The Geography of Buzz

The Spatial Information Design Lab at Columbia University has been documenting the geography of buzz.




Those are maps of Manhattan. According to SIDL (as I assume they call themselves):

In the summer of 2008, the Spatial Information Design Lab set out to analyze the unique spatial and social dynamics that are created by the arts and entertainment industries in New York City and Los Angeles. Working with Elizabeth Currid from the University of Southern California, the lab used a database of arts and entertainment event photography by Getty Images as a proxy for social interaction in geographical space. Because photographs taken by Getty are tagged with location information, they are transformed into data with an unexpectedly powerful spatial component.

The results of the research showed that both Los Angeles and New York have unique “event geographies”, or locations of interest to Getty photographers that reappear at a statistically high [sic] rate than the rest of the city. While each separate arts industry showed some tendencies toward specific geographic locations the events geographies of all the industries are largely held in very similar locations, suggesting that event geographies appear to be closely linked to iconic symbols in both cities.

Now this is all very interesting, but something about this project strikes me as fundamentally tautological: they are determining which places are culturally significant in terms of which places are most frequently photographed. But a place is only going to be frequently photographed if it is deemed culturally significant. So they're determining what places should be deemed culturally significant by determining which places are already deemed culturally significant.

Don't get me wrong - there's value in that. It's interesting to see a map of what are essentially our culture's collective perception of a place; such documentation is a good way for us to become self-aware about those perceptions. But I don't think we should expect to gain any great new insights from these methods. (There's a lot of theatre activity near Broadway? You don't say.) After all, if there are biases in what we (or, more specifically, Getty photographers) deem culturally significant, then these maps will display the same biases.

Like, look at this, from a NY Times article on the project:

That the buzzy locales weren’t associated with the artistic underground was a quirk of the data set — there were not enough events in Brooklyn to be statistically significant — and of timing. “If we took a snapshot two years from now, the Lower East Side would become a much larger place in how we understand New York,” Ms. Currid said.

That doesn't sound like a "quirk of the data set" to me. That just sounds like a method that is keyed to pick up on the most prominent cultural events, which by definition won't be underground or cutting-edge, and which will therefore be confirming of conventional wisdom.

Still, it's cool to be able to quantify the seemingly unquantifiable. I always say: if the rich texture and ambiguity of emotional experience can't be reduced to the barren certainties of mathematics, then what good are feelings anyway?

No, I kid. To quote a man named Junior, "Saying that the quantification of the seemingly unquantifiable aspects of human experience* undermines 'enjoyment' and the 'human factor' is like creationists saying that evolution takes away the 'wonder' and 'mystery' of the universe. It doesn't. It makes it awesomer."

Indeed.

Bonus map of LA (more here):



*Instead of 'the quantification of the seemingly unquantifiable aspects of human experience,' Junior actually said 'VORP.' But the point stands.