In his book on vision published in 1982, David Marr claimed that bottom-up or neurophysiological approaches to cognition had failed to live up to expectation, and were unable, in effect and in principle, to explain cognition. He advocated a top-down approach where formalization of cognitive operations was to serve as a starting point, followed by implementation of cognitive models, followed only in the last instance by neurophysiological investigation. Other commentators have followed suit. This presentation is an attempt to refute this position by showing that the relationship between neuroscience and cognitive science is in effect, and in principle, a two way street. The validity of bottom-up illumination of cognitive theorizing is demonstrated with numerous examples.