I had no idea that last week's MIT/Stanford Venture Lab panel was such a big shindig. danah boyd's account, would you pay $5? makes pretty clear the audience was the shindig, and the panel wasn't. But, it sounds like it was a fun audience.

In this same post, danah discusses ideas about whether people have one social network or many, and she puts forth the concept that people have (or like to have) a single social network that they can facet as they see fit. I like danah's concept a lot (and, in that vein, I again recommend her Faceted Identity Thesis [pdf]).

The thing I observe is, that pretty much no matter what, each individual connects what would otherwise be separate networks of people (albeit with redundant connections between different people). What this suggests to me is that any "social network" one has to "join" (in a strict membership sense) essentially forces its members to maintain multiple networks.

For example, by default, my place of employment defines a social network that, in some sense, excludes everyone I am connected with who doesn't work there. To some degree, this kind of network segmentation is natural or at least inevitable.

But, I also think danah is right about facets in that we do not tend to define our social relationships strictly according to these kinds of segments. For example, there are people at work whom I am friends with, whether or not we work together.

The main thing though is that each of us connect any otherwise separate networks. From our points of view, they are not separate. But, that connector is we.

What I don't like about "social network sites" is that they (the site) are the fundamental connector in all social relationships that occur within their sites. You may be able to create introductions or add friends to your list, etc., but the site itself is an intermediary of an introduction, and an intermediary of a friend-to-friend relationship, etc. Literally, these things happen in the site.

To me, this essentially forces, at least to some significant degree, a certain submission of what should be one's faceted domain of relationships into the segments of multiple social networks (specifically segmented online by the boundaries between one site and another). In other words, you have to give up your role as connector to a site that doesn't see the same connections as you do.

Part of that is a trade, because these social network sites are largely about getting you connections you don't have, because they do see different connections that you do. In this role, there are good reasons for sites like this (e.g., dating sites like Friendster, business introductions like LinkedIn).

But, I think, we generally need to integrate all of our "lives" together, as danah says, as we see fit. We evolve in our relationships without strictly respecting the segmentation of our world—like Romeos and Juliets who connect outside of the social network segments of Montague.com and Capulet.com.

My social network ideas for the iCite net are about creating a network layer that is independent of the web's principal divisions between websites (i.e., a layer above domain names). With this, your own web presence (website, maybe just email in some cases) would be the social network site—it would be the social network site that sees the same connections as you do.

I just think this is a fundamental component of online social connections, and that it is not yet explicit in the web (and, so, generally missing or left out of other social network developments).

I know Marc Canter is working to create a People Aggregator, which I think reflects a similar orientation to what I am working on with this aspect of the iCite net. In general, I feel the iCite net is akin to other projects looking to create "open" mechanisms of social network connection online, where that mechanism functions independent of any particular site, yet nevertheless can be used by any site.