To the first one proceeds thus. It seems that one habitude is constituted from several habitudes. For that whose generation is completed all at once, but successively, seems to be constituted from several parts. But generation of a habitude is not all at once, but successively from several acts, as was said above. Therefore one habitude is constituted from several habitudes.
Further, a whole is constituted from parts. But to one habitude is assigned many parts, as when Tully posits many parts of fortitude, temperance, and other virtues. Therefore one habitude is constituted from several.
Further, from one conclusion alone is able to be had a kind of knowledge both actually and habitually. But many conclusions pertain to one whole kind of knowledge, as with arithmetic or geometry. Therefore one habitude is constituted from many.
But contrariwise, habitude, because it is a sort of quality, is a simple form. But nothing simple is constituted from several. Therefore one habitude is not constituted from several habitudes.
I reply that it must be said that a habitude ordered to working, which we are now principally intending, is a sort of completion of power. Now every completion is proportionate to its completable. Thus just as power, because it is one, extends itself to many according as they converge on some one thing, that is, in a sort of generic notion of object, so also habitude extends itself to many according as they have ordering to some one thing, such as to one specific notion of object, or one nature, or one principle, as is obvious from what has been said above. If therefore we consider habitude inasmuch as it extends itself to such things, we shall then find in it a sort of multiplicity. But because that multiplicity is ordered to some one thing, to which the habitude is principally related, what follows is that habitude is a simple quality, not constituted from many habitudes, even if it extends itself to many things. For one habitude does not extend itself to many things, save in ordering to one thing, from which it has unity.
To the first it therefore must be said that succession in the generation of a habitude does not happen from the fact that part is generated after part, but from the fact that the subject does not directly acquire firm and hard-to-move disposition and from the fact that it first begins to be incompletely in the subject and is bit-by-bit completed -- as is also the case with other qualities.
To the second it must be said that parts that are assigned to each single cardinal virtue are not integral parts, from which the whole is constituted but subjective or potential parts, as will be obvious below.
To the third it must be said that he who in some kind of knowledge acquires by demonstration knowledge of one conclusion indeed has the habitude by incompletely. But when he acquires by some demonstration knowledge of some other conclusion, another habitude is not generated in him, but the habitude that was previously in him becomes more complete, since it extends itself to several things; because conclusions and demonstrations of one kind of knowledge are ordered, and one is derived from another.
[Thomas Aquinas, ST 2-1.54.4, my translation. The Latin is here, the Dominican Fathers translation is here.]
It's very tempting to think of habitudes as being the sort of thing that could 'congeal' together to form new habitudes; St. Thomas shuts this down here, on the basis that this is not how qualities in general work. Habitudes grow by intension, not aggregation; they decline by remission, not by subtraction.
This article, of course, establishes a key principle of St. Thomas's influential taxonomy of virtues in terms of 'parts'. A potential puzzle with regard to the reply to the second objection is that St. Thomas does in fact assign to cardinal virtues not just subjective parts (i.e., specific versions of a virtue) and potential virtues (i.e., associated subordinate virtues directed to secondary matters), but "quasi-integral parts", which he sometimes just calls integral parts. These are not literally integral parts (i.e., parts in our ordinary sense of components making up a whole), but things that are associated with a virtue in the sense that the principal virtue needs them in order to exercise its act fully.
The reply to the third objection is a useful reminder that extension of knowledge works primarily by intension or intensification, not by addition or aggregation, as one might think if one looked only at the accumulation of conclusions that one can list as known. To take a set of principles and recognize more conclusions as proven by them is to know the principles and their consequence more intensively. To know is an act of the intellect as a power to know; knowing more is the intellect literally knowing more powerfully. This has a number of incidental ramifications worth thinking about -- e.g., knowledge cannot be assumed to be adequately characterized by a list of propositions known, and one may know the same thing to different degrees at times, in the sense that how powerfully or 'securely' it falls within one's knowledge, how 'central' it is to the things one knows, can vary as one's power to know that general kind of thing is built up by proofs. In this sense, we should think of knowing as like a light that, when more intense, reveals more.
