Discourses: Book 1 7 : 3

But of course there are times when we have with sound reasoning granted the premisses, and the inference from them is so-and-so; and, in spite of its being false, it is none the less the inference. What, then, should I do? Accept the fallacy? And how is that possible? Well, should I say, "It was not sound reasoning for me to grant the premisses"? Nay, but this is not permissible either. Or, "This does not follow from what has been granted"? But that is not permissible, either. What, then, must be done in these circumstances? Is it not this, that the fact of having borrowed is not enough to prove that one is still in debt, but we must add the circumstance that one abides by the loan—that is, has not paid it—and just so our having once granted the premisses is not enough to compel us to accept the inference, but we must abide by our acceptance of the premisses? And what is more, if the premisses remain until the end what they were when they were granted, there is every necessity for us to abide by our acceptance of them, and to allow the conclusion that has been drawn from them; . . . for from our point of view and to our way of thinking this inference does not now result from the premisses, since we have withdrawn from our previous assent to the premisses. It is necessary, therefore, to enquire into premisses of this kind and into such change and equivocal modification of them, whereby, at the very moment the question is put, or the answer made, or the deduction drawn, or at some other similar stage in the argument, the premisses take on modified meanings and give occasion to the unthinking to be disconcerted, if they do not see what follows in consequence. Why is it necessary? In order that in this matter we may not behave unsuitably, nor at haphazard, nor confusedly.