Allan Gibbard's Reconciling Our Aims, based on his 2006 Tanner Lectures delivered at Berkeley, is a work of impressive scope for a slim volume. Gibbard's own contribution to the volume runs to a total of 120 pages – three lectures and a response – and is supplemented by an introduction from Barry Stroud and commentary from Michael Bratman, John Broome and Frances Kamm.

Gibbard engages in two reconciling projects – the first in metaethics, the second in normative ethical theory. The first, developed here primarily in Lecture I, will be familiar to readers of Gibbard's previous work, most notably his Wise Choices, Apt Feelings (1990) and Thinking How to Live (2003). The aim is to develop an understanding of our moral (and more generally normative) practices, which at once retains the naturalism that has given non-cognitivist views much of their appeal, but which also does justice to the apparent truth aptitude of moral judgements and the seriousness with which we invest these judgements. His main focus in the lecture is to develop his conception of normative judgement as the expression of a planning-type attitude, while giving an account of the role of intuition in such judgement.

In Lectures II and III, Gibbard embarks upon a second reconciling project, developing a normative ethical framework which aims to encompass the attractions of both contractualism and utilitarianism. One's planning how to live is not something which takes place in isolation; it requires coordination with the plans of others. Gibbard's main aim in these two lectures is to outline the shape that communal planning should take, given people's initially divergent aims and values. (One might already question both the possibility and the desirability of everyone sharing a set of common goals, and doubt whether Gibbard's citing of the need to avoid prisoners’ dilemmas is enough to motivate this.) These divergent aims are best reconciled within a communal plan to promote the overall utility as much as possible. Gibbard seeks to establish this act-utilitarianism using a veil of ignorance argument. Conditions of fair agreement for a communal set of principles to live by are those in which we are ignorant of our own identity and assign an equal probability to the possibility of being any specific individual. Following Harsanyi, Gibbard argues that in such conditions a rational agent will select principles that best promote average utility.

In many volumes that take the form of lectures–comment–replies, the main messages, commitments and motivations of the lecturer crystallize in the response to objections. This book is no exception: all three commentaries, and Gibbard's clarifications, defences and concessions, are invaluable in helping the reader grasp the import of the original lectures.

Bratman focuses on Lecture I and raises several tricky questions for Gibbard's planning theory of normative judgement. What should we say about cases where one judges that one ought to do one thing, yet plans to do something else? Why must plans be ‘subject to a constraint of agglomeration and consistency’ (p. 96) in a way that beliefs are, but desires are not? And can we coherently regard normative disagreement on Gibbard's model as genuine disagreement, rather than as mere differences in our contingency plans about what to do in some situation?

The pugnacious exchanges between Kamm and Gibbard may tread familiar ground in the debate between consequentialists and non-consequentialists, but they are no less valuable for it. Kamm's critique draws from Gibbard some pithy articulation of core utilitarian responses to standard objections, which bear extensive thought and discussion, in particular his recommended treatments of respect and of the legitimate imposition of severe burdens (pp. 154–5). Kamm calls into serious question the extent to which Gibbard's contractualism-cum-utilitarianism retains the core attractions of Scanlon's contractualism.

The most rewarding discussion, however, is between Gibbard and Broome. Before going on to analyse Gibbard's deployment of Harsanyi's welfare theorems, Broome tackles Gibbard's central metaethical picture, questioning Gibbard's close identification of normative judgements with planning attitudes with a striking bluntness:

Part of Broome's worry is that even if we can identify normative judgements as some subset of our planning attitudes, it is unclear how an attitude's property of being a non-cognitive, planning-type attitude can explain its having the property of being a cognitive, belief-like attitude, as Gibbard insists it can. Broome suggests, convincingly to my mind, that we can only interpret someone as having the relevant sort of planning attitude if we have already conceived of her as having the normative belief that something ought to be done.

