1 Frankfurt, ‘Alternate Possibilities and Moral Responsibility,’ Journal of Philosophy 66 (1969) 828-39CrossRefGoogle Scholar

2 Frankfurt, ‘Freedom of the Will and the Concept of a Person,’ Journal of Philosophy 68 (1971) 5–20CrossRefGoogle Scholar

3 Frankfurt, ‘Identification and Wholeheartedness,’ in Schoeman, F. ed., Responsibility, Character and the Emotions (Cambridge: Cambridge University Press 1987) 27–45Google Scholar

4 Some philosophers worry that Frankfurt does not include enough in the ‘motivational mesh’ which grounds autonomy (Watson, G. ‘Free Agency,’ Journal of Philosophy 75 [1977] 316-39Google Scholar; Taylor, C. ‘Human Agency,’ in Mischel, T. ed., The Self [Oxford: Blackwell 1977] 103-35Google Scholar; Wolf, S. Freedom within Reason [Oxford: Oxford University Press 1990])Google Scholar. Others worry that no such mesh, however supplemented by further structural conditions, can do the job without help from historical constraints on desire-acquisition (Slote, M. ‘Understanding Free Will,’ Journal of Philosophy 78 [1980] 136-51CrossRefGoogle Scholar; Fischer, J.M. and Ravizza, M. Responsibility and Control [Cambridge: Cambridge University Press 1998]CrossRefGoogle Scholar; Mele, A. Autonomous Agents [Oxford: Oxford University Press 1995])Google Scholar. I take up the first objection in ‘Hierarchical Motivation and Freedom of the Will,’ Pacific Philosophical Quarterly 62 (1981) 254-368, and the second in ‘Doing and Time: Psychological Structure vs Personal History in Compatibilist Theories of Autonomy,’ forthcoming.

5 Cited in Williams, B. Shame and Necessity (Berkeley, CA: University of California Press 1993), 115CrossRefGoogle Scholar. Evidently, this idea was espoused only by the later Stoics of the imperial period. On the relevant differences within the Stoic tradition, see Annas, Julia The Morality of Happiness (Oxford: Oxford University Press 1993), 174-5, 307Google Scholar; also Nussbaum, Martha The Therapy of Desire (Princeton: Princeton University Press 1994), 11, 339, 353, 361, 363, 395Google Scholar. Moreover, it seems that not even Seneca, an important later Stoic, consistently downgraded the importance of worldly misfortune. Nussbaum cites a rather different sentiment from the one in the text: ‘There is no time for playing around …. You who have promised to bring help to the shipwrecked, the imprisoned, the needy, to those whose heads are under the poised axe. Where are you deflecting your attention? What are you doing?’ (317) She cites several other passages from Seneca which further mitigate the awfulness of the sentiment quoted in the text (325, 342).

6 Frankfurt, ‘Three Concepts of Free Action,’ Proceedings of the Aristotelian Society, Supplementary Volume (1975) 113-25, at 124, emphasis addedGoogle Scholar

7 In fact, I have posed the Stoic thesis in Frankfurt's own language, for the distinction between freedom of action and acting freely is one of the cornerstones of his theory. He famously insists that ‘the principle of alternate possibilities,’ which is associated with the first ('Alternate Possibilities and Moral Responsibility’) is not a condition for moral responsibility, which involves only the second ('Freedom of the Will and the Concept of a Person’).

8 This is a feature of his theory which Frankfurt stresses again and again. See, for example, ‘Three Concepts of Free Action,’ 121-2; and ‘Identification and Wholeheartedness,’ 40.

9 I am not the first to take Frankfurt's theory to be committed to the Stoic idea that a free will can flourish in the inner citadel. In fact, a recent collection of essays on hierarchical motivation theories has precisely this title. (See Christman, J. The Inner Citadel: Essays on Individual Autonomy [Oxford: Oxford University Press 1989]Google Scholar.) Annette Baier also writes that in developing her own views she has ‘departed from both Spinoza and Frankfurt … in rejecting the thoroughgoing fatalism which collapses the actual into the necessary, and the normatively demanded into the one thing possible, at all levels of reflectiveness’ (Baier, ‘Comments on Frankfurt,’ Synthese 53 (1982), 288–289)Google Scholar, a decidedly late Stoic reading of both of them.

