Abstract
I argue that, according to ordinary morality, there is moral inertia, that is, moral pressure to fail to intervene in certain circumstances. Moral inertia is manifested in scenarios with a particular causal structure: deflection scenarios, where a threatening or benefiting process is diverted from a group of people to another. I explain why the deflection structure is essential for moral inertia to be manifested. I argue that there are two different manifestations of moral inertia: strict prohibitions on interventions, and constraints on interventions. Finally, I discuss the connection between moral inertia and the distinction between killing and letting die (or doing and allowing harm).
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
See Thomson (1976).
Following Thomson’s usage in her (1976) (Thomson, as we’ll see, rejects the existence of moral inertia). This concept of moral inertia ought not to be confused with a different, quite common usage of the expression: the “moral blindness” to evils generated by habit.
Malm (1989). The thesis that she argues for is only that sometimes there is a moral difference between killing and letting die even if all other things are equal (although there are other cases where other things are equal and there is no difference between killing and letting die). As indicated, Malm’s focus is the ordinary distinction between killing and letting die; she doesn’t attempt to give a theoretical account of the distinction. Similarly, my focus is the ordinary distinction between intervening and failing to intervene.
There are other important differences between my proposal and Malm’s, which will be apparent later (see, in particular, no. 18 below).
See Malm (1989, p. 238).
This is a variation on Judith Thomson’s “health-pebble” example in Thomson (1976, p. 84). Thomson’s example is importantly different in that the health-pebble could be deflected from five people to one person.
Malm (1989, p. 248).
Could he demand that you flip a coin or use some other random decision procedure to decide who dies and who lives? I consider this possibility below in the text.
Similarly, Malm argues, Y’s relatives would have legitimate grounds for complaint if you decided to flip the switch but X’s relatives would not, if you decided not to do so (Malm 1989, p. 246).
What if the bystander by the switch panics and, being at a loss about what to do, and with the best intentions, flips the switch? (Perhaps, while in such panicky state, he flips the switch back and forth, and the train happens to go through the switch when it’s set for the right-hand track.) Do we really want to say that the bystander acted wrongly? I think that, in fact, we do. That was the wrong thing to do from an objective perspective. Of course, the agent’s immoral behavior can still be excusable: perhaps the facts of the situation weren’t fully clear to him at the time, perhaps everything happened so fast that he didn’t have time to think it through, perhaps everyone else would have had similar trouble figuring out what to do in a similar situation. But this doesn’t change the fact that his act was wrong. It seems to me that, were the facts completely apparent to us at the time, most of us would not flip the switch and would consider flipping the switch, and thus harming Y, morally unjustifiable. (Thanks to Karen Bennett for pressing this objection.)
Chance already played a role in determining X and Y’s position on the tracks. So why should we leave it up to chance again? For further argumentation, see Malm (1989), Section II.
People like Tooley would presumably disagree. I discuss Tooley’s view in Sect. 2.3 below.
The fact that the ordinary killing/letting die distinction seems to be compatible with naturalism about human agency is important. In particular, it blocks a serious objection to the moral significance of the distinction: the charge of anti-naturalism that, e.g., Howard–Snyder raises to Donagan’s account of the distinction in terms of the “course of nature” (see Donagan (1977) and Howard-Snyder (2002)). In addition to naturalism, a related reason why the distinction between killing and letting die fails to yield a clear verdict in some cases is that it is unclear to what extent it lines up with the action/omission distinction. There is good reason to believe that the two distinctions don’t completely coincide (many people have argued for this), but the boundaries and areas of overlap between the two distinctions are not clear.
See, e.g., Kagan (1998, p. 99). Certainly, there are other cases of killing and letting die for which the moral difference is much more obvious (say, sending poisoned food to starving children in developing countries versus not sending them aid of any sort), but in those cases it’s less clear that other morally significant factors are sufficiently “equalized” (and thus that condition (b) above is met). The same seems to be true of Foot’s famous “Rescue I” and “Rescue II” examples (see Foot (1984)). In her (1989), Malm calls cases like Runaway Train “simple conflict examples,” which she distinguishes from “comparison examples.”
Moreover, why should the deflection of the threat from X to Y suggest a bad intention on the agent’s part? Why assume that the agent deflected the threat because he wanted to kill Y, not because he wanted to save X?
Tooley (1994, p. 108).
Note that it’s still true that Mary would have died if I hadn’t pushed the button. This is a reason to think that deflection scenarios cannot be characterized in purely counterfactual terms, as Malm attempts to do in her (1989), no. 11.
Intervening in the following way is also impermissible: a bullet has already been launched towards Mary; one could launch a new bullet towards John, which would kill John and also deflect Mary’s bullet away from its path. Although in this case Mary is already under the scope of a threat, we kill John by launching a new threat onto him, not by deflecting the old threat. So this isn’t a deflection case. But I think that cases of this kind are less convincing as a refutation of Tooley and in support of moral inertia. For, as I pointed out earlier, starting a threat typically carries with it a worse intention than merely failing to stop a threat. Hence Tooley could say that the reason it seems morally wrong to intervene in this case is that we can’t help but see a bad intention underlying the act.
