Skip to main content Accessibility help


  • Marcelo Ferrante (a1)


I offer in this paper an argument in support of the orthodox view that resultant luck should not affect judgments of blameworthiness—and so, for example, that we should not blame the successful assassin more than the attempted assassin who equally tries but fails. This view, though widely held among moral philosophers and legal scholars, has been severely challenged as implying either the implausible rejection of moral luck or an equally implausible theory of wrongness according to which actual consequences may play no wrong-making role. The argument I offer, however, assumes both challenges to be true and shows that the orthodox view is consistent with holding them. Indeed, I argue that all other things being equal, successful offenders are no more to blame than their unsuccessful counterparts, even though agents are responsible for what they actually do (and therefore are subject to moral luck), and successful offenders do more wrong than their unsuccessful counterparts do (and therefore consequences do play a wrong-making role). The reason is that the difference in the amount of wrong done by one and the other offender, I show, is counterbalanced by a difference in the degree to which the successful offense and the unsuccessful one are attributable to their respective agents—blameworthiness being a function of both amount of wrong done and degree of attributability.



Hide All

1. The expression is Nagel's. See Thomas Nagel, Moral Luck, reprinted in Mortal Questions (1979).

2. See Michael S. Moore, Placing Blame 191–247 (1997); Zimmerman, Michael J., Taking Luck Seriously, 99 J. Phil.553 (2002).

3. McCarthy, David, Actions, Beliefs, and Consequences, 90 Phil. Stud.57 (1998).

4. Id. at 74.

5. I think that Thomson's objections in her Imposing Risks article still hold after McCarthy's argument. See Judith Jarvis Thomson, Rights, Restitution and Risks 173–191 (W. Parent ed., 1986).

6. I play here with a sentence of Davidson's with which he characterizes basic or primitive actions: “We never do more than move our bodies: the rest is up to nature.” Donald Davidson, Agency, in Essays on Actions and Events 59 (1980).

7. An action φ is a basic action of an agent S if and only if there is no other action φ such that S does φ by doing φ as part of doing φ. See Arthur C. Danto, Basic Actions and Basic Concepts, reprinted in The Body/Body Problem 45–62 (1999).

8. The impact of luck is higher if, say, an agent shoots in a game Russian roulette than if she shoots after checking that the gun is fully loaded. The luckier Alvarez is, the less extraordinary the corresponding case of Borges will be. Depending on the details of the case, Alvarez could be more or less lucky—but lucky she always is.

9. Peter Strawson, Freedom and Resentment, reprinted in Free Will 59 (G. Watson ed., 1982); Fischer, John Martin, Recent Work on Moral Responsibility, 110 Ethics9395 (1999).

10. Copp, David, Defending the Principle of Alternate Possibilities: Blameworthiness and Moral Responsibility, 31 Noûs441, 452 (1997).

11. See Gary Watson, Two Faces of Responsibility, reprinted in Agency and Answerability 260 (2004). Watson's argument is that attributability corresponds to a kind, or a face, of responsibility, the other kind (or face) being accountability. Other philosophers understand responsibility just in terms of attributability; see Fischer, supra note 9, at 96, citing Derk Pereboom as an example. Whether attributability exhausts the conditions of response-worthiness I need not adjudicate here. I content myself with the view that attributability is at least part of the conditions of responsibility.

12. Nozick's framework is introduced in Robert Nozick, Anarchy, State, and Utopia 59–63 (1974); and later elaborated upon in Robert Nozick, Philosophical Explanations ch. 4.III (1981).

13. This is not to say that attributability may not link agents with other items. All that I am saying in regard to the attributability of actions may be applied, duly adjusted, to other things we may claim responsibility for, such as beliefs or emotions. I restrict my focus to responsibility for actions just because my interest in this essay is to elucidate the issue of the relative blameworthiness of agents in success-failure pairs who differ from each other only in what they do.

14. I am assuming a coarse-grained view of the individuation of actions. Under a fine-grained view of the individuation of actions, this point will not be in order. Nothing of importance turns on the choice of the act-individuation approach.

15. I refer throughout to an instance of φ-ing that meets the first two conditions as an action of S's even though it fails to meet the third condition—and so it is not S's own in the strong sense that would warrant a moral response. I do not mean to make any point about the use of the English language; I am just trying to isolate and then focus on the control condition in any action for which an agent may be responsible. Whether what remains after we thus subtract the control condition is something we can still properly call an action is a question for which I have no answer. Maybe the concept of action entails some nonzero degree of attributability so that it makes no sense to refer to a φ-ing as an action of S's if the control condition is not met. Perhaps, as Davidson suggests (see Davidson, Agency, supra note 6), we should say that φ is something S does—a deed of hers—but not an action of hers. Again, I have nothing interesting to say in this respect.

