top of page

Self-organization in Biological Systems

ABSTRACT. The concept of self-organization is based upon the order-from-noise principle, which explains how organisms make use of noise (in the meaning of the theory of information), as a organization factor. The author describes the seIf-organizing systems, and its importance for the understanding of life.

After the establishment of cybernetics as a science, several scholars turned their attention to living beings, trying to classify them as machines whose performances, however extraordinary, were the result of principles ordered in continuity and identity with the general principles of cybernetics. These studies resulted in the concepts of self-organizing system and self-reproducing automaton as cybernetic ways of designating biological organisms. According to Henri Atlan (2), the consequences of this conception can be established as follows:

* The specificity of living organisms is linked to organizational principles and not to irreducible vital properties.

* Once these principles have been discovered, nothing should stop us from applying them to artificial automata, whose performance will then become the same as that of living organisms.

In order to understand self-organizing systems, it is important to establish the concept of reliability. A living machine is made up of molecules that wear out, cells that degenerate, in other words, unreliable elements. However, as a whole, it is an extremely reliable structure, capable of self-repair and functioning even when it breaks down. Morin (11) states that "the artificial machine, once constituted, can only begin to degenerate, while the living machine is, even temporarily, non-degenerative, that is, capable of increasing its complexity". For him, internal disorder, "noise" or error always wears down the artificial machine, whereas organisms always function with a portion of noise, and the increase in complexity of the living being improves its tolerance to it, and there is, up to a certain limit, an intimate generative relationship between the increase in noise or disorder and the increase in complexity.

Structures capable of functioning with this type of cybernetic logic are called self-organizing systems. The appearance of self-referential descriptions in current science bears witness, according to Heinz Von Foerster (1974), to the emergence of new concepts that he calls second-order concepts. According to him, we come across such concepts when a name, whatever it may be, appears preceded by the "auto specifier", as in the following examples: "self-organization", "self-replication", "self-repair", "self-regulation".

The ability to integrate noise, i.e. incorporate it into its structure without being destroyed by it, a characteristic property of self-organizing systems, led mathematician J. Von Neumann to establish that this ability was the result of a fundamental qualitative difference in the system's organizational logic. His work (15), plus that of Winograd and Cowan (16) and Cowan et al. (7) defined the necessary conditions for establishing automata with reliability greater than that of their components:

1- Redundancy of components 2- Redundancy of functions 3- Complexity of components 4- Displacement of functions.

Such conditions occur, for example, in the human organism, which is able to withstand intense variations in organic homeostasis without this leading to the disintegration of the system. The facts demonstrate the need, for complex structures such as organisms, for a certain degree of indeterminacy so that the system can adapt to a certain level of noise. Heinz Von Foerster (13) states that "in self-organizing systems with a sufficient degree of redundancy and reliability, when noise is introduced, it is enriching rather than disruptive or destructive". In the DNA molecule (the fundamental constituent of all living beings), with its capacity for self-replication and faithful transmission of the "noise" (errors) introduced randomly in the transcription of the genetic message, which every quantum system is subject to, noise sometimes leads to mutations, which natural evolution selects because they enrich the organism's performance. Von Foerster (13) also showed that we can only understand the unique properties of organisms by attributing to them, in addition to the property of resisting noise effectively, the property of also using it as an organizing factor! In this way, he established a principle of order from noise.

We know that since Erwin Schrödinger's work, What is life, published in 1944, thermodynamics has established the existence of two different principles by which ordered events can be produced: the statistical "mechanism" that produces order from disorder, and another that produces order from order, that is, complex organization and information that would explain living matter, and whose priority of understanding, according to Schrödinger himself, must be claimed for Max Planck, who, in a short work entitled The dynamical and statistical type of Iaw, already established, this distinction.

However, Von Foerster believes that this "principle of order from order" is not enough to explain living beings. Schrödinger stated that living systems "feed on negative entropy" from order, to which Von Foerster adds the statement that "they also find noise on their 'menu'!... It is not a bad thing that there is noise in the system. If a system gets stuck in a particular state, it is unadaptable, and this final state can be extremely bad. It will be unable to adjust to anything that is an unsuitable situation". This fact, i.e. immobilization in a definitively established order as shown by Atlan (2), characterizes one of the two possible forms of death for an organism, the other being the immobilization of the vital process due to total disorder. W R Ashby (apud Atlan (2)), through a series of works, established a law called the "law of indispensable variety" which reveals a relationship between the variety of response disorders and acceptable states. According to this law, the greater the variety of available responses, the greater the variety of disturbances and the smaller the variety of acceptable states.

