Moral status and the impermissibility of minimizing violations

Philosophy and Public Affairs

Volumn 25, Issue 4, 1996, Pages 333-351

  Lippert Rasmussen, Kasper ^a  

^a NONE

Author keywords

[No Author keywords available]

Indexed keywords

EID: 34248125570     PISSN: 00483915     EISSN: 10884963     Source Type: Journal    
DOI: 10.1111/j.1088-4963.1996.tb00046.x     Document Type: Article

References (37)

1. The appropriateness of the phrase "agent-relative restrictions" comes from the fact that the reasons for action in which these restrictions consist have a strength that depends on who performs the violation. I have a strong reason not to kill an innocent, but a reason of much less strength to prevent others from killing an innocent. Hence, the fact that by violating a restriction I will prevent, say, five similar violations by others does not mean that I have reason, all things considered, to perform that violation.
2. Samuel Scheffler, The Rejection of Consequentialism (Oxford: Clarendon Press, 1982), p. 5.
3. The kinds of killings I have in mind here are killings done by persons (as opposed to killings that result from natural threats, e.g., "the plague killed thousands").
4. Robert Nozick, Anarchy, State, and Utopia (Basil Blackwell: Oxford, 1974), p. 30.
5. Samuel Scheffler discusses the problem in more detail in Scheffler, Rejection,
6. and in "Agent-Centred Restrictions, Rationality and the Virtues," Mind 94 (1985): 409-19:
7. reprinted in Scheffler, ed., Consequentialism and Its Critics (Oxford: Oxford University Press, 1988), pp. 243-60.
8. See Scheffler, "Agent-Centered Restrictions."
9. Note that this explanation of the term does not determine which violations qualify as minimizing in Prisoner Dilemma-like situations (i.e., it is true of each of us that whatever the other does each will violate fewer persons if he does A rather than B. However, if we both do A each of us will end up violating more persons than if we both had done B). I ignore this complication, since it is irrelevant for my present purposes.
10. Scheffler, Rejection, p. 99.
11. See Frances M. Kamm, "Harming Some to Save Others," Philosophical Studies, 57 (1989): 227-60;
12. and Kamm, "Non-consequentialism, the Person as an End-in-Itself, and the Significance of Status," Philosophy & Public Affairs, 21, no. 4 (Fall 1992): 354-89.
13. The DDA, as I shall here construe it, says that harming another is morally wrong. The DDE says that conducting oneself with an intention that harm comes to others is morally wrong.
14. Now, there are situations in which it is permissible, according to commonsense morality, to harm a person and impermissible, according to utilitarianism, to harm that person (e.g., punishment of a criminal that does not maximize aggregate utility). This means that it is crucial (for someone who defends commonsense morality through considerations about inviolability) to say something about how we identify the kinds of circumstances that make up the metric of inviolability. It is not clear how this can be done. (To deal with the problem about punishment, one could say that only the impermissibility of harming innocents is relevant to inviolability and then hope that there are no kinds of cases in which commonsense morality permits and utilitarianism forbids harming innocents.) In fact, it is not even clear that one can say that, all things being equal, the impermissibility of minimizing violations means a higher degree of inviolability. To do so presupposes that minimizing violations (perhaps subdivided into different kinds of minimizing violations differing in terms of the amount of badness prevented) is included in the metric for inviolability. But that requires an argument. This problem, I believe, threatens to unravel The Inviolability Account (as well as the The Independence Account that I shall later propose) from the very beginning. However, I shall continue as if our intuitive sense that the impermissibility of minimizing violations makes us more inviolable/independent can be cashed out. At any rate, since my conclusion in this article is a critical one, it does not require a solution to this problem.
15. Kamm, "Non-consequentialism," p. 383.
16. Kamm, "Non-consequentialism," Ibid., p. 386.
17. It becomes clear in item 7 below why the example is misleading.
18. This supposition might be false. It may lead to more violations if we choose the moral system that permits minimizing violations. However, nothing in Kamm's argument hinges on this and I henceforth ignore it.
19. Note that if agents impermissibly performed minimizing violations, we would have the benefit of being inviolable in the pertinent respect and yet enjoy the higher chances of survival that only less dignified beings - i.e., beings who can permissibly be violated to save more from being violated - can permissibly enjoy.
20. Kamm concedes (in response to the kind of challenge just sketched) that "simple talk about inviolability cannot be all that we need to explain the presence of a constraint" (Kamm, "Non-consequentialism," p. 384).
21. Kamm, "Non-consequentialism," Ibid., p. 387.
22. Kamm first discussed inviolability in connection with her Principle of (lm)Permissible Harm (henceforth: "P(I)/PH"). Kamm believes the P(I)/PH accounts for many of our moral intuitions that others think are accounted for by either the DDA or the DDE. The P(I)/PH says that: "(i)t is permissible to cause harm to some in the course of achieving the greater good of saving a greater number of others from comparable harm, if events which produce the greater good are not more intimately causally related to the production of harm than they are to the production of the greater good (or put another way, if events which produce the greater good are at least as intimately causally related to the production of the greater good as they are to the production of the lesser harm)" (Kamm, "Harming Some to Save Others," p. 232). I believe there is reason to be skeptical about the P(I)/PH (apart from any doubts one might have about the possibility of providing a satisfactory account of what causal intimacy amounts to). Here is a counterexample to the P(I)/PH: Wagon - A trolley rolls down the hillside. The track forks. The trolley is heading for the left track, where five persons are standing. They will be killed unless a bystander switches the trolley to the right track. Unfortunately, one person is standing on the right track and the trolley will, if diverted, kill him. Diverting the trolley will also cause a cargo wagon to detach from the trolley and continue down the left track. Fortunately, the bystander can press yet another switch which will erect, on the left track in front of the five, a barrier that can stop the wagon but not the trolley and the wagon. I suggest that it would be permissible to kill the one by diverting the trolley in Wagon. If so, there is something wrong with the P(I)/PH (as far as its intuitive fit is concerned), since the death of the one in Wagon has a more intimate causal relation with the turning of the trolley than the turning of the trolley has with the saving of the five. After all, the five are not safe simply because the trolley is diverted.
