Clay’s right - i’m a huge skeptic, although i don’t attribute it to the academy at all. My first reaction to hype is and always was critique (unless, of course, i’m doing the hyping). This has resulted in me always ::raising eyebrows:: over everything from the best bands to “i just met the best girl in the world” stories.
I’m not actually in disagreement with Clay about classification - i am, after all, in a librarian school. My first indoctrination was “classification is impossible - here are a bazillion techniques that we use to try to get better schemas.” So, when i critique folksonomy, it is not in comparison to formal structures of classification. My critical reaction comes from any and all concerns that folksonomy is the panacea to hundreds of years of librarian woe. I know that formal systems are screwed, but i think that folksonomy has its own set of problems.
While i acknowledge the comparisons that can be made about the problematic similarities between folksonomy and formal classification, i also think that the effort towards ‘accuracy’ is actually clouding a few major differences. The differences are not that surprising, but very important. It comes down to benevolent dictator vs. crowd behavior. Sometimes the benevolent dictator goes way wrong, but also, sometimes crowds are scary.
There’s a problematic feature to crowds - they like to homogenize. Yes, the guy with the mohawk can assert his independence, but folks might trample him. Or he might be left to his own planet. Should he be given more attention than others because he is different? Should a classification schema be concerned with frequency/popularity or the full range? What does it mean to classify things that are rare viewpoints? Who gets to decide? That’s a heavily contested domain in classification.
Folksonomy isn’t asking the questions about the implications of collective action classification. Who benefits? Who becomes marginalized? What priorities bubble up? How does pressure to homogenize affect the schema and the people involved? How are some people hurt or offended by decisions that are made? Should moderation of classifications occur? If so, what are the consequences?
I totally appreciate the just-do model that is often espoused here, but i don’t subscribe to it. I believe that you have to go into the doing with the questions always at hand and always in check. What makes formal classification interesting is not its end result, its “technology” but the huge discourse around it, trying to figure out the implications of any and all decisions. Those questions have been around for years and i think that it’s important that we use those questions, those concerns, not for comparison but as a guideline for our hyping.
In short, i love tagging and folksonomy. But once it is taken serious and people are talking about ‘accuracy’ and being offended, questions that must be asked despite the hype - “folksonomy is better” is not good enough for me.
1. Joshua Porter on January 31, 2005 10:32 AM writes...
Danah, could you explain a little bit more about what you mean by homogenization?
To me, homogenization is about sameness, a consistency throughout. Similar to Clay's comments about how folksonomies actually provide more insight into some of the cultural issues you brought up before, I think we can include homogenization, too.
Rigid classification schemes are homogeneous for users. Items are classified according to the same categories. The problem is that over time things change, and the costs of modifying classification schemes (or the discourse concerning them) is too high to keep up. I think change is good.
In folksonomies, users might homogenize their own material...for instance by tagging all articles about folksonomies with the same three tags. But this is OK, because the aggregate behavior will show slight variances. It is at these variances--the one extra word that someone used to tag something with--that the guy with the mohawk will be saved from the "wisdom" of the crowds.
In a rigid classification scheme those leaves aren't seen. They're homogenized back into some larger category with less meaning.
Yes, a classification scheme should be concerned with popularity. That's one way to define quality, isn't it? Rare viewpoints are interesting, sometimes, but sometimes they're not. They should be included, of course, but not popularized unless many people deem it so.
Users don't have to be protected from their own opinions. They can decide what is good, and if we aggregate those individual decisions we'll get good overall results. If we start suggesting to users how they should interpret content (kind of like the media/white house does), then the results will be sullied.
Who gets to decide? Everyone.
Permalink to Comment