1 References to Hume are to the Selby-Bigge/Nidditch edition of his A Treatise of Human Nature (Oxford: Oxford University Press, 1978). This work is hereafter cited as T, with page numbers given in parentheses. The following alternative method of citation is also used: T, 1.1.1, meaning Treatise, Book 1, Part 1, section 1.Although my discussion will focus mainly on Hume's views in the Treatise, I will also make references to his Enquiry Concerning the Principles of Morals in theGoogle Scholar, Selby-Bigge/Nidditch edition of the EnquiriesConcerning Human Understanding and Concerning the Principles of Morals (Oxford: Oxford University Press, 1975). The citation will be E, with page numbers given in parenthesesGoogle Scholar.

2 The basis for this charge is Reid's well-known remark that Hume destroys both the corporeal and the spiritual worlds, leaving only impressions and ideas without a subject in which to inhere. See his Essays on the Powers of the Human Mind, printed for Bell & Bradfute (Edinburgh, 1819), Vol. 1, p. 267. For more on Reid's interpretation of Hume, see the following:Google ScholarRobison, Wade L., “Hume's scepticiam” dialogue, 12, 1 (1973): 87–99; alsoCrossRefGoogle ScholarRobison, Wade L., “Hume's Ontological Commitments,” Philosophical Quarterly, 26 (1976): 40–41. Other useful sources for the early interpretations of Hume's philosophy include:CrossRefGoogle ScholarSmith, Norman Kemp, The Philosophy of David Hume (London: MacMillan, 1941), p. 3–5;CrossRefGoogle ScholarPopkin, Richard H., “The Early Critics of Hume,” in The High Road to Pyrrhonism, edited by Watson, Richard A. and Force, James E. (San Diego: Austin Hill Press, 1980)Google Scholar; Norton, David Fate, “Hume and His Scottish Critics,” in McGill Hume Studies, edited by Norton, David Fate, Capaldi, Nicholas and Robison, Wade L. (San Diego: Austin Hill Press, 1979), p. 309–24; andGoogle ScholarMoss-ner, Ernest C., The Life of David Hume (Oxford: Clarendon Press, 1980), chap. 10Google Scholar.

3 Foot, Philippa, “Free Will as Involving Determinism,” Philosophical Review, 66 (1957): 439–50. Admittedly, Foot's criticism is made in a much wider context than I have indicated here: she is questioning the claim, historically associated with Hobbes and Hume, that the idea of moral responsibility is intelligible onlywithin a determinist framework and that free will requires determinism for ascriptions of moral responsibility. Nevertheless, she rejects Hume's conception of responsibility along the lines I will present.CrossRefGoogle Scholar

4 Cf. Bricke, John, “Hume's Conception of Character,” South-Western Journal of Philosophy, 5 (1974): 107–13; and also hisCrossRefGoogle Scholar“Hume's Theories of Dispositional Properties,” American Philosophical Quarterly, 10, 1 ( 01 1973): 15–23Google Scholar.

5 As if to emphasize the significance of cognition in the matter, Hume reiterates elsewhere that agents are morally responsible only for actions that they perform with “thought and deliberation” (T, p. 412). And he gives as morally exculpable forms of behaviour, actions performed in ignorance, in haste and unpremeditatedly. In other words, he believes that the agents of such actions are to be excused. It might be contended that Hume remarks, in T, p. 348–49, thatthe durable qualities or dispositions of an agent need not be exerted intentionally in order to draw praise or blame on the agent. I am inclined, however, to read this passage differently. What Hume meansby the above remark is that, given that we already know the nature of the moral quality of character that directs a person's action, we feel an aversion or inclination towards the person without seeking further to enquire about his intention in acting in any particular instance. On the other hand, in the absence of such knowledge we must enquire about the nature of the person's intention. There are difficulties with such a position which I cannot pursue here. For example, we can very easily make mistakes in our assessment of a person by showing approval or disapproval of him as we have done in the past, although he may have changed his character trait.

