| Abstract: | Rigsby (1986) states translaryngeal harmony is observed in Gitksan. But the data I obtained from Eastern Gitksan consultants suggests that it is not just translaryngeal: nonhigh vowels in the stem can be copied onto the pronominal suffix across not only a glottal stop but also a uvular consonant. The discrepancy between these two observations can be seen as an ongoing change of the transguttural harmony in the present-day Gitksan. In an Optimality Theory (Prince & Smolensky (1993)) analysis, it will be shown that the system with laryngeal transparency and the system with guttural transparency are attained with the reranking of a HARMONY constraint, along a language-specific markedness constraint hierarchy. This approach limits variations among sets of transparent consonants among gutturals, and tells us not only 'how' but also 'why' harmony is being extended in the way it is. |
