D'Souza vs. Dawkins

I hereby confess to finding fascinating the "debate" between those who believe that religious people are ipso facto irrational (i.e. that they are simply subject to "wishful thinking" or "blind faith"), and their pious opponents, who believe that while they themselves are composed of rubber, atheists, in contrast, are composed of glue, with predictable results where accusations of irrationality are concerned. Even so, the prospect of hearing a debate on this issue between Richard Dawkins and Dinesh D'Souza instills in me a profound desire to punch myself repeatedly in the face. That link, I should point out, takes you not to such a debate, but instead to D'Souza's challenge ("It's time to find out whose position is truly based on reason and evidence and skepticism and science").

I only mention it now because in recent bloviations derived from (and, not coincidentally, flogging) his new opus (What's So Great About Christianity), D'Souza invokes none other than Immanuel Kant in support of his position. Naturally our pedantic pundit ties himself into knots, but they are interesting knots nonetheless, clinically speaking, so I thought we might take a look at them.

According to D'Souza, dogmatic atheists such as Dawkins and Daniel Dennett are in the grip of a fatal fallacy, one with Capital Letters:

The Fallacy of the Enlightenment is the glib assumption that human beings can continually find out more and more until eventually there is nothing more to discover. The Enlightenment Fallacy holds that human reason and science can, in principle, unmask the whole of reality. In his Critique of Pure Reason, Kant showed that this premise is false. In fact, he argued, that human knowledge is constrained not merely by how much reality is out there but also by the limited sensory apparatus of perception we bring to that reality.

So the idea is that "atheists, agnostics and other self-styled rationalists" literally believe that they know everything, or they will once the upstate returns come in, and that what has been or will inevitably be discovered leaves no room for rational people to indulge themselves in the narcissistic fantasy of religious faith. Such dogmatic certainty is clearly fallacious; and, in fact, not even Richard "God Delusion" Dawkins claims otherwise. We hardly need the great Immanuel Kant to tell us this.

What is actually true is this: a) what science has already discovered, while indeed incomplete, leaves no room for rational people to indulge themselves in certain narcissistic fantasies often associated with religious faith; and b) far from being constituted by such inanities, religious faith is better off without them. Of course, to say this is not to end the debate (Dawkins would counter with an invocation of the No True Scotsman fallacy), but to begin it (i.e., a real one). My point in mentioning this here is instead to contrast it with D'Souza's response. As required by his polemical strategy, D'Souza has his naturalist flatly deny the existence of anything beyond our ken, while he himself is open-minded enough to allow the possibility. Typical debating tactic, but again, not exactly, well, enlightening (plus he's got stones calling other people glib).

The natural opponent of dogmatism is skepticism. (You're so sure that _______? Well, think again.) It is for this purpose – to counter fallacious Enlightenment dogmatism – that D'Souza turns to Kant. At first it's not clear why. If we want a skeptic to counter dogmatic overreaching, why Kant? Why not, say, Hume, or even Descartes? The idea that we can know reality whole is Hume's very target; indeed, it is rigorously empiricistic skepticism rather than naive dogmatism that most readily characterizes the scientific enterprise. All that reason demands, say Enlightenment types, is that evidence be submitted for claims about reality; and as Carl Sagan liked to say, "extraordinary claims require extraordinary evidence." This response to religious claims, rather than dogmatically affirming their negations ("we know there isn't anything non-material"), simply denies that they have been proven ("we don't know that there is"). This allows the religious person to parry the dogmatic-atheist thrust, even if to make a thrust of his own (whether of the form "here's your evidence" or "here's why I don't need any") requires some further work. Still, if our concern is to get scientists to acknowledge limits on their knowledge, Hume would seem to be the natural place to go.

Of course, Hume himself was a notorious atheist, or at least a freethinker of some sort; plus there's that stuff about "committing [metaphysics] to the flames, for it is nothing but sophistry and illusion." This renders him somewhat unappealing as an authority for the "greatness" of Christianity (unless you're playing the "even X admits ... " card). On the other hand, Descartes's skepticism is too sophomorically goofy to pass the giggle test in a debate involving non-philosophers (what if we were, like, in the Matrix, man?). Kant, however, is the very model of Protestant probity, and a recognized philosophical titan, with a firmly deontological moral theory to boot. (Not only that, as we'll see, his complex, poorly explicated views make it fairly easy to exploit his writings for rhetorical purposes. But let's not get ahead of ourselves.) Still, for deflecting dogmatic atheism, and thereby making room for the possibility of theism, it seems that some sort of skepticism would be required.

