ONE morning a few years ago an envelope arrived from my parents containing the bill from New Rochelle Hospital for my delivery, in 1952. The contents of a basement or attic were being culled, and the bill had turned up in one of the many cardboard reliquaries that have long lent a kind of ballast to my childhood home. The hospital's total charge for a five-day stay, including drugs and phone calls, came to $187.86. I was amazed at the cost, to be sure. But I was also struck by something else: that among all those decades' worth of family documents my parents had looked through, the delivery bill was the only thing they thought of sufficient interest to pass along.
This episode came to mind last fall when the U.S. National Archives and Records Administration announced its decision about whether, in addition to the vast numbers of government documents that it was already required to store, it would keep copies, paper or electronic, of all the government's E-mail. The decision came after years of debate among historians, public-interest groups, and government officials over a question that individual American households also ask themselves: How much stuff is worth saving? Opponents of E-mail storage argued that it would preserve at great expense a lot of useless clutter. Proponents noted that these days E-mail is sometimes the only "paper" trail that survives, and pointed to the role it played in the case of Oliver North and the Iran-contra arms-for-hostages deal.
Last fall the proponents won: the government's E-mail is now being saved.
At some point most of us realize that having a personal archival strategy is an inescapable aspect of modern life: one has to draw the line somewhere. What should the policy be toward children's drawings and report cards? Toward personal letters and canceled checks? Toward family photographs and wedding mementos? Toward favorite but no longer usable articles of clothing? People work out ad hoc answers to such questions, usually erring, I suspect, on the side of over-accrual. My father, who is an artist, still has all his art-school sketchbooks from when he was in his early teens, and he has some 10,000 Polaroid photographs of himself that he took over the years in order to capture details of lighting and drapery. He has a file of newspaper clippings about Fordham football games from the 1930s. Almost everyone seems to save -- or "curate," as archaeologists say -- issues of National Geographic. That is why in garbage landfills copies of that magazine are rarely found in isolation; rather, they are found in herds, when an entire collection has been discarded after an owner has died or moved.
I happen to be an admirer of the archiving impulse and an inveterate archivist at the household level. Though not quite one of those people whom public-health authorities seem to run across every few years, with a house in which neatly bundled stacks of newspaper occupy all but narrow aisles, I do tend to save almost everything that is personal and familial, and even to supplement this private hoard with oddities of a more public nature -- a calling card of Thomas Nast's, for instance, and a baseball bat of Luis Aparicio's, and Kim Philby's copy of The Joy of Cooking.
I cannot help wondering, though, whether as a nation we are compiling archives at a rate that will exceed anyone's ability ever to make sense of them. A number of observers have cited the problem of "information overload" as if it were a recent development, largely the consequence of computers. In truth, the archive backlog has been a problem for millennia. The excavation of thousands of cuneiform tablets in the ancient archives of Ebla, in what is now Syria, was hugely important, but it will be many decades before the tablets are fully translated, and by then further discoveries will no doubt have dug scholars more deeply, as it were, into a hole. A few years ago a Vatican official spent a morning taking me through the rich labyrinths and frescoed recesses of the Vatican Library. "Do you even know what you have?" I asked at one point. He shrugged, and said that although the name of every item probably existed in the records somewhere -- "Here, like this," he said, pulling out an eighteenth-century ledger and pointing to an entry in an elegant hand -- he guessed that no one had actually opened up and looked at two thirds of the collection.
Writing's great advantage over memory has ever been that it allows one to "remember" what one can then forget about -- an invitation to warehousing. The process keeps speeding up, and Roy Williams, a researcher at the California Institute of Technology's Center for Advanced Computing Research, has attempted to calculate how fast. He notes that the amount of information now stored in all printed sources everywhere in the world is roughly equivalent to 200 "petabytes," a petabyte being a quadrillion bytes. In contrast, Williams has calculated, the amount of information that will have accumulated in online media alone by the year 2000 -- that is, in the course of a mere couple of decades -- is expected to be two and a half times as much as that, and he concedes that this figure may be a gross underestimation.
