FREQUENCY COUNT

views updated

FREQUENCY COUNT. An attempt to discover the number of occurrences of particular units in particular contexts of language use, principally WORDS in TEXTS. Such counts have usually been undertaken to provide a statistical basis for word lists used in the teaching of subjects like SHORTHAND and English as a foreign language (EFL). During the 20c there have been several large-scale frequency counts for English, particularly in the US under the inspiration of the psychologist Edward L. Thorndike, as in The Teacher's Word Book (1921). This was a list of 10,000 words that American children could expect to meet in their general reading. His list was derived from 41 different textual sources which provided 4m running words: 3m from the Bible and the English classics, 0.5m from letters, 0.3m from elementary school readers, 90,000 from newspapers, and 50,000 from general reading. The list was widely acclaimed as a breakthrough in the study and control of vocabulary and inspired many imitators and developers. It was considered a valuable objective measure of the appropriateness of vocabulary in schoolbooks and a basis for the construction of achievement tests in reading, spelling, and vocabulary. Although not so intended, it was also used as a basis for EFL word lists. One such list influenced by Thorndike was Michael WEST'S General Service List of English Words (1953), which helped a generation of EFL lexicographers to develop the notion of a basic ‘defining vocabulary’. The classic counts of the first half of the 20c were done with little or no mechanical assistance. More recently, the use of computers has made the gathering, analysis, and processing of data less laborious and time-consuming and has enlarged the body of texts (the CORPUS) which can be sampled in this way. Thus, the Brown/Lancaster/Oslo/Bergen corpus (started in 1967) has been used to confirm hunches about the predominance of certain features of WRITING, GRAMMAR, and SEMANTICS in particular varieties of English, and the Birmingham COBUILD corpus (started in 1980) has provided lexicographers with new information about COLLOCATION, grammar, and MEANING on which to base their decisions on how to structure DICTIONARY entries. See FREQUENCY OF OCCURRENCE, VOCABULARY CONTROL.