neurophysiology: an integration 



1921 



richness of behavior must some day be explainable 

 (although, like a particular storm, not necessarily in 

 unique detail) in terms of neurophysiology and, later, 

 of neurochemistry. There have been marked recent 

 advances, both in characterizing the behavior to be 

 explained and in developing neural models to explain 

 it. 



the machine. Knowledge of the nervous system has 

 come a long way from ancient times when the soft 

 gray stuff that filled the calvarium was regarded as a 

 reservoir of nasal mucus en route from its source in the 

 pituitary through the cribriform plate to its sink in 

 the nose. Yet it was only three centuries ago (1660) 

 that Schneider proved this view erroneous. First 

 the anatomist, acting almost as a taxonomist, identi- 

 fied and named the major bumps and hollows, the 

 gray islands and white channels. The primitive 

 physiologist or pathologist soon laid down the rudi- 

 ments of physiological anatomy by observing the 

 more obvious consequences of irritating or damaging 

 one part or another. Later, the microscopical 11101- 

 phologist, with his hardening fluids and staining tinc- 

 tures and microtomes, described cells and fibers, con- 

 nected them with each other (with a great assist from 

 experimental embryology), and is still proceedim; with 

 the enormous task of diagramming the complete wir- 

 ing circuits in a citv of 10 billion inhabitants, extended 

 in all three dimensions and reduced to the size of a 

 soft ball. The chemical morphologist, identifying and 

 locating the molecular elements, came much later 

 and his studies are largely for the future, along with 

 those of the chemical traffic of the living machine at 

 rest or when active. Analytic neurophysiology, con- 

 cerned with the mechanisms of action rather than 

 with their loci, is an offspring of this century and 

 remains in the early exponential growth phase; 

 analytic developmental neurology is still being born. 

 It would seem to behoove us, then, to lie hopeful 

 rather than impatient if today we can only adumbrate 

 the processes whereby the organ of mind plays the 

 tunes of behavior. 



behavior. Behavior is also seen today in a vastly 

 different setting than a few decades back. Just as 

 the picture of inheritance in terms of compressed 

 homunculi in germ cells gave way to one of interacting 

 developmental processes, so did that of behavior 

 relinquish the anima or soul or consciousness, a 

 mental homunculus that exercised volition, in favor 

 of interacting streams of information. Laboratory 

 and life experiments are defining the quantitative 

 relations between the amount, kind and multiplicity 



of input of information and the speed, precision and 

 complexity of performance — including the reports of 

 subjective experience. The studies distinguish between 

 the properties of input and output transducers, 

 transmitting channels, coding schemata, memory 

 stores, computer elements and even selector (or 

 valuing) devices. Attention, perception, ideation, 

 reason, performance and the like are being given 

 precise and operational meanings. It is one matter to 

 offer a neurophysiological interpretation of a slower 

 response to a more difficult discrimination, but quite 

 another to develop a neural model that gives reaction 

 time as the logarithm of information (proportional to 

 the number of bits presented), or 0.12 sec. .is the 

 time per bit for each input or output decision, or 

 perceptual capacity .is seven bits in the first dimen- 

 sion of "discrimination space,' fewer in successive 

 dimensions. The integrative action of the nervous 

 swem, the handling of allied and antagonistic 

 situations, is being examined as thoroughly at the 

 level ol language and logic as it was earlier at tlii- 

 lexel of spinal reflexes, and appropriate brain mecha- 

 nisms must l>e elaborated to account for the new rich 

 phenomena. 



COMF1 iik viodels. The great computers and the 

 attendant development of mathematics of relation- 

 ship (rather than of magnitude) will aid in model 

 building as the studv ol neural behavior has served 

 in understanding computers. The power sources of 

 muscles and of cranes and like social effectors are 

 unrelated, but both the musculoskeletal and the 

 wire-strut sv stems obey the same laws of mechanics. 

 The photochemical processes in eve and camera 

 differ, but the optics arc the same. The programming 

 and storing and decision-making mechanisms of 

 brain and computer have perhaps nothing in com- 

 mon, but the insights from information theory and 

 set theory, from stochastic analysis and topology, 

 from queueing theory and probability theory play 

 freely in both domains. 



Aside from the use of these rapidly evolving 

 'social brains' to simulate various models of a nervous 

 system, or to solve particular equations applied to 

 actual operations, the development of computer 

 theory has exhibited sharply the processes involved 

 in man's decision-making. Attempts made by Ledley 

 & Lusted (169), to build a computer program for the 

 diagnosis of disease, for example, have involved the 

 following steps: building a table of all signs and 

 svmptoms on one axis, of all diseases on a second, 

 tilling all squares with a plus or zero, and committing 

 this background information to memory. Further, 