6 Of course I do not mean to suggest that Hume thinks statements about intentions are lit-efaftly identical in meaning, with statements about individgaj characters. In the first place, intentions are of a much shorter duration than character, and this temporal difference must be reflected in statements about the two. Second, an intention is a cognitive state of which one is necessarily aware. A person cannot have an intention to perform an action and not know it. By contrast, to have a character is at least to have a certain disposition to behave in certain relevant ways when appropriate circumstances obtain. Third, intentions sometimes just pop into our heads; we cannot explain how we come to have them. But we acquire most character traits by habit, training, practice and education—a point wellnoted by Hume. These differences notwithstanding, Hume wishes to saythat intentions are the particular instantiations of character.

7 Doubts may perhaps be raised over the interpretation I have proposed of Hume's conception of the relation between intention and character; in particular, my suggestion that the relation is necessary. However, such doubts will disappear as soon as it is realized that Hume considers our knowledge of an agent's intention to be conclusive evidence of the agent's character. If Hume conceived of intentions as logically distinct and distinguishable from individual characters, then it would follow that a person with a virtuous character (begging the question of how we ever come to know that a person is virtuous or vicious) would consistently act from a wicked intention. Similarly, a vicious character may consistently perform actions from a good intention. In short, the relation between character and intention would be contingent, in which case anything may arise from anything. But this is precisely the kind of situationthat is pre-empted by Hume's aforementioned remark that the morality of an action is a function of the morality of the intention that gives rise to the action. Where an intention is logically connected with, and hence serves as evidence of, a person's character, a good intention necessarily manifests a good (or virtuous) character and a bad intention necessarily reveals a bad (or vicious) character.

8 I am certainly aware that Hume would not describe as virtuous an action that appears to be so, but which has been motivated by, say, a wicked intent. The action would be anything but virtuous. I wish only to remark here that, on a strict interpretation of his contingency thesis of causation, according to which anything might arise from anything, Hume should say that moral actions can be produced by even non-moral features of persons. I have discussed this issue in my “Hume on Character, Action and Causal Necessity,” Auslegung: A Journal of Philosophy, 16,2(Summer 1990): 149–64Google Scholar.

9 Even the most recent discussion of his account of moral responsibility, that of Lloyd Fields, is defective in tMs matter. See his “Hume on Responsibility,” Hume Studies, 14, 1 (04 1988): 161–75CrossRefGoogle Scholar.

10 Foot, “Free Will as Involving Determinism,” p. 448; emphasis in text.

11 Elements of the conception of self as relatively enduring seem present in the very Book One in which Hume proposes the bundle theory. Consider his analogy of the relation between a republic and its members, which he uses to illustrate the relation between self (as bundle) and its constituent perceptions. As one and the same republic may change its members, as well as its laws and constitution, says Hume, so can one and the “same person…vary his character and disposition, as well as his impressions and ideas, without losing his identity.” Continuing, Hume adds: “Whatever changes he endures, his several parts are still connected by the relation of causation. And in this view our identity with regard to the passions serves to corroborate that with regard to the imagination, by the making our distant perceptions influence each other, and by giving us a present concern for our past or future pains or pleasures” (T, p. 261). Hume undoubtedly begs the question about “same person.” Moreover, it is now an issue whether self is something over and above its constituent perceptions, notwithstanding Hume's aforementioned remarks to the contrary. However, I will not attempt to address these matters here. What is significant for my present purpose is that he advances, albeit implicitly, the notion of self as a continuing entity. Some useful discussions of the view of self as permanent are offered by Mclntyre, Jane L., “Is Hume's Self Consistent?,” in McGill Hume Studies, edited by Norton, David F., Capaldi, Nicholas and Robison, Wade L. (San Diego, CA: Austin Hill Press, 1979), p. 79–88; and alsoGoogle ScholarMcLntyre's, “Further Remarks on the Consistency of Hume's Account of the Self,” Hume Studies, 5, 1 (04 1979): 55–61. Unlike Mclntyre, I have not tried to reconcile Hume's apparently conflicting accounts of the self. I have simply tried to show that, granted that Hume holds different views of the self, heis not precluded from speaking of character as a feature of persons that qualifies us for moral responsibility. For a criticism of Mcln-tyre's position, seeGoogle ScholarBiro, John, “Hume's Difficulties with the Self,” Hume Studies, 5, 1 (04 1979): 45–54.Google Scholar Finally, Robison, “Hume' s Ontological Commitments,” p. 40–41, contain a very illuminating discussion on the topic.

