Charles Camosy's Peter Singer and Christian Ethics is a very good book. It is well-written and clearly argued, and it is about an important question, namely whether Peter Singer's ethical work can be reconciled with Christian ethics. The subtitle of the book gives Camosy's answer, ‘Beyond Polarization’. He thinks that the similarities between Christians and ‘Singerites’ are more important than the differences (though the latter are also real).

One quibble is that the book is not really about Christian ethics as a whole, but about Roman Catholic ethics. The authoritative sources cited are from Aquinas or various papal documents or the Catholic Catechism (though I should add that he cites Protestant sources as well, including my own work). I have no quarrel with this limitation, except for the faulty advertising. But there is an opportunity missed here. In particular, Singer's work belongs in a trajectory from Scotus through the Reformation to Hutcheson and the Utilitarians. If the project was to show the prospects of rapprochement, it would have been better to treat at greater length those parts of the Christian tradition closer to Singer's own heritage. One symptom of the problem here is that Camosy treats natural law theory, and the accompanying eudaimonism of Aquinas (holding that all our actions and choices are directed at happiness) as the conversation partner with Singer. But in fact Singer is closer to the account of obligation from Sidgwick and Kant which puts it in tension with eudaimonism, and which descends from a version of divine command theory. The relation of Singer to natural law is indeed interesting, but only part of the picture.

The key question here is whether to accept a single or a double account of motivation, and Singer is committed to a double account. For Aquinas, following Aristotle in the first sentence of the Nicomachean Ethics, we pursue everything that we pursue for the sake of our happiness. But for Scotus, following Anselm, there are two basic affections or drives of the will. The first is the affection for advantage, a drive towards our own happiness and perfection. The second is the affection for justice, a drive towards what is good in itself independently of any relation to oneself. Scotus thinks we have both affections, and it is good that we should, but the question is how we rank them. For Singer likewise there are two basic motivations. He takes his language here from Sidgwick, not Scotus, but the distinction is the same. Reason gives me both the end (purpose) of my own well-being and the end of the well-being of all those affected by my actions. There is nothing irrational, he says, in preferring myself and my own interests, and the interests of those in special relations to me, but morality requires me to overcome this preference. Christian ethics has both the Thomist and the Scotist traditions within itself, and there has recently been a rebirth of lively interest in the relations between them.

The chapters of the book which deal with medical ethics, with the treatment of non-human animals, and with our obligations to the poor, are uniformly helpful. I would recommend them to anyone who wanted to know about Catholic teaching in these areas, whether she was interested in Peter Singer or not. For reasons of limited space, I will confine my comments to the final two chapters, both of them about the ethical theory that Singer applies rather than about the application of the theory.

Singer is, as Camosy notes, a student of R. M. Hare, and it is Hare's theory of universal prescriptivism that Singer applies. There has been, however, a change since Singer read Derek Parfit's On What Matters, and Singer no longer endorses a Humean account of practical reason (as Camosy acknowledges). I want to comment on the history here, because I think it illuminates the current relation between Singer's ethics and Christianity. R. M. Hare, in Moral Thinking, endorsed a two-level version of preference utilitarianism. At the ‘intuitive’ level we work by intuitions about what is right and wrong, for example the intuition that it is wrong to break a promise. These intuitions we largely inherit from our culture and upbringing. But he thought that we cannot think exclusively at the intuitive level, for three reasons: our intuitions may be incomplete (not covering new situations we encounter), or in conflict (if I do not break my promise, I will cause another person intense suffering), or just plain wrong (I have grown up in a racist culture). At the ‘critical’ level, we try to approximate the thinking of a figure R. M. Hare called ‘the archangel’, who decides what to prescribe by looking at how many of the basic preferences of the affected parties will be satisfied, and then prescribing their maximum satisfaction. I will explain what I mean by a ‘basic’ preference in a moment. The archangel has unlimited access to the relevant information and unlimited sympathy, and also decides what principles for action to prescribe for those who are not unlimited in these two ways. R. M. Hare sometimes used the term ‘God’ rather than ‘archangel’. He did not propose, however, that this version of preference utilitarianism covered the whole of moral thinking. In particular, it did not cover ideals, and he was not able in his subsequent work to develop a theory which did cover them. The difference between ‘basic’ preferences and ideals is that a basic preference is either now for what I should experience now or at some future time for what I should experience at that time. Ideals cover preferences for what should happen to me at some future time, whether I then still have the preference or not. For example, I may have the ascetic ideal of living without material entanglements, and I may now want that for myself even if I subsequently deteriorate into self-indulgence and stop caring about this sort of purity. Ideals also cover preferences for states of affairs I never experience. For example, I may have the ideal of a faithful relationship with my spouse, and be harmed by her unfaithfulness, even if I never learn about it and the experienced quality of our lives together remains the same. R. M. Hare put his theoretical difficulty by saying that he could not accommodate (within his theory) preferences for states of affairs which are not presently within the experience of the person having the preference. ‘I am inclined to think that we ought [to accommodate them], but that I cannot at present do so. There is obviously unfinished business here’ (Moral Thinking, Oxford: Clarendon Press, 1981, p. 104).

