66th IFLA Council and General
Jerusalem, Israel, 13-18 August
Code Number: 034-130-E
Division Number: IV
Professional Group: Classification and Indexing
Joint Meeting with: -
Meeting Number: 130
Simultaneous Interpretation: No
A new classification for the literature of religion
University College London,
The paper examines general problems in the classification of religious literature, such as the evident Christian bias of existing systems, the proliferation of specialized vocabulary, and the variation in understanding of the meaning of terms between different religions and cultures. The capacity of a facet structured classification to deal with some of these problems is considered, and the newly revised. Class 2 of the UDC is presented, with examples of schedules, and practical implementation.
Classification of religion in the major general schemes of classification
The subject of religion is dealt with in detail in all of the major general systems of library classification. In every one it is accorded main class status, often collocated with the main class for Philosophy, and includes, or is proximate to, related subjects such as Ethics, the Occult and Mysticism. In the Bibliographic Classification it is regarded as a social science, and this perhaps the nearest understanding of the class as it is presented here.
As in all recent revisions of the UDC, this class draws heavily upon the Bliss Bibliographic Classification 2nd edition for both structure and vocabulary. It also represents a modification of the proposal for the revision of Class 2 published in the 1993 volume of Extensions and Corrections to the UDC. While the latter constituted a considerable advance on the existing UDC, both in the introduction of a faceted structure, and in its attempts to correct the very considerable Roman Catholic bias, it overall lacked rigour in the facet analysis and still demonstrated some evident leaning toward Christianity. The version now developed addresses both of those criticisms, and offers a more radical and a more even treatment of the world's religions.
The problem of bias in the classification of religion
A major difficulty in constructing a classification for religious literature is that of avoiding bias (whether real or apparent) toward some specific religion or denomination. In some cases this is unavoidable, and not necessarily undesirable; for example the major 'general-special' published classification for theological books, the New York Union Theological Seminary Classification exhibits a strong Roman Catholic bias - but it is a classification for a Roman Catholic collection and tailored to its content. However, a scheme intended for universal application should be as far as possible free of such imbalance and steps must be taken to eliminate it.
Bias occurs, or is perceived to occur, in three main areas:
- an illogical order, or distribution of notation, that causes one system to appear as dominant
- use of vocabulary that has a strong flavour of one system or is special to that system
- inadequate provision of detail other than for the 'favoured' religion
The classification of religions and faiths
Fundamental to a classification of religious literature is the classification of religions and religious systems themselves. There are numerous different criteria for inclusion and for the arrangement of systems, all of which fall short of the ideal to some degree.
Sociological classifications may base arrangement on demographic criteria, but this is rejected as a principle here, since the statistical evidence for numbers of adherents can be unreliable or inclusion criteria inconsistent.
The approach adopted by most schemes of documentary classification is some variant on chronological order of appearance, combined with 'philosophical' association, so that, for example Indic religions or the monotheistic faiths appear contiguous. A major problem arises in the 20th century when a number of 'new' religious movements appear, young in years but old in their theological provenance. Accordingly, those religions that are clearly developed from an older faith are placed alongside that faith; those that represent a substantial move away from any other belief system, or are predominately original in nature are regarded as independent.
There is no concept of value or priority attached to the order of faiths; each is regarded as having equivalent status, even where this is not reflected notationally. The UDC notation is normally expressive of hierarchical structure, and can be seen to imply subordinate or lower status of a topic with a longer classmark. This is not the case here. Because of the large number of religions included it is not possible to give each one a short class number. The 'big five' faiths (Hinduism, Buddhism, Judaism, Christianity and Islam) are allocated two figure classmarks; others may be given longer notations.
Theoretical problems of classification in religion: terminology
Perceived bias towards a particular system can occur as a result of the vocabulary employed. However the subject matter of the class is handled, a particular natural language has to be used to express it, and this language will inevitably have cultural associations which are to some extent linked to a particular faith. In the case of the modern European languages in which most international business in information is conducted, this background faith is Christianity. The words used for religious concepts in French, English, German or Spanish are words for Christian ideas, and it is virtually impossible to find neutral alternatives.
Awareness of the problem goes some way to addressing it, but it is hard to correct the bias without becoming very 'wordy' and substituting descriptions of concepts for the terms themselves. By introducing into the schedule excessive amounts of supplementary text in the form of scope notes, definitions and explanations, or creating entries that are unmanageable from the point of view of the alphabetical index, new problems are substituted for the old ones, and one is no further forward.
Indicating the 'equivalence' of concepts from different cultures can effect a moderately successful compromise. While there is no absolute correspondence in meaning between the Buddhist idea of Bodhisattva and the Christian concept of sainthood, they do occupy the same conceptual niche in their respective faiths; for practical purposes in classification they can be regarded as equivalent in meaning. This understanding of the variability of concepts underlies the 'clustering' of associated (though not necessarily identical) ideas at a particular location. In this way the notation can be used as a switching language between cultural ideas in the same way that it is used as an inter-lingual switching device, or as a standard between indexing systems.
