Philippa Foot, in her famous article, “The problem of Abortion and the Doctrine of the Double Effect,” (Philippa Foot, “The Problem of Abortion and Doctrine of the Double Effect,” Oxford Review 5 (1967).) among many other examples introduces these two cases, Transplant: A transplant surgeon can save the lives of five of his patients by killing a healthy individual to use the victim’s body parts to save his five patients; and Trolley (“Tram” in Foot’s example) Driver: A driver of a trolley which is headed toward five workmen in one track can save the lives of the five only by redirecting the trolley toward the other track which unfortunately another person is working on it and will be killed if the driver turns the trolley onto him. Foot argues that considering the doctrine of double effect one could explain the moral intuition that the surgeon act to save the lives of the five by killing one person is impermissible while the driver act of steering the tram out of the straight track in order to save the lives of the five persons while killing one other person is permissible.  While both cases are similar in killing one non-threatening person to save some other persons’ lives, the difference according to Foot, is between the worseness of killing one rather than letting five die(in Transplant) and worseness of killing five rather than killing one (in Driver). In other words, negative duty of not killing is more important morally than positive duty of not letting die (or positive duty to aid). So she concludes that though the driver may kill the one, the surgeon may not.

Judith Thomson as the one who has called this problem the Trolley Problem, in a response to Foot’s argument had introduced a case which made trouble for it, namely the Bystander: An out-of-control trolley is headed toward five track workmen and will kill them unless a switch is thrown, by a bystander standing next to the switch, but there is one track workman on the other track, and if the bystander throws the switch, the workman will thereby be killed.  Thomson had shown that according to Foot’s analysis the bystander may not kill the one because in this case the bystander must choose between killing one and letting five die, and he may not kill the one and he ought to let the five die. However, she argued, it is permissible to the bystander to turn the switch and thereby kill the one. So she concluded, in her earlier writing, that Foot’s analysis based on difference between not letting to die and not killing is not sufficient to explain the difference between Transplant and Driver. In her recent paper she has changed her mind.( For her earlier work see Judith Jarvis Thomson, “The Trolley Problem,” Yale Law Journal 94 (1985); for her recent work see “Turning the Trolley,” Philosophy and Public Affair 36 (2008).) Now she comes to be agreed with Foot that the bystander in Bystander may not switch the trolley toward one person. She thinks that turning the trolley in Bystander case is the same as using the workman in the other track as a means toward saving five others.

Kamm in her Tanner Lectures (that constitute this book), delivered at Berkeley in 2013, explains in detail her thinking about the nature of the trolley problem, its moral significance, its various forms and its possible solutions. In lecture I, she defines the trolley problem as a problem for non-consequentialists who think there should be a sort of side-constraint on permissible harm to innocent people in order to reach greater goods (p. 12). So the trolley problem according to Kamm is a challenge to explain exactly what this side-constraint on harming amounts to, and what its forms is. In order to reach a more subtle view about this sort of constraint in lecture I, she focuses on the question concerning who turned the trolley or in other words who is permitted to harm an innocent people and in lecture II, she tried to answer the question about how was the trolley turned, or in other words, in which way one is permitted to harm a person for reaching some greater good, namely saving more people’s lives. She extensively focuses on Foot’s and Thomson’s views in this regard and tries to criticize them.

As mentioned above, Foot and Thomson argue that the driver may permissibly turn the trolley because it is his only way of not committing the worse action of killing five, but a bystander may not turn it because killing one is worse than letting five die (worseness thesis). Kamm thinks that the worseness thesis could not explain why the driver may kill one person by turning the trolley but not by pressing a switch to topple a fat man from a bridge in front of the trolley to stop it from its way toward killing five other people (Driver Topple case). This example shows that the simple view that someone who otherwise kills five people may kill fewer ones would permit too much (p. 16).

Kamm’s other objection is due to Thomson’s thesis that “if A wants to do a certain good deed, and can pay what doing it would cost, then—other things equal—A may do that good deed only if A pays the cost himself,” (Thomson, “Turning the Trolley,” 365.) namely the Self-Cost Claim. Consider in the case of Bystander the switch available to the bystander can be thrown in two ways: He can turn the trolley either toward the workman or turn it toward himself, or otherwise he can do nothing and let the five workmen die (Bystander Three Option Case). Thomson thinks that in this case if the bystander does not want to sacrifice his life and perhaps morally need not do so, he should let the five die instead of turning the trolley toward the workman. However, Kamm responds that while it is true that one has in general no duty to kill herself in order to save others lives but nevertheless bystander indecent action in turning the trolley not toward himself but toward the workman seems not be impermissible. For suppose a case in which very small cost have to be paid to save the lives of five persons and still the bystander does not want to pay such a small cost. In this case it seems that he can permissibly turn the trolley toward the workman. Hence, Kamm concludes, the Self-Cost Claim does not seem to be true (pp. 24-25).

