Why Is It Okay to Save Your Mom First?
The Fire Rescue: A Puzzle About Reasons
Imagine you are in a burning building. You can run through one doorway and save your mom. Or you can run through another and save five strangers. Most people feel a strong pull toward their mom. But why? Saving five lives seems better than saving one. Yet something about that choice feels like a reason that belongs especially to you.
Philosophers try to explain that feeling. They say some reasons for acting depend on who you are. You have a reason to save your mom because she is your mom. That reason is tied to you in a way that a reason to save five strangers is not. When a reason is bound up with a particular person’s identity, relationships, or projects, it is called an agent-relative reason. When a reason applies to anyone, no matter who they are, it is called an agent-neutral reason. This distinction helps us make sense of why loyalty to our own families, friends, and promises feels different from a duty to help people in general.
Thomas Nagel and the “Free Agent” in Your Reason
The American philosopher Thomas Nagel, in his 1970 book The Possibility of Altruism, first drew the line clearly. He called agent-relative reasons subjective and agent-neutral reasons objective. His key idea was that you could spot the difference by looking at the logical form of the principle behind a reason.
Nagel believed that any reason can be turned into a general rule: whenever certain conditions hold, anyone has a reason to promote that action or state of affairs. For example, if you have a reason to phone a friend because doing so would make someone happy, you might think the rule is: “If an action makes someone happy, then the agent has a reason to perform it.” That rule does not mention any specific person. The happiness could belong to anyone. So the reason is agent-neutral.
But suppose you revise the rule: “If an action makes the agent’s friend happy, then the agent has a reason to perform it.” Now the rule contains a reference back to the very person who has the reason — the agent. Nagel said that this kind of back-reference acts like a free variable, a slot filled by whoever the agent happens to be. That makes the reason agent-relative. So egoism (the idea that you have reason to promote your own welfare) is an agent-relative theory, while utilitarianism (the idea that you have reason to promote happiness in general, no matter whose) is an agent-neutral theory. The British philosopher Derek Parfit, writing in 1984, gave the terms “agent-relative” and “agent-neutral” the names we still use today.
When a Good Reason Can Turn Bad: The Particularist Challenge
Nagel’s approach works elegantly on paper. But a group of thinkers called particularists raised a serious problem. They pointed out that a single kind of fact — say, that an action will cause pleasure — is not always a reason. If your friend’s pleasure comes from watching you suffer, that pleasure is a terrible reason to act on. In that situation, the presence of cruelty defeats the reason. Reasons, the particularists argued, are holistic: whether a consideration counts as a reason at all depends on the whole context.
This clashes with Nagel’s assumption that any reason must work the same way in every possible case. If a rule says “whenever there is pleasure, you have a reason to promote it,” and that rule fails when the pleasure is sadistic, then the rule is false. But does that force us to give up on the agent-relative/agent-neutral distinction? Not necessarily.
Philosophers can fix the problem by using default principles. A default principle adds a safety clause: “If an action would promote pleasure, and no other feature of the situation explains why it is not a reason, then it is a reason.” This lets us admit that defeaters can cancel a reason, exactly as particularists say. And we can still read off agent-relativity from the principle’s structure. If the description of the reason inside the default principle involves a non‑trivial reference to the agent (like “the agent’s friend” or “the agent’s own pleasure”), the reason is agent-relative. If it involves only features that could apply to anyone, the reason is agent-neutral. So the core distinction survives — it just becomes flexible enough to handle real‑life messiness.
Why It Matters: From Stealing Books to Saving the World
You might wonder why sorting reasons into these two boxes matters outside of philosophy class. One big answer is that it helps us understand a famous fight in ethics: the battle between consequentialists and deontologists.
Consequentialists say the right action is the one that produces the best overall outcome. Deontologists say some actions are forbidden no matter how good the consequences — for example, you must not steal, even if stealing would prevent five other thefts. Consequentialists often accuse deontologists of being irrational: if stealing is bad, shouldn’t less stealing be better? Deontologists can reply by understanding their rule as an agent-relative reason. The rule is not “minimize theft in the world,” but “do not commit theft yourself.” That second rule is indexed to you in just the way agent-relative reasons are. James Dreier, a philosopher working in the 1990s, even argued that almost any moral theory can be recast as a form of consequentialism if you allow the values you are trying to maximize to be agent-relative. That means the agent-relative/agent-neutral distinction might be more fundamental than the old labels of consequentialist and deontologist.
This idea reaches right into your own life. Why does it feel right to keep a promise to a friend, even when breaking it could do more good for strangers? Why do we feel we have a special duty to our own families? Those instincts are not just fuzzy emotions. They reflect the fact that some reasons are tied to your particular position in the world. The distinction between agent-relative and agent-neutral reasons gives you a sharper vocabulary to think about choices that will face you again and again — from sharing your lunch to standing up for someone you love.
Think about it
- If a doctor could save either their own child or three other children, would it be selfish to choose their own child? Why or why not?
- If you believe lying is always wrong, but telling a lie could save someone’s life, can you imagine a way that the rule “don’t lie” could still make sense as an agent-relative reason?
- Think of a commitment you keep only because it matters to you — like looking after a sibling’s secret. Would that still feel like a reason if it applied to everyone equally, or does part of its force come from your special connection?
