CRANV1P1 Formation of Index Languages chapter Cyril Cleverdon Jack Mills Michael Keen Cranfield An investigation supported by a grant to Aslib by the National Science Foundation. Use, reproduction, or publication, in whole or in part, is permitted for any purpose of the United States Government. -68 - Should a synonym list accept this also as a type of synonymity? In view of the fact that separate measurement was being made of these other devices it was decided to interpret synonymity as strictly as possible, but the joint use of synonyms and word- forms (for example) was also measured. Another weakness, showing itself as a result of the single term basis, has already been referred to - the inability to cater for synonyms which appear only at the 'concept' level of phrases; e. g. , Flexural centre and Shear centre, Surface friction drag and Skin friction, Uniform surface temperature and Constant wall temperature. These considerations led inevitably to the recognition of 'quasi-synonyms' as a variant of ipure' synonymity. Confounding of quasi-synonyms In this device, those terms are confounded which on some occasions, but not all, are used synonymously; e.g. Subsonic and Subcritical; Compressor, Impeller and Pump; Blunt, Blunted, Bluff and Rounded; Medium, Environment and Surrounding; Region, Atmosphere and Material. A certain overlap appears here with the device of confounding word-forms; on many occasions, different word-forms would be used in a report indiscriminately to convey the same notion. The same overlap would appear, of course, with 'true' synonyms if 'conveying the same meaning' were the sole criterion. But in a single-term index language, the extra element of context is lacking; although the phrases 'Seasonal density variation' and 'Variation of density with season' were virtually synonymous in the reports indexed, if the single terms 'Seasonal' and 'Season' are taken alone they cannot be regarded as synonymous. With quasi-synonyms this restrictive rule did not apply, since acknowledgement of differences conveyed by variations in context is the basis of the device. So variant word-forms were accepted, where applicable, as one type of quasi-synonym, e.g. Flexural and Flexible, Flow and Flowing. The establishment of synonyms and quasi-synonyms was done as the indexing progressed, with the aid of glossaries, classification schedules, etc. We have already noted that theoretically, the only sure way of tracing synonyms is by a close classi- fiction of the field, utilizing the defining functions of classification to expose synonymity between terms used. However, the compilation of classification schedules was a much longer task and, for clerical reasons, the list of synonyms was compiled first, in the manner indicated above. Again, for clerical reasons, the synonyms were worked out fully only in the case of those which were demanded by the search programmes - i. e. , the starting terms from the questions. Confounding of word-forms There is little to be said of this device, which was the simplest of all to estab- lish. The expanded classes consisted of a comprehensive aggregation of all the various forms a given word-root could take, whether with prefixes, suffixes, parti- ciples or gerunds, etc. Examples are: Angle, Angled, Angular and Angularity; Asymptote, Asymptotic and Asymptotically; Axial, Axially and Axis; Blunt, Blunted, Blunting and Bluntness; Bound, Boundary, Bounded and Bounding. Their relations to synonyms and quasi-synonyms have been mentioned and their place in one-place single term hierarchies will be considered in the next section. The terms used in searching, together with their synonyms, word endings and quasi-synonyms, are given in Appendix 5.2.