Hostname: page-component-cd9895bd7-mkpzs Total loading time: 0 Render date: 2024-12-26T04:49:53.018Z Has data issue: false hasContentIssue false
  • English
  • Français

J. S. Mill's Language of Pleasures*

Published online by Cambridge University Press:  26 January 2009

Robert W. Hoag
Get access Rights & Permissions [Opens in a new window]

Extract

A significant feature of John Stuart Mill's moral theory is the introduction of qualitative differences as relevant to the comparative value of pleasures. Despite its significance, Mill presents his doctrine of qualities of pleasures in only a few paragraphs in the second chapter of Utilitarianism, where he begins the brief discussion by saying:

utilitarian writers in general have placed the superiority of mental over bodily pleasures chiefly … in their circumstantial advantages rather than in their intrinsic nature.… [B]ut they might have taken the … higher ground with entire consistency. It is quite compatible with the principle of utility to recognize the fact, that some kinds of pleasure are more desirable and more valuable than others. It would be absurd that while, in estimating all other things, quality is considered as well as quantity, the estimation of pleasures should be supposed to depend on quantity alone (U, II, 4).

Type
Research Article
Information
Utilitas , Volume 4 , Issue 2 , November 1992 , pp. 247 - 278
Copyright
Copyright © Cambridge University Press 1992

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Footnotes

*

Work on this project has been supported by a Newcombe Fellowship from the Woodrow Wilson National Fellowship Foundation, a stipend from the National Endowment for the Humanities, and a sabbatical leave granted by Berea College. I also thank Joel Feinberg, John Simmons, John Marshall, Alan Fuchs, David Gilbor and Eric Pearson for helpful comments on earlier versions of this paper.

References

1 All quotations of Mill use editions published in The Collected Works of John Stuart Mill (Toronto). References are provided parenthetically as follows: to A System of Logic by volume (vii or viii), and page number of the Collected Works edition, and to Utilitarianism by chapter and paragraph number following ‘U’.

2 Schneewind, J. B., ‘Concerning Some Criticisms of Mill's Utilitarianism, 1861–1876’, James and John Stuart Mill/Papers of the Centenary Conference, ed. Robson, John M. and Laine, Michael, Toronto, 1976, p. 52Google Scholar; Sidgwick's Ethics and Victorian Moral Philosophy, Oxford, 1977, pp. 185–6.Google Scholar

3 The Methods of Ethics, 7th edition, London, 1907, pp. 94–5.Google Scholar

4 Prolegomena to Ethics, ed. Bradley, A. C., Oxford, 1899, pp. 185–201.Google Scholar

5 Ethical Studies, 2nd edition, Oxford, 1926, pp. 117–20.Google Scholar

8 Principia Ethica, Cambridge, 1903, pp. 77–81.Google Scholar

7 Representative examples include the following critics: Anschutz, R. P., The Philosophy of J. S. Mill, Oxford, 1953, pp. 18–19Google Scholar; McCloskey, H. J., John Stuart Mill: A Critical Study, New York, 1971, pp. 64–71CrossRefGoogle Scholar; Plamenatz, John, The English Utilitarians, Oxford, 1949, pp. 141–2Google Scholar; Quinton, Anthony, Utilitarian Ethics, New York, 1973, pp. 42–3CrossRefGoogle Scholar. Defenders of Mill include: Dahl, Norman, ‘Is Mill's Hedonism Inconsistent?’, Studies in Ethics, American Philosophical Quarterly Monograph Series, vii (1973), 37–54Google Scholar; Dryer, D. P., ‘Mill's Utilitarianism’, in Essays on Ethics, Religion, and Society, ed. Robson, John M., Toronto, 1969Google Scholar, Collected Works of John Stuart Mill, x. pp. lxxxviii–xciiGoogle Scholar; Martin, Rex, ‘A Defence of Mill's Qualitative Hedonism’, Philosophy, xlvii (1972), 140–51CrossRefGoogle Scholar; Raphael, D. D., ‘Fallacies In and About Mill's Utilitarianism’, in Mill's Utilitarianism, ed. Sosa, Ernest and Smith, James, Belmont, 1969, pp. 161–72Google Scholar; West, Henry R., ‘Mill's Qualitative Hedonism’, Philosophy, li (1976), 97–101.CrossRefGoogle Scholar

