The Loop Trolley
In this trolley variant, the side track loops back to join the main track. The one person on the loop is the only thing that would stop the trolley from circling back to kill the five. Knowing that, do you still pull the lever?
Judith Jarvis Thomson introduced this variant in 1985 to test whether the permissibility of diverting the trolley depends on intention and means, or simply on redirecting a threat. It forces a confrontation between consequentialist and deontological explanations of our ordinary trolley intuitions.
Thomson, J. J. (1985). The Trolley Problem. Yale Law Journal, 94(6), 1395–1415.
How the loop changes things
In the original trolley problem, you redirect the trolley from five people to one. The one's death is a side effect. The trolley would have killed the five regardless; you just changed its path.
In Thomson's loop variant, the track curves back to the main track. If the trolley diverts to the side track and no one is there, it circles back and kills the five anyway. The one person on the loop is not just a bystander who happens to be in the path. Their body is causally necessary to stop the trolley from returning. Remove them, and the five die. Leave them there, and the trolley stops.
This matters morally. You are not simply redirecting a threat. You are, in effect, using the person as a brake. The structure of the action now looks less like diverting harm and more like the bridge case in the original problem: the one person is a means to saving the five, not merely an unfortunate bystander caught in redirected danger.
What the loop reveals about the original intuition
Many philosophers thought the lever case was permissible because you were redirecting the trolley, not deploying a person as a tool. The loop variant tests whether that explanation was doing real work.
If the trolley's path is all that matters, the loop should be equivalent to the original: one dies, five are saved, pull the lever. But many people who were comfortable with the original case feel less comfortable here, or reason through it differently, even though the arithmetic is identical.
Thomson's point was not to produce a verdict but to expose what is actually driving the intuitions. If your discomfort with the loop is genuine, you are probably relying on something like the doctrine of double effect, the distinction between effects you intend and effects that are merely foreseen side effects. In the loop, the person's death cannot be a mere side effect; it is the mechanism.
The challenge
The loop variant sits between the lever case and the fat man case without resolving the tension between them. It does not have an obvious right answer.
If you pull the lever in the original but not in the loop, you owe an account of what morally distinguishes them beyond a description of track geometry. If you treat them identically, you need to explain why the causal role of the person doesn't change anything.
This is the kind of case that reveals how much work our intuitions are doing and how little we understand about why.
Discussion questions
- Does it change things morally if the only reason the five are saved is because of the one person on the track?
- What is the difference between using someone as a means and having their death be a side effect?
- Does the 'using as a means' distinction feel like a principled moral line to you?
Take it to the dinner table.
Get 3 thought experiments for memorable conversations, designed for dinner, with friends, at events, or anywhere small talk has gone on too long.