Govor, Vol. 40 No. 2, 2023.
Original scientific paper
https://doi.org/10.22210/govor.2023.40.11
Speech disfluencies in the framework of forensic phonetics
Ana Žagmešter
*
Iva Bašić
orcid.org/0000-0002-2171-9300
; Faculty of Humanities and Social Sciences, University of Zagreb
*
* Corresponding author.
Abstract
Speech is defined as optimal human communication (Škarić, 1991). The optimality of speaking and listening is questioned when the fluency of speech is disturbed. Until recently, disfluencies in speech have been considered unaesthetic, sometimes inappropriate, and according to some authors even negative aspect of speech production. However, recent studies approach disfluencies as a part of a complex speech production process and therefore assess them as functional constituent of speech production. From the speaker’s point of view, they contribute to the eased speech production and form the listener’s point of view they facilitate speech perception. This paper presents disfluency in the framework of forensic phonetics, based on the student speech corpora. The distribution of different categories of disfluencies was analyzed from the speech sample collected during four tasks; the so-called continuous speech tasks – description of one’s working day, reading a sample text, description of a picture, and speaking as fluently as possible on a topic of one’s choice for 30 seconds. Based on the results from the analyzed corpus, a new disfluency classification was made, dividing disfluencies into two larger groups: hesitation and self-corrected errors. The calculated data of the temporal parameters of speech – speech tempo and articulation tempo – was correlated with the frequency of disfluency. The results showed that speakers are disfluent from 1% to 15% of the total speaking time, depending on the different speaking style. The highest number of disfluencies (141) was found in the task that required the highest speaking fluency, and almost half of the determined disfluencies (198 out of 397) refers to the prolonged [ə]. The findings of this study suggest that the type of speaking style and according to gender influence on the distribution of disfluencies. Also, individual characteristics of speaking disfluencies were observed as idiosyncratic markers in speech. The recorded data were analyzed following the procedures used for speakers’ recognition in the field of forensic phonetics, which sought to highlight the importance of phoneticians and their contribution to forensics.
Keywords
disfluencies; forensic phonetics; speech tempo
Hrčak ID:
314515
URI
Publication date:
22.2.2024.
Visits: 870 *