses). I am closer to the pragmatists in that area, though I am not a Pragmatist as such (I don't define truth in terms of utility).And "the traditional empiricist counter-move is disallow Descartes' doubts about experience" may be true (the history of modern epistemology is not my area of expertise), but if it so then I am not a traditional empiricist. I don't think it is a valid move to simply "disallow" Cartesian doubts. We have to advance reasons for rejecting them in favor of some particular alternative interpretation of the same observed facts, which is why I devote a lot of words to this very project in my book (as you can see by looking up "Cartesian Demon" in the index).(5) Ultimately, I agree completely with you on what you describe as your view:"I think people should reason, but they should reason from within their own belief systems and abandon the beliefs against which there is good evidence. So different people start from different places, but we can all think together toward the common goal of consensus, even though that 'Omega point' seems pretty far in the distance on a lot of issues."I completely concur. In fact, if ordinary naturalism is true, it is impossible that it could be any other way. There is a clear difference between how we got to where we are and how we justify staying where we are--the effort to do the latter, if conducted honestly and rationally and without cognitive impairment (e.g. you are not being systematically lied to, by others or your own senses), will result in everyone relocating themselves to roughly the same place ideologically. But that project cannot begin at birth. We have to develop and be educated into a particular position or worldview before we can start questioning that position or worldview and finding out how to get from there to where we ought to be--i.e. where the truth is (or where it most probably is, so far as we can know). I think we are both on the same page there.And again, there are certainly numerous epistemological problems to solve no matter where you begin or end up. Hence naturalism must provide an answer to PS, too. I am not saying it doesn't. Rather, I am saying that Naturalism can provide an answer that does not require presupposing God exists, but the AfR requires you to presuppose God exists before you can establish that rational inferences exist--*unless* supplementary argument is provided. My criticism of your book was simply that it did not provide that supplementary argument, not that you couldn't have provided it. I strongly suspect you could. Which is why I set it aside in the critique and moved on.(6) You argue in conclusion of this one blog element that "if there are no rational inferences, then the whole process of evaluating arguments, which is what Carrier and I are both doing, makes no sense. Without the use of rational inference, we would have no way of using critieria to determine which arguments are persuasive and which are not." That is not obviously true. Animals make decisions about what it is best to do, and according to you, I presume, animals are not engaging in rational inferences, so reacting to information in a utilitarian way would not, on that assumption, entail the existence of rational inferences. Of course, I would argue that they are using rational inferences (even if in a sense slightly different than we normally mean when we talk about human reason), but again that this debate exists means the truth here is not obvious.But more to the point, a PS proponent would argue that he does not believe that, say, arguments of type X are persuasive, but that at present he finds it convenient to act as if they are, for the simple reason that doing so realizes what he wants (he actually finds himself persuading people), regardless of why or whether it is in any sense true that "arguments of type X are persuasive." Now, I have a whole line of argument in response to this, against the PS position just stated. So I am not saying his position is necessarily defensible. But to reject it, one must actually argue the point somehow. It isn't enough (at least for anyone who takes the truth seriously) to just sweep the PS position under the rug as something merely ridiculous. It may be ridiculous (that's a matter of taste). But it is not self-evident why it is wrong.And more importantly, if you accept the AfR (as defined under (1) above), the number of ways you can rebut the PS proponent here is highly constrained--in a way it is not constrained for the naturalist or even the ordinary worldview agnostic--and therefore defending the AfR puts a new burden on you that is not shared equally by those who reject the AfR. But let me reiterate in closing that I find this to be the least important objection to your book that I raised in my critique--especially since, even if it were the case that you could not solve this problem, if everything else you say about the AfR remained true, then it is still the case that we are not warranted in believing naturalism is true.I hope that clarifies a lot of things.Be well.--Richard C. Carrier, M.Phil.Columbia Universitywww.columbia.edu/~rcc20

1 comment:

Jason Pratt said...

Hey Richard;

I don't remotely have time to go into this in detail ('work' work, editing, holiday weekend, etc.)--but I do have a few comments. (Me being me, my "few" runs 2500-ish words; but much of what I've written below is in total agreement with substantial portions of what you wrote.)