8 Berger, Fred R., Happiness, Justice, and Freedom: The Moral and Political Philosophy of John Stuart Mill, Berkeley, 1984, esp. chap. 2Google Scholar; and, Gray, John, Mill on Liberty: A Defence, London, 1983, esp. pp. 42–8 and 70–3CrossRefGoogle Scholar. See also Hoag, Robert W., ‘Happiness and Freedom: Recent Work on John Stuart Mill’, Philosophy and Public Affairs, xv (1986), 188–99Google Scholar; and ‘Mill's Conception of Happiness as an Inclusive End’, Journal of the History of Philosophy, xxv (1987), 417–31.CrossRefGoogle Scholar

9 Berger, , p. 38.Google Scholar

10 Ibid., pp. 38–40.

11 Ibid., pp. 71–3.

12 Skorupski, John, John Stuart Mill, London, 1989, p. 305.Google Scholar

13 Gray, , p. 71.Google Scholar

14 Ibid., p. 305, note 12.

15 Not only is Mill silent in Utilitarianism, but, despite his contemporaries' criticisms (see footnote 2), Mill never defends or clarifies his doctrine of higher pleasures. Sumner, L. W., ‘More Light on the Later Mill’, Philosophical Review, lxxxiii (1974), 504–27CrossRefGoogle Scholar; Schneewind, , Sidgwick's Ethics, p. 178.Google Scholar

16 This is too seldom noted about Bentham. In An Introduction to the Principles of Morals and Legislation, ed. Burns, J. H. and Hart, H. L. A., London, 1970Google Scholar (The Collected Works of Jeremy Bentham), Bentham speaks of pleasures' quantity and quality (e.g., ch. 3, §.x), he classifies pleasures into kinds (ch. 5), and he considers both quantity and quality as properties relevant to determining the lot of punishments (e.g., ch. 15, §.i).

17 In addition to Bentham (note 16), Plato, Aristotle, Spinoza, and Hutcheson use similar language in talking about pleasures. See Edwards, Rem B., Pleasures and Pains: A Theory of Qualitative Hedonism, Ithaca, 1979, pp. 70–2, 82–92Google Scholar; and Gibbs, Benjamin, ‘Higher and Lower Pleasures’, Philosophy, lxi (1986), 31–41.CrossRefGoogle Scholar

18 This interpretive approach is justified by several considerations. Logic has now often been shown to be fundamental and central to Mill's philosophical system: Skorupski, , p. xiiiGoogle Scholar; Ryan, Alan, The Philosophy of John Stuart Mill, 2nd edition, Atlantic Highlands, 1987, p. ixCrossRefGoogle Scholar. Mill himself saw the project of the Logic as important, in part, for its significant moral and political implications: Skorupski, , pp. 30–1Google Scholar; Ryan, , pp. ix–xGoogle Scholar. Indeed, the culminating chapter of Logic explicitly refers to Utilitarianism for ‘an express discussion and vindication’ of happiness as the ultimate end of life (viii. 951n); and in defending his hedonic account of desires in Utilitarianism (IV, 11)Google Scholar Mill explicitly refers to the Logic discussion of free agency (viii. 836–43)Google Scholar. Furthermore, accounts in Logic have been useful in understanding other aspects of Mill's moral theory, such as moral rules and his formal conception of happiness. See, for example, Berger, , pp. 86–120Google Scholar; Gray, , pp. 19–42CrossRefGoogle Scholar; Hoag, , ‘Mill's Conception of Happiness’, 418–24.Google Scholar

19 These reasons and the importance of the account of names are detailed in Skorupski, chs. 2–6.

20 Mill's logic antedates Frege: it is syllogistic and all propositions are of subject-predicate form. Skorupski, , pp. 50, 86.Google Scholar

21 Mill does entertain the possibility that some names do not, in fact, denote anything; however, he does not explore the implications of this possibility (Skorupski, , pp. 50, 55Google Scholar). Mill analyses the meaning of propositions in terms of the semantic content of constituent names, thereby departing from Bentham's use of paraphrasis, or contextual definition. Bentham, unlike Mill, sees propositions or sentences to be the fundamental semantic unit. This important difference in approaches to semantic theory may explain Mill's ignoring paraphrasis as a way of dealing with non-denoting names. For a helpful discussion of the importance of Bentham's theory of meaning, including paraphrasis, see Harrison, Ross, Bentham, London, 1983, esp. pp. 53–74.Google Scholar

