Responsibility, reason, and irrelevant alternatives

Philosophy and Public Affairs

Volumn 28, Issue 3, 1999, Pages 205-241

  Hurley, Susan L ^a  

^a NONE

Author keywords

[No Author keywords available]

Indexed keywords

EID: 0141550082     PISSN: 00483915     EISSN: 10884963     Source Type: Journal    
DOI: 10.1111/j.1088-4963.1999.00205.x     Document Type: Article

References (61)

1. John Martin Fischer, The Metaphysics of Free Will: An Essay on Control (Oxford: Blackwell, 1994), chaps. 7, 8;
2. also Martha Klein, Determinism, Blameworthiness and Deprivation (Oxford: Clarendon Press, 1990).
3. Michael Otsuka, "Incompatibilism and the Avoidability of Blame," Ethics 108 (1998): 685-701.
4. Fischer, Metaphysics of Free Will;
5. John Martin Fischer and Mark Ravizza, Responsibility and Control: A Theory of Moral Responsibility (Cambridge: Cambridge University Press, 1998);
6. cf. Susan Wolf, Freedom Within Reason (New York: Oxford University Press, 1990);
7. and R. Jay Wallace, Responsibility and the Moral Sentiments (Cambridge: Harvard University Press, 1996).
8. Harry G. Frankfurt, "Alternate Possibilities and Moral Responsibility," Journal of Philosophy 66, no. 23 (1969);
9. "Freedom of the Will and the Concept of a Person," Journal of Philosophy 67, no. 1 (1971): 5-20;
10. "Coercion and Moral Responsibility," in Essays on Freedom of Action, ed. Ted Honderich (London: Routledge & Kegan Paul, 1973), pp. 65-86;
11. "Identification and Externality," in The Identities of Persons, ed. Amelie Rorty (Berkeley: University of California Press, 1976);
12. "Identification and Wholeheartedness," in Responsibility, Character, and the Emotions, ed. Ferdinand Schoeman (Cambridge: Cambridge University Press, 1987), pp. 27-45.
13. See David Shatz, "Free Will and the Structure of Motivation" Midwest Studies in Philosophy 10 (1986): 451-82, Sect. VII, for a discussion that locates contemporary reasonsensitivity accounts of responsibility in the broader context of contemporary approaches to responsibility.
14. Fischer, Metaphysics of Free Will, and others;
15. Klein, Determinism, also draws this distinction.
16. See, for example, Klein, Determinism.
17. At some points, Wolf's criticisms of the Real Self View may seem to commit her to a regression requirement. But since Wolf does not ultimately accept a regression requirement, her criticisms of the Real Self View should be either recast or reinterpreted so that they clearly do not depend on a regression requirement.
18. This assumption might be unwarranted if objective reasons for acting can come into ineliminable conflict with each other, so that a correct view of how to resolve the conflict does not eliminate the conflict or subsume the force of the overridden reason without remainder. (I argue that this is possible in S. L. Hurley, Natural Reasons: Personality and Polity [New York: Oxford University Press, 1989], chap. 7.) Where there is such a conflict, someone might be weak-willed and do the wrong thing, all things considered, but nevertheless act for the conflicting objective reason.
19. Although, as indicated, her view in effect expresses an asymmetrical form of the irrelevant alternative intuition, which I claim is also at work in Frankfurt cases.
20. Fischer, "Responsiveness and Moral Responsibility," in Responsibility, Character, and the Emotions, ed. Ferdinand Schoeman;
21. Fischer, Metaphysics, p. 208.
22. In Fischer and Ravizza, Responsibility and Control, Fischer revises his 1994 view in various ways. On the revised view, weak reason-responsiveness is too weak a condition for blameworthiness. See note 20 below.
23. This and other varieties of reason-responsiveness are explained further in the following section; see Fischer, Metaphysics, p. 244n;
24. Fischer and Ravizza, "Responsibility and Inevitability," Ethics 101 (1991): 258-78, p. 277;
25. Fischer and Ravizza, Responsibility and Control.
26. Metaphysics of Free Will, pp. 151, 205;
27. though Fischer appears to have changed his mind about this in Fischer and Ravizza, Responsibility and Control.
28. See Klein, Determinism, who argues this, on the assumption that we are not responsible for the causes of our acts. The regression requirement can be regarded as a generalization of Klein's U-condition, though she does not use the term "regressive."
29. "Even an agent who acts against good reasons can be responsive to some reasons" (Fischer, Metaphysics of Free Will, p. 167). Is this because the reasons in question need not be objective, or because the responsiveness in question needs only to be loose? Does pluralism about reasons allow that there are bad but objective reasons? In one example, Sam's reasons for killing are bad reasons, but "they are his reasons," and Sam is viewed as responsible and blameworthy
30. (Fischer and Ravizza, "Responsibility and Inevitability," p. 258). If the counterfactual intervenor tracked objective reasons tightly, but these were not the agent's subjective reasons, would the alternate-sequence mechanism count as reason-responsive? For more recent discussion,
31. see Fischer and Ravizza, Responsibility and Control. Also relevant here are the questions raised in note above: can objective reasons for action come into ineliminable conflict? Can it be the case that when someone acts in a weak-willed way against his correct better judgment he nevertheless acts on and for an objective reason?
