The principal difference between the two is
that the latter use notations in addition to verbal
expressions to support arrangement of information
objects on shelves or their metadata records in a
catalogue. There are also significant differences in
terms of how subjects and their inter-relationships are
displayed in the two. Right from the times of Cutter
and Dewey the relative merits and demerits of the
two have been debated. Even the first Aslib-Cranfield
study sought to compare the two. It is now fairly
clear that the two are complimentary; classificatory
schemes require alphabetical indexes (Dewey’s
‘Relativ Index’ is perhaps the best example of what
an alphabetical index can do to supplement and
compliment classificatory structures; Ranganathan’s
‘Chain Indexing’ is based on ‘Relativ Index’ and
achieves the same purpose when applied to indexing
a classified catalogue). Alphabetical schemes can
be significantly enhanced by classificatory structures
(the hierarchical display of the vocabulary is now
fairly used by many alphabetical thesauri such as
‘MeSH’. All traditional KOS are artificial languages
and differ from natural languages with respect to their
vocabulary, semantics and grammar. The vocabulary
of KOS is normalised and controlled, and employs
terms with well-defined semantics; in contrast it is
not uncommon to find synonyms and homographs in
natural languages. Just as natural languages classify
words into categories (parts of speech), KOS also
categorise terms in their vocabulary.