Speech prosody is essential for conveying communicative intentions. Although neurophysiological data has shown that communicative functions conveyed through prosody are processed rapidly in the human brain, it is still unclear when and to what extent prosodic information is needed for the conscious speech act recognition as speech unfolds. Using a gating paradigm, we investigated the point at which listeners recognise the function of identical Italian sentences – whether they express a question or statement – based on vocal intonation. Comparing cross-spliced and natural sentences, we found that, rising or falling nuclear accentual movement on the sentence-final word seems to be the primary cue for recognition, with questions identified slightly later than statements. Furthermore, we discuss the limitations of splicing techniques in filtering out natural prosodic variations, the presence of a “statement bias” in perceiving incomplete sentences, along with a visual examination of interindividual responses. These findings offer valuable insights into the timing of conscious recognition of different communicative functions based on speech prosody.
