The sisyphean task: The classification of a unitary universe of knowledge in a post-modern world
The idea of a unitary universe of knowledge has persisted throughout most of mankind's recorded history. The influence of new movements in thinking, such as cognitive science and post-modernism, have questioned the old idea of a fundamentally ordered universe of knowledge and introduced new ideas on the "miscellaneous-ness" of information, how and why humans are driven to categorize their world, and the possibility that there is more than one order of the universe or no definitive order at all. How the idea of the universe of knowledge has evolved and been challenged by various disciplines, as well as the impact of this debate upon library classification, is the subject of this paper.
The idea of a unitary universe of knowledge, of a world of natural order that can be quantified, classified, and defined in a specific way, has persisted throughout most of mankind's recorded history. Not only has this been the case for scientists and philosophers thinking on broader terms, but for the many library scientists who believed that the discovery of the true order of the universe of knowledge would yield the best framework for the classification of works in a library. It has only been within the last one hundred years that the idea has been challenged to any great degree, by library scientists, psychologists, and many others. The influence of new movements in thinking, such as cognitive science and post-modernism (the multi-faceted reaction to modernism that basically rejects universals and any kind of objective reality) have questioned the old idea of a fundamentally ordered universe of knowledge and introduced new ideas on the "miscellaneous-ness" of information; how and why humans are driven to categorize their world; and the possibility that there is more than one order of the universe, or no definitive order at all. How is it possible to discover whether there is a unitary universe of knowledge? And why does it matter to library classificationists?
In his essay "The Myth of Sisyphus," Albert Camus tries to make sense of a world dissected by scientific methods and theories, but still feels that the deeper we dig to find "truth" in the world, the less certain we become. He writes that his attempts at dissecting, classifying, and describing himself and the world are "nothing but water slipping through my fingers" (Camus, 1991 p.19). The task of dissecting, classifying, and describing the universe of knowledge - all that can be currently or possibly known by human beings - can easily be compared to the task of Sisyphus, the mythological figure Camus evokes in the title of his essay. According to Greek mythology, Sisyphus was forced to push a boulder up a hill that would keep rolling back down for all eternity as punishment from the Gods. Though ordering the universe of knowledge is not as unpleasant as Sisyphus' labor, coming to an agreement as to what the definitive order should be, or if there is such a thing as a "definitive order" at all, has become equally as elusive and unceasing.
Before the mid-twentieth century, it was taken for granted by most scholars that there was an inherent order to the world. It was believed that the closer we come to uncovering and understanding this order, the more perfect our knowledge should be. In 1605, Francis Bacon, in Of the Proficience and Advancement of Learning Divine and Human, devised a classification system that he felt captured this order. Even though Bacon was a politician and attorney, and not a trained scientist, he nonetheless felt that a new approach to science was needed. Bacon's classification system was revolutionary in that it was the first known system to categorize knowledge in a comprehensive and systematic way. Not only did Bacon feel that there was an inherent order to the universe, but he felt that illuminating this order would allow humans to control nature and improve the lot of humankind for the better. As Robert Flint remarks in his History of the Classification of the Sciences, Bacon's scheme "was a comprehensive and attractive…sketch of the intellectual world, indicating in a striking way, difficult to forget, not only what provinces had been acquired by the human mind, but where, and in what manner, new conquests were still to be made" (Flint, 1904, p.105).
Bacon's achievement in producing his classification scheme is also noteworthy in its influence of later library classification schemes. Early American thinker and statesman, Thomas Jefferson, used Bacon's scheme as inspiration for his library, which he later donated to the Library of Congress - his books as well as his classification system. Though Jefferson's Baconian-influenced organizational system is not currently used by the Library of Congress, it was used until the first decade of the twentieth century. Bacon's system also influenced Melvil Dewey in his quest for the proper and efficient ordering of books. However, Dewey viewed the task of organizing books as a purely practical pursuit and he noted the difference between classifying books and classifying knowledge in general. He did not necessarily believe that the classification of all knowledge was unattainable, however, "the impossibility of making a satisfactory classification of all knowledge as preserved in books, has been appreciated from the first, and nothing of the kind attempted" (Miksa, 1998, p.38).
According to Francis Miksa, the growth in scientific and specialist information production and use gave rise to "the creation of modern library classification theory and technique" in the nineteenth and twentieth centuries (Miksa, 1998, p.56). Among the noteworthy thinkers in the library science field, five in particular have been highly influential and will form the focus of this paper: Ernest Cushing Richardson, Robert Flint, W.C. Berwick Sayers, and Henry Evelyn Bliss, and S.R. Ranganathan.
