ABSTRACT
This paper discusses the treatment of pure agraphia in classical models of written language processing. At the transition from the 19th to the 20th century, the syndrome of isolated agraphia was not accepted by all aphasiological authors. A crucial issue in the debate was the unit of analysis in graphemic processing, the letter or the word. Closely related to this were questions of localisation, in particular whether specialised cortical centres should be postulated for reading and writing or whether these functions were subserved by more general optic and motor areas. Following from a decision on this question, some authors adopted a unilateral left-hemispheric representation of graphemic units whereas others assumed bilateral processing.
