«TROLLEYS AND PERMISSIBLE HARM Thomas Hurka, University of Toronto I’d like to start by quoting something that appeared in The Globe and Mail a few ...»
TROLLEYS AND PERMISSIBLE HARM
Thomas Hurka, University of Toronto
I’d like to start by quoting something that appeared in The Globe and Mail a few years
ago. The newspaper had run a book review that mentioned the trolley problem, and this prompted
the following letter to the editor:
The ethical dilemmas involving a runaway trolley illustrate
the uninformed situations that cause people’s eyes to glaze over in philosophy class. Trolleys and trains are unlikely to run away because they’re equipped with a “dead man’s pedal” that applies the brakes if the driver is incapacitated.
The potential rescuer would not have the choice of “throwing the switch” because track switches are locked to prevent vandalism. And the rescuer’s response would depend on the speed of the trolley. If the speed were less than 15 kilometers an hour, the rescuer could jump onto the trolley, sound the bell and save all five lives. If the speed were less than 30 km/h, then the rescuer (with a switch lock key) could throw the switch and kill only the one person on the branch line.
If the trolley were moving faster than 30 km/h, throwing the switch would cause it to derail, which would injure or kill the passengers but save the workers on the tracks. So the better choice is to allow the occupied trolley to run through on the main track and, regrettably, kill, the five workers.
– Derek Wilson, former CN Rail transportation engineer and project manager, Port Moody, B.C.
At the risk of annoying any retired rail engineers who may be reading, I’ll take the opportunity afforded me by Frances Kamm’s stimulating Tanner Lectures to say something philosophical about trolleys.
Kamm’s widely recognized brilliance as a moral philosopher has two sides. She’s an incisive critic of other moral views, pointing out implications their proponents didn’t realize they have and challenging them with arresting counterexamples. But she’s also a creative philosopher who’s proposed many novel moral principles and ideas. Her critical side is to the fore in her first Tanner Lecture and her constructive side in her second. I’ll focus on a positive claim about the Trolley Problem from her second lecture; this will involve my trying to play the role of critic.
The Trolley Problem asks why it’s impermissible to save five people by throwing a fat man in front of a trolley but permissible to save them by diverting the trolley to another track where it will likewise kill one person. This is a problem because the natural explanation of why it’s wrong to throw the fat man – that then you actively do something that causes his death rather than merely allow him to die – implies that it would also be wrong to divert the trolley, which isn’t what most of us intuitively believe. And while the doctrine of double effect seems to yield the right results here, Kamm thinks it has counterintuitive implications in other cases and should therefore be rejected. Hence her proposal, both in earlier writings and in the first part of her second lecture, of a novel Principle of Permissible Harm that she says avoids these pitfalls.
Like the view that distinguishes doing from allowing, Kamm’s Principle looks at the causal process that produces the death of the one rather than at any facts about your mental state.
But, unlike that view, it doesn’t consider how you relate to his death. Instead, it focuses on a causal distinction that arises later in the causal process, after you act. That’s one reason for calling it a “downstream” principle: its vital distinction occurs after or downstream from you.
What is this distinction? Both when you throw the fat man and when you divert the trolley your act produces an evil, namely one person’s death, but it also produces a greater good, which Kamm describes as “the five being saved.” I think this is an ambiguous description, and will return to this issue later; for now I’ll accept the description. What her Principle of Permissible Harm says is that it matters how the lesser evil is caused. If this evil results from a causal means to the greater good, then the act producing the greater good is impermissible.
That’s why throwing the fat man is wrong. He’ll be killed by his body’s colliding with the trolley, which is a causal means to stopping the trolley and thereby saving the five. But an act that produces the same evil may not be impermissible if the evil results from the greater good itself or from something that’s not a causal means to it but instead has the greater good as what Kamm calls its “non-causal flip side,” so the two are in effect the same thing. That’s why, according to her, diverting the trolley is permissible. Here what saves the five is the trolley’s moving away down the other track, and that’s not a causal means to their being saved but the same thing under a different description; as she puts it, “the five being saved is simply the trolley moving away” (Lecture II, p. 6). In both the fat man and diversion cases you actively do something, and your act leads, by a series of intermediate steps, to a good and also to an evil, namely the death of the one.
In the first case the evil results from one of the intermediate steps, which is why throwing the fat man is wrong. But it in the second case it results from the good itself or from something effectively equivalent to it, which makes diverting the trolley permissible.
As so applied the Principle of Permissible Harm seems to yield the right result in the trolley cases. But it has what I think are highly counterintuitive implications in other cases and is therefore open to the same kind of objection Kamm makes to the doctrine of double effect. A case of this type first occurred to me while I was thinking about the ethics of war, so let me start with it.
There’s a munitions factory that’s a legitimate military target and that we want to bomb, but that has civilians living near it who’ll be killed by our attack. Most moral views say that if the number of civilians who’ll be killed is not disproportionate to the factory’s military importance, the bombing is permitted. But Kamm’s Principle says that whether that’s so depends on how the civilians will be killed. If they’ll be killed by a flying piece of the bomb, the bombing is morally impermissible. That’s because the good our act can achieve is the destruction of the factory, the bomb’s exploding is a causal means to that, and causing even a lesser evil by a means to a greater good is forbidden. But if the civilians will be killed by a flying piece of factory, the bombing is permitted. The factory’s exploding isn’t a causal means to the destruction of the factory; it’s effectively the same thing, or has the destruction of the factory as its non-causal flip side. So the principle that allows the diverting of the trolley also allows bombings in which civilians are killed by flying pieces of factory but doesn’t allow bombings in which the same civilians are killed by flying pieces of bomb.
