In Depth: Edward Feser On Biological Reductionism

In the previous installment of our look at Alex Rosenberg’s The Atheist’s Guide to Reality, we began to examine what Rosenberg has to say about biological phenomena.  This time I want to take a brief detour and consider some of what Rosenberg says about the subject in his book Darwinian Reductionism.  I noted that while Atheist’s Guide pushes a generally uncompromising eliminative materialist line, Rosenberg resists the “eliminativist” label where issues in the philosophy of biology are concerned, and presents his views in that field as reductionist.  Darwinian Reductionism (a more serious book than Atheist’s Guide, and of independent interest) explains why.

I have emphasized that though Rosenberg offers no serious criticisms of theism, though his own positive philosophical claims are preposterous, and though his core argument for the scientism on which his entire position rests is worthless, he is nevertheless a more serious thinker than the likes of Richard Dawkins, Jerry Coyne, or other New Atheists.  The reason is that Rosenberg is more consistent than these other writers, and he is more consistent because he understands (as they do not) the grave philosophical challenges facing naturalism.  In particular, he understands that a consistent naturalist must take a radically eliminativist line vis-à-vis intentionality [This is a very key point] — that the naturalist must deny that meaning of any sort exists, even at the level of human thought and language.  And he understands that the reason why the naturalist must take this line is that it follows from the claim that there is no teleology or final causality inherent in the natural order.  Or at least, once you make that anti-Aristotelian move — a move which (as I have argued at length) was definitive of modern philosophy — and you affirm also that the natural order is all that exists, there is no way consistently to affirm that intentionality is a real feature of the world.  For intentionality essentially involves “directedness” toward an object, as a thought is “directed” toward what the thought is about or a word is “directed” toward what the word means.  And to deny that there is any teleology or final causality immanent to the natural order just is to deny that there is any “directedness” of any sort in it [Essentially he is saying you cannot straddle the fence when it comes to believing in purpose. Either the tendency/orientation of objects exists or it exists merely in our heads, there is no gray area where it exists in some objects and not others]— that there is anything that points beyond itself to some end, goal, or object.  (For more on intentionality, see the relevant posts among my many posts on the mind-body problem.)

The intentionality of words is commonly said to be derived, insofar as, apart from human interpreters and their linguistic conventions, the sounds and ink squiggles we think of as words would be entirely devoid of meaning.  “The cat is on the mat” would, apart from us, have no more semantic content than “blah blah blah.”  The intentionality of a thought, by contrast, is commonly said to be original insofar as it is inherent to our thoughts in a way it is not inherent to words.  We use otherwise meaningless ink squiggles and noises to convey meaning, but no one is using our thoughts as instruments to convey meaning.  They just have it, and are the source of the meaning of words and the like.  The eliminativist is committed to the claim that not only what we think of as words, but even our thoughts are really as meaningless as “blah blah blah” is.  For if original intentionality exists, then there exists something irreducibly “directed” toward an object, which there can’t be if there is no teleology in nature.  And without original intentionality, no derived intentionality can exist either.

This position informs Darwinian Naturalism, where Rosenberg writes:

The problem of naturalistically explaining the original intentionality of the human (and infrahuman) brain is perhaps the most serious fundamental challenge facing neuroscience and its philosophy.  No one has yet solved it…  Indeed, if human intentionality turns out to be derived from some evolutionary process as yet unimagined (and it will have to be unimagined so far, if it is to prove “unmysterious”), it will turn out that both artifacts and genomes will be on a par, neither of them deriving their intentionality directly from something with nonderived intentionality, and both tracing their intentionality back to evolution by natural selection.  Of course, as I indicated above, I am dubious that natural selection can actually produce original intentionality in the brain or anywhere else, and so it cannot produce derived intentionality either.  Both will, on my view, turn out to be illusions, like the purposes we overlay on nature and that natural selection has dispelled. (p. 108)