In other words, for a system subjected to a wide variety of aggressions, a wide variety of responses is indispensable if this system is to remain in a limited number of acceptable states. Hence Atlan's statement that "in an environment that is a source of diverse and unpredictable aggressions, a variety in the structure and functions of the system is an indispensable factor of autonomy". Allan himself demonstrates this law in another work (1), with the example of an organism subjected to bacterial infections of different species, to which, in order to survive, it must respond with appropriate antitoxins. "If the bacterial species," he says, "each require a different antitoxin, then obviously the defence system must have as many antotoxins in its repertoire of responses as there are bacterial species, in order to be able to produce the only acceptable state, characterized here in a very general way by the survival of the organism" (p. 54).

It is also Atlan who shows us that Ashby revealed the deep kinship between the law and Shannon's noise path theorem, demonstrating that in this context the law of indispensable variety establishes that the capacity of the system cannot exceed its capacity as a variety communication path, that is, its capacity as a transmitter. Interestingly, Atlan observed that this form of limitation did not manifest itself in the physical and chemical sciences - which would explain its success - due to two particularities of the systems studied by these sciences, namely:

External degree of homogeneity of the constituent elements.

Relative lack of structural interrelationships compared to integrated biological systems.

According to Ashby (apud Atlan (1)), "these two qualities of complex systems - heterogeneity of parts and richness of interactions between them - have the same implication: the quantities of information circulating, whether from the system to the observer or from part to part, are much higher than those circulating when the researcher is a physicist or a chemist. And it is because the quantities are so high that it is true that the limitation becomes apparent in the selection of the appropriate scientific strategy.

It is quite possible that the quantities of information involved exceed the capacity of the researcher - or the group of researchers - as a transmitter". Ashby also demonstrated the logical impossibility of self-organization in a closed system, i.e. without interaction with the environment. Allan (4) also demonstrates this when he states that "the notion of self-organization sensu stricto is contradictory if we consider organization as the law of functioning of a system. The system cannot change due to an internal determinism alone, because it is this determinism that constitutes its constant law of functioning." André Bejin, in the introduction to the chapter "Auto-organization et connaissance", in the book L' Unité de L' Homme, says: "The self-organizing system 'feeds' on order (Schrödinger) and noise. In this way, as Ashby has shown, it can only be cybernetically open, and complexify itself normatively, using 'materials' that are not only those of its own operation." A self-organizing system can therefore only be modified by factors outside itself. In this way, there are two possibilities for modification:

Through a pre-established program that is injected into the system. This does not characterize noise in the sense of communications theory.

Through random factors that are introduced into the system in such a way that no pattern can be established that allows a program to be discerned.

In the latter case, we could talk about self-organization, even if it's not in sensu stricto, i.e. the system organizing itself without outside intervention, since despite the existence of external action on the system, this is totally random.

In order for noise to enter a system without destroying it, and to be able to have an enriching effect, the system needs to be made up of a complex cybernetic network, whose structure the noise penetrates by causing alterations that do not destroy the coherence of the infinite structural and functional interrelationships that control its performance. Atlan (1, 2) demonstrated that one of the possibilities for the destructive effect of noise - ambiguity-destruction - in a system to be overcome by the enriching effect - ambiguity-autonomy - is the "existence of an alphabet change with an increase in the number of letters, when we move from one type of subsystem to another, as a way of communicating between them".

According to him, this would be "a possible explanation for the change of alphabet observed in all living organisms when going from nucleic acids, written in a language of four symbols (four nitrogenous bases), to proteins, written in a language of twenty symbols (twenty amino acids)". Manfred Eigen (apud Atlan (I. 2)), through studies of chemical kinetics, arrived at results that seem to derive from the principle of order from noise, showing that the mechanisms of DNA replication, protein synthesis and enzyme regulation can be analyzed both in terms of the total amount of information in a population of macromolecules and the amount of information in the different species of macromolecules synthesized. In the same work, Atlan shows us that one of the problems raised by this approach is "the conditions under which certain information-bearing macromolecules can be selected at the expense of others, in a system where the only restriction is that the synthesis of these molecules is carried out by copying identical molecules. For the first time, the concept of selection with orientation, the foundation of theories of evolution, acquires a precious content that can be expressed in terms of chemical kinetics, unlike the usual vicious circle we fall into when we describe natural selection as survival of the fittest, the latter of which can only be defined by the fact that they survive!"

And further on, "one of the most spectacular results that Manfred Eigen reached, applying this theory to systems made up of a coupling of two subsystems with complementary properties, which are the sets of nucleic acids and the sets of proteins, is a possible explanation for the universality of the genetic code: it would be the inevitable result of evolution, where only this code could be selected".

Summarizing the concepts studied, we can establish that if a self-organizing system presents:

1º A high level of structural redundancy, i.e. the components of the whole are repeated a large number of times.

2º High functional redundancy, i.e. the ability of a logical function to be carried out at the same time on several levels of the whole that can control each other.

