The role of phonotactic range in the order of acquisition of English consonants

This paper is concerned with quantifying the factors that affect order of acquisition in phonological development in English. As Menn and Stoel-Gammon (1995) observe, the question of the order of acquisition of the phonemes of English is problematic: acquisition varies between children learning the same language; related allophones are acquired at different times; acquiring a phoneme may not mean that the relevant phonemic contrasts have been acquired; position of the target phoneme in the syllable, together with the other contents of the syllable, affects pronunciation. Nevertheless, the literature does contain observed orders of acquisition of English, together with claims concerning the factors determining that order. Below, we first employ multiple regression analysis to determine the extent to which a reported order of acquisition can be accounted for by phonological and distributional factors. We then compare this conventional statistical modelling with a connectionist modelling approach. In the light of the caveats listed above, the current explorations will determine whether a particular reported order of acquisition is susceptible to a parsimonious statistical description – can a large part of the variance be accounted for by only a few variables? The account, if successful, will stand as a falsifiable model which may then be tested against more detailed observational data taking into account the dimensions of the problem discussed by Menn and Stoel-Gammon.