The ontology of character traits in Hume

Abstract

This paper argues that Hume can account for character traits as lasting mental qualities without violating his reductionist account of the mind as a changing bundle of ideas and impressions. It argues that a trait is a disposition to act according to certain passions or motivations, explained entirely with reference to the ideas and impressions constituting one's current self. This account is consistent with Hume's view of the mind, and relies solely on his accounts of the association of impressions and ideas, and of the relationship between belief and passion, to establish relations that can properly be called lasting mental qualities.

Keywords::

• Humean self
• character
• mental qualities
• association of ideas

Notes

^1. In some contexts, the term “character” may refer more generally to the whole of a person's “character traits.” In this paper, the focus is on what a character trait is, and I am not concerned with the question of what it is to have a “character” more generally. I use the term “character” therefore, in this context, only to refer to “character traits.”

^2. Ainslie also notes that this aspect of Hume's philosophy is “remarkably underexamined” (2007, 80 note 6).

^3. Ainslie (2007, 80 note 6) also notes Baier (1991, chapter 8) and Dees (1997) as exceptions to the general neglect, but both focus on the issue of trait evaluation, which, for reasons to be explained, is pre-empted by the more primary issue of the possibility of possessing traits at all.

^4. Purviance (1997) has also addressed the issue, but does so with the aim of eschewing any ontological account of traits. She has the explicit goal of separating the metaphysics of the self from Hume's practical morality (and so arguing that the former is non-essential for the latter, in a Kantian sense of metaphysical claims being utterly beyond our comprehension and so independent of our practical concerns). Instead, she offers a “Fact of Agency View,” which gives real status to self-assertions without accounting for an ontologically distinct referent. I do not discuss her view directly because it is based on a couple of false assumptions about this debate, and so to articulate her argument would take us far afield of the current discussion. In short, she misreads McIntyre's view and by proxy any ontological account as requiring a “self-identical ‘I’” across time (205), which my view does not entail (nor, I think, does McIntyre's, though it comes closer). My second primary concern with Purviance's argument is that she assumes character judgements are made about a whole character, and so a bundle account cannot support a character judgement (208). There is a lot of evidence against this assumption in Hume's texts, if one looks at the character assessments he gives of historical figures. Hume clearly allows for trait-by-trait judgements of virtue, so this is not a criterion an account of the self needs to accommodate. Since she fails to see room for an ontological account that does not assume a re-identified single, unified self over time, and since she offers a false criterion for an ontological account of self to support character ascription, her argument does not offer reasons to reject the view I offer here. For reasons I will offer, I think we do need some ontological account, so Purviance's explicit initial assumption is untenable.

^5. For example, “mental qualities”: T 3.3.1.4–5; SBN 575; T 3.3.4.1; SBN 606–607, T 3.3.5.1; SBN 614; “qualities”: T 2.1.7.3; SBN 295, T 2.2.1.4; SBN 330, T 2.2.3.4; SBN 348-9, T 3.3.3.9; SBN 606.

^6. “many of those qualities, which all moralists … comprehend under the title of moral virtues, are equally involuntary and necessary, with the qualities of the judgment and imagination …. I might say the same, in some degree, of the others; it being almost impossible for the mind to change its character in any considerable article, or cure itself of a passionate or splenetic temper, when they are natural to it” (T 3.3.4.3; SBN 608).

^7. For example, in his essay “The Sceptic,” Hume writes:

Let a man propose to himself the model of a character, which he approves: Let him be well acquainted with those particulars, in which his own character deviates from this model: Let him keep a constant watch over himself, and bend his mind, by a continual effort, from the vices, towards the virtues; and I doubt not but, in time, he will find, in his temper, an alteration for the better. (E-Sk 170)

See also T 2.2.3.4; SBN 348–349, T 2.3.2.7; SBN 411–412.

Radcliffe (2007) has a developed account of how a naturalistic framework like Hume's can allow for character development, in which she argues that inspiring rhetoric, education, and forms of punishment and reward can all contribute to the possibility of character change even within such a framework; these claims are fully compatible with the account given here, and I am inclined to agree with Radcliffe that this change is possible. Nevertheless, her account also allows for the claim that such change can be quite difficult.

