While arguing that God is not in a genus, St. Thomas raises the objection:
Whatever exists in itself is a substance (or, we call things substances which exist in themselves)
Substance is a genus.
God exists in himself.
The major premise appears to be simply what substance means, since the account follows the way substance is usually explained. The usual philosophy 101 approach to explaining what “a substance” is involves pointing out how accidents never exist by themselves (who hasn’t been told something like “you never see redness just walking down the street”?) and then defining substance in opposition to this.
Now in fact, Aristotle’s own account of substance was never this simple- substances can be either first or second substances, and second substances are universals. This provides at least an opening to argue that if something could never be a true universal that it could not be a true substance either. We can interpret STA’s arguments as making this point, since he explains that what he means by God is not something that admits the possibility of belonging to many, and therefore “God” cannot be a possible universal. That said, Aristotle never argues that what cannot be spoken of as a second substance cannot be a true first substance either, and it seems like an odd claim to make, as though real being (first substance) was contingent on intentional or logical being (second substance).
St. Thomas, however, resolves the objection by saying that substance cannot be what exists of itself, since this would be to claim that the genus of substance is simply being, while being is not a genus. The answer raises the problem of how being can first mean substance on this account, which Aristotle insists on, but it is more interesting to consider STA’s more precise account of the genus of substance as a quiddity which is apt to exist in another. On this account, the main difference between substance and accident is not whether one exists in another or not, but that there is a quiddity of substance but not of accident. This is a reference to Aristotle’s famous arguments about the “snub nose”, where he argues that, since “snubness” includes reference to a nose that if “snubness” existed of itself then we would have to speak of a “snub-nosed nose” (for clarity, Aristotle needed an accident that was said of only one subject). More problematically, the “snub” we just said would itself divide into “snubness of a nose”, giving us “snub-nosed nose nose” and so on ad infinitum. The problem with such a regress is that the putatively independent property never shows itself independently.
But how can we define substance as something apt to exist in another? Even if “existing in itself” was not a complete definition of “substance”, it seems strange to come to the point of defining it by the opposite property, that is, “being apt to exist in another” (presumably, “the other” in the case of substance is the subject or hypostasis.) A deeper problem arises from accounting for substance as a quiddity as opposed to a subject, since this certainly seems to make substance a part or principle of the substance-quiddity composite. But wouldn’t the composite be the substance here? It is by no means clear how, if we account for substance in this way, that we can say that a substance is given in experience. It is strange to insist that we coin a word like “substance” in order to speak of a part of some composite and not the existing composite itself given in experience.
As far as I know, only Rousselot deals with this problem, by saying that, given our mode of existence as subjects that do not exhaust what we are, we are only attuned to things so far as they are this sort of thing. For any natural subject to exhaust the whole of what he is would require that everything that was of the same nature as he collapse together into a single individual – a sort of angel-human or angel-dog – at once an individual and a complete exhaustion of the reality of the nature. As it stands, the natural world itself is a failure to be what it is.