As of 22 April 2009 this website is 'frozen' in time — see the current IFLA websites

This old website and all of its content will stay on as archive – http://archive.ifla.org

IFLANET home - International Federation of Library Associations and InstitutionsAnnual ConferenceSearchContacts

60th IFLA General Conference - Conference Proceedings - August 21-27, 1994

Sociological aspects of classification

Gerhard J. A. Riesthuis
University of Amsterdam


1. Introduction

Collins Cobuild English Language Dictionary gives under the lemma classification the following text:

1. Classification of things is the activity or process of classifying them into different types. eg The cataloguing and classification of all the plants on the island took many months.

2. A classification is a division or category in a classifying system. eg Your insurance group classification changes when you buy a bigger or more powerful car. Under classify we read:

To classify things such as animals, plants, or books means to divide them into groups that have similar characteristics. eg She disapproved of a library that actually classified books under light romance... Books are classified according to subject area.

Most things can be classified according to many different aspects. Books can be classified according to number of pages, to language of the title, to colour of the binding, the subject, and many other characteristics. If we use language as criterium then we can make two categories like English and Unreadable languages, but also more then 200 categories, like in class 8 of the Universal Decimal C lassification (UDC).

The choice between the two possibilities - 2 or 200 classes - is not made objectively as many claim. A lot of prejudices can play a role. Why are all languages, except English, unreadable and why disapprove some people of libraries that possess book classified as ‘light romance’? In this paper we argue that questions like these are sociological questions, as are several other aspects of classif ying and classifications.

2. Bibliographical classifications

In libraries and documentation the most important classifications are the bibliographical classifications. These are classifications build to classify documents (or descriptions of documents) in groups based on the subject, the bibliographical form, and the geographical and time aspects of the documents (Mills, 1977, ch.5). In most cases ‘document’ means in the context of bibliographical classif ication the content of a document. But in the case of the bibliographical form document can also mean physical document, the carrier of the data content. In the list of Formschlagwörter of the RSWK (1986) the following terms can be found: Aphorismus; Arbeitstransparent, Bibliographie, Diskette [Aphorism; Overheadsheet; Bibliography, Diskette]. Two of these terms refer to content and two to t he physical form. A bibliography on a diskette is possible as is an aphorism on a overhead sheet. In the general auxiliary table for bibliographical form of the UDC are the two kinds of form not clear distinquished. The second edition of the Bliss Classification (BC) makes a clear distinction between physical form and the rest in the auxiliary table for bibliographical form.

In the context of bibliographical classifications geographical and time aspect means the aspects of the content. For a book on China in the fourteenth century publish in the United Kingdom in the nineteenth century the time aspect is the fourteenth century and the geographical aspect China. United Kingdom and nineteenth century are dealt with in descriptive cataloguing.

Why do we place the physical form itself of a document in the systematic catalogue and the geographical and time aspects of the same physical form in descriptive, main catalogue?

The answer is simple if one remembers that till some ten years ago most catalogues were card catalogues, and that for each type of entry a separate card system had to be build. Further one should realize that till some fifty years ago (and for many libraries till may be ten years ago) there was in practice only one kind of physical document: the printed document. If librarians wanted to make entr ies for the physical form in order to find the non-printed documents they had the chioce to build a new separate cardcatalogue or to sort the necessary entries into one of the existing cardsystems. The systematic catalogue was clearly more suited then the alphabetical entry catalogue ordered on authors and titles.

In the online age data about the biliographical form of the content and the bibliographical form of the physical document can and have to be separated, put in separate fields of the record structure used. But old habits have a long life! In many online catalogues and bibliographical databases the means for subject searches are to be used to find document of a special physical kinds, e.g. computer programs on CD-ROM's. Resistance to change is quite common and that is a sociological phenomenon.

3. Subject classification

The most important function of bibliographical classifications is giving access to documents with a sought content. Especially in this function sociological aspects are very important. Several different aspects can be distinguished:

  1. Which subjects are mentioned in the classifications used?
  2. How are subjects ordered in the classifications?
  3. Which classifications are used?

3.1. Subjects mentioned in classifications

For bibliographical classifications it is typical that not all subjects that are possible in theory, are mentioned in the schedules. Hulme stressed that “subject-matter is almost indefinitely divisible. For the power of the mind to frame distinctions is practically without limits. (...) A classification based upon this principle would in practice lead to a universal index of minutely divided subj ect headings and to the abolition of all general headings — a scheme revived from time to time by indexing enthusiasts, but which for library purposes may safely dismissed as an economic absurdity. (...). A class heading is warranted only when a literature in book form has been shown to exists, and the test of the validity of a heading is the degree of accuracy with which it describes the area of subject-matter common to the class..." (Hulme, 1911).

