in something necessary without a cause, before it is shown that a totality of necessary entities, comprising causes and effects, is subject to the same principle as a totality of possible entities.”

I say:

When the term “possible” is explained as that which can either exist or not exist, this can apply only to something nonexistent. As Ibn Rushd mentioned, this is the philosophers’ terminology. Anything whose nonexistence is impossible cannot, in this sense, be capable of either existence or nonexistence. Rather, it is necessarily existent, whether it is said to be necessarily existent in itself or through another. If some such thing is necessarily existent through another, eternally, without beginning or end, then calling it possible requires proof.

Thus, dividing what exists into the necessary and the possible in this sense requires proof, as Ibn Rushd mentioned. This differs from a nonexistent thing that can exist: such a thing is known to be possible.

For this reason, Ibn Sīnā and those who followed his approach needed to establish that the celestial spheres are possible in themselves, not necessarily existent in themselves. True though this may be, their proof is exceedingly weak, since it rests on the claim that every body is possible, and their proof of that claim is very weak indeed, as has been explained elsewhere. Abū Hāmid explained its weakness, and Ibn Rushd, despite his efforts to refute him, agreed with him that this approach was weak.

The correct conclusion is that this may legitimately be called “possible” in ordinary language, in the sense that, in itself, it does not