Linguistic segmentation with systemic functional linguistics (Gravelin et al., 2023)
Purpose: Extemporaneous speech tasks provide an ecologically valid sample to examine speech acoustics, but differing methodologies exist in the literature for segmentation. Therefore, the purpose of this study was to examine the utility and reliability of a segmentation approach for extemporaneous speech specified by systemic functional linguistics (SFL) and its potential research and clinical applications.
Method: Ten speakers without communication disorders served as participants in this study, and they responded to self-selected extemporaneous speaking prompts. Two expert analysts and one clinician analyst utilized a segmentation procedure specified by SFL to segment the extemporaneous speech samples into clauses and clause complexes. Intra- and interrater reliability were calculated for each analyst and pair of analysts. Acoustic measures of duration, speech rate, and intercomplex pause durations were calculated for each clause complex.
Results: Analyses for both intra- and interrater reliability revealed high percent agreement that was significantly greater than chance for expert and clinician analysts and between each pair of analysts (p < .001). Acoustic analyses revealed expected variation in number and duration of spoken syllables of clause complexes between and within speakers.
Conclusions: The segmentation approach for extemporaneous speech specified by SFL is a reliable method for trained analysts that is informed by lexico-grammar and allows for acoustic measurement of speech production. It is also a reliable method for clinician analysts for speakers without communication disorders, and future work will investigate its utility for speakers with motor speech disorders.
Supplemental Material S1. Step-by-step segmentation sample guide.
Gravelin, A. C., Archer, B., Oddo, M., & Whitfield, J. A. (2023). Reliability of a linguistic segmentation procedure specified by systemic functional linguistics to examine extemporaneous speech. Journal of Speech, Language, and Hearing Research, 66(4), 1280–1290. https://doi.org/10.1044/2023_JSLHR-22-00554
