Library technology: the black box syndrome
Wilson Library Bulletin (R), (February 1983)PD: Bibliography; Illustration.Abstract
TXT: AUTHOR: Norman Stevens TITLE: Library Technology: The Black Box Syndrome SOURCE: Wilson Library Bulletin 57 475-80 F 1983 COPYRIGHT: The magazine publisher is the copyright holder of this article and it is reproduced with permission. Further reproduction of this article in violation of the copyright is prohibited. To contact the publisher: http://www.hwwilson.com/ What's a black box? It is some ingenious invention designed to revolutionize the way in which we work. In recent years, there has been a widespread if not universal acceptance and use in American libraries of network systems, commercial database searching systems, and turnkey systems from commercial vendors. As those and other library uses of automation have developed and matured, we have all come to accept them as solutions to our problems. Despite the real advantages of these systems, we need to regain some of the skepticism toward mechanical solutions to our problems expressed by Ralph R. Shaw almost thirty years ago. We should heed his admonishment to "think kindly ... of those who take pity upon our benighted state to solve all our problems with machines they have not yet thought about."(FN1) The first black boxes A true library historian could undoubtedly locate a description of some mechanical wonder designed to provide quicker access to the papyrus at the Alexandrian library. As only an amateur library historian, I have not been able to locate such an early example. The earliest example of a black box that I have been able to locate dates only from 1588. Since 1959, the Council on Library Resources has used in its annual report a picture of a medieval scholar at his book-wheel. That picture shows the scholar seated at a large circular device that appears to hold as many as a dozen books as a time, any one of which he can bring to his line of vision by using a foot-activated pedal. This may be the earliest known example of the black box, but it is by no means the last. Our library landscape is littered with a profusion of such devices, most of which, if they are from the distant past, now appear to be somewhat quaint, if not ridiculous, although they may be fundamentally no different in their approach from the devices we now regard as our possible salvation. In the mid-nineteenth century, Charles Coffin Jewett proposed the use of stereotype plates as a means of producing and distributing catalog cards prepared at a central point. At about the same time, Henry Stevens was proposing the use of what he called photobibliography, which involved the direct use of photographs of the title pages of books as a part of catalog entries. While advanced for their time, and probably for that reason never implemented, both of those ideas at least had some merit, because they did seek to address the intellectual problems of sharing cataloging information and of developing catalog entries through the aid of mechanical devices. The late nineteenth century did see in both England and America the development of more mechanical solutions to problems that are illustrative of the black box syndrome. The Indicator Public libraries in England continued to operate with closed stacks during the nineteenth century, and the librarians of the time, while not yet prepared to allow readers direct access to their collections, were concerned about finding some way of indicating to patrons which books in the catalog were available for borrowing. The solution was the "indicator," a device now found only in museums, the first of which was invented in 1863. Quickly thereafter, as many as ten or a dozen various kinds of indicators became available and were put into use. One model of the indicator, typical of the device in general, contained hundreds of small metal shelves placed in a heavy metal frame on a wooden base and covered by a sheet of glass. On each shelf was a miniature metal book representing a book in the collection. At opposite ends of the book were different colored tabs with the catalog number: one color indicated that the book was charged out, and the other color indicated that the book was available for use. After the reader had obtained the catalog number, he or she was thus able to determine the availability of the book without having to bother the library staff. This was the first public inquiry terminal. The indicator also served as the basis for the circulation system: when a book was charged out, the information about the borrower was recorded with the indicator's metal book. For all of its seeming advantages, the indicator had one major flaw. It should be obvious that for a collection of any size the amount of space required for the indicator soon became unmanageable. A collection of about four thousand volumes might require about six to eight feet of counter space for the indicator, but a collection of thirty thousand volumes might require as much as thirty-eight feet. It was the eventual adoption of open stacks that brought about the demise of the indicator. The real problem with the indicator, as is so often the case with black boxes, was not its advantages or disadvantages but the way in which it influenced the thinking of librarians. Writing in 1900, L. Stanley Jast, one of the most prominent British public librarians of that time, addressed "The Mechanical Appliance Craze." Building on an earlier paper by Mr. McAlister, Jast suggested that: The amount of time and attention that is now bestowed upon this side of library work, to the inevitable detriment of its higher and nobler sides, is astonishing and lamentable. Now there is a subtle and fell fascination about these things, which if once allowed full play becomes a positive disease ... I have nothing to say against these things in themselves; it is a question of perspective with which I am concerned... Let us be librarians, not mechanics. That the danger of which I speak is neither imaginary or exaggerated is shown by the strenuous efforts which some of us are making to condition the whole economy of the lending library by the indicator. Those of us who run the indicator, and find it a meritorious piece of mechanism, which serves our purposes satisfactorily, may well object when it becomes a question of the indicator running us.