Holly Smith

For centuries it has been a mainstay of European and American moral thought that keeping promises—and the allied activity of upholding contracts—is one of the most important requirements of morality. On some historically powerful views the obligation to uphold promises or contracts not only regulates private relationships, but also provides the moral foundation for our duty to support and obey legitimate governments. Some theorists believe that the concept of keeping promises has gradually moved to center stage in European moral thought. They see this movement as part of an historical shift from a moral conception in which an individual’s duties are mainly externally imposed and unalterable, to a conception in which duties are largely chosen by the individual.

ConclusionWe have now examined two versions of UG — that proposed by David Lyons, and a new version derived from his. Both were found to be extensionally nonequivalent to AU, but both were also found to be unacceptable moral theories (Lyons' version generated counter-intuitive prescriptions, and the derived version failed to generate any prescriptions at all in an important range of cases). Naturally, these results cannot be taken to show that no satisfactory form of UG will be forthcoming. But at least some of the conditions to be met by future proposals are now clear: any successful version of the theory must interpret the ‘generalized performance’ of an action to include only the actions of agents who face similar ranges of alternatives, and (in some cases) must select as relevant a description of the action which does not refer, either explicitly or implicitly, to the actions of other agents. Reflection shows that the same remarks hold true, mutatis mutandis, for rule utilitarianism as well: unless the rules are formulated with extreme care in these respects, they may yield unacceptable recommendations, or else fail to recommend any action at all in the very cases they are designed to resolve.17.

Because we must often make decisions in light of imperfect information about our prospective actions, the standard principles of objective obligation must be supplemented with principles of subjective obligation (which evaluate actions in light of what the agent believes about their circumstances and consequences). The point of principles of subjective obligation is to guide agents in making decisions. But should these principles be stated in terms of what the agent actually believes or what it would be reasonable for her to believe about her prospective actions? I show that there are many decisions for which “reasonable belief” principles can’t be used by the decision-maker, especially in cases in which whether (or how) the agent investigates or deliberates affects the nature of the prospective action itself. I conclude that subjective rightness depends on what the agent actually believes, not what it would be reasonable for her to believe.

Recently two distinct forms of rule-utilitarianism have been introduced that differ on how to measure the consequences of rules. Brad Hooker advocates fixed-rate rule-utilitarianism, while Michael Ridge advocates variable-rate rule-utilitarianism. I argue that both of these are inferior to a new proposal, optimum-rate rule-utilitarianism. According to optimum-rate rule-utilitarianism, an ideal code is the code whose optimum acceptance level is no lower than that of any alternative code. I then argue that all three forms of rule-utilitarianism fall prey to two fatal problems that leave us without any viable form of rule-utilitarianism.

The requirement that moral theories be usable for making decisions runs afoul of the fact that decision makers often lack sufficient information about their options to derive any accurate prescriptions from the standard theories. Many theorists attempt to solve this problem by adopting subjective moral theories—ones that ground obligations on the agent’s beliefs about the features of her options, rather than on the options’ actual features. I argue that subjective deontological theories suffer a fatal flaw, since they cannot appropriately require agents to gather information before acting