In any system that has a goal, any part or character that contributes to reaching that goal, however mediated this contribution might be, can be said to have a function as Nagel understands the term. And because a particular item may have more than one effect in a system and thus make more than one contribution to its goal, it is often better to speak of a function of the item rather than the function. Nagel recognizes “that every effect of an item will have to count as one of its functions, if it should turn out that each effect contributes to the maintenance of some goal or other.” This position has had numerous adherents. Nagel is quite generous in attributing functions to things; he doesn't demand the existence of some kind of feedback loop from the function Y to the function bearer X. But he does place certain restrictions on the effects that can be taken as functions insofar as he appeals to the characteristic activity of the system. He is not interested in why X is there but in what it does when it is there. We shall see that the further development of the dispositional view by Cummins to a “capacity” interpretation makes it somewhat more generous in attributing functions; and the more radical propensity view proposed by Bigelow and Pargetter and others turns out to be genuinely profligate in ascribing functions.