ABSTRACT
The brain is an enormously complex system. The total number of possible states in a network comprising some 1011 neurons is exceedingly large. In fact, only a vanishing fraction of them will actually occur during a lifetime. This complexity has important strategic implications, both for experimental and theoretical approaches to brain function. In such a system, we cannot expect that the underlying principles will simply pop out from merely observing the neuronal activity during an experiment. A formal theory is needed to work out testable predictions regarding the functioning of the system. These predictions, in turn, lead to the design of new experiments that can critically test the theory. At the same time, a theory of such complex and only partially observable system must incorporate the relevant biological constraints. Otherwise, it runs the danger of degenerating into
a sterile formal game. Thus a close interaction of experiment and theory provides the optimal research strategy to make substantial progress in brain research.
