The original distinction, in the opposing views of René and Ferdinand de Saussure, between views of word structure based on the combination of elementary, atomic signs (or ‘morphemes’) on the one hand and relations between complex words on the other, is reviewed. Early work in American linguistics associated with Boas and Sapir is noted, and the later emergence of clearly morpheme-based views in the Bloomfieldian tradition (especially as continued by Harris, Hockett, and others) is reviewed. This picture was essentially taken over unchanged in early generative grammar, although Chomsky (1965) provided (now forgotten) arguments in favor of an alternate non-morphemic view. The re-emergence of interest in morphology in later work has led to a situation in which the two views that can be identified originally in the work of the de Saussure brothers continue to characterize two conflicting scholarly positions.