The Harmless Torturers
If a million people each cause a victim one ten-millionth of a unit of suffering, and no individual causes any discernible harm, has anyone done anything wrong?
Derek Parfit introduced this scenario in 1984 to expose a gap in standard moral frameworks. The problem is not just practical but structural: theories built around individual acts and their consequences struggle to account for harms that are real and severe but distributed across so many actors that no single one seems responsible.
Parfit, D. (1984). Reasons and Persons. Oxford University Press.
The scenario
One thousand torturers are connected to a single device. A victim is strapped in. Each torturer presses a button that sends a tiny electrical pulse, imperceptible on its own. No individual pulse causes any discernible pain. No torturer can observe any change in the victim's condition.
Collectively, the million pulses add up to agony.
Parfit's question: has any individual done something wrong? Standard act-consequentialism evaluates actions by their consequences. Each individual act produces no bad consequence. Each torturer, considered alone, has caused no harm. Yet the victim suffers intensely.
Why this breaks standard frameworks
The consequentialist account says no individual has done anything wrong, because no individual's action had a bad outcome. This conclusion is hard to accept.
Deontological frameworks fare no better. A duty not to harm others requires that your action constitute a harm. If the threshold for harm is discernible suffering, and no individual act crosses that threshold, then no individual has violated a duty. The million torturers each walk away clean.
This is not a case of moral uncertainty or competing interests. The problem is that collective harm does not reduce to individual harm in the way moral theories assume. The frameworks were built to evaluate individual acts. They were not designed for cases where harm is real, severe, and causally distributed.
What it implies about collective responsibility
The harmless torturers suggest that moral philosophy needs to operate at more than one level. Individual-level analysis is insufficient when harms are aggregative.
Some philosophers respond by arguing that each torturer is wrong because they participate in a collectively harmful practice, even if their individual contribution is imperceptible. This requires a framework that assigns moral weight to participation and contribution, not just to individual causal impact.
Others argue the scenario shows that our threshold-based concept of harm is too coarse. If harming requires noticeable effect, we have built an escape clause for diffuse wrongdoing into the definition. Pollution, climate change, and industrial exploitation all have this structure: each individual act seems negligible, the collective result is catastrophic.
Discussion questions
- If every individual action in a collective harm is imperceptible, does that mean no individual is responsible?
- Can you think of a real case where you contribute to collective harm in a way that feels inconsequential by itself?
- Should moral responsibility track individual impact or participation in the group action?
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.