Naturalism and its Discontents
What is the difference between science and pseudo-science? The criterion by which our current practices distinguish the two is falsifiability, but what is inherently valuable about falsifiable hypotheses? Presumably, the goal of science is the discovery of truth. If an unfalsifiable method predicted data more reliably than a falsifiable one, shouldn’t we adopt the unfalsifiable method? Leon Wieseltier, literary editor of The New Republic, is untroubled by this puzzle or myriad similar puzzles. Or perhaps he has solved them all. That would at least justify the oracular certainty with which he proclaims, in the first sentence of his choleric review of philosopher Daniel Dennett’s new book Breaking the Spell: Religion as a Natural Phenomenon (“The God Genome,” NYT, 2/19/06), that “The question of the place of science in human life is not a scientific question.” Wieseltier seems to believe that only a superstitious dogmatist could possibly deny that statement, and he has a debate-stopping epithet for such a view: “scientism.”
Surely, being an anti-dogmatist, Wieseltier would not object to having his own beliefs subjected to scrutiny. Determining what counts as a scientific question is just as problematic as determining what counts as science. Since Dennett’s project is an evolutionary account of the origins of religious belief, Wieselter might be content with giving the precise statement of his thesis as, “The question of the place of science in human life is not a biological question.” Perhaps so. What about, “The question…is not a question of chemistry”? That seems surer. “…not a question of physics.” We stand on firmer ground still. Consider, however: “The question of the place of science in human life is not a question of cognitive science…of clinical psychology…of sociology…of anthropology.”
Suddenly the thesis no longer goes down like honey — it loses some a priori pull, as philosophers say. What is happening is that the languages of individual sciences become more compatible with questions concerning human life and its values the more that their subject matter is experienced and practical rather than abstract and fundamental. But applied science is not fundamental science; fundamental science is, and what makes a science fundamental is that its truths make the truths of all other sciences true. Sugar doesn’t make coffee sweet if e doesn’t equal mc2. Nor, if the laws of physics fail, does Bloomsday celebrate Ulysses or any priest give spiritual guidance to his flock. (What’s left? The truths of logic.) We use different scientific languages to suit different purposes, but there is no metaphysical difference between what each scientific language describes — it’s all just the same fundamental reality. Any bearing that a truth of one science has on a question is a bearing all science has on that question. So either no scientific practice can inform its own value in human life to any degree or else Wieseltier’s indubitable truism about the role of science in human life not being a scientific matter is simply false.
Don’t take my word for it. Ask David Hume, who wrote the following in the introduction to his Treatise of Human Nature: “’Tis evident, that all the sciences have a relation, greater or less, to human nature; and that however wide any of them may seem to run from it, they still return back by one passage or another. Even Mathematics, Natural Philosophy, and Natural Religion are in some measure dependent on the science of Man.” (His italics, incidentally.) Hume reverses the terms of debate; on his view, the special sciences should inform the practices of the theoretical sciences (he would have suspected a modifier like “fundamental” of invoking sinister metaphysics). Notice that Hume simply takes it for granted that mathematics and theology, as well as physics and ontology, both subsumed under the old concept of “natural philosophy,” belong to the same category of inquiry. Calling science by different names doesn’t change its essence.
One plausible way to understand the origins of what has come to be known as “analytic philosophy,” and has become the hugely dominant mode of doing philosophy in the English-speaking world, is as a way of both systematizing Hume’s insights about the relation of science to human experience, and then carrying out Hume’s programmatic imperatives. On the systematic front, the great Harvard empiricist W.V.O. Quine, in “Two Dogmas of Empiricism,” an essay still fresh and relevant some five decades after its publication, constructed a philosophical theory of why trying to draw metaphysical boundaries between different domains of knowledge is futile (Wieseltier wasn’t the first to try). He concludes:
Ontological questions, under this view, are on a par with questions of natural science. Consider the question whether to countenance classes [e.g. the class of red things] as entities. This…is the question whether to quantify with respect to variables which take classes as values. Now [Rudolf] Carnap [a leading logical positivist] has maintained that this is a question…of choosing a convenient conceptual scheme or framework for science. With this I agree, but only on the proviso that the same be conceded regarding scientific hypotheses generally.
