The point of all this folksonomic participation is the selection of information, capturing relevance and separating signal from noise. A network of active readers is creating a vast knowledge base that offers an alternative to search engines, online directories, and editorial sites. Attention-getting is the problem of publishing online. The medium simply doesn’t offer a more efficient means of extracting value than by using popularity. Popularity, as we know, is only a hair’s breadth from what deToqueville described as “the tyranny of the masses.” Every vote, site submission, tag, or link increases the probability that the very-same submission will be viewed by another user, and the vote, seconded. In systems-theoretical terms, the process is auto-poetic. It self-selects from a pool of information, each selection reducing the (future) pool of information. The selections are the application of value distinctions, are the filter by which one bit of information is distinguished from another. But the moment one selection is made, probabilities that further selections will follow increase (see concepts like the power law and the long tail).
The rub of this massive reduction in selections, is that we want this, if only because it creates what can seem like a conversation. The intrinsic conformity&emdash;that members of a community are all reading the same things&emdash;focuses attention and indeed produces signal from noise. But the value that rises is shown to be valued by a community, even if that value is nothing more than a form of social confirmation/validation. This social signal differs little from that which selects the news: that it is new. To paraphrase a point made by sociologist Niklas Luhmann on the mass media: while these media are designed to observe events and record cultural memory, their memory is designed to quickly forget.
In systems theory, a system’s ability to handle new information is contingent on its internal complexity and differentiation. Folksonomies and social networks tend to push up the peak of the curve, connections being thickest where they are also the most redundant. In time, I hope our ability to apply social mechanisms to the selection of information will improve. We will delineate information by its value to individual users, to expert communities, to those in related communities of practice, and so on. Selections and associations between information sources (links, relevance) will improve. Tags, which are now useful in capturing a community’s boundaries will suffer less from the speed and effects of social selections. But the fundamental process by which information is selected in a social medium will remain: a reflection of a culture’s self-identity, executed on the part of those whose votes serve to include them as members.
Technorati tags: social software, interaction design, folksonomy, web 2.0, tagging, social media