Abstract:
This paper presents a case study of what we will call ergative/nominative paradoxes, which we claim are found in the Tsimshianic language family, but which are possibly found in other languages as well. Such paradoxes are said to arise when both nominative and ergative morphology is simultaneously indexed or related to the same subject. Although these languages manifest this subject marking paradox in different ways, we conjecture that the paradox itself is the result of an ergative system decaying into an accusative one, and that one strategy for languages to relieve the pressure of such a paradox is to develop new paradigms of differential subject marking.
