In Physics Books VII and VIII, Aristotle lays out a number of arguments for the existence of a “prime” or unmoved mover. The Prime Mover is ultimately responsible for all change (motion) and so for the forward progress of time itself (given that time is the measure of change).
Aristotle’s argument build on Plato’s early argument for prime movers in Book X of The Laws. Plato argued that only souls (the principle of life within organism) can be self-movers, and that any chain of causation of motion must terminate in such self-movers. Hence, he argued, souls or live must be “older” than all non-living things.
Aristotle’s arguments deviate from Plato’s in two important respects. First, Aristotle argues that, strictly speaking, there are no self-movers. What we call (loosely) self-movers (like living organisms) are actually complex entities, containing some parts that move other parts. In some cases, the soul itself moves some part of the body, which moves other parts, eventually resulting in the apparently “self-moving” behavior of the whole organism. But nothing in the organism is truly self-moving. Even the soul must be moved (changed) by something external to it, such as some change in the animal’s nervous system.
Second, Aristotle rejects the idea that we can use the argument from motion to establish that all motion had a beginning in time. Aristotle thought that the universe was in a “steady state,” an eternally cyclical, repeating pattern, without beginning or end. He nonetheless accepted that there cannot be an infinite regress of movers. He argued that any motion requires an ultimate mover that is not itself moved. These two aspects of his arguments seem (prima facie) to be in tension or even in contradiction with each other. If the past is infinitely long, won’t there be actual infinite regresses of motion-causation?
Medieval philosophers, such as Avicenna (ibn Sina) and St. Thomas resolved this contradiction by distinguishing between two kinds of infinite regress: those that are per se infinite and those that are per accidens infinite. They were not saying that there were two kinds of causation (per se and per accidens) of such a kind that per accidens causation permitted infinite regresses and per se causation did not. (That way of resolving the contradiction was first introduced by Duns Scotus, but it has been adopted by nearly all contemporary Thomists). There is, in this context, just one kind of causation of motion in view. But some regresses are essentially infinite while others are only accidentally so.
What would make an infinite regress only accidentally infinite? Many Thomists (again following Duns Scotus) suppose that, for a regress to be accidentally infinite, it is both necessary and sufficient that the regress involve an infinite series of times. Accidental infinite regresses take up an infinite number of instants (each successively earlier), while essentially infinite regresses are simultaneous. I think this is entirely wrong: an infinite succession of times is neither necessary nor sufficient for the regress to be per accidens infinite.
When Thomas explains what he means by the per se/per accidens distinction (Summa Theologiae I Q45, a2 ad 7 and De Veritate Q2, a10), he focuses exclusively on whether the causal structure of the series is infinitely complex or variegated. The paradigm of an accidentally infinite series is one in which a hypothetical blacksmith uses an infinite number of hammers (wearing at one after another) to make a horseshoe. In this case, the causal structure is essentially finite: blacksmith-hammer-shoe. A per se infinite series would be one in which each cause is merely an instrument of its predecessor. This would mean climbing to progressively higher and higher levels of instrumental causation without ever reaching a principal cause (e.g., something analogous to the blacksmith in the accidentally infinite case).
We could have a per accidens series in which all of the causes are simultaneous, as long as the essential structure of the series is finite, with a principal agent. Conversely, it’s far from obvious that we couldn’t have an essentially infinite series in which the causes stretch out infinitely far into the past.
Given the fact that Aristotle considers the past to be infinitely long, he must be able to treat all temporally successive series as irrelevant to the explanation of present motion, whether the series is per se infinite or not. Indeed, in Physics Book VIII, 4, Aristotle explicitly states that the mover and moved must be simultaneous. Many interpreters assume that both Aristotle and St. Thomas are relying here on an outdated theory of mechanics, one lacking Newton’s concept of inertial motion.
In fact, this is wrong. Aristotle’s physics did include principles of inertia or impetus, and Thomas clearly recognized this. There are three cases of locomotion to consider: the eternal rotation of the heavenly spheres, the natural motion of heavy bodies (downward) and light bodies (upward), and the violent motion of projectiles (e.g., an archer shooting an arrow).
Each of the heavenly spheres rotates in the way that it does as a result of its own internal nature. No sphere is pushed or pulled continuously by any other body. It is “moved” by its own celestial intelligence, but this intelligence and its inclinations are part of the very essence of the sphere itself. Thus, each sphere rotates as a result of its own internal impetus.
Heavy bodies move toward the center of the earth because that is their nature. Similarly, light bodies move upward by their intrinsic nature. These motions are completely unconnected mechanically with the rotation of the spheres. If the spheres do anything to the sub-lunar world, it is a matter of their stirring up some turbulence in the upper atmosphere. But this turbulence has nothing to do with the natural motions of heavy and light bodies, which Aristotle describes as natural and not by “force” (bia).
