2pSC25. An MRI and EPG examination of /r/ and /l/.

Session: Tuesday Afternoon, December 3


Author: Maureen Stone
Location: Div. of Otolaryngoloy, Univ. of Maryland Med. School, 16 S. Eutaw St., Rm. 500, Baltimore, MD
Author: Darryl Ong
Location: Johns Hopkins Univ., Baltimore, MD
Author: Andrew Lundberg
Location: Johns Hopkins Univ., Baltimore, MD


This study characterized airway shapes for retroflex /r/ and /l/ reconstructed from multiplanar MR images, and compared them to acoustical and EPG data. Observing the production gestures in light of the sounds' acoustic similarities were of interest. The results indicated that the /l/ and /r/ airways had remarkably similar area functions with small pharyngeal volumes and large oral volumes. At the tongue tip there was much closer tongue--palate contact for the /r/ than the /l/ and an anterior cavity for /r/. The similarities of general shape may account for the similar F1 and F2 frequencies. The extra anterior cavity and elevated top for /r/ may affect the F3 difference between the sounds. EPG/acoustic data were collected for bunched and retroflex /r/. The EPG data confirmed the /l/ and /r/ differences and distinguished between the two /r/'s. Palatal contact for the bunched /r/ was lateral and began at the highest point of the tongue and posteriorly (last four rows). For the retroflex /r/ the contact was posterior to the bunched /r/ (last two rows) and was made with the tongue margins posterior to the freestanding tip. Acoustically, bunched /r/ had a higher F3. [Work supported by NIH.]

ASA 132nd meeting - Hawaii, December 1996