Determinism in new-dialect formation and the genesis of New Zealand English

In this paper we use an extensive archive of early New Zealand speakers, to gether with comparisons with the other Southern Hemisphere varieties of Eng lish, to argue that dialect mixture and new-dialect formation are not hapha zard processes. We demonstrate that, given sufficient linguistic informatio n about the dialects which contribute to a mixture, and sufficient demograp hic information about the proportions of speakers of the different dialects , it is possible to make predictions about what the outcome of the mixture will be. We also argue that we have arrived at a probabilistic solution to the problem of randomness in the transmission of dialect features from one generation to another in such situations.