Table 1.
Summary of the 24 test items used.
Duration (in ms) and fundamental frequency (F0, in Hertz) at three normalised time points as well as the overall average of the tone space. Each speaker produced a minimal pair. Note that while duration differs across minimal pairs, the duration remains constant within each minimal pair.
Fig 1.
Pitch contour of the 8-step Male /na/ training continuum.
Bold black lines represent the endpoints of the continuum (Token 1 and Token 8) while dashed grey lines represent the intermediate tokens (Token 2 to Token 7). Note that the pitch contours shown here represent the tone space of the vowel, in which the first 15% and the last 15% of the vowel were excluded to remove possible effects of coarticulation from the preceding consonant and creakiness, respectively.
Fig 2.
Unimodal and Bimodal distribution.
For the Bimodal distribution, Token 2 and Token 7 were heard four times more often than the endpoints (Token 1 and Token 8), while for the Unimodal distribution, Token 4 and Token 5 were heard the most often relative to the endpoints.
Fig 3.
Comparison of Pretest and Posttest scores by Distribution Condition in Experiment 1A.
Test items are coded based on what the participant was trained. For example, if a participant was trained on Male /na/ training continuum, then Trained represents Male /na/ test items and Generalised Syllable (Gen. Syllable) represents Male /kha/ test items. In general, Posttest scores were higher than Pretest scores. Error bars represent 95% confidence intervals.
Fig 4.
Comparison of Pretest and Posttest scores by Distribution Condition in Experiment 1B.
Since only Male /na/ training continuum was used in this experiment, Trained, Gen. Syllable, Gen. Gender and Gen. Syllable & Gender test items represent scores on Male /na/, Male /kha/, Female /na/ and Female /kha/ test items, respectively. In general, Posttest scores were higher than Pretest scores. Error bars represent 95% confidence intervals.
Fig 5.
Comparison of Pretest and Posttest scores by Distribution Condition in Experiment 2.
Just as in Experiment 1B, Trained, Gen. Syllable, Gen. Gender and Gen. Syllable & Gender test items represent scores on Male /na/, Male /kha/, Female /na/ and Female /kha/ tets items, respectively. In general, Posttest scores were higher than Pretest scores. The Bimodal group outperformed the Unimodal group on Novel Syllable test aspect at Posttest relative to Pretest. Note that the Bimodal group’s Pretest score on Female /na/ (Gen. Gender) was relatively high compared to that of the Unimodal group. Error bars represent 95% confidence intervals.