Now of course the problem with skepticism is our unshakeable conviction that we do in fact have real, objective knowledge of the external world. Naturally Kant believes this too, so it's not surprising to hear him say just that (this is the force of his "empirical realism"). D'Souza uses this to concede knowledge of a sort to science, thus adroitly cutting off a possible counterattack: as the moderate voice of sweet reason, he's a modern scientific fellow, neither a loony pre-modern fundamentalist nor a daffy postmodern relativist. Remember, all he needs for his burden-of-proof move is that such knowledge be conceded to be essentially incomplete, not that it be false or uncertain. But again, we hardly need Kant for this; this is standard-issue empirical skepticism, straight out of Hume (or common sense; after all, we'll never be omniscient). So why is D'Souza bringing Kant into it? Watch and learn.

It is essential to recognize that Kant isn't diminishing the importance of experience or what he called the phenomenal world. That world is very important, because it is the only one our senses and reason have access to. It is entirely rational for us to believe in this phenomenal world and to use science and reason to discover its operating principles. But Kant contended that science and reason apply to the world of phenomena, of things as they are experienced by us. Science and reason cannot penetrate what Kant termed the noumena: things as they are in themselves.

Perhaps for lack of an editor, but in any case obscurely, Kant does indeed speak in a number of places of phenomena and noumena as distinct ontological realms. As the first move in his rhetorical conjuring trick, D'Souza shamelessly palms this Kantian distinction off as the Platonic one between a life of mere deception in the Cave and true knowledge of a timeless reality which transcends and underlies human experience. Unlike Kant, Plato believed that the latter was to some degree possible for us through the philosophical use of reason. Kant is concerned instead to establish the limitations on reason; for him, belief in a Platonic realm is the result of "transcendental illusion."

Of course, D'Souza too is concerned with the limits of reason; but his limits are less Kantian than they are Cartesian. Unlike Plato, Kant and Descartes (in his skeptical mode, at least) both deny that reason can provide any knowledge of a realm "beyond appearances." The Cartesian skeptic limits our knowledge to sensory appearances, leaving us in doubt about what (if anything) lies beyond, while Kant's talk of "noumena" seemingly allows a blank affirmation of the existence of that mysterious realm. Thus the appeal to Kant instead, which sets up the theistic punch line to come. For all that's been said so far, though, the only consequence of our "limited sensory apparatus" is the skeptical one we cannot simply rule out dogmatically the existence of a "supersensible" reality on the basis of our empirical knowledge to date. After all, that point constitutes D'Souza's criticism of "self-satisfied atheism" – that it is precisely by so doing that dogmatic materialists rule out the possibility of rational (i.e., non-irrational) religious faith. His conclusion, however, is significantly stronger.

[T]he new atheists and self-styled "brights" can do their strutting, but Kant has exposed their ignorant boast that atheism operates on a higher intellectual plane than theism. Rather, as Kant showed, reason must know its limits in order to be truly reasonable. The atheist foolishly presumes that reason is in principle capable of figuring out all that there is, while the theist at least knows that there is a reality greater than, and beyond, that which our senses and our minds can ever apprehend.

So dogmatic claims of (potentially) universal knowledge constitute an "ignorant boast," and skepticism rightly cuts them down to size by pointing to necessary limits on our knowledge. Fine; but in D'Souza's hands this hard-won knowledge that doxastic modesty is warranted – that our knowledge is limited to (as Kant puts it) the objects of possible experience – is magically transmuted into positive warrant for supposed knowledge of a distinct transcendent realm. Once this slide is made, Kant's criticism of metaphysics licenses a naked affirmation of metaphysical doctrines of the most unashamedly "pre-critical" kind:

Kant's philosophical vision is entirely congruent with the teachings of Hinduism, Buddhism, Islam, Judaism and Christianity. It is a shared doctrine of those religions that the empirical world we humans inhabit is not the only world there is. Ours is a world of appearances only in which we see things in a limited and distorted way, "through a glass darkly," as the apostle Paul writes in his first letter to the Corinthians 13:12. Ours is a transient world that is dependent on a higher, timeless reality. That reality is of a completely different order from anything we know, it constitutes the only permanent reality there is, and it sustains our world and presents it to our senses. Christianity teaches that while reason can point to the existence of this higher domain, this is where reason stops: it cannot on its own investigate or comprehend that domain.

