Online research in philosophy

Weakness of will has been widely discussed from at least three points of view. It has been examined historically, with Aristotle recently occupying centre stage. It has been analysed conceptually, with the question of its nature and possibility in the forefront. It has been considered normatively in relation to both rational action and moral character. My concern is not historical and is only secondarily conceptual: while I hope to clarify what constitutes weakness of will, I presuppose, rather than construct, an account of it. My chief aim is to assess the bearing of weakness of will on the rationality of actions that exhibit it--incontinent actions. Philosophers have tended to assume that incontinent action is a paradigm of irrationality, and none to my knowledge has seriously criticised this assumption. I challenge it and in doing so try to clarify rationality in general.

Strict akratic actions, by definition, are performed freely. However, agents may seem not to be selfgoverned with respect to such actions and therefore not to perform them autonomously. If appearance matches reality here, freedom and autonomy part company in this sphere. Do they? That is this article's guiding question. To make things manageable, it is assumed that there are free actions, including strict akratic actions. Two theses are defended. First, the combination of (i) an intentional action's being uncompelled and (ii) its being - or executing - in appropriate informational circumstances, a sane decision that, as the agent recognizes, is for a course of action that she believes to be inferior to an alternative course of action open to her is sufficient for the action's being freely performed. (Condition (i) provides elbow room allegedly needed for free action, and (ii) encompasses freedom-level psychological sophistication.) Second, the same combination is sufficient for an intentional action's being autonomously performed.

INCONTINENCE, HONOURING SUNK COSTS AND RATIONALITY According to a basic principle of rationality, the decision to engage in a course of action should be determined solely by the analysis of its consequences. Thus, considerations associated with previous use of resources should have no bearing on an agent’s decision-making process. Frequently, however, agents persist carrying on an activity they themselves judge to be nonoptimal under the circumstances because they have already allocated resources to that activity. When this is the case, agents are said to be honouring sunk costs. Honouring sunk costs is thus typically viewed as irrational economic behaviour. Two striking findings of the psychological literature on the sunk cost effect are that it is, in general, not lessened by having taken previously courses in economics, and that agents who did take such courses and that do honour sunk costs seem to be frequently aware of a cognitive dissonance in their behaviour. Davidson’s account of the possibility conditions of incontinent action renders continent action materially unfeasible. The very idea of incontinent action becomes thus a hollow one. I’ll present an alternative description of these possibility conditions that renders continent action materially feasible, and both, continent and incontinent action, cognitively plausible. In a nutshell, my proposal rests upon the drawing of an essential cognitive contrast between explicit processes of deliberative reasoning and lower level heuristic procedures. I conclude the paper arguing that my redescription of incontinent action accounts for the above mentioned cognitive dissonance effect better than the alternatives and that viewing at least an important subset of behaviours of honouring sunk costs as incontinent actions will enable us to consider them as peculiar manifestations of a more general and meaningful pattern in human behaviour.

Philosophical discussions of akrasia over the last fifteen years have focused on certain skeptical arguments which purport to question the possibility of a kind of akratic action which, following Pears, I call 'last ditch akrasia' (Pears [38]). An agent, succumbing to last ditch akrasia, freely, knowingly, and intentionally performs an action A against his better judgment that an incompatible action B is the better thing to do. (See Audi [1] for a detailed analysis.) Last ditch akrasia is not the only kind that has been discussed. Some philosophers (Mele [32], Scaltas [48]) have been concerned with a more extreme form of akratic action, viz. one in which the agent not only judges that action B is best, but in addition intends (chooses, decides) to B. Some have even questioned whether freely acting against one's better judgment is sufficient for akratic action (Schiffer [49]: 201-3).1 Weaker types of akratic action have been discussed, though to a much lesser extent, since they are thought less problematic. Pears distinguishes last ditch akrasia from what he calls, "motivated irrational action" ([38]: 160). In cases of the latter, the akrates' rebellious desire infects his prior reasoning and thinking in such a way that his contemplated action seems to him warranted, and he acts accordingly. (For a taxonomy of cases of akratic action, see Rorty ([44])). Nor has the discussion of akrasia been restricted to akratic action. Philosophers have discussed whether akrasia can be exhibited in the formation of intentions, wants (See Audi [1]: 181-185), and beliefs (Mele [32]: ch. 8; Heil [19]). The primary focus in this paper is on last ditch akratic action.

The occurrence or apparent occurrence of incontinent actions challenges several influential views in ethics and the philosophy of mind, e.g., Hare's prescriptivism and the Socratic idea that we always act in the light of the imagined greatest good. It also raises, as I shall explain, an interesting and instructive problem for proponents of causal theories of action. But whereas Socrates and Hare attempt to avoid the difficulties with which akrasia confronts them by denying - wrongly, I shall argue - that there are akratic actions ([15], 352a-358d; [13], Ch. 5), the causal theorist need not take this unhappy tack. In this paper I shall argue that the truth of a causal theory of action (CTA) is compatible with the occurrence of akratic actions and, in particular, with akratic actions against what I shall call a "here-and-now" intention - i.e., an intention of the agent to do an A here and now. I shall suggest that akratic actions of this type plainly do occur, and I shall attempt to explain how they might be accommodated by a causal theory of action.

That incontinent action is possible, I have argued elsewhere. The purpose of the present paper is to ascertain whether such action can ever be free.

Although much human action serves as proof that irrational behavior is remarkably common, certain forms of irrationality--most notably, incontinent action and self-deception--pose such difficult theoretical problems that philosophers have rejected them as logically or psychologically impossible. Here, Mele shows that, and how, incontinent action and self-deception are indeed possible. Drawing upon recent experimental work in the psychology of action and inference, he advances naturalized explanations of akratic action and self-deception while resolving the paradoxes around which the philosophical literature revolves. In addition, he defends an account of self-control, argues that "strict" akratic action is an insurmountable obstacle for traditional belief-desire models of action-explanation, and explains how a considerably modified model accommodates action of this sort.

Brian McLaughlin, in “Incontinent Belief” (Journal of Philosophical Research 15 [1989-90] , pp. 115-26), takes issue with my investigation, in lrrationality (Oxford University Press, 1987), of a doxastic analogue of akratic action. He deems what I term “strict akratic belief” philosophically uninteresting. In the present paper, I explain that this assessment rests on a serious confusion about the sort of possibility that is at issue in my chapter on the topic, correct a variety of misimpressions, and rebut McLaughlin’s arguments as they apply to the psychological possibility of strict akratic belief and to the etiology of beliefs generally.

Alfred Mele has recentIy attempted to direct attention to a neglected species of irrational belief which he calls ‘incontinent belief’. He has devoted a paper and an entire chapter (chapter eight) of his book, Irrationality (Oxford University Press, 1987) to explaining its logical possibility. In what follows, I will appeal to familiar facts about the difference between belief and action to make a case that it is entirely unproblematic that incontinent belief is logically possible. In the process, I will call into question the philosophical intercst of incontinent belief. If what I say is correct, incontinent belief does not warrant the attention of philosophers of mind.

Elsewhere, I have argued for the possibility of strict or full-blown akratic action - roughly, free (or uncompelled), intentional action against the agent's better judgment.' My aim in the present paper is to defend and account for the possibility of an analogous variety of akratic feeling.