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. - 87 - However, this is not borne out by an examination of the Thesaurus, which suf- fers some inconsistency on this point. For example, Wind tunnel nozzles gives, quite correctly, Nozzles as a BT. But the previous term Wind tunnel models fails to give Models as a BT and under Models gives Wind tunnel models as a Related term {RT). i.e. , a non-generic relation. Again, the term Materials has eighteen RTs; some of them are kinds of material based on Structure {Composite, Granular, etc.) some based on Properties or Behaviour {Radioactive, Magnetic, etc.) some based on Use (Structural, Molding, etc. ). It would appear that the first charac- teristic, at least, designates true species. Again, under Plastics are listed numer- ous resins as NTs. The term Resins itself is treated as a synonym of Polymers, however; but no reference of any kind connects Plastic to Polymers or vice-versa, although there is limited duplication between the NTs for each of these terms. Similarly, there is no connection established between Pumps and Compressors (the latter being treated as a synonym for Air compressor) - although Pumps has num- erous NTs in the form of pumps of particular application, e.g. Fuel pump. (3) RT {Related terms) These are designed toshow non-generic relations (as defined above) and the rules state that it is undesirable to make RTs to 'more specific' terms. However, there are numerous examples in E. J. C. of RTs which do not observe this. For example, Hydraulic equipment has numerous references to particular types of Hydraulic equipment {Hydraulic brakes, Hydraulic presses, etc. ) Apart from the fact that there were cases of true species (e. g. of Hydraulic equipment) being included in the RT framework, problems arose regarding the references from terms like Quartz (as a heat shield); if we assumed that Heat shield is in the relation of RT to Quartz, should we, under Heat shield, have added Quartz as an RT? By analogy with Insulation {say), which gives as RT the material Mag- nesium Oxide, Quartz should have been added. But if it represents (as it does) Quartz as designating a kind of heat shield according to material, it is a 'more specific' term and such references are not encouraged. Generally speaking, E. J.C. observes the old rule of Subject Heading lists which avoids references to adjacent headings on the score that their juxtaposition makes further reminders to the searchers (the question programmers) unnecessary. For example, there is no reference between Shock waves and Shock tubes. So this should lead to the avoidance of a reference from Ionization to Ionosphere (which was, nevertheless, made). Also under Molding materials there are references to four RTs which are adjacent entries beginning with 'Molding . . . ' . Apart from these efforts to observe consistency and method in making refer- ences, the usual variety of relations appeared to be permissible and consequently RTs were added for new terms and for existing E. J. C. terms when these had inade- quate connectives. Examples of the latter situation were fairly*common since Aerodynamics is not a particularly favoured subject in E. J.C. For example, there are no connections between Vibrations and Elasticity, or between Supersonic flow and Shock waves; in the case of Poiseuille flow, {which may be defined as viscous, laminar flow in pipes or between closely parallel planes), the term is rejected and referred to Laminar flow but without any references linking it to the notion of pipe flow, which in this context is just as important as laminar flow. References from post coordinated terms. E. J.C. provides a number of instances in which a term is distributed between two or more wider terms; e. g. , Pressure gas welding [OCRerr]-', had BTs which include