A problem is slowly emerging for historians in the form of librarians discarding books from their collections, a procedure that has potential long-term consequences for scholars doing research in the years to come. We need to understand the features and magnitude of the problem and begin to address it today.

Librarians, primarily in the United States, but also to an expanding extent in western Europe, are faced with space constraints in storing books and with declining budgets for maintaining the collections they have today. To address these twin problems they have either created off-site storage facilities and are learning how to make those conveniently accessible to scholars, or they are culling their collections of materials that patrons do not appear to use frequently. They are disposing of materials largely through three mechanisms: by making them available to other librarians, offering them for sale at local book sales in their communities, or simply selling them to book dealers.

The pace of disposing of such materials is about to pick up sharply over the next few years because Google is rapidly scanning tens of millions of volumes, with the intent of making these available online. Large research libraries are willingly participating by making their materials available to Google: Harvard, Michigan, and Oxford Universities, to name a few. Their goal is noble: to make millions of volumes of information available online in a convenient fashion and, soon, searchable. That last function—"searchable"—means enabling a Google search through all the scanned pages for information; for example, one that could list every reference made to oneself.

Once the scanning project is well underway, the temptation for librarians to dispose of their paper copies of books will be enormous because of lack of space and budgets to keep the originals. Their arguments will be exactly the same as what we heard over the past decade with magazines and journals: easy access, convenience, and so forth. The limitations of that strategy will also be the same, most notably the loss of the serendipitous effect of walking down an aisle of books on a topic of interest or the ability to work with the original artifacts as read in their day, compromising our effectiveness as researchers.

One might say, "Impossible, librarians won't do that, they are book people." We have the well-documented ugly case of the mid-20th century when they hastily microfilmed millions of pages of newspapers, in some cases so poorly that one can barely read them, not completely photographing an issue, or all the multiple issues of a paper that appeared daily, and doing it in black and white and losing the rich colors, for example, of late 19th- and early 20th-century iconography. This act was followed by the physical destruction of well over 90 percent of all collections of paper-version newspapers in the United States.1 And now it seems that when we go back to our libraries each fall, an increasing number of journals have disappeared off the shelves over the summer, replaced with a tiny notice to go look them up on the Internet. In short, librarians have proven perfectly capable of disposing and destroying physical copies of vast quantities of materials and of being confident that their reasons were sound and noble.

There are some instances where this strategy can have some significant unintended consequences not yet appreciated by historians, let alone librarians. Take the example of books on computing. Nothing seems so out-of-date than a user manual for PC-DOS, or a book on how to write in a programming language published in 1960, let alone a volume on designing computer architectures published in 1958, or 1978, or even in 1988. All of these are routinely discarded because they are "out-of-date," and, to be sure, rarely checked out, let alone even looked at. But they are the ephemera of one of the early days of the emergence of what historians will certainly someday conclude was one of the most important technologies ever developed by humankind, and clearly of the 20th century.

I have a collection of over 2,000 of these books, dating from the 1930s, a collection put together in support of my research on how computers were used across 40 U.S. industries over the past 60 years.2 These publications discussed the state of the technology in different periods, how they were or should be used, and reflected the attitudes of computer scientists, engineers, users, and the managers who paid for all of these. Because it is nearly impossible to find all of these volumes in convenient collections (or at all), I had to put together this collection to conduct my research. Of those volumes published before 1990 acquired from secondhand book dealers, over 50 percent were discards from academic libraries. I recognize that many of these could be duplicates, but in some cases I seriously doubt they had two copies in the first place. These included volumes that are now considered rare, or minor classics in the field, such as the works of Maurice V. Wilkes, an early computer builder in Great Britain; Edmund C. Berkeley, who published the first trade book on computers in 1949; and Norbert Wiener, who coined the term "cybernetics."3 In the years to come we can expect others to join the list, such as possibly early editions in the Idiot and Dummy series that were sold by the millions but which are also discarded by the millions. We all know that someday historians will need to consult such books to discuss the role of computing in 20th-century life. How many other classes of materials will be lost if already visible trends are allowed to continue? We do not have to worry about 200-year-old books as librarians recognize the need to hold on to most of those. The problem is and will continue to be 19th- and 20th-century publications. Already librarians are reminding us of the poor quality of 19th-century pulp paper and, of course, their desire to move toward digital libraries.