22 Mill's use of ‘feeling’ parallels Descartes's introduction of cogito (think) as a technical term to designate diverse sorts of mental states: doubting, understanding, affirming, denying, willing, imagining, desiring, and sensory perceptions. Meditations on First Philosophy, trans. Cottingham, John, Cambridge, 1986, p. 19Google Scholar. I use capitalization, ‘Feeling’, to indicate use of Mill's broad, technical term.

23 Although Mill clearly classifies both pains and pleasures as Feelings, I restrict my attention to pleasures only. It is not correct to assume that what is true of pleasures applies mutatis mutandis to pains, and Mill explicitly recognizes this. He does not even discuss quality or quantity of pains in Utilitarianism: indeed, he hardly mentions pains at all, but does say ‘pain is always heterogeneous with pleasure’ (U, II, 8). In Logic Mill denies that pains are merely the absense of pleasure or that pains and pleasures are positive and negative ‘values’ of some common dimension: The word “unpleasant”, notwithstanding its negative form, does not connote the mere absence of pleasantness, but a less degree of what is signified by the word ‘painful’, which, it is hardly necessary to say, is positive’ (vii. 42). Scepticism about the mutatis mutandis claim is also supported by Bentham's elaborate classification of pleasures and pains (IPML, (CW), ch. 5): some sorts of pleasures presuppose or require pains (i.e., the pleasures of relief experienced when a pain ceases or abates (§.xvi)), some ‘pains are grounded upon pleasures’ (i.e., pains of privation resulting from the thought of not possessing any of the various sort of pleasures (§§.xvi–xx)), and some pleasures are without any corresponding pain (e.g., those of novelty (§.xxii, fn. 1)).

24 Utilitarianism, unlike Logic, is an apologetic written for a popular publication, Eraser's Magazine. Thus, Mill here uses ‘feeling’ in a non-technical, popular sense to signify emotions (vii. 51). The context usually makes clear whether Mill is using the technical term.

25 Another indication of Mill's broad, technical use of ‘pleasure’ in Utilitarianism is apparent from his hedonic theory of desires. Recent scholarship shows that, contrary to traditional views, Mill does not hold that pleasures and only pleasures are the object of all desires (Berger, , pp. 12–16, 32–5Google Scholar; Hoag, , ‘Mill's Conception of Happiness’, 425–6Google Scholar; ‘Happiness and Freedom’, 189–90Google Scholar; Skorupski, , pp. 295–301Google Scholar). Furthermore, in the fourth chapter of Utilitarianism, Mill says:

Those who desire virtue for its own sake desire it either because the consciousness of it is a pleasure, or because the consciousness of being without it is a pain, or for both reasons united

to desire anything except in proportion as the idea of it is pleasant is a physical and metaphysical impossibility (U, IV, 8, 10; emphasis added).

Although there is disagreement about Mill's precise view (see, for example, the exchange between Skorupski, and Berger, in Philosophical Books, xxvi (1985), 194–9, 202–6Google Scholar), both passages presuppose the possibility that the consciousness or idea of something can itself be a pleasure: some ideas are pleasures. This reflects the view advanced in Logic: for Mill, pleasures are not only sensations or emotions, but include ideas and other species of conscious states. The broad use of ‘pleasure’ suggested in the Logic appears to play a role in both Mill's normative standard and his moral psychology.

26 This reflects an empiricist tradition. John Locke includes pleasures and pains among the simple ideas of both sensation and reflection, the latter being ideas the mind has from observing its own actions, from turning ‘its view inward upon itself’ (An Essay Concerning Human Understanding, London, 1706, Book I, chapters 6, 7Google Scholar). And Berkeley, who Mill so admired, insisted that one is aware of one's own mind and one's own mental activities Dialogues, iii, 4–5Google Scholar). Mill's own views about self-knowledge are not developed in Logic; however, Mill is puzzled when he later turns to the topic in chapter 12 of An Examination of Sir William Hamilton's Philosophy. See Skorupski, , ch. 7, esp. pp. 236–8Google Scholar; and Ryan, ch. 6.

