Resumo:

This article has two objectives: (i) to present the methods employed in an analysis of the variable pretonic midvowel height in the speech of 32 migrants from the Northeastern states of Alagoas and Paraíba living in the Southeastern state of São Paulo/Brazil, in a dialect contact situation; and (ii) to call attention to two limitations of variationist studies, namely, the great amount of time spent in identifying, extracting, and coding data, and the subordination of research questions to a restricted set of methods.