In his clear, gracious, and helpful critical notice of Brute Rationality, Sergio Tenenbaum raises a number of worries about the way in which the central distinction in that book — the distinction between the requiring and justifying roles of practical reasons — is brought in to explain some commonplace normative data.² He also suggests two alternate ways of accounting for the same data, one of which derives from Sidgwick and has recently received its most sophisticated presentation in Derek Parfit's forthcoming Climbing the Mountain, and the other of which is inspired by Kant's distinction between perfect and imperfect duties. In this response to Tenenbaum I would like to do three things. First, of course, I would like to defend my own account. The cases Tenenbaum uses to argue against my use of the justifying/requiring distinction all have a common characteristic; roughly put, they are all ones in which it seems that some sort of maximizing principle applies, but is beyond the resources of my account. The same criticism has been raised in a number of other reviews.³ I would like to take this opportunity to make clear a general strategy that is available for dealing with all such cases. Next, I would like to show that both the Sidgwickian and Kantian alternatives [End Page 463] either yield wrong conclusions or change the subject. And finally, I would like to show that both the Sidgwickian and Kantian alternatives in fact incorporate as robust a justifying/requiring distinction as I would wish to argue for.

I The Justifying/Requiring Distinction

Anyone reading this response to Tenenbaum's 'Brute Requirements' will already be familiar with the justifying/requiring distinction. But to avoid some possible confusions, and to facilitate later points, it may be worth making some quick remarks about that distinction. First, the distinction is not between two kinds of reasons, but between two roles that the very same reason might play. The distinctive thesis of Brute Rationality is best put as a thesis about the strengths of reasons in these roles: the thesis that many reasons have a great deal more justifying strength than requiring strength. Of course, since most views implicitly identify justifying and requiring strength, they have no need to distinguish the two roles. Consequently, it was important to distinguish and clarify them. But, despite a general lack of attention to the distinction, it is almost completely uncontroversial that reasons play the two roles, as I define them.⁴ For to play the requiring role is simply to do the following: to make it rationally required to perform some action that otherwise would not be rationally required. And to play the justifying role is simply to do the following: to make it rationally permissible to perform some action that otherwise would not be rationally permissible. Consider the reason 'that the action is necessary if I am to avoid significant pain.' This reason clearly can play the requiring role, since I can be required to perform some actions to which this reason applies, when, had the reason not applied, I would not have been required. Now consider the reason 'that the action will save a stranger from significant pain.' This reason clearly can play the justifying role, since it can be rationally permissible for me to perform some actions to which this reason applies, when, had the reason not applied, it would not have been rationally permissible. In light of these examples, it ought not be controversial that reasons can play both a justifying and a requiring role. Controversy enters at two subsequent points. First, I claim that (up to a certain degree of vagueness) we can assign any given basic reason degrees of strength in each of these roles: a justifying strength and a requiring strength. This should [End Page 464] not be overly controversial, since it amounts to little more than a denial of the controversial thesis of particularism about practical reasons. The...