32. However, Fischer and Ravizza, "Responsibility and Inevitability," did endorse a different asymmetry, between responsibility for acts and for omissions. On their past view, responsibility for an act does not require the freedom to refrain from performing the act, while responsibility for failure to perform an act does require the freedom to perform (p. 271;
33. cf. Fischer, Metaphysics).
34. They argued that since omissions do not involve "actual causal control" of movement, but responsibility does require reason-responsive control in some form, responsible omissions must instead involve "regulative control," an alternate sequence requirement of outright freedom to perform. In their 1998 book Responsibility and Control they have changed their minds. The earlier view can be corrected as follows. Reason-responsive control involves dispositional properties of operative mechanisms in the actual sequence of events. It requires that a different response would be made under different conditions, for example, if the reasons were different or if the circumstances different and so differently related to the same reasons. But it does not require an alternate possible sequence: that a different response could be made, holding all else constant. These points apply to omissions as much as to acts. If the operative mechanisms when you omit to act are so disposed that you would have acted under different conditions, and these dispositions characterize a reason-responsive control system, then your omission may nevertheless count as reason-responsive and as responsible. For example, you may be responsible for omitting something you should have done in virtue of reason-responsive dispositions characteristic of weak will, or of dedication to evil. For further discussion, see Sects. III and IV below.
35. Fischer and Ravizza (Responsibility and Control) no longer require for responsibility for omissions a kind of control that involves alternate possibilities. However, they still think that there is such a kind of control, indeed the kind typically associated with moral responsibility, which they call "regulative control" (pp. 20, 24, 37, 338). By contrast, I hold that no kind of control requires the outright possibility of an alternate sequence, all else constant (see note 26 below and also
36. Wallace, Responsibility and the Moral Sentiments, pp. 86-87, 181, 189-90, 220-21, 262-63, who argues for powers of reflective self-control as conditions of accountability and that such powers are compatible with determinism). A different and more promising asymmetry, which Fischer's view does support, is one between talents and handicaps with respect to reason-responsiveness. Omissions to act in worthwhile ways are often explained not by the operation of mechanisms that are reason-responsive in these blame-supporting ways, but rather by sheer lack of reason-responsiveness. Sheer lack of reason-responsiveness can itself be regarded as a kind of handicap, and reason-responsiveness as a generic talen. People handicapped by lack of reason-responsiveness may make poor quality decisions, both to act and to omit, for which they are not responsible, even though talented, reason-responsive types are responsible for their acts. This point may be relevant to the roles of responsibility in theories of justices
37. He uses the terms "strong" versus "weak" reason-responsiveness (Metaphysics, pp. 164ff).
38. See ibid., pp. 165-67. As explained in the text, Wolf's view is not a pure actual sequence, reason-responsiveness view; her ability condition makes alternate sequence requirements. She thinks that it is possible to be able to act in accord with the True and the Good, but fail to do so nonetheless, thus that weakness of will is possible (Freedom within Reason, pp. 88-89).
39. Further consideration of weakness of will suggests one way to do this (cf. Fischer and Ravizza, Responsibility and Control, chap. 3). Distinguish all-things-considered reasons to act ("It's the right thing to do") from specific reasons to act ("That would be the kindest thing to do, though not necessarily the fairest"). Assume that the latter may conflict, even if they are objective reasons (see
40. Hurley, Natural Reasons, chaps. 7, 8 on pro tanto reasons). A weak-willed agent does not act in accord with the reasons she has, all things considered, but she does act for one of the conflicting specific reasons she has. Suppose distinct specific reasons conflict over whether to do x or y, the agent deliberates and judges correctly that all-things-considered the right thing to do is x, and instead she does y. Moreover, she does y for the specific reason there is to do it, even though her all-things-considered deliberated judgment to the contrary took account of this reason. In this familiar type of situation, intermediate reason-responsiveness might require the mechanism on which the agent acts not just to respond to reason in some other possible worlds, but to be responsive to deliberation in the face of conflicting reasons in some other possible worlds. More specifically, it could require the mechanism on which the agent acts to satisfy various combinations of the following conditions: (a) that it responds to specific reasons (of the types in play in the actual world) in some other worlds where reasons of those types obtain; (b) that it responds to all-things-considered, deliberative reason in some other worlds where specific reasons conflict; (c) that it responds to all-things-considered, deliberative reasons in some other worlds when specific reasons of the types in play in the actual world obtain and also conflict. Each of these conditions could be strengthened to require 'most other worlds' rather than 'some other worlds.' (See and cf.
