What we call computational neuroscience involves construction of mathematical and numerical models for understanding cognitive phenomena. This issue is devoted to showing how it can also be used to help in the analysis of cognitive defects. Although the models may seem abstract to clinicians, they are based on the reality of brain anatomy. The theoretical papers presented here are connectionist: They posit a network of cells connected by synapses whose weights are modified during learning. Architecture of connectionist models has progressed and ramified considerably since they were first introduced, and we include some examples of the current state of the art. The final work presented here is concerned with the connection of the constructed models with clinical experience and experiment.