1 All page references in the main text of the article are to this book. An earlier version of this paper was read at an ‘Author Meets Critics’ Session at the 2006 Pacific Division Meeting of the APA. I would like to thank the audience for their questions and comments, and especially Joshua Gert for his insightful replies on that occasion. I would also like to thank Tom Hurka, Jonathan Peterson, Fred Schueler, and two anonymous referees for this Journal for comments on earlier drafts. Research on this paper was partly funded by a grant from the Social Sciences and Humanities Research Council of Canada.

2 I am assuming, of course, that there are no other reasons that would make rational for me to sing now.

3 See Brute Rationality, 76 and many other passages

4 See Brute Rationality, especially 37-8.

5 I find this comparison with morality unpersuasive. Although I can't go into this issue here, it might be worth quickly mentioning at least one problematic feature of this analogy. Even if one were to agree that ordinary morality has the requisite structure, moral reasons will only have this particular structure if it is true, as Gert supposes, that morality is concerned (almost) exclusively with our duties to others. But on this interpretation, morality by definition ignores any reasons that require the pursuit of our self-interest. So it is no surprise that our self-interest does not morally require that we act in any way. There is no similar feature of reasons in general that Gert could exploit in this context.

6 Cf. examples on 22-3.

7 I'll argue in more detail for this point later.

8 For different suggestions along these lines, see Thomas Hurka, Perfectionism (New York: Oxford University Press 1993), 87; Donald Regan,‘Value, Comparability, and Choice’ in Ruth Chang, ed., Incommensurability, Incomparabüity, and Practical Reason (Cambridge, MA: Harvard University Press 1997); and Ruth Chang's Introduction in Chang, Incommensurability. Note that Chang's ‘on a par’ relation can also account for (e₂) and (f₂).

9 I am not claiming great historical accuracy here.

10 ‘Virtually’ and ‘almost’ are important qualifiers here.

11 Since, arguably, money is just an instrumental good, such considerations strictly speaking cannot be basic reasons in Gert's view. However given that it is so easy to compare monetary amounts, I'll keep using these examples. But one could substitute gastronomic pleasures, for instance, as long as one made sure that one of the pleasures was clearly greater than the other, and each great enough to justify the actions in question.

12 Clause (b) is necessary in order to rule out causes of objective irrationality that are unconnected to the agent's will, such as ‘blindness and clumsiness.’ (Brute Rationality, 161).

13 In fairness to Gert, he does examine a somewhat similar approach, namely Raz's claim that there are (or might be) exclusionary permissions. However, I find Gert's criticism of Raz unpersuasive, particularly if applied to any of these views. Very briefly, here are my misgivings. Gert asks about the ‘ontological Status’ of exclusionary permissions or their ‘justification.’ I don't see why anyone should be more worried about the ontological Status of any of the materials in these proposals than about the ontological Status of two dif ferent kinds of strength that reasons may have. Moreover in terms of justification, as we pointed out above, there's no independent justification for the claim that the same reason has widely divergent requiring and justifying strengths; on the other hand, it is relatively easier to see the rationale for accepting that an agent-relative conception of rationality is as compelling as an agent-neutral one, or that there should be ways in which it is rational to restrict the Claims that others’ projects make on our own.

14 Gert raised this objection in ‘Author Meets Critics’ Session on his book in the 2006 APA Pacific Division Meeting.

15 Although I am connecting this rationale to a Kantian view, it is not an exclusively Kantian one. With the exception of the emphasis on policies, it is quite similar to the rationale given by Scheffler to agent-centered permissions in The Rejection of Consequentialism (New York: Oxford University Press 1994).

16 Although in the last case the combination of attitudes would turn out to be rational only if either (a) the change of heart is a relevant change in the circumstances of the agent or (b) the sense of ‘rationality’ in question is the subjective one.

17 See Brute Rationality, 14. Gert only says that one has to show that requirements of rationality to which the Kantian appeals are inescapable, not that they are inescapable independently of prior commitment to morality. But I take it he would consider it trivial that they are inescapable in light of one's commitment to morality. Gert also says simply that Kantian is committed to show that the requirements can be described in nonmoral terms. But given that Gert thinks that requirements of rationality are such that it does not make sense to ask about them why we should comply with them, I take it that he would think that this implies the requirement I describe here.

18 Or at least the later Kant. A particularly striking passage is the following footnote in the Religion: ‘The most rational being of the world might still need some incentives, Coming to him from the objects of the inclinations, in order to determine his power or choice. He might apply the most rational reflection to these objects … without thereby even suspecting the possibility of such a thing as the absolutely imperative moral law…. Were this law not given to us from within, no amount of subtle reasoning on our part would produce it or win our power of choice over to it.’ (Immanuel, Kant Religion Within the Boundaries of Mere Reason, Wood, Allen and Giovanni, George di eds. (New York: Cambridge University Press 1998),Google Scholar 27n.). It is more contentious whether the Groundwork is committed to such a project.

19 Notice that this cannot be explained in Gert's view by saying that some of these actions are objectively rational but subjectively irrational. After all, each of these actions should count for Gert as objectively rational.

20 Gert also appeals to the connection between irrationality and moral responsibility to bolster his case. But I must confess that I am not clear on the nature of the case here. (In f airness to Gert, he grants that the case is inconclusive. See Brute Rationality 82-3.) As Gert himself notes, not all cases of irrationality excuse or even attenuate one's responsibility. Many cases of irrationality are also cases in which the agent is not responsible, but there might be ‘common causes’ for the irrationality and the lack of responsibility (for instance, the agent is incapable of acting from the right reasons). Moreover actions done out hatred or revenge (and not for any kind of pleasure), or actions in which one puts others at risk by reckless endangering one's own life (reckless driving, or smoking heavily around one's children) could all (plausibly) come out irrational in Gert's account. And yet the agent bears the full responsibility of the härm that she may cause the others. I don't see how Gert can explain the attribution of responsibility in these cases if the connection between irrationality and moral responsibility is as tight as he thinks it is. Again, it is important to note that none of what I said in the defense of a wider conception of rational requirements depends on the details of the Kantian strategy; in fact, it is available to anyone who holds that moral reasons are (requiring) reasons sans phrase.