Geeking out as instructional model?

Mike Caulfield has the wonderful habit of, about once a month, posting something that makes you sit up and think.  Last week he did it again, discussing “Geeking out as a conversational paradigm.”  Learning that Mike did grad work on conversation analysis illuminates many of the things he’s written recently and his work on Federated Wiki.

The specific thing that Mike’s post has set me to thinking about is to what extent geeking out is like and not like various dialog-centered instructional models.  My off the cuff response is that geeking out is a sort of socratic method inter partes, where rather than a single leader/teacher asking most of the questions, everyone is socratically questioning everyone else. That feels different to me than Mike’s description of a database synchronization operation, but this may be because the terms come from different domains and not actually mean that the two things are different.

I also think the inter partes  part is important.  Does a geek out work differently when there are big differences between the background knowledge of the various participants (when the databases are way out of sync) or does the connectivist nature of the enterprise allow each participant to plug in where they can and extend the conversation with whatever knowledge they can bring to bear? I look at Mike’s carefully described geek out and wonder, ” What happens to a person at the table who knows nothing about the Joss Whedon oeuvre?”  By #6 (singer/songwriters) we have reached a different and broader topic, which doesn’t require Whedon awareness. This suggests to me that geek outs, because they tend to be wide ranging, can include people with diverse knowledge sets.

Coming back to my first question of how this could be an instructional model, geek outs seem to push against the prevailing educational model of disciplinary silos that focus on deep knowledge in narrow domains.  If your classroom default was a geek out, what would that look like?  Is it amenable to meeting a pre-specified set of learning outcomes. or is the potential for the conversation to go anywhere (think rhizomatic learning) part of the point?