A phenomenon approach to faceted classification

English abstract of a Japanese paper presented at the 53th conference of the Japan society of Library and information science (JSLIS), Keio university campus, 22-23 October 2005

Nowadays, faceted classification is a promising answer to the challenges of Internet information retrieval. However, some completion of traditional faceted classification is needed for modern use. In faceted classification schemes, the main classes have usually been structured according to disciplines, that is, the meaning of facets depends on the context of its discipline. Disciplinary approach is not the only possible one: after discussing historical studies of the phenomenon approach, the paper claims that for Internet information retrieval there is a need of freely faceted classification, that is, faceted classification by phenomena approach.

I will focus on the integrative level theory of CRG (Classification Research Group), giving comparison studies of main advantages and disadvantages of the theory, and analyzing why CRG failed to execute the theory further into practice. I will argue that phenomenon-based faceted classification for Internet information retrieval is needed, given considerations about the development of interdisciplinary information and the importance of the user factor in information organization.

Traditional discipline-based classification is not flexible enough to adequately manage multi-disciplinary/cross-disciplinary subjects. A cross-disciplinary subject is that of a study about some particular phenomenon, which is considered from the multiple viewpoint of different disciplines, without separation of discipline boundaries. Coates, in 1969, noticed that traditional disciplinary boundaries could become unsatisfactory with classification of objects. Beghtol, in 1998, argues how the present development of interdisciplinarity makes disciplinary approach insufficient.

Phenomenon classification is proposed. According to Broughton, phenomenon classification is a classification of activities themselves rather than subjects and fields of study. One pioneer of the phenomenon-based approach was J.D. Brown. In the introduction of his Subject Classification, he disposed that the regular or existing subject should be preferred to the [?more general] standpoint of disciplines. But in mainstream library science, Mills strongly criticized to the applicability of phenomenon approach, one of the main arguments being that it would be inappropriate for humanistic disciplines. As he said: "Classification has been described roughly as the bringing together of like things: there is much greater likeness between the economics of petroleum and the economics of the furniture industry, say, than there is between the economics of the petroleum industry and the chemistry of petroleum". After pointing out the three main problems of disciplinary approach, possible application of phenomenon classes was argued by Austin again.

The phenomenon approach was studied theoretically in great detail by the CRG, with reference to the integrative level theory. The concept of integrative levels has appeared in the work of philosophers since the 19th century. Comte was said to be the first to notice the order among the sciences. To him, they could be set in an ascending order from the simplest to the most complex: mathematics, astronomy, physics, chemistry, biology, and sociology. This is the arrangement known as "the integrative levels in nature". Several other philosophers also have pointed out that reality is ordered into a number of integrative levels. Among them are Nicolai Hartmann (1964) and James K. Feibleman (1954). The latter and Joseph Needham have been considered as a basis by the CRG.

CRG interpreted the theory of integrative levels as that the world of things develops from the simple towards the complex with adding of new and different properties. The CRG believed that the theory of integrative levels could be used in a classification system to divide a list of things into a sequence of groups. But CRG never fulfilled it into the practice of a classification scheme.

In theory, integrative level theory can produce several benefits. The phenomenon approach makes classification more natural: a subject is located in the scheme according to its place in the arrangement of the world, rather than limited to a specific discipline, which in time could develop and change. In fact, changes in the ways of scientific communication, or of research itself, could make classifications based on disciplines partially inadequate, while they seem to affect less classifications based on the natural order of phenomena.

CRG later discovered a major problem of integrative level theory: integrative levels do not only involve a linear upward progression, but also branching occurs as a mean of further dividing and sub-dividing. However, a linear order is definitely necessary only for shelf organization in the libraries; digital documents on the Internet can liberate themselves from a unique physical position.

An objective comparison between the two approaches is hard to carry out, for there is no general classification based on the phenomenon approach. But theoretically the two approaches are compatible. According to Feibleman, the external world can be analyzed into the integrative levels: physical, chemical, crystallographic, biological, psychological, and cultural; and science is that division of culture which conducts the search among facts through the method of hypothesis, experiment and verification, leading to the forecast of events and the control over phenomena. So the relation between phenomena and disciplines could be considered in the following way: disciplines can be seen as a particular kind of phenomenon at the cultural level. Hence, if we now think phenomena firstly, we can see disciplines as different approaches/facets of a phenomenon.

Possibility of shifting from disciplines to phenomena as the basis for the ordering of main classes in classification has been brought into light again in recent years, for example, integrative level theory is brought again in 2004 by Gnoli. The ILC project is working on the possibility of a classification based on phenomenon approach instead of disciplinary one. Tests are being done in classifying information in specific domains, for example, animal communication, industrial innovation, and faceted classification itself. Improvement can be obtained by applying some integrative level theory to the well known faceted classification. That is, some effort which is based on faceted classification but with a phenomenon approach instead of traditional disciplinary approach.

A phenomenon approach to faceted classification / Hong Mei = (ILC) — <http://www.iskoi.org/ilc/phenomenon.php> : 2005.09.05 - 2011.07.20 -     [phenomenon.htm until 2011.07.20]