I believe that a fruitful area of research in this areas would involve insights from a game-theoretic approach to social interactions to analyze how the rules of various kinds of software people are using to interact influence social behavior, sometimes in predictable ways, and sometimes in unpredictable ways. In a totally unmoderated forum, where the members of the community have no means or ability to exclude people, the only remaining form of punishment for bad behavior seems to be more of the same kind of bad behavior: flaming. Wikis introduce some great benefits for the creation of a less hostile social environment. Flaming comments can be easily hidden by other community members before they cause a general flame war. Editors can be banned by the community in a transparent and open way. Pages can be temporarily locked to allow for a cooling down period. But there is much room for research here, I think, because each change to the software induces further changes to the social norms. What is the impact in a wiki setting of turning on or off the ability for not-logged-in users to edit? (In practice this mostly means allowing people to edit anonymously rather than pseudonymously... an important distinction.) What would happen if adminship were given automatically based on number of edits? What would happen if adminship were given through a strict voting process? What would happen if adminship were given through a strict voting process by existing admins? What if everyone was an admin by default? --Jimbo