27 I characterize these pleasures as associated with exercising one's mental faculties. First, the language of ‘faculties’ is odd with respect to emotions and moral sentiments. But Mill clearly links emotions with the activity of the imagination, and he sometimes speaks of a moral capacity or moral faculty (e.g., U, HI, 8). Second, I use the vague expression, ‘associated with’, to express the relationship between activities and pleasures. Mill does not specify the relationship, and he does not consistently describe the relationship as, for example, a causal one. It is possible, for example, that the relationship is a complex one of the sort defended by Aristotle in Books VII and X of the Nichomachean Ethics. Mill's precise understanding of the relationship is not important to my aims here.

28 Warner, Richard, Freedom, Enjoyment, and Happiness: An Essay on Moral Psychology, Ithaca, 1986, pp. 119–20Google Scholar. This is also a feature of ordinary usage: see Perry, David L., The Concept of Pleasure, The Hague, 1967, pp. 61–3.Google Scholar

29 Once one abandons ‘the project of identifying pleasure with a certain kind of mental element, there is no reason not to take the most liberal alternative and consider the quality of pleasantness attachable to any sort of conscious state’ (Alston, W. P., ‘Pleasure’, Encyclopedia of Philosophy, vi. 343Google Scholar). It is not clear how far to take this suggestion. At one point Mill seems to assume that a ‘manner of existence’, a way of life, can be a pleasure. After stating a criterion for estimating two pleasures (U, II, 5), he immediately applies this criterion to establish that ‘the manner of existence which employs… higher faculties’ is superior to the life of ‘the fool, the dunce, or the rascal’ (U, II, 6). Thus, Mill apparently holds that something as extended as a way of life can be called a ‘pleasure’.

30 IPML (CW), ch. 5.

31 Mental state accounts contrast with views of utility as a state of the world that fulfils a desire or preference. Griffin, James, Weil-Being: Its Meaning, Measurement and Importance, Oxford, 1986, pp. 7–8Google Scholar. Other terminology is sometimes used to focus on a similar distinction: for example, Alston refers to mental state accounts of pleasure as ‘conscious-quality theories’, contrasting them with ‘motivational theories’ (pp. 341–7); and in Pleasure and Desire: The Case for Hedonism Reviewed, Oxford, 1969Google Scholar, J. C. B. Gosling contrasts mental state accounts with ‘adverbial views’ of pleasure (ch.4).

32 Contemporary discussions often adopt this requirement. See, for example, Warner, , p. 120Google Scholar; Griffin, , pp. 13–20Google Scholar; Glover, Jonathan, Causing Death and Saving Lives, New York, 1977, pp. 63–5Google Scholar. Mill's experience requirement is a strong one, though, for he identifies the pleasure with that which is immediately experienced, a mental state. But this stronger version of the Experience Requirement follows from Mill's metaphysics more than from the Experience Requirement per se: one can accept the Experience Requirement without reifying, as mental entities, the immediate objects of that experience.

33 Such examples are briefly discussed by Warner, , p. 120.Google Scholar

34 Berger, , p. 305, note 12.Google Scholar

36 In Logic Mill says Feelings, and thus, pleasures, must be part of a mind's experience (vii. 51). So either animals have minds or, contrary to Utilitarianism, animals are not capable of even the pleasures of mere sensation. Given that Mill regards a mind as ‘an unknown recipient or percipient… of sensation’, as a capacity for being conscious (vii. 63–4), the former is a plausible interpretation. It is a ‘popular perversion’ to restrict ‘mind’ to intellect (vii. 51–2). Humans and animals are distinguished by different mental capacities, not by the absence or presence of consciousness or of a mind itself; indeed, ‘human creatures’ sometimes retain only the capacity for an animal's pleasures, those of sensation (U, II, 7).

36 For discussions of these relational accounts of pleasure, see Gosling, , pp. 54–85Google Scholar; and Ryle, Gilbert, The Concept of Mind, London, 1949, p. 132.Google Scholar

37 Mental state and motivational accounts are not mutually exclusive. Most motivational accounts analyse pleasures in terms of a relation where one place is something other than a mental state. But accounts of pleasure can be distinguished with respect to their logical structure (e.g., pleasure analysed as predicate or relation) independently of their metaphysical status (e.g., mental states or states of the world).

