Prosodic prominence of modal verbs in narratives

The study deals with accentuation (presence or absence of realized lexical stress) of modal verbs in continuous spoken texts. The material was taken from audiobooks where Czech professional actors (8 male + 8 female) read out various narratives produced by renowned authors. A corpus of over 17,000 words was used. The recordings were annotated with special attention to presence or absence of materialized stress (i.e., accent). The chief purpose of the study was to provide descriptive data on modal verbs in connected speech, namely in the genre of narrative monologue. The results showed that modal verbs were deaccented in more than a quarter of their occurrences, but that monosyllabic and polysyllabic verbs behaved differently from each other. The infinitives associated with modal were also inspected. They were much less often unaccented and, importantly, the influence of monosyllabicity followed a different pattern than in modals. Additionally, information on mutual position of modals and associated infinitives is provided and an observation of negative forms of modals is made. The data can be further used in follow-up research to find out how structural and communicative requirements interact to produce the actual prosodic forms, or how modals in narratives differ from those in other communicative genres.