Suppose, to simplify the case by abstraction, that the fulfilment of a promise to A would produce 1,000 units of good for him, but that by doing some other act I could produce 1,001 units of good for B, to whom I have made no promise, the other consequences of the two acts being of equal value; should we really think it self-evident that it was our duty to do the second act and not the first? I think not. We should, I fancy, hold that only a much greater disparity of value between the total consequences would justify us in failing to discharge our prima facie duty to A. After all, a promise is a promise, and is not to be treated so lightly as the theory we are examining would imply. What, exactly, a promise is, is not so easy to determine, but we are surely agreed that it constitutes a serious moral limitation to our freedom of action. To produce the 1,001 units of good for B rather than fulfil our promise to A would be to take, not perhaps our duty as philanthropists too seriously, but certainly our duty as makers of promises too lightly.
(W. D. Ross, The Right and the Good [1930; repr., Indianapolis/Cambridge: Hackett Publishing Company, 1988], 34-5 [footnote omitted])
Note from KBJ: Let me put this quotation in context. Ross is a moderate deontologist. This means two things: first, that there are certain types of act, such as breaking a promise, that are intrinsically wrong (i.e., wrong in and of themselves, independently of their consequences); and second, that it is sometimes right, all things considered, to break a promise. When is it right to break a promise? When, by keeping it, one would lose a great deal of utility. Ross tells us here that losing one unit of utility is insufficient to justify breaking a promise. In other words, he has a high—but not infinitely high—threshold. Moderate deontologists differ among themselves by setting different thresholds. An absolute deontologist such as Immanuel Kant (1724-1804) agrees with Ross that there are certain types of act, such as breaking a promise, that are intrinsically wrong, but disagrees that it is sometimes right, all things considered, to break a promise. According to Kant, it is always wrong, all things considered, to break a promise. Do you see why Kant is called an absolutist? He allows no exceptions to his rules. He has an infinitely high threshold. Certain acts must not be performed, no matter how good the consequences of doing so, and certain acts must be performed, no matter how bad the consequences of doing so. For Kant, utility is irrelevant to rightness. For Ross, utility is relevant to rightness but not dispositive.