41. Tony Honore: "a person guilty of fault must have, besides a general capacity for decision and action, the ability to succeed most of the time in doing the sort of thing which would on this occasion have averted the harm." "Responsibility and Luck," Law Review Quarterly [1988]: 530-53, p. 531; see also 550-51.)
42. Fischer is sensitive to worries about how mechanisms are individuated, and to parallel worries in epistemology ("Responsiveness and Moral Responsibility," pp. 81-106; Metaphysics;
43. Fischer and Ravizza, Responsibility and Control). For present purposes we simply help ourselves to his use of the idea of the dispositional character of the mechanism that actually operates when an agent acts. See discussion in Sect. IV.
44. Fischer and Ravizza, Responsibility and Control, pp. 38-40, 40n, 46-47;
45. see also Wallace, Responsibility and the Moral Sentiments, p. 190n, on mechanism worries.
46. This is the kernel of truth in Wolf's asymmetry thesis. However, it turns critically on the tightness of the reason-responsive disposition; a looser formulation will miss Wolf's point (as in Fischer's weak reason-responsiveness; cf. Fischer and Ravizza, "Responsibility and Inevitability";
47. Fischer, Metaphysics, pp. 156ff).
48. Perhaps she also satisfies an intermediate requirement of responsiveness: she may be generally responsive to the results of rational deliberation about what should be done, all things considered, when objective reasons conflict.
49. Someone could be both evil and weak-willed, but that does not raise any different issues. Reason-responsiveness that is both loose and subjective is not considered separately.
50. It may be tempting to ask at this point: when we stipulate that a different mechanism could have operated but did not, was whether it operated or not under the agent's control? But this question runs the danger of confusing an alternate-sequence condition with an actual-sequence control condition. Control of X requires that X would have been otherwise under certain conditions, not that it could have been otherwise all else constant. Whether a different mechanism could have operated is a question about an alternate sequence, about outright possibility, all else constant. But control, like reason-responsiveness, involves dispositional properties of the actual sequence. If the agent controls which mechanism operates, then the agent maintains the state of affairs with respect to mechanism-operation at some target value in the face of exogenous disturbances, by adjusting her behavior accordingly to compensate for those disturbances (see for example William T. Powers, Behavior: The Control of Perception [Chicago: Aldine, 1973]). Control depends on dispositions to respond so as to maintain the target value under various counterfactual conditions. Moreover, if you are controlling something, you only do otherwise if circumstances are different: different exogenous disturbances demand different responses to maintain the target value of what you control.
51. (Fischer and Ravizza, Responsibility and Control, reinvent some of the wheels of control theory, e.g., at p. 120.) If your variation in behavior is independent of those conditions, you are not maintaining the target value, hence not controlling whatever takes that value; target maintenance is essential to control. (Hence what Fischer and Ravizza, op.cit., still call 'regulative control' is not a kind of control at all.) You may in turn control the target value, but that just pushes the conditional-dispositional character of control back a step and begins a control hierarchy. And similar points apply to control of the target value and to control hierarchies. Meta-control no more requires alternate possibilities than does simple control
52. (cf. Fischer and Ravizza, Responsibility and Control, reinvent some of the wheels of control theory, e.g., at op. cit., p. 31). Nevertheless, the dispositional character of control is compatible with indeterministic causation; control is neutral with respect to determinism
53. (Randolph Clarke, "Indeterminism and Control," American Philosophical Quarterly 32, no. 2 [1995]: 125-37);
54. cf. Otsuka, "Incompatibilism," Part I.
55. But, again, since Wolf's view is not a pure actual-sequence view this remark does not strictly apply to her. See note 19 above.
56. Is this just a new way of elaborating the old intuition that indeterminism cannot make for responsibility?
57. For the role of this supposition in Frankfurt's work, see, e.g., "Freedom of the Will and the Concept of a Person," p. 19;
58. "Alternate Possibilities and Moral Responsibility," p. 837. This article can be regarded as emphasizing and trying to explain further those few passages in Frankfurt in which the "wouldn't have done otherwise even if could have" formulation appears.
59. Causal mechanisms and dispositions can make counterfactuals true in both deterministic and indeterministic worlds (see Clarke, "Indeterminism and Control," for related points).
60. This phrase suggests an analogy to intuitions supporting independence or sure-thing axioms in rational choice, which would be interesting to pursue but which is not developed here.
61. See Klein, Determinism, on the distinction between regression and could-have-done-otherwise principles.