In ‘Ought and moral obligation’, Bernard Williams argues that ‘ought’ has more than one meaning, including the moral ought and the practical ought. However, he argues that in both these meanings, and perhaps in others too, ‘ought’ has the same logical form. He says it is a propositional operator, which expresses a property of a state of affairs. He denies the alternative view that ‘ought’ describes a relational property of an agent – specifically a relation between the agent and a possible act.
The logical form of a simple ‘ought’ sentence, according to Williams is always O(p). We may read ‘O’ as ‘it ought to be the case that’, so Williams would represent ‘Mary ought not to go so near the edge’ as equivalent to ‘It ought to be the case that Mary does not go so near the edge’. But this is not an adequate account of all oughts, since sometimes an ought is a property of (‘is owned by’, I shall say) of a particular agent. Often the owner is the grammatical subject of ‘ought’, but sometimes it is not. It is not, for instance in ‘You ought to get a medal for that’. Nor is it in ‘Mary ought not to go so the near the edge’ when Mary is an infant and the sentence is said to her mother.
I shall argue that the ownership of oughts cannot be accounted for with Williams’s logical syntax. For owned oughts we need the logical form NO(p), where ‘N’ names the owner. It is perhaps surprising that Williams should take the approach to oughts that he does take. Ownership is a particular sort of agent-relativity, and Williams often stresses the agent-relativity of normative properties. In ‘Ought and moral obligation’, he does indeed say that the practical oughts is relative to the interests of the agent, but in his mind this relativity evidently does not extend as far as ownership. I shall look for an explanation of this feature of his thought.
I shall also examine the relation between moral oughts, practical oughts and others.
Taking off from some of Bernard Williams's ideas in his collection Truth and Truthfulness, the paper will study a strange example (albeit fictional) in which it seems to be both utterly unimportant and at the same time enormously important whether or not a certain account is true. The account in question is the famous 'Atlantis' story from the hyperarchaic past, and the fictitious stage on which it is presented is Plato's dialogue-complex, the Timaeus-Critias. It is hoped that the study would contribute to a general discussion of why truth matters (when it does).
In prescribing against stealing and other actions, morality presupposes the existence of certain conventions: in this example, conventions of property. But any such conventions are bound to be morally contestable, so that a comprehensive morality ought to provide a basis for their assessment and, in particular, a basis for assessing them against an open-ended range of alternatives. Only a consequentialist philosophy, however, can meet this constraint. Any form of non-consequentialism has to privilege an agential point of view — say, that of 'we the people' — and this will rule out alternatives in which the political principal changes identity: alternatives, for example, in which peoples merge or divide under novel political arrangements.
It is sometimes said that the problem of moral luck is that we are attracted both to affirm that ‘We are morally assessable only to the extent that what we are assessed for depends on factors under our control’ and to affirm judgements which are inconsistent with it. Most writers on the subject focus on a narrower problem. They grant that possibly various moral assessments do apply to us in virtue of factors beyond our control. They believe, however that we are morally responsible for X only if X is under our control, or that we are responsible only because, and to the extent that X has aspects which are under our control. I will call this The Control Principle.
A simple way to understand the question of moral luck is as a question about the truth of The Control Principle: If it is true then there is no moral luck. If false then there is moral luck. Though things are not that simple, this question will provide the organising principle of the present essay. I will suggest that properly understood The Control Principle does not have some of the consequences it is often thought to have. For example, it allows for different degrees of blame to attach to wrongful intentions, wrongful attempts and wrongful actions. It also allows for blame to be attached to some beliefs, emotions and other attitudes. I will then offer an argument for rejecting The Control Principle.
Toleration is often said to be a puzzling or paradoxical value. Bernard Williams has suggested that in the very circumstances where it is most needed it may well seem impossible. The appearance of paradox arises in particularly acute form when one tries to provide a general justification of tolerance, that is, a general argument as to why people ought to be tolerant of one another. Important is the issue of justification is, however, this paper will concentrate on a slightly different issue. It will consider the question of what exactly is good about toleration, or, to put in another way, why so many people consider toleration to be an important value in its own right. What features of the practice of toleration enable it to attract the allegiance of its supporters? Clearly, this question is closely related to the question of justification. However, questions about the good of toleration are in one way less ambitious than the question of its justification, because the features of toleration that enable it to earn the allegiance of its supporters may not suffice to justify it to others. At the same time, attempts to provide a general justification of tolerance sometimes neglect the less ambitious question with which I am concerned. In attempting to provide reasons, acceptable to all, for endorsing a regime of toleration, they sometimes neglect the question of why some people find toleration an especially good or valuable feature of a society. In addressing this question, I will argue that, for many people, the phenomenon of being linked to others through a practice of mutual deference to one another’s values – that is, through shared participation in a practice of toleration – is experienced as a form of fraternity, and I will offer an explanation of why this should be so.
In 'Internal and external reasons' Bernard Williams famously argued that normative reasons must satisfy what's come to be called the 'internalism requirement'. As Williams himself pointed out in a subsequent paper, one of the main advantages of such a conception is that it lays bare the connections between what it is for someone to have a reason for action, on the one hand, and what it is for him to be responsible for what he does on the other. The question to be addressed in this paper is whether Williams's own Humean conception of internal reasons gets these connections right, or whether that requires us to embrace a more rationalist conception.
Consider a teenage girl who is contemplating motherhood. Prior to her becoming pregnant, it seems that she might truly judge that it would be a bad thing on balance to have a child at this stage in her life. After giving birth as a teenager, however, she might also truly judge that it is not a bad thing on balance that her child exists. These attitudes have commonly been thought to be in tension with each other. I argue, however, that when correctly interpreted in deliberative terms, as judgments about the agent’s reasons, the apparent conflict disappears. Giving birth changes the girl’s situation, in ways that give rise to corresponding changes in her reasons for action and for various emotional responses.
A consequence of this analysis, however, is that there may be mistakes or errors in deliberation that the agent is unable to regret having made. The teenage mother ought not to having conceived and given birth to a child at that stage in her life; and yet, as a mother, she can hardly regret having made the wrong decision in this particular matter. This raises large questions about the relation between justification and regret. Williams argues in “Moral Luck” that our decisions can be justified or “unjustified” retroactively through intervening circumstances that make regret either impossible or unavoidable. I challenge Williams’ assumption that justification and regret are necessarily connected in this way, and show that the things that drive a wedge between justification and regret need not have anything to do with epistemic luck.
One of Bernard Williams’s most famous and frequently discussed lines is his complaint that the man who considers whether it is morally permissible before choosing to save his wife (rather than a stranger) has one thought too many. As many of his readers have recognized, the claim, on the face of it, concerns the role of explicitly moral thought in our practical decision-making – specifically, it concerns the limits of that role.
While Williams’ point, thus interpreted, poses a challenge, especially for Kantian theories of morality, it is a challenge that sophisticated Kantians may be able to meet, saving their theory from the larger attack on such theories that William purports to be making in the essay.
In this paper, I shall suggest a less literal but more far-reaching interpretation of the intent behind Williams’s famous line. It is in the spirit of Williams’ discussion, I shall argue, to object to the pervasive concern for moral permissibility, not only in contexts in which immediate action is called for, but also in the context of cool reflection upon hypothetical cases. A husband who would so much as hope that he would act on behalf of his wife whenever, but only whenever, it was morally permissible to do so, may also have one thought too many - or, at any rate, a husband who had no such hope may be just as admirable (and certainly as loving) a man. Understood this way, Williams’s discussion is a forceful contribution to his more general attack on the idea that morality, with or without deliberation, should override or constrain every other commitment.