B004R9Q09U EBOK (27 page)

Dewey’s relentless efforts to create a unified national library system, magnified by his considerable ambition, would prove a potent mix that yielded lasting consequences for American libraries. Dewey’s obsession with efficiency and his strong bent for hierarchical management made him an ideal agent of the industrial age. While his energetic efforts exerted a lasting impact on libraries—many of them salutary—his hyper-controlling personality exerted an unfortunate influence over the subsequent history of American librarians, who have long struggled with excessive bureaucratization and a processcentric work culture that regularly leaves libraries struggling to adapt in a world of fast-changing information technologies. Dewey’s tire
less efforts may have ultimately yielded many public benefits but they proved a mixed blessing for librarians, many of whom spend their careers chafing under the often stultifying management culture that Dewey played a large part in fostering.

When Dewey introduced his cataloging system for the first time in 1876, he tried to attract interest by touting its economic benefits. “The usefulness of these libraries might be greatly increased without additional expenditure,” he wrote, “for with [the Dewey Decimal System’s] aid, the catalogues, shelf lists, indexes, and cross-references essential to this increased usefulness, can be made more economically than by any other method which he has been able to find.”
¹⁰
For American libraries to move forward, Dewey believed, they had to abandon idiosyncratic local practices and adopt common standards. The economic argument was apparent on its face; the continuing explosion of printed books demanded efficiency and promised libraries the opportunity to expend their resources on public service rather than maintaining homegrown catalogs. The adaptation of national standards promised a simple economy of scale. By supplanting the unpredictable judgment of willful librarians with the logic of a centrally administered system, libraries could function more efficiently, expanding their reach by yielding some of their autonomy. It was a classic industrial value proposition, straight from the factory floor: constricting individual autonomy in service of the greater ideal of productivity.

In its original incarnation the Dewey Decimal System called for organizing books into nine top-level “classes,” each with a corresponding beginning number (1 for Philosophy, 2 for Theology, and so on). Each class was further subdivided into ten “divisions,” which in turn were subdivided into 1,000 distinct headings. The system’s original divisions included:

The top-down determinism of the Dewey Decimal System has drawn its share of critics, who deride the classification as overly simplistic and riddled with cultural bias. Nowhere are its biases more explicit than in its self-evident favoritism toward Christianity. For example, within the Theology class, divisions 200-280 are devoted to Christian denominations, while Judaism and Islam must make do with a single number each (296 and 297, respectively). Dewey, in his own defense, readily admitted that his system was far from perfect. “Theoretically, the division of every subject into just nine heads is absurd,” he wrote. “Practically, it is desirable.”
¹¹

Whatever its shortcomings, the Dewey Decimal System has proved remarkably resilient over the years. Although it undergoes periodic revision, its basic contours have remained unchanged for more than a century. It remains the standard in public libraries throughout the United States (and in many other countries). While many academic libraries use the more complex Library of Congress Cataloging system, the Dewey system has outlasted numerous competing systems. It has also survived the transition from the old card catalog to the new online catalogs now in service at most public libraries.

While physical card catalogs have all but disappeared from most libraries, the imprint of Dewey’s industrial library lives on. Most electronic catalogs are in truth simply digitized card catalogs, employing the same bibliographic standards (known in library circles as the MARC format) and cataloging techniques devised for the old card catalogs. The seemingly modern Internet-connected online catalog is a living relic of the nineteenth century.

In recent years Web pundits have invoked Dewey’s catalog as an exemplar of the kind of anachronistic, top-down, institutional on
tology the Internet so effectively subverts. While Dewey’s system surely represents a vestige of industrial age mind-set, it is somewhat unfair to dismiss it as simplistically hierarchical. Although the system does provide seemingly rigid categories, it also allows for access through multiple points of entry, such as cross-referenced subject headings; and critics who deride the one-dimensionality of its numbering scheme tend to overlook the fact that the system actually does provide multidimensional access. But the system’s enduring strength lies in its simplicity and transparency. Almost anyone can grasp the system; it is easy to implement and easy to use. Like the most successful products of the industrial age, it enjoys mass appeal. As we move into the digital era, however, the value of mass appeal is starting to diminish.

In an era of networked systems, rigid hierarchies are becoming less practical than fluid systems able to adapt to the needs of emergent communities. The industrial virtues of mass appeal and transparency are ceding ground to the digital virtues of flexibility and reconfigurability. On these counts, Dewey’s system falls short. In the age of the Internet search engine, the Dewey catalog looks increasingly anachronistic, and almost certainly pressure will continue to mount on libraries to reengineer their old industrial ontologies.

