Thesaurus

A thesaurus may be a compilation of some or all words symbols, phonemes or other units of one or more natural or artificial abridged or unabridged languages. Originally the thesaurus was monolingual, the oldest modern example being the nineteenth century English compilation of synonyms (with some antonyms) by Dr P Roget. These synonyms were published under nearly a thousand thematic or analytic headings based on a classification of knowledge similar to a library retrieval system. More broadly a monolingual thesaurus came to mean a compilation, potentially, of all the semantic units of a language, with its structure reflecting a definition of relationships of the units. In practice, as thesauri developed for computerized information retrieval, the entries or 'descriptors' became the focal lexical units. Synonymous descriptors are correlated with each of the primary ones and in addition, relational descriptors are added and coded, such as higher class or top term, or sub-class, lower term. These hierarchial, generic vs specific relationships which are logical, are distinguished from caused or constraining relationships such as first, following or means, end, and also from associative descriptors such as man, dog. A catalog or index of descriptor types might indicate as many as seven or eight frames of reference, and possibly more. In its philosophical or epistemological aspects a thesaurus may contain information about some aspects of reality and also information about information, metainformation, that allows artificial intelligence to add to its store of semantic units and relationships.