Teleology revisited

Over at his own blog, One Brow (who sometimes comments on this blog) responds to my recent post on the Wright-Coyne-Manzi debate. I appreciate his thoughtful remarks, but I must say that I find his post a bit frustrating. He admits that he has not read The Last Superstition, but puts forward criticisms of its arguments anyway – criticisms that I have already addressed at length in the book! I know that some readers think I drop references to TLS into various blog posts simply as an act of self-promotion; and if it helps to sell a few copies, I certainly don’t mind. But the main point is always to indicate to interested readers where certain issues relevant to the topic of a post have been addressed at greater length. I cannot reasonably be expected to recapitulate the arguments of TLS every time I revisit one of its themes. That’s why people write books: To develop at length arguments and ideas that cannot adequately be dealt with in shorter contexts (e.g. blog posts).

Anyway, one of One Brow’s comments that I would like to address here concerns the central topic of the post to which he is responding, viz. the way in which talk of “algorithms,” “information,” and the like in biology evinces, if it is meant seriously, a tacit commitment to the reality of teleology or final causes. One Brow says:

terms like algorithm and information means [sic] very different things when applied to biological objects or other non-human-created systems than to computer programs. Algorithms are merely cycles enacted and altered by external stimuli and ended, if at all, by other stimuli (possibly external or internal), while information represents how easily a string can be compressed by interpretations [sic] functions that are not aimed specifically at that string. Neither concept has any recognition of teleology or lack thereof, they are statements of content, not purpose.

The problem with what One Brow is saying here – if I understand him correctly, anyway – is that he is just mistaken in assuming that his claims are necessarily inconsistent with the claims I made. And what this reflects is a basic failure to understand what the Aristotelian-Thomistic (A-T) tradition means by “final cause" or “teleology.” (Like Coyne, One Brow seems annoyed that defenders of Aquinas are always complaining that critics don’t understand his arguments. And yet rather than show that the charge is false, the critics’ responses only ever seem to provide further evidence for the charge! What can I say? Don’t blame the messenger, guys…)

Let me make some general remarks about what the A-T tradition does mean, then, before coming back to One Brow’s comment. If you are going to understand Aristotle and Aquinas, the first thing you need to do is put out of your mind everything that you’ve come to associate with words like “purpose,” “final cause,” “teleology,” and the like under the influence of what you’ve read about the Darwinism vs. Intelligent Design debate, Paley’s design argument, etc. None of that is relevant. If you think that what Aristotelians or Thomists mean when they say that teleology pervades the natural world is that certain natural objects exhibit “irreducible specified complexity,” or that some inorganic objects are analogous to machines and/or to biological organs, or that they are best explained as the means by which an “Intelligent Designer” is seeking to achieve certain goals, etc., then you are way off base. I realize that that’s the debate most people – including writers of pop apologetics books – think that arguments like the Fifth Way are about. They’re not. Think outside the box. “What hath Thomas Aquinas to do with William Paley?” Nothing. Forget Paley.

In fact Aristotle and Aquinas are concerned with something far less high-falutin’ than all that. The core of the A-T “principle of finality” can be illustrated with the simplest sort of cause and effect relation you might care to take. As Aquinas sums it up: “Every agent acts for an end: otherwise one thing would not follow more than another from the action of the agent, unless it were by chance” (Summa Theologiae I.44.4). By “agent” he doesn’t mean only conscious rational actors like ourselves, but anything that serves as an efficient cause. For example, insofar as a chunk of ice floating in the North Atlantic tends, all things being equal, to cause the water surrounding it to grow colder, it is an “agent” in the relevant sense. And what Aquinas is saying is that given that the ice will, unless impeded, cause the surrounding water to grow colder specifically – rather than to boil, to turn into Coca Cola, or to catch fire, and rather than having no effect at all – we have to suppose that there is in the ice a potency, power, or disposition which inherently “points to” the generation of that specific effect. That the ice is an efficient cause of coldness entails that generating coldness is the final cause of ice. And in general, if there is a regular efficient causal connection between a cause A and an effect B, then generating B is the final cause of A.

Now already I can hear some readers – for example the sort, like the sands of the sea for multitude, who made snotty and uncomprehending remarks in response to Manzi’s response to Coyne (as you’ll find if you can stomach plowing through Coyne’s combox) – sputtering replies like the following: “So what divine ‘purpose’ is the ice supposed to serve, then? To chill our martinis? To give furriers a market for their products? What superstition! And what about that iceberg that sank the Titanic? What about hypothermia, frostbite, and the ‘brain freeze’ I suffered through the last time I had a Slurpee? Where’s the omni-benevolence of your Flying Spaghetti Monster sky-god now, huh? HUH?!”

