prev "Autoformalisation - Knowledge acquisition of professional skills" by Gregory Gromov,
   Microprocessor Devices & Systems, Moscow, 1986, N 3, p.80--91, Chapter 3


Economy of Computerization

    But where was the rational kernel of the ‘cybernetic euphoria’ of the 1950's that made this new field so attractive to many of  those outstanding scientists, well-known for their constructive results in hard science, who found themselves at least initially drawn into its orbit?

Evidently, this was brought about by the fact that the appearance in the mid-20th century of computers, these information processing machines, made the recording and long-term storage of mathematically formalized professional knowledge possible in a form that—for the first time in human history—allowed this knowledge to directly affect the operation of production equipment, thus eliminating the intermediate stages of knowledge’s interaction with humankind. The process of recording previously formalized professional knowledge in a form capable of directly influencing the operation of machinery was called ‘computer programming’. One might expect that the appearance of tools for the immediate production implementation of vast volumes of professional knowledge accumulated by humanity would result in an upsurge in labor productivity.

  However, we know that the first thirty years of the computer era had no such effect. The slowdown in labor productivity growth that we have witnessed in industrially-developed countries during the last few decades was mainly caused by the continuous withdrawal of specialists and labor from material production to the information sectors of our national economies. The swelling ranks of ‘knowledge workers’ was stimulated by the continuously increasing complexity of industrial society and, in consequence, the increasing volumes of information in circulation. However, whereas the machinery and automatic control systems in the material production field were continuously being upgraded, contributing to higher labor productivity there, information processing, which involved research and office workers, specialists, and managers of various levels, was more difficult to automate. That is why by the beginning of the 1980s the manpower engaged in the information sphere in most industrially-developed countries made up nearly 50 percent of the total manpower engaged in all branches of national economies, and was still showing rapid growth. Considering all this, one has to ask: why did the average level of labor automation in information processing remain almost unchanged over 30 years in the computer era?

The Formalized Knowledge Barrier

As a matter of fact, the first generations of computers were intended to solve clearly-stated mathematical problems, primarily problems of a purely calculative nature. “It is unworthy of human race to spend long hours, like slaves, on computation," mathematicians ·had been complaining for centuries. The computer, designed to solve such problems, in time proved to be a useful tool in a great number of other applications. Computers were mainly used in the fields where it was necessary to accurately solve clearly-stated, formalized problems. As a rule, these were high-priority problems in economics or defense, belonging to the most mathematically advanced branches of hard science. Attempts to use the computer in solving so-called information problems that usually came down to automation of storage and the speedy processing of great volumes of data with well-developed structure, based on standard algorithms (patent information retrieval, searching for bibliographic references, certain bank transactions, airplane ticket sales, etc.) were also successful.

The success or failure of attempts to separate professional knowledge from its divinely-selected repositories has until recently been determined by the possibility or impossibility of formalizing this knowledge using mathematical methods. The fields of professional knowledge that lent themselves to such formalization were called ‘hard sciences’.

But all attempts to find effective applications for the computer beyond the limited area of previously accumulated formalized problems and the funds of collated data built up in the process of economic development ran into considerable difficulties which grew rapidly as the understanding of the given field grew deeper.

As Shannon remarked, that computers are similar to learned academics. In calculating long lines of arithmetic operations, the computer leaves Man far behind. But when one tries to make computers do non-arithmetic operations, they appear clumsy and ill-equipped for such a job.

In the early 1980's, all this made it necessary to regard all previous development in a more critical light. As Shannon had predicted, “it will be all too easy for our somewhat artificial prosperity to collapse overnight ..."

Besides the above example of sacred knowledge, kept secret for thousands of years, being broken down by ‘astrologer-priests’ into laws of celestial mechanics, available for practical use, there are many other equally convincing examples to be taken from various fields of hard science that show how the process of knowledge formalization, if successful, resulted in the elaboration of mathematically rigorous techniques for disseminating ‘in-shop’ professional secrets. However, of no less significance is the fact that these frontal attacks, using all the formalization techniques that traditional mathematics had to offer, proved ineffective against a great many fortresses of professional knowledge.

For instance, in medicine, one of the most important branches of modern science, current methods of professional knowledge communication are in many aspects similar to those used a thousand years ago. And this cannot be explained by any lack of attention shown by hard scientists for this ancient field of knowledge. As the academician Israel Gelfand said, evaluating the results of 15 years of work in medical diagnostics and prognostication carried out by a group of mathematicians and medical specialists under his supervision, “the application of mathematical methods in medicine, though it has a relatively long history, is still in its initial stage. When we started working with real medical data, it became clear that the reliable general principles that mathematicians had applied in solving physical and technical problems were of little use in this new field.” The situation is evidently very much the same in other fields where mathematical methods are not traditionally used.

Further developing Gelfand’s idea of generalization, we can conclude that medicine is far from being the only ‘hard nut to crack’ for traditional methods of professional knowledge formalization. Evidently, this can be explained by the different levels of complexity in calculating a ballistic trajectory or predicting the temperature of a gas, on one hand, and describing the functional mechanism of a living cell, or the human organism as a whole, or a manufacturing enterprise or entire sector of industry, on the other.

“It is quite natural," Jacques Hadamard noted in this regard, “to speak about a more intuitive type of mind for which the area of idea combinations lies more deeply, and a logical type of mind, when this area is situated closer to the surface."


 "Autoformalisation - Knowledge acquisition of professional skills" by Gregory Gromov,
   Microprocessor Devices & Systems, Moscow, 1986, N 3, p.80--91, Chapter 3

  Copyright © 1986-2011 Gregory Gromov