DEONTOLOGICAL MORAL OBLIGATIONS AND

Many claim that a plausible moral theory would have to include a principle of beneficence, a principle telling us to produce goods that are both welfarist and agent-neutral. But when we think care-fully about the necessary connection between moral obligations and reasons for action, we see that agents have two reasons for action, and two moral obligations: they must not interfere with any agent’s exercise of his rational capacities and they must do what they can to make sure that agents have rational capacities to exer-cise. According to this distinctively deontological view of morality, though we are obliged to produce goods, the goods in question are non-welfarist and agent-relative. The value of welfare thus turns out to be, at best, instrumental. Many theorists think that two related claims will occupy centre stage in any plausible moral theory. The first is that we should bring about more rather than less of what’s of intrinsic value. The second is that welfare has intrinsic value. Putting these two claims together, they suppose that any plausible moral theory will tell us that we should produce more rather than less welfare. Richard Arneson, who is typical of the theorists I have in mind, puts the point this way: The concept of intrinsic value is not merely a building block in consequentialist theories, and if this concept (or the best revi-sion of it we can construct) is found wanting, the loss would have wide reverberations. More is at stake than the status of consequentialism. I suspect any plausible nonconsequentialist morality would include as a component a principle of benefi-cence. In a consequentialist theory some beneficence principle is the sole fundamental principle; in a nonconsequentialist theory beneficence would be one principle among others. Whatever its exact contours, a beneficence principle to fill its role must rank some states of the world as better or worse, and direct us to bring about the better ones within the limit