A living open book

How to cite this book chapter: Suber, P. 2014. A living open book. In: Woodward, H. (ed.) Ebooks in Education: Realising the Vision. Pp. 113–117. London: Ubiquity Press. DOI: http://dx.doi.org/10.5334/bal.l This is a case study of my short book, Open Access (Suber 2012a). The book is not “enhanced” in the way that a growing number of digital academic books are enhanced. It has no graphics, no multimedia, and no interactivity beyond links, and does not offer different layers or pathways for readers at different levels. From that point of the view the book is conventional and text-oriented. But it has two other enhancements worth highlighting. First, the full text is open access, which benefits authors and readers, and sometimes also publishers. Second, the book has a companion web site of open-access updates and supplements, which benefits all three groups.

I wanted the book to be open access for the same reasons that I want all research literature to be open access. In the book itself (Section 5.3), I discuss the reasons why open access is more difficult for books than articles, and nevertheless why it can be viable for authors and publishers. I won't digress here on the advantages of open access. The book should speak for itself on those. I'll merely say that my readers and I have realised the benefits for which I' d hoped. I'm grateful to MIT Press for its willingness to make this possible.
I'll focus here instead on the consequences for the press. A good number of anecdotes and studies suggest that open-access full-text editions can stimulate a net increase in the sales of print books. I collect the evidence in the book itself at pp. 109-110, especially notes 8-10 at pp. 200-202, and in the updates and supplements for pp. 109-110. The evidence is promising enough that dozens of academic presses now experiment with open-access editions that coexist with priced print editions.
The same anecdotes and studies suggest that the "net boost to sales" effect works for some books, or some kinds of books, and not for others. To see whether it worked for my book, MIT Press analysed its sales data before and after the first open-access editions appeared in mid-2013.
The open-access editions did not boost the net sales of my book. On the contrary, the month after the open-access editions appeared, print sales were one-third of what they were the month before, and were never as high again. However, two months out, print sales were higher than they were one month out. Instead of an immediate spike in sales followed by a slow decline, we saw an immediate drop followed by a mild bump.
The open-access editions caused a deeper drop in digital sales than in print sales. Not surprisingly, people who already prefer to read digital editions also prefer free digital editions. If the sales of print editions declined more slowly, part of the explanation lies in the appeal of print itself.
Ellen Faran, the director of MIT Press, and I have discussed why the "net boost to sales" phenomenon works for some books, including some academic books, but didn't work for my book. Ellen also wonders whether the book is perceived as a reference book more than a monograph. That would matter because the evidence suggests that the "net boost to sales" phenomenon works better for monographs than reference books. The thinking is that when readers have an open-access option, they're more likely to buy a book they want to read in full than a book they only want to consult for snippets. I admit that as I wrote the book, I thought of it more as a monograph than a reference book. But I've since seen evidence to support Ellen's theory. Four months after she voiced it, Choice named the book an outstanding academic title for 2013 in the category of reference books.
Several universities (Cornell, Georgia Tech, Harvard, and the University of Florida) bought print copies of the book in bulk to give to new faculty members. Ellen believes that bulk purchases became harder to justify after the open-access editions appeared. We have no data on whether the open-access editions might have increased individual purchases of the print edition, while decreasing bulk purchases, with the net effect of decreasing print sales. If so, it's particularly regrettable because, as Ellen notes, giving out print editions has more impact than giving out links to an open-access edition.

Updates and supplements
While the book became open access one year after publication, I launched the book's web site on the very day of publication. I created it as soon as I could because I already had a growing collection of updates and supplements to post. These were of two kinds: evidence that I had before publication in mid-2012 but couldn't fit into a short book, and evidence that only appeared after I submitted my manuscript in mid-2011.
The book was supposed to be short. But because open access suffers from wide and deep misunderstandings, and because an overriding purpose of the book was to lay these misunderstandings to rest, I was especially concerned to document my claims in full. In the end, I exceeded my word count in order to include 41 pages containing 153 small-font notes. I'm grateful to MIT Press for letting me do so. Nevertheless, I still had to cut many relevant studies from the manuscript, and wanted to restore them as soon as possible through my online supplements. Moreover, the world of open access is fast-moving, and many new studies appeared during the year when the book was in production. This is one reason why some books are (nearly) obsolete on they day they are published. I didn't want that fate for my book. The online updates and supplements turned a short book into a long one. More importantly, they turned a frozen and dated book into a living book.
I could have kept the book alive in the traditional way by giving public talks and writing new articles. I did both, but I wanted all the updates and supplements to be easy to find and use. I wanted them all in one place, searchable, with links to their sources, and links back to the relevant passages of the book. If I wrote about new evidence in a new article, that article would reach one kind of reader. But at the same time, I wanted to write a paragraph-length summary of the new evidence, with a full citation and link, in the form of a footnote or endnote, for the book web site.
Another way to keep a book alive is to publish a second edition, or even a series of new editions over many years. I could ask MIT Press to consider this. But it might decline and it might be right to do so. Even if the new material was not voluminous, the sales might not justify new editions. More importantly, I might not want this solution myself. For example, I might want to write voluminous supplements. (It turns out that I do.) I also want to post my updates in real time, not just every few years. Above all, I want to write the updates that I want to write, without running them by an editor or publisher. The book itself went through the standard vetting process at MIT Press, but the updates and supplements are entirely under my control. This gives me a freedom that book authors seldom have. As a bonus, this control also means that I could decide the access terms on my own. Open access came to the main text one year late, but the web site was open access from the start.
Despite the fact that author supplements are not vetted by the publisher, the publisher can still point to them as an enhancement to the published book. MIT Press links to my book site from its own web page on the book, and added the site URL to the second printing of the print edition. Beyond letting me post updates and supplements, the book web site lets me link to reviews, translations, and other editions, including open-access editions. Before the open-access editions appeared, the supplements could only cite the main text by page number. But now that the whole book is open access, each supplement deep-links to the page it supplements. Moreover, readers can deep-link to each individual thread of supplements, and use them as public footnotes summarising the evidence for a claim they may want to assert themselves.

Recent developments
As I write this case study (late April 2014), the book web site has been viewed more than 89,000 times. It links to eight open-access editions, as well as the MIT Press page on the book. It links to 19 reviews with short excerpts. It points to ten translations in progress, and one already online. At least three of the translations are now under way because the translators saw my call for more translations on the web site itself.
Above all, the site contains 236 updates and supplements supporting 121 arguments and assertions from the book. The updates and supplements alone come to more than 38,000 words, or more than 80% of the word count of the original book. And of course the site grows in real time.
Here are three quick examples of the kind of updates and supplements I've been adding: Six months after the book came out, I wrote an article generalising the idea (Suber 2012b) of public footnotes growing in real time to keep the documentation current on important assertions in an article, book, or field. The result is a new container or structure for digital research. I called it an "evidence rack" because it allows any argument or assertion to become a hook on which to hang new evidence as it emerges. My book web site fulfils only part of the vision of an evidence rack, but I'm consciously letting it evolve in that direction. For me, the book web site shows that a book can be a finished product of a certain length with an unfinished library of supplements of indefinite length. It lets me make additions in real time without delay or compromise. It lets me link to each growing thread of evidence, and because supplements are open access, it lets other scholars do the same. It also lets me expand the documentation and deepen the inquiry of the original book, thinking only about relevance, not length. If I had to keep my supplements short enough to fit into a new edition, then I' d have to omit most relevant new studies, which would recreate the problem that made supplements necessary in the first place. Finally, it lets me do justice, or try to do justice, to a rapidly growing field, together with its nuances and complexities, in a way that no single, short, dated book ever could.