Key points are not available for this paper at this time.
This paper proposes an analysis of nasal harmony within the framework of Optimal Domains Theory (ODT), and demonstrates that transparency and opacity derive from principled constraints that limit the realization of Nasal on potential anchors. The analysis differs fundamentally from the autosegmental analysis in two respects: it does not treat harmony as feature spreading, and it does not use feature specification or feature geometry to distinguish transparent and opaque segments from segments that undergo harmony. The ODT approach can account for the presence of inherently nasal segments that are transparent to harmony, as demonstrated in our analysis of Terena, unlike the autosegmental analysis, which incorrectly predicts that nasal segments will always trigger or be opaque. We also discuss why obstruents are typically opaque to nasal harmony, in light of the notion of contrast and the need to preserve contrast in harmony systems. The ODT analysis is based on the notion of the feature domain and the articulation of constraints which govern both the size and the composition of the feature domain, in this case for the feature Nasal.
Cole et al. (Sat,) studied this question.