Monday, January 28, 2008

Seeing Like a Librarian

Blogging will be light for the next two weeks, as I will be traveling. However, I'm going to take advantage of being trapped in airports, airplanes and other Internet-free places to slack off on my paying work and do something I've been meaning to do for awhile: re-read James C. Scott's Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. I read this book 9.5 years ago as an undergrad and haven't re-read it since, but conclusions from it keep popping into my head in relation to the library-land discussions on privacy and on tagging/folksonomies. At its heart the book is a critique of the high modernist tendency in 20th century global politics, but (as I remember) it has some really interesting things to say about both the dangers of governmental attempts to keep close statistical tabs on citizens and the dangers of ignoring local folk knowledge when constructing a view of the world. So, hopefully I'll be back soon with something insightful to say about all of that.

Feel free to get the book yourself and read along if you're so inclined. It's quite accessible to people who aren't into political theory, and it's actually really interesting.

Wednesday, January 23, 2008

More Great Free Ad-Supported Content

The Atlantic has joined the trend: they will now make all of their content available free online.

Corporate Trust and Business Information

Apparently I'm in good company when I trust Google more than the government: a new study shows that amongst the college-educated elite in 18 countries, trust in business is higher than trust in government or the media. (Link goes to the Financial Times, so it will go behind a pay-wall at some point.)

Another interesting tidbit from the article: “When young US opinion leaders were asked to choose the most credible source for corporate information, a surprising 55 per cent mentioned Wikipedia.”

Monday, January 21, 2008

Why I Trust Google

Siva Vaidhyanathan, proprietor of the blog The Googlization of Everything, has a brief comment today on Google's open-source data project that I blogged about yesterday. As usual, I completely disagree with his comment, but it got me thinking about something.

Vaidhyanathan seems to think that it would have been better if the government would have set up something like this as a public service, rather than having a company such as Google do it. The broader version of this sentiment—that the government is more trustworthy than corporations—is pretty common, I think, and I find it very curious. The vast majority of the time, the government poses a far bigger danger to you than any corporation ever could, simply because the government has the guns and the jails and the authority to use them on you—“a monopoly on the legitimate use of violence,” to use the political science term for it. Corporations that make use of data mining or collaborative filtering or other tools that let them learn about your tastes and habits may be able to annoy you with eerily targeted advertisements, but only the government can take that data and decide that the pattern indicates that you're a criminal and ruin your life with it.

The exception to that dichotomy, of course, is the corporations that collaborate with the government in the “deciding that you're a criminal and ruining your life” department—think the RIAA, MPAA, and Microsoft. But for the most part the corporations that collaborate with the government in that way are the corporations that sell digital products directly to consumers and who need the threat of government punishment to keep consumers from pirating their stuff. Google (wisely, I think, for reasons that I will discuss in a subsequent post) has realized that more money can be made more easily by selling advertisements to third parties rather than selling stuff to customers. In that situation there's no reason for Google and its users to have an antagonistic relationship with each other, because economically they're both on the same side: both benefit when the users get what they want, which is free and copious access to Google's content and services.

So that's the short version of why I trust Google with my data: because they don't themselves have the power to harm me with it, and because they have no incentive to collaborate with governmental agencies that could harm me with it. The fact that Google has a history of going to court to fight the government when it tries to get data from them shows, I think, that Google itself known on which side its bread is buttered.

Along those lines, I'm interested to see how Google's expansion of its DC lobbying operations plays out. I'm mildly concerned that Google might wind up a little too cozy with the government, but I'm more intrigued by the possibility of having Google running around DC fighting on behalf of its users (who, remeber, are on the same side as Google) against the government and the corporations that collaborate with it.

Sunday, January 20, 2008

Another Good Thing from Google

I have forgiven Google for (possibly) mining my e-mail to recommend new blogs to me. (Well, I was never really all that upset with them about it in the first place.) (It's still only recommending library-related blogs in its top 3 recommendations, by the way.)

Why is Google back in my good graces? Because they're opening a new service that will host large scientific datasets for free access. Being a data geek, I'm all in favor of having more data to slurp up and play with. Although, I wonder if they're ever going to have data in the social sciences or psychology? I suspect not, unfortunately, because of the potential human subjects/ethics problems. But still, this is very cool!

Sunday, January 13, 2008

Learn Something New Every Day

Yesterday's new thing that I learned: the technical term for the sort of recommender tools I blogged about a couple of weeks ago is “collaborative filtering.”

Today's new thing that I learned: Google Reader is now using collaborative filtering to recommend new blogs to me. Curiously, the top three recommendations that it had for me were all library blogs, despite the fact that library-related blogs are a distinct minority in my blog subscriptions. However, my Gmail account is absolutely overflowing with library-related stuff, because I subscribe to a whole bunch of library-related listservs. So now I'm wondering . . . is Google using information skimmed from my e-mail to suggest blogs to me? The Google Reader FAQs say no, but I remain suspicious—and really curious about the data and algorithms they're using.

Monday, January 7, 2008

Thursday, January 3, 2008


I admit to being a critic of the cataloging status quo much of the time, but this has me even more baffled than usual. Does anybody know why the Library of Congress is reclassifying Scottish literature as English literature? I mean, not only is this problematic for all of the reasons listed in the article, but it is distinctly possible that Scotland will become independent from the U.K. in the next 10-15 years. (Not inevitable, but distinctly possible—more likely than Quebec becoming independent, but less likely than Kosovo, let's say.) Which seems rather problematic, no?