You wrote:

{{a skeptic can come along and say "Okay, then rational inferences don't exist." How can you refute him? You cannot present a rational argument to the contrary without assuming rational arguments exist (hence begging the question...}}

(I know you wrote more, even in that sentence, but I'm focusing on this for the moment. I'll mention the rest of it later, I promise.)

Now, I totally agree with this (and always have--it's been part of my version of the AfR since the beginning; Lewis', too). But I would agree with this whether or not we were accepting (provisionally for purposes of discussion, or otherwise) that some version of the AfR is true.

Yes, it is entirely impossible to inferentially refute an assertion that rational inferences don't exist, without begging the question in favor of their existence. Similarly, it is entirely impossible to inferentially refute the existence of rational inferences without tacitly presuming the failure of the refutation attempt.

The best that our sceptic can do, then, is merely _assert_ that rational inferences don't exist. So what? He can do that all day long; but the moment he tries to _use_ that assertion as a ground for anything, he's toast, because he'll be tacitly (and I've even seen it happen sometimes explicitly) refuting his own assertion.

What is the best that _we_ can do then? ('We' meaning those of us who accept 'the existence of rational inference'--whatever we decide _should be_ meant by that, in this-or-that case and/or across the board.)

Well, obviously we can't (legitimately) prove that rational inference exists. We'll be begging the question in our own favor if we do. While I have criticisms of Victor's book, I don't recall offhand that I ever caught him doing this; but I'm entirely prepared to agree that if I _did_ catch him doing this, I'd have to call foul (on that point at least, plus procedurally on whatever he tries to subsequently build on it afterward.)

What we _can_ do, is prove what consequences follow, either from accepting or denying this as a presumption. (And of course the presumption should be examined and carefully refined.)

Granted, when we do this we will be presuming that rational inferences exist (one way or another). I insistently agree that this isn't the same thing as thereby _proving_ rational inferences exist, though.

What I find, however, when I follow the exercise through, is that I am going to be presuming in _any_ case, whether tacitly or explicitly, that I can make rational inferences. And since, in _any_ case, I am going to be presuming the truth of this (whether or not the premise is actually factual), then I choose to accept the premise on what might be called the most foundationally basic working provision: I really have no working option of doing otherwise, even though technically it may be false after all. The technical possibility of its falsity is not even remotely a possible ground for asserting the opposite premise instead (as your exemplary sceptic is doing), much less using that opposite premise as grounds for any conclusion--the moment I try _that_, I am instantly back to affirming the premise instead: I _can_ (presumably) in fact do rational inferences.


(All of this, of course, requires accepting that reason does _NOT_ "demand belief in infinitely many contradictions", as one intrinsic sceptic recently tried to quote to some of us. But if reason _did_ in fact demand this, then there would be absolutely no point in debating _against_ anything, including against any particular purported "Christian" belief--which the intrinsic sceptic I am thinking of routinely insists on doing.)


Anyway, the long and the short of it, is that _whatever_ I do I'm evidently going to be presuming my own rational inference capability (even when I try to deny it); so I should therefore be following out and accepting the implications of this presumption.

"God exists", is, I find, one of those implications. More precisely, I find that I should deduct atheism out of the philosophical option list, accepting not-atheism instead, which survives the deduction.

This is what the AfR debate is supposed to be about. (Obviously I'm _not_ reproducing the actual argument here, just some pertinent details about it.)

The argument _isn't_ supposed to be formulated as "rational inferences cannot exist if a particular kind of god (hereafter God) does not exist, therefore etc." That's how Van Till tried it, admittedly--which he then, with something like perverse triumphancy, called "glorious circularity" {grimace}--but even some of his greatest fans (like Sproul) strenuously disagree with that procedure. (Me, too... and I'm far from being a fan of his. {wry g})


Now, if Victor is indeed trying that, then yes I agree he should be thwapped. Personally I don't recall that he is, and I'm entirely certain he isn't trying to do it on purpose (unlike CVT for instance); but I'm willing to agree as strongly as _you_ like, that _that_ procedure is fallacious.