(FN2) The card catalog American librarians were fortunate to have escaped the black box solution of the indicator, perhaps because their libraries moved earlier to open shelves or perhaps simply by accident. The one event, more than any other, which seems to have precipitated American librarian-ship into the black box syndrome was the shift from the printed to the card catalog in the late nineteenth century, when it became apparent that the continuing rapid growth of library collections required a more flexible means of providing information about a library's holdings to its users. We did not miraculously adopt the card catalog as we know it today; the literature of the period seems to be replete with mechanical solutions to this problem. The Library Journal for December 1893, for example, reports on the discussion held earlier that year at the Massachusetts Library Club on the question, "Is There an Impending Revolution in Library Cataloging?" According to that report, the discussion centered largely around mechanical issues, including a consideration of the problems encountered with the card catalog because so few users can consult it at one time. At the meeting, "Mr. Lane said that by using trays which can be taken out, this trouble is partially obviated." In another note, appearing in The Library Journal in 1900, no less an authority than Melvil Dewey refers to such issues as the substitution of single trays for double drawers and the fact that "trays, drawers, pockets, and various devices for convenient work have multiplied with wonderful rapidity." As librarians and libraries sought the black box solution that would create the ideal card catalog, some ingenious devices were forthcoming. On December 8, 1891, Charles N. Judson, who seems to have been an efficiency expert and a frustrated user, not a librarian, patented a card catalog drawer with the ever-popular double rod, designed "to facilitate the examination of the several cards of which the card-catalogue is composed." His solution consisted of utilizing indented cards, with the indention and the text shifting from left to right on a card-by-card basis. As Judson carefully explained in his patent application: "In consulting card-catalogues it is necessary to separate one card from another by inserting the finger or thumb between the cards, pulling the first card away from the pack in order to consult the next, and as the cards are frequently packed in the case quite tightly and stand close to each other it is often difficult to get them apart, causing great annoyance and irritation when haste is desired. The object of my invention is to so form the cards as to enable this to be done readily and at the same time they shall be so inscribed that they can be consulted in the easiest manner possible." The Library Journal for May 1892 contains a fascinating, if somewhat incomprehensible, description of "The New Library Drawer" designed by Frederic Badger, an assistant in the Harvard College Library. "It is not supposed," the article boldly states, "that old libraries can at once adopt it; its method effects too complete a revolution in the old ways; but it is confidently offered to new libraries and those now entering on library work. It claims to surpass all appliances heretofore used by libraries." From the description, for no copy of the actual drawer seems to have survived, it must have been very similar to the kardex units that most of us now use for our serials records. Undoubtedly the most famous and widely promoted black box of that era was Alexander J. Rudolph's fantastic "continuous indexer," which was first announced to the world at the American Library Association's meeting in San Francisco in 1891. Perhaps the most amazing aspect of Rudolph's invention was not the device itself but the extent to which it was promoted, with some success, as capable of solving all cataloging problems, even though it was little more than a different device for displaying catalog cards. In many respects similar to the scholar's book wheel, Rudolph's continuous indexer was simply a means for displaying a series of catalog cards on a revolving belt that could be cranked around in a cabinet until the appropriate cards appeared under a glass plate. Just as the indicator had become, for Jast, a symbol of the black box in England, so it seems this discussion about how to display catalog cards helped set American librarianship off on its quest for mechanical solutions to its problems. Badger's library drawer, Judson's card catalog, Rudolph's continuous indexer, and other black boxes of that time seem slightly ludicrous. We would like to think that common sense ultimately prevailed and that the solution selected, as represented in our present-day card catalog, was the correct one. But was it? Much of the discussion of such devices in the 1890s was ably summarized in an editorial in the December 1892 issue of The Library Journal. "The question recurs, of course, whether the card catalog is, after all, the final form of the library catalog, and whether, having reached the millenium of the card catalog, we shall not have to begin over again on an improved system." Far from having reached a millenium, the card catalog is now in the process of being phased out after less than one hundred years of existence. Fascinated by technology Having resolved, for the time being, the problem of the card catalog, American librarians were able to spend the next half-century or so pursuing their fascination with machines along other avenues. Jast's warning never received much attention in this country, just as it had not in his own country. Fortunately, or unfortunately, no major black box project captured our imagination during that period, although, to some degree, circulation systems bid fair to do so with such splendid innovations as audio-charging. Rather, our imaginations were captured by a variety of lesser problems and less dramatic mechanical solutions to those problems. The space allotted to such solutions in the professional journals, the existence of