The upshot is that classifying a question as scientific, or not, is an extrinsic matter of choosing a conceptual scheme, and the factors weighing on such a choice are nothing but our subjective preferences about what we hope to achieve with our framework. And this is just Hume’s view paraphrased more generally and formally. In short, science is a human practice, and the claim that its role in human life is not a scientific matter is either an absurdity or a tautology.
The empiricist portrait of the world that Quine inherited from Hume has proven to be compelling to more contemporary philosophers than any other. One of its adherents happens to be Daniel Dennett. By a mistaken interpretation of Hume, based on downright embarrassing scholarship (more on that later), Wieseltier manages to unload a rhetorical artillery barrage on empiricism itself; Dennett is just collateral damage. Underneath the rhetoric and far out of proportion to its volume, Wieseltier has two arguments to make against naturalism. (To be sure, ‘empiricism’, ‘naturalism’, and if one must, ‘scientism’, denote distinct if related concepts. But Wieseltier respects neither these conceptual distinctions nor any others. This is his language game we’re playing.)
Wieseltier’s first argument attempts to prove that naturalism is necessarily committed to biological reductionism, and therefore false, though the last inference doesn’t quite follow. Affirming two observations from Dennett, to the effect that human beings are animals, but that we also have “creeds” and other features that make us different from other animals, Wieseltier quotes Dennett once more: “But it [our difference] is itself a biological fact, visible to natural science, and something that requires an explanation from natural science.” Then comes the victorious thunderclap:
As the ancient rabbis used to say, have your ears heard what your mouth has spoken? Dennett does not see that he has taken his humanism back. Why is our independence from biology a fact of biology? And if it is a fact of biology, then we are not independent of biology. If our creeds are an expression of our animality, if they require an explanation from natural science, then we have not transcended our genetic imperatives. The human difference, in Dennett’s telling, is a difference in degree, not a difference in kind — a doctrine that may quite plausibly be called biological reductionism.
Indeed, whose authority, if not that of the ancient rabbis, is one inclined to draw upon in determining the boundaries of scientific inquiry? If Wieseltier is familiar with the move in analytic philosophy from reductive to non-reductive forms of physicalism and naturalism, he does not say so. But if it were true that features of the world like rationality and belief could be real and irreducible to biology only if no scientific fact explained their existence, then either rationality and belief are just “expressions of our animality” or else some kind of Cartesian substance dualism is true. So if Wieseltier is right that naturalism entails reductionism, he has an either/or choice to make, between becoming a biological reductionist himself or upholding a truly preposterous metaphysics in order not to take back his humanism. (In which case, who is it exactly who believes in a superstition?)
Fortunately for Wieseltier, it need not come to that. The reason to reject fully reductive naturalism, first of all, is not that it would entail unpleasant consequences for our beliefs about having “transcended genetic imperatives” — a truth isn’t any worse off for our uneasiness with it — but because the empirical verdict on full-blown reductionism is already in, and it doesn’t work. Even the most promising candidates for reduction classes and “bridge laws” (as Ernest Nagel called them) turned out to be woefully inadequate to the task. So the claim that all explanation is natural needs to be refined. Non-reductive naturalism is perfectly willing to concede that some truths do not have an ultimate explanation that can be expressed scientifically, but does not concede, and in fact whole-heartedly repudiates, the notion that any truths are not completely dependent on scientific truths. There can be, and undoubtedly are, infinitely many truths — mental, moral, political, aesthetic — that science cannot explain, but a complete statement of the truths that science can explain fully determines all the others.
“Determination without reduction,” as the motto goes, turns out to provide a compelling picture. It respects both our intuitions that, as Dennett puts it, “we are different,” as well as our intuitions that there are no Cartesian minds floating around independent of the laws of physics. Try this thought experiment: Suppose two individuals situated in the same environmental and social contexts are subatomic particle-for-subatomic particle duplicates of one another, hence all their physical properties and their neuronal histories are identical. Is it not obvious that they would have qualitatively identical memories, beliefs, and hopes? In other words, they would in fact have qualitatively identical minds, minds that are no worse off, nor any less real, for being dependent on brains and bodies. The only unreal minds are the ones that are independent of bodies — they don’t exist.