As I have said, I think this eliminativist position is the one a consistent naturalist has to take.  For to say that “directedness” is real but external to the natural order is essentially to adopt a Cartesian position; while to say that it is real and intrinsic to the natural order is essentially to return to an Aristotelian position. [This is the Continental divide that is so significant. Why should I call it a divide? Once water lands on one side or the other, it will trickle downhill towards either atheism or theism, the logical conclusions of either point of view. There is a lot that is needed to back this claim up, which may be presented in due time. Theism and reductionism are not logically exclusive but it would require the theist's belief in God to be based on faith merely on faith (which many appear to be quite happy with…) and can't be considered science in the original sense of the word (scientia–knowledge). In the same way, seeing the teleology of objects does require one to believe in a causa causarum (cause of call causes) unless by the 'exception' of atheistic faith]  And either way, naturalism will have been abandoned.  A serious naturalist, then, either has to find some way out of this dilemma — the dilemma of having to choose either eliminativism or some form of anti-naturalism — or follow Rosenberg in adopting eliminativism and then try to find a way to make eliminativism something other than the incoherent mess that I (and many others) have argued that it is.  Whatever his faults, Rosenberg faces up to the problem in a way that ignorant hacks like Coyne do not.

Or at least he does where the mind is concerned.  While his position also entails (as I suggested in my previous post) a kind of eliminativism about organic phenomena, Rosenberg in general tries, as I have said, to hew to a less radical, reductionist line in the philosophy of biology.  Not that even that line isn’t radical.  Rosenberg’s complaint in Darwinian Reductionism is that most of his fellow naturalists are not sufficiently reductionist.  Like him, they tend to be physicalists, holding that the physical facts fix all the facts; but unlike him, most of them resist reductionism in biology.  Rosenberg (rightly, in my view) regards this position as incoherent.  But rather than taking that as a reason to abandon physicalism (which is what it is), he takes it as a reason to endorse reductionism (committed as he is to physicalism — though there is even less in the way of argument for physicalism in Darwinian Reductionism than there is in Atheist’s Guide).  His aim in Darwinian Reductionism is to argue that biological reductionism does not have the unhappy implications that have led other naturalists to resist it.  

One of those implications is that reductionism would seem to conflict with the evident fact that where biological phenomena are concerned, “the whole is more than the sum of its parts.”  Rosenberg maintains that the sense in which this slogan is true is a sense perfectly consistent with reductionism.  He offers the wetness of water as an example of a feature that the parts of a whole do not have — individual H₂O molecules are not wet — but which is nevertheless reducible to relations between the parts.  And purportedly irreducible biological phenomena (Rosenberg implies) are no different.  

Now since it is the distinctively organic properties of biological phenomena that are supposed to pose a challenge to reductionism, appealing to the wetness of water hardly addresses the biological antireductionist’s concerns.  Moreover, the claim that even water and its properties are “reducible” to H₂O molecules and their relations is more controversial than Rosenberg (and many other naturalists) suppose.  (See e.g. Oderberg’s Real Essentialism and van Brakel’s Philosophy of Chemistry for discussion of the relevant issues.)  But let’s leave all that aside and consider Rosenberg’s response to the specific objection that the feeling of wetness isn’t reducible to relations between H₂O molecules:

But the feeling of wetness is a complex relation between the molecules and our neurological system — more molecules, of course.  It’s not an isolated property of the water.  Reducing the feeling of wetness of water is another matter altogether different from reducing its wetness.  Reducing feelings, sensations, is one that science has yet to accomplish, owing to the incompleteness of our understanding of neurology.  But it would be blatantly question-begging to assert that no macromolecular — that is, reductionistic — explanation of our sensations can ever be provided.  Making such an unargued assumption is very far from taking on the burden of proof.  It is, in fact, a matter of shifting the burden of proof onto the reductionist, and demanding an impossibly high standard of proof: that science should complete the reduction of human neurology in order to show that the wetness of water is equal to the relations among molecules.  Although reductionists cannot accept so high a standard, they can and do argue that the whole history of science since the seventeenth century has been a continuing empirical vindication of reductionism. (p. 13)

Now, I do not necessarily have a problem with the suggestion that the feeling of wetness is a complex relation between water itself and our nervous systems, nor with the idea that the feeling is a material feature of the human organism (though my own, Aristotelian conception of “matter” is not the same as Rosenberg’s materialist conception, and “molecules” are certainly not the whole story).  But the rest of this passage reflects a delusion common in those beholden to naturalism, scientism, and (by extension) modern atheism — the “picture that holds them captive,” as Wittgenstein might have put it.  

