I

The ethics of Bernard Williams has been characterized in many ways. A good example of a variety of interpretations gathered in one place is the collection of essays in his honour, World, Mind, and Ethics, where Williams’ ethics is examined from a range of very different perspectives. Charles Taylor focuses on aspects of Williams’ ethical thought which can be seen under a Communitarian light;Footnote ¹ while Martha Nussbaum recruits Williams for her project, conceived within a liberal horizon, of producing an ethics which has the philosophy of Aristotle as its beacon.Footnote ² Williams has also been compared to David Hume. For example, in The Sources of Normativity Christine Korsgaard puts Hume and Williams – and John Stuart Mill – on a par, since according to her they all explain normativity as a form of ‘reflective endorsement’;Footnote ³ but normally Williams is considered a Humean above all because of his instrumental explanation of reason in ethics.Footnote ⁴ Finally, in his last work, Truth and Truthfulness, Williams plays still another role, that of the Nietzschean, pursuing an inquiry which makes use of a genealogical method.Footnote ⁵

It does not seem that he recognizes himself in any of these philosophical traditions; most likely the truth is that Williams’ thought is one of a kind which cannot – and must not – be compressed in one single model. This should not prevent us, however, from trying to acquire a better understanding of his position through the philosophical exercise of comparing it to other traditions and authors, to discover if, and in what measure, it has been influenced by them. In this article I shall read some specific aspects of Williams’ ethics in the light of the philosophy of David Hume. Starting from some critical observations made by Williams regarding a Kantian conception of morality, I shall try to show that he and Hume converge on many important points; their stance on ethics is similar and goes well beyond what is commonly held to be their shared definition of practical reason as simply a means–end device.

II

Williams’ ethical stance is known to be an anti-theoretical one. He criticizes most of contemporary moral thinking for its tendency to conceive of ethics as an object which is to be organized in the form of a moral theory composed of a certain number of principles, procedures and rules.Footnote ⁶ Williams maintains that the ambition of moral philosophy to provide objective grounds on which to base ethical life is doomed to failure. Philosophy alone cannot untangle ethical problems; philosophical reflection may appear as the most abstract and general kind of thought, hence suitable to embrace the whole sphere of ethics in all its aspects, but in fact it reveals itself to be inadequate in dealing with the extreme multiplicity which characterizes ethics. In particular, Williams thinks that the alleged universality which moral philosophy claims to possess is actually fictitious, for it is the result of a peculiar ethical conception: one which used to have a Christian foundation, which in modern times has changed into a morality of duty exemplified at its best by Kant,Footnote ⁷ and which he considers to be the paradigm of the great majority of present-day moral theories. These theories, even though they may appear to be very different, are seen by Williams to share some common features.Footnote ⁸ First, they depend upon the idea of the autonomy of ethics, whereby ethics is conceived as a self-sufficient domain. Second, they are governed by the notion of obligation, and they are consequently structured into rigid systems where obligations are given a determinate ranking, such that it is always possible to say which superior obligation prevails over which inferior ones. Moreover, they consider only voluntary actions to be expressions of true moral actions. And lastly, the only properly affective moral reactions admitted by them are blame (from other people) and remorse, regret or guilt (from oneself). These traits, Williams thinks, are serious flaws in a correct view of ethics, and may be traced back to the Kantian conception.

What are the main characteristics of Kantian morality? According to Williams, its main feature is its purity.Footnote ⁹ Kantian morality acquires its authority by being completely autonomously related to the situation in which people are placed, separate from their personal emotions, sentiments, desires, and aims; this authority has nothing to do with the contingencies of the individual lives of people, since it springs from that common reflective nature which detaches human beings from their specific identities and qualifies them as moral subjects par excellence. From this perspective, human beings are primarily rational creatures; their rationality enables them to reflect upon themselves and to impose upon themselves a morality in the form of a categorical imperative, that is to say a non-mediated command, a law whose only motivational power is its transparency. Furthermore, because of their rationality, human beings conceived in Kantian terms are radically free; no matter what the particular circumstances in which they find themselves, or the particular identities they have, they can see their own condition as accidental and not necessary for the determination of their actions. This allows them to think about themselves as always in principle capable of acting otherwise. Finally, their shared rationality makes human beings all equal, since they are all governed by the same moral law produced by their common rational essence, which can thus be said to be universal. Hence, Kantian morality, by appealing to a notion of rationality which contains in itself the notions of freedom and equality, claims to have found a final foundation for ethics, which is all the more strong, since it appeals to nothing outside itself for its legitimation.