Now we have our answer. We needed the skeptical moment ('scuse my Hegelian) in Kant to combat dogmatic scientism, but that only gets us a wash: neither side can prove its claims. To complete the slide to a positive argument for supernaturalism, we need Kant's positive reaction to Humean skepticism. However, (as Kant shows!) a reaction to skepticism need not be a recoil into dogmatism. Kant indeed restores empirical knowledge (even of scientific laws, which it seemed Humean inductive skepticism had threatened) in the face of the acknowledged limitations of our senses, the significance of which Kant believes both skeptics and traditional metaphysicians misconstrue. But D'Souza goes farther: the Cartesian epistemic limitations of our senses (now curiously identified with "reason") – taken together with Kant's supposed demonstration of the existence of a world beyond them – leave the field open for knowledge to be provided by a distinct faculty not subject to the relevant limitations, one which, in providing knowledge of this transcendent realm, ipso facto shows things as they really are ("face to face," as Paul puts it), in a way that "merely" empirical science cannot. (I qualify this accusation a bit below.)

[UPDATE: I don't mean to imply that D'Souza gets Paul right here, or that Paul is indeed endorsing the Platonic picture. See the comments below for more about Paul.]

In other words, D'Souza's alleged "congruence" of Kant's transcendental idealism with traditional religious doctrine (so construed) is not in the Critique at all (Schopenhauer, maybe). In D'Souza's version, that "congruence" is instead a doctrine of full-on Platonist metaphysics with a Cartesian epistemological twist, complete with a characteristic equivocation on whether we have any real knowledge (resulting from equivocation on the nature of "appearances", to complement that above on "reason"). Even if it were Kant's view – and it's true that this was once a standard reading of Kant (one to which my undergraduate Kant teacher impatiently responded with "come on, read the book a little bit") – those who do attribute this view to Kant have near-universally taken its equivocation about knowledge to constitute what is obviously wrong with it. Here's Kant scholar Henry Allison on the matter:

The most basic and prevalent objection stemming from the standard picture is that by limiting knowledge to appearance, that is, to the subjective realm of representations, Kant effectively undermines the possibility of any genuine knowledge at all. In short, far from providing an antidote to Humean skepticism, as was his intent, Kant is seen as a Cartesian skeptic malgré lui. Some version of this line of objection is advanced by virtually every proponent of the standard picture, including Strawson. [For example,] Prichard construes Kant's distinction between appearances and things in themselves in terms of the classic example of perceptual illusion: the straight stick that appears bent to an observer when it is immersed in water. Given this analogy, he has little difficulty in reducing to absurdity Kant's doctrine that we know only appearances. His [...] main point is simply that this claim is taken to mean that we can know things only as they "are for us" or "seem to us" (in virtue of the distortion imposed by our perceptual forms), not as they "really are." Since to know something, according to Prichard, just means to know it as it really is, it follows that for Kant we cannot really know anything at all. Clearly, such a conclusion amounts to a reductio of the Kantian theory. (Kant's Transcendental Idealism: An Interpretation and Defense, 1983 ed., pp. 5-6, my emphasis)

If science discovers the truth about empirical reality, as all parties claim to admit, then we cannot take this reality qua reality and stick it back beyond the veil in the Platonic manner, let alone Cartesianize it into in-principle unknowability by "reason" (i.e. even by philosophy). Such epistemic nihilism is the very "scandal for philosophy" Kant is determined to overcome.

The essential point is that the incompleteness of our knowledge has nothing to do with its truth. Of course incompleteness is all D'Souza needs for his insipid anti-dogmatic conclusion. But, again, that wouldn't get him what he really wants, which is an a priori demonstration of the existence of a supernatural realm accessible only by "faith." I scare-quote that word because D'Souza here indulges the popular notion (shared, alas, by many of his scientific-rationalist interlocutors) of religious faith as a distinct mode of epistemic access to reality, differing from the senses only in being, well, extra-sensory, like ESP. In fact, that's exactly how he presents the issue. A tape recorder, he says, captures only sound and knows nothing of visible reality; so "[w]hat makes us think that there is no reality [which] lies beyond our perception, reality that simply cannot be apprehended by our five senses?". For D'Souza, what distinguishes the "intellectual plane" of theism from its rival can't amount only to a boring agnosticism about the unknowable; instead, once open to the possibility not of unknowable, but of extra-sensory reality, we may find ourselves presented with as yet unimagined experiences, with accompanying convictions of contact with ultimate reality. (Of course he doesn't say this, but that's the only way to make sense of his position.)

