Limites e contribuições do uso da EME para pesquisas de cunho qualitativo na aquisição do PB

The main objective of this paper is to answer the following question: how can we draw a parallel, during the analysis of the data, between children who are at different moments in the language acquisition process or who belong to different populations (children that are monolingual, bilingual, with language disorders, etc.)? In order to answer this question, we are discussing the advantages and limitations of adopting a measure like MLU, from English Mean Length of Utterance, or EME as it was named in Brazil, proposed by Roger Brown (1973). We intend to verify the effective contribution of these measures to the studies in the area and propose an adaptation in the studies that use a qualitative approach. Firstly, we recall the study of the criteria proposed by Brown (1973), Scliar-Cabral (1976), Parker & Brorson (2005), Parisse & Le Normand (2006), and Araújo (2007), and by scholars of Brazilian Portuguese morphology, like Câmara Jr (2009), Monteiro (2002), and Kehdi (1990), among others. Secondly, from a dialogic-discursive point of view (BAKHTIN, 2006; BAKHTIN/ VOLOSHINOV, 1992, 1981), we provide as an example the research on the plural in the language acquisition process (HILÁRIO, 2010, 2011b, 2012). The reflections made indicate that EME may be a useful measure for the researches on language acquisition. Nevertheless, adaptations should be done and its inherent problems should be taken into account, especially if other factors are considered – like the age, the number of types/occurrences, the maximum limit of items per utterance etc. – although it is not the only parameter of analysis. Considering that EME may vary according to the conditions of production and may not reflect the qualitative changes in the children’s discourse, other factors should be analyzed, like the context of the discourse, the participants involved etc.