The transition from /a/ to /i/. The superimposed dotted arrow shows how the constriction would move if its constriction location moved continuously.
(a) An example trajectory of the tongue body circle from /a/ to /i/. Superimposed on the figure is a b-spline of the path of the center of the tongue body circle. The gray scale denotes time, with the first tongue trajectory denoting /a/ in black and the last one denoting /i/ in light gray. (b) 13 trajectories of the center of the tongue body. Each V-V transition lasted 350 ms. The most concave is in black and the most convex is in lightest gray. The particular trajectory in (a) is for Step 6.
Location of point of minimum constriction in the vocal tract as a function of time (in percent). Upper panel: Steps 1–5, all of which have identical placement of CL as a function of time. Middle panel: Step 6. Lower panel: Step 7 (darkest gray) to Step 11 (lightest gray).
First panel: Curvature of tongue trajectory as a function of step. Second panel: CL discreteness as a function of step. Third panel: Pivoting index as a function of step. Fourth panel: Degree of synchronization of palatal constriction formation and pharyngeal constriction release.
(a) Trajectories in F1-F2 space (frequency in barks). Step 1 is in black and Step 11 is in lightest gray. (b) Curvature of F1-F2 trajectories as a function of step. (c) F2 maximal velocity as a function of step. (d) Time (in percent) at which F2 maximal velocity occurs.
Results of perception Experiment 1 as a function of step: Naturalness rating (black), three-step discrimination (gray), and 1-Step discrimination (lightest gray).
Results of perception Experiment 2 as a function of step: Naturalness rating (black) and 5-Step discrimination (gray).
Results of perception Experiment 3 as a function of step: Naturalness rating (black), three-step discrimination (gray), and 1-Step discrimination (lightest gray).
Left: Probability histograms of the chance of occurrence of each of the measured acoustic quantities in the observed 34 tokens of /ai/. Right: Posterior probability of each of the steps on the scale of synthetic stimuli, given the likelihood of observing that measured quantity in the observed data.
Relation between articulatory, acoustic, and perceptual variables. First ten rows show for each two pairs of variables. Pairs with explained variance higher than 85% are in bold large type. Last two rows show the difference (in -scores) for each of the articulatory and acoustic variables between Steps 6 and 1 (row 11) and 11 and 6 (row 12), for comparison with discrimination data. Variables that show insignificant difference for the 1–6 pair and a highly significant difference for the 6–11 pair are in bold large type. Abbreviations: of tongue trajectory, discreteness, index, sync, curvature, Max, Max Time, 1 naturalness rating, and 3 naturalness rating.
Article metrics loading...
Full text loading...