I have a little bit of personal experience with meltdowns in intentional communities. I was a member of the Kerista Commune in the 1980s, and I watched it melt down in the early 90's. I was an administrator at the Internet Infidels Discussion Board (as SingleDad, still considered insane by some of the few who still remember me), part of my own community meltdown, and I followed the contretemps closely when EverlastingGodStopper was banned.
An intentional community is just some socially interacting group of people where membership in the group is directly by virtue of some explicit social intention, both on the part of the founders as well as the participants. It's a community a member explicitly chooses because of the social qualities of the community itself... and it's one where the community chooses members based on their compatibility with the existing social qualities.
Intentional communities stand in contrast to (more-or-less) open communities, such as geographical communities like cities, where social intention is not a criterion (or a very weak criterion) for community membership. When I moved to my fair city, for example, nobody asked me about my political, religious, or social views. Unlike an intentional community, such beliefs are irrelevant to my status as a member of the this specific geographical community. (Of course, legal requirements apply, but they apply to all geographical communities.)
Intentional communities stand in contrast also to workplace communities. While a workplace does have selection criteria, those criteria are typically much looser in a social sense, while being much tighter in an economic sense.
Part of the problem is that while they're not precisely new, people don't have a lot of experience coexisting in intentional communities. The members of the community tend to apply thinking appropriate to geographical communities, while those who founded and are administering the community tend to apply thinking appropriate to workplace communities. The latter is especially seen in intentional communities on the internet, where someone needs to pay the server bills, and someone usually owns a trademark on the name of the community, an important asset.
All the community meltdowns I've seen have started with those who administer the community (usually by virtue of ownership of assets, but sometimes (as in Kerista) by the founder's authority) start exercising their authority to maintain the intention of the community. The members — even and sometimes especially those members who are in fact aligned with the community's intentions — resist this exercise of authority in the same terms they would resist an authority in an open community.
Neither side typically addresses the actual situation: the community is neither a workplace nor an open community. Since everyone is pretty much ignoring important truths, and since it's much harder to reconcile fantasy than truth, the controversy spins out of control. The community dissolves, or a chunk of people leave with bad feelings. Those who remain seem always diminished, guarded and less trusting.
The truth is that an intentional community must maintain some sort of intention. An intentional community cannot afford and does not benefit from the absolute freedom of speech that applies to and benefits open communities. There are many open communities on the internet. If anyone wants to belong to a community with true freedom of speech, there are any number of venues for that purpose, notably unmoderated usenet newsgroups.
Once the notion of absolute freedom of speech is abandoned within an intentional community, the question becomes not whether to restrict speech, but how, to what degree, and most importantly on whose authority.
Another truth is that an intentional community is a community, and the social quality of the community is the only quality on which the community can be judged. An intentional community does not have the external, objective constraint of profitability that constrains authority in a workplace community. Nobody is getting paid to be there, and nobody is there to make money. When the administrators, moderators and leaders of an intentional community start acting in ways appropriate to executives and managers in the workplace, they ignore that they are not in fact in a workplace.
This truth is a bitter pill for those who own assets important to the community. The name itself of the Internet Infidels Discussion Board or the Richard Dawkins chat room does half the job of making the community what it is. The name draws in new members, and these names are reasonably and fairly owned by their respective foundations.
But the name does only half the job. The members of the community do the other half. If they do not share authority to maintain the intentionality of the community, they will not be members of a community, they will be consumers of a service. If the owners of a trademarked name wish to provide a service for people to consume, that's their prerogative, but if they truly wish to form a community, they have to sacrifice a great deal of control.
To effectively manage an intentional community, we can borrow a page from the political science of democracy.
The founders and owners of an intentional community should establish a "constitution", a statement of the basic intentions and processes of the community. The constitution should specify most of these basic intentions in an objective way: It should be objectively determinable whether any member is or is not in compliance. The founders should enforce the objective provisions of the constitution directly, but they should leave any vague provisions to the membership.
The constitution should be difficult to change without the consent of both the owners and most of the members of the community.
The members should be responsible for day-to-day operation and the fine details of the community's standards, either directly by means of issue-by-issue votes, or indirectly by delegating authority by election.
Other than provisions in the constitution, the members must be in control of the membership, using a process specified in the constitution. The owners should be able to unilaterally expel a member only for objectively determinable violations of the constitution (or legal violations), and only after member-driven processes have failed.
In 2000, I became an administrator at the Internet Infidels Discussion board. I was chosen by the owners, and my job was to represent their interests. My first real crisis was Eternal, the greatest troll I've ever seen, before or since. To this day I'm not sure whether the guy (?) was sincere and completely stupid, brilliant and completely insincere, or just plain nuts. In any event, his posts were contributing very little of substance and generating enormous ill-will and bad feeling.
After discussing the issue among the owners and other administrators, we (mostly me) decided to ban him. This was our (mostly my) big mistake.
I think if I had put it to a vote, I could have gotten a majority (or perhaps even a super-majority) in favor of banning him. But because I exercised non-democratic authority, the move was seen as autocratic and indifferent to the feelings of the members.
I can't guarantee that as community will follow my advice; I offered it to IIDB and it was politely ignored. I can't guarantee that any community that does follow my advice will survive: Every time we solve one problem, two more spring up in its place. That's life. But I believe that a community run by its members will have fewer controversies, problems and outright blowouts than one run autocratically, however benign and enlightened that autocracy.