Sayre-McCord inspects 'deontic logic' and finds it wanting. He approves of its motives: to capture the underlying structure of our moral discourse, and to provide a neutral battleground for moral debate. According to his analysis, however, these two laudable desires prove incompatible.

The first attack on deontic logic, one already formulated by others, is concerned with moral dilemmas. The idea that there can ultimately be no moral dilemmas has been assumed as fundamental. Apparent dilemmas are held to be due either to an inability to discern our true obligations, or to the severity of the sacrifices which the situation demands. This belief apparently stems from the plausible intuition that a logic which permits conflicting conclusions is inconsistent and therefore fails.

But this intuition reaches too far. It is not, in fact, illogical to say that we ought to do both A and not-A, given B -- unless one also makes two additional assertions. The first is that 'ought' implies 'can', for of course one cannot do both A and not-A. And second, a 'principle of deontic distribution' would have to hold as well: that if one ought to do A (given C) and one ought to do B (given C), then one ought to do both A and B (given C). But neither of these supplementary principles can be taken for granted. Therefore, a deontic logic that allows for moral dilemmas is not necessarily false, although there is a potential trap here.

There is another possible problem having to do with permissibility, too. If we assume that all obligatory acts are permitted, and that an act is permitted only if we do not have any obligation to refrain from it, then moral dilemmas could not be accommodated logically. But again, if either of these two assumptions is denied, deontic logic can allow for moral conflicts.

One other principle to be considered is the one Sayre-McCord calls the 'principle of deontic neutrality', which states that deontic logic is to be the common ground on which all moral theories are tested. This principle is one that both proponents and opponents of deontic logic have supported. But it turns out that it lies at the very root of the problem. For Sayre-McCord is able to show that while deontic logic can be amended in such a way as to be neutral between moral theories that permit or do not permit moral dilemmas, this is a very demanding requirement. He identifies several specific problems. In the revised deontic logic which he describes, there is the doubtful assumption mentioned before, that 'ought' implies 'can': yet, not all moral theorists would accept this assumption. Second, it is presumed that when something ought to be, its logical consequences ought to be; this, too, is arguable.

If we further revise deontic logic to accommodate these disagreements as well, it becomes so anemic as to be virtually worthless. Thus to enforce the principle of deontic neutrality leaves us with no deontic logic at all. Perhaps this should not be a surprise though. It indicates only what we already knew -- that our various and competing moral theories are thoroughly controversial.

It is possible, admits Sayre-McCord, that some alternative logic could be developed to escape these difficulties. But the prospect is dim. Our next task ought to be to define some set of moral theories which offer the most reasonable accounts of obligation, and develop a logic that works at least for them. But identifying the moral theories to be included is a matter for moral theory itself, not for logic.

Contents

Index