III

If ethics were in effect based on these Kantian grounds, it would have such a strength and inevitability that we would always find the right solution to our ethical dilemmas. For Kantians, in fact, rationality is the constitutive aspect of human beings conceived as moral beings; they can be said to be so exactly because they share that common rational trait which makes them all ‘citizens of the Kingdom of Ends’. Since reason is seen as an intrinsic characteristic of the nature of moral subjects, they possess in themselves a faculty which can ensure both the prescriptive element of reasons and their motivational force. In this sense, the normativity of morality is guaranteed by the practical rationality of the agents, which can motivate them by referring exclusively to itself.

According to Williams however, we cannot rely on such premises: in fact, following the same line as Hume, Williams is convinced that ethics has to do with individuals whose human nature is basically specified by its being essentially sentimental, rather than rational. It is indeed true that human beings are rational creatures; but they are not just that. Besides rationality, there is a whole series of passions, sentiments and emotions which guides the conduct of people and marks their identities in a way which prevents us from being able to single out a pure rational nature as separate from the rest. Furthermore, even supposing it were possible, it is doubtful that rationality alone could guarantee the obligatoriness which characterizes moral prescriptions. While Kantians believe that morality deserves this name only if it is grounded on motivations which are ‘universal, rational and of principle’, Williams agrees with Hume that, on the contrary, ‘there are desires of different degrees of generality, involving degrees of remoteness and independence from the subject: the slide towards morality is a slide along these continua’.Footnote ¹⁰ Specifically, Williams presents a model which, he claims, is inspired by Hume's conception of rationality in ethics, based on questioning the existence of a single rational deliberative process which provides correct reasons to act, capable of presenting themselves as obligatory for all agents. According to this model, we can claim that an agent has a reason to Φ – ‘(where “Φ” stands in for some verb of action)’Footnote ¹¹ – only if we can work out a ‘sound deliberative route’Footnote ¹² whereby we can trace back the agent's reason to her ‘subjective motivational set’ (S).Footnote ¹³ By S Williams does not only mean the agent's desires broadly conceived, but also her ‘dispositions of evaluation, patterns of emotional reaction, personal loyalties, and various projects, as they may be abstractly called, embodying commitments of the agent’.Footnote ¹⁴ For Williams, then, a normative reason is such only if it can be traced back to some element already present in the agent's S; that is, only if the agent, starting from S, knowing all the relevant facts, and deliberating rationally (‘rationally’ in a strictly means–end sense) is moved to Φ.

The agent's sound deliberative route has nothing to do with rationality as conceived in Kantian terms, but it amounts to an heuristic and creative process, by which we figure out how to satisfy our ends, our desires, or our values; it consists of an exercise of the imagination which calls into question our whole S, and at the end of which new elements become part of S, whereas others are dismissed. This process is fundamentally indeterminate: ‘there are no fixed boundaries on the continuum from rational thought to inspiration and conversion’.Footnote ¹⁵ Williams does not deny that someone who is indifferent to a given normative reason may be blamed. She may not, however, be charged with being simply irrational; such a person may be blamed, rather, for being cruel or selfish or thoughtless – i.e. she can be said to be an insensitive person, who is not irrational.

IV

According to Williams therefore, deliberating correctly is a process which unfolds from the agents’ present subjective motivational states, is entirely dependent on them and cannot be governed by external rational criteria for their correctness. We would not be able to account for one of the most important aspects of ethics – namely that moral reasons endowed with practical force are capable of moving us to action – by simply appealing to hypothetical conditions of rational practical thought; instead, rationality should be regarded as one of those faculties which, together with a complex range of passions and sentiments, forms a crucial part of a human psychology which is far more elaborate than is suggested by the Kantian picture of morality.

Like Williams, Hume recognizes that a properly ethical inquiry must provide a convincing answer to the motivational problem, and he also believes that it is impossible to find a satisfying solution to that problem if we do not produce an explanation of the phenomenon of ethics which takes the form of a moral psychology, in which the passional and the imaginative dimensions of human nature are central.Footnote ¹⁶ Reflecting on ethics in the light of such assumptions brings Williams’ approach very close to Hume's sentimentalistic one. From Hume's sentimentalistic perspective, ethics finds its explanation as well as its justification in human sentiments. For him, ethical thinking is part of a broader examination of the mechanisms which structure human nature. Now, even though Williams believes that no particular conception of human nature will ever provide us with a definitive foundation for ethics, nonetheless he also says that reflecting on human nature ‘is at any rate a comprehensible project, and . . . it represents the only intelligible form of ethical objectivity at the reflective level’.Footnote ¹⁷ If we want to understand ethics, says Williams, we cannot do that independently from developing a theory of human nature in which ethical reflection makes sense. This project, however, can be developed only through the convergence of different disciplines – history, sociology, psychology – whereby we form a picture of human nature which is not abstractedly conceived, but is always attentive to the specific reality in which particular individuals live. So, if we want to find out what human nature is, and if we want to use this knowledge for ethical purposes, we ought to regard it as something which can never be severed from its specific realization in a specific, contingent reality.

