This study is based on the general finding that the linguistic reality of a bilingual community is complex and that the two languages are not compartmentalized into any particular spheres of social life. It uses this finding to explore a theoretical position that treats facts regarding language functions and usage as if these, in and of themselves, could explain language shift, loss, or maintenance. By examining several defining criteria for diglossic and bilingual situations, it tests the theoretical power of the view that without diglossia, bilingualism is only transitional, and that with diglossia, the less politically powerful language or variety has a better chance of being maintained. In addition, problems of describing and analyzing language choice and vitality are raised as issues that need to be addressed, particularly in terms of data gathering techniques. Beyond empirical methods, problems of inquiry and analysis are examined, especially those based on abstraction divorced from socio-historical and economic conditions. (Author/AMH)