The problems of subject search on the Internet are no different in principle: search engines may permit easy location of verbally expressed topics, b… - Brian Campbell Vickery

Scientific information is faced with the following problem. On the one hand, we have the world’s literature of science and technology, past and present, in many languages; on the other, and enquirer with a question. How to select, from a vast mass of words, the few that are the most closely relevant to an enquiry? It is this selection process that makes use of classification.

The classification of subject matter may be carried out for all sort of special purposes - to arrange books on shelves, to group inventions in patents, to classify the raw materials, intermediates and products of importance to a particular manufacturer, and so on. All such arrangements have their particular uses and their particular problems.

Throughout the nineteenth century, apart from the division in theoretical sciences and arts, classifiers attempted to divide the sciences into two groups. Already they had before them the examples of Francis Bacon (speculative and descriptive) and Hobbes (quantitative and qualitative). For Coleridge, the sciences were either pure (Grammar, Logic, Rhetoric, Mathematics, Metaphysics) or mixed. Arthur Schopenhauer’s similar groups were called pure and empirical, Wilhelm Wundt in 1887 called them formal and empirical, Globot mathematical and theoretical, and the St. Louis Congress of Arts and Sciences (1904) normative and physical. made similar division of the sciences into abstract and concrete