3º High reliability, i.e. the ability to operate using degradable constituent units which, despite the disorder and noise introduced into the system, do not cause an increase in the system's entropy and can even be regenerative and enriching as a whole.

To repeat, if a self-organizing system has all the above characteristics, we can infer that it will be able to react to the random effects of the environment by reducing its redundancy and reliability, without stopping functioning. The continuation of its operation could then give rise to greater variety and heterogeneity (as a result of the reduction in redundancy), making it capable of more improved regulatory performance. This ultimately consists of generating order (information) from noise.

We know that living beings are fundamentally structured from two groups of macromolecules: proteins and nucleic acids.

Proteins are responsible for the macroscopic structure of organisms, through their fibrillar form; and for controlling and catalyzing the thousands of chemical reactions that take place within them, through globular forms or enzymes (Monod).

Nucleic acids are molecular systems capable of self-replicating and transmitting genetic information, ne varietur, generation after generation.

The whole has a high level of structural and functional redundancy, allowing extraordinary reliability to be achieved, as it is organized in hierarchical stages through successive integrations of subassemblies (The Integrons, by François Jacob).

As a result, biological systems survive and reproduce, "evading the fall to equilibrium" (Schrödinger), cheating Carnot's principle, which "is a decree of death" (Brillouin), because they have "extremely highly complicated" forms of molecular organization, according to Von Neumann's expression, which allows the structuring of a bioarchitecture in integrated stages, and the occurrence of self-regulation.

Finally, we should also remember that living systems are thermodynamic systems with an extremely complex metabolic organization which, in addition to extracting negentropy from the environment and using noise as an enriching factor in its own cybernetic network, is subject, like any non-living system, to the principle of minimum energy which guides all chemical reactions towards the reduction of free energy, "the type of energy capable of producing work under conditions of constant temperature and pressure" (Lehninger).

In addition to the characteristics of biological systems explained above (architecture in stages, self-regulation, minimum energy), every living structure is the consequence of an evolutionary process that selects it because of its positive performance and, consequently, the survival of the species.


1 – Atlan. H. 1972a L ‘organization biologique et la theorie de l’information. Hermann, Paris.

2 – Atlan. H 1972b. Du bruit comme principe d’auto-organisation. Communications 18: 21-35. Centre d’Etudes des Communications de Masse École Pratique des Hautes Études. Le Seuil, Paris.

3 – Atlan, H. 1974a. Conscience et desirs dans systemes auto-organisateurs. In Edgar Morin, massimo Piatelli-Palmarini. L’unité de l’homme.Auto-organisation et connaissance: 449-465. Le Seuil, Paris.

4 – Atlan H. 1974b. Le principe d’ordre à partir de bruit, l’apprentissage non dirige et revê, In Edgar Morin, Massimo Piatelli-Palmarini.L’unité de l’homme. Auto-organisation et conaissance: 469-475. Le Seuil, Paris.5 – Bejin, A. 1974. Presentation. In Edgar Morin, Massimo Piatelli-Palmarini. L’unité de l’homme, Auto-organization et connaissance: 447-448. Le Seuil, Paris.

6 – Brillouin, L. 1959. Vie matiere et observation Albin Michel, Paris.

7 – Cowan, J. D. 1965. The problem of organismic reability. In Wiener e Schade, org. Progress in brain research. Cybernetics and the nervous system. 17: 9-63. Elsevier Publ. Amsterdam.

8 – Jacob, F. 1970. La logique du vivant. Édition Gallimard, Paris.

9 - Lehninger, A. L. 1976. Biochemistry, vol. 1 Molecular components of cells, Editora E. Blucher Ltda. São Paulo.

10 - Monod, J. 1971. O acaso e a necessidade, Editora Vozes, 2nd edition.

11 – Morin, E. 1973. Lê paradigne perdu: la nature humaine, Le Seuil, Paris, trad. Br. 1975, O enigma do homem, Zahar Editores.

12 – Schrödinger, E. 1976. What is life? Cambridge University press. Cambridge (first published 1944).

13 – Von Foerster, H. 1960. On self-organizing systems and their envionments. In Yovitz e Cameron org. Selff-organizing systems. Cameron, Pergamon. Nova york.

14 – Von Foerster, H. “Remarques introductives”. In Edgar Morin, Massimo Piatelli-Palmarini. L’umité de l’homme. Theorie de la cognition et epistemologie de l’observation: 400 Le Seuil, Paris.

15 – Von Neumann, J. 1966. Theory of self reproducing automata. Ed. W. Bruks, University of Illinois Press, Urbana, Illinois (apud Atlan, 1972b).

16 – Winograd, S. And Cowan, J. D. 1963. Reliable computation in the presence of noise, MIT Press, MIT Cambridge, Mass (apud Atlan, 1972b).


bottom of page