The time course of first formant movement in normal and loud speech

Abstract

Louder speech tends to be slower than typical speech. For some vowels (viz., low and lax ones), louder speech may also be associated with higher first formant frequencies (F1s) at vowel midpoint. Possibly, increased duration allows speakers more time to reach a lower jaw position, thus a higher value of F1, in loud speech. In earlier work, we found that some loudness differences appeared at 25% and 75% of vowel durations, i.e., they were seen for much of the vowel. The current study presents data from 11 German-speaking women producing naturally elicited typical and loud speech. Target words included high, low, tense, and lax vowels with surrounding consonantal place controlled. We will present full vowel trajectories, i.e., from preceding to following consonants, to assess the time course of F1 differences across vowel types and loudness conditions. F1 will also be considered within normalized time intervals to control for durational differences between conditions. We expect that F1 formant trajectories will show minimal loudness differences throughout the vowel for high tense vowels, but time-varying differences will be more apparent for low and lax vowels, and appear rather early in the vowel.