This delusion — the delusion that “science has now explained everything else, and so it’s only a matter of time before it explains X (where X = qualia, or intentionality, or some other feature which poses a difficulty for naturalism)” — is one I have elsewhere called “the materialist shell game,” and it works like this.  First, “science” is (implicitly if not explicitly) defined in such a way that no explanation that makes reference to irreducibly teleological or qualitative features of the world is allowed to count as “scientific.” [Want a recent example? Look HERE.] Second, seemingly irreducibly teleological and qualitative features of the world — apparently goal-directed natural processes, say, or colors, sounds, heat, cold, and the like as these manifest themselves to ordinary experience — are re-described as mere projections of the mind onto external reality and not allowed to count as truly “material.” [Why? It is not useful to count them as such, they contribute nothing to our immediate manipulation of material reality for technology and these features are often confused with  truly subjective properties. If a man is color blind and sees red as green, is he seeing red or green. The answer is that he is seeing green because the nature of color is based in the perception of it. However this does not mean that the light waves, by their very nature, should be considered entirely without a reality because it is in the nature of our eyes to perceive the lightwaves in a certain way. This tendency to REALLY see red when our eyes are functioning properly should not be considered any less real just because one eye is not functioning according to its nature. Now of course the next question, is what right do you have to say that the eye has a nature then?]  Teleology, color, sound, heat, and cold as we experience them do not (so the story goes) really exist in the material world itself, but only in our subjective mental representations of it; objectively there are only colorless, soundless, purposeless particles in motion, which by virtue of their motions cause us to experience them as if they had the characteristics common sense attributes to them.  (Color, sound, etc. as physical properties are accordingly also re-defined, in terms of surface reflectance properties, compression waves, and the like.)  Third, it then asserted that “science has explained” such-and-such external material phenomena in a way that makes no reference to irreducible teleological or qualitative features.  What is not acknowledged is that this claim is atautology, since (again) nothing that made reference to such features would be allowed to count as “scientific,” and (again) no features that couldn’t be described in non-teleological and non-qualitative terms would be allowed to count as “material.”  Fourth, it is fallaciously inferred that since this methodological sleight of hand has “shown” that irreducibly teleological and qualitative features do not exist in the external material world, we have every reason to believe that it will also “show” that they do not exist in the mind either — that they will either be reduced to non-teleological and non-qualitative features of the brain, or eliminated altogether.

This is like saying that since we have gotten rid of all the dirt in the room by sweeping it under the rug, we have good reason to think that the dirt under the rug can also be gotten rid of in the same way, and even for thinking that it never really existed in the room in the first place.   Such an argument would be doubly farcical.  For first of all, none of the original dirt has really been gotten rid of at all, but merely relocated.  And second, the “sweep it under the rug” method is for obvious reasons the one method that cannot in principle work when applied to the dirt under the rug itself.  Those naturalists — and they are like the sands of the sea for multitude — who confidently appeal to the historical “success” of scientific reductionism in providing “reductive explanations” of “everything else,” as evidence that the mind too is bound to yield to the same mode of explanation, are committing a fallacy no less egregious.  (And we have already seen why an appeal to the predictive and technological virtues of the methodological stipulations in question is no less fallacious if intended to establish the metaphysical completeness of the reductionist’s picture of the world.)

In Darwinian Reductionism no less than in Atheist’s Guide, then, the scientism that undergirds Rosenberg’s entire position rests on little more than an appeal to ill-founded contemporary academic prejudice.  But there are other problems too.  As I have indicated in earlier posts in this series, one problem with reductionist and eliminativist accounts of this or that natural phenomenon is that they fail to do justice to the irreducible causal powers (arguably) manifested by the phenomenon.  In the case at hand, a reductionist account of biological phenomena would seem implicitly to deny that there are any genuine causal powers at anything higher than the molecular level, and thus seems to entail eliminativism about the biological (as, in my previous post, I suggested Rosenberg’s position does).

