Methodological Challenges to Corpus Linguistics in Language Acquisition in Brazilian Portuguese

I will discuss the methodological challenges for the organization of three corpora belonging to a child aged 20 months and 21 days, aged 22 months and 20 days and aged 26 months and 8 days, when acquiring Brazilian Portuguese. The methodology was developed by R. Brown (1973), when he organized Adam, Eve and Sarah"s classic corpora. I used the same measures, such as Mean Length Utterance, MLU and Maximum Upper Bound Limit calculation, whose rules I describe. The results are compared with the 19 reports listed by Brown (1973: 66) regarding different languages acquisition, including non-Indo-European ones. The criteria applied for selecting utterances, which will make up the corpora are also explained, for building up the emerging grammars. The child"s total utterances number, in phonetic transcription, was 5530; that of adults, canonically transcribed, when addressed to the child, was 5664. All recordings, as well as phonetic transcripts, glosses, morphological analysis (mor line) and lexicons are available on the CHILDES Platform, the world's largest database on language acquisition.