Every skill anyone has was acquired by getting it wrong and adjusting

The argument

The case that people learn from their mistakes rests on how competence is actually acquired. Nobody walks, speaks, reads, drives, cooks, plays an instrument or does a job well on the first attempt. In each case the pattern is the same: an attempt, an outcome worse than intended, an adjustment, and an attempt that is slightly better. Proponents argue that this is not an occasional feature of learning but its basic mechanism, and that it is so pervasive it goes unnoticed. The question is usually asked about dramatic failures, where learning is contested, and almost never about the countless small errors from which every ability a person has was assembled. Proponents argue that the appearance of not learning usually turns out, on inspection, to be something else. Sometimes the person did not know the outcome was bad, because no feedback reached them. Sometimes the feedback arrived too late or too noisily to be attached to the action that caused it. Sometimes the person learned a different lesson from the one an observer expected, and acted on that instead. Sometimes they learned the lesson but the incentive to repeat the behaviour outweighed it. Each of these is a failure of the conditions for learning rather than evidence that people cannot learn. The positive claim is therefore conditional but substantial: given a clear outcome, attributable to the action, received soon enough to connect the two, and no overwhelming reason to persist, people revise. Proponents point out that this is why apprenticeship, practice, rehearsal, drafting and review work at all, and why institutions deliberately manufacture those conditions when they want behaviour to change. On this view the human capacity to learn from error is reliable enough that whole professions are built on arranging for errors to be made where they are cheap.

Premises

Counter-arguments

Sceptics argue that the premise about skills proves something narrower than the conclusion. Learning to play an instrument involves correcting motor errors with immediate, unambiguous feedback, which is nothing like the moral and strategic mistakes the question usually concerns, where the harm is diffuse and the lesson is contested. They also argue that the conditional defence is unfalsifiable: every case of someone repeating a mistake can be reattributed to a missing condition, so the claim never risks being wrong. Sceptics add that the pattern of repeated error in relationships, finances, addiction and institutional decision-making is too consistent to be explained by feedback problems alone, and points instead to limits on how much people revise at all.

Rejecting the premises

[Rejecting P1] Acquiring a motor or technical skill involves immediate unambiguous feedback, which is not the kind of mistake the question usually concerns. [Rejecting P2] Reattributing every failure to learn to a missing condition makes the claim unfalsifiable rather than supported. [Rejecting P3] The listed conditions are rarely all present in the decisions that matter most, so the concession largely gives the case away. [Rejecting C] A capacity that requires conditions seldom met in real life is small comfort, since what people actually do is what the question asks about.