cannot be nonexistent. Whatever cannot be nonexistent is not contingent, and whatever is not contingent is necessary. The aggregate is therefore necessary, and whatever is necessary does not depend upon anything. Your statement that the aggregate depends upon the aggregate means precisely what someone means by saying that it is necessary in itself. For what is necessary in itself is not independent of itself; rather, it cannot do without itself. Since you have acknowledged that it is necessary in itself, your statement that it depends upon something is invalid. This is unlike the aggregate of contingent causes, which cannot be necessary in itself, because none of its members exists through itself. If nothing within the aggregate exists through itself, then the aggregate is all the more incapable of being necessary in itself.

The objection raised by this man is of the same kind as the objection raised by al-Āmidī—indeed, it is the very same objection, and perhaps one took it from the other. It is that the whole has its existence determined by its individual members, each of which has its existence determined by another, without end. Al-Āmidī answered it in one of his two books, but said in the other that he knew of no answer to it. He reported that some people maintained that the aggregate was necessary in itself on this account, and he regarded this objection as sophistical.

What they all intend is that the aggregate of effects extending without end depends upon nothing beyond its successive individual members. The falsity of this is self-evident once it is properly conceived. It has only appeared problematic to those who found it so because they failed to conceive it fully. For when someone says, “Causes