The second creation

A conversation about the theory of autopoiesis and its intellectual history

Humberto R. Maturana (* 1928) first studied medicine in Chile, then anatomy in England, was awarded a Ph.D. in biology at Harvard in 1958, and subsequently worked at the Massachusetts Institute of Technology (M.I.T.). In 1960, he returned to work with the University of Chile at Santiago, which he only left for occasional periods of research and teaching abroad. He is particularly well known for his theory of autopoiesis (self-creation) that he began to develop in the late 1960s. This theory provides a novel feature of living beings going beyond the traditional criteria of biology - reproduction, mobility, etc.


According to Humberto R. Maturana, a circular, autopoietic form of organisation distinguishes living beings, from the amoeba to humans. Living systems form a network of internal and circularly enmeshed processes of production that make them bounded unities by constantly producing and thus maintaining themselves. Autopoietic systems are autonomous. Whatever may happen inside them, whatever may penetrate and stimulate, perturb or destroy them, is essentially determined by their own circular organisation. The concept of autopoiesis began to gain greater popularity in the early 1980s.

In the meantime, it has exploded in academic circles and become a synonym for an autonomous form of reality production. It has taken on a vital life of its own as a universally exploitable trendy concept in journals on systems thinking and family therapy or at the conferences of sociologists and media scholars - even in the face of resistance by its creator. Humberto R. Maturana is still active as a professor of biology who seeks to promote a theory of cognition in the context of the natural sciences. Until his retirement, he was director of the Laboratory for experimental epistemology and the biology of cognition at the University of Chile in Santiago, which he had founded himself.