Correcting bias using facet analytical techniques
Facet analysis is a powerful tool for determining the conceptual structure and managing the vocabulary of a subject. It is also effective in establishing the co-ordinate status of all the members of a facet, and in providing equal levels of detailed subdivision for each (through the medium of systematic construction of compound classmarks).
Amongst the Humanities, religion is particularly suitable for facet analysis. The vocabulary lends itself well to categorization, and the facet structure is evident, with a clear preferred citation order. The principal facets, given in inverted schedule order are as follows:
(Theory and philosophy)
Religious concepts. Theological ideas.
(Evidences of religion). Sacred books. Scriptures
(Agents) Persons in the subject
Social customs and practice. Social theology
Ritual practice and observance
(Structure of religions)
Religious hierarchy and governance
Religious organizations. Associations
Sects and movements
(Religions by various characteristics)
(Systems). Specific religions and faiths
The new Religion class in UDC
The most recent revision of Class 2 exhibits the logical structure outlined above. The main schedule will consist of the enumeration of the systems facet i.e. the listing of the world's faiths. Other facets will be contained in a single special auxiliary schedule for use throughout the class.
The notation used in the proposal of 1993 proved not operationally feasible in terms of consistency of filing order, and was in any event not altogether easy to apply. A single facet indicator (the hyphen), which is used to introduce all of the terms in the auxiliary, therefore replaces it.
Although examples are not frequently encountered, UDC allows two or more terms from a special auxiliary schedule to be added in sequence to a main class number. Where the special auxiliary represents the whole vocabulary of the subject (apart from the principal facet) as is the case here, the special auxiliary can act as a faceted, inverted and retroactively applicable classification scheme for that subject.
When more than one term is compounded with a given religious system, the terms are added in reverse schedule order, (descending numerical order), to maintain the implicit citation order.
Examples of synthesis(using provisional notation):
It is quite clear that the notation used in this way achieves a high degree of specificity with relatively short classmarks.
- Judaism 27 + circumcision -531 + interpretation of Torah -26 = 27-531-26
- Hinduism 23 + monastic orders -77 + celibacy -452 = 23-77-452
- Sikhism 235 + feasts and festivals -56 + food rites and customs -422 = 235-56-422
- Assyrian religion 252.3 + New Year festival -563 + ritual drama -527 + texts -2 = 252.3-563-527-2
- Anglican church 283 + synod -73 + attitudes to marriage -46 = 283-73-46
For works which deal with topics from the auxiliary schedule in a general or comparative way (i.e. where there are several faiths considered, or none in particular) the hyphen subdivisions should be added directly to the base number 2 (e.g. Mysticism 2-58, Religious ethics 2-41).
In order to give guidance on the application of this innovative use of notation, and in the handling of the very many cases of system specific vocabulary, many more examples of combination will be given in the published schedules for Class 2 than is normal practice in UDC. It is hoped that this too, in demonstrating the synthetic capacity of the system, and indicating the great potential for the generation of compound concepts, will confirm the efforts made to eradicate religious and cultural bias in the new class.
Donin, H. H. To be a Jew; a guide to Jewish observance in contemporary life. New York, Basic Books, 1972.
Ellis Davidson, H. R. Gods and myths of Northern Europe. Harmondsworth, Penguin, 1964.
Hinnells, J., ed. Dictionary of religions. Harmondsworth, Penguin, 1997.
Humphreys, C. Buddhism Harmondsworth, Penguin, 1962.
Hurbon, L. Voodoo; truth and fantasy. London, Thames and Hudson, 1995.
Klostermaier, K. A concise encyclopaedia of Hinduism. Washington, D.C., Oneworld Publications, 1998.
Ling, Trevor. A history of religion East and West. London, Macmillan, 1968.
Mills, J. and Broughton, V. eds. Bliss Bibliographic Classification 2nd edition Class P Religion, the Occult, Morals and Ethics. London, Butterworth, 1977.
Netton, I. R. A popular dictionary of Islam. London, Curzon Press, 1992.
Owen Cole, W. and Sambhi, P. S. A popular dictionary of Sikhism. Richmond, Curzon Press, 1990.
Sangharakshita Who is the Buddha? Birmingham, Windhorse Publications, 1994.
Sen, K. M. Hinduism. Harmondsworth, Penguin 1961.
Smart, Ninian. The World's religions. Cambridge, C. U. P. 1989.
Sykes, Egerton. Who's who in non-classical mythology; new edition edited by Alan Kendall. London, J. M. Dent, 1993.
"Universal Decimal Classification proposal for Class 2 Religion". Extensions and Corrections to the UDC Vol.15 1993, 45-63.
Wouk, Herman. This is my God . London, Jonathan Cape, 1959.