Introduction

Perhaps the most dramatic demonstration of variation in the development of the auxiliary is to be found in Wells (1979a). Using the Bristol child language corpus, Wells studied sixty children's emerging auxiliary forms and meanings between the ages 1;6 and 3;6. For each child the transcripts of nine recordings made at three-monthly intervals were analysed. The children had been selected to provide a representative sample balanced for family background, sex, and season of birth, and naturalistic data were obtained using radio-microphones and tape recorders pre-programmed to record 24 x 90 second samples between 9am and 6pm.

Wells attempted to distinguish between piecemeal development and syntactic rule learning by measuring the interval between the first emergence of auxiliaries (defined as the occurrence of a single form, but excluding negative imperative ‘don't’) and evidence of rule-based auxiliary use (defined as the occurrence of five different major forms). For most children (forty), the figure obtained is the equivalent of one or two recording intervals (i.e. three or six months). Four children, however, attain the criterion on a single occasion (in less than three months), while a further three take at least four occasions (twelve months or more). The fact that over a third of the sample reach the criterion in three months or less is taken by Wells as evidence of rapid rule learning. For the remainder, a more piecemeal approach was indicated.

If variation in the time which elapses between first emergence of auxiliaries and a criterion of mastery really does reflect a single dimension of piecemeal versus analytic learning, this would clearly be of considerable interest to an investigation into individual differences.