I find this implication very hard to accept. Whatever the morality of collateral harm in war ultimately says, I don’t think anything in it can turn on whether civilians are killed by one kind of flying object rather than another. Nor do I think the world’s militaries have a strong duty to develop weapons that will kill civilians only indirectly, by what they make explode, rather than by their own explosive force. Even if Kamm’s Principle’s yields attractive results in the trolley cases, it seems extremely implausible here.
I presented this case to her at a workshop on the ethics of war some years ago, and in reply she gave me a winning smile and said, “They all laughed at Christopher Columbus.” She was acknowledging that her Principle has the implication I pointed out and saying that, however surprising the implication may be, we should accept it. And she’s continued to acknowledge that it follows from her Principle. In her recent Ethics for Enemies book she says that, if we applied the Principle in war, “it might be permissible to drop bombs on a military factory if the sideeffect deaths of civilians were caused by the destruction of the factory itself but not if they were caused by the bombs.”1 But accepting the implication would be a strange thing for her of all people to do. Her primary method in moral theory has been what she calls the “method of cases.” It establishes moral principles by appealing to our intuitive judgements, but the primary intuitions concern particular examples. We don’t first decide that certain principles are correct and then settle cases by applying the principles to them. We first consult our intuitions about cases and accept principles only if they match those intuitions.
She’ll therefore sometimes reject an attractive-sounding principle because it conflicts with just one or a few particular judgements. In an earlier paper Judith Jarvis Thomson proposed that it’s less morally objectionable to redirect an existing threat, as you do when you divert the trolley, than to create a new threat, as when you throw the fat man.2 Kamm has rejected this proposal because it yields what she thinks is the wrong result in the Lazy Susan case described in her second lecture, where rotating a lazy susan with five people on it to remove them from the path of a trolley causes a rockslide that kills another person. Even though the rockslide is a new threat, she thinks rotating the lazy susan is permitted.
On a scale of Ingenuity in Philosophical Examples, I’d say this Lazy Susan case comes pretty high up – I certainly couldn’t have invented it myself. But precisely for that reason, I find it less than completely decisive. Even if, once I’ve understood how it works, I share Kamm’s intuition that rotating the lazy susan is permitted, I’m not very confident about that intuition just because the example is so far from reality.
But my factory bombing case comes much lower on the scale. In recent years there have been thousands of bombings of military targets in which thousands of civilians have been killed, some directly by the bomb and some indirectly by its effects, for example by the collapse of a targeted building. Yet no one has thought that in these real-life cases Kamm’s distinction makes a moral difference. In Ethics for Enemies she notes that her distinction isn’t recognized in standard just war theory, but I don’t think that’s only because the theory’s framers didn’t think of it. I’m pretty sure that if they had thought of it, they would have given it no weight and would have been very confident in doing so. However much force the Lazy Susan case has against Thomson’s proposal about redirecting threats, I think the factory bombing case has more force against Kamm’s Principle of Permissible Harm.
It may be relevant, however, that the case occurs in war, and this suggests a possible response for her. In Ethics for Enemies she suggests that war may be a special context that alters the morality of harming in certain ways, so perhaps a Principle that yields the right results in trolley cases isn’t properly applicable in war, in which case a counterexample set in war doesn’t tell against its use elsewhere. And her book sets her Principle aside for just this reason when discussing issues about war.3 I don’t myself believe that war is a special context. Like others such as Jeff McMahan, I think the morality of war is just an extension of the everyday morality of self- and other-defence.
And even if war is a special context, we’d need a specific argument why one of its effects is to make a Principle of Permissible Harm that’s otherwise applicable no longer so, and it’s hard to see how that argument could go. Finally, war can presumably be a special context only for those involved in it, as soldiers or at least as citizens of a warring nation. But imagine that our enemy has located its munitions factory on its border with a neutral country, so the civilians who’ll be killed are neutrals. Our duties to them surely haven’t changed, yet I still think it makes no difference whether they’re killed by a flying piece of bomb or a flying piece of factory.4 Moreover, parallel cases can be constructed outside war and even involving a trolley.
Imagine that an out-of-control trolley is hurtling toward me and four other people and, if we do nothing, will kill us. We can’t divert the trolley but have a bomb we can throw to stop it. There’s no one on the trolley but there’s a bystander next to the track who’ll be killed if we throw the bomb. May we do so?
People may disagree about this case. I think Thomson will say we may not throw the bomb; we’re not permitted to kill the bystander but must instead let the trolley kill us. My own view is that we may throw the bomb, and not just because there are five of us. Even if I were alone on the track I’d be permitted to throw the bomb, because in circumstances like these each person is permitted to care somewhat more about his own good. So I think we may act.
But Kamm’s Principle implies that whether or not we may throw the bomb depends on how the bystander will be killed. If she’ll be killed by a piece of bomb, we’re not permitted to save our lives because then her death will result from a causal means to the stopping of the trolley. But if she’ll be killed by a piece of trolley, we are permitted, because the trolley’s exploding just is its stopping or has its stopping as its non-causal flip side. Kamm implicitly acknowledged this in her first lecture. The case I’ve just described is close to her Bomb Trolley Case, and about it she said, “it seems to me impermissible to set off a bomb that will stop the trolley from hitting the five when a piece of the bomb will kill a bystander as a side effect” (Lecture I, p. 5; also pp. 6-7). Why the reference here to a piece of the bomb killing the bystander? Why not just say you may not set off the bomb? I think Kamm was recognizing that, given her Principle, using a bomb to stop the trolley isn’t wrong if the bystander will be killed by a flying piece of trolley.