The work of Ernest Cushing Richardson was "the first concerted attempt by an American librarian to identify basic philosophical principles underlying library classification" (Miksa, 1998, p.57). For Richardson, the universe of knowledge has an inherent order that can be discovered through the study of science. Organization does exist in the world and the goal of science is to uncover that perfection. Nevertheless, not every human endeavor to organize the world necessarily follows the scientific arrangement of knowledge, and that is perfectly acceptable - it is man's way of expressing an artistic vision. Though the classification of library collections should aspire as much as possible to follow the "real order" of knowledge, Richardson admits that this may not always be possible. What may be the most logical order of books may not necessarily be the most practical order. This assertion is probably one of Richardson's most important. He admits that even though such an order in a library would be wonderful, practicality should be a classificationist's and a librarian's primary objective. This is the main difference between theoretical classification and library classification and the reason why he considers the latter an art, "like classification of specimens in a museum" (Richardson, 1935, p.23).
Though the belief in a static and knowable organization of the universe of knowledge was prominent among some library classificationists of the late nineteenth/early twentieth centuries, attitudes began to shift and become more cynical towards the idea. Although science was held in higher regard than most other disciplines when it came to the capacity to best organize knowledge, library theorist and philosopher Robert Flint felt that "no scheme of the sciences can be final and perfect so long as new sciences remain to be formed" (Flint, 1904, p.68) (in Flint's work, it is assumed that he equates "the sciences" to knowledge in general). Nonetheless, he felt that the many attempts throughout history to classify knowledge were certainly not useless.
W.C. Berwick Sayers examined the history of classification systems, and agreed with Flint that "no scheme, by whatever means contrived, can fix permanently the relationships of All-Knowledge" (Sayers, 1958, p.xx). However, Sayers believed that the main classes of knowledge, such as history and science, are fixed and will likely remain unchanged, even though we continue to learn and discover new knowledge. Subclasses will always fluctuate; classes do not. Along these lines, Sayers attempted to make clear the distinction between "natural" and "artificial" qualities of objects. "Natural" qualities of objects are most commonly based upon similarities of function and structure among like things. "Artificial" qualities are often based on subjective features or on "some partial or accidental quality or qualities" (Sayers, 1958, p.10).
Sayers pointed out that using either of these qualities as a basis for classifying knowledge is acceptable. However, the objective of the classification must always be kept in mind. It is not the particular qualities used that are important, but whether the classification scheme created is useful for those using it to find the desired information.
Henry Evelyn Bliss shared Sayers' opinion that general classes of knowledge are relatively fixed and "the more consistent the general plan is with the established systems of science, the more efficient and the more permanent the system of classification will prove to be" (Bliss, 1929, p.xiii). He criticizes "the established systems of classification for libraries" because they are merely practical. Those that created these systems "did not see that the better the classification conforms to the system of science the more serviceable it will be, the more efficient, the more practical" (Bliss, 1929, p.411-412).
For Bliss, the universe must be organized in some specific way (where system is "nearly" synonymous with organization, the universe is the "supreme system of systems" (Bliss, 1929, p.169)), because otherwise, our world would be chaotic and almost unlivable. He expresses the need for organization and the need to study the organization of the universe through science, as these are just as important and necessary as food, shelter, and tools for the "continued sustenance, the desired welfare, and the developing civilization of humanity" (Bliss, 1929, p.3). However, it is difficult for humans to discover the exact order because the human mind is influenced by ideals, backgrounds and experiences that do not necessarily contribute to our understanding of the true organization of the universe. None of these factors add up to experiential knowledge, according to Bliss, and will only lead to disorganization if not kept in check. As Bliss felt that none of the library classification systems at the time measured up to the true organizational system of the universe, he created his own classification system called the Bibliographic Classification or Bliss Classification.
In order to support his argument for the stability of organization in the universe, Bliss attempted to pin down the definitions and differences between classes and concepts. He wanted to be able to say what a class is and why we should be confident in using the term to organize the universe of knowledge. He began by defining class as "things related by some likeness" (Bliss, 1929, p.119) and concept (or class-concept, which seem to be the same thing in his mind) as "the mental correlate of the class" (Bliss, 1929, p.120). Class refers to the kinds of objects in the world themselves while concept refers to how we perceive these objects. A concept is subjective; a class, though possibly "conceptual and relative" (Bliss, 1929, p.123), is not necessarily so.
W.C. Berwick Sayers's student, S.R. Ranganathan, felt differently from Bliss in regards to the necessity of a fixed, hierarchical (or, enumerative) classification system for library materials. Ranganathan, an Indian mathematician who studied librarianship in England, designed his "colon" classification scheme in response to a "Eureka!" moment he had while visiting a department store in London. He encountered a Meccano set, a collection of construction pieces (such as wheels, screws, rods, and string) that children could use to construct an infinite number of toys. The Meccano pieces reminded Ranganathan of language itself - how the alphabet is arranged and rearranged to form words, sentences, and, ultimately, works. "These ideas," Ranganathan wrote, "gave me the courage to think that there was nothing wrong in building up class numbers as in a Meccano set…[and] I was encouraged to pursue the designing of the Colon Classification as a type of Faceted Scheme for Classification" (Ranganathan, 1965, p.15).