Historians obviously have problems when information is scarce, but it's not hard to see a very different problem emerging as source material becomes spectacularly overabundant. The great historian R. G. Collingwood addressed this issue in 1926 in his Lectures on the Philosophy of History, pointing out why historians at the time approached ancient Greece differently from ancient Rome. For example, he wrote,
Greek history in the fifth century B.C. is a valuable study for the beginner in historical work because there are so few sources for it that the beginner can grasp them as a whole, and proceed to the work of interpreting them for himself.... When, on the other hand, he deals with Roman history of the early Empire, he is embarrassed by the immense mass of the available sources, especially those derived from epigraphy; here, therefore, he is confronted with the opposite problem, the problem of acquiring a sound scholarship or acquaintance with the sources, and the work of interpreting them falls comparatively into the background.
Imagine how much more daunting the work of interpretation becomes with the development of modern national bureaucracies and the national archives they have supplied. The historian Derek Beales, writing recently about the immense task facing any chronicler of the nearly sixty-eight-year reign of the Emperor Franz Josef of Austria-Hungary, observed, "What remains of the remorseless bureaucratic record is far too large for any one person to master." Leave aside the task of assessing an entire epoch and consider what is required in purely physical terms to preserve even a single prominent person's lifetime documentary output. Benjamin Disraeli's correspondence survives down to the level of what today would be an E-mail message: "My darling, I shall be home for dinner at 1/2 pt 7. In haste, Your, Dis." Woodrow Wilson left so much behind that the historian Arthur S. Link spent his entire career at Princeton University annotating and publishing Wilson's personal papers, in sixty-nine volumes.
The National Archives in Washington now has about five billion documents in storage. Nationwide, federal repositories have another 19 million cubic feet of them, which in boxes set end to end would stretch from coast to coast. (Thirty years ago the boxes would have reached only from Washington to Wichita.) There are thousands of other data hoards, public and private, all around the country. Electronic storage, despite its incomparable vastness, may eventually make finding some specific things easier, if one knows what one is looking for, and the Library of Congress and the National Archives have ambitious projects under way. But electronic storage also has some major drawbacks, such as the fact that deterioration and technological obsolescence will require that all electronic data be copied onto new systems every ten years. In any event, the cadre of people available to look into all this material is not getting much larger. The number of professors of American history, for instance, has been relatively constant for years. (There are currently about 6,000.)
Is it preposterous to begin thinking of some of our archives as the new tels? Tels are the mounds that layer upon layer of former cities make; they are everywhere in the Middle East, harboring the archaeological record of thousands of years of human history. But there are too many of them for more than a few ever to be excavated systematically, and understanding what's in even those few takes decades if not centuries. For the rest, the occasional exploratory shaft or trench must suffice. I thought of those shafts and trenches recently when a historian friend mentioned that some of his colleagues, confronted with too great a bulk of sources, resign themselves to "sampling," and others increasingly narrow the focus of their academic inquiries until they find something manageable.
Don't get me wrong: I am not proposing that we discard anything at all. One rarely knows in advance what will turn out to be of interest or importance and what should have gone directly into the oubliette. It is always delightful when something is discovered unexpectedly, as the biblical city Urkesh of the Horites was just a few months ago (inside a tel). But information does have its natural predators, and it may be that sometimes natural processes work out for the best.
A professor of mine once said that the only thing enabling him to complete his study of medieval France was that so many of the records he would otherwise have had to look at were destroyed during the Wars of Religion and the French Revolution. I called him up recently to ask if I remembered his remark correctly. He said yes, I did. And he recalled with particular gratitude a certain bonfire in Carcassonne.
The Atlantic Monthly; May 1996; Backlogs of History; Volume 277, No. 5; pages 20-22.