The End of History?

Ed. Note: This is an abbreviated version of a talk given at the Ranke Conference cosponsored by Syr­acuse University and the American Historical Association on Oct. 16–18, 1986.

The United States at one time took a proud lead in making its records openly and promptly accessible to the public. It was assumed that in a democracy the government’s records were the public’s. They should be open for use, and there should be no copyright in government records. The great wars of this century temporarily interfered with this concept as secrecy came to be considered essen­tial. Security classification closed many records, but the basic belief in the prin­ciple of open records reasserted itself in the post–World War II era.

A series of executive orders first brought coherence into the classifica­tion system and then began to reinstate some reasonable procedures for opening the files. They all went in the same general direction under Republican and Democratic administrations alike. Until 1980 there was a steady development toward the application of the principle that within a reasonable period of time, the people should have access to the records of their government, and that practical administrative and budgetary procedures had to be instituted to make this principle effective in reality.

That process was dramatically and emphatically reversed on August 1, 1982, when E. O. 12356, a new execu­tive order on security classification, took effect. Designed to get people off the government’s back by keeping records closed to public access, this order did more than put a drastic halt to the process underway in the preceding thirty-five years. It invited the reclassification of previously opened records. It shifted all the presumptions from favor­ing opening to favoring the closing of records. It called for favoring longer over shorter closing periods.

Furthermore, the original White House draft called for no time limits at all; it showed the administration’s pref­erence for the perpetual closing of records, a cherished goal up to now only reached by the perpetual exemption of the operational records of the Central Intelligence Agency from the application of the Freedom of Information Act. Along with the reversal in the pur­pose of the rules came an equally dra­matic reversal in the budgetary imple­mentation. A majority of declassifica­tion positions were abolished. The intended result has been a great and continuing reduction in the declassification process and an equally enormous upswing in the annual net increase in classified documents.

It was appropriately in the Orwellian year of 1984 that the Information Security Oversight Office, the agency charged with supervising this process, in its annual report hailed the reduction in the opening of archives with “special delight” and asserted that it “looks for­ ward to even greater progress” in the same retrograde direction—a prediction fully borne out by subsequent experience. The most recent report shows that the declassification process currently leaves a net increase of almost 7 million pages of classified material in the stacks annually! For Orwell’s fans in the Information Security Oversight Office, the signs of “progress” are dramatic; for the other 330 million of us, they are dis­heartening indeed.

What this means in practice is that serious research in American history will be halted chronologically at some point from the early 1950s to the indefinite future. It also means that historians in other countries, who in the past have often used the American example as an argument for earlier opening of their own government’s records, will no long­er be able to do so.

Instead of British historians arguing with their own government to shift from a fifty- to a thirty-year rule so that they do not have to cross the Atlantic to study their own nation’s history, there may well be influences moving the other way. The determination of the Ameri­can government to keep its own records closed will influence American allies in the same deplorable direction. And this problem is accentuated by the quantity of modern records with the attendant enormous costs.

Another area for concern is the physi­cal deterioration of the acidic paper used for most records and books in the twentieth century. This paper is practi­cally guaranteed to disintegrate chemi­cally after thirty to sixty years, depending on original quality and the condi­tions under which it has been kept.

What this means is that the records of this century, unlike those of the eighteenth and earlier centuries, simply will not survive. A couple of centuries hence, people may well conclude from the survival of a minuscule number of books and scanty archives from the 1980s, that the overwhelming majority of people in our age were illiterate and that, as in most ancient times, only a minute minority could read and write, using acid-free stationery and printing books on acid-free paper.

One should note the relationship be­tween keeping records closed for long periods of time on the one hand and their physical deterioration on the other. The present administration in Wash­ington is, we are sometimes told, begin­ning to concern itself with President Reagan’s place in history. They need not worry; if they have their way, his place will be lost. If the records are kept closed for as long as the administration prefers, they will yellow and crumble into dust by the time historians are finally allowed access to them.

Unless historians take these matters seriously and begin to work on them energetically and successfully, the future of the discipline is dim indeed.

Far from being a peculiarity of this country, the problem is worldwide. When scholars are eventually allowed access to Soviet archives, they will dis­cover that modern paper deteriorates in Eastern as in Western Europe depend­ing on its acidity and conditions of stor­age rather than on the political and social system that produced it. Those European countries that are currently carefully excluding their own and foreign scholars from their recent archives are in fact destroying the basis for their own historical record in this process.

