Beware of sneaky malicious agents in moral thought experiments

In moral philosophy, thought experiments are supposed to use and clarify common intuitions to help us distinguish between acceptable and unacceptable states and actions in morally problematic situations. Moral frameworks are currently built mostly by trying to make these intuitions as coherent as possible, as people have various (interpersonally different but, all things considered, fairly convergent) intuitive responses to ethical dilemmas. In addition to these normative intuitions, the end results of a thought experiment also rely on intuitions about what exactly is *going on* in the situation in the first place, and subtle misunderstandings on the descriptive level could easily damage the reliability of our responses to the experiments.

So I’m pretty sure that suffering caused by violence or other malicious acts subjectively feels at least somewhat worse than suffering originating from impersonal causes, even when the tissue damage is equivalent: it certainly feels vastly more terrifying, disgusting, and unacceptable from a third-person perspective (to me at least). So, to evoke the sense of absolutely maximal suffering in moral thought experiments, it’s useful to describe the suffering as torture or violence of some kind, as this maxes out the intensity of the suffering current human brains are capable of (and capable of empathizing with, for the purposes of the thought experiment). But there’s a tradeoff here: yeah, we probably make the experiment more emotionally effective, which all things equal gives us a better understanding of the relevant moral intuitions – but we also lose some clarity regarding the full implicit consequences in these scenarios in a way that I’m pretty sure will bias our judgment.

This is because torture (or other personal, malicious causes of suffering) doesn’t actually happen in a consequentialist void where the torture is the only consequence and correlate of a choice, and no otherwise alarming or threatening events follow. Most importantly, this has basically never happened in the history of human evolution that has shaped all of our intuitions, which means they may not be well equipped to pretend there is such a void even when it’s required by the experiment. In our descriptive-level intuitive interpretation of the situation, the presence of torture or violence probably implies there is an obviously callous, probably unpredictable agent around who thinks it’s OK to hurt others, an unknown unsafe environment where such agents apparently are born, no one around with enough empathy and power to stop the violence, possibly something that poses a threat to us too, etc. (Though note that this is exactly the kind of idle evo-psych speculation your mother warned you about; I don’t know if anyone knows how intuitions really work.) Yes, often other causes of intense suffering, such as starvation or disease, also imply there’s something wrong on a community-affecting level – but probably not quite as saliently and alarmingly as personal violence does, because violence requires direct, immediate action. Whatever the exact cause is, I think there’s an intuitive emotional reaction against violence that isn’t purely reducible to its direct consequences (the suffering it causes), because it has always been such a different kind of a problem to solve, I guess?