neurophysiology: an integration 



1955 



properties of an over-all statistical population of 

 neurons, especially when arranged in sheets as in the 

 cortex; and the properties of waves of activity traveling 

 in such a population are leading to exciting be- 

 havioral models. Further, the influence of sheer 

 numbers of neurons is beginning to be apparent, and 

 the ideas of a functional neuron reserve and of an 

 optimal activity level help interpret many phenomena 

 in human behavior, especially such reversible or 

 enduring changes as are associated with stress, age, 

 personalitv and mental illness. 



Additional models are appearing rapidly which 

 help account for information handling in terms ot 

 channel capacity, availability and redundancy, and 

 of storage, noise and similar communication concepts. 

 Attention is being given to programming mechanisms 

 in the nervous system, and stud) here will surely 

 yield great insights; but this is still in its infancy. 



Finally, as the individual neuron controls its 

 metabolism and so its function, so the whole nervous 

 system is exquisitely protected by homeostatii 

 mechanisms so that its milieu, like that ol a fini 

 chronometer, is kept highK constant. Regulation "I 

 the cerebral circulation, a blood-brain barrier, 

 active transport mechanisms, special chemoreceptors 

 in the nervous system, special chemoeffectors tying 

 the nervous system to the pituitary and endocrine 

 mechanisms- these and many other highly-evolved 

 devices attest the importance of the central nervous 

 system to higher organisms. 



. 1« Information Model 



A recent volume by Broadbent (32) summarizes 

 extensive studies on perception, attention and per- 

 formance, and derives from these data a formal 

 model of the nervous system as an information- 

 handling machine. To some extent the entire nervous 

 system acts as a single communication channel and, 

 therefore, exhibits a limited capacity for which 

 sensory events must compete. (This capacity is 

 clearly related to the functional neuron reserve, 

 considered above, and so to the size of the cortex 1 

 The channel is not captured at random by a few of 

 the vast number of sensory events clamoring for it; 

 certain classes of events — for example, sounds of like 

 pitch or spatial localization — tend to be selected 

 together for transmission. Further, certain types of 

 classes are more likely to be selected than others; 

 which ones gain preference is related to the properties 

 of the stimuli and to the state of the organism. Among 

 the former, increase in intensity or in time since last 



accepted favors acceptance, and certain modalities 

 are prepotent over others, as hearing versus vision or 

 touch versus temperature in the dog. (Some constant 

 relations, as pitch and laterality, are reflected in 

 auditory cortex potentials, according to Ades.) On 

 the side of the organism, the presence of a particular 

 'drive state' favors selection of events related to its 

 reinforcement (satisfaction) — a hungry animal at- 

 tends to food. Further, experience is stored, in the 

 sense that the probability of one selected signal 

 following another helps determine future selection; 

 conditioning can alert or habituate attention. 



Perhaps more exciting than the preceding more 

 formal statements of common experience is evidence 

 that incoming information may be held in temporary 

 storage for seconds before passing through the limited 

 capacitv channel, along with other events of the same 

 i lass. The duration ol temporary storage can be 

 increased beyond this short time by passing the 

 information through the limited capacity channel 

 and returning it to temporary storage, a cyclic 

 process that can be continued lor some lime, but at 

 the expense of using up part of the capacity. Long- 

 term storage, by contrast, adjusts the internal coding 

 to the probabilities of external events and does not 

 encroach upon the dynamic communication channel. 

 (Hen- again the neurophysiological picture of a 

 dynamic memory, involving messages reverberating 

 in closed neuron l< ■< >| >■- , and ol .1 fixed memory, 

 involving material changes in and altered physio- 

 logical properties of such loops, is in congruence.) 



Of especial interest is die conclusion that a shift 

 from one class of events to another, in the selective 

 process, requires a time of significant duration in 

 relation to that lor continuous How of one class of 

 events. (Bartlev supplies evidence lor afferent channels 

 becoming relraclorv lor 0.1 see., with phasic filtering 

 of visual, auditor) and other inputs.) Attention 

 "blinks' or 'nods' while messages of one sort are 

 received during which another type is transmitted. 

 (This is based on studying the reception of con- 

 flicting auditor) messages, but is reminiscent of the 

 lading of pattern vision when jiggling of the retinal 

 image is prevented, as described by Eldred and 

 Whitteridge. Removal of the frontal intrinsic cortex 

 leads to •flickering' of attention, according to Pri- 

 bram, which would relate the selective filter to this 

 brain mechanism.) Finally, there is tentative evideni e 

 that information from one type of event is sampled 

 for some minimal time before it results in action; 

 decision waits upon some minimal accumulation of 

 relevant data. (This would fit the suggestion of 