After examining too many hypothetical cases, Kamm ends her first lecture by suggesting that if both the driver and the bystander are permitted to turn the trolley, but neither may do certain other things like toppling a fat man, nor pressing a button that causes a bomb to annihilate the driver and trolley together, nor…, then the trolley problem is not merely about who is sometimes permissible to kill innocent people rather than let others die, but it is about the way the person is killed. This is the question concerning of how the trolley is turned rather than by whom it is turned (p.38). This question leads Kamm into her second lecture in which she proposes her principle regarding the permissibility of harming innocent non-threatening people to save others. According to Kamm “actions are permissible if greater good or a component of it (or means having these as a non-causal flip side) leads to lesser harm even directly. Actions are impermissible if mere means that produce greater good (like the bomb or second trolley) cause lesser harm at least directly, and actions are impermissible if mere means cause lesser harms (such as toppling people in front of a trolley) that are means to producing greater goods” (p.66). This is what she called the Principle of Permissible Harm (PPH) (p.67).

What PPH is saying? PPH says that it matters how the lesser harm is caused. If the lesser harm is caused by a means that prevent the higher harm then the prevention of the higher harm is the causal effect of  the lesser harm, so it is impermissible to act accordingly. But if the lesser harm is caused by constitutive relata of prevention of the higher harm then the action is permissible. For example “the relation between the trolley moving away and the five being saved in Driver is not a causal relation but, rather, seems to be a constitutive relation” so turning the trolley is permissible (p.62).  Kamm concludes the second lecture by insisting that an impartial non-consequentialist should consider the side-constraint on permissible harm which is to act only through means that do not treat other people as a mere means.   

The book is not finished here. Judith Thomson, Thomas Hurka and Shelly Kagan, have written subtle comments on Kamm’s lectures. At the end of the book Kamm tries to respond to commentators as well. Thomson thinks that what Kamm’s analysis of the trolley problem lacks is a focus on a moral theory which enables one to explain what features of some killings and lettings die conduce more strongly than others to impermissibility or permissibility. She says that what is needed in order to answer the questions regarding the trolley problem isn’t dozens of hypothetical killing and letting die cases but a moral theory (p.128). She also adds that PPH leads one to allow some impermissible cases and so is not acceptable. Consider in the case of Bystander a situation in which if nobody were on the alternative track the trolley will go round in loop and hit the five persons from the rear and kill them but if the trolley hit anyone it will be stopped (Loop case). Thomson argues that if Kamm accepts that the bystander in Bystander may throw the switch then she has to accept that the bystander in Loop may throw the switch, because the bystander may not know or even cannot know that he is the bystander in Bystander or in Loop (p.130). In other words, it seems that knowledge of the bystander about the track that is looped or not is relevant to the fact that whether the innocent workman is a means that causes the five people saved, or not. This counterintuitive implication is what Kamm’s PPH commits to.

Hurka in his comments also criticized Kamm’s PPH. Consider the case of a munitions factory belongs to ISIS for example, which we have to destroy it by bombing in just war, but nevertheless, some civilians nearby will be killed by our attack. According to Kamm’s PPH the permissibility of our bombing depends on how the civilians will be killed. If they’ll be killed by the bomb itself the bombing is impermissible while if they be killed by the expulsion of the factory our killing is permissible. Hurka finds this implication very hard to accept because here the way the civilians will be killed seems to be morally insignificant, at least in this case. The main problem of PPH according to Hurka is that it makes an arbitrary distinction between items in a sequence of means like bomb parts and the factory parts (pp.142-146).

Kagan while appreciates Kamm’s efforts to solve the trolley problem, but he also criticizes her for putting some morally insignificant or even irrelevant criterion in her PPH. He says that “when directly consider the difference between the harm being caused by the saving of the many (or its non-causal flip side) and its being caused by a mere means to the saving of the many, it isn’t at all obvious—to me, at least—why a difference like that should matter morally” (p.158). In response Kamm may appeal to the distinction between subordination and substitution. In subordination case like the Driver Toppling case, one person is considered as a mere means to save others while in substitution like in Driver case, no such subordination is involved.  Kamm may claim that the subordination/substitution distinction corresponds and matches with causal/non-causal flip side role of the victims in those cases. However the main question remains that why this correspondence is the case, or in other words, why the causal inter-victim relation makes this a case of subordination.  

In response to her commentators Kamm tries to answer these objections and many other objections. And these remarkable exchanges between some leading contributors make this book really invaluable. I suggest reading this book to every graduate student or moral philosopher who wishes to know lots of hypothetical cases in which the trolley problem happens. I think it fair to say that no one has showed so many hypothetical cases in order to solve the trolley problem, but the book with those too many examples seems to me more puzzling than illuminating.

 © 2016 Ebrahim Azadegan

Ebrahim Azadegan ,Sharif University of Technology