Autopoiesis for Beginners

Autopoiesis means self-creation, but why is that concept so interesting?

It was coined by the Chilean biologist Humberto Maturana around 1970. It was picked up by familiy therapists, social constructivists, organizational psychologists and, famously, the German sociologist Niklas Luhmann.

But what does it mean? I have come across countless incomprehensible explanations. In what follows I’ll share my understanding, developed when I was associating with some of Maturana’s and his pal Heinz von Foerster’s groupies in California in the early 1980’s. Forgive me if my terms are a little rusty, I haven’t been keeping up.

Daydreaming as autopoiesis
To get us started, I will use the simplest terms possible, like this analogy:

My teenage daughter Clara sits in the couch, daydreaming. I tell her for the third time: ”Pick up your coat and school books!” She bends over and then stops, saying, ”You know, dad, the boy I talked to after school today?”

My failure to get through to her is what Maturana invented the concept of autopoiesis to explain (remember, the example is an analogy. Maturana was originally concerned with the biology of cells and organisms). Biological systems like my daughter do what they do, like daydreaming, largely unaffected by their environment. She is busy creating her own little world, probably involving that boy, and this process of self-creation Maturana designated autopoiesis (from the Greek, auto = self, and poiesis, as in poetry, to create or produce).

The sense in which her mind is closed to the rest of the world he called organizational or operational closure. This not to say that she is not open to the world in other respects. She clearly hears what I say; it just doesn’t penetrate. So, in this other sense, she is open and related to the world, not cut off from it.

This relates to what Maturana called the organism’s structure, to be clearly distinguished from its organization or set of operations. The structure of my daughter includes the molecules and energy that goes into her, as well as the particular configurations between them that obtain at any particular moment. Sound waves from my vocal chords do enter her ears; she is structurally coupled to the world, as Maturana would have it.

By this combination of the concepts of structural coupling and organizational (or operational) closure Maturana can explain why Clara hears me, yet does not hear me. She is open to my input, yet closed. I cannot tell her what to do, I cannot instruct her, because what she hears is so influenced by what is going on in her mind that she twists my input to suit her purposes, as it were: My mentioning her coat and books makes her bring up the subject of the boy that’s on her mind.

Goodbye, “input” and “output”
When input is received by an autopoietic system, ”input” is no longer an appropriate term, says Maturana. We can talk about a coin being put into a vending machine as an input, because the input determines the output: a can of Coke is released. But in Clara’s case I clearly fail to determine her subsequent behavior. To make us remember this point, Maturana uses the term ”perturbation” instead (in Danish, ”forstyrrelse” is used nowadays).

My words are a perturbation of Clara’s daydreaming autopoiesis. They affect it in hard-to-predict ways—ways that are clearly much more determined by what she is thinking about at the time. This response is a ”compensation,” not an output, says Maturana. In this sense, Clara starts to produce the output intended by my attempt to instruct her, that is, she bends to pick up her stuff, but then her autopoiesis takes over, as it were, and she talks about the boy instead (this being the compensation).

An autopoeitic system is structurally coupled to the world, yet organizationally closed, and so responds to perturbations by making compensations. This defines the system’s autonomy, another key word with Maturana.

Incidentally, he doesn’t like the term system, either, because it smacks of input-output determinism. In the 1960’s it certainly did. Then, much systems thinking was about inputs crossing boundaries into systems that were like black boxes inside, spewing out output. This type of thinking was only a minute step away from the behaviorism of the 1950’s with its stimulus-response models.

Maturana wanted to put miles between himself and simplistic input-output systems. So, instead of system, he used the word ”unity” to designate an autopoietic sys…, oops, thing. This term never picked up among his followers, who happily use the word ”system” today (reasonably enough: No systems thinker today would insist on input-output determinism).

Cells and organisms
Now, let’s put the analogy of my daughter aside and talk about what Maturana addressed originally: biological cells and organisms. A classical problem in biology is how organisms maintain their identity despite their enormous throughput of energy and molecules. I may contain none of the molecules that made up my body ten years ago, yet I am the same person. What defines that identity?

An answer that was emerging in the 1960’s involved homeostasis, as expressed in cybernetics and systems thinking: An organism pursues goals, such as survival (and hence identity), and its input is used for this purpose. Maturana wanted none of that. Too mechanistic. Applies to a thermostat, fine, but not to a living cell.

It’s a system’s autopoiesis that defines its identity. It is by way of its organizational closure that an organism or a cell keeps doing what is does. All the nutrients ingested and secreted by an organism are part of its structural coupling with the world, so in this sense the system is open. However, all matter and energy entering the system are used and kept in check by the cell’s organizational closure. Or, rather, the fact that matter and energy are used appropriately to maintain a system the way it is is evidence of its organizational closure.

Sure enough, all the ferocious activity of billions of molecules interacting inside a cell is certainly not determined by the nutrients and other molecules crossing its membrane every second. The activity is somehow self-determined, and the entering molecules become part of that show, but in a role that is subordinated to the system’s own goings-on. (Note the informal language I use here. Maturana wanted precision and hence opted for the many technical terms; hence his lingo often sounds contrived and strained, even affected.)

