William H. Walters has an intriguing new article forthcoming in the November 2008 College & Research Libraries entitled Journal Prices, Book Acquisitions, and Sustainable College Library Collections. His thesis is that academic libraries, at least smaller ones or ones that cater to undergraduates, should largely abandon serials in favor of actual books. This is simultaneously provocative and conservative, and I think it's worth seriously considering.
A puzzling piece for his argument is e-books. Walters is no Luddite; he casually yet approvingly cites the use of e-books by science and engineering majors and the inclusion of e-books in Choice's annual list of top books. I take this to mean that he proposes that e-books are part of the solution for college libraries. But one of Walters's major concerns is with "sustainable access" - the ability of libraries to retain rights to previously purchased intellectual property even if they let a subscription lapse. Full fair use and ongoing access through format changes has been major concerns with e-books, so such off-the-cuff promotion of e-books seems at odds with the argument against journals from sustainable access grounds. Open source format e-books are an important resolution to this tension - if you keep your e-books in something simple like .pdf or even .txt format, then this is a lot less of a concern.
I think Walters's inherent conservatism is causing him to miss out on the future of Open Access scholarly research distribution. A good friend of mine who's a physicist in string theory has told me that all the top researchers in physics barely read journals at all any more; they're just there for external validation. All the real physics research is done through the arXiv. (The X is a Greek letter "chi" - get it?) The Social Science Research Network is taking off, now joined by a Humanities Research Network.
On the one hand, Walters argues that librarians must assume "that technological advances, however impressive, will not alter the basic economic, cultural, and legal underpinnings of the current system" - an assumption I think is already false. On the other hand, if we base our budgets on this assumption, but then find ways to add a bunch of free (as in both speech and beer) content, then we come out ahead - lots of books, plus journal-like content.
In the end, I buy much of Walters's argument. There's no need for undergraduate libraries to have every minor journal; the top three or four in each field are probably plenty, with the rest supplemented with ILL and consortia. If this boycott forces database marketers to un-bundle the high-impact from the obscure, all the better.
But I think Walters is grossly overselling college students' preference for books. He says that students see books as more authoritative and resulting in better work than other sources, and that they use those other sources only when they're pressed for time. Newsflash - they're always pressed for time. As Alison J. Head argues in another forthcoming paper, procrastination is ubiquitous on college campuses. Even if you keep the library open 24/7 during finals week, the students will still rely on electronic resources at 2AM the night before the paper is due. This might bring us back to e-books, but I think it's also a powerful argument that online journals are vitally important, because they can be accessed from a dorm room at midnight and at least they're an improvement on Wikipedia.
To sum up, I think that Walters is right that college libraries should probably shift much more of their budgets back to books - not because we can abandon electronic content, but because I think we can use a combination of (as he suggests) market power and (as I and many others suggest) open source and open access to bring the price of journals and other serial content back down.
4 hours ago