But is this not remote from Hume's Enlightenment project of presenting a conception of human nature which is as refined as possible – human nature as it is ‘in itself’? Less so than may appear to be the case at first glance. In fact, in the Introduction of Hume's Treatise of Human Nature we find a declaration of intent that is very similar to the one Williams adopts; Hume's study of human nature is presented as always field-tested, as it were. Hume says

Hume's project is that of providing a scientific explanation of human nature, but from the very beginning this is understood to be a morally laden enterprise, in the sense that he states clearly that we cannot hope to arrive at any definition of human nature in a pure, abstract form; instead, human nature can be described only by examining sentimentally characterized people, placed in a social context from which they cannot be removed. This conviction is already clear in the Treatise, and it will become stronger in the two Enquiries and in the Essays.

Thus, Hume and Williams see ethical reflection as one aspect of a wider survey of human nature, pursued by examining how human beings act in the particular reality in which they find themselves. An ethics which hinges on the notion of reason alone is to be discarded: they are both convinced that rationality represents only a small part of this picture of human nature; sentiments, emotions and passions are at the forefront, and carry more weight. It is only by leaving aside the ambition of establishing morality on certain structural conditions of rational practical thought, and starting instead from a richer, basically sentimental psychology, that it is possible to explain the practical value of normative reasons and to give a convincing account of our central ethical concepts.Footnote ¹⁹

V

Like Williams, Hume was convinced that reference to reason often conceals an idea of ethics characterized in religious terms, which cannot but result in a form of impositional control; and the whole work of Hume can be read as a critique of the influence of religion on human affairs. Far from considering ethics as something which has to do primarily with commands and obligations – no matter whether imposed by God or by ourselves – Hume shows a distinct preference for a conception of ethics which recalls the classical one, as a study of those virtues and vices which compose our ethical universe. If we read the first section of the Enquiry Concerning the Principles of Morals we see that the aim of ethical reflection is that of formulating that list of virtues and vices which constitutes our moral vocabulary. Hume says

This catalogue can be obtained by applying ‘those universal principles, from which all censure or approbation is ultimately derived’,Footnote ²¹ and which enable us to distinguish virtues from vices: discriminating between personal qualities which are or are not of use to ourselves or to others, or immediately pleasurable or painful to ourselves or to others.Footnote ²²

Williams’ intentions are not very different: the moral philosopher should study people in their specific circumstances, so as to determine that system of values – that is, that series of virtues – which allows human nature to flourish.Footnote ²³ We may therefore classify Williams’ attitude towards ethics as a kind of virtue ethics;Footnote ²⁴ and it is worth noting that, in so far as Williams may be considered a representative of this method of ethics, he is not an Aristotelian (as many contemporary virtue ethicists are),Footnote ²⁵ rather he follows the naturalistic line exemplified in Hume. In fact, Williams argues that following the Aristotelian path today is an arduous project, since it does not seem possible to be an Aristotelian without accepting the finalistic metaphysics on which the Aristotelian system is based; a metaphysics which, considering modern evolutionary theory, Williams is convinced should be given up.Footnote ²⁶ For his own part, Hume – who in turn can legitimately be considered an exponent of a tradition of virtue ethicistsFootnote ²⁷ – refuses all appeal to final causes, thinking that a complete survey of those passions in which human nature expresses itself can be presented by acting as an ‘anatomist’ of human nature,Footnote ²⁸ i.e. by applying an experimental method to the study of human nature similar to the one used by Newton in his study of the physical world.