23. Following Shelly Kagan (The Limits of Morality [Oxford: Oxford University Press, 1989, p. 84]),
24. say that one countenances harm to another whenever another is harmed and one could have prevented that from happening (Shelly Kagan (The Limits of Morality ibid., p. 84).
25. In effect, Jennifer's deed reminds us of two facts. First, a minimizing violation of a constraint need not be intentional under that description for it to be a minimizing violation. Second, the impermissibility of a minimizing violation need not be the impermissibility of a minimizing violation intended under that description for its impermissibility to add to our inviolability.
26. I called this schematic case "the standard case," since there are exceptions. One exception is this: some gangsters have set a building ablaze. Philippa can save the many people who are trapped inside the building only by turning on the sprinkler system. Unfortunately, doing so will drown Andy, who has taken refuge from the fire in the cellar, where he would otherwise have survived. In this nonstandard case it is not true that if Philippa does not perform a minimizing violation, i.e., turns on the sprinkler thereby killing Andy, there might be no violations. Still, one might think it highly significant that in nonstandard cases the intentions of the villains were alterable. Thus, in the case at hand there was a time such that even if Philippa had not done what turned out to be a minimizing violation, there need not have been any violations at all. Full compliance with the demands of morality is a possibility in standard and nonstandard cases alike.
27. Strictly speaking, there could be situations in which an agent can perform a minimizing violation even though everyone complies with the demands of morality (e.g., through no fault of my own, I have got into a situation where, whatever I do, I will either violate V1 or violate V2-V6. Violating V1 would qualify as a minimizing violation). I ignore such cases below, because I do not think commonsense morality has any quarrels with this species of minimizing violations.
28. For a related point, see Peter Strawson's classical paper "Freedom and Resentment," repr. in Gary Watson, ed., Free Will (Oxford: Oxford University Press, 1982).
29. At this point, I am entertaining the idea that the difference between intention-dependent and intention-independent truths provides a rationale in the Schefflerian sense of the impermissibility of minimizing violations. Nancy Davis seems to be making a similar suggestion in "Utilitarianism and Responsibility," Ratio, 22 (1980): 21.
30. Why not say that V2-V6's being harmed results from A1's refraining from harming V1? Surely, the fact that A1 did not harm V1 is part of the causal explanation of V2-V6's being harmed. I use the phrase "x results from y's conduct" in a more narrow and moralized sense here. If x would not have occurred under full compliance, and x occurs, and if y is the one whose frame of mind is such that full compliance is impossible, then x results from y's conduct. Anyway, nothing substantial hinges on this particular usage.
31. I mention this to indicate how The IndependenceAccount explains the time-relativity of constraints (i.e., why one should not violate a constraint now to prevent one's own future and more numerous violations: see Kamm, "Non-consequentialism," p. 382). To keep things simpler I shall ignore the time-relativity of constraints below except in note 28 and when I define independence and unsacrificability.
32. I need this qualification for the following reason. Suppose it is impermissible for A1 to countenance harm to V1-V6 if everyone were fully complying. Suppose now that not everyone is fully complying. A2-A6 will countenance harm to V2-V6 unless A1 countenances harm to V1. If I did not add the qualification "through the will of another," the impermissibility of A1's countenancing harm to V1 would not make persons more independent than its being morally required that A1 countenances harm to V1. In the former case, A1 could permissibly (through the will of A2-A6) countenance harm to V2-V6. In the latter case, A1 could permissibly countenance harm to V1 (though not through others' wills).
33. One can argue in a similar way with respect to the issue of the time-relativity of constraints. Briefly, suppose I should neither violate the DDA now to prevent my more numerous future violations, nor fail to fulfill a duty to save now in order to prevent more future failures to fulfill my duty to save. In that case, it seems that the reason why there is a time-relative reason against violating the DDA has nothing to do with the very nature of doing harm. This indicates that if there is a time-relative reason against doing harm, then there is also a time-relative reason against intending harm.
34. Suppose you are required under circumstances of full compliance to save V1. Unfortunately, if you save V1, then A2-A6 will choose to kill V2-V6. In this case The Independence Account suggests that you should save V1: we would be less independent if you were permitted not to save V1. The Inviolability Account does not say anything about this case, since the permissibility/ impermissibility of not saving V1 does not affect our degree of inviolability (unless the death of V1 is intended).
35. But is it not possible to provide an account of why independence is more important than unsacrificability based on the need to reflect the difference between intention-dependent and intention-independent conditionals? One reason why not is this. Although it is true of some cases involving minimizing violations that A2-A6 do, but A1 does not, make full compliance impossible for the other party, it need not be true of all cases involving minimizing violations. Recall that not all minimizing violations are performed with the intention of minimizing the number of violations. A1 might have been bent on harming V1 whatever A2-A6 intended to do. It just happens that A1 thereby saves V2-V6 from being violated. In this case, full compliance is, given A1's intentions, not possible whatever A2-A6's intentions are. Thus, there is no (systematic) asymmetry with respect to who makes full compliance impossible for whom that can account for the impermissibility of minimizing violations.
36. The Inviolability Account faces a similar problem.
37. See Immanuel Kant, Groundwork of the Metaphysics of Law, trans. H. J. Paton (London: Hutchinson, 1987), p. 96.