One of the most striking characteristics of David Hume as a philosopher is his thoroughgoing naturalism, which runs through his epistemology and treatment of religion, and extends, crucially, to his theory of morality. In this paper, I examine Hume's naturalistic theory of morality and evaluate whether it is compatible with an objective theory of ethics.
It is important to begin by noting that Hume claimed nowhere to be building a theory of prescriptive ethics. His project was to give a naturalistic, scientific account of human cognition and morality. The name of his first and greatest work, the Treatise of Human Nature, bore the revealing subtitle Being an attempt to introduce the experimental Method of Reasoning into Moral Subjects, indicating his intention of doing for moral subjects what Newton had done for natural philosophy. In the third book of the Treatise, which is a treatment of morality, his aim was to explain the process whereby we come to make moral judgments, if only to counter two schools of thought on this subject that were prominent in his time: on one hand, the moral rationalism of men such as Samuel Clarke and William Wollaston, who held that morality could be explained as the product of reason, and on other hand, the ‘selfish’ philosophies of Hobbes and Mandeville, who saw all human morality as having its origins in considerations of self-interest.1 Hume’s concern was to replace these two philosophies, which he saw as misconceived, with his own empirical account, supported by the evidence of psychological introspection. That being said, I shall argue that Hume’s account of moral psychology consists of both positive and negative theses that, if accepted, narrowly constrain any theory of ethics.
Although Hume’s second Enquiry is devoted to explicating the principles of morals, the definitive statement of Hume’s theory of morality is found in the third book of his Treatise of Human Nature. Hume immediately makes it clear that he places himself firmly in the anti-rationalist camp, in opposition to moral rationalists such as Clarke and Locke. Moral properties are not inherent in objects and their relations, and thus cannot be inferred by our powers of reasoning. Human actions “do not derive their merit from a conformity to reason, nor their blame from a contrariety to it” and therefore “moral distinctions … are not the offspring of reason.”2 This exclusion of reason in the formation of moral judgments is due to Hume’s particular conception of reason. The powers of human reasoning alone cannot motivate us because reason is passive. It is, to speak anachronistically, the processing faculty of the mind. It can devise and evaluate means to some end, but cannot provide us with the ends themselves. Our ultimate ends, far from being the product of reason, are determined by our passions, which govern the ends we seek. Hence he maintains that “reason is and ought only to be the slave of the passions.”3
If reason itself cannot give rise to moral distinctions, perhaps moral distinctions are an empirical matter. This Hume flatly denies. Reason judges either matters of fact or matters of relations, but morality, Hume asserts, is never found in any matter of fact which it is possible to perceive or grasp by reason alone. We can examine the empirical world around us as much as we like, but we will never find any intrinsic moral facts ‘out there’:
Take any action allow’d to be vicious … and see if you can find that matter of fact … which you call vice. In which-ever way you take it, you find only certain passions, motives, volitions and thoughts. There is no other matter of fact in the case. The vice entirely escapes you … till you turn your reflection into your own breast, and find a sentiment of disapprobation.4
Since morality is not to be found in matters of fact, perhaps it lies in the nature of relations between objects. This too is impossible, Hume maintains, since the same relations can obtain between animals and plants as between humans, and yet former is never the subject of moral approbation and disapprobation while the latter often is. Relations alone, therefore, cannot give rise to moral distinctions. But if moral distinctions are found neither by reason nor in empirically ascertainable matters of fact or relations, then where do they have their origin?
Given Hume’s epistemological commitments, morality must be rooted either in matters of fact (i.e. empirically discoverable properties), relations of ideas (i.e. reasoning), or impressions -- the fundamental, primitive units of sense-experience that include both the input from our sensory organs and our emotional experiences. Since the first two have been ruled out, the process of elimination leaves us only with impressions. Hume therefore concludes that it is “by means of some impression … that we are able to mark the difference betwixt [vice and virtue].”5
Where do these impressions come from? They are “deriv’d from a moral sense”6 which provides us with impressions of moral approval and disapproval in evaluating the character traits of human beings, and the actions which these traits give rise to. Moral sentiments are basic units of sense-experience, and as such, they can be neither true nor false. One either experiences a certain moral impression or one does not, and that is all. It is important to note that Hume does not classify the moral sense as a proper sensory organ, apprehending properties and qualities that are “really out there”, much as vision apprehends the shape and outline of physical objects. The moral impressions caused by contemplating vice and virtue “may be compar’d to sounds, colours, heat and cold, which, according to modern philosophy, are not qualities in objects, but perceptions in the mind.”7 Although this is highly ambiguous and open to a number of interpretations, Hume’s position is clarified somewhat in the Enquiry Concerning the Principles of Morals, where he explicitly compares moral sensations to our sensations of beauty.8 The feeling of beauty is a sentiment which beholding something gives rise to, not a part of beholding it per se, and the same is the case with our moral sentiments. Virtues and vices are simply those traits which arouse our approval or disapproval in contemplating them, and we consequently approve or disapprove of actions based on our evaluation of the virtues or vices we believe to have given rise to them.