There are thus two ways to construe Dennett’s proposal to explain the existence of humanity’s difference-making creeds through natural science. On one hand, one might emphasize the element of explanation, and conclude that Dennett is simply proposing a reduction of those creeds to biology. If so, Dennett is a biological reductionist. Alternatively, one might emphasize Dennett’s ostensible realism about those creeds. In that case, the object of his explanatory proposal is the subset of physical facts that determines the fact that those creeds exist; and such a proposal is paradigmatic non-reductionism. In no case is Dennett’s doctrine one that “may quite plausibly be called biological reductionism.” It either simply is or is not. Dennett, to be sure, is guilty of equivocation on this point; some of his writing suggests the first construal is the correct one, some of it suggests the second. But what does that matter in the end? If cognitive science, evolutionary psychology, and any other relevant fields could discover the set of facts among the physical ones that is coextensive with the set of facts relating to human mentality and rationality, science might then be able to discover regularities and laws applying to those facts and not others. The sort of research Wieseltier opposes could give us deeper insight into what it means to be human than we have ever imagined possible. This is what it takes to be a humanist?
So much for the notion that naturalism entails reductionism. Wieseltier does not acknowledge his debt to the true author of his second argument against naturalism, but the distinguishing features of a subtler approach to philosophy are readily apparent. Rather than claim on the basis of a few armchair reflections that a probable majority of practicing philosophers subscribe to belief in a contradiction so glaring that it flabbergasts the ancient rabbis in their graves, Alvin Plantinga, in his “evolutionary argument against naturalism,” brought to light a deep, subtle inconsistency for which naturalism has no easy or fully satisfactory answer. There is, however, a steep theoretical price to pay for following Plantinga’s argument to its end, a price too steep for many people upon taking its full measure. Plantinga, to his credit, has the courage of his convictions, an attribute conspicuously absent from Wieseltier’s strategy of co-opting Plantinga’s rejection of naturalism while avoiding its consequences.
Wieseltier poses the argument as a rhetorical question: “[I]f reason is a product of natural selection, then how much confidence can we have in a rational argument for natural selection?” So far, there is only a naked appeal to intuition, an appeal that neither would nor should move any naturalist. The first axiom of naturalism is that reason is a product of natural selection and we can have confidence in it anyway. But Plantinga’s argument, fully developed, can move past an intuitive stalemate. Here is a capsule summary: Modus ponens (if p then q, p, therefore q) and modus tollens (if p then q, not q, therefore not p) are the two atomic forms of inferential reasoning, the building blocks of deduction. Now Plantinga and naturalists both believe that we have a priori knowledge of the validity of inferential reasoning, but the naturalist has the additional belief that inferential reasoning is the product of evolution through natural selection. However, to have a priori knowledge of the validity of inferential reasoning, one would have to have a deductive argument to that effect. And any deductive argument for the validity of inferential reasoning produced by naturalistic evolution would entail recourse to modus ponens or modus tollens, hence begging the question. So the only argument left for the validity of inferential reasoning produced by naturalistic evolution would be an inductive argument. Which would mean that we cannot have a priori knowledge of the validity of inferential reasoning; so naturalism ends in a reductio ad absurdum, on pain of abdicating a claim to a priori knowledge of the truths of deductive logic.
The trouble with Plantinga’s argument is that it is too powerful. His alternative to naturalism is supernaturalism, and he bases his claim to a priori knowledge of the truths of logic on receiving them from God. Very well, but what is the argument for the a priori validity of inferential reasoning given by God? And will it invoke modus ponens or modus tollens? Plantinga believes that the special nature of God lifts the standard constraints on the acquisition of a priori knowledge. But that is a leap of faith, not a rationally justified belief. The evolutionary argument against naturalism exposes the fact that to have any epistemology at all — that is, to have any theory of explanation — one must assume the axioms of that epistemology as primitives. The alternative is nihilism about the possibility of explanation.