In the year 1944 the physicist Erwin Schrödinger published a small book that has since become a classic in the history of science. It carries the title: What is Life? Your own thinking has been intensely concerned with this question. You have developed a description of the living - as a biologist -, the theory of autopoiesis, which is still causing excitement in the scientific world. However, let us start at the beginning. Why have you been so deeply fascinated and obsessed by the question of what it is to be living? Was there a particular incident, some key intellectual experience?
Humberto Maturana: In fact, there have been various incidents and different key experiences that have inspired me. You must realise that I was often very ill as a child; death was a constant companion in the days of my childhood. I fell ill with tuberculosis several times, and the threat of this disease made me think about the relation between life and death quite early. I remember writing a poem at the age of 14 years, which deals with the difference between a corpse and a rock, the corpse being different from the rock because it had been alive. The fact of being alive was, therefore, not a property of matter - but what was it then, I asked, if one can lose it?
You are describing a dialectical pattern: In the encounter with our own death, we become aware of our craving for life.
Humberto Maturana: You could say so. In the year 1949 I was in a sanatorium in the mountains, being ill with tuberculosis again, and I had strict orders not to exert myself in any way; the prescribed therapy was, in fact, not to do anything at all. In secret, however, I read two books. In Nietzsche's Thus Spake Zarathustra I discovered that wonderful story of the metamorphosis of the spirit, the spirit being transformed first into a camel, then into a lion, and finally into a child. The child is described as the first movement: If I ever got out of the sanatorium alive, I thought, I would be like a child, starting from scratch, at the beginning again. Towards the end of Julian Huxley's book Evolution: The Modern Synthesis, I came upon a chapter in which I read that evolutionary progress meant increasing independence of a living being from its environment. Human beings, therefore, appear to be the most independent and most advanced living beings. So there I was lying in my bed, completely dependent on my medium, unable to leave the sanatorium, ill, close to death, and knew clearly that Julian Huxley could not have been right.
If I understand correctly, the confrontation with death led you to ask the question of the nature of life. Moreover, Nietzsche and Huxley offered answers, which you could relate to your own situation.
Humberto Maturana: Indeed. Life, I said to myself, has no meaning, no sense, and does not follow any programme of evolutionary progress. My conclusion had a tautological ring: the sense and the purpose of a living being is just to be what it is. The purpose of a dog is being a dog, the purpose of a human being to be a human being. Anything affecting a living being and happening to it, it became clear to me, had to do only with itself. When a dog bites me because I have stepped on its tail, then it bites me because it wants to avoid pain. This means that living beings are autonomous, that they have defined limits, that there is a boundary marking what belongs to them and what does not.
It has become customary in biology to answer the question of what being alive means by drawing up a list of necessary properties. To be alive means, it is said for example, to be capable of reproducing and to be able to move around. Why did you not find any such list satisfactory?
Humberto Maturana: Because such a procedure leaves in the unclear how we could possibly establish that the catalogue of possible features and criteria is complete. In the year 1960 there was another decisive key experience; a student asked me during a lecture what actually began four billion years ago that would justify our claim that it caused the beginning of life. The question embarrassed me considerably because I could not answer it. Therefore, I asked the student to come back a year later; I would then be in a position to answer his question. I kept asking myself, though, as I continued to ponder the question, how I might be able to decide to have actually found the proper answer. How could I be sure to have defined life appropriately by listing features like reproduction or locomotion, special chemical composition, or a combination of such features?
The problem is how we might be able to prove that we have found all the central features.
Humberto Maturana: Drawing up a list of features presupposes, strictly speaking, knowledge of all the potential features. Only those who believe that they already have the answer, although they are still looking for it, could possibly know when their list is complete. I was, however, searching for an understanding of living systems, which did not require the enumeration and classification of all the components and processes involved. I was looking for a form of organisation common to all living systems, which had to be independent from their particular components and their particular structures.
How did you come to develop the theory which has become widely known under the catchword autopoiesis?
Humberto Maturana: My own thinking went through various stages. At first I spoke about systems without an external purpose, whose activities make sense only within their own being. These self-referential systems were distinguished from allo-referential systems, whose essential feature was that they served a purpose external to them. (A car, for example, is an allo-referential system: its sense and purpose is to serve as a vehicle of locomotion from one place to another.) The concept of reference did not really appeal to me, though, because it always involves a relation between different elements - and I did not want to describe a pattern of relations; I wanted to understand the processes of a system through the system itself. Therefore, I looked for a concept that would highlight the processes that ultimately resulted in the phenomenon of self-reference.
You wanted your theory of the living to be alive itself.
Humberto Maturana: I was both obsessed and fascinated by a specification of the living that could not be separated from the actual realisation of the living. Although I had read Erwin Schrödinger's book, my question was not what life was, but what essentially constituted a living system. I wanted to discover the configuration of processes, the specific molecular dynamics, which produced, as a result, a living system, a cell, for instance. What must happen for such a system to arise? Conceptually, at least, I wanted to create a living system; that was my goal.
