ABSTRACT This project continues the documentation of Chiwere, a Siouan language in the late stages of obsolescence, a language that has fewer than 15 fluent speakers. Probably as a result of the disuse of Chiwere as an every-day form of communication for more than a generation, all present-day speakers of Chiwere, even fluent ones, exhibit some gaps in their knowledge of the language. These inadequacies appear to be patterned. This project addresses the problem of integration of divergent and partial knowledge toward recovering a complete grammatical system. Accommodation theory, a theory most usually applied to gain understanding of dialect variation and shift in synchronic settings, appears to offer a guide to integrating these various less-than-perfect competencies into something resembling a relatively complete picture of Chiwere as it was in its full vibrancy of life. The specific goals for this project are (1) to develop this method for recovery of the complete linguistic system from several incomplete models, thus enhancing the description of Chiwere; (2) to devise an explanatory model of language change in Chiwere that may be more generally applicable for the whole of the Great Plains languages; (3) to reconstruct the paths of language death, i.e., to make a case histroy of language death in Chiwere; (4) to test hypothese suggested by accommodation theory about the pattern of decline in Chiwere and its relevance for indicating important features of the language.
