Analysing rhythm in ritual discourse in yucatec maya using automatic speech alignment

Over the years, research in ethno-linguistics contributed to gather corpora in a wide range of languages, cultures and topics. In the present work, we are investigating ritual speech in Yucatec Maya. The ritual discourse tends to have a cyclic structure with repetitive patterns and various types of parallelisms between speech sections. Previous studies have revealed an intricate connexion between a speech’s structure and vocal productions, in particular through temporal aspects including rhythm, pauses and durations of different speech sections. To further investigate our findings by relying more strongly on the acoustic recordings, automatic speech recognition tools may become of great help, in particular to test various linguistic and ethnolinguistic hypotheses. Unfortunately, Yucatec Maya, with less than one million native speakers, is an under-resourced language with respect to digital resources. As a total, 24 minutes of ritual speech from three performances were manually transcribed by expert linguists in Yucatec and a basic pronunciation dictionary for Yucatec was created accordingly. The transcribed acoustic recordings were then automatically time-aligned on a phonetic and lexical basis. Automatic segmentations were used to measure tempo changes, durations of breath units as well as to examine their link with the structure of the ritual text.