Again, the point about our cognitive limitations would be okay if all it entailed were the triviality that we cannot know that there's nothing out there that we simply can't know about; but then D'Souza spoils that point by conflating it with his Cartesian appeal to perceptual illusion, where what we're missing is not additional knowledge, but actual knowledge of the really real. This allows him to construe faith's independence of the five fallible bodily senses as showing it to be of another epistemic order entirely. The "supersensible" can thus be equated, as required, with the "supernatural," and that with the otherwise inaccessible "reality," and this, finally, with truth itself. Now of course we know that there is a truth about what we believe (that is, that our beliefs are true or false, the meaningful ones anyway), whether it is known to us or not. So what started out as a wash (science can't say whether or not there is a supernatural realm, so it's not irrational to regard the question as at least open, in spite of the methodological materialism of modern science) is now changed, in an instant as it were, to its being irrational to believe that there isn't such a thing – and thus not (at least potentially) to have faith. But of course it hardly takes religious faith to believe that our beliefs have truth values, and only a conjuring trick can make it look like it does – a conjuring trick in which Kant had no part. For Kant, no additional quasi-sensory modality can get us knowledge of things as they are in themselves, so D'Souza's Cartesian talk of tape recorders oblivious to vision is not at all to the point.

D'Souza will reply that he has been careful not to claim knowledge for faith. Kant's argument, he says, "is entirely secular: It does not employ any religious vocabulary, nor does it rely on any kind of faith. But in showing the limits of reason, Kant's philosophy "opens the door to faith," as the philosopher himself noted." He's been careful all right; but the question is not whether it takes faith to "open the door" to faith, but instead what faith does when it comes through that door. And of course the whole point of shoehorning Platonism into the Kantian argument is that the religious person is thereby entitled to a doctrine that "[o]urs is a transient world that is dependent on a higher, timeless reality [which] is of a completely different order from anything we know, it constitutes the only permanent reality there is, and it sustains our world and presents it to our senses." It's true that as religious doctrine goes, this is pretty abstract (if also highly speculative!); but the only way consistently to abjure knowledge of a realm "beyond reason" is to say, with the East, that "the way that can be spoken of is not the way," and respond to any attempt to say more (i.e. add any content whatsoever to our doctrine) with a smack upside the head. As he is committed to a particularly doctrine-laden version of neo-platonism, your typical Christian is in no position to do this. For him, he "knows by faith" about all kinds of things. With respect to knowledge claims, the bumper sticker "God said it; I believe it; that settles it" is only an extreme version of religious dogmatism (that is, cognitivism).

If, as the religious person believes, we have experience of the divine, then Kant's limitation of knowledge to possible experience does not touch religious belief at all. That may seem to defend it against Kant's strictures, as required; but in fact what it means is that Kant's argument here is completely impotent for D'Souza's purposes. The Kantian connection of the intelligibility of belief to possible experience means that in the relevant sense, both religious belief (which if true amounts to knowledge) and empirical belief are on a par, as being knowledge of "appearances" rather than a "transcendent" realm. Only a further equivocation on the notion of "belief" (and the above one on "transcendent") can disguise this. Essentially, D'Souza wants to get for free – from the nature of reason itself – what only further detailed explanation of the nature of faith and knowledge (or, in Hegelian again, Glauben und Wissen) can accomplish. For his argument to work, faith needs both to supply knowledge unavailable otherwise, on the one hand, and on the other, not to supply knowledge at all. (Of course, Kant himself, as the author of Religion within the Boundaries of Mere Reason, has plenty more to say, but as I haven't read that book, I can't comment; but that title sure is suggestive, isn't it? Maybe someone more knowledgeable than I will help out ...)

I mentioned this earlier (plus it should be obvious by now), but it bears repeating. My intention is not to side with Dawkins here. Of course I agree with D'Souza's weaker claim, i.e., that religious faith is (religious believers are) not ipso facto irrational. I merely take issue with his crude and tendentious (and yet ingeniously slippery!) exploitation of subtle philosophical issues in order to score debate points. I see also that I have helped myself, toward the end there, to my own idiosyncratic views about belief qua doxastic commitment, among other things. Much more would have to be said, about faith and knowledge both; but such a real discussion could only begin by rejecting the very idea of this pointless debate.

Breaking news I: D'Souza will debate (not Dawkins, but) Dennett at Tufts on 11/30. Dare I hope that Kant will not be mentioned?

Breaking news II: D'Souza has just unloaded a new, well, load, about how of course we have free will ("I can knock [my] coffee mug onto the carpet if I choose") and how this proves the existence of the immaterial soul. Maybe later, eh?