10 Whether hierarchical motivation theories should incorporate some kind of historical condition is a large and vexed issue which is too complex to tackle here. (For some of the relevant literature, see note 4.) Suffice it to say that Frankfurt has hitherto been reluctant to take this step for reasons which are not entirely clear. He may share the suspicion of some philosophers (Double, R. The Non-Reality of Free Will [Oxford: Oxford University Press 1991], 56–61)Google Scholar that it is harder, perhaps even impossible, to defend a compatibilist version of a hierarchical motivation theory which makes a serious concession to history. However, other philosophers argue that this is not a good reason to resist history, because compatibilism has the resources to tame it. (See, for example, Mele, 158-9.) However that may be, Frankfurt should not resist the version of the restricted PAP defended in the text, because it is an historical condition in a very weak sense which falls far short of any of the historical constraints on preference-acquisition which he rejects. Like his unsupplemented hierarchical theory, it places the principal emphasis on what an agent who has been manipulated in the past does now at time t+1 to deal with it and why he strives for wholehearted identification now. I clarify the dispute between structuralists and historicists about autonomy and what it would take to resolve it in ‘Doing and Time.'

11 Fischer, The Metaphysics of Free Will (Oxford: Blackwell 1994)Google Scholar

12 Fischer's condition is quite acceptable, with whatever qualifications are necessary to deal with unintentional action, which is action proper. If I unintentionally shoot you, mistaking you for a deer in the underbrush, I have shot you, not merely made some movements. Thus, some pro-attitude or volition will enter into the causal provenance of my action even though specification of its content will not make essential reference to ‘my shooting you’ but rather (say) to ‘my shooting the deer making that rumpus in the underbrush.'

13 I include this observation to allay a worry registered by an anonymous referee for the Canadian Journal of Philosophy.

14 I pursue this theme in greater detail in ‘Tie Me Up, Tie Me Down: The Freedom and Unfreedom of Person-Parametric Decisions,’ forthcoming. See especially the case of ‘Pran.'

15 Wertheimer, Alan defends an account of duress as justification in his Coercion (Princeton: Princeton University Press 1987), 166-9.Google Scholar

16 I emphasize that the sketch offered in the text is just that, merely a sketch. There are many complications in a purely first-order analysis of the excusing force of compliance with coercion which we cannot consider here. However, to blunt one immediate source of concern, I will note that such an account does not entail that harsh, credible threats always relieve the recipient of moral responsibility for complying. For example, I would have a hard time getting anyone to excuse my having killed a man who reneged on his debt to a loan shark by explaining that I did it solely because Sammy the Bull threatened me with serious harm if I did not comply. (Suppose he carefully explained to me something like: ‘Kill the welshing bastard or I'll break both your arms and legs … slowly.’) It is perfectly compatible with a first-order account that it require that P coerces Q in a responsibility-relieving way only if Q is morally permitted to comply with the terms of P’ s threat. This neither reduces the non-blameworthiness of coerced actions to justification nor departs from the main line of a first-order account of the distinctly excusing force of compliance to some harsh, credible threats. (I include this observation to allay the worries of an anonymous referee for the Canadian journal of Philosophy.) For appropriately detailed accounts of first-order theories of coercion, see Nozick, R. ‘Coercion,’ in Morgenbesser, S. etal., eds., Philosophy, Science and Method (New York: St. Martin's 1969)Google Scholar; Feinberg, J. Harm to Self (Oxford: Oxford University Press 1986)Google Scholar; and Zimmerman, D. ‘Coercive Wage Offers,’ Philosophy and Public Affairs 10 (1981) 121-45.Google Scholar

18 Frankfurt cites Dworkin's ‘suggestive discussion of these issues’ (in ‘Coercion and Moral Responsibility,’ note 16), but in doing so mistakenly runs together two different ideas: 1) not wanting to be in a situation in which one is forced to act on a certain first-order desire, and 2) having a second-order desire not to have a certain first-order desire. Dworkin's account of acting unfreely makes reference only to 1), whereas Frankfurt's central condition for non-autonomy also involves 2).

19 I offer such an account in ‘Coercive Wage Offers.'

20 Hare, R.M. The Language of Morals (Oxford: Oxford University Press 1952)Google Scholar and Freedom and Reason (Oxford: Oxford University Press 1963); Gewirth, A. Reason and Morality (Chicago: University of Chicago Press 1978)Google Scholar

21 To be sure, the Stoic vision of liberation from coercion is not thereby shown to be incoherent on its own; for that we also need at least the hierarchical theory of the unfreedom of coerced action. But they were made for each other; in fact, since the latter entails the former, the existence of the paradoxes also entails the incoherence of this aspect of Stoicism, a result to be applauded.

22 Slote's ‘Understanding Free Will’ is much-cited in the literature on autonomous agency, but not much of it is much-cited, for commentators only mention the brief last section in which Slote urges that hierarchical motivation theories must supplement their structural condition with a historical constraint on desire-acquisition. He may well be correct about this, but there is an unappreciated irony in his urging this course on Frankfurt, because an historical condition would enable Frankfurt to escape from the very Stoic encumbrances which Slote in the bulk of his essay urges him to embrace!

23 In preparing this paper, I have been helped by comments from Ishtiyaque Haji, Martin Hollis, Alfred Mele, and two anonymous referees for the Canadian Journal of Philosophy.