It is known in literature that North American speakers pronounce the medial /nt/ cluster either as [nt] or [n], which is considered alternative realization. Thus, in this study, we proposed several optimality theoretic constraints to account for the a ...
It is known in literature that North American speakers pronounce the medial /nt/ cluster either as [nt] or [n], which is considered alternative realization. Thus, in this study, we proposed several optimality theoretic constraints to account for the alternativel realization of such cluters in Englsih. Even though we could provide an analysis by which the optional pronunciaiton could be explained by re-ranking relevant constraints, it is not clear what optional pronunciation is, which means we wanted to know how native speakers actually pronounce the medial /nt/ clusters in English. So we selected 40 English words containing the /nt/ in their medial syllable boundary. We confined the research to the heterosyllabic /nt/ clusters in English because the other possible heterosyllabic NC clusters do not have alternative pronunciations like /nt/. We asked the 40 native speakers of North American English, who are teaching Englilsh in a city in Korea consisting 30 Americans and 10 Canadians, to read the 40 words as comfortably as possible. We recorded 1600 tokens in the lap-top computer since we were not conducting a serious acoustic phonetic analysis. Before we asked that native speakers of English to pronounce the words, we asked them to check wheter each given word is "Familiar" or "Unfamiliar." We evaluated the 1600 tokens and classified them into two groups: familiar and unfamilar words. The subjects of the research marked 1310 words as familar while 290 as unfamilar. From the total of 1600 tokens, they dropped the /t/ in 173 (10.8%) words out of 1600. However, the native speakers dropped the /t/ in 160 familar words out of 173 t-dropping words, which accounts for 92.5%. On the other hand, they deleted the /t/ in 13 unfamiliar words out of 173 t-dropping words. This shows that the native speakers have a tendency of dropping a segment if they recognize a word as familar while they maintain or pronounce all the segments if they recognize a word as unfamilar.
When we consider the relationship between word familiarity and nationallity of the subject, we could not find any significant difference in the pronounciation of words. When we consider the gender factor and word familiarity, we could see some interesting differneces between male(25 people) and female(15people). Even though female speakers makred words as familiar by 2% more than that of male speakers, male speakers dropped the /t/ about 5% more than female speakers in the familar word group. In the unfamiliar word group, however, female speakers dropped the /t/ about 3.5% more than male speakers. When we consider the age factor in pronunciation, we also could find an interesting result. We divided the subjects into older group (aged over 30: 13 people) and younger group (aged below 30: 25 people) in which two of the subjects did not tell us their age because of the privacy. While the older group marked more words as familar than the younger group, the younger group dropped the /t/ about two times that of the older group. When we considered the subjects' teaching experiences, this factor also revealed some interesting figures. We divided the subjects into two groups. People in one group have more than 3 years of teaching experience while the other fewer than 3 years. Interestingly, the subjects with more teaching experience marked 10 % more words as familiar than that of less experienced teachers. However, the subjects with less teaching experience dropped the /t/ about more than two times that of the more experienced teachers. The quantitative evidence we provided in the study is meaningful in that even though it reflects various facts of the /nt/ pronunciation by native speakers, but one thing in common is that they drop the /t/ if a word is familiar: otherwise, they rarey do it. So the quantitative evidence supports the assumptions we made before conducting the pronunciation test.