Ethics Thought Experiments

The trolleyology tradition

The trolley problem, introduced by philosopher Philippa Foot in 1967 and refined by Judith Jarvis Thomson, launched an entire sub-discipline. 'Trolleyology' is the study of when it's morally permissible to harm one person to save many. Foot's original version asked whether a bystander could divert a runaway trolley from five people to one. Thomson's footbridge variant added a twist: what if you had to physically push someone off a bridge to stop it?

The numbers are the same in both cases. One death saves five. But most people have very different reactions. They'll pull the lever but not push the person. One explanation is the doctrine of double effect, a principle going back to Thomas Aquinas: when you divert the trolley, the one person's death is a side effect, but when you push, their death is your means. Many people feel that distinction matters morally, even if they can't say exactly why.

The variants proliferate from there. The loop trolley, where diverting still hits five people unless one person's body stops it. The fat villain, where the person on the side track caused the danger. The transplant surgeon who could save five patients by harvesting organs from one healthy person. Each one probes a different intuition, and taken together they make it hard to claim your moral thinking is as systematic as you assumed.

Consequences, duties, and the space between

Behind the trolley variants lies a deeper conflict between two traditions in moral philosophy. Consequentialism, most famously associated with Jeremy Bentham and John Stuart Mill, holds that the right action is the one that produces the best overall outcome. Five lives saved is better than one, full stop. The math is what matters.

Deontological ethics, associated above all with Immanuel Kant, disagrees. Some acts are intrinsically wrong regardless of their consequences. Kant's categorical imperative asks you to act only on principles you could will to be universal laws, which means you can't treat persons merely as means to ends, even if the numbers favor it. Pushing someone off a bridge treats them as a tool. Their rights create constraints that consequences can't override.

The trolley problem forces most people to notice they hold both views at once, and that those views conflict. Kant's murderer at the door asks whether you must tell a murderer where your friend is hiding if directly asked, even if lying could save your friend's life. Kant himself said yes: lying is categorically wrong. Few people agree. The experiment shows what happens when you apply an absolute rule in a world that generates extreme cases.

What do we owe to strangers?

Peter Singer's drowning child asks: if you walked past a shallow pond and saw a child drowning, and could save them by wading in and ruining your expensive clothes, would you be obligated to? Almost everyone says yes. Singer then draws the implication: the geographical distance between you and a child dying of preventable disease doesn't change your moral situation. The child's need is just as real. Your ability to help is just as real. The only difference is visibility.

This argument unsettles most people more than trolleyology does, because it doesn't stay hypothetical. If it's sound, it implies most affluent people are doing something seriously wrong every day. Singer calls for giving to effective charities until giving more would harm you as much as the person you're helping. The drowning child converts a philosophical argument into a personal one.

The lifeboat variants explore what happens when resources are genuinely scarce. If a boat can only hold ten people and twenty are drowning, who do you choose? There's no clean answer, which is part of the point. These cases bridge individual ethics and political philosophy by asking not just what you should do, but what rules a society should have so you're never in that position.

God, morality, and the Euthyphro dilemma

Plato's Euthyphro dialogue presents one of the oldest unresolved problems in moral philosophy. Socrates asks: is something good because God commands it, or does God command it because it is good? Either answer is uncomfortable. If things are good simply because God says so, morality seems arbitrary. God could have commanded cruelty and cruelty would have been right. But if God commands things because they're independently good, then goodness exists prior to God, which undermines the idea that God is the ultimate source of moral authority.

The free will defense addresses the problem of evil: if God is omnipotent and perfectly good, why does suffering exist? The classic answer is that God permits evil to preserve human free will. But this raises further questions. Could God not have created beings capable of freely choosing good without the option of evil? These aren't just theological puzzles. They're about the relationship between freedom, goodness, and what it means to be a moral agent.

Population ethics and its paradoxes

Some of the hardest ethics thought experiments concern people who don't yet exist. Derek Parfit's mere addition paradox asks: if we add a large number of people to the world whose lives are worth living but much worse than average, is the world better or worse? Intuition resists saying worse, because we haven't harmed anyone. But if each addition is a small improvement, we seem forced to accept the 'repugnant conclusion': a world of billions of people living just barely worth-living lives could be morally preferable to a smaller world of flourishing people.

The non-identity problem is stranger still. Future people who might exist are different people depending on what choices we make today. A person born twenty years from now into a depleted world is a different person than the one who would have been born had we acted differently. Can we harm someone by choosing to bring them into existence in difficult conditions? Do we owe anything to people who, unless we act wrongly, will never exist at all?

These questions don't have clean answers, and that's informative. Our moral intuitions weren't built to handle cosmic scales, long time horizons, or hypothetical populations. They may be genuinely inadequate for the decisions about climate, technology, and population that matter most right now.

Ethics experiments at the dinner table

What makes ethics thought experiments good for conversation is that they force a choice. You can't answer 'it depends' and leave it there. Do you pull the lever? Do you push? Do you give until it hurts? The question won't let you off the hook. That pressure is what makes them useful, not as tricks to catch people in inconsistency, but as prompts that surface what you actually value when you have to commit.

The richest conversations happen when someone changes their answer midway through a variant, and then has to explain why. That moment, 'wait, I would pull the lever but I wouldn't push, so what am I actually committed to?', is where the philosophy stops being abstract.