b. Proof of the Second Conclusion

36. The proof of the second conclusion [n.29] is that a respect is essentially a relation between two extremes, therefore, just as to take away the term that the respect is to is to take away or destroy the idea of respect, so to take away that which the respect is of is to take away the respect itself and to destroy the idea of respect. Not therefore because a relational accident is an accident does it for this reason require a subject or foundation; but because a respect is a respect for this reason does it require that of which and to which it is a respect (even in the case of divine reality).

37. If it be argued against this [second] conclusion [n.29] that it seems to contradict the preceding one [n.28], because, according to the preceding conclusion, the idea that ‘to be accident to’ or to ‘inhere’ signifies is a respect coming from outside [n.32], therefore it cannot belong to any respect denominatively, because a respect is not a subject or a foundation of a respect (for then there would be a regress to infinity); therefore ‘paternity’ or ‘likeness’ cannot inhere in or be accident to, because ‘to inhere in’ or ‘to be denominated by inhering in’ requires that the thing denominated be the subject or foundation of the inhering in. I reply that one respect can well be founded on another, as posterior on prior (just as was touched on above in a question about character, and was proved about proportion and proportionality [Ord. IV d.6 nn.300-305]). And so I say that just as whiteness can be denominated from inhering in, or from what can be accidental to, so can paternity be denominated; and ‘like’ is a denomination in this regard.

38. But if you argue that therefore a respect is denominated by itself - the solution to this will be given in the third article [nn.39-45, cf. nn.67-68].