A well-known doctrine of Aristotle’s is that of the unmoved mover, the first cause of the eternal motion of the universe. Aristotle’s key argument for the necessity of an unmoved mover is presented in Physics, Book VIII, chapters 4–5, where he argues, first, that everything that is in motion must be moved by something (ch. 4) and, second, that a thing cannot move itself (ch. 5). I have long been puzzled by this argument, and lately I have come to think it’s not just my ignorance—there really is something wrong with the argument. My aim here is to explain what it is.
To assess Aristotle’s argument for the unmoved mover, it is helpful to know how the unmoved mover fits into Aristotle’s cosmology. To review, Aristotle believes the universe is a finite sphere with the earth at the center. There is a fundamental division between the sublunary region and the superlunary region. Below the moon’s orbit, natural motion is linear, or, more exactly, radial in straight lines to and from the center. The four sublunary elements, earth, air, fire, and water, naturally move up and down along these lines. Earth moves naturally down until it reaches the center or is obstructed by something intervening (such as other chunks of earth). Fire moves naturally up to the extremity of the sublunary region unless something obstructs it. Earth and fire are thus “absolutely” heavy and light, respectively. Water and air, by contrast, are heavy and light “relatively”—to each other and to earth and fire. For example, water is heavier than air but lighter than earth. Left to themselves, the elements would form concentric spheres of earth, water, air, and fire. Their motions up and down are natural in that they result from a principle within themselves. In such motions, they realize their actuality (entelecheia) in a process like that of nutrition and healing, which operate naturally if nothing obstructs them (De Caelo, IV.3).
Beginning with the orbit of the moon, natural motion is circular, revolving around the center. Now, circular motion is one of the two simple motions. The other is rectilinear motion. But Aristotle thinks that circular motion is primary, since a circle is simple and complete, whereas the line of rectilinear motion is not. For, if the line is infinite, it is incomplete, having no endpoints, and if it is finite, it can always be extended. (Cf. De Caelo, I.2 and Physics, VIII.9.) But if rectilinear motion is simple and is the natural motion of simple bodies (earth, air, fire, water), and if circular motion is simple and prior to rectilinear motion, then circular motion must likewise be the natural motion of some corresponding simple body. But the simple, natural body of circular motion cannot be any of earth, air, fire, or water, because their natural motions are rectilinear, and a distinctive, internal principle of motion is the defining essence of any natural body. The simple body of the superlunary region therefore must naturally move in a circle. For various reasons, it also is ungenerated, indestructible, unalterable, and not subject to increase or diminution (De Caelo, I.3). People have given it the name aether, though Aristotle doesn’t seem to much care what it is called and rarely uses that word.
Thus, the superlunary region contains spherical bodies—the heavens—which are composed of aether, that rotate eternally. On these are mounted the observable heavenly bodies: the fixed stars, sun, moon, and planets. There are multiple spheres, whose rotation is semi-independent, in an arrangement that accounts for the semi-independent observable motions of sun, moon, etc. The outermost sphere, called the first heaven, carries the fixed stars, and it supplies the primary motion, all the other spheres being yoked to it. All of this is uncreated. It had no beginning and will have no end. Beyond the first heaven there is nothing. There are no other worlds. This is the universe.
Now, what moves the first heaven? This is the question that motivates the argument for the unmoved mover. The argument can be summarized as follows.
- Everything that is in motion is moved by something.
- A thing can’t be moved by itself.
- So, everything that is in motion is moved by some other thing. (from 1, 2)
- A thing that is a mover may itself be in motion or unmoved.
- A mover that is in motion is moved by some other thing. (from 3)
- On pain of endless regress, not every mover can be moved by some other thing.
- So, on pain of endless regress, not every mover is in motion. (from 5, 6)
- So, there is an unmoved mover. (from 7)
This argument seems perfectly intelligible—and, I am inclined to think, valid—given premises 1 and 2.
I also think that the case he makes for premise 2 is straightforward, whether or not it is true. Simplifying somewhat, the basic argument stems from the idea that motion is imparted by something that already possesses that motion. Thus, A heats B by being itself hot, and one teaches geometry by knowing it oneself. So, if a thing were to move itself, it would have to—at the same time and as a whole—both undergo and cause the same motion, which is impossible. Again, for Aristotle, motion is a progress from potentiality to actuality, and actuality is not reached until the potentiality is fulfilled. But the mover is in actuality already. What does the heating is already hot. So, continuing the heating example, what heats itself would have to be both hot and not hot at the same time and in the same respect. But this is impossible. The only way out of this would be to split the body in two, C and D, and say that C does the moving and D is moved. But now C must itself either be moved or unmoved. If C is moved, then the problem recurs for C. If C is unmoved, then D alone is moved and we can no longer say that a thing moves itself. So, nothing moves itself. (See Physics, VIII.5 257b1–258a26.)
My problem is with premise 1, which says that everything that is in motion is moved by something. And the problem is this: if the natural motion of aether is to rotate, why does it need a mover? The whole point of natural motion is that no external impulse drives it. It moves by itself because that is its activity. For example, fire goes up because that is its activity (De Caelo, IV.3 310b30–311a10). Fire is light, and lightness just is the principle of going up. Fire doesn’t need to be moved up. Likewise, earth is heavy, and heaviness just is the principle of going down. Earth doesn’t need to be moved down; it goes by itself. That is its activity. Now, if the natural motion of fire is to go up and the natural motion of earth is to go down and they don’t need to be moved up or down, and if the natural motion of aether is to go around in a circle, why can’t it just go around in a circle? Why does it need to be moved around in a circle?
Of course, in Physics, VIII.4, Aristotle provides a general argument for why, after all, everything does need a mover, including natural bodies such as earth, air, fire, and water. The reason, in the case of such natural bodies, is that their motion begins only when they are brought into being or, if they are already in being, then when whatever is obstructing their natural motion is removed. For example, when water evaporates into air, it rises of itself, but something else had to cause its evaporation, and that something is the mover.
So, too, with heavy and light: light is generated from heavy, e.g., air from water (for water is first such potentiality), and air is actually light, and will at once realize its proper activity unless something prevents it. The activity of lightness consists in the light thing being in a certain place, namely high up: when it is in the contrary place, it is being prevented. (255b8–12)
By “being prevented,” he means, for example, a wineskin full of air might be held under water by a stone. If something removes the stone, the wineskin at once rises. But although it rises of itself, it still has a mover—namely, whatever removed the stone (255b24–31). (See also De Caelo, IV.3 311a1–14.)
This is an admirably hardheaded doctrine. There are constraints on motion and hence on possible explanations of motion. And to be clear, by “motion” here Aristotle means practically any change; not only locomotion, but change of quality, change of quantity (increase and decrease), and generation and destruction. There is no spontaneous motion. All initiation of motion requires a mover. And transmission of motion is by contact.
Nevertheless, it is not quite right to conclude, as Aristotle does, that “all things that are in motion must be moved by something” (256b3). Rather, what he has shown is that all initiation of motion requires a mover. Once air comes into being or the wineskin is released, it simply moves with no further need of a mover.
Now, the thing about the circular motion of the aether is that it doesn’t need to be initiated. It has always been and will always be. So, why does it need a mover?