1 See, for instance, ‘The Modal Argument for Incompatibilism,’ Philosophical Studies 53 (1988), 227-44; ‘Freedom, Necessity, and Laws of Nature as Relations between Universals,’ Australasian Journal of Philosophy 68 (1990), 371-81; ‘Freedom, Causation, and Counterfactuals,’ Philosophical Studies 64 (1991), 161-84; ‘Stop Me Before I Kill Again,’ Philosophical Studies 75 (1994), 115-48; ‘Free Will Demystified: A Dispositional Account,’ Philosophical Topics 32, Agency, (2004), 427-50; and ‘Compatibilism, Incompatibilism, and Impossibilism,’ in Contemporary Debates in Metaphysics, Theodore Sider, John Hawthorne, and Dean Zimmerman, eds., Oxford: Blackwell 2008.

2 ‘Freedom, Foreknowledge, and the Principle of Alternate Possibilities,’ CanadianJournal of Philosophy 30 (2000), 1-24

3 In this paper, I will be using ‘was able to do otherwise,’ ‘was free to do otherwise,’ and ‘could have done otherwise’ interchangeably. I follow Frankfurt in making no assumptions about what we mean, or should mean, when we use these locutions. (But see notes 6, 7 and 10.)

4 A ‘back-tracking’ argument is one with the following pattern of reasoning: ‘ If the present were different in way D, then the past would have been different in way E; if the past were different in way E, then the future would be different in way F; therefore, if the present were different in way D, the future would be different in way F.’ Although the argument I criticized is a ‘back-tracking’ argument, I did not use that term in my CJP 2000. I will say more of this later.

5 ‘Freedom, Foreknowledge, and Frankfurt: A Reply to Vihvelin’

6 I understand the ability/opportunity distinction in the way that it's understood in ordinary English. When we say that someone has the ability but not the opportunity to do something, we are invoking a contrast between a person's contribution to the facts that enable her to do something and the contribution made by the person's environment or situation. More precisely, the contrast is between the contribution made by the intrinsic properties of a person and the contribution made by the extrinsic or relational properties of a person. To have the ability to do something, is, roughly, to have ‘what it takes’ to do that thing (e.g., bike-riding skills and unbroken limbs and properly working brain and…); to also have the opportunity is to be in a situation which provides you with what you need in order to exercise your ability (e.g., a bike and…) and in which nothing extrinsic to you would prevent you from exercising your ability. (e.g., no prison walls or chains or Black lurking in the background). In saying that the ability/opportunity distinction is embedded in commonsense, I am not claiming that commonsense can settle the question of whether we really have the abilities and opportunities we think we have or whether these abilities and opportunities are compatible with determinism.

7 ‘Ability’ is sometimes used in a weaker way, to mean only that someone has the know-how, skills, or competence, required to do something. (In the literature, this is sometimes called a ‘general ability.’) Given this weaker sense of ‘ability,’ a person with a broken leg retains the ability to ride a bicycle. It should be fairly obvious that Black doesn't rob Jones of his bike-riding ability in this weak sense. What is less obvious, but also true, is that Black doesn't rob Jones of his ability to ride his bicycle in the stronger sense that we have in mind when we say, on a particular occasion, that a person has the ability to ride a bicycle: in addition to having bikeriding skills and competence, it is also true that the person has ‘what it takes’ to exercise those skills, then and there. (See note 6.)

8 The story I just told is not intended to be a Frankfurt story. A Frankfurt story is a story where i) we are supposed to agree that a person could not have done anything other than what he actually did; and ii) we are supposed to have the intuition that he is nevertheless morally responsible for something that he did. My story doesn't satisfy the first criteria. Black's interest in Jones is limited to his desire that Jones not ride his bike that year, so he leaves Jones as free as he ever was to perform other actions, including mental acts like deliberating about bike-riding, deciding to ride his bike, and so on. My aim in telling this story is not to make any point about moral responsibility, but, rather, to make the following point about counterfactuals and ability: we are not entitled to infer, from the fact that someone wouldprevent you from doing X, that you in fact lack the ability to do X. It may be that you retain the ability, and only lack the opportunity. And, for all that we've seen, it may be that you retain both ability and opportunity. I will say more of this latter possibility later.

9 Harry Frankfurt, ‘Alternate Possibilities and Responsibility,’ Journal of Philosophy 66 (1969).

10 I say ‘unfortunately’ because Frankfurt's choice of name has led some philosophers to suppose that the commonsense platitude that PAP is supposed to capture is to be understood in terms of a person's ‘alternate possibilities.’ But of course alternate possibilities need not be actions, let alone actions that the person has either the ability or the opportunity to do. Recognition of this has led to talk of ‘robust’ alternatives, or ‘genuine access’ to alternatives, but these locutions are metaphorical and unhelpful. We can avoid this kind of confusion if we understand PAP as saying that a person is morally responsible for what he has done only if the person could have intentionally acted in some alternative way, and if we understand ‘could’ as ‘was able to,’ leaving it open whether this should be understood as ‘had the ability’ or ‘had the ability and also the opportunity.’

