We have focused here on classification systems. But complementary
advances in other KOS are also desirable. In particular, a universal
thesaurus – or at least interoperable thesauri – would complement a
universal phenomenon-based classification. Users could then enter any
search terms and be guided to controlled vocabulary. If we wish to fully
avail ourselves of digital possibilities, and especially the SemanticWeb,
then greater clarification of hierarchical (distinguishing “kind of” narrower
terms from “part of” narrower terms), equivalence (distinguishing different
degrees of similarity) and especially related term relationships (there are
several distinct kinds of related terms recognized in ISO standards, but these
are rarely distinguished in thesauri) are called for [16].
The sort of approach to classification suggested above is much easier in
a digital world. It would have been extremely difficult in an age of card
catalogues to provide multiple subject entry points for a single document.
And facilitating user curiosity in multiple directions would have been even
more challenging. It is possible that the sort of approach recommended here
might complement classificatory approaches developed during an age of
card catalogues. Or perhaps the digital imperative will lead us to the most
significant change in classificatory practice for well over a century. Cutter
and Dewey may not have foreseen the staying power of their classifications;
we may well be at another historical turning point for KOS; how well we
handle it will reverberate throughout society.