Biological Principles of Information Storage and Retrieval
Heinz von Foerster · 1967
Biological Principles of Information Storage and Retrieval Heinz Von Foerster BCL fiche #095, files #64–89 Published in Electronic Handling of Information: Testing and Evaluation, Kent, Taubee, Beltzer, and Goldstein (ed.), Academic Press, London (1967), pp. 123–147. Dr. Ruth Davis, in her keynote address (Chapter I), tossed out to all of us a very important challenge. If I translate this challenge into my own language, then es- sentially she said that all the large information handling organizations are, de facto, functional isomorphs of liv- ing organisms. And since this is so why don’t we, in our analysis and synthesis of these large organizational sys- tems, look first at the structure and the functions of these living organisms. Of course she didn’t use exactly the same words as I did right now. For instance, she didn’t say “living or- ganisms” she just simply said “people.” Or, for instance she didn’t say “functional isomorphs,” she just said “du- plicative mimicking.” But essentially I think she meant exactly that because she illustrated with her series of ex- amples very beautifully precisely that point. If I remember correctly, as one example she cited the large military organization, another one was a business concern and the third one was, I think, a large data re- trieval system. However, she showed that in these orga- nizations there still remain a lot of tenacious bottlenecks which apparently are always present. They don’t disap- pear. They’re just shifted around, pushed from one side to the other side, and the load of the responsibilities are just transferred from one shoulder to another, but never resolved. On the other hand, if one looks at biological systems, how they handle information, one may see that some of these bottlenecks are not at all present. All re- sponsibilities are beautifully distributed and the system functions very smoothly. With her challenge to consider structure and function of biological systems and the way in which they handle information, I think she put her fin- ger on an extremely important point in the whole field of information storage and retrieval. Of course at the same time she raised an extraordinarily difficult and very com- plex issue. Nevertheless, it is precisely to this issue that I would like to address myself this morning. Unfortunately, I can do this only in a very superficial fashion, hoping that at least you get a feeling of what we believe is going on in H EINZ V ON F OERSTER rather complex living organisms and of the way in which they handle information, for these systems handle infor- mation in an entirely different way than we do in our in- formation storage and retrieval systems. Moreover. they process the information fast, effectively and reliably. Perhaps you may get already a hunch of the funda- mental difference between the way in which living or- ganisms handle information and the ways in which we do this now by contemplating for a moment the fact that living organisms perform better and more efficiently with increasing complexity of their tasks, while of course, our information systems at the moment become increasingly slow and hopelessly deficient if they are confronted with tasks of even moderate complexity. For example, we all know that a three-year-old child easily speaks a natural language; a three-year-old child recognizes its makers in about one hundred milliseconds; however, a three-year-old child has great difficulty in multiplying 269 by 3,625. However, a computer can do this very fast, as you know. This is a very trivial task and hence it does it in a jiffy. But of course, he has tremen- dous difficulties and he’s very clumsy indeed if he has to communicate with us in natural language. In addition. I have never seen a computer recognizing its maker. Have you seen a computer wiggling its tail when the head of the computer department enters the lab? Clearly, there seem to be some fundamental differences. The problem to which I would like to address myself now is the basic issue in all the information processing of living organisms and I shall attempt to develop the various features of information processing in living or- ganisms from a single first principle. This first principle is essentially the principle of inductive inference, as you will see in a moment. But let me build up to that point. I just mentioned it so that you know what my intentions are. You know that the essential problem of a living or- ganism is not so much to live. It is essentially to sus- tain life, it is the problem of survival. That is to live for the next instant also, not to live only for the present mo- Biological Principles of Information Storage. . .