Comment on Downing and Oliver's (1973–1974) Experimental Procedure for Establishing a Child's Conception of ‘Word’

Abstract

This article comments on the experimental procedure-followed by Downing and Oliver (1973–1974) which aims at establishing to what extent a child possesses the conception of ‘word’. The procedure consisted of a ‘Yes–No’ game during which a child was asked to respond to phonemes, syllables, phrases, sentences, short words, long words, and non-verbal sounds, after having been trained to respond with a ‘Yes’ if he heard ‘one of something’. Because low internal consistencies were found in two samples (56 English-speaking and 82 Afrikaans-speaking Grade 1 children) factor analyses were carried out to determine how items in the tests group together. In both samples three factors were extracted. Examination of these factors led to the conclusion that children use different criteria for classifying stimuli at different times. Sometimes they listen for ‘one word’, sometimes for ‘one sound’, depending upon the uncertainty which the stimulus generates. It is stated that it is not possible to know whether the child listens for one word or one sound. It is suggested that the solution lies in presenting the child with long words only, which have been established as being representative of the lexicon of a particular age group.