The impressions from moral sense – our feelings of approval or disapproval – are caused, according to Hume, by our ability to sympathise. Sympathy is a natural propensity to share the feelings of other human beings. This sympathetic faculty is the source of moral impressions and consists in a psychological mechanism whereby one person experiences the sentiments of another through an act of the imagination. Moral impressions, the product of sympathy, are sentiments with their own, uniquely identifiable phenomenological qualities, distinguishable from other feelings of pleasure or displeasure.
In contemplating Hume’s naturalistic account, as outlined above, we immediately perceive a problem. If moral distinctions are the product of “a feeling or sentiment” in our “own breast,”9 then it is open to skepticism in much the same way as other senses. Assuming that an internal moral sense is the source of all our moral judgments, how can we know that this sense does not deceive us? Just as the vision of man suffering from jaundice deceives him as to the proper colour of things, so a skewed moral sense might be misleading. We can easily imagine someone who sees colours in an inverted spectrum, and likewise, we can quite easily imagine a person with a warped or inverted moral sense. From a purely naturalistic perspective, there is nothing inherently contradictory or paradoxical about the existence of someone who feels strong approval in contemplating the basest villainy, and violent disapproval upon beholding the spectacle of warm-hearted benevolence. We can even conceive of a person completely devoid of any moral sense whatsoever. This is not a problem that Hume expressly tackles, but it is never the less one which arises as soon as we seek to construct any normative theory of ethics on the basis of his moral sense account.
In basing his account of morality on an innate moral sense, Hume was inspired by the philosopher Francis Hutcheson. Hutcheson, too, was a Scotsman, and like Hume he regarded morality as having foundations in human nature. Unlike him, however, he was a minister and a religious man, and could therefore make use of means unavailable to Hume. Within a religious framework, a benevolent deity can serve as a firm foothold against relativism: If the moral sense is a natural part of man, it must have been bestowed on him by God. The authenticity and the rightness of the moral sense therefore has divine sanction. God, as a benevolent and omnipotent creator, would not have given man a moral sense that led him much astray from the truth in matters of right and wrong. Even if the moral sense could be warped in some way by experience, it is inherently disposed to be accurate. Just as Descartes maintained that “all our ideas or notions contain in them some truth, for otherwise it could not be that God, who is wholly perfect and veracious, should have placed them in us,”10 so Hutcheson could defend the veracity of his moral sense on similar grounds. If our moral sense tells us that something is wrong, we can infer that it really is wrong, because God would never endow us with a deceiving moral sense. In a similar way, Lord Shaftesbury, the founder of the moral sense tradition, saw man’s moral sense as part of a wider, harmonious system.12 It goes without saying, however, that Hume’s uncompromising naturalism (and, arguably, atheism) precluded such religious and metaphysical ways of establishing the absolute ethical authority of the moral sense. In dispensing with notions of God and harmony, Hume opened to the door to moral skepticism and relativism, for if a moral sense is the ultimate source of morality, then it seems impossible to establish moral facts, except insofar as the person with whom we seek to establish these fact shares our particularly constituted moral sense.
Although Hume’s naturalistic moral sense theory in this way raises the spectre of moral relativism, there are several ways he could have evaded such conclusions. Unfortunately, he thoroughly eliminated these options in his critique of alternative accounts of morality. Take the problem of the inverted or skewed moral sense posed earlier: One way Hume could tackle this problem is to maintain that those whose moral sense does not function in the same way or to the same extent as that of other people are in some sense ‘unnatural.’ In this scheme, ethics get their supposed objectivity from the premise that there is such a thing as a ‘natural’, ‘normal’ moral sense. The ‘unnatural’ morally blind or morally malformed person should be held to the standards and evaluations of the ‘natural’ moral sense which almost everyone else possesses. Ethics are thus established by intersubjective consensus – they are ‘normative’ in an absolutely literal sense of the word. Hume, however, explicitly closes this avenue by saying that it is “impossible … that the character of natural and unnatural can ever … mark the boundaries of vice and virtue.”13 This means that an intersubjective consensus – the ‘natural’ moral sense shared by most – cannot establish ethical truths by being contrasted with the ‘unnatural’. His rejection of the use of ‘natural’ and ‘unnatural’ to distinguish the correctness of moral judgment is made even more explicit in his famous statement that one cannot derive an ought statement from an is statement. Nonmoral premises combined with reasoning cannot lead to the discovery of moral properties. It is always necessary to smuggle in input from the sentiments in order to reach normative conclusions. But if we want to ground ethics on sentiments, we need an answer to ‘Whose sentiments?’, a question which Hume does not even pose, much less answer.