38 Such mental state accounts of pleasure are properly and influentially criticized by Ryle, , pp. 107–9Google Scholar, and Gosling, , pp. 28–53.Google Scholar

39 Griffin, , p. 8Google Scholar; Alston, , pp. 341–3.Google Scholar

40 The qualification excludes considering individual names. As Mill acknowledges in an early draft of Logic (viii. ‘Appendix A’, 984)Google Scholar, it may be possible to signify a pleasure by definite description or proper name. But in published editions of Logic this possibility is not explicitly recognized, and at one point Mill implies that all names of feelings are concrete general names (vii. 104). Furthermore, since the doctrine of qualities of pleasures is expressed using general names, it is not necessary here to address complexities arising from Mill's treatment of individual names. For a discussion of these complexities, see Skorupski, , pp. 54, 67–9.Google Scholar

41 As Mill himself explains, he uses ‘abstract’ and ‘concrete’ ‘in the sense annexed to them by the schoolmen’, not in the modern sense, exemplified by Locke, whereby ‘abstract name’ is applicable to all general names (vii. 29).

42 Ibid., p. 78.

43 See discussions listed in notes 2–7 above. Even those not concerned with the traditional consistency question have also made this assumption. See, for example, Edwards, , pp. 33–4.Google Scholar

44 Berger suggests this usage is confused: p. 305, note 12.

45 The example is from Gosling, , p. 28Google Scholar. The accompaniment view is notably criticized by Ryle, chapter 4.

46 Mill's account has phenomenological support. See, for example, Duncker, Karl, ‘On Pleasure, Emotion, and Striving’, Philosophy and Phenomenological Research, i (1940), 389–101.Google Scholar

47 Skorupski, , pp. 52–3, 95–8Google Scholar. This metaphysical point holds even if, as Skorupski argues (pp. 63–7), Mill's understanding of attributes is inadequate for semantic theory.

48 Skorupski, , pp. 94, 208.Google Scholar

49 Recent, revisionist scholarship shows that the doctrine of qualities of pleasures is central to Mill's moral theory. But Mill seldom uses the language of pleasures in his moral and political writings. In Utilitarianism pleasures are discussed in two places: the doctrine of qualities of pleasures (U, II, 1–10) and in conjunction with the proof of the principle of utility (U, IV). Except for these two occasions, ‘happiness’ is used almost exclusively in the first four chapters of Utilitarianism; in the fifth chapter on justice and in On Liberty, Mill uses ‘well-being’.

50 See note 15 above.

51 As even the brief texts above suggest, Mill sometimes suggests that all names of any feelings (simple or complex) connote only an unanalysable resemblance, and that names of objects are those which connote an analysable resemblance (vii. 70–1, 102–3). It does seem, however, implausible to hold that names of complex feelings are indefinable, especially given Mill's phenomenalistic analysis of the language of objects (vii. 70–1).

52 ‘Natural Kinds’, Ontological Relativity and Other Essays, New York, 1969, pp. 116, 119, 121Google Scholar. Note that Quine uses ‘kinds’ here in a broader sense than Mill, for Mill denies that all sortings (classes) are also kinds (see Section V below).

53 Ibid., pp. 116, 121–31.

54 Ibid., pp. 121, 123.

55 In effect, Mill holds that, with respect to simple Feelings, the name of the connoted attribute is an abstract name that does not itself have a connotation. Mill clearly admits that some, but not all, abstract names do connote (vii. 32).

