ABSTRACT
In almost all fields of scientific enquiry the professional practitioners are divided into two camps. There are those who are temperamentally inclined to start off with formal ideal models and work down to the facts of the case, and there are those who prefer to start with the facts and work cautiously upwards to a generalized abstraction. One branch of anthropology in which we can observe this process at work is the study of kinship terminology. Anthropological studies of kinship terminology share one extraordinary peculiarity. This chapter draws attention to certain features of Firth's discussions of 'the language of kinship'. It uses an argument based on structural transformation to justify a mixture of levels of analysis, and this fact may turn out to be of some theoretical significance. The chapter investigates a part of the Kachin language of kinship, exploring whether 'the [Kachin] classificatory system of counting relationships' reveals itself as a convenient, flexible and commonsense piece of social mechanism.
