So-called identity theories that postulate the identity of mental phenomena with brain states are usually associated with materialistic ontology. However, the historical picture of the actual attempts at spelling out the mind–brain identities is more complex. In the eighteenth century such identities were most enthusiastically proposed by dualists , whereas non-reductionistic materialists such as Diderot tried to get along without them. In the nineteenth century physiologists such as Broca, Charcot and Wernicke, who postulated discrete and localizable neural correlates for ideas (...) and mental images, ended up, despite their professed materialism, defending theories that resembled Cartesian schemes and the theories of their eighteenth-century dualistic predecessors. Some problems resulting from the use of these theoretical models postulating local and discrete traces, images or representations in the brain in the context of materialist ontology are obviously related to their dualistic origins and to the fact that they are necessarily bound up with dualistic presuppositions. In the context of the criticism of the nineteenth-century memory trace paradigm it is, however, important to bear in mind that functional localization as such is not dependent on the notion of localizable representations. Nor are materialist theories of the mind necessarily dependent on postulating specific type–type identities between the mental and the physiological. (shrink)