A coleta de dados em Rikbaktsa (Macro-Jê) e a pesquisa de campo falante-falante

This article describes the experience and results of data collection in Rikbaktsa (Macro-Jê) as part of the CNPq's Pró Humanidades project. The project used Dayal's (in)definiteness and genericity questionnaire (forthcoming) as a guiding thread. In the first stage, we translated the contexts and examples into Portuguese, adapted them to the local culture, developed corresponding sentences in the language, and tested hypotheses. This primary data collection counted with the consultant and one of the researchers who is a native speaker. The second stage was developing a list of sentences in the language that, paired with illustrative situations, allowed for positive and negative evaluations. Using these, we created a sentence-context game. The consultant then conducted field research with other speakers, functioning as a game. This third stage resulted in new positive and negative evaluations, as well as corrections from the speakers. This led to a fourth stage, in which this material was taken to the school. The data collection contributed to the creation of a database of over 200 sentences in the language, accompanied by evaluations, the raising of original semantic questions, field research involving speakers of the language, and, finally, a critical analysis of the limitations of the instrument used, the Dayal questionnaire. These activities proved very fruitful to better understand the semantics across languages, to awaken speakers' metacognitive awareness of their own language, and to develop data collection material that can be transformed into teaching material.