Google: A Company, Not a Religion.

Navigation: Main page

Author: Crawford, Walt1 WCC@RLG.ORG

Have you ever searched something on Google yielding more than 1,000 results? What a silly question. I'd guess most readers have seen results in the millions. Heck, even a vanity search may yield tens of thousands of results. Here's a better question: are you sure you've ever retrieved more than 1,000 results for a Google search?

I'm not being facetious, and this isn't an attack on Google, except to those who seem intent on sanctifying my neighbors (Google is 0.6 miles straight down the road from my office here at RLG).

At this writing, Google never returns more than 1,000 results from any search--typically somewhat fewer. That's easy to verify: Set your preferences to 100 results per page and search any modestly general term. Say, for example, "EContent." On July 20, 2005, that gets "about 512,000" results--and they run out at 587(!) with this message attached: "In order to show you the most relevant results, we have omitted some entries very similar to the 587 already displayed. If you like, you can repeat the search with the omitted results included." Click on that underlined phrase and you get "about 528,000"--but the result runs out at 994. It's not just Google; most open-Web search engines stop returning results after some point, anywhere from 250 to just over 1,000.

This isn't news to anyone who's ever tried to go all the way through a result set. Google admits that it limits returns to that size. Which means, to be sure, that you're relying on Google's word that there are more results and that they're "less relevant" than what you've seen. For that matter, you're always stuck with Google's word as to which results are relevant, and its algorithms have grown less transparent over time--of necessity, given the extent and expertise of gaming ("search engine optimization") to warp those rankings.

I can almost see the emails now. "Why on earth would anyone ever want to go through that large a result set?" Well, if you're a researcher, you might think of reasons. "Most people look at only the first 50 results anyway, so who cares?" Well, most people on the planet don't read EContent either, so why does it exist? The ones that will really get my goat will go something like this: "Why are you trying to tear down Google?" and "The public has spoken. Google is all they want, all they need, and that's the final answer." These are all paraphrases but not straw men. I've seen exactly these lines of reasoning on a library discussion list and elsewhere, sometimes from people who really should know better.

Google is a company that produces some very good products and services. It is not a religion. Its services are neither perfect nor necessarily the final word in any area--and, more to the point, they're not universally applicable to all uses and all users. For that matter, Google doesn't even represent the majority of open-Web searching, although some people seem to think it's the only search engine left.

I use Google (some of the time). I like the clean home page (for most uses). I don't happen to believe that Google is manipulating search results to favor advertisers or anyone else, although I believe some sites are getting good at "optimizing" their placement despite Google's best efforts.

That's not the point. The point is that there's more to searching than Google, and there always will be. (I could have written this column quite a few years ago--but AltaVista would have been the subject, not Google. Remember AltaVista? It was the last word in what search needed to be, for a while.…)

Unfortunately, too many people--including librarians and other professionals--like easy answers. Think how often you hear this kind of logic from software developers: "People love Google, so all searching should work just like Google." Really? Think a scholar won't mind if the English Short Title Catalog shows some of the books published in Oxford between 1700 and 1750, stopping at an indefinite point along the way? Or that a medical researcher will be happy if Medline provides "the most relevant" articles on a disease, with no way of knowing how relevance was judged--and no way to retrieve the others? I'm guessing some lawyers would be dissatisfied if the big legal databases yielded only "the most relevant cases" for a search, with no way to retrieve all of them.

There's plenty of room for specialized databases and search engines alongside Google and its open-Web competitors. At least there should be plenty of room. "Google's mission is to organize the world's information and make it universally accessible and useful." That sounds great as a corporate mission statement--but that's all it is. It's not a universal truth. You probably wouldn't want one company to be the access point and organizer for all information. When someone quotes that mission statement and says, "What room does that leave for X?" where X could be libraries, LexisNexis, publishers, or what have you, they're making a fundamental error. It's an error that hurts us all, Google included.

~~~~~~~~

By Walt Crawford

WALT CRAWFORD (WCC@RLG.ORG) IS A SENIOR ANALYST AT RLG. HIS ONLINE PUBLICATION, CITES & INSIGHTS, IS AT http://CITES.BOISESTATE.EDU. COMMENTS? EMAIL LETTERS TO THE EDITOR TO ECLETTERS@INFOTODAY.COM.