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. - 74 - original concepts, based closely on the natural languages of the documents indexed, reflected a degree of precoordination which was excessive for our purpose so over- elaborate phrases were now broken up into smaller units, e.g., Biconvex circular arc cross section was broken up into Biconvex cross section and Circular arc cross section; Dissociated frozen hypersonic laminar boundary layer became Dissociated boundary layer, Frozen boundary layer, Hypersonic boundary layer and Laminar boundary layer. If search were subsequently necessary for the original concept, it would still be possible by postcoordination. Meanwhile, this splitting up allowed maximum freedom in distinguishing facets and subfacets (arrays). In other words, the rigidity attending the excessive precoordination typical of the older classification systems (resulting in the obscuring of the multiple relations between facets and sub- facets) was avoided. Formation of concept hierarchies This task proceeded in the normal way, by the now well-established process of facet analysis. However, some of the problems which occur when making a special classification were absent or greatly reduced; at the same time, the unusual basis of the schedules (the 'natural language, concepts, already embodying a certain degree of precoordination) raised some new problems of presentation. These p,,ints are discussed later. The procedure was as follows: the concepts (mostly short phrases like Tumbling entry, Centre of rotation, Crossed flexure pivot, but with some single words, e.g., Strips, Trajectory, Pivot, Inclination) were first grouped into the following major subject areas: Aircraft types and parts Bodies (Aerodynamic) Non-aerodynamic structures Flight: flying operations Fluids, gases, atmosphere Fluid flow: Kinds, Elements (vortices, jets, etc. ) Aerodynamic forces and loads, processes and properties Aeroelasticity, flutter Aerodynamic reference parameters (angle of attack, pl[OCRerr]nform, etc. ) Mechanics, dynamics Heat Research: Experiment, Theory General properties and processes. No particular significance attached to this order; for convenience of reference it ap- proximated to the order of terms in the original Cranfield Facet Classification. Generally speaking it reflected the citation order used in locating concepts; a concept containing notions from more than one area was located under the one appearing first in the above sequence; e. [OCRerr]., Wing-body interference went under Wing-body, no.t Interference; Spherical segment nose went under Nose (Aircraft parts) not Spheri- cal segment (Bodies); Le'ading edge stall went under Leading edge. But where a clear relation, explicit or implicit, existed between two elements of a concept, and reflec- ted a clear precoordinate indexing principle (e. g., subordination of agents to the operations or processes they serve) this was observed, even if it ran counter to the broad rule above; e. g., a Shielding mechanism is a structure (non-aerodynamic)