A continuum of social models
In my earlier view, I only identified two stages: social networks, and communities. The former describes a loose network of people who generally keep in touch with each other for a common purpose, whether purely social or for business purposes. The people within the network usually know the others, and have some degree of strong ties with each other, but they don't really define themselves in detail as a group, have well-defined goals or ideology. This contrasts against the next level of networks, the community, which do have a common identity and purpose, and where the network of people generally work together for the direction of the community. Beyond the community is the organization, a very strongly oriented group of people, often with defined objectives, budgets and even some level of hierarchy as to how they should work together.
Somewhere in between the social network and the community are other social models. What Howard Rheingold keeps describing as "smart mobs", a loose social network of folks who gather on an "instant" basis for a specific purpose. This is slightly different than the pure social network since folks in a smart mob have a (vaguely) defined purpose: anyone who is interested in X, let's meet up to do something. Other descriptions for this are often used in events as Birds-of-Feather (BOF) sessions, and Meetups. Unlike the more personal connection of a pure social network, the members do not necessarily know each other from prior contact and simply have a common interest. This can also be a transitional idea; one the meetup is over, the members disperse taking the knowledge gained from the event. Sometimes, they do keep in contact and evolve into a longer running community of interest, where membes return to keep working on that interest.
Another model that is described aptly in The Starfish and the Spider (see my book list). Groups like the Apaches of North America (circa 1800s); Alcoholics Anonymous, a peer-group organization, etc. These have some of the characteristics of a social network in that each group is fairly independent with no overall leader. However, they also go the next step along where there is an organizing principle or ideology, and there are local leaders on a decentralized basis. I call this as something different--some will say I'm just splitting hairs--because the overall larger picture of the group is never centrally coordinated. For that matter, something like this is hard to converge into a realistic distributed model, unless there is a strong need for and common belief amongst the members that the goals and purpose are what they share. Otherwise, it tends to be a localized organized that never grows beyond it's domain.
On the other end, the centralized community, seems like a good idea to try out at the beginning but over time this can grow into an inherent bureaucracy instilled in trying to centralize activities. It has some advantages in that there is less tendency to deviate from the mission of the community across the distributed organization. The Starfish model on the other hand has resilience, but truly works on a decentralized basis if the core goals and beliefs are truly of common interest on a distributed scale, and has been tested over time. Otherwise, it can break down into separate factions as each group goes its own way.
The key to group involvement in any of these types of people networks is still strong leadership and influencers. There are good practices and models for each level that can help people work faster or better, but the idea still depends on a having an interest in being a leader and keeping that interest and momentum going over time. I'm surprised by how many folks keep imagining that they can instantly grow a network from complete zero to success in a very short time. Such situations are pretty rare, and usually, new ideas that work in this vein really leverage existing relationships, population moods or past history to get there, rather than truly working from zero.