This essay argues for a new conception of bad faith based partly on Harry Frankfurt's famous account of personal autonomy in terms of higher‐order volitions and caring, and based partly on Sartre's insights concerning tacit or pre‐thetic attitudes and “transcendent” freedom. Although Sartre and Frankfurt have rarely been connected, Frankfurt's concepts of volitional “wantonness” and “bullshit” (wantonness about truth) are similar in certain revealing respects to Sartre's account of bad faith. However, Sartre leaves no room for Frankfurt's central point that (...) people can volitionally commit themselves to develop certain motives and reduce others, thereby shaping an authentic identity through volitional “identification” and “alienation” of motives. Frankfurt in turn minimizes the role that a sense of liberty to form and alter higher‐order volitions plays in experiences of autonomy, which Sartre recognizes in the transcendence of the “for‐itself.” When these points are combined, we get a tri‐level theory of intrasubjective volitional stances that explains Sartre's cases better than his own theory can, and also clarifies how sincere or procedurally autonomous self‐relation is possible. (shrink)

Readers familiar with Harry Frankfurt’s argument that we do not need leeway-liberty (or the power to bring about alternative possible actions or intentions) to be morally responsible will probably also know that the most famous and popular response on behalf of leeway-libertarianism remains a dilemma posed in similar forms by David Widerker, Robert Kane, and Carl Ginet: either the agent retains significant residual leeway in Frankfurt-style cases, or these cases beg the question by presupposing causal determinism. In the last few (...) years, there have been several different attempts to defend Frankfurtian critiques of PAP in response this dilemma. In a novel approach, Derk Pereboom and Michael McKenna present cases in which all deliberatively relevant or “robust” alternatives are blocked, but the agent’s act or decision is not determined. Pereboom and McKenna argue that any plausible leeway-condition on responsibility must characterize the required alternatives as robust in two ways: being voluntary performances and having a practical relevance accessible to the agent’s mind. I agree with the requirement of robustness, and argue that we can build this notion into a complex concept of agent-possibility, or “agentive-can.” However, I argue that both McKenna’s and Pereboom’s conceptions of robustness are too demanding: they exclude alternatives that are intuitively relevant. Moreover, I argue that the alternative of refraining from deciding, or voluntarily failing to decide, is robust in the right sense. In agreement with a tradition running from Ockham back through Scotus to Aquinas, I argue that this robust alternative is necessary for responsibility. If the Frankfurt-controller eliminates it, then the agent’s responsibility is undermined. In particular, I argue that Pereboom’s tax evasion cases do not refute this leeway-condition on moral responsibility. (shrink)