Changing classifications

When teaching and writing about the classification of documents (‘bibliographic classification’), I try to remember to make the point that document classifications very often draw from, and less often contribute to, more general classifications and taxonomies of knowledge, and of entities in the physical world. So it is worth keeping an eye on classification in general, beyond the confines of the document collection. Two recent developments are worth mentioning in this context.

One example, illustrating how different criteria can lead to different classifications. comes from the natural sciences, specifically mineralogy and geology. There are several long-standing classifications of minerals, based on their chemical composition and crystalline structure. A new approach is to consider also the ways in which minerals form, including weathering, geological activity, and the effects of microscopic life, and creating a mineral taxonomy on the basis of a mineral’s ‘kind, a combination of its chemical composition and structure with its origin. The result is to recognise many more distinct minerals than in the current widely-used classifications. The taxonomy was produced with the aid of machine-learning analysis of large volumes of scientific literature, by contrast with ‘expert consensus’ approach to deriving the earlier taxonomies; which are, of course, still useful for many purposes. An intriguingly different to classification development from that to which we are used in the documentation area.

A second, and very different example, is a presentation of the historical developments in classification of knowledge at large scales in the form of an ‘interactive historical atlas of the disciplines‘ from the University of Geneva, from the earliest times to the present day. Such divisions of knowledge on a disciplinary basis are the basis of the most widely used bibliographic classifications, and it is fascinating to see the many and varied ways in which the disciplines have been arranged, most of which have never influenced the classification of documents; 255 separate classifications are presented in the atlas.

For anyone interested in the development of classification, here must, one would think, be inspiration to be found in these many ways of understanding the structures of disciplinary knowledge.