I'm going to use what I've heard of Pres. Miller's talk as a jumping-off point for talking about something I've been thinking about for a while: adhocracy and stigmergy, and the relationship between them.
Adhocracy is a way of organizing a community effort that allows anybody to assert initiave, control, and leadership at any time (but doesn't compel anyone to follow them). Hierarchies spring up when needed, fall down when not. Actually, all bureaucracies are really ad-hocracies... they just don't advertise that they're socially constructed and that their longevity comes from everyone believing that they're The Way Of Things instead of a malleable system. what I'm calling "adhocracies" here are communities that advertise and make explicit that they are ad-hoc.
Stigmergy is the distributed construction of an artifact where the creators use the construction of the artifact itself as a means of communication - you "talk" about making changes to the thing by actually making those changes to the thing. Wikis are a good example, as are really rough foam prototypes in UOCD, as is a good improv troupe performance.
Stigmergic things are usually constructed by adhocracies, but many exceptions exist - think of a bubble-gum/graffiti wall (stigmergic but not ad-hocratic, since it lacks a cohesive community). Ad-hocracies make stigmergic things by definition, since ad-hocracies themselves are stigmergic (so at the very least, you're talking about changing the structure of your community... by changing the structure of your community.)
I bring these terms up in this very ill-thought-through post because at Olin, I often found myself at a loss for the right words to describe the systems I saw, and wanted to expose more folks to the terminology in case it's helpful.
System dynamics (particularly human system dynamics), education (particularly things studying informal systems of learning, as in self-studies and unschoolers) and the sociology of hacker cultures (Eric S. Raymond's stuff is classic, if showing its age somewhat, and only giving a freeze-frame snapshot of one possible world of hackers) are other good places for getting this stuff. Epistemology, too - software testing is one weird place where that particular area of study started making sense to me in the context of how organizations like Olin grow and change and move.
I'll also note that sometimes the best thing you can do to encourage "change" and "grassroots involvement" (or whatever you want to call it) is to make these tacit structures explicit. The first procedure in a set of procedures should always be for a way of revising those procedures (see: CORe constitution and "abolish the Honor Board.").
That's it. Hopefully this incoherence will make someone think of something - spout me back something (mallory.chua at alumni) if you'd like to toss ideas like this around.