Tuesday, November 26, 2013

What Is the Future of Scholarly Books in the Digital Age?

Many of us entered graduate school imagining that the milestones of our career would consist mainly of the books we would write -- each book a tangible object, printed by a university press, that would enter the discourse when pulled off a shelf at a library or academic bookstore or arriving in a cardboard box from Amazon.  But it now seems that new scholarly books will enter the discourse primarily by digital means.  In this post, I will discuss the mechanics -- and the intellectual significance -- of this transition.  (Please note that my discussion pertains to "scholarly books" in the sense of monographs aimed at audiences of faculty and students.  It’s of course true that academics may write trade books aimed at a broader audience, but here I’m not focusing on those.) 
Digital access to recently-published scholarly books involves three kinds of players: (1) university presses, which produce the books, (2) university and college libraries, which seek to make the books accessible to their faculty members and students, and (3) digital interfaces between the presses and the libraries.  The digital interfaces include Project MUSE, Books at JSTOR, University Press Scholarship Online (UPSO), University Publishing Online, EBSCO eBooks, Ebrary, and ACLS Humanities E-Book.  Some of these interfaces have been around for a while, but several (as noted below) have launched in just the last one or two years. 

Depending on the interface and the terms of its subscription or purchase agreement with the university or college at which the potential reader works or studies, there is potential for an incredibly high degree of digital access to books, approaching the level to which we’ve become accustomed for journal articles.  For me, the exemplar is the purchase that my employer (Yale) has made of the e-book collection at Project MUSE, which launched in 2012.  For recent scholarly monographs from the 80+ presses participating in Project MUSE -- including those at Cornell, Duke, Georgetown, MIT, NYU, the University of California, U. Michigan, U. Penn, and UVA -- Yale faculty members and students can read the entire book online and can download each and every chapter as a distinct PDF document, thereby obtaining their own personal copy, much as if they’d scanned or xeroxed the Yale library’s dead-tree copy.  There are, of course, legal restrictions on what you can do with the PDFs, just as there are legal restrictions on what you can do with a chapter that you scan by hand from a dead-tree library book.  But this level of access is far beyond what’s offered by Google Books or Amazon “Look Inside” (in which you get only a fragmentary view of the book) or by a trade-book interface like Kindle (in which each reader must pay individually for the book and read it on a custom-built machine). 
A serious limitation of Project MUSE e-books is that some major university presses -- including Cambridge, Harvard, Oxford, Princeton, U. Chicago, and Yale -- don’t participate.  But these presses do participate in other interfaces.  For example, the presses at Harvard, Princeton, and Yale are among the many participants in Books at JSTOR (launched in 2012).  Meanwhile, Oxford University Press owns and participates in UPSO (launched in 2011), as do the presses at U. Chicago, Yale, and elsewhere.  Cambridge University Press owns and participates in University Publishing Online (launched in 2011), as do several additional presses, mostly British. 
Participation by the presses is only half the battle: it’s also necessary for your university or college library to subscribe to, or purchase from, the interface.  At present, the interfaces are so new that there isn’t yet an industry standard for which interfaces a serious university library should have. For example, both the Harvard library and the Yale library have purchased all of the e-book collection at Project MUSE, providing broad access for their faculty and students to recent books from the numerous presses therein, but those libraries have not yet bought access to Books at JSTOR.  My guess -- not based on any inside information -- is that this is likely to change soon and that it’s going to become standard, at least for big university libraries, to buy access to all the big interfaces so as to give faculty and students broad digital access to recent books from the whole panoply of university presses. 
Thus, if you write a scholarly book, it’s increasingly likely to be discovered and read in digital form, through an interface like MUSE or JSTOR.  I think this is a happy development for the quality and effectiveness of research and intellectual discourse.  The readers who would find your book relevant if only they knew about its content will be much more likely to discover the book and use it.  This will make an especially big difference for research that crosses boundaries between disciplinary communities.  When an author and a reader aren’t in the same kind of department, don’t have the same mentors, don’t read the same journals, and don’t attend the same conferences, a digital search is often their most realistic hope to “meet” one another. 
