Pollock offers a strikingly coherent account of the nature of moral reasoning.

His first main topic is the naturalistic fallacy, which tells us that we cannot define the truth conditions of moral statements in terms of nonmo-ral concepts. Pollock acknowledges the validity of this rule, but points out that similar situations prevail in other areas of philosophy. We can hardly ever define the truth conditions of any concepts, as the efforts of mid-twentieth century philosophy have shown. The way to escape from this difficulty is to look beyond the logical functioning of concepts to their epistemological functions. Not all reasons involve entailment. Many are prima facie reasons that are not by any means certain, but in the absence of conflicting evidence seem to be legitimate.

It will come in handy to pause to explore what sorts of evidence can take the wind out of a prima facie reason. Pollock specifies two kinds. The first is a 'rebutting defeater1, which provides evidence contrary to the proposition in question. The second is an 'undercutting defeater', which attacks the connection between the reason and the conclusion.

Pollock now begins to apply these ideas to moral judgments. Convin-

CURRENT PHILOSOPHY 23

cing arguments have been offered against the idea that correct moral judgments are based on what on 'Ideal Observer1 would decide. Commission of the naturalistic fallacy is among the problems with this approach. But if we see the authority of the Ideal Observer as a prima facie reason for moral judgments rather than a logically conclusive reason, we may be able to avoid these difficulties. But how can we frame such a version of an I-deal Observer theory?

We have to shift gears to get at the specifics. Pollock addresses the nature of 'abstract desires'. We can desire certain things in imaginary situations. We can likewise approve (or disapprove) of things in imaginary situations. What Pollock suggests is that when we judge a concrete situation, especially one in -which we are not impartial, we do it by considering an abstract situation exactly like the one we are in except that we imagine that we are impartial. This abstraction is flexible in that we can narrow it as necessary to include all relevant (and perhaps conflicting) considerations that apply in the real-life situation. But it remains an abstraction after all. And because it is an abstraction, it has the critical property of being able to be made universal.

Furthermore, because -we are dealing with prima facie reasons here, we are always subject, at least in theory, to run up against conflicting reasons. Now this might not occur if the only available 'defeaters' were rebutting defeaters. But undercutting defeaters play a key role. We may initially believe that something is morally acceptable, but then we discover that others -- perhaps many others, or perhaps only a single other whose judgment we respect -- react otherwise. This fact then casts doubt on our original moral judgment and gives us grounds to revise it. Another way in which our moral judgments develop is through what Pollock calls a 'personal ignorance defeater'. In these cases we learn through what others tell us that we have not been abstracting the moral situation correctly due to our inability to understand or appreciate some aspect of it.

Pollock realizes that what he has given us is a way to make moral judgments, but no clue as to what the content of those judgments should be. We have a normative ethics ■without any metaethics. Does this then make moral truth entirely subjective? Pollock is hopeful that it doesn't. He has given us a legitimate way to revise our moral judgments based on how others judge a situation, and he points out that just because not all moral differences have been resolved we are not free to conclude that they cannot be resolved. If we could get everyone to think rationally, we could probably get them to agree on moral issues.

We can hope, therefore, to approximate the results of moral objectivity. But we would still like more. We would like to get beyond moral consensus to assurance of moral truth. Yet this is asking too much. Pollock believes. We can no more get at absolute truth in moral matters than we can in other areas of philosophy. We will have to be satisfied with a moral theory that can give us grounds both for our making and for revising moral judgments.