the American Library Association's Committee on Equipment and Appliances, and some of the contemporary applications of equipment are all testimony to our quest for the perfect black box that would provide for one aspect of library service or another. Neither circulation systems nor other mechanical devices were a real challenge to American ingenuity. So, to some degree, the black box syndrome languished. But we had only to wait until the late 1930s for a new machine challenge, far greater than any seen before, to emerge in the shape of electronic data processing. All of that has led to the current fascination, vastly superior to the mid-nineteenth-century British enchantment with the indicator, that we now have with computers and other applications of contemporary technology to library problems. By the time we were well on our way to the initial stages of today's computer craze, Ralph Blasingame reflected on our love of gadgets, echoing Jast's earlier concerns. "The common aim," he wrote in 1956, "in introducing tools or machines into almost any process is to reduce the time or energy required to perform some operation or to produce a more uniformly satisfactory result. To stick to this rule in the application of what may be referred to as 'gadgets' is sometimes difficult; many of these devices have a kind of fascination for some librarians which occasionally obscures the true economics of their application.... Besides, another danger of the indiscriminate use of mechanical miscellanea is that work may actually be created for them."(FN3) Perhaps, in this last statement, Blasingame foresaw one of the major dangers of the computer that was to come. From the late 1930s through the early 1970s, the potential library applications of computer technology presented an unreal world of their own. Most often, as Ralph Shaw suggested in 1956, we failed to adequately distinguish between the form and the substance, or the word and the deed.(FN4) There were many valuable experiments, many useful developments, and even a few good operational systems. To read or recall the literature of that period, however, one would think that magnificent solutions to all of our problems had been developed. Shaw, who was one of the few contemporary critics of library automation, was frequently chastised for his skepticism, but to read his articles on the subject and compare their contents with the developments of the time only serves to reinforce his wisdom. In reality, a large percentage of the articles, books, and reports published at that time described systems, many of which were as bizarre as the card catalog solutions of the 1890s, that never existed, had serious flaws and were abandoned, or were as widely used as Judson's catalog drawer. "Machines that they have not yet thought about" As bizarre as those systems may seem and as interesting as it might be to examine them in some detail, they somehow seem perfectly sensible in contrast to the visions of the library future built during that same period around the black box "machines that they have not yet thought about." It is far more interesting to examine some of those solutions, which parallel today's black boxes of tomorrow. The advent of computers and their potential application to library services brought with it a fascination with solving the problems of library management by users knowledgeable about machines and frustrated by librarians' failure to respond adequately to the challenge of those machines. Mr. Judson, with his "great annoyance and irritation," was the forerunner of what was, for a period of time, a veritable flood of well-wishers. The first and most famous solution was proposed by the distinguished American scientist, Vannevar Bush, in 1945. Bush foresaw ä future device for individual use, which is a sort of mechanized private file and library. It needs a name, and, to coin one at random, 'memex' will do. A memex is a device in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility."(FN5) Bush describes in some detail the physical appearance and inner workings of a memex, which would be a kind of desk, and its construction around microfilm. "The Encyclopaedia Britannica could be reduced to the volume of a matchbox. A library of a million volumes could be compressed into one end of a desk." Fortunately, Bush did not discuss either the costs of developing a memex or the time when it might become available. Unfortunately, his approach served, in large measure, to suggest that this kind of black box solution to library problems was around the corner. In the early 1960s, under the sponsorship of the Council on Library Resources, J. C. R. Licklider of IBM investigated what he titled Libraries of the Future. Much of his work described the size of the body of human knowledge in mathematical terms, and analysis that is not easy to comprehend. He did predict, in more manageable terms, what we might expect based on certain assumptions about the growth of random-access memory capabilities. "It would be possible," Licklider stated, "to put all of the solid literature of a subfield of science or technology into a single computer memory in 1985.... All this refers to fast, random-access digital memory. How fast? ... only an optimist would hope for access shorter than 0.1 microsecond."(FN6) Project Intrex, conducted at the Massachusetts Institute of Technology in 1965 (but not, I hasten to add, within the library there), was yet another attempt to describe the library of the future, in this case in an effort to describe what the information transfer process at MIT might be like by 1975. By that time, according to calculations, a computer-based system would contain a restricted field of äbout 5,000 books at 300 pages per book and 100 journals, the entire contents of each of which may since its inception approximate 2,500 pages of text."(FN7) By 1971, John G. Kemeny, president of Dartmouth College, was proclaiming his vision of the library of the future. Reverting back to Bush's concept of microfilm, Kemeny suggested that with ultra-microfiche "1,000 miniaturized volumes can be stored in one ordinary card file a foot long," and ä million such volumes could be kept in 1,000 card drawers, with the drawers numbered and the cards numbered within the drawers." Such a simple solution, not even as ingenious as Rudolph's continuous indexer or the indicator, would provide faster access than the conventional library "but, obviously, an automated system can do much better."