The first philosopher to appreciate the problem of induction fully was not a Christian apologist like Plantinga, but the consummate atheist David Hume, who casts his shadow over any discussion of naturalism and the limits of naturalistic explanation because it is his epistemology that provides the theoretical foundation of the actual practice of science. To enter into the discussion, therefore, a working knowledge of Hume’s epistemology is absolutely indispensable. The root source of all of Wieseltier’s trouble is that he gets Hume’s epistemology completely, utterly wrong, and the cause of that error in turn is that, undeterred by a surfeit of biographical evidence and a consensus in Hume scholarship to the contrary, Wieseltier attributes to Hume a belief in the existence of God on the basis of an argument that contradicts the essential character of Humean philosophy. “His God was a very wan god,” asserts Wieseltier, understating matters to the point of absurdity. “But his God was still a god; and so his theism is as true or false as any other theism.” The meager evidentiary basis of that claim is a single sentence outside either of Hume’s two major philosophical works:
The whole frame of nature bespeaks an intelligent author; and no rational enquirer can, after serious reflection, suspend his belief a moment with regard to the primary principles of genuine Theism and Religion. (The Natural History of Religion).
It does look fairly convincing on first glance that Hume both believed in the existence of God, and held that belief on the basis of an argument from design. Indeed, passing off this remark without further consideration of the major themes of Hume’s corpus is so suspiciously convincing that it precludes the possibility that it is just an honest mistake. Quite simply, either Wieseltier has been defrauded himself or he is attempting to defraud his readers. For in An Enquiry Concerning Human Understanding, nearly universally regarded as Hume’s masterwork, and again in the Dialogues Concerning Natural Religion, he gives a devastating counterargument against affirming the existence of God from observations about design. Though Hume’s argument against theism-from-design has been cleaned up and formalized over the centuries, it has never been substantively improved upon.
Would Wieseltier have us believe that the man who constructed the definitive rebuttal to the argument from design nevertheless upheld the argument from design himself? Does Wieseltier believe Hume was a schizophrenic? Not even schizophrenia could make the notion of Hume-the-theist remotely plausible: Hume’s general methodological principles provide rules for rebutting all arguments of the type of which the argument from design is a token, and they also helpfully reveal what is actually going on in the passage from The Natural History of Religion. In contemporary philosophical discourse, “Humeanism” denotes the doctrine, as described by the metaphysician David Lewis, that “all there is to the world is a vast mosaic of local matters of particular fact, just one little thing and then another.” The foundation of Hume’s epistemology is the denial of necessary connections anywhere in nature. In the Enquiry he sets out to demonstrate that all we can ever have knowledge of is the conjunction of one event with another; reason then applies the concept of causality to our experiences and tries to deceive us into thinking that causality is something real, “out there,” rather than a cognitive illusion:
The bread, which I formerly eat, nourished me; that is, a body of such sensible qualities was, at that time, endued with such secret powers: but does it follow, that other bread must also nourish me at another time, and that like sensible qualities must always be attended with like secret powers?
“No,” goes the answer resoundingly. The evident dependence of the existence of causality on the necessary constancy from one moment to the next of invisible “secret powers” should tip us off to the fact that nothing makes it so that uniformities in nature are necessarily so. Belief to the contrary is based on phantoms in the minds of those whom reason has successfully misled. So Hume does not think there is any justification for inferring the necessary existence of cause-and-effect relations from observing nature. The suggestion that Hume believed God’s existence could be inferred from the same method is farcical.