One reason I'm entirely certain that Victor _isn't_ trying to go this route (aside from the fact that he occasionally says he isn't {g}), is that he readily and vigorously insists that a conclusion of "God exists" from a successful AfR, does _NOT_ thereby drastically narrow down the list of divine attributes. It doesn't (of itself) solve for supernaturalistic theism instead of naturalistic theism (i.e. pantheism), for instance; it doesn't solve against various sorts of deism; it sure doesn't solve in favor of the Trinity (so far as it goes). I myself doubt it even strikes against some sort of cosmological dualism.

This shows that _procedurally_, Victor isn't beginning with "a particular kind of god (hereafter God)". Whatever mistakes he may be making, he isn't salting the scale by assuming some particular kind of God from the getgo, which his theory wouldn't work without.

I'm a little dubious that he would even agree he was trying to say that _either_ rational inferences _don't_ exist, _or_ else God does exist. (I know I'm not.) It looks to me more like he's saying (more-or-less as I am): rational inferences _do_ exist; therefore, insofar as we accept this as true, we should go on to conclude that God does exist. (That's oversimplifying a lot, but generally that seems to be the shape of his arguments.)

He did precisely this, recently, in his June 29th comment on Dennett's "well-known paper about intentionality". (Actually, he pointed out that an atheist fellow graduate student came up with the argument he inverts against Dennett; but I know Victor well-enough to know he means much the same thing in what he himself is writing.)

Observe, in this example: what is Victor's _premise_?

_Not_ "2x. A particular kind of God exists".

Instead, his premise is "2. Human beings do possess original intentionality."

He accepts what he judges to be Dennett's argument ("1. If [atheistic] naturalism [and thus hereafter] is true, then humans cannot possess original intentionality.") By changing the subsequent test-premise, Victor arrives at "3. Therefore, naturalism is false."

Whether Dennett is actually doing what Victor reports or not, it should at least be clear that Victor is doing exactly the opposite of what you're faulting him for (and indeed is attributing precisely the same methodology you're faulting, to Dennett instead.)

Granted, I don't think Victor's rephrased argument there strictly holds water (sympathetic toward its result though I am). The original if/then argument of point 1 does not, in itself, entail that the falsification of its result (by premise or whatever) requires the falsification of its 'if'. And obviously there is at least one major branch of atheists who would strenuously disagree that the truth of naturalistic atheism _necessarily_ entails the falsity of original intentionality. (Heck, if it comes to it, Dennett himself strenuously disagrees with it on a regular basis--whenever he insists on receiving credit for his own work, for instance. {g})

Still, I think in fairness you should be prepared to agree, that Victor (at least sometimes) is agreeing _with you_ about the faultiness of a particular methodology; and that he is giving us at least _some_ textual evidence of doing something very different, in _his_ AfR, from the methodology you're attributing to him. (As far as I recall, he's doing this in CSLDI, too.)



In summary: I agree it's entirely proper to fault an AfR proponent if he begins by _assuming_ the existence of a particular kind of God in order to justify rational inference capability, along the way to concluding thereby that we should believe God exists. I don't think this is what Victor is doing (usually anyway, certainly not on purpose if ever he does make a slip on it); but I totally agree that you should have insuperable problems with it. I totally agree, because:

1.) it begs the question at least once, and maybe twice;
and
2.) it's just about the total reverse of an argument _FROM_ reason. {g} (i.e. you're describing a theistic argument _to_ reason from theism, thence to theism--which in fact has been attempted by some apologists, I'm sorry to say.)

I also agree that a more rigorous AfR proponent (such as myself, and I think Victor, too), should _not_ attempt to _prove_ (whether sooner or later) that we can make rational inferences. On the contrary, the most rigorous AfR proponents will be admitting and insisting that the capability of rational inference (and/or of 'original intentionality') cannot be proved per se. The best we can do is prove that we're going to be presuming it anyway for any argument we make, even if we try to argue to-or-from its denial. I simply choose to accept that I can never legitimately escape accepting its truth, and then I go onward to work out the consequent implications: if true, then what else should I be believing to be true? (Note the difference however between what I'm doing here, and a claim that a belief in rational inference is "properly basic", at least in the sense of being undeniable.)