Networks of interactions
Maturana saw the cell’s activities as a kind of network of interactions, where all the molecules going into and coming out of the chemical reactions would link up in a circular fashion. Molecules resulting from one reaction would go into the next, and so on, looping back, so that the network would be producing itself in a recurrent kind of way. The network creates its own constituents, the very material molecules of which it consists. Maturana wanted to focus attention on the circularity of molecule production and network reproduction and used ”autopoiesis” for that.

A machine cannot create or maintain itself, it produces stuff that is other than itself, it is an allopoeitic system, said Maturana (allos, Greek: other). I forget whether he went so far as to say that life is defined by autopoiesis. This would be a sensible thing to say. And in the early 1970’s, a most radical thing, too, when biology was entering its now love affair with the genome (which Maturana didn’t much like either: a little program that sits inside every cell and tells it what to do. Not very autopoietic, much too deterministic and one-way, not circular enough.)

Maturana in his own words
Anyway, we are now ready to tackle some of Maturana’s own dense jargon. This paragraph is from an early publication co-authored by his student Francisco Varela (who, by the way, went on to become a scientist of considerable interest). By machine and unity they mean system, and the components mentioned are what I just called molecules. Think cell while you read this.:

"An autopoietic machine is a machine organized (defined as a unity) as a network of processes of production (transformation and destruction) of components which: (i) through their interactions and transformations continuously regenerate and realize the network of processes (relations) that produced them; and (ii) constitute it (the machine) as a concrete unity in space in which they (the components) exist by specifying the topological domain of its realization as such a network." (Maturana and Varela, 1973, p. 78)

In much simpler and more inaccurate terms: A cell is a thing with a internal network of interacting molecules that (i) reproduce this network of interactions as they do their thing and (ii) thus make up the cell as a concrete little networky thing that sits right there.

Stripped of its dense jargon, we can see Maturana’s real contribution here: his distinction between the cell’s (i) organization, closed as it is, and its (ii) structure, its concrete existence as a material thing that happens to look the way it does right now. It is the ”organization”, realized through autopoiesis, that makes a body what it is, that is, all the invariants that persist as the body changes and ages. But the structure is the body you see in time and space.

Perception is autopoetic, too
Maturana went on to apply these ideas to perception, particularly the neurophysiology of vision. The basic idea is that vision is not determined by signals arriving like input to the retina. Such signals are perturbations that affect the autopoietic network of ongoing processes in the central nervous system. The visual images that are formed in the brain are as much a product of the goings-on in the visual system as it is a product of light entering the eye.

To Maturana, this is true of all “input” entering the senses: sounds, words, music, smell, touch etc. Everything is filtered through what is already known, that is, by all the constraints and background knowledge and the whole way our minds are structured (note that these are my colloquial terms, not his). Other traditions in philosophy (phenomenology, hermeneutics, pragmatism, to name a few) and in psychology (Piaget, gestalt, what have you) have said exactly these things, of course, and to my mind this just adds to the general sense that Maturana is making with the concept of autopoiesis. It may provide a way of linking up these philosophical and psychological traditions with biology.

Bear in mind also that he arrived at these position from the desire to describe the cell and the organism in ways more congenial to living systems than the mechanicism of input-output systems thinking of the 1960’s.

Autopoiesis elsewhere
A most conspicuous application of Maturana’s thinking is, of course, the work on social systems by the German sociologist Niklas Luhmann, which is currently picking up fame. I’m neither happy with this application nor very knowledgeable about it, but see my blogpost “Did Luhmann write over Maturana’s dead body?”.

My sister, Inger Ravn, did her master’s thesis in biology at RUC on Maturana in the mid-1980’s. She co-wrote a lengthy introduction to the Danish edition of Maturana’s and Varela’s book “The Tree of Knowledge” (Kundskabens træ. Ask, 1987). Her introduction took a Maturana quote for its title: “This is a cosmology and as such, it is complete”. During that research she really came to dislike Maturana for his intellectually totalitarian pretensions, his cosmology, as in “I can explain everything and if you disagree, I can explain that, too”.

He certainly spoke like a guru in the 1980’s and had wrapped himself into his own little self-referring network of terms, his terminological autopoiesis, as it were, which is also what makes him so tedious to read.

Jesper Hoffmeyer and I interviewed him at DISPUK north of Copenhagen in the early 1990’s (the text, “Kærlighedens biologi”, was published in Hoffmeyer’s magazine OMverden, 1991, no. 7, pp. 17-19.). On that occasion, he appeared to be a reasonably humble man who made good sense, and that is how I left him.

The transfer model of learning
Recently, he has popped up again in my mind, as I have tried to reconceptualize the knowledge sharing for my work on “learning conferences” at Learning Lab Denmark. I have found his basic terms, as described above, quite useful in thinking about the transfer model of education and communication and why it falls short.

A lecturer wastes most of his energies because his words are not picked up by the audience as transmitted. A listener has a mind that is organizationally closed while structurally open; he receives the sounds but makes unpredictable kinds of sense (if any) of the speaker’s message. This is not due to defects in the listener’s mind, but simply the way minds work. They did not evolve to accommodate classroom teaching.

I don’t think Maturana ever really got round to deciding what could be done to improve communication in situations like these. I have found little use for him in that regard and have instead turned to pragmatic concepts of knowledge and a humanistic view of human nature and psychology.