Whoa, whoa, whoa. Slow down, and calm down. Nobody said anything about either human purposes or divine purposes. Indeed, there is nothing whatsoever in the specific claim under consideration that has anything to do with “purposes” at all, if what is meant by that is the idea that the ice or the coldness serve some end beyond themselves in the way that a bodily organ functions for the good of the organism of which it is a part, or a machine serves the ends of its designer. To be sure, each of the latter examples would involve teleology of a sort; but it is not the sort in question here. The claim so far is only that where there is an efficient causal connection between A and B, then generating B is the final cause of A in the sense that A inherently “points to” B or is “directed at” B as its natural effect. That’s it.

So far, then, nothing has been said about either “design” or a “designer,” because the point has nothing to do with design. Nor does it have anything to do with complexity, “specified” or otherwise. We’re talking about ice here – ice! – not the bacterial flagellum, eyeballs, or any of the other hoary chestnuts of the Darwinism-versus-ID dispute. Indeed, we’re talking about something many naturalistic philosophers have come to endorse in contexts far removed from philosophy of religion or the Darwin wars – albeit without realizing that they are more or less reviving a Neo-Scholastic philosophy of nature. When a mainstream naturalistic philosopher like David Armstrong speaks of the “dispositions” physical objects possess as manifesting a kind of “proto-intentionality,” and when a mainstream naturalistic philosopher like George Molnar argues that the causal powers of material objects exhibit a kind of “physical intentionality,” they are certainly not claiming that there is an intelligent designer who made the world with certain ends in view. But they are (even if unwittingly) more or less stating in modern jargon what the A-T tradition meant by the principle of finality. (As usual, see TLS – and, now, Aquinas – for more.)

As Christopher Martin notes in his important book Thomas Aquinas: God and Explanations, modern philosophers tend to think that, where teleological arguments for God’s existence are concerned, getting from the existence of teleology to the existence of God is easy, but establishing that there really is any teleology in the natural order in the first place is difficult or impossible. But as Martin also notes, this is more or less the reverse of the view taken by thinkers like Aquinas. For Aquinas, it is easy to show that teleology exists; for without it, efficient causation becomes unintelligible. (As I have noted many times, the moderns’ abandonment of final causality is the source of all the puzzles about causation that have plagued modern philosophy since Hume.) What takes work is showing that the existence of teleology entails the existence of God. After all, Aristotle himself, even though he firmly believed both in final causality and in the existence of an Unmoved Mover, did not think that final causality needed an explanation in terms of the Unmoved Mover, or indeed any explanation at all. He took it to be just a fundamental feature of the natural world; his argument for the Unmoved Mover begins instead with the existence of change or motion, not the existence of teleology.

Aquinas disagrees with Aristotle here. But, just as when arguing for the existence of teleology, so too when arguing from the existence of teleology to the existence of God, Aquinas does not appeal to “irreducible complexity,” to the way biological species are adapted to their environment, to the “fine tuning” of the laws of physics, nor to any other of the evidences emphasized by modern proponents of the “design argument.” Nor does he argue from a purported “analogy” between the universe and the products of human design. Nor does he weigh probabilities or argue “to the best explanation.” Again, you need to put Paley and Co. completely out of your mind. And again, the basic idea is much simpler than all that. It is essentially this: For a cause to be efficacious – including a final cause – it has actually to exist in some way. It’s not just that for A to be the efficient cause of B, A must exist – as it obviously must – but also that for B to be the final cause of A, B must also exist, in some sense, otherwise, being nonexistent, it could not be efficacious. Hence for the “coldness” that the ice generates to function as a final cause, it has to exist in some way; for an oak to function as the final cause of an acorn, it too has to exist in some way; and so forth.

Now there are only three options here: B must exist either in the natural world; or in some Platonic heaven, as a Form; or in an intellect which “directs” A towards B as A’s natural end or goal (as a carpenter has the table in his intellect as the end or goal of his hammering and sawing). Now by hypothesis, B does not exist in the natural world: the whole point is that the coldness that the ice will produce, or the oak that the acorn will grow into, have not yet come about but are initially merely “pointed” to by the ice or the acorn. Nor does B exist as a Platonic Form – at least not if, like Aquinas, one endorses moderate (or Aristotelian) realism about universals, instead of Platonic realism. The only place left for B to exist, then, is in an intellect; and it must be an intellect that exists outside the natural order altogether. For the causal relations in question are totally unintelligent: ice and acorns do not have intellects, nor is there any intelligence at the level of the even more fundamental causal processes studied by basic physics and chemistry. And all the intelligence that does exist within the material world – in us, for example – presupposes the operation of these unintelligent causal processes (since the existence of our bodies, and thus of us, presupposes them). So, there is no place left for the intellect in question to be than outside the natural order. That is to say, all the causal relations that exist in the natural order exist at all only because there is an intellect outside the natural order which “directs” causes to their effects.