Singer, though faithful to universal prescriptivism in most ways, was not faithful to this qualification in scope. He claimed, by contrast, that preference utilitarianism could cover the whole of moral thinking without remainder. One place he explored this question is his reply to R. M. Hare in Douglas Seanor and N. Fotion (eds), Hare and Critics (Oxford: Clarendon Press, 1988), pp. 147–59. It is this ambition that he is now recanting. The influence of Derek Parfit is crucial here. On What Matters is a sustained defence of moral realism against any attempt, like Hume's, to explain values as projections of our desires or volitions. This means that claims about the content of those values have somehow to be justified. If we are to avoid relativism, the justification cannot simply be by consensus. A justification from our nature as given by biological evolution does not seem likely to succeed, because we have been given by this evolution a mixture of good and bad. Rejecting his own previous rejection of R. M. Hare's modesty about the scope of basic-preference satisfaction puts Singer back in the position of R. M. Hare, of not knowing how to combine in a single theory the agent's consideration of maximising basic-preference-satisfaction with the agent's ideals (many of them, in R. M. Hare's view, religiously based) for what should be the case whether I experience it or not.

The relevance of this point about ideals to the relation of Singer and Christianity is that R. M. Hare's uncertainty here (and Singer's newfound perplexity) recapitulates the inclusion within Christian ethics of both God's impartial benevolence, which can be expressed in terms of formal justice, and God's creating and perfecting in Christ human life, which is a substantive ideal. Here the Christian ethicist has an important contribution, in the tradition of divine command theory. From within the framework of this faith we can say that there are values and obligations which have been revealed to us. Suppose we grant that God's benevolence is to be understood at least in part in consequentialist terms: God brings all things to good. What does this mean about our own ideals and their relation to consequences? It is important to see that the good towards which God brings all things is not itself to be understood either in terms of the satisfaction of our preferences or in terms of purported moral absolutes like the prohibition on lying. The good is, to use Scotus’ terms again, that we become co-lovers (condiligentes), entering into the love which is between the members of the Trinity. Prohibitions on lying (and, I would argue, the rest of the second table of the Ten Commandments) will not be relevant any longer to this final good. Our own ideals for our lives on this earth should be seen as preparatory for this final good. But then we are left again with the tension found in both Singer and R. M. Hare, because we see only fragmentary prescriptions for how to prepare for this final good, and the fragmentary prescriptions are not consistently practicable. We see that we have to respect people's agency, because it is through this agency that they become co-lovers; and we see that respecting this agency requires satisfying their preferences, other things being equal. But we also see that loving God requires loving the character of God, as far as we can see it, and this requires loving such things as faithfulness and truth. Holding our ideals for life on this earth as preparatory does not solve the ethical problems, but it gives them a context.

Camosy should be commended for his open-minded exploration of the similarities and differences between Singer and the traditions of Christian ethics. The remarks in this review about Singer and R. M. Hare should be seen as supplementary to Camosy's achievement, not as corrective to it. But if they are on target, they show how much more work is still to be done.