56 Locke and Berkeley differ with respect to the precise criteria for identifying epistemological simples. See Winkler, Kenneth P., Berkeley: An Interpretation, Oxford, 1989), pp. 53–64.Google Scholar

67 Early associationist accounts (e.g., Hartley's) stressed the similarities among people, not, as Locke emphasized, their differences. In as much as associationism must presuppose that people have certain capacities, it maintains that human nature has some fixed features. See Mandelbaum, Maurice, History, Man, and Reason, Baltimore, 1971, pp. 147–62.Google Scholar

58 IPML (CW), ch. 5, §.xxxiii, fn. and §.ii, respectively. Note that Bentham's list of simples (§.ii) includes a number of logically complex things, e.g., skills: what is psychologically simple can still be logically complex in that one can conceive of its having aspects or parts. For a discussion of related complexities about simple and complex, see Winkler, , pp. 53–64.Google Scholar

59 IPML (CW), ch. 5, §§.ii–iii. Parenthetical citations are to sections of this chapter.

60 Bentham omits pleasures of the feelings, a neglect of imagination and emotion for which Mill criticized Bentham's impoverished conception of human nature. See Mill, 's ‘Remarks on Bentham's Philosophy’Google Scholar and ‘Bentham’, in Essays on Ethics, Religion, and Society, CW, x. 3–18 and 75–116, respectively.Google Scholar

61 Mill's conception of mind in terms of active and passive faculties parallels views held by Berkeley, Plato, and Aristotle. See Turbayne, Colin M., ‘Lending a Hand to Philonous: the Berkeley, Plato, Aristotle Connection’, in Berkeley: Critical and Interpretive Essays, ed. Turbayne, Colin, Minneapolis, 1982, pp. 299–304Google Scholar. But even if sensation is itself an activity or even if sensation involves some modicum of mental activity in constructing experience (e.g., through judgement), Mill can still maintain that sensation is relatively passive, involving a distinctly different sort of mental activity than imagining or contemplating.

62 The example is from Mill's notes on his father's work, An Analysis of the Phenomena of the Human Mind. The example is discussed by Feagin, Susan, ‘Mill and Edwards on the Higher Pleasures’, Philosophy, lviii (1983), 246–7Google Scholar. As will become apparent, I do not follow Feagin's understanding of the distinction between higher and lower pleasures.

63 Berger, , pp. 37–40CrossRefGoogle Scholar; Gray, , pp. 70–2Google Scholar; Hoag, , ‘Mill's Conception of Happiness’, 426–9Google Scholar; ‘Happiness and Freedom’, 190–5.Google Scholar

64 Mill, , On Liberty, ch. 3, para. 3Google Scholar. Gray pragmatically justifies his talk of pleasures as activities in terms of its consistency with Mill's claims in On Liberty (pp. 70–2)Google Scholar. I return to this in Section VII.

65 One might object to my account on grounds that the higher pleasures include not only those ‘of the intellect, of the feelings and imagination’, but also relevantly complex experiences like ‘the moral sentiments’ (U, n, 4). Indeed, Mill analyses the moral sentiment of justice into its component parts (U, V, 16–21). But Mill properly uses ‘moral’ equivocally: in a broad sense opposed to amoral, to distinguish moral feelings from aesthetic ones, for example, and in a narrow sense to signify natural sentiments appropriately modified by considerations of the general happiness (Berger, , pp. 19–23Google Scholar). The former, broad sense signifies other-regarding feelings such as sympathy (U, V, 19) or ‘the social feelings of mankind … to be in unity with our fellow creatures’ (U, III, 10). When Mill characterizes the higher pleasures as including those of the moral sentiments (U, II, 4), he cannot use ‘moral’ in the narrow sense because, as his analysis of the sentiment of justice demonstrates (U, V, 21), that use presupposes a conception of the general happiness and an aim of the doctrine of higher pleasures is to establish a conception of that standard. Therefore, the higher pleasures, including moral (as opposed to amoral) sentiments, can plausibly be construed as simple in the relevant sense.

66 As throughout, I attempt to interpret Mill independently of his phenomenalism. Just as some questions of logic can be insulated from metaphysics, so some questions of ethics can be insulated from Mill's phenomenalist metaphysics. Furthermore, since all pleasures are conscious states (Section I), all attributes of pleasures are grounded exclusively in those Feelings. Mill's phenomenalist account of attributes, then, does not affect discussion of pleasures' attributes.

67 In his discussion of estimating pleasures' values, Bentham denominates all relevant aspects as ‘circumstances to be taken into account in estimating the value of a pleasure’ (IPML (CW), ch. 4, §.i; emphasis added). Thus, one ought not assume Mill uses ‘quantity’ in Utilitarianism to signify Bentham's dimensions of pleasures' values. Indeed, Mill lists remoteness and sequence among relations (vii. 67–8, 70), thereby eliminating from quantity (a non-relational attribute) Bentham's conceptions of propinquity, fecundity, and purity.

