There are many ways in which a physical activity may be so modified as to cause a human observer's perception of it to depart widely from the indications of physical apparatus. We are all aware of the existence of sensory threshold, adaptation process, and nonlinearity, all of which represent departures of perception from the regular variations of magnitude in a physical stimulus.
A further process is inhibition, whose effects are equally unexpected and often even more profound than these objects.
(...)
The important thing in a communication network is not the output level attained but the signal-to-noise ratio, for it is the ratio that determines our ability to recognize the signal a distinct from the noise. In the nervous system also it was found that "noise" is always present, in the form of general background of spontaneous activity, and a sensory effect has to be identified in the presence of this background (Hoagland, 1932). This problem is still with us.
(...)
A decade ago there was a development of communications and information theory. Already we have too much of this theory, and it is necessary to develop an inhibition theory. We shall need to discover a way of measuring the loss of information caused by a given amount of inhibition. A possible measure is the number of bits lost when the information is passed through a system divided by the number of bits introduced at the input. Such a measure I consider more important in physiology and psychology than in communications engineering.
We know that any information that we have at hand contains a number of small disturbances. These are usually eliminated by a statistical treatment. The mean value of a series of measures represents the inhibition of many small unwanted bits of information, but this procedure does not go far enough. What we are interested in is direct from of inhibition that cancels out a whole of unwanted information. The problem is of far greater scope than statisticians have ever dreamed of.
The simplest way to get rid of information is to reduce the sensitivity of the receptors. This method s used effectively in all complex living systems. For example, we know that the organ or Corti of the ear is sensitive to displacement. This sensitivity is so great that one can almost hear the Brownian movements of molecules. (...) the organ of Corti, as living tissue, requires a constant blood supply (...) hence we should expect to hear our own heartbeat with tremendous loudness. We do not so because, for one thing, there is a factor of frequency differentiating between external sound stimuli and the sound of the heartbeat. The circulatory pulsations are of low frequency, and the solution that nature made to the problem was to reduce the sensitivity of the ear to lower frequencies while leaving the sensitivity unimpaired in the range between 1000 and 4000 cycles per second (cps).
Thus it is not surprising to find that the threshold sensitivity for frequencies around 20 cps is nearly 10,000 times less than for frequencies around 1000cps. These relations may be seen in Fig.3.
Figure 3
(...)
Adaptation is a common process in living systems for the reduction of the effect of a stimulus. It is seen mainly as a progressive loss of sensitivity during a period of stimulation.
(...)
The retina seems to lose its "pattern recognition ability" quickly if the pattern is maintained on one portion of its surface. As has been proved by Riggs and others (1953) an image that is made stationary on the retina disappears partly or completely in a few seconds.
(...)
That the transients of a stimulus are of utmost importance in vision was shown in an objective way in experiments on the eye of the horseshoe crab, Limulus, by Ratliff, Hartline, and Miller (1963). Even a moderate change in the stimulus intensity, such as doubling or halving, produced an immediate change in the discharge rate of a single unit of the optic nerve, as seen in Fig. 10. This figure shows the effect of a doubling of the light stimulus, and then a return of the stimulus to its former level. The response frequency increases rapidly to more than twice its original value when the light is increased, and in about 0.5 sec returns to the base line. Then when the light intensity is dropped to its initial level, the frequency falls abruptly and soon rises to the base level once more.
Figure 10
No comments:
Post a Comment