ABSTRACT
One major problem of psychology, raised by Lashley, and raised again by artificial intelligence is how is it possible to produce complete and integrated percepts and sequences of behaviour on the basis of data from the outside world which is fragmentary and when the components of the brain are unreliable. Lashley's answer was rather unsatisfactory. He continually pointed out what was wrong with other people's formulations, convincingly arguing, for instance, that any version of connecting responses to stimuli was hopelessly inadequate. But his own solution, mass action and equipotentiality, is scarcely a solution. It simply restates the problem. We are now in a position to look more deeply into the principles which make mental processes reliable in the face of unreliability or poor quality of components and data.