Obviously this line of argument raises all sorts of questions: Why accept the metaphysical assumptions underlying the argument? Why assume that there is only one such intellect directing efficient causes to their effects, or that it has all the various divine attributes? Why should we believe that an intellect could be something outside the natural order, and thus something immaterial, in the first place? All good questions, and all dealt with in The Last Superstition and (in greater detail) in Aquinas. But the point for now is to give a sense of how very different is the argument summarized in Aquinas’s Fifth Way – and like all the Five Ways, it was only ever meant to be a brief summary, not a self-contained one-stop proof – from Paley’s “design argument.”

In particular, in addition to the differences already noted, there is this crucial one: To reject Paley’s divine designer is ipso facto to reject the “design” Paley claims to see in nature. But to reject Aquinas’s notion of a divine intellect is not ipso facto to reject the existence of teleology. One could instead adopt Aristotle’s view that teleology is just a basic feature of the natural order requiring no explanation. To be sure, this may not be a defensible position at the end of the day – that teleology ultimately entails a divine intellect is precisely Aquinas’s claim. But the point is that, as Aquinas acknowledges and Paley and his successors do not, the inference from teleology to an ordering intelligence is not immediate. There is logical space for an alternative understanding of teleology, and it requires significant philosophical work to rule that alternative out. Establishing the existence of teleology in the natural order is a necessary condition for the success of an argument like the Fifth Way; it is not a sufficient one.

Now, let me return, at last, to One Brow’s remarks. One Brow says that to describe natural selection and other natural processes as “algorithmic” is simply to note that they are “cyclical.” And what he means by this, I gather, is that they embody regular causal patterns (in particular, patterns of what Aristotelians call efficient causation). Competition between species leads to changes in the gene pool, planets tend to orbit stars in patterns roughly conforming to Kepler’s laws, and so forth. But there is, One Brow says, no “purpose” being served by any of this. The purpose of natural selection is not to lead to better organisms, because it has no purpose; and the purpose of planetary orbits is not to generate seasonal changes on the planets themselves, because they have no purpose either. A causes B in a cyclical pattern, with no external purpose being served by that pattern; and that’s that.

If this is what One Brow means, though, he is not saying anything that is incompatible with what I have been saying. For whether natural selection, planetary orbits, or anything else serves a purpose in the sense in question is irrelevant to the existence of teleology. The claim isn’t that the fact that A causes B in a cyclical pattern entails that there is some plan or purpose outside the cycle that the pattern exists in order to further. The claim is that the mere existence of the cycle is all by itself a manifestation of teleology. The argument isn’t “A tends to cause B; therefore there must be some purpose outside of both A and B the realization of which this causal relationship exists in order to further.” It’s rather “A tends to cause B; therefore, causing B must be inherent or natural to A.”

If that claim sounds obvious and trivial, then terrific: You’re starting to understand Aristotle and Aquinas, because it’s supposed to be obvious and trivial. Or rather, it would be trivial if not for three factors: First, if Aquinas is correct, this obvious and seemingly trivial fact cannot be explained unless there is a divine intelligence directing causes to their effects. Obviously that is a substantive claim, not a trivial one. Second, when you work your way up through ever more complex levels of material reality, you find correspondingly more complex manifestations of final causality or teleology, and in the case of human beings this has significant moral implications. (Indeed, from an A-T point of view it is a precondition of there being any such thing as morality at all.) Third, what is obvious and trivial to common sense has become obscured by 400 years worth of intellectual squid ink, as modern philosophers have moved ever farther away from the classical tradition and indulged in ever more bizarre exercises in what P. F. Strawson called “revisionary metaphysics.”

One Brow also denies that attributing “information” to natural processes implies the existence of teleology. But there are two problems with this. First, if he means that “information” in the technical, Claude Shannon sense doesn’t by itself entail semantic content of the sort we associate with purpose, then he’s right, but that doesn’t undermine the claim at issue. For where A carries Shannonian information about B, that is only because there is a causal connection between A and B, so that we are back to the original A-T point that such an (efficient) causal connection presupposes final causality.

Second, it is not at all clear that scientists who speak of “information” and the like really do confine their usage to the narrow, Shannonian sense of the term. As John Searle (among others) has complained for decades now, fast-and-loose computer science and information-theoretic talk pervades contemporary intellectual life, and has afforded materialistic explanations (e.g. of the mind) a specious plausibility they would not have if the relevant terminology was used more precisely. The thing is, some of this talk – that is, some of the talk that attributes something like semantic “information” to material processes – is by no means unmotivated. As Daniel Dennett likes to say, there are “real patterns” in nature underlying the “intentional stance” we find it useful to take toward certain natural phenomena. And that these patterns seem to require an intentional or semantic description is evidence of even richer levels of teleology than the sort I’ve been describing in this post.

Such richer levels of teleology – in complex inorganic systems, in biological phenomena, and in human thought and action – are, from an A-T point of view, certainly real. And they are, of course, the sort that Paley and Co. tend to focus on (though even here their understanding of teleology is very different from the A-T conception). But I have focused on the most rudimentary sort of teleology – the sort manifest in even the simplest causal connections – because that is all that is required for an argument like Aquinas’s Fifth Way.