Social Fiefdoms

I don’t know if I’ve coined the term ‘Social Fiefdom’ or not, but I thought of it the other day and I think it does a great job of describing the social respect meme that we humans naturally create and follow. I’ve heard several people discussing parts of this phenomena and I wanted to start a conversation about it. It goes like this:

Whenever a group of people get together and there is a specific topic, a Social Fiefdom is immediately created. Certain people will carry respect from others based on their expertise in the area of discussion. For example, let’s say 20 of us get together and we begin discussing blogging. Someone who has never blogged before will be relegated to the social status of a ‘serf’ during the conversation. The value of their input is minimal and the quantity of their input is as well. In fact, much of what they might have to say will be noise. Someone who blogs occasionally would be more of a baron, they will have some valuable insight, but their thoughts will not likely be taken as key information. Someone who is a voracious blogger and has tons of traffic would be more of a lord. Their insights would be valuable, but not alwasy respected. Someone who is either a ‘celebrity blogger’ or has a business that revolves around blogging would be royalty, what they had to say would be listened to carefully, given much weight and respect, and would likely influence how others thought about blogging.

Let’s extend this example with some names. In our little gathering are 4 people who don’t really blog, occasionally read blogs and don’t really have much to say about them. Since they are here, we’re assuming that they are interested in blogging. Therefore, they may ask questions or make a comment, but they aren’t going to be listened to very often – they’re the serfs. There are 4 more people who blog regularly and their blogs are occasionally or regularly read by the rest of the people in the group. What they have to say will be listened to, and their questions addressed, but they won’t be leading the discussion. Now let’s say that Scott Beale, Chris Messina and Tara Hunt are there. The three of them blog regularly, have a fairly decent readership and have some weight behind what we say. In addition to that, they are in companies that are very involved with blogging. Their thoughts will likely be given some weight, they’ll be listened to and the’yll also be the ones in closest communication with the ‘royalty’ now, for the royalty – Robert Scoble, Dave Winer and Doc Searls are in the group. Whatever they have to say will essentially override anything Chris, Scott, or Tara have to say. Their thoughts are going to be keenly listened to by all in the group. They will be given the utmost respect by the group and while some may argue with their points, it will be very difficult to get a group consensus on that arguement.

Now, let’s say that whole conversation turns to facial recognition – now Tara is the royalty and Scoble has very little input. His thougths will be listened to, but what Tara has to say will have weight. Now we talk about Firefox – several in the group may have some weight, but Chris will be the royalty in that topic. And so on.

This is a Social Fiefdom. Humans understand it, what I’ve just described should be very logical to everyone. We do it naturally. So …

How do we make computers understand this???

This leads to a lot of things Web 2.0 related (in my mind at least) several of the factors invovled include: Identity, Attention, Reputation, past history, tagging, ranking, web based social groups, etc.

In order for a computer to understand Social Fiefdoms and to be able to represent them, it has to know several things. It has to know the Identity of each person in the group. It has to know what each person is interested in and what they pay Attention to. It also has to know what expertise each person has and what Reputation they have with that expertise. It needs to know When that was achieved and it needs to know how much each member of the group knows about the other members of the group. In order to programatically determine who is royalty in the group and who is a serf, it will need algorithms to conclude who knows the most about the topic and is the most respected. It will then need to recalculate and shift the Fiefdom when the topic of conversation changes.

Humans do this instinctively. It would be very difficult to get a computer to do this. Mostly because it relies heavily on information that may not be available. There are several sites/services that attempt some of this. There are several sites/services that provide some of the needed pieces. To my knowledge, there is nothing that provides a Social Fiefdom service. The closest thing I can think of is memorandum, and it falls pretty far from the mark.

So, anyone interested in creating a Social Fiefdom system? There is of course another problem – achieving web ubiquity with such a service. If anyone wants to tackle something like this, or has some good ideas about it, I would be very, very interested in hearing from you.