Citizen Google

At the risk of being dismissed as an anti-Google crank, unable to comprehend the Realpolitik behind their human rights position in China and biased against any private sector solutions to public record problems, I'll break my vow of silence to point out some questions that I have about the new partnership between Google and the U.S. Government.

Certainly, the search engine giant has become more beloved than baseball, apple pie, and grandma within many communities of practice in both academia and technological research. But since there are still some advocates for a digital public trust and critics of corporate proprietary conduct and secrecy like Siva Vaidhyanathan (who openly talks about "A Risky Gamble with Google") I might as well revisit the issue one more time.

For those who haven't been following recent events, Google will be digitizing parts of the Congressional Record, which I'm sure rival Readex won't be happy about. Google Video will also putting government films from the National Archives online.

I tried out the Google U.S. Government Search for myself today. Although I found that the results were delightful in their eclecticism, I wondered how well the ranking algorithm or the metadata procedures were working, based on the seemingly scattershot results.

For example I put in "videogames" as a search term, knowing that there have been several hearings on the subject in recent months and that there was still testimony and evidence that I was looking for. I got no closer to data that I could use as a scholar, and as a citizen I found the oddball results were largely out of date: the first result was about a game developed in 1958 and the secod was about a 2000 commemorative postage stamp. Now, I happen to find those things very cool as cultural artifacts but not particularly useful to my understanding of public policy decisions.


Of course, I also had to try out my favorite subversive query in the realm of government documents -- for the keyword "masturbation" -- and got almost the same results that I would have with a regular Google search restricted to .gov URLs.

As far as the basic information design, I thought that the prominence of "American Forces," "White House," and "Washington D.C." headings might discourage participation by citizens looking for information from their local legislators, given the emphasis on centralized, hierarchical, top-down forms of political organization that U.S. Government Search instantiates. I couldn't understand why there also wasn't a way to search for particular types of documents (like podcasts) in their tutorial page.

(In the interest of full disclosure, I have to point out yet again that Google owns this Blogspot software and server, and that Google Book Search now even has its own Blogspot blog. Today Book Search is trumpeting its work digitizing the works of Shakespeare. Of course, MIT got there first by almost a decade.)