68 It follows that not only is Mill's appeal to qualities of pleasures a non-consequentialist defence of their value (i.e., not an appeal to their ‘greater permanency, safety, uncostliness, etc.’ (U, II, 4)), but that contrary to some commentators (e.g., Raphael, , p. 354Google Scholar), quality cannot be understood as reducible to quantity. Mill's defence of the mental pleasures as superior rests on a ‘higher ground’ in two respects: it relies only on intrinsic features and the ground of the estimation is independent of quantity.

69 This follows from Mill's conception of any attribute as a natural property subject to scientific study (note 47 above).

70 Confusion understandably arises from the ambiguity of ‘quality’ in ordinary English and Mill's use of the term in estimating things. Though perhaps unfortunate, Mill's choice of terminology is not without foundation: etymologically, ‘quality’ derives from the Latin, qualis, meaning, ‘of what kind’ (Gibbs, , p. 41Google Scholar). As ‘the quality of wine’ ambiguously suggests both a descriptive and normative claim, so ‘the quality of pleasure’ is similarly ambiguous. No such confusion appears in the Logic account of quality as an intrinsic attribute of things: ‘quality’ is, in some respects, a technical term for Mill, and recognition of this illuminates the discussion in Utilitarianism.

71 Mill emphasized ‘kind’ in all four editions of Utilitarianism as well as in the original essay published in Fraser's Magazine. But some popular editions of Utilitarianism mysteriously omit Mill's use of italics to emphasize ‘kinds’ (i.e., in U, II, 4), while accurately including Mill's emphasis of other terms (e.g., in U, II, 11). Note, for example, editions published by Hackett (ed. Sher, George, 1979Google Scholar) and Bobbs-Merrill's Library of Liberal Arts (ed. Piest, Oskar, 1957Google Scholar). As my discussion reveals, D. E. Cohen, for example, is incorrect to identify classes and kinds in Mill's account of pleasures (‘J. S. Mill's Qualitative Hedonism: A Textual Analysis’, The Southern Journal of Philosophy, xviii (1980), 151–8).CrossRefGoogle Scholar

72 Skorupski, , pp. 87–90Google Scholar; also, see pp. 59–63, 95–8.

73 Mill expresses this point in several different ways: in the case of mere classes, all distinguishing attributes are ‘in some way dependent on, or connected with’ the connoted attribute (vii. 122); all individuating attributes are either connoted, implied by, or derivative from some few fundamental differences (vii. 122, 124); the differences all refer to some common cause (vii. 124). The fundamental idea seems to be that in the case of kinds there is no small set of characteristics which account for (logically or causally) all other class-differentiating characteristics. This account clearly opposes Locke's, reflecting Mill's rejection of essentialism and general substances. See Skorupski, , p. 394.Google Scholar

74 Nothing in Mill's account of kinds precludes the possibility that differences in kind are not sometimes associated with differences in quantity alone or with differences in any combination of attributes. But in Utilitarianism Mill is clearly concerned with a difference in kind grounded in qualitative differences alone (U, II, 4–5), and I restrict my discussion accordingly.

75 The connotations of all general names are vague, but the difficulty of indeterminate connotations is particularly acute for names of kinds (viii. 668–70).

76 See note 63 above as well as U, II, 4, 6. It follows from Mill's account that, for example, Jan Narveson is incorrect to identify quality and kind: not all differences in kind are differences in quality (note 74), and not all differences in quality need be also differences in kind (Morality and Utility, Baltimore, 1967, p. 82).Google Scholar

77 For example, see discussions listed in note 7 as well as Robson, John M., The Improvement of Mankind, Toronto, 1968, pp. 156–8CrossRefGoogle Scholar. The importance of the circularity problem is pursued by McPherson, Michael S., ‘Mill's Moral Theory and the Problem of Preference Change’, Ethics, xcii (1982), 263–7.Google Scholar

78 One might not only view ‘I mean’ as colloquial, but, given the common use of similar language of pleasures (note 17 above), Mill plausibly emphasizes what I mean as opposed to, for example, Bentham's or Hutcheson's use of terms like ‘quality’.