What sense, then, can we make of the solitary line Wieseltier takes as dispositive of Hume’s theism? Quite the opposite, in fact, of what Wieseltier takes away from it. Consider precisely what it is Hume says: No “rational enquirer” can suspend his belief in theism and religion. But Hume is not, in his own idiom, a “rational enquirer”; he is the champion of empiricism, and rationalists are his philosophical antagonists. Of course a rationalist of the sort Hume is criticizing cannot suspend belief in God. Rationalism takes as indubitable the postulate that what pure reason makes out of perception is reality. Anyone laboring under that false doctrine, and who perceives nature as bearing marks of design, would be powerless to resist fallacious inferences from the appearance of design in nature to the reality of the existence of God. In a line from the Treatise I would find it hard to believe Wieseltier has never come across, Hume makes his thoughts about the role of reason overt: “Reason is, and ought only to be the slave of the passions, and can never pretend to any other office than to serve and obey them.” The rationalism Wieseltier believes is common ground between himself and Hume is in other words the precise object of Hume’s intellectual scorn. Hume is not affirming the argument from design, but laughing at those who do.
There are, of course, alternatives to Humean science, and Plantinga points the way to one. Call it Kierkegaardian science — believe in God and, by virtue of the absurd, the science will follow. Aristotle’s science dominated most of the history of Western civilization, until Galileo and Copernicus embarrassed its geocentrism, Newton embarrassed its mechanics, and Darwin embarrassed its notion of biological species as eternal and unchanging. But the undoing of Aristotelian science is its method, not its conclusions. A science according to which penicillin cures bacterial infections because it possesses an antibiotic virtue is not a science capable of discovering penicillin in the first place. Moreover, it takes Kant, not Aristotle, to provide a principled basis for erecting the sorts of walls between science and philosophy and between individual sciences that Wieseltier proposes. The theoretical cost of doing so is accepting Kant’s theory that space and time are nothing more than “forms of sensible intuition,” and consequently that not even the images captured by the Hubble telescope advance us one inch towards an understanding of “things in themselves,” i.e. true, transcendental reality. Wieseltier wants Kantian science without Kantian metaphysics, a possibility ruled out not by the sinister scientistic machinations of the likes of Dennett, but by the minimal requirements of intellectual defensibility.
However, the science Wieseltier actually lends his support to is nothing so dignified as Kant’s, but the only science that could result from the self-parodying rationalism he mistakenly attributes to Hume (and here is where the political implications of Wieseltier’s arguments become apparent). There is unfortunately no shortage of bullies who claim to have proved, on the grounds that it seems to them that “nature bespeaks an intelligent author,” that such an author necessarily exists. The name of that peasant revolt against knowledge is “intelligent design theory,” and Wieseltier, for all his erudition, is its oblivious footsoldier. “[W]hy must we read literally in the realm of religion,” wonders Wieseltier, approximating candor, “when in so many other realms of human expression we read metaphorically, allegorically, symbolically, figuratively, analogically?” What a silly question. Of course we may read any way we choose to, and no one has suggested otherwise. All that naturalists ask is that we not mistake our right to read metaphorically for the power to make metaphors into literal truth by believing in them strongly enough. The occasional stridency Dennett displays in reminding us that the universe is indifferent to our thoughts about what it should be is nothing compared to the metaphysical hubris involved in self-righteously refusing to pay heed to those reminders.
Such hubris, on Oedipus’ part, was tragic; on Wieseltier’s part it is farce. “There are concepts in many of the fables of faith, philosophical propositions about the nature of the universe.” The claim is dangerous nonsense even ignoring the elementary confusion of concepts and propositions (a freshman in introductory philosophical semantics wouldn’t get off so easily). Propositions expressed in fables are categorically not propositions about the nature of the universe; they are at best mimetic representations of the universe. This is why Wieseltier’s charge that Dennett’s book repudiates philosophy comes to nothing but vocus flatus in the end, and why his nihilism is not essentially epistemological, but ontological. For philosophy begins, as Heraclitus and Parmenides knew, with the distinction between appearance and reality, the fundamental and everything else. By imprecating that distinction, Wieseltier’s “humanism” abolishes the very possibility of a distinctly human being, simply because it abolishes the necessary conditions of any being at all.
Daniel Koffler recently
graduated from Yale University with a B.A. in philosophy, and is currently
working on a book about academic cults of personality, tentatively entitled
Contagion of the Gown.