The idea that a classification scheme could incorporate multiple ideas and subjects, or facets, for one information object was a radical one, but Ranganathan felt that this was the best way to categorize information objects that may fit into several different subject areas. The problem with classification, Ranganathan believed, is mainly due to language, which is a barrier to proper expression of our ideas. Through notational language that "could at least theoretically function as a system of material signs signifying aspects of individual experience" (Rafferty, 2001, p.188), Ranganathan saw that classification of knowledge, as well as library materials, could benefit from the reduction of "natural language" from the equation, as well as a recognition that within subjects there is a repetition of terms and ideas. There is not enough time and space to delve too deeply into faceted classification here, but suffice it to say that the idea of "building up" an information object's class as opposed to forcing it into a fixed system changed the landscape of library classification. However, Ranganathan's Colon Classification has not gained the popularity or wide-spread use of the Dewey Decimal Classification system due to its complexity and the difficulties in applying it to information entities (Broughton, 1929, p.34). Nonetheless, the influence of facet analysis can be seen in the Dewey Decimal Classification and the Universal Decimal Classification schemes, which are no longer wholly enumerative like the Library of Congress Classification scheme.
The growing discomfort with fixed, hierarchical categories of knowledge in the library world in the twentieth century mirrored that of other investigations into the universe of knowledge in general. Around the mid-twentieth century, the newly developing and blossoming field of cognitive science initiated debates concerning how the universe of knowledge should be classified using more experimental methods. Instead of simply assuming that the universe must be ordered in a certain way, many cognitive scientists examined psychological, anthropological, and linguistic sources and performed experiments on human subjects to ascertain if we are even capable of discovering the true order of the universe, assuming that there is one.
What is generally called the "classical view of classification" by cognitive scientists within the literature was the favored theory towards the beginning of the movement. According to George Lakoff (1987), the classical view espouses that "things are categorized together on the basis of what they have in common…things were assumed to be in the same category if and only if they had certain properties in common. And the properties they had in common were taken as defining the category" (p.5, 6). In addition, according to the classical view, categories are defined by essential properties and exist outside the human mind and, therefore, it is possible that "there is a correct, God's eye view of the world - a single correct way of understanding what is and what is not true" (Lakoff, 1987, p.9).
Lakoff and other cognitive scientists address the many problems with the classical view, but it was Eleanor Rosch and her prototype theory that probably did the most damage to the idea of a unitary universe of knowledge. After many experiments using human subjects, particularly in determining how humans from different cultures perceive colors differently, Rosch concluded that the insistence on defining features of categories does not take into account how human minds actually view categories (Rosch, 1978).
Humans naturally think of particular objects as better representatives of their respective categories because the objects seem to share more features of a particular class with other objects of that class and share fewer features with objects outside their class. When asked to think of a bird, humans are more likely to think of a robin than a penguin or a chicken because "it is more readily recognized as a bird and less likely to be misclassified as a member of another category (such as a mammal or a fish)" (Gardner, 1987, p.346). This is what Rosch calls the "prototype theory" because humans tend to recognize categories based upon how closely an object resembles a prototypical member of a particular category rather than by its defining features, if they can be said to exist at all.
Gardner (1987) points out, "categories reflect the perceptual structure of the perceiver" and not the inherent order of the universe, but if this is true then how can any classification scheme, much less a library classification scheme, be held up as better than any other? The answer, according to David Weinberger (2007), is that no scheme can be held up as better than any other - and that is a good thing because knowledge is not a concrete reality, but a fluid notion of how each individual understands and assesses his or her world over time. Weinberger insists that knowledge is limited mainly by two things: 1) physical limitations of only being able to place something in one physical spot and 2) our mental limitations and our desire to categorize (think of the old saying, "a place for everything and everything in its place" - not only physically, but mentally as well). The ultimate in accessibility to the universe of knowledge is to discard the traditional manner of defining and classifying concepts, particularly the idea that there is a unitary universe of knowledge that is ordered in one particular way. Weinberger notes that "when we draw a map of knowledge, it is all too easy to assume that knowledge is a territory that can be subjugated by applying a rigorous and relentless methodology" (Weinberger, 2007, p.91). This is a typical, post-modern viewpoint that, according to Miksa, "rejects the idea that there is some rendition of the universe of knowledge that reflects some absolute structure of subjects and their relationships" (Miska, 1998, p.86).