You wanted to play God.
A (laughing): I did not want to play God, I wanted to be God.
What happened next in the step-by-step development of your new theory of the living?
Humberto Maturana: In 1963, in the laboratory of a microbiologist friend with whom I regularly discussed the developments in molecular biology, I had the decisive inspiration. The dogma in molecular biology was, at the time, that information travels from the cell nucleus to the cell cytoplasm. We asked ourselves whether it might not also travel the other way, from the cytoplasm to the nucleus; nobody had yet heard of retroviruses, so our question was quite justified. We designed experiments that were never implemented, but one of those days I drew a diagram on the blackboard and said to my friend: "The DNA participates in the synthesis of the proteins, and the proteins participate as enzymes in the synthesis of the DNA." My diagram had the form of a circle. When I looked at what I had just drawn on the blackboard I exclaimed: "My goodness, Guillermo, that is it! This circularity of the processes reveals the dynamics that makes living systems autonomous, bounded, and independent entities." I had found the conceptual basis of the phenomenon that was later termed autopoiesis. From then on, I described living systems as circular systems.
We have now reached the last phase of this short science-historical prelude. How did the concept of autopoiesis finally come to be invented?
Humberto Maturana: It must have been in the year 1970; I had met with my friend José Maria Bulnes who had written a doctoral thesis on Don Quijote. In this thesis he dealt with the dilemma confronting Don Quijote: to choose the path of poiesis (production, creation), or to involve himself in praxis (actual work), without paying much attention to the consequences of his actions. He finally goes for praxis and decides to become a knight errant, and therefore decides against poiesis and writing novels about knights errant. During that conversation it hit me: "That is the word I have been looking for: autopoiesis." It means self-creation and consists of the Greek words autos (self) and poiein (produce, create). I had successfully condensed into a concept my idea of what essentially characterises a living system. There was the additional advantage that the term was completely unknown - in contradistinction to the somewhat cumbersome expression circular systems -, and that it focussed the attention on the result of the constitutive processes of systems that produce themselves as unities through their own operations. The product of the autopoietic organisation of a system is that very system itself.
Can the concept of autopoiesis be specified in greater detail?
Humberto Maturana: Living systems produce themselves within their closed dynamics. They share the autopoietic organisation in the molecular domain. When we examine a living system, we find a network producing molecules that interact with each other in such a way as to produce molecules that, in turn, produce the network producing molecules, and determine its boundary. Such a network I call autopoietic. If we, therefore, encounter such a network in the molecular domain, whose operations effect its own production, then we are dealing with an autopoietic network and, consequently, with a living system. It produces itself. This system is open to the input of matter but closed with regard to the dynamics of the relations that generate it.
Perhaps an example demonstrating the autopoiesis of the living would be helpful at this stage. You have often referred to the cell as an autopoietic system. Would that be a compelling model?
Humberto Maturana: In my terminology the cell is described as a molecular autopoietic system of the first order; consequently, a multicellular entity is an autopoietic system of the second order. The special thing about cellular metabolism is that it produces components, which are in their entirety integrated into the network of transformations that produced them. The production of components is, therefore, the condition of the possibility of an edge, of a boundary, of the membrane of a cell. This membrane, in turn, participates in the ongoing processes of transformation, it participates in the autopoietic dynamics of the cell: it is the condition of the possibility of the operation of a network of transformations that produces the network as an integral whole. Without the boundary of the cell membrane everything would dissolve into some sort of molecular slime, and the molecules would diffuse in all directions. There would no longer be an independent entity.
This means that the cell produces the membrane and the membrane the cell. The producer, the act of production, and the product, have become indistinguishable.
Humberto Maturana: I would say, a little more rigorously: The molecules of the cell membrane participate in the realisation of the autopoietic processes of the cell and in the production of other molecules within the autopoietic network of the cell; and autopoiesis generates the molecules of the membrane. They produce each other, and they participate in the constitution of the whole.
The specification of autopoietic systems of the first order I find convincing but I do not understand how you can say that autopoietic systems of the second order (e.g. human beings) can produce themselves. Could we not say just as well: Human beings, in their everyday lives, essentially produce what is different from them. People work, build houses, bake bread, knit sweaters, etc.
Humberto Maturana: Naturally you can see human beings in this way. When you describe persons in the social domain as workers, then it is certainly possible to consider them as producers of bread or sweaters and characterise them primarily in this way. That they are living systems is practically irrelevant in this context because machines that make the same products could replace them.
To classify an entity as an autopoietic system of the second order might, therefore, be the result of the approach selected and the perspective chosen?
Humberto Maturana: Yes and no. The microphone that we use to record our conversation cannot simply be regarded as an autopoietic system even if we were very much tempted to do so. Only children can do that sometimes. In their play, the non-living may appear to be living. That is play, however, and we know that it is.
