["TESOL Quarterly, Volume 60, Issue 1, Page 406-416, March 2026. ", "\nAbstract\nIn qualitative research, transcription is a significant stage of the data collection process often employed as a method of language data analysis. In this account, however, I underscore the interpretive dimension of transcription in the context of classroom practitioner research, and the decision‐making entailed in this approach. To illustrate this, I present a pragmatic model of transcription synthesized from several theoretical perspectives. This model is geared to minimally guided group discussions focused on developing L2 learners' proficiency in collaborative argumentation and dispositions to criticality. Given these aims, I reflect on the challenges, in transcribing such discourse, of balancing the accuracy of speaker expression and intent on the one hand with readability and reader inference on the other, especially in view of the potential linguistic and fluency limitations faced by L2 participants. I also discuss the ethical and interpretive implications of transcription, considering the putative tensions between the researcher's role in shaping the final text and the importance of representing participants' voices authentically. The significance of this model lies in the deceptive simplicity of its format: a sequence of dialogic exchanges typically seen in narrative fiction. Constructing transcribed text in this way maintains the semantic integrity of the various ideas expressed in discussion, which, I contend, contributes to a more accurate inference of participants' ideational intent. Overall, this account aligns with the notion of the transcript as a constructed text mediated by the researcher's theoretical stance and practical features of the learning environment.\n"]