The blogosphere is humming with news of Google’s new API for Book Search. This basically allows you to link to a Google Book record based on ISBNs, LCCNs, and OCLC numbers in your own dataset (sometimes called a library catalog). Google is pitching and most librarians are thinking of this as a preview tool for books that the library already owns. Most of the books linked to would be under copyright protection and would therefore only have a preview available in Google. There are, however, thousands (millions) of books in the Google project that are in the public domain now and have a complete electronic text available. That’s what I want to link to.
This is somewhat of a conundrum. The API can only create a link if you have the book already cataloged. We don’t own those millions of public domain books nor do we have them in our catalog, therefore, no link. I don’t think Google and their library partners are suggesting that we all catalog the public domain books in the project, but I’m wondering, why not? But how? Are there MARC records available? Not that I have seen. It’s tempting to want to catalog them all and then have a link generated by the Google API. Boom! Increase our library holdings by a couple million! There are a lot of practical reasons why that may not happen. The cataloging department is crying out in anguish right now about the workload. And just because they are free does not mean that we want them.
Mini-fantasy: We should apply the same principles of collection development and selection that we otherwise would use. Subject librarians could search Google Books for authors of interest or relevant topics. There is a search choice to limit to “Full View,” or those books that are in available entirely online. The librarian could then copy the search URL to the cataloging department: “catalog all of these.” Or perhaps they do it on a title-by-title basis. Maybe someone (like one of the current book vendors) could develop a selectors tool that would exploit the Google API–an aid to selection of Google books. This is a weird scenario as far as collection statistics are concerned. Are these really in our collection? As much so as many of the electronic journals (especially from aggregators) in our catalogs.
Alternate world mini-fantasy: Rather than try to add Google books to our catalogs, just use Google as our catalog. Book search has a very nice display for each record. Why not configure Google so it shows the holdings of my local library, much like libraries are now doing with their link resolvers? This may require a third party like OCLC. Alternate-alternate scenario is that WorldCat show a link to all the public domain Google books. We then use WorldCat local to be our public catalog. It shows the print holdings that I have in my library, but also the e-books that are available at Google. I haven’t looked much at Open Library to know if they plan on doing something like this.
Any of these scenarios would increase the visibility of Google books, placing them before the eyes of lots of library users, more than just the API would do.