The bookstore of last resort is an online venue where you can purchase any book that has been offered for sale in recent years, even if you make that purchase years from now. It is a from-this-moment-on archive and preservation service for commercially available works. Perhaps at a future date,
it would also be a from-this-moment-back service as well, but you have to start somewhere. As Cole Porter said, you have to begin with the beguine.
The idea for the bookstore of last resort came to me as I have been studying patron-driven acquisitions (PDA) over the last several months. PDA has implications that don’t apply or are of less importance to the traditional way libraries acquire books, and among these implications is a new wrinkle on preservation.
A step into the wormhole, please, and we are now back in the print era — or, as I saw it spelled in a mailgroup not long ago, the Print Era. Authors write books, which are sent to publishers; publishers (after rejecting many more titles than they publish) put the books into their network of distibutors; the distributors (bookshops, wholesalers, cataloguers, etc.) move the books along until they come to end-users, whether individuals or library patrons. Books for which there is significant demand may get reprinted; virtually all of the profit for a publisher is in those books that get reprinted. Publishers keep a certain number of copies in the warehouse against future demand. How many copies varies by title and publisher and also by one’s interpretation of the U.S. tax code, which addresses how one accounts for unsold inventory. At some point, all but the tiny number of books that have strong ongoing demand run into a an insuperable problem: the cost of reprinting the titles exceeds what can be earned from the by-now diminished demand. So these books — most books — are allowed to go out of print, which makes them commercially unavailable.
A quick digression: the advent of print-on-demand supply chains has enormous implications for the book business, as many people have now come to realize. Books need never go out of print. Instead of being stuck with a minimum printing of, say, 500 units to reprint a title on an offset press, POD makes it possible literally to print a single copy at a time. If you think you can sell only 20 copies of a title each year, offset printing, with its minimum print runs (because small quantities on offset equipment cost so much per unit), is not feasible financially. POD was thus revolutionizing supply-chain management and, by extension, the contracts authors sign with publishers (no rights reversion for books that never go out of print), well before e-books were on the horizon.
What happened to those books in the print era that went out of print? The publishers pretty much walked away from them. Now and then, a reason came up to bring an out-of- print title back into print (e.g., an author becomes increasingly prominent, stirring demand for his or her “lost” early titles), and entire companies were built on the chance that restoring books to the marketplace was a winning proposition. But for the most part, when books went out of print, those titles lived on only on library shelves. Since the titles were no longer available commercially, the libraries set up their own supply chains in the form of interlibrary loans. (In some instances, libraries used ILL to lower costs, but ILL for print books is not cheap.) Thus books entered into a second stage of their lives, with libraries serving as distributors. The print copy sitting on a library shelf somewhere served to preserve the cultural record and make it available to the scholarly and curious.
When publishers attempt to stop libraries from sharing electronic copies of texts with other libraries, they often fail to realize that the library is simply extending a time-honored practice that came into existence precisely because of publisher neglect — rational neglect, as it is impossible to make a case for unprofitable printing in the commercial world, but neglect nonetheless.
When we step out of the wormhole, we are greeted by an entirely new environment in which books sit on servers; books are indexed by search engines; POD is commonplace; e-books now comprise 20% of the consumer book market; we read on tablets and smartphones; and copyright battles are as common as a burrito. This is not your father’s book business. Information technology is being applied to every stage of the value chain, sometimes enhancing it, sometimes disrupting it.
This is the context of PDA. Librarians are examining their supply chain and changing it. In the print era, librarians acquired books just in case someone would want them at some point, now or in the future. Since books routinely went out of print, a library that declined to purchase a book today might find it impossible to procure tomorrow. With PDA, however, the library puts off that purchase (or, in many instances, a rental) until there is a clear sign of demand. And here is the problem: How does the library know that the books in the PDA system will be available commercially?
This is the new wrinkle on preservation. One kind of preservation is the “dark archive,” where digital files are stored in the event that there is some special kind of triggering event, such as the bankruptcy of a publisher, which would otherwise make those files unavailable to people and institutions that had already purchased them. But this new kind is to preserve works that have not yet been purchased. How can we be assured that the titles in a PDA catalogue will be commercially available when library patrons finally get around to requesting them? This is the place for the bookstore of last resort.
