Cunningham's Law

The Wikipedia model. Wikipedia's success is partly built on Cunningham's Law. Rather than asking experts to write articles, Wikipedia allows anyone to write — including with errors. The community of people willing to correct errors is far larger than the community willing to write from scratch, and errors attract more engagement than empty pages. The wiki's design (easy editing, visible errors, public accountability) is optimised for correction-driven knowledge production.

Tension with Brandolini's Law. Cunningham's Law and Brandolini's Law appear to contradict each other. Brandolini says refutation is more expensive than production; Cunningham says people will eagerly expend that energy when they see confident wrongness. The resolution is that Cunningham's Law describes the motivation to correct (which is strong for confident claims), while Brandolini's Law describes the cost of doing it completely (which is high regardless). Motivated individuals will pay the Brandolini cost for emotionally salient wrong claims — but the asymmetry still means that for most misinformation at scale, refutation falls behind production.

The Socratic version. Socrates' dialogue technique — asking questions in a way designed to lead his interlocutor to a false conclusion, then examining it — is a philosophical precursor. By allowing people to state wrong views clearly, he created conditions in which the wrongness could be examined and corrected. Cunningham's Law is the internet-era version: allow the wrong answer into the open, where it can be corrected.

Epistemic humility and fake wrongness. The law can be deliberately exploited by posting a wrong answer you know to be wrong in order to elicit corrections. This is sometimes done in good faith (as a pedagogical technique) and sometimes in bad faith (as a way to get others to do your research). Practitioners who use it ethically acknowledge the answer they get comes from the corrector's knowledge, and treat it accordingly.

1995–2000. Ward Cunningham develops the WikiWikiWeb, the world's first wiki, observing that communities of correction are more reliable knowledge producers than communities of composition. The observation that confident wrongness attracted correction faster than humble inquiry is part of this learning.

2010. Steven McGeady publicly attributes the principle to Cunningham in a New York Times blog post. The principle quickly circulates in technology and knowledge management communities.

2010s. Wikipedia's documented history of correction dynamics provides extensive empirical support for the law. Studies of Wikipedia article improvement show that visible errors are corrected faster than invisible gaps are filled.

2023. Large language models are shown to exhibit a version of Cunningham's Law in human evaluation: users who receive an LLM answer and then state a different answer are more likely to receive extensive elaboration and correction from the model than users who simply ask a follow-up question.

Knowledge production. Understanding Cunningham's Law is useful for anyone who needs to extract knowledge from a community. Asking the right question is not always the most efficient way to get the right answer. Stating a plausible but wrong claim can be more efficient in communities where the correction instinct is strong.

Platform design. Discussion platforms can deliberately exploit Cunningham's dynamics: making the edit and correction of answers easy, making correct/incorrect voting visible, and featuring confident claims where the community can examine and correct them. These design choices, informed by the law, shape how effectively platforms produce reliable information.

Online epistemics. The law reveals something about online epistemic culture: the motivation to demonstrate competence by correcting is stronger than the motivation to help. This shapes what kind of knowledge is produced online: errors are corrected faster than gaps are filled. A community that never makes mistakes also never triggers corrections, and may accumulate invisible gaps.

AI systems as correctors. Large language models now participate in online correction dynamics. They are fast, persistent, and not subject to the motivational barriers that limit human correction. An LLM can correct confident wrongness at scale and without emotional engagement. Whether this is net positive for online epistemics — replacing motivated human correction with mechanised correction — is an open question.

Trolling and misinformation. Cunningham's Law can be weaponised: posting extreme or false claims to generate engagement (corrections, rebuttals, shares) exploits the correction instinct to amplify reach. The viral spread of some misinformation follows exactly this pattern — confident, outrageous wrongness generates more engagement than calm truth.

Stack Overflow and documentation. Stack Overflow's culture of marked correct answers, visible voting, and edit history is Cunningham's dynamics institutionalised. The community polices wrong answers aggressively, because the visible wrong answer is a stronger motivational trigger than the invisible missing answer. The platform's design aligns with the law's prediction.