Jump to ContentJump to Main Navigation
On Custom in the Economy$

Ekkehart Schlicht

Print publication date: 1998

Print ISBN-13: 9780198292241

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/0198292244.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 26 February 2017

(p.289) Appendix C Hume's Law

(p.289) Appendix C Hume's Law

On Custom in the Economy

Ekkehart Schlicht

Oxford University Press

Abstract and Keywords

‘Hume's Law’, as commonly interpreted, states that ethical judgement cannot be derived from factual judgement. This position would lead to ethical relativism. It is argued that this interpretation is wrong and that factual judgement and ethical judgement are tied together.

Keywords:   ethical relativism, Hume's law, judgement, moral sense, naturalistic fallacy

C.1 Surprises With Hume

Hume's Law Surprises With Hume

The account of Hume's theory of property may have surprised psychologists because of its stress on cognition.1 It may have also surprised philosophers and social scientists. Some of them would be inclined to think that ‘Hume's Law’, so familiar to them, and typically associated with ethical relativism, is somewhat at odds with the universalist stance taken by Hume with regard to the foundation of property. They would also find it at variance with his simultaneous stress on property being intimately related to cognitive processes. They are right. ‘Hume's Law’, as usually interpreted, entails a view about an independence of cognition and valuation that runs counter to Hume's ideas. In this sense, the usual interpretation of Hume's Law is misleading.

The purpose of this appendix is to discuss the relationship between ethical relativism, Hume's Law, and the position taken by Hume on ethical issues, which is universalist.

C.2 Ethical Relativism

Hume's Law Ethical Relativism

The position of ethical relativism asserts a fundamental arbitrariness of ethical valuations. Ethical relativism is usually confounded with cultural relativism by virtue of the argument that ethical valuations are determined by culture. Yet cultural relativism and ethical relativism are formally separate issues. The reason is that a social scientist who believes in the possibility of a positive universalist social science may still consider ethical valuations to be independent of factual judgements. He will thus see them as ultimately arbitrary and culturally contingent.2

The position of ethical relativism can be defended along two lines. The first defence is to treat ethical relativism as a special case of cultural relativism; the second is to assert the independence of ethical from factual judgements. The first line of argument has been discussed earlier and will not be dealt with here. The second line of argument is based on Hume's Law.

(p.290) C.3 Two Versions of Hume's Law

Hume's Law Two Versions of Hume's Law

There is no explicit statement of the law by Hume himself. The conventional view of the law is well captured by the following quotation, taken from the Oxford Dictionary of Philosophy:

Hume's law. A name for the contested view that it is impossible to derive an ‘ought’ from an ‘is’, or in other words, that there is no logical bridge over the gap between fact and value. Hume's own statement is in the Treatise of Human Nature, iii. 1.1, where he wrote that it seems ‘altogether inconceivable that this new relation [ought] can be derived from others, which are entirely different from it’. The ‘law’ in fact appears as something of an afterthought to other discussion.1

Another dictionary of philosophy is more decisive:

Hume's law. The increasingly popular label for Hume's insistence that the ‘naturalistic fallacy’ is indeed a fallacy; and hence that conclusions about what ought to be cannot be deduced from premises stating only what was, what is, or what will be—and the other way round.2

A brief statement of the first version of Hume's Law is as follows:

The Loose Law: An ought cannot be derived from an is.

This ‘law’ can, however, be further elaborated. Granted that an ought cannot be derived from an is, it is still possible to derive an ought from an is in conjunction with another ought. The normative statement ‘You ought to cure your cough’ and the positive statement ‘Your cough will disappear if you keep warm’ imply the normative statement ‘You ought to keep warm.’ The last statement is a derived valuation.3 Any value that is not derived is termed basic. This gives rise to the second formulation of Hume's Law:

The Tight Law: Any value is either basic or derived.

These laws look very similar. Both are consistent. They differ only in that the Tight Law introduces a distinction between basic and derived values that is not to be found in the Loose Law. This distinction is to be found in Hume, although he has not used the terms ‘basic’ and ‘derived’.4

(p.291) C.4 The Loose Law, the Tight Law, and Ethical Relativism

Hume's Law The Loose Law, the Tight Law, and Ethical Relativism

Consider how these versions of Hume's Law relate to the issue of ethical relativism.1

The Loose Law suggests the arbitrariness of values, because values cannot be derived from factual judgements. This is precisely the position of ethical relativism. A relativistic reading would take the law as vindicating all kinds of ethical relativism. There would be no factual basis for universal human rights. The very idea of rights, morality, good, and bad would appear arbitrary, a matter of taste or cultural contingency.