In response to this sort of objection (which, as he notes, parallels what is called the “causal drainage argument” in the philosophy of mind, since reductionism seems to make causal power at all higher levels of reality “drain away” to the lowest physical level), Rosenberg tells us that “reductionism does not deny that biological kinds have causal powers — the physical ones; it reveals them” since “’higher-order’ [functional] terms… name the same properties which ‘lower-order’ — macromolecular — terms name” (p. 196).  In other words, it’s not that biological phenomena lack causal powers; it’s that they have no causal powers over and above those of their molecular parts, no powers that are not “physical powers,” which are “the only causal powers there are” (Ibid.)  For biological or functional terms, Rosenberg claims, “do not identify distinct ‘higher-level’ kinds with distinct ‘higher-level’ causal properties.”

This is like saying that Feuerbach’s conception of God as a mere projection, or Freud’s theory of religion as wish-fulfillment, are not really forms of atheism, but are rather merely “reductionist” versions of theism which do not “eliminate” God but merely “reveal” the true nature of His causal relationship to the world.  Obviously, to say that a “higher-level” term like “God” does not “identify a distinct ‘higher-level’ kind with distinct ‘higher-level’ causal properties” but instead names “the same properties which ‘lower-level’ terms name” — viz. human psychological properties like projection and wishful thinking — would just be a roundabout way of saying that there is no God and that only the psychological processes themselves are real.  Similarly, to say that biological terms do not really name any kinds or causal powers over and above the molecular ones is just a roundabout way of saying that the biological kinds and causal powers do not exist and only the molecular ones do.  Rosenberg’s biological reductionism — like the bogus Feuerbachian or Freudian “reductionist” “theism” just described, like reductionism in the philosophy of mind, and indeed like reductionism generally — really is just a thinly disguised eliminativism.  

Except, that is, when it is a thinly disguised Aristotelianism.  In one of the more interesting sections of Darwinian Reductionism, Rosenberg addresses the question of whether genes really can be said to carry “information” and to embody “programs,” as they are commonly said to do.  As I emphasize in The Last Superstition, this sort of talk doesn’t sit well with the naturalist’s official rejection of immanent teleology.  Now, some philosophers of biology regard talk of “programs” and “information” (in anything but the thin Shannonian sense of “information”) as merely metaphorical.  But as Rosenberg indicates, the problem cannot be so easily evaded:

Molecular biology is, of course, riddled with intentional expressions: we attribute properties such as being amessenger (“second messenger”) or a recognition site; we ascribe proofreading and editing capabilities; and we say that enzymes can discriminate among substrates… Even more tellingly, as we have seen, molecular developmental biology describes cells as having “positional information,” meaning that they know where they are relative to other cells and gradients.  The naturalness of the intentional idiom in molecular biology presents a problem.  All these expressions and ascriptions involve the representation, in one thing, of the way things are in another thing…  The naturalness of this idiom in molecular biology is so compelling that merely writing it off as a metaphor seems implausible.  Be that as it may, when it comes to information in the genome, the claim manifestly cannot be merely metaphorical, not, at any rate, if the special role of the gene is to turn on its information content.  But to have a real informational role, the genome must have intentional states. (pp. 99-100)

But how can a naturalist accept such descriptions of biological phenomena as more than metaphorical [This is exactly what I would talk to him about if we were face to face.] — especially a naturalist like Rosenberg, who takes an eliminativist line on intentionality?  

Rosenberg’s solution is to split the difference.  “The crucial question,” he says, “is not intentionality but programming” (p. 108).  What does this mean?  Rosenberg rehearses John Searle’s famous Chinese Room thought experiment, and allows at least for the sake of argument that it shows that running a program is not sufficient to generate original intentionality.  But even if the intentionality of a program is at most derived, the program can still be efficacious.  After all, Searle in the Chinese Room still gives out the right answers in response to questions put to him, even though he has no idea what the symbols he’s manipulating mean; and he does so by virtue of running a program.  Therefore (Rosenberg seems to infer), we can say that genes embody programs, and do their distinctive work by virtue of embodying them, even if we deny that they really possess original intentionality. 

