V. J. Boucher
Dept. of French, Univ. of Windsor, 401 Sunset Ave., Windsor, ON N9B 3P4, Canada
The object was to isolate, across changes in speaking rate, categorical patterns in the relative timing of aspects that can relate to voicing distinctions. In a first series of observations, French sujects were asked to produce sequences [tapetape...] and [dapedape...] at different rates. Categorical ratios were observed in the timing of releases and closures when these were measured with reference to a CVC or CV interval but not with reference to VC. Only the ratios of release times were categorical for all subjects and CVC was more successful in revealing categorical ratios. Second, to determine if CVC provides a reference frame in perception, a test was constructed from multisyllabic sequences containing a voicing distinction [t]/[d]: [apapatapapa]/[apapadapapa]. Thirty milliseconds was spliced away from various portions of the sequence containing [d] and the results of an ABX test showed that perception shifted markedly from [d] to [t] when the 30-ms splice was applied to both the C and V surrounding [d] but not when the splicing occurred one syllable away or when it was applied to either the C or V surrounding [d]. This suggests that perception and production of categorical patterns in timing aspects operates by reference to CVC.