Is There an Addressed Memory in the Nervous System?

Abstract

A memory is presumably a useful facility for the organism. As with any equipment, if we want to find how it works, it is as well to begin by asking for what it is used. The value of the memory for the organism is to increase the probability that actions taken will lead to survival. We shall expect to find the memory, then, connected on the one hand with the taking of actions and on the other with some system for assessing the results of these actions. This, which we might call a biologist’s way of looking at memory, is rather different from the more usual treatment of the memory as a record. There is indeed a sense in which the memory is a record of the past, but it is important to realize that it is probably not in any such form as a tabulated record that is consulted item by item. Von Foerster [11] has made this important point wittily by calculating the size of the record that would be needed to write out a multiplication table for 10-digit numbers (10¹⁰ x 10¹⁰). He estimates that the table would need books occupying a shelf 10¹⁵ cm long, which is 100 times the distance of the earth from the sun. However, he continues, by means of a little desk calculator with 20 wheels, he can obtain the answer to the multiplication of any two 10-digit numbers in a few seconds.