Some acoustic properties of boundary signals
Karen M. Forrest, N. Umeda · The Journal of the Acoustical Society of America · 1975
In a previous study, listeners' perception was used as an indicator of boundary signals in continuous speech [J. Acoust. Soc. Am. 56, Suppl. S33(A) (1974)]. While each subject had his own definition of the term “boundary,” there was high agreement among listeners about boundary locations. On the basis of those results, we investigated some acoustic parameters (relating to speaker's production) associated with boundary placement. The change in fundamental frequency was measured, in the word preceding the boundary, from the last stressed vowel to the end of the word. For comparison, measurements were also made in content words that were not followed by a boundary. Preliminary results show a greater fundamental frequency change (downward, upward, or both) at points where boundaries were perceived. Most of the perceived boundaries have a typical fall-rise pattern, i.e., from a maximum, occurring at the stressed vowel, the pitch decreases to a minimum level and then begins to rise again. However, at a small number of boundaries very little change in pitch was observed. Measurements were also made to determine the duration of the fundamental frequency excursion. It was found that the excursion was two to three times longer for a word preceding a boundary than when no boundary was present. The data suggest specific cues, both within and across words, that are relevant to boundary perception.