ABSTRACT
For a linguist or anthropologist or theorist who grasps the methodological starting point of modern linguistics, the sine qua non of its development, Sapir's (1925) "Sound patterns in language" is the explicit key. There he set forth the "thought experiment" (for which there are in fact many empirical examples) of two languages that agree in the presence of certain sounds. When the function of the sounds in the system of the language is investigated, however, the languages differ. Again, he considered the case of two languages that differ as to the presence of certain sounds, but which, when function is investigated, prove to be alike.
