- There are ordered series of efficient causes
- We perceive that there are effects
- Nothing can be the efficient cause of itself
- Therefore, the effects have efficient causes
- An ordered series of efficient causes cannot have an infinite regress
- In ordered series of causes, there are first, intermediate, and ultimate causes
- If a cause is removed, so is its effect
- Therefore, if there were no first cause, there would be no intermediate or ultimate causes
- If there were an infinite regress, there would be no first cause
- There are intermediate and ultimate causes
- Therefore there is a first cause
An efficient cause is an agent cause, that which produces an effect. If the effect is a building, the efficient cause is the builder. Parents are the efficient cause of their children.
We perceive by experience that there are indeed efficient causes in nature. Or, as C. F. J. Martin puts it, we perceive that there are “subjects of efficient causality” which necessitate that there are efficient causes. For instance, an artist sculpting a statue from bronze. We perceive the bronze being fashioned, and we know it must have an efficient cause. Why? Because nothing can be the efficient cause of itself. In order to be the efficient cause of itself, a thing would have to be “prior” to itself, as Saint Thomas says, which is impossible. In the example of the statue, we are seeking the cause of its coming into being. A cause must exist in order to exert causal activity. So the statue could not be its own cause, because then it would exist before it exists, which is absurd.
So we know that there are series of efficient causes. Now a series of causes will either terminate in a first cause, or else have an infinite regress. As we saw in our post on the first way, a per se causal series cannot have an infinite regress. In the second way, Saint Thomas gives a further defense of this: if a causal series does not have an infinite regress, and rather terminates in some first cause, then the series will have the following order: first cause, intermediate cause, ultimate cause. Imagine a man moving a rock on the ground with a stick. The rock is the “ultimate” cause (or the effect), the stick is the intermediate cause, and the man’s hand is the first cause.
If a cause is removed, Saint Thomas explains, its effect will cease as well. So if one were to remove the first cause, the intermediate causes would also be removed, and there would be no effect. But if there were an infinite regress, there would be no first cause; and if no first cause, no intermediate causes, and no effect. But there is an effect, so there cannot be an infinite regress. There must be a first cause.
In the first way, it was argued that only per se causal series cannot have infinite regresses. Granted that there are causal series of efficient causes, are there per se series? This is a good question, because it is not as immediatley obvious that there are instances of per se series of efficient causality in general as there are for series of motion specifically. A good example of efficient causality is a parent begetting a child; but as we mentioned in the post on the first way, this is a per accidens causal series, not per se.
A per se causal series is one in which the effect is here and now continuously dependent upon the activity of the cause. Think, for instance, of an artist sculpting a statue from bronze. This is why C. F. J Martin’s terminology of “subjects of efficient causality” is helpful. If we just consider “effects” of efficient causality in general, we’re likely to think of things like children having been begotten by parents, in which case it’s more difficult to think of there being per se series of efficient causality. But if we consider subjects of efficient causality, we see more clearly that there are indeed per se series. A piece of bronze that is here and now being fashioned into a statue here and now requires an efficient cause to be sculpting it. And that efficient cause — the artist — here and now depends upon numerous factors to sustain him in existence and in activity. For instance, he depends upon oxygen, gravity, etc. This is clearly a per se causal series.
Just because some causal series might require a first cause, why think that this means there must be a single, ultimate first cause of everything? Why think, in short, that this first cause is God? Efficient causality consists of the actualization of a potential. Think again of the artist scultping a statue from bronze. The bronze has the potential to be shaped into a statue before the artists begins to work on it; and as the artist works, he actualizes the potential of the bronze to be molded.
But as we saw in the post on the first way, if there is a per se causal series of actualization of potentials, the first in the series must be purely actual with no potentiality. Such a being of pure act will necessarily be one, and will necessarily have the following attributes: immateriality, eternality, timelessness, immutable, incorporal, and will be the ultime first cause of all series of efficient causality. These are the traditional attributes of God, and are sufficient to demonstrate the falsity of naturalism.
How, then, is the second way actually different from the first way? The first and second ways are remarkably similar, in both structure and content. The main difference is that the first way applies only to motion (which is a specific kind of efficient causality), while the second way applies to, and starts from, general efficient causality itself.
Thomas Aquinas. Summa Theologiae. I, Q. 2, Art. 3.
C. F. J. Martin. Thomas Aquinas: God and Explanations. Edinburgh University Press, June 30, 1997.