One way to respond to skeptical counter-examples of varying moral senses is to claim that human nature is in fact the same everywhere, and that all human beings are endowed with an identical moral sense which never varies. A belief in the universality of human nature was, after all, one of the more striking characteristics of the Enlightenment philosophy of man.14 This would give any ethical theory a solid grounding in the universal, uniform nature of man. However, it is difficult to imagine the cautious, skeptical and empirically-minded Hume making a factual assertion of such scope. He does not, at any rate, do so explicitly in any of his philosophical works.
As these considerations indicate, Hume’s negative theses narrowly constrain the sort of ethical theory that it is possible to reconcile with his moral psychology. In a way, his treatment of morality bears a certain resemblance to his treatment of the problem of induction. After raising the problem of induction and demonstrating the impossibility of justifying that the future will be anything like the past, Hume proceeded to explain why we in fact employ this method of inference: It is the product of habit, hard-wired into human beings, and nothing more needs to be said. In the case of morality, too, we happen to have a moral sense, also hardwired, which is the source of our moral judgments. In both these cases, we have explanations rather than justifications, as Hume’s analyses turn practical and common-sensical when the resources of philosophical reasoning are exhausted. It may very well be true that “no more may be requisite to the regulation of our conduct and behaviour”15 than the possession of a moral sense. However, this still does nothing to establish any moral facts. It is an explanation of morality, not a justification.
At this stage, it should be clear that Hume's moral theory leads logically to moral relativism and subjectivism. If the innate moral sense is the source of our moral approbation and disapprobation, and there is nothing further to appeal to, then there is no framework by which to determine ethical rightness. The only rational answer to “Why is X wrong?” is “Because I am constituted in such a way as to feel it is wrong,” or possibly “Because I and most (or all) others are thus constituted.” Those in possession of a moral sense that is different to ours exist on a different moral plane, since they make moral judgments that are incommensurable with those of others. That is a clear case of moral relativity.
In conclusion, Hume’s account of morality has some striking strengths. As an assault on the egoistic moral philosophy of Hobbes and Mandeville, it hits the mark. The presence of a natural sympathy in men is a more plausible theory than the “selfish” philosophies, since it can better accommodate observable facts such as human benevolence. Furthermore, his skeptical arguments against reason as the source of morality raised grave problems for ethical systems based on man’s powers of reasoning; these problems still remain very much pertinent today. Viewed, then, as a critique of and substitute for the rationalist and self-interest moral accounts of his time, Hume’s theory of morality is a success. However, if we accept his naturalistic moral sense as the all-encompassing source of morality without supplementing it with some additional, non-subjective source, we are compelled to accept the moral relativism that this entails. Nowhere in his writings does Hume spell out the relativist consequences of his theory, but he must surely have been aware of them. Perhaps he was a moral skeptic, and believed that there was no way to establish objective moral truth. If that was the case, he was certainly much too prudent and cautious a man to have said so overtly.
1. Mackie, J. L., Hume’s Moral Theory (London: Routledge, 1980), p. 82-83.
2. Hume, David, A Treatise of Human Nature (London: Penguin, 1985), p. 510.
3. Hume, Treatise, p. 494
4. Hume, Treatise, p. 520.
5. Hume, Treatise, p. 522.
6. Hume, Treatise, p. 522.
7. Hume, Treatise, p. 520.
8. Hume, David, An Enquiry Concerning the Principles of Morals (Indianapolis: Hackett, 1983), p. 13.
9. Hume, Treatise, p. 520.
10. Descartes, René, Discourse on Method (London, 1946).
11. It is worth noting, however, that basing the rightness of the moral sense on the benevolence of God raises in turn a whole host of problematic theological and ethical questions.
12. Mackie, p. 14.
13. Hume, Treatise, p. 527.
14. Berlin, Isaiah, The Age of Enlightenment: The Eighteenth-Century Philosophers (Oxford: Oxford University Press, 1956).
15. Hume, Treatise, p. 521.
Baillie, James, Hume on Morality (London: Routledge, 2000).
Berlin, Isaiah, The Age of Enlightenment: The Eighteenth-Century Philosophers (Oxford: Oxford University Press, 1956)
Cohon, Rachel (ed.), Hume: Moral and Political Philosophy (Aldershot, England and Burlington: Ashgate, 2001).
Cohon, Rachel, "Hume's Moral Philosophy", The Stanford Encyclopedia of Philosophy (Winter 2004 Edition), Edward N. Zalta (ed.), URL =
Descartes, René, Discourse on Method (London, 1946).
Hume, David, A Treatise of Human Nature (London: Penguin, 1985).
Hume, David, An Enquiry Concerning the Principles of Morals (Indianapolis: Hackett, 1983).
Mackie, J. L., Hume’s Moral Theory (London: Routledge, 1980).