(FN8) These are a few examples, selected not entirely at random, of black box solutions. They foreshadow the work of those who now tout videodisc technology or who forecast a paperless information society. One of the most recent of these solutions promotes the idea that videodisc/microcomputer technology could store the entire National Union Catalog on one disc and provide an average access time of 2.5 seconds. All this at a cost, including hardware, of less than $5,000--that is, after the initial production costs for the transfer of the information to the videodisc. What went wrong? Why have the solutions of Bush, Licklider, Project Intrex, and Kemeny failed to materialize? Why do their projections already seem as quaint, in their own fashion, as the card catalog solutions of Badger, Judson, and Rudolph? Can we really expect a videodisc of die National Union Catalog to appear in the near future, heralded by a brochure advertising all of its wonders, including the hardware, for less than $5,000? Will the paperless information society be here by the year 2000? The answers seem obvious. For those visions of the future, suggested such a short time ago, many things have gone wrong. Technology has neither developed along the lines the visionaries suggested, nor as quickly as they thought it would. It never does. Microforms have real advantages for many library applications and serve us well, but that medium has not yet begun to, and it is never likely to, serve as a substantial substitute for the printed word. Above all, the demand for these solutions has not been there. The planners failed to take into account all of the costs of their systems, especially in terms of how all of that information was to be transformed miraculously into some new storage medium. Those costs can be staggering and are beyond the present realm of reality for libraries, given their role in society. Project Intrex, for example, in looking at the information transfer system at MIT in 1975, proposed that it would take an annual operating budget of $15 million to provide the kinds of services it envisaged; in 1980-1981, the MIT libraries, which manage traditional library systems involving some use of automation, had a total operating budget of just over $5 million. The best analysis of this aspect of the problem has come from Nicholas Alter of University Microfilms International. Looking at only a comparatively small body of literature, akin in many ways to what Licklider and Project Intrex examined, Alter describes some of the problems. He suggests that "to put full-text computer storage costs in perspective, consider the cost of mounting an online full-text distribution service for doctoral dissertations ... The texts of the thirty thousand new dissertations that UMI receives annually from American universities would take roughly two and a half million hours to key and proof for computer entry, at a labor cost of several million dollars--clearly not a practical option." According to Alter, the additional costs of storage and access "quickly escalate into a multimillion dollar annual investment."(FN10) It is all too easy to be misled by our desire for these black boxes to think that they will somehow materialize without cost. The videodisc of The National Union Catalog will probably fall by the wayside because of the minor matter of the initial production costs for the transfer of the information to the videodisc. That is always the way. We tend to look at the glamour of such solutions and not at the practical reality of how those solutions are to be accomplished. If there is a moral to this examination of the black box syndrome through the library ages, it is a twofold one. First, it is abundantly clear, the black box solution to our problems will always be with us. As the technological base of society increases, so can we expect the scope and vision of those solutions to increase. Second, and this is the true moral, we need to continue to be skeptical of the claims and promises of technology. We have come a long way since Alexandria, since the scholar's wheel, since the indicator, and since the card catalog. We have even come a long way since the first applications of computer technology to library operations and the first visions of the library of the future. Automation and technology have done much to improve library operations in the 1980s; they will do much to further improve library operations in the next two decades. But we should not expect miracles. I, for one, suspect that libraries will not be all that much different by the year 2000 from the way they are now. ADDED MATERIAL Norman Stevens is university librarian at the University of Connecticut, Storrs and a regular columnist for WLB. Scholar's book-wheel Cotgreave's library indicator Footnotes 1. Ralph R. Shaw, "From Fright to Frankenstein," D.C. Libraries 24:10, 1953. 2. L. Stanley Jast, "Some Hindrances to Progress in Public Library Work," Library Association Record 2:83-4, 1900. 3. Ralph Blasingame, "Gadgets: Miscellanea, But Not All Trivia," Library Trends 5:239, 1956. 4. Ralph R. Shaw, "The Form and the Substance," Library Journal 90:567-71, 1965. 5. Vannevar Bush, Äs We May Think," The Atlantic 176:106-7, 1945. 6. J.C.R. Licklider, Libraries of the Future, Cambridge, M.I.T. Press, 1965, p.17-18. 7. Planning Conference on Information Transfer Experiments, Woods Hole, MA, 1965. Intrex Cambridge, M.I.T. Press, 1965, p.106. 8. John G. Kemeny, Man and the Computer, New York, Scribner's, 1972, p.89-90. 9. R. Kent Wood and Robert D. Woolley, An Overview of Videodisc Technology and Some Potential Applications in the Library, Informtion and Instructional Sciences, Syracuse, ERIC Clearinghouse on Information Resources, Syracuse University, 1980, p.25. 10. Nicholas Alter, ÜMI Looks at New Storage and Retrieval Systems," University Microfilms International Newsletter No. 14, (Winter 1981), p.1.
Description
test export from refworks