180 Knowl. Org. 28(2001)No.4 P. Rafferty: The Representation of Knowledge in Library Classification Schemes The Representation of Knowledge in Library Classification Schemes Pauline Rafferty School of Information Studies, University of Central England in Birmingham, Perry Barr, Birmingham, [email protected] Pauline Rafferty is a senior lecturer in the School of Information Studies, University of Central Eng- land in Birmingham. Her research interests include critical and cultural theory and social semiotics, and the application of these research approaches to information retrieval. She is currently working on a novel indexing methodology project, democratic indexing, with SIS colleague, Rob Hidderley. Pauline Rafferty. (2001). The Representation of Knowledge in Library Classification Schemes. Knowledge Organization, 28(4). 180-191. 18 refs. ABSTRACT: This article explores the representation of knowledge through the discursive practice of ‘general’ or ‘universal’ classification schemes. These classification schemes were constructed with- in a philosophical framework which viewed ‘man’ as the central focus in the universe, which believed in progress through sci- ence and research, and which privileged written documentation over other forms. All major classification schemes are built on clearly identifiable systems of knowledge, and all classification schemes, as discursive formations, regulate the ways in which knowledge is made accessible. Of particular interest in determining how knowledge is represented in classification schemes are the following: – Main classes: classification theorists have attempted to ‘discipline epistemology’ in the sense of imposing main class struc- tures with the view to simplifying access to knowledge in documents for library users. – Notational language: a number of classification theorists were particularly interested in the establishment of symbolic lan- guages through notation. The article considers these aspects of classification theory in relation to: the Dewey Decimal Classification scheme; Otlet and La Fontaine’s Universal Bibliographic Classification and the International Institute of Bibliography; Henry Evelyn Bliss’s Biblio- graphic Classification; and S.R. Ranganathan’s Colon Classification. The Representation of Knowledge in Library wonderment of this taxonomy, the thing we ap- Classification Schemes. prehend in one great leap, the thing that, by means of the fable, is demonstrated as the exotic Foucault began ‘The Order of Things’ by quoting a charm of another system of thought, is the limi- passage from Borges, on the monstrous classification tation of our own, the stark impossibility of of animals in ‘a certain Chinese encyclopaedia’. In this thinking that.’ (Foucault 1974, xv) monstrous classification animals are divided into: Foucault argued that within this taxonomy what ‘‘(a) belonging to the Emperor, (b) embalmed, really ‘transgresses the boundaries of all imagination, (c) tame, (d) sucking pigs, (e) sirens, (f) fabulous, of all possible thought, is simply that alphabetical se- (g) stray dogs, (h) included in the present classi- ries (a, b, c, d) which links each of these categories to fication, (i) frenzied, (j) innumerable, (k) drawn all the others.’ (Foucault 1974, xvi) The use of the al- with a very fine camelhair brush, (l) et cetera, phabetical symbols suggests that relationships exist or (m) having just broken the water pitcher, (n) should exist between the categories thus linked, be- that from a long way off look like flies.’ In the cause of the conventions of the Arabic alphabet, but Knowl. Org. 28(2001)No.4 181 P. Rafferty: The Representation of Knowledge in Library Classification Schemes in relation to our system of thought, that does not fication system developed into a tool consisting of a appear to be happening. For Foucault the other ‘mon- system of classes, made up of Main Classes, each of strous quality’ running through Borges’ taxonomy is which was divided into increasingly specialised sub- the fact that these categories could only exist in and classes; a series of symbols, called notation, which op- through language. They could not be juxtaposed in erate as signs signifying the classes; and an index any other sense, in any other place, in any material which links subject terms and the notational signs. site. Borges’ monstrous encyclopaedic classification is The classification notation in effect becomes a sym- an important starting point for Foucault in his inves- bolic language built on what Ranganathan terms tigations into the historical contingency of discourses ‘natural language’. Method, order and objectivity, of order, power and control. Foucault’s investigations which carry with them the connotations of ‘science’, of the operations of power through localised institu- are achieved in the general library classification tions and practices are the starting point of this paper scheme through the rational structures and conven- which explores the order and juxtaposition of ideal tions of Main Classes which assert forms of logical concepts through the discursive practice of classifica- taxonomy, and through the artificially constructed tion schemes. symbols which bear with them connotations of alge- The philosophy of librarianship might be charac- braic languages. These taxonomies are the product of terised as positivist, embracing science and research, rational, and often pragmatic and functionalist, and modern, rooted in an optimistic interpretation of worldviews. science and society. In the early days of the profes- Berwick Sayers whose highly influential ‘Manual of sion, librarians tended to take the view that progress Library Classification’ was first published in 1926 ar- would be inevitable if knowledge is made accessible. gued that the arrangement of books by men [sic] was During this period knowledge organisation solutions ‘merely a part of the divine instinct that ‘order is took the form of large-scale library classification sys- Heaven’s first law’; for arrangement, or classification tems which were described as general, or in the case of as we call it, lies at the base of every well-managed life UDC, universal. These classification schemes were and occupation.’ (Berwick-Sayers 1955,1) For Berwick constructed within a philosophical framework which Sayers the process of classification is natural, but that viewed man as the central focus in the universe, natural operation is the function of a divine interven- which believed in progress through science and re- tion. Men [sic] are thus defined in relation to an as- search, and which privileged written documentation sumed god. Unsurprisingly, given the interest in edu- over other forms. cation, the explosion of knowledge, and the modern Librarians have traditionally combined high pur- faith in science and method, the late 19th and early 20th pose and idealist aspirations with pragmatic practical centuries was a particularly fertile time in the history solutions. Their creative genius embodied in library of general classification systems. During this time classification schemes attempts at one and the same Melvil Dewey produced the Dewey Decimal Classifi- time to map knowledge, to represent ‘the order of cation scheme which was first published 1876 and is things’ in relation to the ideal, and to organise the ma- probably still the world’s best known scheme; Otlet terial – books on shelves – within the everyday con- and La Fontaine’s International Institute of Bibliogra- fines of the library building. phy was set up in 1895; Henry Evelyn Bliss’s Biblio- graphic Classification was published between 1940 and 1953; and perhaps the most intriguing of all the The General Classification Schemes general schemes, S.R. Ranganathan’s Colon Classifica- tion was published in 1933. General classification schemes for libraries are con- The creators of these systems had in common a de- cerned with mapping knowledge so that subjects are sire to facilitate access to knowledge in books and a differentiated from each other and the relationships belief in the value of science, organisation and between subjects are spatially represented. Classifica- method. They often emphasised the practical purpose tion theorists believe that there is some sort of ‘order of their classification schemes. In the introduction to of things’ and that the order, which relates to the ab- the 1961 abridged English edition of UDC, the editors stract world of ideas, can be made material in the claimed that the UDC should ‘not be regarded as a form of highly conventionalised, symbolically anno- philosophical classification of knowledge, nor is the tated classification schemes. The general library classi- order of subjects of primary importance.’ (BS 1000A 182 Knowl. Org. 28(2001)No.4 P. Rafferty: The Representation of Knowledge in Library Classification Schemes 1961, 6) But UDC was built on a system of knowl- language’, that is socialised language developed in edge: UDC main class order was built on top of the and through communities of speakers, as unscien- order of main classes developed by Melvil Dewey tific and imprecise. which in turn was based on Francis Bacon’s rationalist representation of orders of knowledge. Commenta- tors on classification theory have tended to adopt Main Classes and Disciplining Epistemology fairly ambivalent views about the relationship be- tween library classification and philosophical systems Although in his Canons of Classification (Berwick of knowledge. Berwick Sayers argued that classifica- Sayers 1955,