Developing a meaningful account of the acquisition of English past tense

Abstract

On the whole, forming the past tense in English is predictable (add '-ed') but there are many exceptions that children need to learn (go-went, think-thought). The frequency of the verb, its meaning, and its sound pattern influence children’s production of errors. However, these factors are not well understood, and they change as children grow up. In this project, an investigation of the earliest stages of past tense acquisition in English will be carried out using two tasks that encourage 2-to-5-yr-olds to produce past tense forms. The data will be analysed with reference to a wide range of frequency, sound and meaning properties of verbs. Analysing all of these factors in a single study will provide a much more accurate picture of the acquisition process than is currently available, and allow evaluation of current theories of past tense acquisition. A database documenting the children's performance on 300 different verbs and categorising the verbs on a large number of established and new properties will eventually be made available to the research community. Acomputational model of past tense acquisition will be produced using the data from the study to improve understanding of how the various verb factors interact to influence production of the past tense. The longer term aim is to use the model to evaluate possible SLT intervention programmes.

Duration of the project

June 2011 - May 2014

Funding body

ESRC grant RES-062-23-3062