Idea-Based Collaboration

One of two kinds of collaboration Marsh distinguishes, the other being action-based, and the split determines how each should be organized. Where the purpose is to develop or verify an idea, the work should be open to all contributions, since restricting who may contribute restricts what can be found. Where the purpose is action in a specialized field, only those peer recognized as qualified can take the action, though meaningful input can still come from a broad section of the user group and be evaluated and promoted by those qualified. Idea-based work therefore requires extensive feedback and peer review, both to identify signal from noise and to supply knowledge bridges between elite levels of knowledge and everyone else. The critical rule is that promotion attaches to ideas rather than to people. Idea based systems should promote ideas instead of people, a body of laws instead of a judiciary, which allows an idea to be audited without all the unrelated distraction of attached personalities. Marsh's argument for this is partly defensive. An idea based culture, where seeming majority opinions are rejected in favour of facts and individual assessment, is the only real defence against astroturfing and persona management software used to simulate mass support where none exists, and it is also the only way marginalized voices get heard, since bad data can come from good places and good data from bad ones.

Open peer review, open source development and Wikipedia's verifiability standard all implement versions of the same principle, that a contribution is assessed on its content rather than on its source, and Karl Popper's account of objective knowledge as public and criticizable is the philosophical ancestor. Collective intelligence research studies when crowds outperform experts and when they do not. Marsh's contribution is the pairing rather than the principle, since separating idea-based from action-based collaboration shows why open contribution is right for one and wrong for the other, which resolves a tension that open collaboration advocates usually have to argue their way around case by case.