VI

Having virtues and vices as the principal target of moral inquiry, the attention of both Hume and Williams is focused on evaluating the sources of virtue and vice, viz. the dispositions of individuals. The objects of moral examination are mainly the motives which move people rather than their actions, and what is morally pregnant is the consideration of those character traits which define people as moral agents. This is why individuals are the core interest of ethical reflection for both figures; they become morally relevant as bearers of a particular character, the virtuousness or viciousness of which makes them objects of moral appraisal.Footnote ²⁹

Seen in such a light, the people with whom Hume and Williams deal are very different from the Kantian rational moral agents in another important respect. We have said that, for a Kantian morality, the very fact that the agent is rational makes her unconditionally free; and her freedom, in turn, is what enables her to act in a truly moral way. For a Kantian morality, and also for the majority of contemporary moral theories, the sine qua non of moral dominion is that agents can express radically free actions. People may be deemed responsible for what they do, and may thus be evaluated morally, only by presupposing that they are capable of acting freely, i.e. that their actions are not the product of the circumstances in which they find themselves. On the contrary, Hume and Williams share the conviction that people cannot avoid what Williams calls ‘constitutive luck’.Footnote ³⁰

For Williams, the idea that a person can be considered truly responsible only if she is immune to luck is due to a misconception: that of believing that being an agent exclusively means being capable of acting voluntarily according to a decision. Specifically, people are deemed to acknowledge an action as their own – which is to say, something of which they are the cause – when it takes the form of a self-imposed command. Under these conditions, an action can be considered free, and thus moral. But Williams notes that the awareness we have of our own acting differs from the description that can be made of it from a third-person perspective. With respect to our will, although we perceive ourselves to be in command, it has to be acknowledged that what we represent to ourselves as the conclusion of active deliberation should also be regarded as demonstrating obeisance to non-voluntary factors. As Williams says, referring to Nietzsche, ‘what is called “willing” is a complex of sensations, thinking, and affect of command’; we are at the same time the one who commands as well as the one who obeys, and that these two aspects of the self should be kept separate is a premise which is not self-evident.Footnote ³¹ This consideration could have been directly taken from a famous note of the first Enquiry, where Hume says that, even though we may imagine that ‘the will itself is subject to nothing’ because we feel we are intrinsically free – that is, we feel that, notwithstanding our having acted in a particular way, we could have acted otherwise – somebody who knows us can predict our actions as the result of our having a specific character.Footnote ³²

According to Williams, moral luck plays a fundamental and inevitable role in ethics, determining its substance as well as its aims. Ethics is intrinsically a posteriori, and passivity is one of its fundamental characteristics.Footnote ³³ The hypothetical capacity of people to act freely plays a minor role in explaining and justifying their moral responsibility, and when we have to judge ourselves and others in ethical terms, we refer to an authority which goes well beyond intentional agency. Ethical judgement is ‘a “whole person” response’,Footnote ³⁴ which certainly has to do with voluntary actions, but also has to do with aspects of character which are not under our control. Hume also conceives of ethics as something which cannot be circumscribed within clear and definitive limits by appealing to free will. On the contrary, he thinks that what should be done is

And what results from these reflections is that it is not possible to distinguish a priori between virtues and talents, or between vices and defects – that is, between the voluntary aspects of our personality and the non-voluntary ones.Footnote ³⁵ The object of moral judgement proves to be a virtuous or vicious character whose formation and whose moral status is largely independent of the exercise of free will. If freedom plays a role in ethics – as indeed it does – that is limited to a ‘liberty of spontaneity’ which is explained exactly by conceiving people within a causal dimension which determines them. ‘Liberty of indifference’ – i.e. freedom from any causal chain whatsoever – cannot be admitted, because if we did that, we would not be able to trace people's actions to the exercise of their character, and thus to attribute responsibility for what they do.Footnote ³⁶

VII

Hume and Williams are on the same side in believing that our recognition of people as moral subjects – as bearers of characters which may be assessed on moral grounds – does not surpass the specific sphere in which people move and make their choices. A moral subject acquires depth not by appealing to a transcendental moral dimension, but by living in a shared context in which her sentimental psychology can reveal itself. A moral subject becomes what Williams calls a ‘mature agent’Footnote ³⁷ when she is aware that her identity is the product of more than just her deliberating self and she is thereby properly responsible for what she does. A mature agent is a concrete individual who moves within the context of a familiar reality by which she is determined, and within which she is able to find her bearings by being able to perceive its morally relevant aspects – that is, she is able to understand what Williams calls ‘thick ethical concepts’.Footnote ³⁸ None of this would be possible without a continuous interchange among people. Moral subjects, by measuring themselves against others within a particular social world, by reflecting on themselves in other people's judgements and by having to account for their actions, can elaborate a stable image of themselves and recognize themselves as ethical beings.Footnote ³⁹ That is why, on Williams’ account, the sentiment of shame plays a key role in explaining the moral identity of people – a role which is surprisingly similar to the one played by the passion of humility in Hume's moral psychology.

