Aging on a neural network

Summary form only given, as follows. Much of the data on the slowing of human information processes with advanced age can be modeled by neural networks that have been corrupted in ways suggested by neuropathology. The author assumes that a sequential, loop-free network carries out some information processing task. Two types of age defects are introduced on the network, a progressive breakdown in neural connectivity and a progressive attenuation in neural conduction. The consequence of these defects in terms of processing time are derived. Actual human response data are then reviewed, both as a function of task type and as a function of age. Tendencies in the data indicative of both types of network defects are seen. Which information processes are afflicted in which ways cannot yet be determined.<>