Neurophysiologists are often accused by colleagues in the physical sci-
ences of designing experiments without any underlying hypothesis. This
impression is attributable to the ease of getting lost in the
ever-increasing sea of professional publications which do not state
explicitly the ultimate goal of the research. On the other hand, many of
the explicit models for brain function in the past were so far removed
from experimental reality that they had very little impact on further
research. It seems that one needs much intimate experience with the real
nerv-. ous system before a reasonable model can be suggested. It would
have been impossible for Copernicus to suggest his model of the solar
system without the detailed observations and tabulations of star and
planet motion accu- mulated by the preceeding generations. This need for
intimate experience with the nervous system before daring to put forward
some hypothesis about its mechanism of action is especially apparent
when theorizing about cerebral cortex function. There is widespread
agreement that processing of information in the cor- tex is associated
with complex spatio-temporal patterns of activity. Yet the vast majority
of experimental work is based on single neuron recordings or on
recordings made with gross electrodes to which tens of thousands of
neurons contribute in an unknown fashion. Although these experiments
have taught us a great deal about the organization and function of the
cor- tex, they have not enabled us to examine the spatio-temporal
organization of neuronal activity in any detail.