In the latter half of the nineteenth century, the localizationist doctrines became closely associated with the memory trace paradigm. The analysis of the texts dealing with the localization and the nature of 'the loss of articulated speech' (motor aphasia) by Bouillaud, Lordat, Dax, Broca, Trousseau, Baillarger, Charcot and Wernicke shows how the biological paradigm of localization presented by Gall and based on the notion of organ-function correspondence was transformed into a model based on localizable memory traces. This change resulted in the theoretical unification of the mechanisms of motor and non-motor forms of aphasia. These forms, which the earlier authors tended to separate in their analyses of the underlying mechanisms, were now regarded as involving similar mechanisms related to the loss of mnestic images. The crucial step in this development was taken by Broca who presented the hypothesis that the faculty of coordination of speech movements, which according to his predecessors was the faculty lost in motor aphasia, was actually an intellectual faculty and a specific form of memory, and motor aphasia consequently a selective kind of amnesia. Theorists like Charcot and Wernicke generalized this idea into a comprehensive theory of the nature of localization based on the notion of memory traces. Thus, the localization of function was reduced to the localization of representations. Instead of biological paradigms, this model of localization is rooted in the epistemological tradition of psychology represented by Locke and Condillac, who were primarily interested in the problem of representation. In physiology, this approach usually resulted in attempts at localizing representations instead of functions.