Corner vowels in males and females ages 4 to 20 years: Fundamental and F1-F4 formant frequencies

Abstract

The purpose of this study was to determine the developmental trajectory of the four corner vowels' fundamental frequency (f_o) and the first four formant frequencies (F1-F4), and to assess when speaker-sex differences emerge. Five words per vowel, two of which were produced twice, were analyzed for f_o and estimates of the first four formants frequencies from 190 (97 female, 93 male) typically developing speakers ages 4-20 years old. Findings revealed developmental trajectories with decreasing values of f_o and formant frequencies. Sex differences in f_o emerged at age 7. The decrease of f_o was larger in males than females with a marked drop during puberty. Sex differences in formant frequencies appeared at the earliest age under study and varied with vowel and formant. Generally, the higher formants (F3-F4) were sensitive to sex differences. Inter- and intra-speaker variability declined with age but had somewhat different patterns, likely reflective of maturing motor control that interacts with the changing anatomy. This study reports a source of developmental normative data on f_o and the first four formants in both sexes. The different developmental patterns in the first four formants and vowel-formant interactions in sex differences likely point to anatomic factors, although speech-learning phenomena cannot be discounted.

FIG. 1.

(Color online) (a)–(d) Display of the vowel mean data of each speaker for f_o and F1–F4: /i/ (a), /u/ (b), /æ/ (c), and /ɑ/ (d). Female speakers are shown in the left panel, and male speakers are shown in the right panel. For each frequency, the variance function regression (VFR) with fifth degree polynomial fits is displayed using thick dashed and thick solid line lines for females and males, respectively, with ±1 standard deviation for each denoted by thin dashed and dotted lines for female and male speakers, respectively. Vertical lines reflect the four pubertal-stage cohorts (years;months) as described in Sec. II (Methods): pre-pubertal (4;0–7;11); peri-pubertal (8;0–10;2); pubertal (10;3–14;5); and post-pubertal (14;6–20;0).

FIG. 1.

(Color online) (a)–(d) Display of the vowel mean data of each speaker for f_o and F1–F4: /i/ (a), /u/ (b), /æ/ (c), and /ɑ/ (d). Female speakers are shown in the left panel, and male speakers are shown in the right panel. For each frequency, the variance function regression (VFR) with fifth degree polynomial fits is displayed using thick dashed and thick solid line lines for females and males, respectively, with ±1 standard deviation for each denoted by thin dashed and dotted lines for female and male speakers, respectively. Vertical lines reflect the four pubertal-stage cohorts (years;months) as described in Sec. II (Methods): pre-pubertal (4;0–7;11); peri-pubertal (8;0–10;2); pubertal (10;3–14;5); and post-pubertal (14;6–20;0).

FIG. 1.

(Color online) (a)–(d) Display of the vowel mean data of each speaker for f_o and F1–F4: /i/ (a), /u/ (b), /æ/ (c), and /ɑ/ (d). Female speakers are shown in the left panel, and male speakers are shown in the right panel. For each frequency, the variance function regression (VFR) with fifth degree polynomial fits is displayed using thick dashed and thick solid line lines for females and males, respectively, with ±1 standard deviation for each denoted by thin dashed and dotted lines for female and male speakers, respectively. Vertical lines reflect the four pubertal-stage cohorts (years;months) as described in Sec. II (Methods): pre-pubertal (4;0–7;11); peri-pubertal (8;0–10;2); pubertal (10;3–14;5); and post-pubertal (14;6–20;0).

FIG. 2.

(Color online) (a)–(d) Estimated VFR mean frequency and variance changes for vowels: /i/ (a), /u/ (b), /æ/ (c), and /ɑ/ (d). The left panel shows estimated mean (±1 standard deviation) for male (thick solid and thin dotted lines) and female (thick dashed line and thin dashed lines) speakers by age. The right panel shows the estimated VFR log residual variance for male and female speakers by age. Note that when there are no significant sex differences in variance, the model predicted variance plots are identical. See the caption for Fig. 1 for a description of the vertical lines.

FIG. 2.

(Color online) (a)–(d) Estimated VFR mean frequency and variance changes for vowels: /i/ (a), /u/ (b), /æ/ (c), and /ɑ/ (d). The left panel shows estimated mean (±1 standard deviation) for male (thick solid and thin dotted lines) and female (thick dashed line and thin dashed lines) speakers by age. The right panel shows the estimated VFR log residual variance for male and female speakers by age. Note that when there are no significant sex differences in variance, the model predicted variance plots are identical. See the caption for Fig. 1 for a description of the vertical lines.

FIG. 2.

(Color online) (a)–(d) Estimated VFR mean frequency and variance changes for vowels: /i/ (a), /u/ (b), /æ/ (c), and /ɑ/ (d). The left panel shows estimated mean (±1 standard deviation) for male (thick solid and thin dotted lines) and female (thick dashed line and thin dashed lines) speakers by age. The right panel shows the estimated VFR log residual variance for male and female speakers by age. Note that when there are no significant sex differences in variance, the model predicted variance plots are identical. See the caption for Fig. 1 for a description of the vertical lines.

FIG. 3.

(Color online) Graphic and numeric display of VFR-based mean frequency (Hz) measurements of ages 4–20 for female and male speakers, and corresponding differences and statistical significance p-values by frequency and corner vowels.

FIG. 4.

(Color online) Graphic and numeric display of VFR-based age-centered variance (Hz²) of ages 4–20 for female and male speakers with corresponding differences and statistical significance p-values by frequency and corner vowel.

FIG. 5.

(Color online) Inter-speaker variability. Significant changes in variance of f_o, F1–F4 across the four age-cohorts for each of the corner vowels are displayed for male and female speakers with filled and open horizontal triangular bars, respectively. The base of the triangular bar represents the maximal variance at the pubertal-stage age-cohort. The pointed apex of the triangular bar reflects the direction of change in variance. Grayed-out plots indicate that the F-test was not carried out since the model predicted variance remained stable as a function of age (see Fig. 4).

FIG. 6.

(Color online) Intra-speaker variability. Significant differences in f_o, F1–F4 across the four age-cohorts for each of the corner vowels in males (shaded triangles) and female (open triangles) speakers. The base of the triangular bar represents the maximal difference at the pubertal-stage age-cohort, and the pointed apex of the triangular bar reflects the direction of change in difference.