Williams observes that generally the ‘autonomous moralities’Footnote ⁴⁰ opt for guilt rather than for shame when they have to mention moral sentiments. Within such conceptions, guilt is identified as the typical moral sentiment, since it does not appear to be compromised with our personal desires or needs, nor with those of any concrete person in particular, but it expresses our respect for an abstract moral law. Williams himself discards guilt for shame. He describes shame as that uneasiness we feel when we are seen by the wrong people at the wrong moment.Footnote ⁴¹ When we feel shamefaced we feel naked, and exposed to the gazes of others; our reaction is to cover or to hide ourselves. To feel ashamed corresponds to the consciousness of having lost power; a loss originally due to being physically exposed to other people's judgements, but which in the ethical context turns into a more general feeling of being at a disadvantage with regard to someone who is not even necessarily present.

When we feel ashamed, Williams says, it is as though we were before somebody in whose eyes we have failed.Footnote ⁴² This ‘imaginary observer’Footnote ⁴³ poses obligations to which we feel bound; it is somebody, rather than anybody, whose opinions we respect, and whose authority derives from the fact that it represents genuine social expectations. The imaginary observer has an identity which is independent from that of the agent, but the latter has internalized the judgement of the former, has made it her own and has thereby become capable of understanding herself as somebody who can act morally. The claims of the imaginary observer represent the requirements of an ethical reality into which individuals are immersed, and shame gives the coordinates by which individuals are able to shape a character, relate with others and with the world, and hence recognize themselves as moral selves.

If we move to Hume, we find a very similar psychological mechanism in the passion of humility. Hume is generally considered to be the philosopher who most powerfully denied the existence of personal identity, claiming that the self would be nothing but ‘a bundle or collection of different perceptions’.Footnote ⁴⁴ However – as many scholars have convincingly arguedFootnote ⁴⁵ – Hume regains a notion of a unitary self when he moves from the survey of the understanding in book one of the Treatise to that of the passions in book two, and then of morals in book three. That is due in particular to the indirect passions of pride and its opposite, humility. Whereas as an object of the understanding we are not allowed to speak of any self-consciousness whatsoever, thanks to the passions of pride and humility, people become conscious of themselves as unitary individuals. From this different, passional point of view, human beings are no longer considered as a heap of discrete perceptions, but as physically identifiable flesh-and-blood individuals, and capable of feeling, and being the objects of, peculiar passions.

What emerges from Hume's treatment of these passions is that they are characterized as fundamentally social.Footnote ⁴⁶ Hume says that ‘Our reputation, our character, our name are considerations of vast weight and importance; and even the other causes of pride; virtue, beauty and riches; have little influence, when not seconded by the opinions and sentiments of others.’Footnote ⁴⁷ That is because human beings ‘are mirrors to one another’;Footnote ⁴⁸ by being proud or humble – but, following Williams, it may be more correct to say by feeling shame, even though Hume does not use this expressionFootnote ⁴⁹ – human beings react to each others’ opinions and thus form a conception of themselves as positively or negatively evaluated by those who surround them. In Hume then, the passion of humility – and of pride – plays a comparable role to that of shame in Williams insomuch as it is considered a fundamental factor in determining the identities of human beings, regarded as individuals who are related to and dependent on others. When their feelings of pride or humility are caused by the evaluation of character traits which are considered as virtues and vices, human beings become conscious of themselves as moral subjects; that is to say, as the bearers of particular characters which – by being the objects of other people's scrutiny, as well as their own – mark them as morally distinguished agents.

VIII

Humean pride and humility, and Williams’ sense of shame, are the intersubjective sentimental mechanisms which relate people to each other and with reference to which the moral identity of any given individual may be framed. Thus, Hume and Williams develop ethical reflections which definitively have more in common than is normally believed. They both centre on people conceived as concrete individuals, who are the protagonists of an ethics understood as being based on the flourishing of the potentialities of a common human nature always seen within the specific context of its realization. Their shared conception of the moral subject cannot in principle be translated into a single feature which is supposed to express the moral subject's true moral essence. Any attempt to do that will inevitably end up with an impoverished description of people, as the Kantian case well demonstrates, and the morality which results cannot but be gloomy and, far from being self-chosen, will be a fundamental imposition.Footnote ⁵⁰