Rating the ratings, and the end of neutrality

When I search for my name on major search engines, I'm satisfied with the results. (I hope this modest admission doesn't spur any readers to change that situation.) But many individuals and companies can't make the same claim. Scads of self-help sites advise them to take deliberate action to raise their profiles, and SEO firms as well as others offer paid services to take such action.

Meanwhile, Wikipedians castigate individuals and companies for changing their own Wikipedia entries. No other behavior could reasonably be expected. Passive participation in the online reputation game is not an option.

The operational adage in this game is: "If I am not for myself, who am I?...And if not now, when?" to cite the over-cited Rabbi Hillel. (Now, that was a dude with some reputation. Too bad his Golden Rule isn't observed in cyberspace.)

Could another level of indirection--a rating of ratings--improve this situation? Probably not. But thought experiments are worth considering--especially since the stakes are even higher than I've indicated. Now only are we increasingly forced to take part in the ratings game, but we are prevented from being neutral.

The end of neutrality

As social networks and other sites push ratings, many observers have pointed out what an awkward position that puts us in. If somebody we casually know asks us to friend him (note that this verb is has significantly different connotations from the good old Anglo-Saxon word "befriend"), we know that saying no is a slap in the face, whereas accepting the request gives an implicit endorsement we may feel is unwarranted.

Even worse is when we respect the person on one level (say, for the quality of his work) but not on another (perhaps his personal interactions). Some sites allow us to distinguish different types of reputation, but that only makes things worse. ("Are you accusing me of treating my coworkers better than my family?")

Some people advise you to deal with this situation by accepting a request and then quietly unfriending the person later. But this ruse could be discovered all too easily. And it leads to new types of anxieties among the friend and others: are you just being a sneak, or did some unknown incident cause you to change your status?

I have a related problem: I respond to friend requests but don't initiate them. The reason I don't initiate them is that I built up many other ways to find the people and information I need, long before social networks hit the scene. However, that means that the friends I have on these networks reflect the needs of people connecting to me, not my needs.

Don't think I'm complaining. These contacts are all wonderful people I want to keep up with. I'm just noting that potentially, I will lend my efforts to building up social bonds that are of less value to me than to others. As I said, passive participation in the online reputation game is not an option. So I may at some point create a standard response message: "I'm sorry, but until I find time to actively seek out contacts, I am not accepting any more friend requests."

Meanwhile, the very fact that we have contacts--or that we make postings on forums, or that we engage in other activities--is increasingly being sucked up into Web 2.0 aggregators and used to create ratings. Passive participation is not an option.

Thought experiment: rate the ratings

Scads of Internet sites already try to rank information of value to you by measuring the preferences of people you know and trust. del.icio.us and Technorati are well-known examples. But these are silos that have little impact on the general searches I mentioned at the start of the article, and they favor information from people you already know, or at least have heard of.

Suppose some protocol allowed people to self-identify as communities. Then, some search engines could check which community each link or recommendation comes from. Thus, links or recommendations about a book or site on ecology could be treated differently depending on whether they come from environmentalists who are progressive Democrats, environmentalists who are evangelical Republicans, or purported environmentalists associated with the Wise Use movement.

If you're in one of those communities, ratings from your community would rise in importance. Each link or recommendation would be displayed along with the community from which it originated. Sites that don't participate in the protocol at all would be downgraded in importance.

In order to achieve validity, links and recommendations within the community would have to be factored into the super-rating, so that results wouldn't be distorted by the vocal participation of people who don't really represent mainstream views in that community.

People would be motivated to join these communities in order to get more relevant ratings and to see to it that their own ratings have a greater effect on people who agree with them. They would also find people they didn't previously know who they have affinities with. They might even seek out people they don't agree with and start constructive dialogs!

I know very well there are risks to adding a new level of indirection in situations like these. They're all familiar.

t folksonomies can produce meaningful categories. It could lead to fights over who is a legitimate representative of community values. It introduces new sources of inaccuracy, not to mention opportunities to game the system.

But it does break down silos and introduce a general solution to a problem that dozens of social networks are trying to solve in isolation. Maybe it's worth trying. Or maybe our efforts are better spent at searching for ourselves every few days to see what turns up, and marshaling our friends to try to improve results.