The same thing is true for what is often called the Third World. A disproportionate number of Third World countries are in areas of higher temperature and often of higher humidity than the more developed countries, with the result that their records deteri­orate even more rapidly. Where humid­ity levels are naturally lower, such as Egypt, major hydroelectric projects have created humidity problems. Fur­ther hastening the deterioration process is the fact that many developing nations simply lack the resources for proper care of records. If these countries keep their recent records closed, they face an ironic situation: eventually the only records of their past available to historians are likely to be those papers of their former colonial masters, which were generated before the introduction of wood acetate paper in the late nineteenth century.

There is more bad news to come. The turn to computerization in the last thirty years hastens the danger of societies being left with little or no material for the study of their own past. There is, quite obviously, the disappearance of preliminary drafts and stages in the preparation of important documents. But this is not all. It is not just the drafts that vanish with the push of a button; it is the final clean copy that will disappear as well. Several aspects of the computer revolution combine to create this situa­tion.

First there is the problem of changing hardware. Machine readable records presuppose machines that are literate in those records. Given even the present rate of change in this field, it is safe to predict that by the end of this century almost no machine readable cards, disks, tapes, wires, etc., generated up to now will be readable. There will be no machines capable of reading them, and if any machines survive, there will be no spare parts for them, and if there are spare parts, no one will know how to make them work. This situation already exists. One can work with American census schedules of the nineteenth cen­tury, but not with those of the 1960 census. Of the two machines left in the world that can handle those tapes, one is in Japan and the other has been retired to the Smithsonian Institution (Commit­tee on the Records of Government, Report, Washington, DC, March 1985, p. 9).

The various types of tapes, wires, and disks not only change at such a rate that the new generations of hardware cannot handle them after a few years, but also they deteriorate. We do not really know how long the new materials will last, but the proud promises on the cartons of floppy disks—promising that these disks are guaranteed to last five or as many as ten years—provide a clue. No doubt some of the disks and tapes will last longer, but it is unlikely that many of them will be in physical condition for use in fifty years.

It will not do to answer that there are currently known methods of transferring existing machine-readable infor­mation onto new tapes, disks, etc. For the most part, transferring is not done, and is further complicated by the changes in technology. Those who seri­ously believe that the possibility of con­ version to new or more permanent for­mats will solve the problem are invited to consider the extent to which this society’s ability to produce acid-free pa­per has had any practical effect on their own correspondence and the records they see around them.

There is still another aspect of the machine readable records disaster. The information maintained in one or an­ other form of automated storage is frequently accessible only through index­ing or other entry devices that are in the form of software programs or special scan-sensitive markings on the type or other material. If these software and indexing scan systems are lost or cannot be manipulated by a later generation of machines, the originals are to all intents and purposes lost even if they survive physically. Examples of calamities of this type are already upon us; two in­ volving important records from the Vietnam War are described in the very suggestive recent report of the Commit­ tee on the Records of Government (March 1985, p. 31). There will certainly be more incidents of this kind.

There are two possible ways of engag­ing the prospect unfolding before us. If no basic changes are made, one way will be for future historians to be  trained and to work on the twentieth century essentially the way ancient history is done now. Combining careful analysis of a few fragmentary surviving texts with archaeological and numismatic evi­dence, historians will have to attempt to recover a society characterized by the generation of enormous volumes of rec­ords. They will be working with surviv­als that will be proportionately smaller than those surviving from Athens or Rome and minuscule when contrasted with the comparatively voluminous pa­pyri and cuneiform tablets of ancient Egypt and Mesopotamia. Only the occa­sional university press book on acid free paper will remain of the millions of tomes in the libraries and only letters on rag bond paper will survive out of the vast paper production of the world’s Pentagons.

There is a second possibility. Histori­ans can work together with archivists and librarians in a major effort to publi­cize the problem and possible solutions. There exist feasible solutions for each aspect of the basic issue: long-term clos­ing of records, disintegration of records and the effective disappearance of machine readable records. The answers to each of these that I shall now describe briefly are not necessarily the only possi­ble ones; it is just that no other currently feasible ones have as yet been proposed. The only realistic answer to the prac­tical problem of declassifying the enor­mous volumes of records generated by the modern state is the imposition of a requirement for self-declassification, a requirement that in this country would mean amending the Federal Records Act. Every document that is classified must have included in the initial classifi­cation a declassification schedule with dates in it. No other form of classifica­tion would be valid and no penalty would be imposed on the publication of documents not carrying such a declassi­fication schedule.

