The Myth of the Paperless Office

By Abigail J. Sellen and Richard H. R. Harper. Cambridge, Mass. and London: The MIT Press, 2001. xi, 231 pp. Index. Bibliography. $24.95. ISBN 0-262-19464-3.


Review essay published in American Archivist (Vol. 65, No.1, Spring/Summer 2002)


On the day I first started to work on this review, the New York Times had a front page article on the anxieties generated by the U.S. Patent and Trademark Office's (USPTO) decision to discard paper files after their digitization. Entitled "Ingenuity's blueprints, into history's dustbins," the piece hit some familiar notes: Thomas A. Edison, heritage, marginalia, history being lost, Nicholson Baker. No archivist was quoted or consulted, but a representative of the American Library Association supplied a carefully qualified, if somewhat ungrammatical, claim, "If in fact some if this information is being thrown away and it has not been completely digitalized, that's an issue for us."

The very next day, the Times, carefree of any corporate memory, published an article about why the USPTO must reduce the amount of time it takes to process a patent application, which the paper of record deemed a critical issue in the age of fast developing technology. Among other things, it noted, "Automation of the entire patent application system is a central tactic in the agency's effort to reduce pending time." As I pondered the familiar difficulties of serving two masters, both Ned Ludd and George Jetson, what suddenly came to mind was the realization that, again, no archivist was quoted or consulted. Two days, two articles, two prominent instances of recordkeeping quandaries that archivists could help to solve and, yet, there was no reference to us.

Our inconspicuous absence also haunted my reading of The Myth of the Paperless Office. In this book, the authors seek to answer the question of why paper usage and consumption has gone up steadily during all the years when information technology developments and information technology prognostications have led us to believe that paper was on its way out. Studying work places, work processes, and work flows in a number of organizations, the authors come to the conclusion that paper has certain functionalities ("affordances" is the term used) that make it desirable, valuable, and necessary. We can make better use of paper, they argue, and, by studying its qualities and its applications, we can make much better use of information technology, but we have no reason yet to make a "paperless office" a practical goal.

This assessment has some especially interesting implications for those of us involved with electronic records. The implications of this particular book for everyone in the archival profession, though, should be noted immediately, because of the eerie resonance with the Times' articles about patents. Despite dealing with a host of issues intimately involved in our special area of expertise, Sellen and Harper seem unaware that we exist. This is a research project into a topic that totally neglects all that we have done in the field. All the paraphernalia of our particular intellectual expertise—journals, programs, Ph.D.s etc.—are so much undiscovered territory.

The authors, of course, are not archivists, and they work here primarily on an anthropological model—they go into the field to study records creators in their native habitats—so their lack of awareness of any archival literature might be understandable. But what seems staggering is that in none of the workplaces they visit does any informant mention archivists or records managers or refer to archival or records management practices and policies. Of course, not every organization has a formal records management program, but every organization has some sort of de facto policy in place, if only to deal with financial or human resources records that have high legal profiles. But apparently either nobody said anything at all or nobody said anything that Sellen and Harper thought worth recording. This isn't like the Sherlock Holmes story of the dog that didn't bark. This is more like some alternative universe, where the young Jenkinson was hit by a bus or Schellenberg got into law school, and archivists never left the rare books room.

In our absence, and this is where the book gets really interesting, Sellen and Harper do not neglect records and the concerns they raise; they note them and deal with them—they invent a wheel, even going so far as to devise a graphic of the document life cycle (p. 203). They had to come up with something, as the situations they encountered were so dire: "Banks, law firms, and records offices require enormous amounts of space for their archives. Much of this is to preserve a paper trail of past actions and events. Every letter, every transaction is kept just in case it may ever be needed. In fact, most of these documents are never accessed and never needed. Some of these paper files are kept as legal necessity. Many are kept because they provide a kind of emotional security blanket" (p. 28)

In some ways, this is comfort, albeit cold. Pace Voltaire, it seems, even if archivists didn't exist, it would be necessary to invent them. Just as a practical matter, Sellen and Harper demonstrate that somebody has to resolve that costly tension between legal necessity and angst-driven obsession. And on that intriguing note, this book presents us with the opportunity to evaluate a novel perspective on archival functions. What, we can ask, did these two authors write on their blank slate?