11 For a classic statement and defense of the so-called ‘Conditional Analysis’ of ‘could have done otherwise,’ see G.E. Moore, ‘Free Will’ in his Ethics (Oxford: Oxford University Press 1912). For critique of the Conditional Analysis, and an alternative compatibilist account, see Keith Lehrer, ‘“Can” in Theory and Practice: A Possible Worlds Analysis,’ in Action Theory, M. Brand and D. Walton, eds. (Dordrecht: Reidel 1976). For more references, see note 16.

12 Ibid., 8

13 ‘Freedom, Foreknowledge, and the Principle of Alternate Possibilities,’ this Journal, ibid.

14 If it makes sense to talk of the opportunity to begin or try to act, then Jones retains the opportunity to begin or try to act contrary to Black's plan.

15 I don't mean to imply that I take back what I said in my CJP 2000 about free indeterministic agents. It is controversial whether counterfactual intervention is possible where Jones is a free indeterministic agent, but I continue to insist that if it ispossible, it does not work as advertised.

16 Due to the enormous influence of Frankfurt, this view is now a minority view, even among those philosophers who call themselves compatibilists. But see David Lewis, ‘Are We Free to Break the Laws?’ Theoria 47 (1981) 113-21; Michael Smith, ‘A Theory of Freedom and Responsibility,’ in G. Cullity, ed., Ethics and PracticalReason (New York: Clarendon Press 1997); Michael Smith, ‘Rational Capacities,’ in Ethics and the A Priori: Selected Essays on Moral Psychology and Meta-Ethics (New York: Cambridge University Press 2004); Joseph Keim Campbell, ‘Compatibilist Alternatives,’ Canadian Journal of Philosophy 35 (2005) 387-406; Kadri Vihvelin, ‘Free Will Demystified: A Dispositional Account,’ ibid, and Michael Fara, ‘Masked Abilities and Compatibilism,’ Mind, forthcoming.

17 Fischer, 335

18 Fischer, 334-5

19 This is not intended as either a definition or a necessary condition of opportunity but as defeasible evidence of opportunity. And it states a condition stronger than we ordinarily think is required for opportunity; we would ordinarily grant that Jones has the opportunity, as well as the ability, to ride his bike if we believe that if he tried to ride it, he might succeed, or would have a reasonably good chance of succeeding (or something like that).

20 David Lewis, ‘Counterfactual Dependence and Time's Arrow,’ Nous 13 (1979), 455-76

21 Lewis calls this ‘the standard resolution’ (Lewis, ibid.), but that's because he wants to use counterfactuals to analyze causation, and because he ambitiously hopes to do this without making any assumptions about causation; in particular, he does not want to assume that the standard resolution for evaluating counterfactuals is the one appropriate for causal counterfactuals. If we reject this ambitious aim, then we can arrive at Lewis's widely accepted similarity metric for ‘standard’ counterfactuals by a different route: we can say that our default or standard way of evaluating counterfactuals is the way that is appropriate for causal counterfactuals, and Lewis's similarity metric is the right one for causal counterfactuals. See note 23.

22 This is Lewis's example. Lewis says: ‘It is right to say either, though not to say both together. Each is true according to a resolution of vagueness appropriate to some contexts’ (Lewis, ibid., 34).

23 John Collins, Ned Hall, and L.A.Paul, ‘Counterfactuals and Causation: History, Problems, and Prospects’ in Causation and Counterfactuals,’ Collins, Hall, and Paul, eds. (Cambridge, MA: The MIT Press 2004). See also David Lewis, Counterfactuals (Harvard: Harvard University Press 1973); Lewis, ‘Causation,’ Journal of Philosophy 70 (1973) 556-67; Lewis, ‘Counterfactual Dependence and Time's Arrow’; Lewis, ‘Are We Free to Break the Laws?’ and many other papers.

24 Fischer, 341

25 Fischer, 335

26 The chief additions are my use of the ability/opportunity distinction and my discussion of the uses and abuses of ‘back-tracking’ counterfactual arguments.

27 Fischer, 335

28 Fischer, 337

29 Fischer, 337

30 Fischer, 337

31 Fischer, 338

32 Jonathan Bennett, A Philosophical Guide to Conditionals (Oxford: Clarendon Press 2003), 159-63 and 172-6. Bennett says that the invalidity of Antecedent Strengthening is ‘the securest thing we know about subjunctive conditionals’ (169).

33 Counterfactuals, 32. See also 10-18.

34 Fischer, 338

35 Lewis 1973, 33-4. Whether or not Lewis is right about this is a disputed point in the logic of counterfactuals. See Robert Stalnaker, Inquiry (Cambridge, MA: The MIT Press 1984), 130-2, for a discussion of a putative counterexample due to Pavel Tichy. And see Jonathan Bennett, 298-301, for a discussion of the semantic differ ence between theories of counterfactuals that accept the validity of this argument pattern and those that reject it.

36 Fischer, 338

37 Fischer, 340

38 For a survey and discussion of criticisms of the Conditional Analysis, see Bernard Berofsky, ‘Ifs, Cans, and Free Will: The Issues,’ in The Oxford Handbook of Free Will, Robert Kane, ed. (Oxford: Oxford University Press 2002).

39 I am grateful to Joe Campbell, John Carroll, Charles Hermes, Terrance Tomkow, four anonymous CJP referees, and the commentators at the first Online Philosophy Conference for helpful comments on earlier versions of this paper.