The Great Geographic (and Political) Popularity Contest

One of the most expensive, time-consuming and difficult parts of doing academic or commercial research is collecting good data to analyze.  But new sources of data are emerging that creative researchers are tapping into and much of this data comes from the internet.  Social networking sites, website usage statistics, and online buying-habit statistics are three big sources. 

One of the articles we read this week by David Crandall and Noah Snavely (1) describes their use of a growing volume of geo-coded data associated with photos that people have posted on the global photo archive, Flickr.  Crandall and Snavely mapped the geo-codes, as well as other attached metadata, from Flickr to do research on the meanings of these geographical photo-groupings.  They used the data to figure out what places in the world were the most popular travel destinations, and which routes were the most traveled.

But this analysis raises some interesting questions about the meaning of these results – questions that I’m sure Crandall and Snavely considered.  For example, might there be other more popular places in the world?  Mecca comes to mind.  Crandall and Snavely’s analysis led them to conclude that “the top most photographed cities in the world according to Flickr are New York, London, San Francisco, Paris, and Los Angeles, while the five most photographed landmarks on Earth are the Eiffel Tower, Trafalgar Square, Tate Modern, Big Ben, and Notre Dame.” (p. 241) But Crandall and Snavely are careful to say “according to Flickr” and “most photographed” because there are clearly people who don’t post to Flickr, and people who don’t post – period.  There are also vast populations in the world who don’t photograph (not owning a camera, although this number is undoubtedly, and happily, shrinking) or who don’t vacation.  Still, for those that vacation, photograph and post on Flickr, these are not surprisingly the most popular destinations.

The analysis of the most traveled routes is also interesting, and especially from a historical standpoint.  The most traveled routes are also the routes that are most convenient places for people to be hanging around with their cameras.  Major highways.  Train routes.  Popular beaches.  Coastlines, in general.  But leaving aside coastlines and rivers whose geographies were created by shifting tectonic  plates, the highways and railways were created by humans.  But what is the meaning of these routes and why were they created where they were?  For the most part they connect up “important” places like cities, natural resources (like mines), recreational destinations (like ski resorts) and commercial establishments (like malls).  And mostly, they follow the most convenient and cost effective paths through valleys and crossing rivers at narrow points.  But they often also follow routes that have been used through the millennia by our earlier ancestors.  It’s interesting to consider that many of these present-day popular routes have been popular places for people throughout the history of humanity, including coastlines.

But these geographical places and routes are more than just interesting – they are also political. People travel to places that are advertised and that have been developed for travel.  Highways are built to have on-and off-ramps according to the political influence of neighbourhoods and businesses.  And even websites that people “travel” to, like Flickr, are political – the sites that pop to the top of our search lists are not random.  For the most part, websites get to the top of the list either by paying for that honour, or by being popular.  But that popularity is also often financially stimulated.

So what does all this mean?  I guess only that we have great new sources of data to analyze, and some of it has a historical continuity.  But it also means that we need to carefully consider the meaning and sources of the data we find.

My $.02

Allison Smith

  1. David Crandall and Noah Snavely Networks of Photos, Landmarks, and People Leonardo , Vol. 44, No. 3 (2011), pp. 240-243