The first note is that, here, records are considered proprietary, in the same sense as a proprietary software application or file format. They are so inextricably linked to their creators that they cannot easily be shared. For example, in a study of records at a chocolate manufacturer, files are deemed so idiosyncratic that "anyone other than the owner would not be able to glean much from the file without the owner's being present to tell them about what was in it and how it was put together." This was because the records "supported rather than constituted the expertise" of the owners (italics in the original) (p. 129). In another company, Sellen and Harper learned the same lesson. Records were of little use (and, in fact, little used) without their creators: "When workers moved on to a new project, the knowledge was in their heads, not in the documents … leveraging this knowledge was best done by bringing together effective project teams as and when necessary" (p. 39). As a result, simply saving files is purposeless, as "documents do not speak for themselves." To create some value, to realize the potential knowledge a record may have, it takes "work to make its meaning, provenance, and importance clear to others" (pp. 133-34).

Sellen and Harper do not spell out how to do that, although they do stress the difficulties. As a result, while not introducing any specifically postmodern gloss to their work, they do hint at records' essential indeterminacy. This comes up in a number of instances. In a discussion of "information ecologies," the authors note that "different forms of information are made useful by their interdependence with other forms of information." Those other forms are various, covering anything from a report, a file, a wall chart or a work plan, but especially "the minds of the people using" the records (pp. 188-89). Those cannot all be retained as archives, so, at this point, we have reached an understanding of a context that cannot easily be captured, with the consequence that fixing meaning to any record becomes entirely problematic or, at least, arbitrary. That possibility is also raised in the discussion of the creation of records. In a study of the compilation of police crime reports, the authors conclude that "accurate crime reporting is a process that by its very nature needs to unfold over time," with numerous participants, phases and considerations (p. 121). What results is an artificial document—not simply evidence of what happened that night, say, but a complex construct of research, dialogue, and social work.

By now, one could reasonably ask, "What's the point of saving anything?" There are, as the book notes, some legal requirements that organizations must respect, regardless of whether the results warrant the effort. Above all, though, Sellen and Harper draw attention to the fact that indeterminacy is a quality that may frustrate objectivity, but can promote use. In that context, what constitutes the archives is less important than how the archives are maintained. Therein lie the disadvantages of paper. "Paper simply does not afford widespread awareness and access for a large audience of potential consumers . . . paper is an outdated, unsuitable technology for preserving and leveraging knowledge of the past" (p. 169). In contrast, technology tells a different story. It fosters widespread and remote access over networks; fast, exhaustive searches of large volumes of information; flexible organization and reorganization of information; links to related material; and easy modification of content (pp. 170-72).

All these qualities promote use, particularly in the sense of establishing a new sense of ownership and of assigning new meaning to information within a new context. They also throw some light on the document life cycle proposed in this work. Sellen and Harper categorize records in terms of "hot, warm, or cold." Hot records are in use now, serve multiple and immediate purposes, need to be readily at hand, and are best utilized in a paper format. Warm records have just served or are just about to serve an immediate need, should also be kept readily at hand and, again, are best in paper (pp. 132-33). Cold records are the "dusty archives" (p. 169). As paper, they have no immediate purpose, nor, as time passes, people leave, and memory fades, do they have advocates in or value to the organization. As represented in the book's vision of a document life cycle, it is only some enabling technology and a conversion to digital form that allows cold records to once again become hot and to reemerge as knowledge, to be researched and re-created for use in other activities and other records (p. 203).

With this, we reach the definition of an archives that justifies its costs, particularly in terms of the return on investment necessary to implement information technology on a significant scale. The authors argue that a retention schedule based on use—hot, warm, cold—lends itself to the intelligent adaptation of technology and work practices. In this schema, paper records always have a place, so we will not at any time soon see a paperless office. But we can and we should, Sellen and Harper think, start seeing paperless archives.

This conclusion does have some tendentious aspects. As noted, the authors are not subject matter experts, particularly in the area of archives; they are acting as ethnographers and thus are largely hostage to the quality of their informants. Obviously, their informants have some faults. None, for example, mentioned any of the larger social issues that animate archivists and routinely justify our work—accountability, heritage, history, etc. None, seemingly, was an archivist or a records manager. What can we gain from reading this book, then?

While we can certainly dispute the fine points and even many of the broad ones, we should appreciate the work's potential value as a medium—it conveys to us the views and opinions of the people we have to work with in any large organization. Sellen's and Harper's informants are the constituents any archivist in a large organization has to reach, especially to ensure that new recordkeeping systems address archival concerns. Their emphasis on the internal usage of records and on return on investment, as the rationale for recordkeeping and for implementing technology, are starting points for discussion and negotiation. If we read this book as a report from a focus group, unmediated by our preconceptions, then we can use it as evidence of the expectations our partners have. Our absence in the picture can give us a clearer view of the landscape into which we have to fit and allow us to refine our analysis of how to implement appropriate recordkeeping strategies.

Robert Horton
Minnesota Historical Society