Similarly, however, I would ask you to totally agree in turn, that _I_ should have insuperable problems when I find a scientist or philosopher (atheist or otherwise) attempting to use rational inferences, either to prove that we can in fact make rational inferences, or else to prove that we cannot in fact make rational inferences but that our apparent 'arguments' can reach trustworthy answers anyway.

Such as when you say "a naturalist (believes he) can argue from basic observations to the existence of rational arguments" and "I assume that... we can produce a justified belief that rational inferences exist." What fascinates me, and I think points to the crux of the question of atheism vs. not-atheism, is why after strenuously agreeing that this is an _improper_ procedure in principle, you go on to say that this methodology is proper anyway.

I think it's because the implications of atheism _require_ that either this, or the other option I mentioned two paragraphs ago, must be attempted sooner or later. (But this is hardly something I've established here.)


Something you mention later has bearing to what I've just discussed, so I'll address it here, too:

{{when you [Victor] say "Can't I just point out the disastrous epistemic consequences of not believing in rational inferences?" you are missing the point of the PS objection. A PS proponent would answer: "Yes, the epistemic consequences are disastrous, but reality does not conform to the way we want things to be, and for all we know, the way things are may indeed be epistemically disastrous and there is nothing you can do about that."}}

Personally, I grant this. And the PS proponent could answer this to my own procedure above, as well.

But it wouldn't be much to the point. The PS proponent is pretty obviously going to go right on ahead believing that he is _not_ an epistemic disaster, precisely because there is nothing he can do about it; more precisely, because he knows quite well that he can't accomplish anything unless he robustly presumes that he can in fact make rational inferences.

More precisely still: the PS proponent and I are _both_ going to go right on ahead believing _my_ presumption, _not_ his, and acting on our acceptance of the truth of it. The difference, is that he's going to dump our common presumption the moment it looks convenient to him (for various reasons) to dump it; which probably feels pretty safe because he knows, deep down, that he isn't _really_ going to dump it. (As exemplified above: the person who declares that "Reason demands belief in infinitely many contradictions" obviously doesn't coherently believe that for a single moment; otherwise he wouldn't _himself_ be making a truth declaration about what "Reason demands", which he expects to be of some use in accomplishing something.)

I can either choose to be consistent with what I discover I'm going to be presuming anyway to be the truth whenever I try to make _any_ argument; or I can be like the PS proponent and do it anyway but say I'm _not_ doing it. {shrug}

A PS proponent can certainly _say_, "maybe there are no mathematical inferences and no one ever does science as you understand it and though there may be all sorts of overpaid people who think they are scientists, it may be that there is really no such thing"; but the moment he goes on to _argue_ anything (such as, for instance, against the AfR reaching a true and useful conclusion) then he's at least tacitly abandoned that list of maybes. (Or at least he's tacitly agreeing there are such things as metaphysical inferences and metaphysics, etc.)

I don't consider toothless bluffs, immediately retracted and denied by the proposer when it comes time for _him_ to do something, to be valid ripostes.

(I remember you said you didn't accept PS, and aren't advocating it here. But I think you're wildly overestimating their relevance in disputing against anything, especially the AfR. Though admittedly not against the goofy inverted VanTillian-AfR you're mainly arguing against above; but even in zorching _that_, which I certainly agree should be zorched, our PS-er is tacitly affirming the central presumption used in a proper AfR, and refuting his own position.)


Fwiw, I think I agree with the foundationalism you're describing; though I'm somewhat doubtful I would be an _empirical_ foundationalist. Still, I'd probably be willing to go pretty far even with that, so long as it doesn't require denying or justifying my inferential capability (or other similar missteps).

aaaannnd... now it's mid-afternoon, and I haven't done editing today. argh. {sigh}

Jason Pratt