GRAPHIC TRANSCRIPTION: challenges and possibilities

desafios e possibilidades

Abstract

Performing graphic transcription of oral texts is a challenge for most language scholars, because each corpus may require a specific methodology, as well as computational resources that can assist in this process. This paper presents some computational tools to perform the conversion from oral to written modality and provides basic subsidies for researchers from different areas to know and select the best method for their studies. In this sense, we describe three programs that have helped the development of research with oral sources by our research group. Paula (2010), Ramilo and Freitas (2001) Labov (1983) and Du Bois (1991) present reflections for the development of a graphic transcription methodology, respecting the speech and its enunciator. The results indicate that, although the researcher makes use of these computational tools, the researchers' commitment to reviewing the transcription is still essential, because if a speech is enunciated overlapping, for example, the computational resource will hardly be able to help with the transcription in a reliable, trustworthy way.