The possibility of changing perspectives is only successful with one direction, then: We may view autopoietic systems as systems that produce something different from themselves. In the case of non-living systems such an approach does not work, they cannot be classified as autopoietic (simply because we feel like doing so).
Humberto Maturana:: Right. If I describe our microphone as a living system then you will certainly want to know how the autopoiesis of this system functions. And I shall not be able to give you a satisfactory explanation.
What do you call systems that create something different from themselves?
Humberto Maturana: Originally I called them, as already mentioned, allo-referential systems, today I call systems of which we can say that the sense and purpose of their operations are external to them, allopoietic systems: the result of their operations is not themselves - just think of cars and computers. This is, however, not at all a depreciative term, and it should not be misconstrued as the expression of a discriminatory hierarchisation. Without my car and my computer, I could not live the life that I would like to be leading.
Is the feature of autonomy central for the realisation of autopoiesis? We could certainly claim that practically all systems are autonomous because they all function according to their own rules. If I shout at the waiter in a café at the top of my voice to order a coffee, he may still fail to bring me one. The same will happen if I ask my coffee maker (an allopoietic system) quite politely to make me a cup of coffee. Coffee will be produced only when a filter is put in, when water is filled in, when the machine is switched on, - in brief, if I play according to the rules of the machine.
Humberto Maturana: Naturally there are various possibilities for different systems to be autonomous, to function according to their own rules. Of course, there are many autonomous systems, which are not living systems. It would be mistaken to see autonomy as the key feature of autopoiesis. The central point is that we have a closed network producing molecules that, in turn, produce the network that produced them. Put in a formula: Autopoiesis is the specific way and manner intrinsic to living systems of being autonomous, of realising their autonomy. Autonomy is the more general notion.
How do we know that autopoiesis, this particular form of circular organisation, really is the crucial criterion of life? How could you prove that?
Humberto Maturana: We could prove it by successfully presenting a series of processes that would as a result produce what somebody wants to have proved. We would have to demonstrate that the realisation of autopoiesis is the direct or indirect source of all the characteristics of living systems and ultimately produces an entity that has all the known and unknown features of a living system.
You designed a computer model that simulates an autopoietic system. In the scientific literature there has been the odd voice claiming that this simulation actually refuted your theory. Your simulation, the argument runs, is evidently not alive, although it may possess the features of a living system.
Humberto Maturana: I can only reply that that model was meant to be an illustration and does not claim to prove anything. It certainly is no living system. The computer functions like the puppeteer in a puppet theatre, it is used to transform the different elements into entities that exhibit in the space of observation, i.e. the graphic space, a dynamics comparable with the molecular dynamics. The computer and its program are used to drive the elements which move autonomously in a living system: molecules need no puppeteer, they need no hidden power to move them, they move themselves - by thermal agitation due to thermodynamic laws. That is precisely what is special about them. However, as you know, immense efforts are invested at present to create artificial life. One day these attempts will undoubtedly succeed, although they hold terrible risks, and we will be able to build autopoietic systems in the molecular domain.
If your prediction is borne out and artificial life is created, God would not simply be dead, as Nietzsche once said, God would have become superfluous, eliminated by the creation of autopoietic systems. Would you agree?
Humberto Maturana: Not at all. Before we can answer such a question or formulate a thesis of this kind, we must have agreed on the meaning of the word God. Yogananda, the great Yogi who came to America, once said: When you think that God is far away then God is far away, when you think that God is near then God will be near. The word God designates a human concept that has gained enormous significance and power in our world. To many people, however, God does not appear to be, as the Christian doctrine prescribes, an intelligent and creative Being in whose image we were made. What is decisive for them is that speaking of God permits them to speak of an inaccessible presence and of a connection with the source of their existence about which we cannot actually speak at all. If I, therefore, see God as the source of everything then he can certainly never be superfluous. The fact that life forms itself under specific conditions is then an expression of God's existence.
: An author managed to let emotions run high in Germany for some time by concluding all his interviews with the same question, - a good question, I think: Does God exist?
Humberto Maturana: I was asked myself at the end of a lecture: "Do you believe in God?" My answer was: "I exist in God's Kingdom." The questioner repeated: "Do you believe in God?" I answered again: "I exist in God's Kingdom." And he repeated the question once more: "Once again: Do you believe or do you not believe in God?" - "Would you like me more or less," I said to him finally, "if I answered Yes or if I answered No?" His insistence was due to a desire for discrimination.
And your answer meant, in fact: God's existence is not a matter of faith.
Humberto Maturana: I would say that people believing in God are plagued by severe doubts.

This dialog is an excerpt from the forthcoming book: "From Being to Doing. The Origins of the Biology of Cognition" by Humberto R. Maturana and Bernhard Poerksen. It is due to be published soon by Carl-Auer Verlag (ISBN: 3-89670-448-6). For more information see:

(Humberto R. Maturana/Bernhard Poerksen)