What about the arrow? Here commentators almost universally take Aristotle to believe that his account of projectile motion has no place for inertia or impetus. It is true that Aristotle believed that the arrow must be continuously acted upon by the surrounding air in order to move upward. This is because the upward movement of the heavy arrow is “unnatural” or “violent”. But what about the movement of the air itself? It is obvious that the columns of air are not continuously moved by the archer or the bow. As Aristotle himself recognized (in Physics VIII, 10), if this were so, then the columns of air would stop moving the instant the bow stopped moving, and, consequently, the arrow would stop moving then as well. The bow must transmit an impetus to the air, so that it keeps moving after the bowstring stops. Each segment of the air changes its location inertially and carries with it the power to convey the same kind of impetus to the next segment of the air column. Thanks to friction, this impetus gradually declines, until we reach a segment of air that lacks the power to convey the power of moving either more air or the arrow. Once the arrow passes beyond the scope of that last powerful segment, the arrow stops moving upward, and its natural motion downward takes over. Of course, Aristotle must have known (by observation) that the arrow continues moving horizontally (in motion that is neither natural nor anti-natural) by inertia.
The motion of the projectile is explained entirely in terms of the local interaction between the bow, the air, and the arrow. There is no mechanical connection between the rotation of the spheres and the flight of the arrow. Why, then, did Aristotle think that all motion in the universe depends on the simultaneous rotation of the heavenly spheres?
The connection between the spheres and sublunar motion is not mechanical but rather metaphysical in nature. The solution the mystery is this: the Prime Mover must keep the heavenly spheres in continuous rotation in order to explain the forward motion of time itself.
In Physics Book VIII, chapters 1 and 6, Aristotle links the perpetuity of time with the perpetuity of motion. This makes sense, since he had defined time to be the measure of motion in Physics IV chapters 11and 12 (219b4–5, 220a4–5, 220a25–27, 221a1–9). Aristotle was very careful to avoid a definitional circle, and so he avoided defining motion or change in terms of time. That is, Aristotle explicitly rejects the At-At theory of change, as proposed by Bertrand Russell. According to the At-At theory of change, a substance undergoes change when it has some quality at one time that it lacks at another time. Instead, Aristotle defines change in Physics III (chapter 1) entirely in terms of actuality and potentiality: change is the actuality of the potential qua potential (201a10–16).
In Physics Book VIII, Aristotle draws the obvious conclusion from these definitions: the passage of time depends on the actuality of motion, not vice versa. In Chapter 1, Aristotle refers to his earlier definition of motion (251a9-10). He then mentions his definition of time as the measure of motion, drawing the obvious conclusion that it is impossible for a period of time to pass during which there is no motion (251b11–16, 27–29).
In Book VIII, Chapter 6, Aristotle extends this line of reasoning to draw the conclusion that there must be an eternal, immobile First Mover. He first emphasizes (258b27–259a2) that motion must be both eternal (beginningless and endless) and continuous (constant, uninterrupted, and unbroken—συνεχουσ).
“We have shown that eternal motion exists of necessity. And for such movement to be eternal it must be continuous (συνεχη), for what constantly exists at all times is as such continuous, whereas the successive (εϕεξησ) is discontinuous (and therefore not eternal). But for the movement to be continuous, it must be unified…” (259a16–19)
He also argues explicitly that no temporal agent and no aggregation of temporal agents has the causal power to account for the continuity of time:
“Neither (as is obvious) can any one of such (non-eternal and potentially intermittent agents) be the cause of the everlasting and uninterrupted process, nor can the whole sum of them; for that the process should be everlasting and uninterrupted is an eternal necessity, whereas the whole sum runs back without limit (so that we never come to a prime cause at all) and is not coexistent but successive.” (258b29–259a2)[1] [We might want to produce our own translation of these crucial lines]
A little later in the same chapter, Aristotle moves explicitly from the continuity of motion to the existence of an atemporal First Mover:
“So that if there must needs be a continuous movement, there must be a primary mover which is not even incidentally moved, if, as we have said, there is to be amongst things that exist a certain unceasing and deathless movement, and the universe is to abide self-contained and constant; for only if the principle abides, the sum of things, being in continuous relation with it, must also abide.” (259b23–28)
As Aristotle concludes the chapter, he draws the conclusion that only a timeless, essentially immovable cause can provide a causal explanation for the continuity of the passage of time (as we argued above):
“The immovable, as we have said, being simple and unchanging and self-constant, causes an unbroken and simple motion.” (260a18–20)
[1] Loeb translation by Philip H. Wicksteed and Francis M. Cornford, Physics V-VIII, Harvard University Press, 1934, pp. 343-5.
The Analytic Thomist 