A similar conclusion seems to emerge from the Tight Law. As basic values, by definition, are not justifiable any further, they are arbitrary and non‐rational. In so far as the law implies the existence of basic values, it would suggest the relativistic interpretation. Yet this is not the case.

The Tight Law says nothing about the existence of basic values. This may be clarified by a formally identical argument about numbers. We may define a natural number as derived if it can be obtained by taking the difference of two other numbers, and as basic if it is not derived. It is true that any number is either basic or derived, but all numbers are derived, and none is basic. Each number is based on another one. For that reason, there exists no basic number. Moral values may formally be related in the same way. All values may be based on others, without any basic value being required or implied.

The tight interpretation does not imply any kind of ethical relativism. This can be seen from an argument in analogy to the argument about basic and derived numbers. Assume that there are two sets of mutually exclusive values, say A, B, C and a, b, c. The value A is assumed to be incompatible with a, in the sense of being ‘just the opposite’. This is denoted by (A) ⊥ a, which reads ‘A is incompatible with a’. Under a similar assumption for the other valuations, it is assumed that

( A ) a , ( B ) b , ( C ) c
holds true.

Assume now that there is a set of factual knowledge, known to be true. Together with this knowledge, each value judgement can be derived from (p.292) others. As a matter of notation, (A, B) → C denotes that ‘factual knowledge allows the derivation of value C from values A and B.’

Assume the following inferences to be possible in this way:

( B , c ) A , ( a ) B , ( A , b ) C , ( C , b ) a , ( c ) b , ( a , B ) c .
All values A, B, C, a, b, c are derived values. They can be derived from other values by means of factual knowledge. Yet the set of valuations is uniquely determined as (A,B,C). This can be seen as follows. Note first that this set of norms violates none of the above implications. So it is a possible system of values, given the factual knowledge. But what about other norm systems like (A, b, c) that could be obtained by combining other alternatives? They can all be ruled out. Consider each of the alternative valuations. If a were right, B would be right. This in turn would imply c to be right. B and c imply A—a contradiction, since a and A are incompatible. So a is wrong and A is right. Assume now that b is right. As A is right, this would imply C to be right, and this would imply a to be right. This cannot be the case, because A has been shown to be right. So B is right, and b is wrong. Assume now that c is right. This implies b to be right, but this cannot be the case, because B has been shown to be right. This proves that the set of values (A, B, C) is uniquely determined by the factual knowledge available.

Thus, the Tight Law does not imply ethical relativism. To be precise, it does not imply anything concerning the issue of ethical relativism. It does not rule out a multiplicity of possible value systems, it does not rule out uniqueness, and it does not rule out the non‐existence that may come about if no value system can be found that is consistent with the postulated inferences.

The above example has shown that the Tight Law is compatible with uniqueness. In this sense the Loose Law, although compatible with the Tight Law, is misleading. The following, after all, is implied by both versions:

First implication: It may be the case that an ought can be uniquely determined from an is under the assumption that an ought exists. This does not presuppose the existence of basic values.

This seems to be a better version than the Loose Law because it does not suggest, incorrectly, that valuations must be arbitrary. There is no reason to make such a contention about values. However, another implication can be obtained from this discussion. That implication is indeed very universalist in spirit: (p.293)

Second implication: Additional knowledge may reduce the set of possible values, but will never increase it.

This implication stresses the dependence of values on knowledge, rather than stating that valuations are arbitrary. Thus, even if it is assumed that Hume's Law is correct, the law certainly cannot be used as a defence of any kind of ethical relativism.

C.5 The Existence of an Ought

Hume's Law The Existence of an Ought

Hume's concern, however, went in a quite different direction. The first section of book III, part I of the Treatise is entitled ‘Moral distinctions not deriv'd from reason’. It is followed by a second section, ‘Moral distinctions deriv'd from a moral sense’. The usual discussion of Hume's Law relates to the first section and neglects the second. The omission leads to misunderstanding. It neglects the way in which the moral sense relates to reason. The ‘law’ is actually a consequence of the way in which Hume distinguishes ‘reason’ and ‘moral sense’. The law is an implication of a theoretical distinction drawn between reason and moral sense—nothing more, nothing less.

