How to Transcribe Phonetically

This workflow describes the successive methodological decisions and operations required to produce a phonetic transcription of oral/spoken data for research and documentation purposes in the social sciences and humanities (linguistics, dialectology, sociolinguistics, psycholinguistics, language documentation, conversation analysis, etc.). It is designed for researchers, students, and archivists who need to convert an audio recording into a phonetically annotated resource, whether for a single case study or for a larger, shareable, and reusable corpus. The workflow was elaborated on the basis of practical teaching and research notes on phonetic transcription (auditory and acoustic-supported transcription, use of Praat and ELAN, IPA/SAMPA conventions, inter-annotator agreement, and data publication), and integrates reference information on the International Phonetic Alphabet (IPA) and its alternative ASCII-based encoding, X-SAMPA. The workflow can be adapted to individual projects or to team-based research infrastructures, and it explicitly foresees the possibility of publishing the resulting transcriptions, together with the source audio, as a citable open dataset (e.g. via Zenodo or a comparable repository).