Suppose that we want to frame a conception of reasons that isn't relativized to the inclinations of particular agents. That is, we want to identify particular things that count as reasons for acting simpliciter and not merely as reasons for some agents rather than others, depending on their inclinations. One way to frame such a conception is to name some features that an action can have and to say that they count as reasons for someone whether or not he is (...) inclined to care about them. The problem with the resulting conception, as we have seen, is that it entails the normative judgment that one ought to be inclined to care about the specified features, on pain of irrationality, and this normative judgment requires justification. The advantage of internalism is that it avoids these normative commitments. It says that things count as reasons for someone only if he is inclined to care about them, and so it leaves the normative question of whether to care about them entirely open. Yet if we try to leave this question open, by defining things as reasons only for those inclined to care about them, we'll end up with a definition that's relativized to the inclinations of particular agents—won't we? Not necessarily. For suppose that all reasons for acting are features of a single kind, whose influence depends on a single inclination. And suppose that the inclination on which the influence of reasons depends is, not an inclination that distinguishes some agents from others, but rather an inclination that distinguishes agents from nonagents. In that case, to say that these features count as reasons only for those who are inclined to care about them will be to say that they count as reasons only for agents—which will be to say no less than that they are reasons for acting, period, since applying only to agents is already part of the concept of reasons for acting. The restriction on the application of reasons will drop away from our definition, since it restricts their application, not to some proper subset of agents, but rather to the set of all agents, which is simply the universe of application for reasons to act. (shrink)

Can there be good reasons for judging one set of probabilistic assertions more reliable than a second? There are many candidates for measuring "goodness" of probabilistic forecasts. Here, I focus on one such aspirant: calibration. Calibration requires an alignment of announced probabilities and observed relative frequency, e.g., 50 percent of forecasts made with the announced probability of.5 occur, 70 percent of forecasts made with probability.7 occur, etc. To summarize the conclusions: (i) Surveys designed to display calibration curves, from which a (...) recalibration is to be calculated, are useless without due consideration for the interconnections between questions (forecasts) in the survey. (ii) Subject to feedback, calibration in the long run is otiose. It gives no ground for validating one coherent opinion over another as each coherent forecaster is (almost) sure of his own long-run calibration. (iii) Calibration in the short run is an inducement to hedge forecasts. A calibration score, in the short run, is improper. It gives the forecaster reason to feign violation of total evidence by enticing him to use the more predictable frequencies in a larger finite reference class than that directly relevant. (shrink)

Argues for the humean theory of motivation on the grounds that rejecting it requires rejecting decision theory.