« Cold run?, part II | Main | 2:08:54/160 »
February 25, 2005
Partial solutions
Michael Gorman has a way with hyperbole. First I saw this in Revenge of the Blog People! on LibraryJournal.com:
Blog People (or their subclass who are interested in computers and the glorification of information) have a fanatical belief in the transforming power of digitization and a consequent horror of, and contempt for, heretics who do not share that belief.
The Blog People subclass in question must not have a very close relationship with real software.
Gorman wrote that as rebuttal to blog entries in response to his Los Angeles Times editorial suggesting the hullabaloo over Google's project to scan old books is akin to personal commuter helicopters and briefcases of microfilm to carry the entire Library of Congress, solutions looking for problems. Right.
But in the first article he also suggests that Google hits "may or may not be relevant," and come back "in no very useful order." Since he didn't include a link, I Googled to his original editorial, Google and God's Mind. It was the third hit after a couple of blog entries at MemeStreams.net referencing the editorial.
Gorman's right, and Google's right. Or they're both slightly wrong. Googling for something when you don't know what you're looking for -- doing research rather than looking something up -- seems as pointless as Gorman concludes. Yet Google provides a fine tool for retrieving something you know is there, but forgot where you or someone else put it on the web.
Google's okay if you already researched the French prison system in the 19th century and need to find that quote about "few murderers in the prisons of Nantes in 1874." It works even better if you cannot quite remember what flag to pass the Linux kernel so it'll boot from CD-ROM over PCMCIA on a Vaio laptop. It does it better than other general purpose search engines, and that's why you use it.
At the same time the folks heralding Google as the electronic equivalent of "the mind of God" have had too much of the CoolAid. If you want to learn something about the French prison system in the 19th century, or about how Linux works, of course you read a book prepared by someone who knows the subject and writes well rather than scan a bunch of Google hits in the order they occur. Retrieval is only a small part of the story. Even someone with a PostIt pile life knows that.
Partial solutions are a good thing, though, for the problems they're originally supposed to solve. It's only unfortunate when they get overhyped, overrated, and oversold. Scanning books, putting them online, and getting them indexed is a good thing. Expecting people to learn what's in the books by scanning the index would be unfortunate.
Posted by Mark at February 25, 2005 08:59 PM