All this is not to say that digital versions of materials are not valuable. Indeed, it is nearly impossible to do research today without consulting them; they are fast to get to, can be easy to use, and in some cases are searchable. But they are only one major source of information that historians have to consult to do their work thoroughly. Could a historian of books appreciate their "application" if they did not see examples, hold them in their hands, and read them? Historians of computing find that they too must touch the machines and use them if they can to appreciate their usefulness when compared to previously available information handling tools, techniques, and devices.

What We Should Do

Historians individually and as a community should help librarians appreciate the value of holding on to individual volumes that make up the ephemera of earlier times and not simply capture an image of those books. One cannot assume that they appreciate the urgency of this issue; assume nothing, and have the obvious discussion with your university librarians about what to save. In short, inject yourself and various historical associations into the decision-making process that determines what is to be saved or discarded.

Each of the specialized historical associations should do this as an ongoing activity, acting as lobbyists on behalf of future historians. Thus, for example, the Society for Spanish and Portuguese Historical Studies Society should protect books on the Iberian Peninsula (many on the Spanish Civil War published in the 1930s and 1940s have recently flooded the market); while diplomatic historians should do the same through their community, others specializing in the history of technology through the Society of the History of Technology, and so forth. In other words, for the foreseeable future, historical associations and individual historians should watch over the fate of books and collections.

The American Historical Association should, in particular, use its prestige, voice, and size to establish a national initiative in the United States to lobby judiciously for the preservation of books before it is too late, and to collaborate with other historical societies in other countries. I am particularly concerned about American and European university libraries and secondarily with those of poorly funded local historical institutes.

The AHA should also apply for grants to fund a major survey of private collections in North America to discover materials that are currently not in the control of librarians, but which can later be acquired once libraries recognize that a particular collection is worth preserving. These collections can include materials on the early history of various technologies of the 20th century (such as computing, telephony, and about the design and use of household appliances) and myriad social practices. Such a survey when done from time to time with the help of book dealers, and data mining of sales patterns of large antiquarian internet portals (such as abebooks.com) could identify valuable collections.

Armed with such information, historians and librarians collaborating could reach out to such collectors to proactively work with them to ensure that there is a plan for the disposition of those collections. In some instances this will mean monitoring growing collections for a decade or more. In the case of books on computers that delay may be a good circumstance because the older the materials become, the more attractive they might be to a university to collect.

In short, a familiar crisis is returning. Instead of newspapers and magazines, the problem is now books. Given where I sit in IBM, I see coming after the books the whole issue of digital archives requiring the attention of historians and not just archivists and librarians. I welcome the arrival of digital files. But I don't look forward to the destruction of important ephemera as the price to pay for digital convenience. Let the pattern of action of historians injecting themselves into the preservation decisions of librarians be set now, before the subsequent battle ensues over archival materials. Let us learn from the history of what happened to newspapers and magazines. If we ignore books now, by the end of this century the number that will have been destroyed or lost in the name of progress, convenience, and fiscal responsibility will have been appalling, mimicking what happened to newspapers. The time to act is now.

—James Cortada has worked at IBM for over 30 years, and has been a member of the AHA since 1970. He is the author or editor of numerous history books. His most recent is a three-volume history of the use of computers in America, The Digital Hand (Oxford University Press, 2004–08). He may be reached at jwcorta@us.ibm.com.

Notes

1. Described in considerable detail by Nicholson Baker, Double Fold: Libraries and the Assault on Paper (New York: Random House, 2001).

2. James W. Cortada, The Digital Hand, 3 vols. (New York: Oxford University Press, 2004–08).

3. Maurice V. Wilkes, The Preparation of Programs for an Electronic Digital Computer (Cambridge, Mass.: Addison-Wesley, 1951); Edmund C. Berkeley, Giant Brains: Or, Machines That Think (New York: John Wiley & Sons, 1949); Norbert Wiener, Cybernetics (New York: John Wiley & Sons, 1948).