When people think of bookstores nowadays, they usually begin — and end — with the Amazon model. You can buy books, print or digital, from any device, usually at surprisingly low prices. But there is much that Amazon does not do, things it does less well, and things it does misleadingly. Amazon, for example, does not provide some of the meta-materials for books (e.g., cataloguing information) that libraries require. Under the “doesn’t do well” heading we have the problem of incomplete metadata for many titles, though one might fairly lay this charge at publishers’ feet. As for misleading, for which Amazon should truly be ashamed, is Amazon’s penchant for steering consumers to merchandise of greater profitability or strategic importance to Amazon against the customers’ own declared interest. For example, a search for a finely edited version of a public domain classic will bring a user to a page that offers that edition and also a free or almost-free Kindle edition — except that the Kindle edition may have the text of the classic without the desired editorial apparatus. Really, Mr. Bezos, have you no sense of decency?
The bookstore of last resort would not do what Amazon does well (for all my cavilling, Amazon is one of the world’s great companies) but for where Amazon does not provide a full and satisfactory solution. In order for librarians to feel entirely comfortable with PDA systems, they need to be confident that there will always be one vendor that stocks or stores all titles that a patron may request, however far into the future that request may be made.
In order to switch from just-in-case to just-in-time, a third party has to take on the role of the just-in-case supplier.
The bookstore of last resort may not be a particularly attractive place; after all, it competes only for desperate customers who have no other place to turn. It will not have comfortable sofas or a coffee bar; its interface may not be stylish, and its fulfillment may lag far behind Amazon. As the vendor of last resort, it may be expensive, and if it displays a bit of arrogance in its trading relations, well, that comes with the territory. The bookstore of last resort, in other words, does not uniformly embrace best practices, but is an advocate of worst practices or, as they may be better termed, selective practices. The bookstore of last resort, in other words only does supremely well what it and no one else does. It is expert about one single thing in an expanding universe of many things. And that single thing is to make books available now and forever.
The bookstore of last resort would of course be a virtual bookstore — go to a Web page and begin your search. It would stock all titles of interest to the scholarly community, and it would stock them forever. But what do we mean by stock? Would it carry print books? Digital books? Would it host its own digital books or would it simply include a link to the digital storefronts of other vendors? All of the above, provided that these options were commercially meaningful. But the core service would be the versions or formats stocked for the long term, which may only be commercially viable in the aggregate.
This service would thus host digital files for every title in its catalogue. The file types would be determined by librarians who specialize in digital preservation. And since some digital files may become unreadable in the future, there would be an accompanying service for POD, as print is a good and useful format even if it lacks many of the affordances of digital text.
The bookstore of last resort is unlikely to have a throng of customers, as many of its titles are available elsewhere, perhaps at lower prices or perhaps implicated in broader marketing programs that promote customer loyalty (e.g., Amazon Prime). For it to be viable, therefore, it may need to have a two-sided economy. Some support for the store comes from its paying customers, who pay top dollar for a hard-to-find book, but the publishers of these books themselves pay a fee for each title hosted by the service. These fees are small, but they guarantee that the books will always be available. One possible model would work like this: initial capital to start the service would come from philanthropic institutions; publishers’ annual hosting fees would pay for day-to-day operations; and the sale of books through the service would generate a profit, to be invested in augmenting the service.
Few publishers would pay to participate in this service unless they have to. Librarians can wield the stick by insisting that no titles would be included in PDA catalogues unless the publisher participated in the bookstore of last resort. Thus libraries can assert their market power as buyers to get publishers to take on some of the responsibility for long-term preservation.
I have previously written about other kinds of online marketing venues for scholarly books and a bricks-and-mortar community space to promote the sale of books. The bookstore of last resort could be integrated with these other services. It may be that the preservation service could become a feature of a broader-based online merchant, or perhaps the bookstore of last resort is the first iteration of such a broader service. Whatever form this ultimately takes, and in whatever order, what is clear is that the new world of digital media will require new infrastructure for preservation, and that the scholarly community will have to invest in creating it.