16. O'Connor, Timothy, Indeterminism and Free Agency: Three Recent Views, 53 Phil. & Phenomenological Res.499, 500 (1993). There is also what is often called alternative-possibilities control. Alternative-possibilities control emphasizes that the φ-ing that actually obtains is one among other possible actions actually open to the agent. Roughly, S has alternative-possibilities control over her φ-ing only if she could do other than φ-ing. Even if necessary for responsibility, alternative-possibilities control bears no impact on the relative blameworthiness of the individuals in success-failure pairs—either both or none of them will lack alternative-possibilities control. The same, I show below, is not the case with agent control. Since my interest in elucidating the control condition is limited to what may impact on the judgments of relative blameworthiness in success-failure pairs, I focus on agent control.

17. Id. at 500.

18. See John Martin Fischer & Mark Ravizza, Responsibility and Control (1998).

19. Id., ch. 2 and 63–64.

20. Id. at 38–39.

21. Id. at 38. See also 46–47 for some observations that make the idea of an action-producing mechanism somewhat more precise.

22. Id. at 69–85; for discussion, see Watson, Gary, Reasons and Responsibility, 111 Ethics374394 (2001), reprinted in Gary Watson, Agency and Answerability 289–317 (Clarendon Press, 2004).

23. Fischer & Ravizza, supra note 18, at 41–42, 69–76. Fischer and Ravizza further distinguish reactivity (as the capacity to not-φ out of reasons to not-φ) from receptivity (understood as the capacity to recognize reasons in the world). I need not make that distinction here.

24. I draw the term from Alfred Mele's and Paul Moser's analysis of intentional action. Alfred R. Mele & Paul K. Moser, Intentional Action, reprinted in The Philosophy of Action (Alfred R. Mele ed., 1997).

25. Id. at 224.

26. Mele & Moser, in contrast, offer the case as an example of a doing that is “too coincidental to count as intentional.” Id. at 225.

27. As the contrast between these two cases shows, reliability (and hence agent control) may depend on knowledge (or justified belief). But it may also depend on skill. Take the archery examples I gave above. It may very well be the case that the beliefs involved in my hitting the bull's-eye in my first shot—in aiming at the target and releasing the arrow, say—are the same as the instructor's beliefs involved in her hitting the bull's-eye. And there is still an obvious difference in terms of reliability and hence in agent control. That is what we call skill. The instructor is a skilled archer, which means that when acting from her normal action-producing mechanism, she exhibits high-reliability measures in making her actions meet the archery-relevant reasons for action—like the reasons there might be for hitting the bull's-eye. See id. at 252–253. For Mele & Moser's notion of skill, see id. at 246.

28. I am thankful to an anonymous referee for helping me see that the arguments I make in this section were called for.

29. See Nagel, supra note 1, at 26–28.

30. I thank Mercedes Etchemendy for helpful comments clarifying this point.

31. See Jonathan Dancy, Practical Reality 56–60 (2000) at 59.

32. For this relation between acting intentionally and acting for a reason, see, e.g., Donald Davidson, Actions, Reasons, and Causes, reprinted in Essays on Actions and Events 6 (Clarendon Press, 1980); Robert Audi, Acting for Reasons, reprinted in The Philosophy of Action 75–105 (Alfred R. Mele ed., 1997).

33. See, e.g., Mele & Moser, supra note 24, at 240–241.

34. Supra note 12.

35. Michael Zimmerman, for one, holds both that blameworthiness does not depend on wrongness and that attempts to do what is wrong are not themselves wrongful. See Michael J. Zimmerman, supra note 2, at 560–561; and Zimmerman, Michael J., A Plea for Accuses, 34 Am. Phil. Q.229243 (1997).

36. This view of what it is for which a person deserves blame is defended in Copp, supra note 10, at 448–451.

37. Walen, Alec, The Doctrine of Illicit Intentions, 34 Phil. & Pub. Aff.3967 (2006). Walen draws a distinction between wrongness and impermissibility that I need not make here. As I use the terms in this paper, “morally wrong” and “morally impermissible” are synonymous expressions.

38. Id. at 66–67.

39. Id. at 39. This is a very rough statement of Walen's argument. See id. at 50–56.

40. In one and the other case, Borges has different beliefs, which lead him to engage in different courses of actions in order to carry out his intention to kill Victor. Given this intention to kill Victor and the different sets of beliefs, Borges is moved to form different, more specific intentions regarding what to do in order to kill Victor. At the level of the chosen means, then, there will be different intentions.

41. See, e.g., Joshua Dressler, Understanding Criminal Law 370–372 (2d ed. 1995).

42. See Thomson, supra note 5, at 173–191.

43. Id. at 184–188.

44. I owe this objection to an anonymous referee.

* I would like to thank Amalia Amaya, Mitchell Berman, Jorge Cerdio, Lewis Kornhauser, Larry Laudan, Stephen Morse, Francisca Pou, Eduardo Rivera-López, Scott Shapiro, Seana Shiffrin, Horacio Spector, and anonymous referees for helpful comments on earlier drafts.


  • Marcelo Ferrante (a1)


Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed