Key points are not available for this paper at this time.
This paper describes the acoustic characteristics of medial /t,d/ in American English as a function of phonetic environment. The data consisted of some 3000 word tokens, each embedded in a carrier phrase, recorded on two separate occasions by six subjects, three male and three female. Quantitative results were obtained on the acoustic characteristics of each stop for all phonetic environments, as well as the difference between /t/ and /d/ for a given phonetic environment. The interaction between certain phonological rules such as flapping and glottalization and low-level phonetic recoding rules such as vowel nasalization and nasal deletion was investigated. Based on the statistics derived from our corpus, probabilities of occurrence were derived for all phonetic realizations. In addition, interspeaker variability was examined and a significant difference was found in the application of phonetic rules between male and female speakers.
Zue et al. (Mon,) studied this question.