If it is said that one of the two parts depends upon the other, or that the whole depends upon each part, and so forth,

the reply is: First, this is not your argument, for you have only claimed dependence upon the other—that one of them produces the other or is its efficient cause. This is necessarily false, since, in contingent composites, no part is the efficient cause of another, nor does it produce another by its own choice. Even if we supposed that some composites had a part that produced another part, this would not hold of every composite. The proposition would therefore not be universal, and the case under dispute would not necessarily fall among those in which one part depends upon another. How much less so if no efficient cause of the other part is to be found among its parts?

If you mean that one of the two parts exists only together with the other, this amounts merely to their mutual entailment and to one being conditional upon the other. This is simultaneous, concomitant circular dependence, which is possible and valid, and necessarily obtains between any two mutually entailing things. This does not contradict the whole’s being necessary by virtue of the whole.

If it is asked of each part whether it is necessary in itself or not,

the reply is: If you mean whether it is something produced, an effect of an efficient cause, then none of the parts is such. Rather, each is necessary in itself in this respect. If you mean: Is there among them—among the parts—anything of that kind? Rather, each is necessary in itself in this respect. And if you mean whether any of them exists without the other existing