^8. One might worry that Hume's scepticism about causal relations creates a problem here. Insofar as Hume gives us some definition of what it means to say that a rolling billiard ball “causes” the ball it strikes to move (by which we mean the two are conjoined, and the mind associates the first with the second), we are, he argues, equally able to say that a character causes a person's actions. The difficulty in the second case is that we rarely, if ever, have a direct impression of another person's character when she is acting, whereas we do of the rolling billiard ball, and so while there is a constant conjunction of two impressions in the case of the billiard balls, there is only the constant conjunction between an action and an inferred motive in the case of human action. However, I think we have grounds for justified inferences of this sort. First, we arguably do have some first-person understand of traits causing actions in ourselves, and once the account of traits offered here is developed, we will see that the features of a person that constitute traits are discernable by her and in principle discernable by others who communicate and interact with her. So though I may not “see” a person's character each time I see her act, I can in principle understand character as causing actions generally, and also know her well enough to be familiar with her character. Furthermore insofar as Hume takes it as a fact that we must rely on causal assumptions in spite of their lack of rational justification, we should no more avoid talking about actions having determinate causes even in cases in which we are unfamiliar with a person's character than we avoid saying the motion of an inanimate body has a determinate cause; the issue here, of course, is distinguishing whether the determinate cause is a character trait, or an impulse, or a motive stemming from misunderstanding of the facts, etc. Elsewhere (in manuscript) I argue that there is a process of distinguishing traits from these other determinate causes, and this process requires further information about the person and the context of the action.

^9. My insistence on the reality of traits in persons is not new in Hume scholarship. For example, Penelhum (1975, 149), while not giving an account of character, has noted at different times the point I make here, that character traits are real features of individuals independently of their being perceived: “At present, however, it is important to add to the condition that moral judgments, to claim truth, must be based on disinterested appraisal that this appraisal must involve the discernment of real qualities and tendencies in the character appraised.” See also Penelhum (2009, 265). McIntyre (1990) is also attentive to this point.

^10. In establishing the first criterion, I agree with Bricke (1973) that Hume can allow for non-reductionist dispositional properties; however, I disagree with Bricke's claim that “it is that enduring thing we call a man's body … which is the carrier, as it were, of those enduring properties, including his character, which are necessary to explain his behavior” (22).

^11. He uses the terms “mind” and “self” interchangeably in this context, and I will do so as well.

^12. The terms “inclination” and “propensities” are understood to still be vague here – these are the terms Hume and we use to explain this behaviour, but the question still to be answered is exactly what these terms mean. (Thanks to an anonymous reviewer for requiring this clarification.)

^13. Though my argument here is independent (and more coarse-grained), I think my comfort in making this claim is in part the congenial, and more fine-grained, argument offered by Cohon (2008).

^14. In regard to a person being prone to calm rather than violent passions, a strong belief in the value of reflective decision-making might counteract the violence of present passions.

^15. Thanks to an anonymous reviewer for suggesting this clarification.

^16. In addition to passages cited in this paper, Hume refers to dispositions such as benevolence or generosity repeatedly in his texts; see, for example, T 3.3.1.21; SBN 585, EHU 8.15; SBN 88, EPM App 2.6; SBN 298.

^17. By “depends” I do not mean anything metaphysically ambiguous – one could say, “explained by,” or “determined by,” or “caused by” to express this same idea.

^18. This view can also accommodate the more unusual behavioural tendencies that Hume classifies as traits, such as cleanliness. A person's mental associations with clutter, or with keeping a clean house, can produce motivating passions. These associations of ideas are not different in kind than those that ground more traditional traits such as loyalty or generosity. Wit might still be seen as a problematic case, although I think it is plausible to characterize wit as a certain ability to associate disparate ideas with facility in ways that draw out 97absurdities or ironies. Even eloquence seems plausibly explainable in terms of a person's wealth of associated ideas and facility in connecting them.

^19. In addition, though this is not the topic of discussion here, the view I have offered quite amenable to Ainslie's more sophisticated answer to the epistemological question of how we recognize traits, and our understanding of character as the object of our passions. It enables us to explain character traits as mental qualities in a proper sense, and so establish their ontological status, while still maintaining his suggestion that on any given occasion, we need not know which mental elements are the source of the exhibition of character traits. That is, we can still separate the traits-as-objects-of-passions from traits-as-causes-of-actions, without leaving the latter unanswered.

^20. Ainslie's (2007, 97) account of traits as the offspring of social affective responses to a person's actions has trouble explaining Hume's claims about virtue in rags, or un-exhibited traits.

^21. By emphasizing beliefs here I do not intend to eliminate the necessary role of passions in producing characteristic behaviour; I merely mean to emphasize that in cases in which these passions never happen to arise, there is still something we can learn, or could in principle learn, about a person's character. (Thanks to an anonymous reviewer for requesting this clarification.)

Additional information

Notes on contributors

Erin Frykholm

Erin Frykholm is Assistant Professor of Philosophy at the University of Kansas. She works on the British Moralists, with particular focus on Hume's first-order moral theory and philosophical issues that arise therein. She also has research interest in overlapping issues in contemporary virtue theory.