Constraints on Variation in the Production of English /r/
| Number | 1327 |
|---|---|
| Year | 2003 |
| Drawer | 25 |
| Entry Date | 01/24/2008 |
| Authors | Gick, B., Iskarous, K., Whalen, D.H. & Goldstein, L. |
| Contact | |
| Publication | In S. Palethorpe & M.Tabain (Eds.), Proceedings of the 6th International Seminar on Speech Production, pp. 73-78. |
| url | http://www.haskins.yale.edu/Reprints/HL1327.pdf |
| Abstract | English /r/ has been shown to vary across dialects and speakers, and even within speaker. One account for this variation argues that acoustic values are stable despite substantial articulatory variations, resulting in a “trading” relationship favoring acoustics over articulation (Guenther & al. 1999). This paper questions the evidence and need for this acoustic representation, arguing that acoustic variation follows articulation. First, correlations in kinematic point-tracking patterns previously raised in defence of acoustic-articulatory trading relations are found to occur not just in /r/, but across other segments lacking /r/’s proposed F3 acoustic “target”, suggesting more general physical constraints on tongue kinematics. Second, using a combination of ultrasound and video imaging, all three component gestures (lips, tongue anterior, tongue root) for English /r/ are observed to be consistent with a model of articulatory targets specifying only constriction locations, suggesting that any stability in the acoustic signal corresponds with these stable articulatory parameters. It is further suggested that articulatory differences previously observed in /r/ variants are overemphasized by point-tracking measurement devices, which are unable to measure actual constriction locations as directly as imaging techniques. |
| Notes |
Search Publications