Apart from the sheer increase in likelihood that books will be matched with interested readers, there are further implications that go to the very nature of books.  Significantly, a digital interface tends to disaggregate a book into its component chapters.  When you do a word search of the full text of all the books on Project MUSE, the results that appear are chapters of books.  This disaggregation may go farther in the future.  As I was finalizing my book with Yale University Press this past spring, I learned that the Press had recently established a policy -- partly in response to advice from the interface people -- of asking each book author to write an abstract and select five “keywords” for every chapter.  These abstracts and keywords are being sent to all the digital interfaces in which Yale Press participates.  Indeed, UPSO presently offers searches of books by abstract and keyword.  Project MUSE doesn’t, but perhaps they will in the future, especially if more publishers provide such info.  (Not yet having access to Books at JSTOR, I don’t know what kind of searches they offer.) 
Writing an abstract and keywords for every chapter of my book -- something I suspect many of you Legal History Blog readers will do for your next book if you haven’t already -- was an interesting exercise.  It’s familiar to people who publish articles in social science journals or post articles on the Social Science Research Network (SSRN), but it’s more novel for scholars oriented toward the humanities and/or toward book-writing.  The chapter-by-chapter nature of the exercise forces you to think about how different parts of your book might interest different audiences.  Of course you hope that people will be interested in the book’s overall thesis, but if (say) your chapters consist of case studies, some readers may latch onto one of them from an angle quite different from yours. 
This is especially true for books that focus (as mine does) on the development of the modern administrative state.  Typically, a book on the administrative state will (1) state a trans-substantive thesis about the general nature of bureaucracy, politics, etc., while (2) supporting that thesis through case studies grounded in particular substantive areas of administration.  My book, Against the Profit Motive: The Salary Revolution in American Government, 1780-1940, analyzes a trans-substantive phenomenon (the fact that public officers once earned their living on a profit-seeking basis but were ultimately converted to salaries) through case studies of particular areas of government work (naturalization, veterans’ benefits, state and local property taxes, federal custom-houses, state and federal prosecutors, naval warfare, and a few more).  For all of these case studies, the book offers extensive primary research that isn’t available anyplace else.  I therefore hope the book will be of interest to readers who care about (say) criminal prosecution but have no interest in official compensation per se.  Chapter-by-chapter abstracts and keywords help researchers recognize that a 500-page book contains (say) 50 pages that’s highly relevant to their work, even though the case-study subject matter of those 50 pages isn’t a big enough part of the book to be reflected in the book’s title or its LC subject headings.  For my book, the LC subject headings are “Fees, Administrative--United States--History” and “United States--Officials and employees--Salaries, etc.--History.”  Neither of these reflects the topics of the case studies -- an omission that can be remedied by abstracts and keywords for individual chapters. 
To be sure, there may be disadvantages to the digital interfaces’ tendency to disaggregate a book into particular components that don’t all have the same audience.  Perhaps there is a risk of undermining the integrity of the book as a coherent intellectual product.  But from what I’ve seen, the interfaces make it easy for a researcher to jump from the search-retrieved chapter to the book as a whole.  And the chapter-by-chapter abstracts make it easy for the researcher to situate herself within the structure of the book, having entered it through a “side door,” as it were. 
Still, as an author, you may feel some ambivalence about the researcher whose general project is orthogonal to your own, but who extractively reads a single chapter of your book because the obscure primary sources analyzed therein prove to be highly (if accidentally) relevant to her project, even though she little appreciates the remainder of what you’ve done.  But in my view, a case-study-based book should ideally make contributions at two levels: first, in terms of its general trans-substantive argument; and second, in each of the substantive areas on which its case studies respectively touch.  In using (say) criminal prosecution as a case study of officers’ monetary incentives, I sought to move things forward within the historical study of criminal justice, and I’d love for scholars specifically interested in that subject to “extract” Chapter 7 of my book. 
Further, I think that an extractive style of reading is inevitably common -- and sometimes (though far from always) optimal -- in primary-source-based disciplines like history.  In the time that it takes you to read one book cover-to-cover, you might instead “read” five books extractively.  The disadvantage of reading extractively is that you’ll engage less deeply with each author’s argument, but the advantage is that you’ll have time to engage with the arguments of more authors -- and benefit from seeing more authors’ primary research.  It may be that the digitization of books will make this style of reading more prevalent, and I think it would be good for readers and authors to be more self-conscious about that possibility.