THE CONNECTIONIST SIMULATION OF APHASIC NAMING

The simulation of language disorders using interactive activation (IA) networks and connectionist systems is discussed. An existing LA accou nt of aphasic naming is described, in which two network parameters (de cay rate and connection strength) are varied to fit the error producti on of an aphasic patient. Fairly similar results can be obtained throu gh modification of additional parameters, including the so-called ''sh ared weight increase factor'' linking lexical and semantic units. This leads us to consider simulation of aphasic naming using connectionist networks which do not require explicit variation of network parameter s. A madular connectionist architecture is presented, in which semanti c-lexical and phonological knowledge are instantiated using self-organ izing Kohonen maps, while connections between them are implemented usi ng Hebbian networks; a linear connectionist network (Madaline) is used to simulate nonword repetition. The Hebbian connections are lesioned in order to reproduce the patient's naming errors. (C) 1997 Academic P ress.