Table 4.
Pairwise difference in mean LA estimated acoustically from 39 speakers of American English vs. measured from XRMB for same speakers and tokens. In each cell, the left member is the difference measured from XRMB data (XRMB) and the right member is from formant+amplitude reconstruction (F+A). Bold numbers indicate a significant difference (p < .05). A black box around a pair indicates discrepancy between the XRMB data and the acoustic reconstruction.
| /i/ |
/I/ |
/e/ |
/ε/ |
/æ/ |
/u/ |
/℧/ |
/o/ |
/ɔ/ |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| XRMB | F+A | XRMB | F+A | XRMB | F+A | XRMB | F+A | XRMB | F+A | XRMB | F+A | XRMB | F+A | XRMB | F+A | XRMB | F+A | |
| /I/ | −.25 | −.13 | ||||||||||||||||
| /e/ | −.99 | −.49 | −.75 | −.37 | ||||||||||||||
| /ε/ | −.99 | −.62 | −.74 | −.49 | .01 | −.13 | ||||||||||||
| /æ/ | −1.66 | −.93 | −1.41 | −.80 | −.66 | −.43 | −.67 | −.31 | ||||||||||
| /u/ | 1.04 | 1.67 | 1.28 | 1.79 | 2.03 | 2.16 | 2.02 | 2.29 | 2.69 | 2.59 | ||||||||
| /℧/ | .55 | 1.03 | .79 | 1.15 | 1.54 | 1.52 | 1.53 | 1.65 | 2.20 | 1.95 | −.49 | −.64 | ||||||
| /o/ | .32 | 1.18 | .56 | 1.30 | 1.31 | 1.67 | 1.30 | 1.80 | 1.97 | 2.10 | −.72 | −.49 | −.23 | .15 | ||||
| /ɔ/ | −.98 | −.09 | −.73 | .04 | .02 | .41 | .01 | .54 | .68 | .84 | −2.01 | −1.75 | 1.52 | −1.11 | −1.29 | −1.26 | ||
| /a/ | −1.71 | −.86 | −1.46 | −.73 | −.72 | −.37 | −.73 | −.24 | −.05 | .07 | −2.75 | − −2.53 | 2.26 | −1.89 | −2.03 | −2.04 | −0.74 | −0.77 |