Is Teleology Conceptually Impossible? And If Not, When Should We Resort to It?

We all know—don’t we?—that modern physics is inconsistent with teleology, at least with real teleology, as opposed to the substitute version that so many of us have tried to fashion for use in “teleosemantics.” But what is it, exactly, that modern physics bans? What form would a teleological theory take? When pressed, most of us would be hard pressed to come up with an answer. (Do pause and think about it, Dear Reader. I can almost guarantee that if you are not familiar with the material I am about to discuss, you’ll get it wrong. I certainly did, and I am supposed to be an expert! Ha!)

This is the target of a brilliant 2006 analysis, “What Would Teleological Causation Be?” by John Hawthorne (Oxford) and Daniel Nolan (ANU). I’ll summarize the paper here, and reflect some more on it in a follow-up post.

(The paper was published in Hawthorne’s 2009 Essays in Metaphysics. I only heard of it recently, when it was cited by Tom Nagel in his recent book and mentioned in Eric’s insightful post on the Principle of Sufficient Reason.) 

It is often said that teleology is impossible because it involves backward causation—Aristotle says that the acorn sprouts “for the sake of” the mature oak, but (so the criticism goes) this cannot be a causal relation because the mature oak cannot make the acorn sprout. (Spinoza: "That which is really a cause, it considers an effect.") H&N demur. Teleology and backwards causation are different things, they show. (This, by itself, is one of the surprising and most valuable parts of the analysis.)

Suppose we observe the following strange occurrence, which I’ll call BAM. . .

. . . An asteroid in free motion through space is suddenly impacted by several smaller bodies simultaneously; these smaller bodies are absorbed by the asteroid with no cratering or shattering—the projectiles hit and immediately disappear within the larger mass of the asteroid. (I have simultaneously simplified and embroidered H&N’s example.)

Taken in the usual way, BAM is compatible with the laws of mechanics, but it would be an extraordinary coincidence to have several bodies collide at the same time, and in this unusual way (no shattering, no cratering, which requires all kinds of special conditions). But it would not be a coincidence if what had happened in fact was that the asteroid had simultaneously emitted several particles !!moving backward in time!!. Prima facie, then, BAM gives us some reason to posit backward causation. Suppose that we discover that BAM-type events occur often, and always in asteroids made out of material PQR, configured as in C. Then the evidence becomes compelling.

That’s backward causation. But it’s not teleology. There are two differences. First, backward causation is continuous: if point c lies on a retro-temporal path between the asteroid and earlier points on the intercepting projectile’s path, then c was backwards caused by the emission event BAM. However, teleology is not continuous: the acorn sprouts for the sake of the mature oak but not for the temporally intermediate sapling.

Second, there’s the matter of intervening causes. (Here again, I am elaborating.) Any causal process can be interrupted. Interrupting teleology means that the end will not be achieved. Chop down the sapling and there will be no mature oak: notwithstanding this, the acorn was for the sake of the oak. Interrupt backward causation and certain earlier points in the retro-temporal trajectory will vanish. If a backward moving particle and anti-particle (originating from two PQR asteroids) were to intersect with and annihilate each other, it would look to us temporally forward moving folks that they had been spontaneously created and sent on their way to the said asteroids.

What about teleology? H&N have a fascinating discussion of the form that teleological laws of change must take. They say that what is needed in order to formulate such laws is a notion of “end-distance”. End-distance is not necessarily spatial distance; rather, it is a measure of how close something is to its teleological end, i.e., that for the sake of which it is. (The sapling is closer to the end of being an oak than the acorn is.)

A teleological law says what kind of path an end-directed thing will traverse, and what sorts of constraints will determine its motion. In Aristotelian physics, for example, a ball released above a bucket will trace a path that is for the sake of coming to rest at the centre of the Earth. (Here, by the way, end-distance is spatial distance to the centre.) The ball gets closer and closer to its end, but it does not scoot around the bucket. Rather, it comes to an abrupt halt when it hits the bucket. On the other hand, released on the side of a hill, it’ll roll down gullies and around ridges, and get closer and closer to the centre until it comes to a low point. The teleological law of motion has to be formulated in such a way that it predicts these trajectories.

When would one resort to positing teleological process? One suspicion is that in modern mechanics, any teleological law of motion may turn out to be just a restatement of a straight efficient law of motion. In fact, this is, perhaps, one way that modern mechanics edges teleology out. For example, the ball might get to the ditch at the bottom of the hill along a path dictated by a reformulation of Hamilton’s Law of Least Action, and come to a halt when it has to roll too far up-hill. Or something of that sort. If this is so, then teleology would be otiose in mechanics.

(Some, including H&N, hint that the Hamilton's Principle of Least Action might be teleological. I don't see how: It's continuous in the way that backward causation is. It's completely derivable from Newonian mechanics.And the end doesn't play a role: rather, where a particle ends up is at the end of the path that has Least Action . . . am I wrong about this??! Incidentally, Hamilton's Law is puzzling for another reason: it quantifies over all possible paths, including physically impossible ones. Jeremy Butterfield discusses.)

When might teleology not be otiose?  That is, under what sorts of circumstances might it be indicated? H&N describe a case, and I’ll discuss it in the second part of this post. But (taking a leaf out of biological teleophiles such as Tim Lewens of Cambridge), canalization may be a crucial indicator. Canalization is the phenomenon (described by C. H. Waddington) where an entity will take alternative paths to its prescribed end-state when its normal path is blocked. Waddington described this in embryology: if an embryo’s normal developmental pathway is blocked, it will go down another one, leading to the same end. Suppose the ball released over the bucket did scoot around it to find another pathway to the centre. This would be a powerful indicator that teleology was involved. 

OK. So is there anything in the real world that indicates teleology? Thomas Nagel thinks so, and Eric suggests that he is using (or misusing) the Principle of Sufficient Reason. Back to this in a later post. This one is for the sake of that one.