As both Broome and Bratman press, there certainly appear to be many cases where one judges that one ought to do one thing while one plans to do something else entirely. In particular, for an agent who accepts the act-utilitarian view which Gibbard goes on to argue for in Lectures II and III, this seeming divergence between ought judgement and plan is likely to be pervasive. The act-utilitarian claims that what there is strongest reason to do is bring about the most utility, impartially considered. But the practical deliberation of an agent who subscribes to this claim will surely rarely take the form of working out what there is strongest overall reason to do. Instead, she will typically lay aside those possible actions which produce the most utility, impartially considered, and deliberate amongst options which are less demanding on her. This is because plausibly, given familiar facts about the preventable suffering which exists in the actual world, the possible course of action with the best expectable consequences much of the time is to embark on a life of very significant self-sacrifice, devoting the vast majority of one's spare time, money, energies, attention and resources to relieving the plight of the some of the world's poor, whether by raising money for effective charities or campaigning for fairer rules of international trade.

As a Gibbard-style act-utilitarian, I may sincerely judge that a life of extreme devotion to these causes is what there is the most to be said for doing; in adopting instead a lifestyle of fairly moderate contribution to these causes, I may judge that those few others who adopt the more self-sacrificial lifestyle are living lives for which there is even more to be said. Seemingly quite distinct from my normative judgement that this lifestyle of extreme devotion is the one I have strongest reason to adopt is my actual practical deliberation about what to do, my planning what to do. When I go about this latter deliberation, I set aside my judgements about what there is strongest reason to do, since I am unwilling to live a life which involves the sacrifice of so many of the projects and relationships I care most about. It may be that Gibbard can still construe my original normative judgement as just a different exercise of planning from the one I more obviously go on to engage in when actually deliberating about what to do, after rejecting the very demanding options recommended by my act-consequentialism. However, much more needs to be said here.

All three commentators question this identification of normative judgements with planning attitudes. From where does the pressure come to identify normative judgement with something from which it is putatively distinct? From Gibbard's perspective, it comes from a demand that we reconcile our understanding of the normative with a naturalistic picture of the world. Our everyday practices of planning do not appear vulnerable to the queerness worries which Mackie raises about moral judgements, and so the planning theory of normative judgement seems like an appealing refuge against his error theory.

Nonetheless, once one concedes, as Gibbard does, that normative judgements can be straightforwardly true, without being so in virtue of their correctly representing some part of reality, it is not totally clear from where the pressure comes to portray these judgements as something other than that which they appear to be. Gibbard may be right to insist that ‘When we look at ourselves as parts of the natural world . . . we won't find a responsiveness to anything non-natural’ (p. 13). But need those who think of normative judgements as sui generis, as irreducible to the expression of some non-cognitive attitude, conceive these judgements in terms of responsiveness? A third way suggests itself which seeks both to avoid committing us to a realm of non-natural facts to which we are responsive and to avoid construing normative judgement as a planning attitude from which it is intuitively distinct. This is the sort of anti-ontology view, or cognitivist irrealism, favoured by Hilary Putnam, T. M. Scanlon and John Skorupski, among others, which holds out the hope of allowing normative judgements to be in perfectly good order while being just what they appear to be. We do after all have a fairly good pre-theoretical understanding of what it means to say that we have reason (or most reason) to do, or to believe, or to feel something. And we have a very high degree of confidence in very many of these normative judgements that include not just practical normative judgements, such as that there is reason not to inflict severe pain for fun, but also epistemic and affective judgements, such as that the fact that everyone outside has an umbrella up gives me reason to believe it is raining, or that the fact that my loved one has died gives me reason to feel sad. Given that understanding, and that confidence, offering an epistemology of such reasons judgements, if it can give us a convincing story about when such judgements are warranted and when they are not, may be sufficient for us to rest easy in seeing normative judgements as being philosophically unproblematic, without having to construe them in the way Gibbard does, as piggy-backing on more fundamental, non-cognitive attitudes from which they appear distinct.

There is no doubting, however, that Gibbard's own naturalistic picture of the normative is exceptionally rich, and the ways in which he develops it in Reconciling Our Aims are fascinating for anyone interested in the bases of ethics and of normativity in general.