Abstention under conditions of awareness and capacity is not morally neutral, and blameworthiness extends beyond the present moment to foreseeable escalation, such that an agent who has the capacity to halt a present evil is blameworthy not only for its immediate continuation but also for its subsequent progression, even where future capacity may be diminished or absent.
Suppose you are confronted with a trolley dilemma much like the trolley dilemma Foot introduced in 1967, but with a twist. You are standing on a trolley whose driver fainted. The trolley is out of control and bolting towards an intersection. On the main track ahead of you, two people are minding their own business. On a parallel track, there is no one. If you do nothing, the trolley will continue on the main track and kill the two people. If you pull the lever, the trolley will divert onto the parallel track, and no one will be harmed. The solution seems simple enough. Why would anyone, in the right state of mind, hesitate to pull the lever and save the two people on the main track. But here is the catch. Each time the lever is pulled and the two people on the main track are saved, a spatiotemporal distortion appears on the parallel track. The trolley is swallowed by this distortion and instantaneously sent back in time to the moment immediately preceding the intersection. However, something has changed. When the trolley reappears, there are no longer two people on the main track, but four. You are no longer standing beside the fainted driver. Instead, a different individual now occupies your position. You are nevertheless fully aware of what is occurring, as if observing the scene through a magical mirror you can follow the unfolding of the events. The process then repeats. If this new agent pulls the lever, saving the four people, the trolley is again swallowed by the spatiotemporal loop and sent back in time. Another random individual replaces the previous one beside the fainted driver. This time, eight people stand on the main track. With each repetition, the structure remains identical.
Every time the lever is pulled, the trolley loops back. The individual beside the fainted driver is replaced and the number of people on the main track doubles (2 → 4 → 8 → 16 → 32 → … indefinitely). If, at any stage, the lever is not pulled and the people on the main track are killed by the trolley, no spatiotemporal distortion appears. The trolley proceeds forward, the deaths occur, and the looping effect terminates.
Moral Tension
The moral question is not whether to save two people but at what point the loop ought to be broken. Saving lives in the present ensures that far more lives will be placed at risk in the future. From the outset, the agent is aware that repeated intervention does not resolve the moral problem. It merely postpones it, magnifies its stakes, and transfers blameworthiness onto others.
What Are You Doing When You Pull the Lever?
The central question raised by the Looping Trolley concerns what an agent at any given iteration is doing by intervening or abstaining, given full awareness that the same structural decision will recur, in identical form, before a numerically larger track, in the hands of a different agent. The question is therefore not about a single agent’s repeated choice but about the moral character of a choice made under the foreknowledge of its iterative structure.
Conclusion
Intervention risks sustaining an escalating structure of harm; refusal permits certain and immediate deaths. Yet when escalation is structurally foreseeable, blameworthiness does not end with the initial moment of decision. An agent who knowingly refrains at a point of capacity may be blameworthy not only for the present harm but also for its predictable worsening. Repeated intervention may preserve moral innocence in the present, but only by transferring moral burden into the future.
This raises the question whether moral innocence preserved through foreseeable escalation of harm can be morally defensible at all.
The experiment therefore presses a more demanding issue. At what point does continued intervention (or abstention) become morally indistinguishable from enabling harm?
Read the full preprint on SSRN: https://dx.doi.org/10.2139/ssrn.6899558
Leandro Loriga PhD