What in the world is weakness of will?

0 R. Holton Department of Linguistics and Philosophy, Massachusetts Institute of Technology , 77 Massachusetts Avenue, 32-D808, Cambridge, MA 02139-4307, USA 1 J. May (&) Department of Philosophy, University of California , Santa Barbara, 5631 South Hall, Santa Barbara, CA 93106, USA At least since the middle of the twentieth century, philosophers have tended to identify weakness of will with akrasiai.e. acting, or having a disposition to act, contrary to one's judgments about what is best for one to do. However, there has been some recent debate about whether this captures the ordinary notion of weakness of will. Richard Holton claims that it doesn't, while Alfred Mele argues that, to a certain extent, it does. As Mele recognizes, the question about an ordinary concept here is one apt for empirical investigation. We evaluate Mele's studies and report some experiments of our own in order to investigate what in the world the ordinary concept of weakness of will is. We conclude that neither Mele nor Holton (previously) was quite right and offer a tentative proposal of our own: the ordinary notion is more like a prototype or cluster concept whose application is affected by a variety of factors. - two quite distinct notions (Holton 1999).1 The first is the idea of acting contrary to ones best judgementthe idea that is to the fore in classical discussions of akrasia. The second is the idea of over-readily revising a resolution. It is plausible that the two typically go together: in acting contrary to their best judgments, agents will typically over-readily revise their resolutions; and conversely, in over-readily revising their resolutions they will typically act contrary to their best judgments. So it is understandable that the two ideas have been conflated. But nevertheless they are clearly distinct. Holton could have rested content with pointing out the distinction and getting clear on the ideas involved. Rashly though, he went further. He claimed that it is just the second idea, the idea of over-ready resolution revision, that corresponds to our ordinary notion of weakness of will. The idea of acting contrary to ones best judgment he took to be a philosophers invention, best labeled with the philosophers proprietary term akrasia. He should have known better.2 Whilst the claim enabled him to formulate his position in a neat sloganweakness of will is not akrasiait left him open to empirical refutation. For the nature of our ordinary concepts is, at least to some degree, an empirical question; and he had done no real empirical work to substantiate what he was saying. Not that it seemed that such work was needed at the time. Even 10 years ago the analysis of ordinary concepts was a leisurely affair, conducted largely by reflection, and the casual interrogation of colleagues, friends and students. Now, however, it is an altogether more rigorous business, and in a recent paper Mele (2010) has used the survey method that has become the hallmark of experimental philosophy to argue that Holton was wrong. As we understand him, Mele claims that the ordinary notion of weakness of will is disjunctiveone exhibits weakness of will either by acting contrary to ones evaluative judgment or by acting contrary to ones plan. On this view, Holton should not have been so restrictive. Our project is to shed light on what in the world the concept of weakness of will is. Is there an ordinary notion here? If so, is it disjunctive as Mele contends? Employing some empirical methods ourselves, we argue that neither the traditional account of weakness of will, nor Holtons, nor Meles, is quite right. Indeed, our findings suggest that no simple account phrased in terms of necessary and sufficient conditions will do the job. The ordinary notion of weakness of will is more like a prototype or cluster concept. There are core cases that possess a number of features. As these features are removed, people are less inclined to describe the resulting cases as ones of weakness of will. Akrasia and resolution-violation are indeed among these features. However, neither is sufficient on its own for an ascription of weakness of will; and other features also play a role, such as the moral valence of the action. 1 A lightly revised version appears as ch. 4 of Holton (2009). 2 In fact it seems that he did: the English language is a plastic instrument he wrote; is it not very likely that the traditional account captures one of our uses of the expression? But he then went on to say that he couldnt help thinking that the traditional account was straight-out wrong (pp. 257258). He should have tried harder to resist. 2 Meles case against Holton In Intention and Weakness of Will (1999), Holton argues that the ordinary concept of weakness of will essentially involves the violation of a certain kind of intentiona resolution. Resolution is something of a technical term for Holton; it refers to an intention or plan one has to stick to a certain course of action in the face of the temptation to succumb.3 For example, consider a smoker who sincerely admits it isnt best for her to continue smoking cigarettes but nevertheless doesnt plan to quit; she recognizes the detrimental effects of smoking but doesnt resolve to quit in spite of her evaluative judgment. Such a person is clearly being akratic. After all, akrasia is a term of art, and this person is by hypothesis acting contrary to what she thinks is best for her to do. But is she acting in a weak-willed manner? Is she displaying weakness of will?4 Holton claimed not. For him, agents display weakness of will only if they violate a resolution.5 To get a case of weakness of will then, the smoker would have to resolve to quit, but then succumb to temptation and continue. Akrasia and violations of resolutions often come and go togetherwe often resolve to do what we think is best for us to dobut they can come apart. Call this the resolution account of weakness of will. In a recent paper, Mele (2010) argues that the ordinary concept of weakness of will involves both the notion of akratic action and the notion of intention-violation. Building on earlier work, Mele distinguishes an agents evaluative commitments (roughly their judgements about which action would be best) from their executive commitments (roughly their resolutions). He holds that there are traditional or orthodox versions of akrasia that involve violations of ones evaluative commitments; but that there are also nontraditional or unorthodox versions of akrasia, that involve violations of ones executive commitments. After summarizing his earlier work, Mele writes: I did not offer full-blown analyses of akratic and enkratic action. Instead I offered sketches of conceptions of both kinds of action designed to accommodate traditional and nontraditional species of them. Are these sketches hopelessly flawed? (p. 394) 3 It seems pretty close to the ordinary usage (consider a New Years resolution to quit smoking). B (...truncated)