14th International Congress of Phonetic Sciences (ICPhS-14)
San Francisco, CA, USA
Traditionally, the Korean preliquid nasalization (e.g. kl → ŋn) has been analyzed by a feeding order of L-nasalization (kl → kn) and prenasal nasalization (kn → ŋn). It is thus predicted that if L-nasalization is blocked, prenasal nasalization will not occur. To test this prediction, we performed experiments in which audio recordings of tokens with an underlying stop-liquid sequence (e.g. /kl/) were analyzed by the author and two other trained phoneticians relying on the spectrographic display and perception. Ten native Korean speakers were requested to pronounce the spelling pronunciation of a post-stop liquid, while reading ten real words (e.g. /tɛhak-lo/). Since the Korean spelling system is phonemic, spellings directly reflect the underlying segments. The results are not compatible with the prediction of the traditional analysis. Finally, we provide an alternative account relying on a historical reanalysis of the triggering segment of prenasal nasalization.
Bibliographic reference. Jun, Jongho (1999): "Nasalization not triggered by nasal", In ICPhS-14, 527-530.