Another interesting example to consider is a third-person “ducking” scenario (Boorse and Sorensen (1988, p. 118). Boorse and Sorensen seem to believe that, just as it would be okay for X to duck a bullet even if he realizes that Y (who was standing in queue behind him) will get hurt, it would be okay for a bystander to push X out of the way in the same circumstances (even if he realizes that Y will get hurt instead of X). This strikes me as wrong, if other things are really equal, e.g., if we know that Y won’t be able to escape the threat in the meantime. My impression is that, although it would be okay for X to duck the threat himself, it wouldn’t be okay for a bystander to push him out of the way if other things are equal, just like, although it would be okay for X to redirect the train away from himself and towards Y (if he had a remote control that he could use to redirect the train), it wouldn’t be okay for a bystander to do the same thing. An additional complication is that, in a ducking case, it seems less clear that we regard the bystander’s act as a deflection of a threat. For he is not directly acting on the threat itself (as in Runaway Train), but on X. This might explain our perceiving the ducking cases in a slightly different way.
So what should we think about the version of Tooley’s diabolical machine case where an evil man is responsible for the setup? I’m not really sure. I certainly see why it’s more tempting to say in this case that it would be permissible to flip a coin. But, still, I’m not sure that this is justified.
Perhaps we should add to the antecedent of the principle: “and P is the only major threatening or benefiting process of its kind in the situation.” For imagine that a train T1 is threatening X and another similar train T2 is threatening Y, and that you could deflect T1 to Y and T2 to X. Intervening doesn’t seem wrong in this case, if all other things are equal. Thanks to Ryan Wasserman for raising this point.
I am indebted to Stephen Yablo for bringing up this kind of scenario in conversation some years ago.
An important objection that I’ll bypass, and that I think needs to be addressed, is this: Assuming that we can make sense of the distinction between intervening and failing to intervene, why think that this distinction might have moral significance? It is tempting to think that it’s because, when an agent fails to intervene with some ongoing state of affairs, the outcome is not “due to the agent” in any way. But this is misguided, because in all the scenarios we have looked at, the agent’s failure to intervene still helps to determine the outcome. (Bennett raises this type of objection against Donagan’s views on killing and letting die in Bennett (1995, Chap. 7).
A causal difference is not the only kind of structural difference that could fill the bill, but it is one obvious candidate. Another option would be to try to distinguish deflection scenarios from other types of scenario in terms of the different kinds of processes that obtain in them.
This is assuming that we regard causation as an objective relation (one that obtains between events in the world independently of our subjective or inter-subjective experience). Ned Hall recently argued for a view of causation according to which, interestingly, causation rests on the distinction between default states of the world and deviant states (departures from the default states). For Hall, this is because causation rests on counterfactual dependence and counterfactual dependence, in turn, rests on the default/deviant distinction. According to Hall, although typically there is a natural and intuitive way to draw the default/deviant distinction, the distinction is not purely objective. As a result, causation doesn’t turn out to be fully objective either (See Hall (2007), especially pp. 126–127. See also Maudlin (2004), which inspired Hall’s view.). If this view had any merit, it would be an obstacle for the project I have described.
References
Bennett, J. (1995). The act itself. New York: Oxford University Press.
Boorse, C., & Sorensen, R. (1988). Ducking harm. Journal of Philosophy, 85, 115–134.
Donagan, A. (1977). The theory of morality. Chicago: The University of Chicago Press.
Foot, P. (1984). Killing and letting die. Reprinted in B. Steinbock & A. Norcross (Eds.), Killing and letting die (2nd ed., pp. 280–289). New York: Fordham University Press.
Hall, N. (2007). Structural equations and causation. Philosophical Studies, 132, 109–136.
Howard-Snyder, F. (2002). Doing vs. allowing harm. The Stanford Encyclopedia of Philosophy (Summer 2002 Edition), E. N. Zalta (Ed.), URL = http://plato.stanford.edu/archives/sum2002/entries/doing-allowing/.
Kagan, S. (1998). Normative ethics. Boulder: Westview Press.
Malm, H. M. (1989). Killing, letting die, and simple conflicts. Philosophy and Public Affairs, 18, 238–258.
Maudlin, T. (2004). Causation, counterfactuals, and the third factor. In J. Collins, N. Hall, & L. Paul (Eds.), Causation and counterfactuals (pp. 419–443). Cambridge, MA: M.I.T. Press.
Paul, L. A. (2000). Aspect causation. Journal of Philosophy, 97, 235–256.
Rachels, J. (1975). Active and passive euthanasia. New England Journal of Medicine, 292, 78–80.
Sartorio, C. (2005). Causes as difference-makers. Philosophical Studies, 123, 71–96.
Thomson, J. J. (1976). Killing, letting die, and the trolley problem. Reprinted in Rights, restitution, and risk (pp. 78–93). Cambridge: Harvard University Press.
Tooley, M. (1994). An irrelevant consideration: killing versus letting die. In B. Steinbock & A. Norcross (Eds.), Killing and letting die (2nd ed., pp. 103–111). New York: Fordham University Press.
Yablo, S. (2002). De facto dependence. Journal of Philosophy, 99, 130–148.
Acknowledgements
I would like to thank the audiences at the 2007 Bellingham Philosophy Conference, the University of Wisconsin-Milwaukee, the University of Notre Dame, and Northern Illinois University, for their helpful comments on earlier versions of this paper. Special thanks to Juan Comesaña and to my commentators at the Bellingham conference, Anne Barhill and Peter Graham.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Sartorio, C. Moral inertia. Philos Stud 140, 117–133 (2008). https://doi.org/10.1007/s11098-008-9229-x
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11098-008-9229-x