The Last Library Is Greater than Google | Peer to Peer Review

Barbara Fister, Gustavus Adolphus College, St. Peter, MN -- Library Journal, 9/3/2009

The other day a short news item about a conference held at Berkeley on the Google Book Search settlement included a phrase that caught my eye. Geoffrey Nunberg, an adjunct professor at UC Berkeley's School of Information and a specialist on language and new technologies, said of the Google project, "this is the last library." [See "Google, 'The Last Library,' and Millions of Metadata Mistakes" for more on Nunberg's comments and the response from Google Book Search engineers.]

The terms of the settlement raise all kinds of issues. Will the Google orphanage unfairly require payments for books that are actually in the public domain? Will there be any privacy provisions for users, or will the Last Library conduct surveillance on your reading habits in order to match your interests to advertising? How expensive will it be to enter the library? Will academic libraries have to sacrifice their book budget so they can subscribe to the One and Only library? And will Google's special relationship with the class of authors and publishers mean we'll never have a second crack at building a digital library that functions differently?

The last library? Really?
I was interested to read an expanded idea of what Nunberg meant by that phrase, "the last library," in an essay published this week in the Chronicle of Higher Education, "Google's Book Search: A Disaster for Scholars." (Though a subscription is required, chances are you can find it in one of your library's databases.)

He argues that Google's project to digitize books is the largest ever and, because nobody else will be in a position to do what they've done, it will essentially be the final word. But what really bothers him (as perhaps befits a linguist) is how mangled that final word is. Google lets you type in a phrase and come into a book "sideways"—landing in the middle of any text where the exact phrase is found. But when it comes to the kind of research scholars do, the metadata is "a mishmash wrapped in a muddle wrapped in a mess." 

He goes on to catalog (no pun intended) some of the characteristic errors that make the Google library so frustrating for a scholar who wants to compare editions or search by subject. (Examples are available in his conference presentation slides.) The confusion he describes is reminiscent of Jorge Luis Borge's short story, "The Library of Babel" in which an infinite library offers everything, an exhilarating prospect until you discover its vastness means it's impossible to find anything. "To locate book A, consult first book B which indicates A's position; to locate book B, consult first a book C, and so on to infinity…."

The library is full of imperfect copies, and even professional searchers are defeated by its chaotic vastness. "If an eternal traveler were to cross it in any direction, after centuries he would see that the same volumes were repeated in the same disorder (which, thus repeated, would be an order: the Order)."

Though Nunberg says Google blames the poor metadata on the publishers and libraries that provide the books they scan, he finds examples that suggest otherwise, describing books with strange metadata published before the metadata schemes used to describe them existed. And some of the metadata from publishers does, indeed, seem decidedly odd.

The Cat Lover's Book of Fascinating Facts is given a BISAC code placing it in Technology & Engineering / General; in the Library of Congress catalog it has the subject heading: cats—miscellanea. Dates are another area of confusion. Google might seize on a date mentioned in the book and assume it is the publication date. It has over 500 books in its collection that mention the Internet supposedly published before 1950.

In spite of all this, Nunberg seems quirkily optimistic that, given time, Google will manage to import good cataloging data, match accurate metadata to the right books, and clear up the date glitches. I'm not so sure, knowing how complicated database cleanup can be for even a smallish academic library. Nunberg makes the assumption that Google has a responsibility to make Google Book Search accurate enough for scholarly research.

If it's going to be the Last Library, and therefore the one that scholars will rely on, they are obligated to get it right. By claiming it's a public good, Google is duty bound to make it good, but what's good enough for Google isn't good enough for scholarship.

Libraries have limits, and it's a good thing, too
I think Nunberg is closer to the mark when he says that Google simply underestimated what's involved in searching for information in books. Books are not like blog posts or web pages. In some ways, Google's massive digitization is like any library that contains millions of books that are poorly cataloged. It's wonderful to have such a lot of information, but frustrating when you actually need to find something and you're not quite sure what it is.

The Google library is handy when you have a precise need. It's not much good when you're doing what students typically do in libraries: explore an unfamiliar idea, get a sense of the landscape, and once oriented, home in on a promising area. As a friend of mine once said, they're not seeking answers in a library; they're learning how to ask good questions.

There is no doubt that a lot is riding on getting the settlement right. The nature of the electronic library is up for grabs—whether each use of a book in the future will be expected to generate a revenue stream, whether privacy will be a luxury we can no longer afford, whether instant access to all books will become a necessity that every library will have to pay for, regardless of the cost. It may be that Google has created the first library with such enormous reach, and the terms of the agreement are ones we'll be stuck with for a long, long time.

But it won't be the last library. We will still need libraries that are more than digitized caches of information. We'll still need places that serve a local community, that curate a collection, that organize it by both subjects and classes, making it approachable from multiple directions. We'll still have to help students learn how to formulate questions, examine the possibilities, and gain a sense of the infinite possibilities encompassed in a library that is not infinite.

Read more Newswire stories:

Library Groups Step Up Criticism of Google Settlement; Some Academic Institutions Support It

Google, “The Last Library,” and Millions of Metadata Mistakes

Clues About Europe's Digital Library Future in Europeana Report

Bumps Along the Road for SMU's Bush Library

Columns:
From the Bell Tower—When Every Student Has a Kindle

Architectural Questionnaire for Academic Libraries

People

Best Sellers in Business–Economics