Skip to content

Ranking vs. Mapping Knowledge

Lately I’ve been worried about the costs of rankings. We all know about the sturm und drang surrounding the USN&WR ranking system, but I think the problem is actually far more general, as these provocative responses to Posner’s efforts to rank public intellectuals show. I explored some of the problems caused by search engine rankings in this paper, and suggested a few potential legal responses. But I think a technological or even aesthetic response might be far more effective: some way of representing data that does not lend itself to the commensurating metric of ranking.

A couple events recently gave me some hope for innovation here. At the Yale Access to Knowledge conference, a World Bank expert presented a “Knowledge Assessment Methodology” that measured certain kinds of development. The “web style” graphic reminds me of Calabresi’s old line about “joint maximization;” that there are often many variables we both want to increase and yet don’t believe can be condensed into any single number.

Perhaps inevitably, the site does boil the results into numbers eventually. But I found a more steadfastly spatial representation of information at a recent exhibit called “Places and Spaces: Cartography of the Physical and the Abstract.” As the exhibit guide notes,

[I]nnovative mapping techniques [can] physically show what and where science is today, how different branches of science relate to each other and where different branches of study are heading, where cutting edge science is erupting as archipelagos in the oceans of the yet unknown, and how it all relates back to the physical centers of research. The world of science is turned into a navigable landscape.

The exhibit showed both scholarship on the topic and examples–my favorite was the dissertation outline as subway map, which recalled Simon Patterson’s tongue-in-cheek London Tube map of saint, celebrity, and sportstar lines.

The creativity behind these projects gives me some hope that search engines’ organization of data won’t be entirely rankings-driven. Imagine, for example, a search for “plastic christmas trees.” Currently, you’ll get a rather jumbled list of sellers promoting them, news stories describing them, and environmental critiques of them. Perhaps some day these categories of interest will be mapped into clusters, like this map does for (many types of) science.

Finally, if you’re interested in applications of these ideas to legal scholarship, this paper on the “web of law” should be relevant.