Hume would accept both implications. He has actually dealt with the possibility of uniqueness mentioned in the first implication:

In order, therefore, to prove that the measures of right or wrong are eternal laws, obligatory on every rational mind, ‘tis not sufficient to shew the relations upon which they are founded: We must also point out the connexion with the will; and must prove that this connexion is so necessary, that in every well‐disposed mind, it must take place and have its influence.1

This highlights what Hume's discussion is aiming at. Right and wrong are not external to human beings and derivable more geometrico: they relate to the will and are active forces. Reason is passive:

The merit and demerit of actions frequently contradict, and sometimes controul our natural propensities. But reason has no such influence . . . [It] is the discovery of truth and falshood. . . . Moral distinctions, therefore, are not the offspring of reason. Reason is wholly inactive, and can never be the source of so active a principle as conscience, or a sense of morals.2

Therefore, reason cannot establish the existence of an ought. The problem is not that of a relativism of values, but that of existence. The existence of an ought cannot be derived from judgements about truth and (p.294) falsehood. Yet reason can contribute to establishing what possible oughts there might be.

This question about the existence of an ought is quite different from the problem about the contents of the ought. Usual discussions, however, refer typically to the latter. A correct version of Hume's Law would read:

The existence of an ought cannot be inferred from a judgement about the true and the false. This is not necessarily the case with the contents of the ought.

In this sense, the usual interpretation given to Hume's Law is wrong, or at least severely misleading. It is especially misleading in its relativistic interpretation, since the problem of the existence of an ought is phrased in a universalistic way. The very question presupposes the ought to be not arbitrary.

C.6 The Moral Sense

Hume's Law The Moral Sense

The interpretation of Hume's Law given above follows modern deductive reasoning, but runs against Hume's view of the way in which moral judgements are formed. He makes the argument that the existence of an ought cannot be derived from reason in order to establish the existence of a ‘moral sense’. In this, he aims to relate moral judgement to psychology. It is interesting that this idea has not been taken up by either philosophy or psychology.1 It would not be appropriate to give a comprehensive account here.2 A sketch of Hume's ethical theory must suffice.

Reason cannot give rise to an ought. Hume concludes from this that moral judgements are brought about by ‘impressions’ arising from the individual's ‘moral sense’ in a direct way:

Thus the course of the argument leads us to conclude, that since vice and virtue are not discoverable merely by reason, or the comparison of ideas, it must be by means of some impression or sentiment they occasion, that we are able to mark the difference betwixt them. . . . The next question is, Of what nature are these impressions, and after what manner do they operate upon us? Here we cannot remain long in suspense, but must pronounce the impression arising from virtue, to be agreeable, and that proceeding from vice to be uneasy.3

Yet these impressions are not just impressions of pleasure or uneasiness; they are of a ‘peculiar kind’: (p.295)

An action, or sentiment, or character is virtuous or vicious; why? because its view causes a pleasure or uneasiness of a particular kind. In giving a reason, therefore, for the pleasure or uneasiness, we sufficiently explain the vice or virtue. To have a sense of virtue is nothing but to feel a satisfaction of a particular kind from the contemplation of a character. The very feeling constitutes our praise or admiration. We go no farther; we do not enquire into the cause of the satisfaction. We do not infer a character to be virtuous, because it pleases: But in feeling that it pleases after such a particular manner, we in effect feel that it is virtuous. The case is the same as in our judgements concerning all kinds of beauty, and tastes, and sensations. Our approbation is imply'd in the immediate pleasure they convey to us.1

Pleasure (or, in today's terminology, utility) is, however, a term that comprises many different impressions:

A good composition of music and a bottle of wine equally produce pleasure; and what is more, their goodness is determin'd merely by pleasure. But shall we say upon that account, that the wine is harmonious, or the music of a good flavor? . . . the satisfaction is different . . . Nor is every sentiment of pleasure and pain of that peculiar kind, which makes us praise and condemn.2

The particular kind of sentiment relates to a detached view of events, later formulated by Adam Smith as the point of view taken by an ‘impartial spectator’. Hume phrased this as follows:

The good qualities of an enemy are hurtful to us; but may still command our esteem and respect. ‘Tis only when a character is considered in general, without reference to our particular interest, that it causes such feelings or sentiment, as denominates it morally good or evil.3

This way of forming moral judgements is deeply rooted in human nature.