79 This opposes G. E. Moore, for example, who improperly attributes to Mill an analytic naturalism whereby desirability is defined in terms of a natural property, being desired (Principia Ethica, pp. 66–8Google Scholar). Mill's naturalism in ethics is not a definitional one and flows from his general approach to epistemology. Everett Hall thoroughly defends Mill against Moore's criticisms of the ‘proof: ‘The “Proof” of Utility in Bentham and Mill’, Ethics, lx (1949), 1–18, esp. 1–7.Google Scholar

80 Mill also develops the distinction in his essay, ‘On the Definition of Political Economy’, in Essays on Economics and Society, ed. Robson, John M., 2 vols., Toronto, 1967Google Scholar, Collected Works of John Stuart Mill, iv. 338–9, 319–20.Google Scholar

81 Berger, , pp. 23–5.Google Scholar

82 Skorupski, , pp. 83–5, 286.Google Scholar

83 Mill does not require unanimity among competent judges, but allows an agreement among a majority to be sufficient evidence for estimating pleasures. Mill recognizes that circumstances may arise where otherwise competent judges are not clearly in the best position to classify pleasures for the purpose of articulating a conception of human happiness as a moral standard. That is, competent judges may be in a state of ‘unhappiness so extreme that to escape from it they would exchange their lot for almost any other, however undesirable in their own eyes’ (U, II, 6), or they may momentarily succumb to ‘the influence of temptation’, or suffer ‘from infirmity of character’ or from a decline in their nobleness of character (U, II, 7).

84 The appearance of circularity lingers, ‘an old threat’ that epistemological naturalists claim to avoid. See Quine, , ‘Epistemology Naturalized’, Ontological Relativity and Other Essays, pp. 69–90, esp. 82–4Google Scholar. The question of circularity for Mill, then, is not peculiar to his doctrine of qualities of pleasures, but a broad, complex issue associated with his naturalized epistemology and its pragmatic strains.

85 See Kraut, Richard, ‘Two Conceptions of Happiness’, Philosophical Review, lxxxviii (1979), 167–97CrossRefGoogle Scholar. Kraut defines a subjective view as one where the standards of one's own happiness are self-imposed (180).

86 Some recent scholarship has attended to Mill's emphasis on activities and practical reasoning as included among the higher pleasures. But the implications of this are yet to be consistently pursued. Gray, John, for example (pp. 71–3)Google Scholar, asserts the higher pleasures are associated with deliberation. The view is offered without textual support and inconsistently maintained. After asserting that the higher pleasures are associated with autonomous choice, Gray shifts to talking about higher pleasures as objects, as things chosen: ‘if the pattern of his autonomous choices change, then so must the content of his higher pleasures’ (p. 72). But the content of the higher pleasures need not change so long as autonomous choosing continues, whatever its pattern.

87 Berger refers to this as the ‘meshing’ problem, pp. 286–7.

88 The point was first made by Berlin, Isaiah, ‘John Stuart Mill and the Ends of Life’ in Four Essays on Liberty, Oxford, 1969, pp. 173–206Google Scholar. Skorupski sees this unifying attempt with ‘happiness’ as the real weakness in Mill, 's moral theory (p. 307)Google Scholar; Hoag is much more sympathetic (‘Mill's Conception of Happiness’, 417–30).Google Scholar

89 Williams, Bernard, Ethics and the Limits of Philosophy, Cambridge, 1985, pp. 17–18.Google Scholar

90 It is a basis of Gray's conclusion that Mill cannot provide an adequate account of liberalism: Liberalisms: Essays in Political Philosophy, London, 1989, pp. 220–4.Google Scholar

91 Griffin, James defends such a view in Well-BeingGoogle Scholar. Particularly interesting and helpful is the discussion of incommensurability (ch.5), and I rely on his approach in the sentences that follow.

92 Revisionist views of Mill maintain that the aim of the doctrine of qualities of pleasures is to estimate pleasures as constituents of happiness in life. See references in note 8.

93 Hoag, , ‘Mill's Conception of Happiness’, 41–120, 430–1.Google Scholar

94 See Hoag, Robert W., ‘Mill on Conflicting Moral Obligations’, Analysis, xliii (1983), 49–54.CrossRefGoogle Scholar