A document would therefore carry a notation, for example, that it is top secret, would go down to secret on 1 January 1997, and would be declassified as of 1 January 2007. No further review of the document would be needed un­less the declassification were either to be speeded up or postponed. The majority of the millions of pages of classified documents created each year would never need to be reviewed again at all, and the burden of examining the end­less pages of classified paper would be placed on those who wish to hasten or slow the previously established pace. Since any extension of classification would also have to have termination dates to be valid, the volume of truly long-term document closures would shrink steadily. The backlog of records classified before the imposition of the self-declassification requirement would at least be a finite quantity and could be addressed over a period of perhaps ten years, with the National Archives given broad authority to act itself on docu­ments thirty-years old.

All modern societies that expect to make their records accessible will eventually be forced by the practical prob­lems posed by the classification mess to turn to some variant of the self-declassi­fication system. They will also find it essential to include the costs of declassi­fication in the budgets of the agencies that create the classified documents just as they now include the cost of safes in their budgets. Again, all modern socie­ties will find that their archival institu­tions will either not have the resources for this or would otherwise have to divert an inordinate and increasing pro­ portion of their resources to it. The management and servicing of the mountains of paper the modern state creates is quite enough for archives; the costs of declassification will have to be considered a part of the cost of classify­ing. The sooner self-declassification be­ comes the legally required norm in any country, the sooner that country can look forward to a day when closed records will start to shrink. The longer the delay the bigger the pile will become in the interim.

Let me turn next to the problem of deteriorating paper. There have been some encouraging experiments with the deacidification of paper as a means of saving books and records from com­plete destruction by halting the process of deterioration. Certainly such tech­niques need to be applied, especially to library holdings (Committee on the Rec­ords of Government, Report, Washing­ton, DC, March 1985, pp. 104–5). It is, however, highly unlikely that such methods will ever be applied on an adequate basis to archives; and since in any case they can only halt, not reverse, the deterioration, the available effective technology must take precedence.

The available effective technology is microfilming. This is the one practical way to cope with the problem. Silver halide film lasts and copies can be made for readers and for eventual replace­ment. The National Archives used to carry forward an excellent microcopy program itself and should resume that procedure instead of turning such mat­ters over to commercial operators, who leave the readers uninformed about the identity of editors and the principles of selection and omission. Microfilming done properly and according to archival preservation standards is done best un­der the control of archivists, not sales­ people; and—with the costs again charged to the records-creating agen­cy—must be seen as a function of the permanent custodian of the film. Micro­reproduction in some form is likely to remain a major part of any preservation program. Whatever changes in technol­ogy are ahead, we can be certain that magnification will always be possible.

This reference to technology brings me to the last issue, that of machine readable records and automated storage and retrieval. There are those who hope that standardization will assure the long-term accessibility and physical sur­vival of the material on which the ma­ chine readable information has been placed. This happy event will take place some decades after the Greek Calends; if it ever does come, there will be all the vanishing material created in the mean­ time. The video disk with laser imprint­ed information under an acrylic shield is supposed to be a possible remedy for all these problems and is currently being utilized by the Library of Congress for purposes of long-term preservation of books as well as other materials. One wonders about the machines of future centuries scanning these disks with a technology we cannot now imagine.

As the National Research Council also concluded in a major study recently, there is no effective alternative to micro­filming that does not change the format of the text. What that means is the requirement that agencies which create records of machine readable disks, tapes, or wires must be required to produce and turn over a hard paper copy together with funds for archivists to prepare a microfilm. There is also the possibility of recourse to Computer Output Microfilm (COM), but how use­ful that will be for many records, unless edited in some way, remains to be seen. Microfilm we know we can preserve and make accessible. By the time any other currently anticipated solution to the problem has been developed technically and applied in practice, the overwhelm­ing majority of the existing machine readable records and those created be­tween now and that ever-receding hori­zon will be inaccessible by available hardware or deteriorated beyond recall or more likely both.

The problems spelled out here are both difficult and massive. They threat­en to put an end to the discipline of history as we have come to know it, at least as it is applied to the last decades of the nineteenth century and to the twentieth century as a whole. Unless histori­ans take these matters seriously and begin to work on them energetically and successfully, the future of the discipline is dim indeed.