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. - 85 - (it) Confounding of opposites: this was used occasionally[OCRerr] as in Continuurn flow, Use Free molecule flow. (iii) Avoidance of precoordination: consistency here was not assisted by the E, J. C. rules, one of which (T-l) warns agains[OCRerr] being too specific and another (T-4) warns against not being specific enough (in the mailer of bound terms). One arbitrary limitation on the degree of precoordination (to 34 characters) is evidently imposed by the three-column format used in printin[OCRerr] the Thesaurus. But only in a few cases did new combinations exceed two words (e.[OCRerr]r. Mass transfer cooling, Blunt leading; edge, Wing-Body-Tail configurations). This policy included the representation of sorne concepts by an instructed co- ordination of single terms; e.g., Aerodynamic noise Use Aerodynamics x Noise (sound), Dynamic systems Use Dynamic characteristics x Systems, Sounding probes Use Sounding rockets x Space probes, Radiating bod.v Use Radiation x Aerodynamic configuration, Reflected wave peak overpressure Use Shock wave x Reflection x Pressure. This device is not very clearly described in E.J.C. (using the[OCRerr]l[OCRerr]and & devices) and some of the examples of precoordination make the policy no clearer; e.g., under the term Pressure is given a large number of precoordinated phrases (Pressure distribution, Pressure measurement, Pressure gradient, etc.). When a 'new' term Pressure plotting occurred, it was not clear whether to preeoordinate or keep separate or confound as a near-synonym of Pressure measurement. Again, a 'new' term Circular wind tunnel might lead to acceptance by analogy with Circular saws, etc. But should Rectangular wind tunnel and Octagonal wind tunnel be simi- larly distinguished? Sometimes, this so[OCRerr] of economy in precoordination avoiding highly specific new terms, led to strange equivalents such as Root section. Use Foundations x Profile. The record of these rejected terms and phrases, together with the ones to be used in their place, grew to large dimensions and constructed a massive 'lead-i[OCRerr]' vocabulary from the terms and expressions of the natural language to those < tie controlled E.J.C. languages. Over !,500 entries were made for the subs, [OCRerr]..[OCRerr].hic}- totalled 350 documents. It should be noted, however, that a number of these rejects were simply word-form variants, e.g., Oscillatory Use Oscillations; Oscillating Use Oscillations; Oscillatory motion Use Oscillations; Elastic Use Elasticity; Edged Use Edges. Selection of references (1) UF (Use for) These have already been considered above as forming a lead-in vocabulary. (2) BT and NT (Broader terms and Narrower terms) The definition of these two reciprocal relations is reasonably clear in E. J. C. The BT reflects a true generic (Thing/Kind) relation excluding not only the obviously non-generic ones, like operations, or Properties, but also, explicitly, the Whole/ Part relation, which is often loosely associated with the generic. The BT also ex- cludes "generic families constructed on the basis of usage", so Platinum, a member of the class Metal, is not regarded as a member of the class Catalysts since it is only sometimes used as a catalyst. This seems to suggest an even stricter inter- pretation by the E.J.C. of the notion of 'c!ass' - i.e., one which excludes from membership aI1 but ,true, species in the sense that they possess permanent and funda- mental characteristics,' uniquely defining them.