Prosody is a central axis of the research, connecting different scales of speech analysis, from acoustic description to discourse organization and language processing, in dialogue with corpora, experimental methods and speech technologies.
Prosody has been a long-standing research axis, with work on production and perception, discourse organization, intonation, speech rate and prosodic boundaries. Current work revisits these questions in narratives and in comparisons between natural and synthesized speech.
Since the systematic incorporation of online methods, my research has connected prosody and processing through eye tracking, behavioral measures and EEG. The 2024–2028 research fellowship project investigates discourse-structure processing in multimodal contexts.
Between 2024 and 2026, the evaluation of synthesized speech became a central line of work. These studies combine naturalness, prosody, acoustic metrics, perceptual evaluation, EyetrackingMOS and the preparation of neurophysiological protocols to investigate what TTS systems preserve — or lose — from human speech.
Work with speech data includes NURC Digital, digitization and annotation protocols, corpus curation and language-documentation projects. Data preparation and traceability are treated as part of the scientific method, rather than as merely technical steps.
Reproducibility, data management, transparency, ethics and open science cut across my research, student training and editorial work. These topics also form part of the graduate course Metodologias de Investigação em Linguística, offered in 2026.2.