These sentiments are so rooted in our constitution and temper, that without entirely confounding the human mind by disease or madness, ‘tis impossible to extirpate and destroy them.4

Moral sentiment relates, further, not to single features of an action, but to overall evaluations of an entire pattern, just like perceptions of beauty. ‘The same principles produce, in many instances, our sentiments of morals, as well as those of beauty.’5 In that, it relates directly to the tendency towards overall clarification discussed above:6 (p.296)

Attend to Palladio and Perrault, while they explain all the parts and proportions of a pillar. They talk of the cornice, and frieze, and base, and entabulature, and shaft, and architrave; and give the impression and position of these members. But should you ask the description and position of its beauty, they would readily reply, that the beauty is not in any of the parts or members of a pillar, but results from the whole, when that complicated figure is presented to an intelligent mind, susceptible to those finer sensations.1

The perception of morality arises from the human constitution, like the perception of ‘sounds, colours, heat, or cold’.2 The impressions are ‘involuntary and necessary.’3 They are universal:

Tho' the rules of justice be artificial, they are not arbitrary. Nor is the expression improper to call them the Laws of Nature; if by natural we understand what is common to any species, or even if we confine it to mean what is inseparable from the species.4

It is, thus, perfectly clear that Hume cannot be considered a defender of any kind of ethical relativism.

C.7 An Afterthought

Hume's Law An Afterthought

Another criticism of Hume's Law can be formulated as follows.

Suppose I have developed a positive theory of ethical judgements which describes and predicts how ethical judgements flow from impressions, tastes, and deliberations. Such a theory must be, by necessity, both positive and normative, in the sense that my own ethical judgements must follow the rules established in my positive ethical theory; otherwise, my positive theory would be false. Thus, a purely positive theory, if true, (p.297) rules out ethical judgements that do not conform to the theory. A positive theory, therefore, carries normative implications.

In conclusion, there are strong reasons for challenging the validity of what became erroneously known as ‘Hume's Law’ and what seems incompatible with Hume's universalism and his writings.


(1) See Ch. 11.

(2) This position is akin to that entertained by many economists. The logical independence of cultural and ethical relativism has been stressed by Herskovits (1958: 267). On cultural relativism, see App. B.

(1) S. Blackburn (1994: 180).

(2) Flew (1975: 146). It must be mentioned that the entry in the same dictionary on the ‘naturalistic fallacy’ is less simple‐minded.

(3) This is in conformity with Hume (1740a: 459, 462); for the distinction between basic and derived values, see also Sen (1970: 59–64).

(4) Hume (1740a: 459, 462).

(1) The following argument is very simple and straightforward but is not well known. It is related to Albert's (1968: ch. 12) discussion of ‘Brückenprinzipien’. Recent contributions on the topic, e.g. Sen (1970: 59–64) and Sugden (1986: 153–4; 175) seem to misrepresent the case. The issue is of some importance for the problem of cultural relativism; therefore I state it here again. (An earlier statement is in Schlicht 1974.) For a good discussion of cultural relativism, see Asch (1952).

(1) Hume (1740a: 465).

(2) Hume (1740a: 458).

(1) The Gestalt psychologists Wertheimer (1935) and Köhler (1938) are exceptions.

(2) For an excellent introduction to Hume, see MacNabb (1967). This article contains an annotated list of critical works on different aspects of Hume's philosophy, including his moral philosophy. The point I am going to make—assimilating Wertheimer (1935) and Hume (1777a)—is somewhat different from the issues discussed in that literature.

(3) Hume (1740a: 470).

(1) Hume (1740a: 471).

(2) Hume (1740a: 472).

(3) Hume (1740a: 472). For Smith's theory involving an impartial spectator, see Smith (1759).

(4) Hume (1740a: 474).

(5) Hume (1740a: 577).

(6) See S. 8.1.

(1) Hume (1777a: 292). This later view, not to be found in the Treatise, may be seen as a development of the relation between beauty and virtue, as elaborated in the Treatise. It is reminiscent of Wertheimer's (1935) view of morality. Wertheimer (1934) would insist, however, on treating logic and morality as being generated by the same principles of ‘Prägnanz’, ‘requiredness’—or, as I say, ‘clarity’. I agree with Wertheimer (see S. 8.1). This does not necessarily imply any substantive disagreement with Hume. I simply think that the theoretical distinction he has drawn for his own philosophical purposes may lead to severe misunderstandings and should be followed with care.

(2) Hume (1740a: 469). These are ‘not qualities of objects, but perceptions of the mind’. The extreme subjectivism expressed here thus pertains not only to morality, but to reason as well: ‘Reason is nothing but a wonderful and unintelligible instinct in our souls, which carries us along a certain train of ideas, and endows them with particular qualities, according to their particular situations and relations’ (Hume 1740a: 179). This does not justify cultural relativism either. Those who read Hume in a relativistic manner should be careful enough to declare that the entire world—including their own existence and their own theoretical statements—is as arbitrary as they declare morality to be. I would not object.

(3) Hume (1740a: 608).

(4) Hume (1740a: 484).