Ladd, D. R., Faulkner, D., Faulkner, H., & Schepman, A. (1999). Constant “segmental anchoring” of F0 movements under changes in speech rate. Journal of the Acoustical Society of America, 106(3), 1543–1554.
Presentation: Sally
Summary: Sarah
The ‘segmental anchoring’ theory of F0 movement was proposed by Arvaniti et al. (1998). Specifically, the notion arose from the finding that the beginning and end points of rising pitch accents are aligned with specific segmental positions. While Arvaniti et al. (1998) manipulated the internal structure of accented syllables, this study aimed to further testify this hypothesis by investigating the effect of speech rate on F0 realization of prenuclear rising accents. The authors predicted that if segmental anchoring is valid, both F0 rise time and slope should be dependent on speech rate, whereas maximum and minimum F0 levels and their corresponding anchoring segmental landmarks should not. Two experiments were included in this study. The first experiment was mainly conducted to see whether speech rate manipulation was feasible for speakers. Preliminary duration measurements showed that speakers were indeed able to produce test sentences with three rates (fast, normal, slow), and the overall durations vary with speech rates as well. In the second experiment, more speakers were recruited. Results showed that in general, at a slower speech rate, rise time increased and F0 slope became less steep. In addition, for most of their speakers, the alignment of the F0 maximum and minimum with segmental landmarks did not significantly differ at three speech rates. These findings therefore supported the segmental anchoring hypothesis. Taken together, this study also implicated that the alignment between tonal targets and segmental landmarks, instead of F0 rise time and slope, should be characterized as the invariant elements of tonal realizations.