Literary warrant is of course a sociological phenomenon. What is published is more or less dependent on what referees recognize as worthwhile. Referees are part of the scientific community and influenced by the opinions in this community.

The consequence of emphasis on literary warrant is that there are no classes for new subjects! Here we have a very nice paradox: "A dynamic information society depends on subject access to pioneering literature from the dominant paradigms and literature from the marginal paradigms, as this literature is central for the innovation processes. Classification systems are made from yesterday’s concepts of the dominant paradigms. Therefore classification systems are normally not suited to providing subject access to literature from marginal paradigms and pioneering literature in the dominant paradigms" (Poulsen, 1990, p. 183).

There is no easy way out. It is impossible to predict the future scientific, technical and social developments. Every classification will sooner or later become inadequate because too many subjects are failing in it. New subjects emerge from developments in sciences, but also from sociological changes in society. Which subjects are dealt with in classifications depends on the literary warrant, bul also on the question which subjects are en vogue in the social environment of the builders and revisors of the classification. This changes with time. Reasons for the changes in interests are scientific developments and changes in social reality, but also fashion and serendipity. Environment was not a important subject 25 years ago. Super conductivity was a very important subject for some years, but the interest has diminished since the great breakthrough did not come. It is possible that some new subjects appear very fast in the bibliographical classifications, for other it can take a very long time. An example: environmental sciences and computer sciences in the UDC.

Super conductivity, AIDS and in vitro fertilization are examples of scientific development. Changes in social reality that are followed by a change in interest for specific subject are numerous. To give some examples: environmental problems (Werner, 1984), refugees, political changes in Eastern Europe, islam and the growth of free time and tourism in many parts of the world.

Also in library and information science many new subjects can be found. Not only the new media but also conservation of documents, user instruction and coordination of collection building (Conspectus!) are subjects that attract far more publications now than twenty-five years ago. In the first edition of the Bliss Classification the terms "proletariat", "labour class" and "bourgeoisie" are not found (in the second edition this terms are present). In the UDC we find this terms in the general auxiliary table for persons, in sociology and in political science.

Which subjects are found in a classification also depends on the kind of documents that are classified with it. In the Library of Congress Classification (LCC) we find a detailed subdivision of brass bands against only one class in the Universal Decimal Classification (UDC). The LCC is unlike the UDC not only meant for books but also for printed music and recordings.

3.2. Ordening of subjects in classifications

In almost all universal classifications the first subdivision is made according to scientific disciplines. Thus the main classes are: philosophy, theology, economy, medicine, chemistry and so on. This does not mean that all classifications mention the same main disciplines. Only the sovjet Library-Bibliographic-Classification (BBK) has "Marxism-Leninism" as first main class with as first subclass "Works of the classical authors of Marxism-Leninism" (works of Marx, Engels and Lenin). Of course the explanation for this phenomenon has to be found in the sociological-political situation in the Sovjet-Union (Goltvinskaya, 1993, see also Dahlberg, 1992).

Also the relative importance of the various main disciplines is not accidental. The just published second International Medium Edition of the UDC (London, British Standard Institute, 1994) has 61,339 notations. The main table has 54,785 notations. Class 1 and 2 (philosophy and theology), twenty percent of the notational space have 1789 notations, that is 3.26 % of the notaions in the main table. Class 5 (mathematics and natural sciences) has 12,445 notations (22,71 %) and class 6 (Applied sciences, medicine, technology) 28,174 (51,41 % !). In the second edition of the Bliss Classification (BC) the two main classes mentioned have not more then about five procent. This difference is the more striking when we realise that the BC is made for libraries and thus for books, and the UDC in the first place for documentation and thus for articles. Theology and philosophy publish relatively far more in books then the applied and technical sciences.

Often is said that classifications keep related subjects together. These is true in a way, but only in a way. Especially in enumerative bibliographic classifications it is important in which order the principles of division are used. In the field of social-economic history I can made the following classification:

N 600 Labour movement
N 610 Roman-Catholic labour movement
N 611 Trade unions for the textile indsutry
N 612 Labour hours
N 613 Strikes
N 640 Socialist labour movement
N 641 Trade unions for the textile indsutry
N 642 Labour hours
N 643 Strikes

The field labour movement is first subdivided according to ideological background of the movements, then according to branch of industry and than according to subject the organizations deal with. So somebody who has to find literature about strikes has to look under each trade union! In the nineteen-fifties this was no problem: social historians were interested in social movements: studies of a p olitical party or a trade union. In the eighties this changed. Studies about strikes in the textile industries or the labour hours became in the centre of attention. The existing enumerative classifications became inadequate. Here also we have an situation where sociological phenomena has a great influence on bibliographical classifications.