The trouble with this is that it takes account of only half, and the less important half, of Searle’s critique of computationalism.  The Chinese Room argument shows only that computationalism cannot be the whole story about the mind. [It can be part…the material and efficient causes]  But Searle’s later argument against what he calls “cognitivism” — the thesis that the brain is a digital computer (as distinct from the thesis that the mind is the software run on the computer) — is intended to show that it isn’t even part of the story.  If the Chinese Room argument succeeds, it shows that even if the brain is running programs, it could not possess original intentionality merely by virtue of running them.  The later argument purports to show that the brain is not in any interesting sense running programs in the first place.  For being a “program” is not (Searle argues) an observer-independent feature of the world; it does not capture anything intrinsic to the physics of a system.  A physical system “runs a program” only relative to an interpreter, who uses the system to run the program, just as ink squiggles and sounds count as words only relative to an interpreter who uses the squiggles or sounds to convey meaning.  And since no one is literally “using” the brain to run programs — no one is literally saying of each of our brains “Let’s let this brain process count as such-and-such a symbol, let’s let that brain process count as such-and-such a transition between symbols, and so forth” — the brain is not running them.

Now as I argue in The Last Superstition [I highly recommend his book for anyone at any level of philosophy] , one could resist Searle’s conclusion here, but not in a way that helps the naturalist.  If we say that computationalist descriptions of the brain capture real, objective features of the brain, then we are implicitly committed to an Aristotelian conception of matter.  For the notion of a program isteleological; and teleology that is intrinsic to a material system is just Aristotelian, immanent final causality.  As Searle emphasizes, we could in principle attribute all sorts of programs to all sorts of physical systems — to use one of his examples, there is a possible interpretation of the microstructure of the wall in his room on which it is running Wordstar.  To say that some particular attribution is privileged, that a physical system is, apart from our interests, running such-and-such a program and not the others, is to say that its states inherently “point” to the realization of that program rather than to the others.  (The Aristotelian implications of computationalist descriptions of nature have also been noted by James Ross and Valentino Braitenberg.)

This puts Rosenberg in a dilemma.  If he wants to insist that matter is utterly devoid of any inherent teleology — his official general metaphysical position — then (given Searle’s critique of cognitivism) he is not entitled to attribute programs of any sort to genes.  If instead he wants to insist that genes really do embody programs even apart from human interests — his official position on this specific issue — then he has to acknowledge that something like Aristotelian final causality is a real feature of the world after all.  

Rosenberg misses the Aristotelian implications of his position because he evidently thinks that the only alternative to reductionism and eliminativism in philosophy of biology is vitalism.  But vitalism is like Cartesian dualism in the philosophy of mind, or the extrinsic, “watchmaker” model of teleology represented by Paley’s “design argument” — it is a corruption of the Aristotelian-Scholastic views the moderns sought to replace, even if a corruption that is sometimes mistakenly read back into those older views.  For the Aristotelian, the right way to think about the soul is in hylemorphic terms, as the form of the living body, not as a complete substance in its own right a la Descartes’ res cogitans.  The right way to think about teleology is as immanent to the natural order, on the model of organic phenomena rather than on the model of machines.  And the right way to think about life is also in hylemorphic terms, as the possession of a certain kind of formal cause, not as a spooky kind of “force” or elan vital.  

(Rosenberg also seems to think that intentionality-with-a-t entails intensionality-with-an-s, so that since — he argues — genes do not exhibit the latter, neither do they have the former.  As Tim Crane and others have emphasized, though — and as I noted in an earlier post — intensionality-with-an-s is not essential to intentionality-with-a-t.  What is essential is just directedness toward an object, which a thing can exhibit even if we can give a non-intensional description its directedness.)

So, Rosenberg’s attempt to find a reductionist middle ground in the philosophy of biology fails.  His position is unstable, equivocating between eliminativism on the one hand and Aristotelian anti-reductionism on the other.  But there is a lot more to Darwinian Reductionism than I have been able to convey here, and in particular much to make it a more serious and interesting book than Atheist’s Guide.  As a philosopher of religion, and as anapologist for naturalism, Rosenberg is hopeless.  But as a philosopher of science and an interpreter of naturalism, he is consistently interesting.  Both naturalists and their critics need to take him seriously.

I might just start referring to this as the continental divide from now on….some one needs to write a book on this. Unfortunately many people, without ever reading Kant, assume that his conclusions of the impossibility of metaphysics as a science are correct. I'm not saying that they aren't but its important to not take a historicist approach…Kant came after Aristotle therefore Kant is correct. The problems need to be thought through because unfortunately philosophy doesn't always work the same way technology does….the latest is the greatest. I can name a few taking this approach in this past century that cost of few hundred million lives…