12 To be fair, Foot may also be read as denying Hume's view that intentional actions are always expressions of individualcharacters, and hence that responsibility-attributions are always and only for character-manifestations. On this alternative reading, mycriticism of her may perhaps not apply. But then, because she will now simply be advancing a position that is antithetical to Hume's, it remains an open question whether she or Hume holds the correct view about the relation between intention and character. In whateve r way the matter is resolved, however, Hume's view that a discovery about the nature of an agent's intention is also a discovery about the moral quality of the agent's character clearly undermines Paul Helm's contention that it is “notpossible precisely to identify an action as ‘in character’ or not.” The reason, says Helm, is that “one cannot.…specify the necessary features of a person's character such that if one of the conditions is not met it is the case that theperson was acting out of character and hence was not responsible or was not as responsible for the action” (Helm, Paul, “Hume on Exculpation,” Philosophy, 42 [1967]: 265–71)CrossRefGoogle Scholar.

13 Hobart, R. E., “Free Will as Involving Determination and Inconceivable Without It,” Mind, 43, 169 (01 1934): 1–27,CrossRefGoogle Scholar reprinted in Free Will and Determinism, edited by Bernard, Berofsky (New York and London: Harper & Row, 1966), p. 63–95. The citation here is to the latter text, p. 68. Emphasis in the original.

14 This objection is similar to that which Molyneux advances against Locke's theory of personal identity. See Allison, Henry A., “Locke's Theory of Personal Identity: A Reexamination,” in Locke on Human Understanding, edited by Tipton, I. C. (Oxford: Oxford University Press, 1977), p. 105–22Google Scholar.

15 I say “sometimes” because we disallow some occasional thoughtless actions as the product of ignorance or accident and, consequently, we do not excuse such actions in a way that excludes responsibility for them. For example, a parent who thoughtlessly leaves a loaded gun lying around when he has children in his care is deemed (morally and legally) accountable for any untoward event that may arise as a result of the children toying with the gun. Similarly, a parent who leaves a child unattended in a car with the key in the ignition is responsible for an accident that may occur if the child somehow starts the car. Responsibility-attribution in either case in made on the presumption that the parent should have known better than to act as he did.

16 Hume certainly does not limit our ascriptions of praiseand blame to actions that are voluntary in the sense of deriving from voluntarily acquired mental features. He contests as groundless the distinction traditionally drawn between moral virtues and abilities. According to Hume, the reason given for the distinction between virtues and abilities is that the former are voluntarily acquired, not the latter. Consequently, the actions we perform as a result of the virtues have moral significance, but not the actions we perform invirtue of our abilities. Hume reacts to this view, saying among other things that, although abilities are gifts of nature and are therefore involuntarily acquired, they nonetheless qualify their possessors for moral evaluation because of their utility or disutility eitherto their possessors or to others. Accordingly, it is not the case that, unless a mental feature is voluntarily acquired, it cannot qualify its possessor for moral appraisal in terms of the action it motivates (Treatise, 3.3.4). But this aspect of his moral theory must await another discussion.

17 I presented a version of this paper in April 1990 at a conference of the Atlantic Society for Eighteenth Century Studies (ASES) at St. Mary's University, Halifax, Nova Scotia. I am grateful to participants at the Philosophy session, in particular Nathan Brett and Susan Tatton, for an engaging discussion of some of the issues I have examined in this work.