Facetclassifications can be an answer to this problem. Put the organizations in one facet and the subjects in another. If postcoordinated searching is possible, we can search for organizations and for subjects independent from each other. When a fixed facetformula is used then for searching there is no difference with a enumerative classification.

3.3. The use of classifications

Nowadays less libraries and documentation centres use classifications for subject access then fifty years ago. Many librarians and documentalists have the opinion that searching with classifications is too difficult for there users, because it necessary first to find the codes. In many sociological studies it was found that users prefer searching with words from a natural language. There are many use studies in which can be read that searching with title words is very popular with users.

In this situation it is understandable that especially in situations where there is no open access as in most european scientific libraries there is a tendence to replace subject access with classification by subject access with keywords, even if there is in the online age an alternative. It is possible to search with text in the classification schedules and indexes, making finding the correc t notations easier (Vizine-Goetz, 1992).

Many managers of libraries put again the question if controled subject access is necessary at all, when most searches are done with title words. This is an old discussion (Roloff,1970), but till now only very few libraries have stopped subject indexing altogether.

In situations where classifications are used, is the choice of the classification often more determined by sociological and economic reasons than by information science considerations. Reading the relevant american journals gives the impressions that Dewey Decimal Classification (DDC) and the Library of Congress Classification (LCC) are not the best classifications in the world. Yet they are heav ely used, just because so many other libraries use them. Another example is or may be was the Russian Library Bibliographic-Classification (BBK) (Šamurin, 1967, Ch. 13; Roloff, 1976, p.110-114). This classification is build on the foundations of marxism-leninism and libraries in the (former) USSR and East-Block States were more or less obliged to use it, even if it was known that there were no classes suitable for Western social science literature and some libraries had big collections of them (Lorenz, 1993).


Dahlberg, Ingetraut. The basis of a new universal classification system seen from a philosophy of science point of view. - In: Classification research for knowledge representation and organization / ed. by N.J. Williamson and M. Hudon. - Amsterdam [etc.]: Elsevier, 1992. - p. 187-197.

Hulme, E.W. Principles of book libraries. - In: Library association record (1913). The citations are taken from Theory of subject analysis: a sourcebook / ed. by Lois Mai Chan, P.A. Richmond and E. Svenonius. - Littleton : Libraries unlimited, 1985. - p.50-51.

Lorenz, Bernd. Systematische Aufstellung in deutschen wissenschaftlichen Bibliotheken. - 2. erw. und überarb. Aufl. - Wiesbaden: Harrassowitz, 1993. - 111 p. - (Beträge zum Buch- und Bibliothekswesen; Bd. 21).

Mills, J., Broughton, V. Bliss bibliographic classification: second edition. Introduction and auxiliary schedules. - London: Butterworths, 1977. - xiv, 209 p.

Poulsen, Claus. Subject access to new subjects, specific paradigms and surveys: PARADOKS-registration. - In: Libri, 43(1990)3, p. 179-202.

Roloff, Heinrich. Aktuelle Probleme des Sachkatalogs. - In: Biblos, 19(1970), p.82-96.

Roloff, Heinrich. Lehrbuch der Sachkatalogisierung. 4. überarb. Aufl. - München, Saur, 1976.

Regeln für den Schlagwortkatalog / bearb. von der Kommission des Deutschen Bibliotheksinstituts für Sacherschliessung. - 2. erw. Ausg. - Berlin, Deutsches Bibliotheksinstitut, 1991. - xxxiii, 288 p.

Šamurin, E.I. Geschichte der bibliothekarisch-bibliographischen Klassifikation. - Leipzig: VEB Bibliographisches Institut, 1967. - 2 Bd.

Vizine-Goetz, Diane. The Dewey Decimal Classification as an online classification tool. In: Classification research for knowledge representation and organization / ed. by N.J. Williamson and M. Hudon. - Amsterdam [etc.]: Elsevier, 1992. - p. 373-380.

Werner, Ingrid. Zur Systematik "Umwelt", zu einigen inhaltlichen und Abgrenzungsfragen bei der Sacherschließung umweltrelevanter literarischer Dokumente. - In: Probleme der Adaption der Bibliothekarisch-Bibliographischen Klassifikation. - Berlin: Universitätsbibliothek, 1984. - (Schriftenreihe der Universitätsbibliothek Berlin; 49).