In this post-modern, digital world, Weinberger's ideas are gaining in popularity. Classification of digital information objects is made easier by the fact that they do not take up physical space, like books in a library. So not only does the idea of an objective, definitive classification scheme of all knowledge seem restrictive and arbitrary, library classification, which has historically concerned itself only with physical information objects, is seen as failing "to represent…documents at a level of specificity that is required and desired by the users" (Mai, 2003, p.9). In this sense, designing one classification scheme that is best for many is less desirable than an infinitely flexible scheme or multiple schemes that cater to the individual user. In libraries, organizing their physical information objects using a classification system in open stacks may give way to a closed system where the various classification schemes are played out on a computer screen and not on the shelves. Users would then be able to locate library materials according to the organizational system that makes the most sense to them.
Despite the appeal of classification systems designed purely with the individual in mind, many still struggle with the idea that there is no universal organization of knowledge because some pieces of "knowledge" seem so obviously defined, so clear in their structure and purpose. For example, we acknowledge that a chair can be used in a variety of different ways (e.g., one can stand on it to reach the top of the bookcase to dust, one can prop it up against a door underneath the doorknob to keep the door closed, etc.). However, it is also generally acknowledged that the primary purpose and use of a chair is to sit upon it. To classify a chair under "things that can be stood upon to reach high places" may not be a satisfactory way of classifying a chair for most people. So even though it may be interesting, and sometimes useful, to classify concepts in a variety of different ways, ultimately, human culture and the way the human mind is wired lead us to order our worlds in similar ways. Eleanor Rosch speaks of this in her experiments concerning the prototype theory. When presented with certain categories or category systems and "if such a system becomes markedly out of phase with real-world constraints, it will probably tend to evolve to be more in line with those constraints - with redefinition of attributes ensuing if necessary" (Rosch, 1978). In other words, humans tend to categorize within their comfort zone, which is usually heavily influenced by culture and individual perceptions of the world. Although classifying concepts in a multitude of ways may be useful in particular circumstances (such as classifying a chair as "something one sits upon" or "something that can be leaned against a door to keep something out"), it is important to keep these classifications within boundaries that are also recognizable and useful for others as well. Otherwise, as Rosch points out, such categorization will likely adapt or fall out of use.
Although scientists may not always agree on how to classify knowledge, classification remains an instinctual activity of the human mind. Librarians do not have a monopoly on how to organize knowledge using a library classification system, but they certainly have a stake in how this task is approached and regarded by others, be they other librarians, cognitive scientists, or those who put all their faith in the new digital revolution.
The classical view of the universe of knowledge as an inherently unitary and ordered system may be outdated and inconclusive, yet many people remain compelled to approach the world around them as though knowledge were stable, recognizable, and possessing an underlying orderliness. However, in this "postmodern world" there are a growing number of people who feel that the traditional manner of classifying knowledge (and by association, library materials - physical or otherwise) is too restrictive.
If Weinberger is correct in his assumption that "we have to get rid of the idea that there's a best way of organizing the world" (Weinberger, 2007, p.10), then librarians must heed the words of W. C. Berwick Sayers and Melvil Dewey now more than ever - in order to remain relevant, librarians must maintain systems that are useful, regardless of whether they fit neatly into the traditional classifications of the universe of knowledge or the subjective (dis)order of the digital world. Like Sisyphus and his never-ending task, librarians must accept that the path to a useful classification of knowledge may never have a definite conclusion; that the universe of knowledge is a continually evolving conversation, not necessarily a puzzle to be solved. This is not to say, however, that traditional classification schemes are outdated and useless. There are many years of thought and experience built into these systems, many of which have evolved significantly over time and continue to do so. The most commonly used classification systems in use today, such as the Dewey Decimal Classification and Library of Congress Classification systems, were designed with the full awareness that knowledge is not static, but ceaselessly expanding and changing. Even if the universe of knowledge is perpetually elusive and continually redefined, library classification must remain true to its roots as it continues to be open to change.
Ranganathan, S.R. (1965). The colon classification. In S. Artandi (Ed.), Rutgers Series for the Intellectual Organization of Knowledge (Vol. 4). New Brunswick, NJ: Graduate School of Library Service, Rutgers.
Rosch, E. (1978). Principles of categorization. University of California at Berkeley. Retrieved 5/13/07 from http://originresearch.com/documents/rosch1b.htm
Karen Snow is currently a doctoral candidate in the Interdisciplinary Information Science Ph.D. Program at the University of North Texas in Denton, Texas. Her research interests include the organization of information, library classification (current and historical), cataloging, and cataloging education.
Copyright, 2013 Library Student Journal | Contact