As entrenched as the Dewey system may now seem, it competed with alternative cataloging systems throughout the late nineteenth and early twentieth centuries. Cutter never liked the system, arguing that “its notation would not afford that minuteness of classification which experience taught me to be needed in our library. I did not like (and I do not like) Mr. Dewey’s classification.”
¹²
In addition to Cutter’s system, other classifications like the Bliss system, the Universal Decimal Classification (see
Chapter 11
), and the Library of Congress system all vied for the attention of libraries—indeed, all of them are still in use somewhere in the world. These systems all shared a similarity of approach, relying on a set of descriptive cataloging rules and a hierarchical schedule of subject headings. They also shared the same fundamental limitation: a reliance on proscriptive subject cataloging. In other words, they relied on the efforts of librarians to identify and anticipate the universe of possible subjects. As the universe of available subjects continued to expand, however, this labor-intensive approach began to show cracks in its conceptual seams.

In the 1930s the visionary Indian librarian Shiyali Ramamrita Ranganathan proposed an entirely new approach to cataloging. While his system would fail to catch on, his idea has since proved portentous. As the librarian at Madras University and a prolific writer on library theory (he coined the term “library science” in 1931), Ranganathan wrote a series of books outlining a new model of cataloging known as faceted (or analytico-synthetic) classification. Faceted classification allows for more nuance and specific categories than simple hierarchical subject schemes, by supporting descriptions based on multivalent characteristics, rather than on a single top-down deterministic list of subjects. Ranganathan’s Colon Classification (see
Appendix D
) described a framework by which any document could be broken down in terms of five facets: personality, matter, energy, space, and time. These facets could then be combined and interpolated to describe any piece of information in almost endlessly reconfigurable ways.

For example, a faceted classification of a book about management of the London stock exchange until 1975 might look like this:

Using the faceted system, subject entries could then be generated by taking any of the major facets as a starting point and interpolating the other facets as qualifiers. It is literally multifaceted. The above example might appear in a catalog in any of the following ways:

The system thus allows for a multidimensional subject classification that assumes no particular orientation on the part of the user. Using Ranganathan’s system, any item can be described in a highly specific way yet without imposing a deterministic subject hierarchy. It is a bottom-up classification, rather than a top-down ontology.

Despite its limited real-world track record, Ranganathan’s model has exerted a significant influence, especially on the development of the Library of Congress Subject Headings, which support a limited implementation of faceted classification. In recent years, faceted classification has also acquired considerable cachet in computer programming circles, largely because it lends itself so easily to relational database implementations. At the University of California at Berkeley’s School of Information, researchers have developed a system called Flamenco that allows users to search and browse a collection of 35,000 images from the fine arts collection cataloged by using hierarchical faceted metadata. Commercial software companies like Endeca are marketing Web-based software tools that apply the theory of facets to practical knowledge management applications. Ranganathan’s vision has proved particularly well suited to Web-based search applications because it lends itself to the kind of iterative querying on which so many Web users have learned to rely.

While Ranganathan enjoys a resurgent reputation among software developers and remains widely admired by librarians, his original Colon Classification failed to catch on. The system has never been fully implemented outside India but remains a kind of Platonic object for many librarians and information scientists. Perhaps the simplistic hierarchies of the Dewey and Cutter systems, for all their ontological limitations, exert a deeper psychic pull, invoking our ancient reliance on hierarchical information systems. Could the epigenetic rules of classification (see
Chapter 2
) dispose us toward simpler top-down systems, even as our rational minds recognize the evident superiority of the endless reconfigurability of facets?

As libraries continued to proliferate in the early twentieth century, many of them also became more specialized. A new breed of narrowly focused libraries began to appear: newspaper libraries, business libraries, biology libraries, and so forth. At the beginning of the twentieth century, a new professional field of “special libraries” began to emerge. While Victorian-era librarians concerned themselves almost exclusively with the organization of books, special librarians saw their roles in a different light, less as book curators and more as active participants in organizational ecologies of information. In 1915 the American librarian Ethel Johnson wrote: “The main function of the general library is to make books available. The function of the special library is to make information available.”
¹³
In 1909 a group of these librarians banded together to form the Special Libraries Association. One of its early members, Aksel G. S. Josephson, predicted that eventually librarians “might come to the point where special libraries will not even have whole books, but only such parts of many books as it needs, treating books as well as periodicals on the principle of documentation.”
¹⁴
That principle, predicated on liberating the contents of books from the confines of their physical volumes—harkening back to Panizzi’s original distinction between the book and the work—anticipates the promise of the digital era, the liberation of texts from their physical confines. A new term began to gain currency: “documentation.”