12.3 Thesauruses

12.3.1 Origins of the English thesaurus and precursors to Roget

The term thesaurus is neo-Latin, based on the classical Greek word for ‘treasury’, i.e. storehouse for precious objects. But among its various applications in English it is rarely found in its classical sense, and its application by lexicographers to an onomasiological lexical reference was not its earliest English use. From the sixteenth century to the eighteenth the term thesaurus appeared in the title of semasiological dictionaries, such as the alphabetically arranged bilingual Thesaurus Linguae Romanae et Britannicae (Cooper 1565). The term’s application to a conceptually ordered presentation of the lexicon began with Roget’s Thesaurus (1852), yet it continued to be applied to alphabetically ordered language references, as in the American Thesaurus of Slang (1942). Since the later twentieth century, the term thesaurus has been extended to structured information systems created for libraries and information technology. Lexicographers themselves use thesaurus in more than one sense, so that alongside its Rogetian sense of being a concept-driven construction of the lexicon at large, thesaurus may refer to a lexical index to a ‘closed corpus’ (Ilson 2010), i.e. a finite language database such as the Thesaurus of Old English, or a published dictionary, as in the case of the Historical Thesaurus of the Oxford English Dictionary. The term thesaurus is thus polysemous, though its use to refer to a strictly onomasiological description of the English language has been trademarked in the UK as Roget’s Thesaurus.

The first truly onomasiological work with reference to the English language was produced in the seventeenth century by members of the British Royal Society, led by John Wilkins (1614–1672). Published as An Essay Towards a Real Character and a Philosophical Language (1688), it included exploratory discussions on the value of creating a symbolic system to express universal human thoughts, a pervasive theme among contemporary European philosophers such as Descartes and Leibniz. To achieve this goal, Wilkins devised a system of ‘real’ characters (non-alphabetic shapes) that were intended to express human thought directly, without recourse to the words of any individual language. Instead the system should name ‘things and notions’, classified into structured semantic sets so as to function within conceptual hierarchies like those of the modern thesaurus. The overarching conceptual systems were set out in detailed ‘philosophical tables’ (Clauss 1982: 540), drawn up by Wilkins’s colleague William Lloyd. The words/names in the tables were painstakingly cross-referenced to the ‘alphabetical dictionary’, the largest and last section of the Essay (Considine 2008: 298). The Essay’s dictionary presents the inventory of 11,500 English words identified for the things/notions of the theoretical universal language, excluding words which named anything that was specific to place or time (Clauss 1982: 544). Despite its originality, Wilkins’s monumental Essay was not particularly well received outside the Royal Society (Clauss 1982: 542–6), and its grand onomasiological design was eclipsed for nearly two centuries by other types of dictionary constructed on a more limited semantic basis, i.e. covering a set of semantic relations or a specific semantic field. One was the eighteenth-century ‘topical dictionary’, whose authors were the pioneers of terminography in applied sciences and contemporary areas of knowledge (e.g. art, health). Their focus on a particular field of endeavour arguably gave their work a conceptual/onomasiological basis, even within an alphabetical macrostructure (Hüllen 1999). Another specialized type of dictionary – dictionaries of synonyms – can also be seen as maintaining the onomasiological approach in the long interval between Wilkins’s Essay and Roget’s Thesaurus. Although the key words to each synonym cluster are alphabetically listed, the microstructure for each entry presents a semantic/conceptual analysis of the cluster and the finer sense distinctions among them. The model for English synonym dictionaries was the work of Abbé Girard (1718), La Justesse de la langue françoise ou les différentes significations des mots qui passent pour synonymes, which was translated and adapted to the English lexicon by John Trusler (1766). An independent and better known perspective on English synonyms was published later in the century by Hester Piozzi, who had in a previous marriage (as Mrs Thrale) been an acquaintance of Samuel Johnson. Her dictionary’s formidable title: British Synonymy: Or an Attempt to Regulate the Choice of Words in Familiar Conversation (1794) resonates with the standardizing impulse of the eighteenth century. Yet inside the covers it was evidently designed as a contribution to the lexical education of middle-class women, to help them converse elegantly, and ‘intended chiefly for the parlour window’ rather than the library shelf. It was lively and well received – republished twice (Gove 1984). The writers of synonym dictionaries after Piozzi (notably William Taylor, English Synonymes Discriminated (1813); George Crabb, English Synonymes Explained (1816)) reverted to more discursive treatments of sense relations traceable back to Girard (Gove 1984). These various lexicographic endeavours suggest growing interest in semantic fields within the English lexicon, while the alphabetical macrostructure remained the default for presenting lexical information.