The Reality of Informationally Closed Entities:

In today’s post, I am looking at the idea of “informationally closed”. The idea of informational closure was first proposed by Ross Ashby. Ashby defined Cybernetics as a study of systems that are informationally tight. Ashby wanted cyberneticians to look at all the possibilities that a system can be in. Here the system refers to a selection of variables that the observer has chosen. Ashby noted that we should not look at what individual act a system produces ‘here and now’, but at all the possible behaviors it can produce. For example, he asked why does the ovum grows into a rabbit, and not a dog or a fish? Ashby noted that this is strictly related to information, and not energy:

Growth of some form there will be; cybernetics asks “why should the changes be to the rabbit-form, and not to a dog-form, a fish-form or even to a teratoma-form?” Cybernetics envisages a set of possibilities much wider than the actual, and then asks why the particular case should conform to its usual particular restriction. In this discussion, questions of energy play almost no part – the energy is simply taken for granted. Even whether the system is closed to energy or open is often irrelevant; what is important is the extent to which the system is subject to determining and controlling factors. So, no information or signal or determining factor may pass from part to part without its being recorded as a significant event. Cybernetics might, in fact, be defined as the study of systems that are open to energy by closed to information and control – systems that are information-tight.

Ashby’s main point regarding this is that the machine or the system under observation selects its actions from a set of possible actions, and this will remain the same until there is a significant event that causes it to alter the set of possible actions. The action of the system is entirely based on its structure, and not because an external agent is choosing that action for the system. The external agent is only triggering or perturbing the system, and the system in turn reacts. This idea of informational closure was further taken up by Humberto Maturana and Francisco Varela. The idea of “informationally closed” is a strong premise for constructivism – the idea that all knowledge is constructed rather than perceived through senses. They noted that as cognizant beings, we are informationally closed. We do not have information enter us externally. We are instead perturbed by the environment, and we react in ways that we are accustomed to. Jonathan D. Raskin expands on this further:

People are informationally closed systems only in touch with their own processes. What an organism knows is personal and private. In adhering to such a view, constructivism does not conceptualize knowledge in the traditional manner, as something moving from “outside” to “inside” a person. Instead, what is outside sets off, triggers, or disrupts a person’s internal processes, which then generate experiences that the person treats as reflective of what is outside. Sensory data and what we make of it are indirect reflections of a presumed outside world. This is why different organisms experience things quite differently. How Jack’s backyard smells to his dog is different from how it smells to him because he and his dog have qualitatively different olfactory systems. Of course, how Jack’s backyard smells to him may also differ from how it smells to Sara because not only is each of them biologically distinct but each has a unique history that informs the things to which they attend and attribute meaning. The world does not dictate what it “smells” like; it merely triggers biological and psychological processes within organisms, which then react to these triggers in their own ways. The kinds of experiences an organism has depend on its structure and history. Therefore, what is known is always a private and personal product of one’s own processes.

Raskin gives an example of a toaster or a washing machine to provide more clarity on the informational closure.

Maturana asserts that from the point of view of a biologist living systems are informationally closed–that is, things don’t get in and they don’t get out. From the outside, you can trigger a change, but you cannot directly instruct. Think of it as having a toaster and a washing machine. And, the toaster is going to toast no matter what you do. And, the washing machine is going to wash no matter what you do. And they both can be triggered by electricity. But the electricity doesn’t tell the toaster what to do. The toaster’s structure tells the toaster what to do. So similarly, we trigger organisms, but what they do has to do with their internal structure–including their nervous system–and the way it responds to various perturbations.

The idea of informational closure forces us to bring a new perspective to how we view the world. How are we able to know about reality? From a constructivism standpoint, we do not have a direct access to the external reality. What we can truly say is how we experience the world, not how the world really is. We do not construct a representation of the external world. This is not possible, if we are informationally closed. What we do is actually construct how we experience the world. As Raskin points out, the world is not a construction; only our experience of it is. Distinguishing experiential reality from external reality (even a hypothetical, impossible-to-prove-for-sure external reality) is important in maintaining a coherent constructivist stance.

All knowledge from this standpoint is personal, and cannot be passed on as a commodity. In constructivism, there is an idea called as the myth of instructive interaction. This means that we cannot be directly instructed. A teacher cannot teach a student with a direct and exact impact. All the teacher can do is to perturb the student so that the student can construct their personal knowledge based on their internal structure. Raskin notes – once people’s internal systems are triggered, they organize their experiential responses into something meaningful and coherent. That is to say, they actively construe. Events alone do not dictate what people know; constructive processes play a central role as people impose meaning and order on sensory data. 

The more interactions we have with a phenomenon, the better we can experience the phenomenon, and it aids in our construction of the stable experiential reality of that phenomenon. Repetition is an important ingredient for this. Ernst von Glasersfeld notes:

Without repetition there would be no reason to claim that a given experiential item has some kind of permanence. Only if we consider an experience to be the second instance of the self-same item we have experienced before, does the notion of permanence arise.

From this point, I will try to look at some questions that might help to further our understanding of constructivism.

What is the point of constructivism if it means that we cannot have an accurate representation of the real world? The ultimate point about constructivism is not about an ontological stance, it is about viability. It is about being able to continue to survive. All organisms are informationally closed, and they continue to stay viable. The goal is to fit into the real world. Raskin explains – the purpose of this knowledge is not to replicate a presumed outside world but to help the organism survive. In Cybernetics, we say that we need to have a model of what we are trying to manage or control. This “model” does not have to be an exact representation of the “system” we are trying to control. We can treat it as a black box where we have no idea about the inner workings of the system. As long as we are able to come up with a set of possibilities and possible triggers for possible outcomes, we can manage the system. A true representation is not needed.

How would one account for a social realm if we are informationally closed? If each of us are informationally closed, and our knowledge are personal, how we do account for the social realm, where we all acknowledge a version of stable social reality. Raskin provides some clarity on this. He notes:

Von Glasersfeld held that people create a subjective internal environment that they populate with “repeatable objects.” These repeatable objects are experienced as “external and independent” by the person constructing internal representations of them. Certain repeatable objects–those we identify as sentient, primarily other people–are treated as if they have the same active meaning-making abilities that we attribute to ourselves. Consequently, we are able to experience an intersubjective reality whenever other people respond to us in ways that we interpret as indicating they experience things the same way we do. Once again, this alleviates concerns about constructivism being solipsistic because people do relationally coordinate with one another in confirming and maintaining their constructions. 

For von Glasersfeld, it means that people construe one another as active meaning makers and consequently treat their personal understandings as communally shared when others’ behavior is interpreted as affirming those understandings. As I stated elsewhere, “when experiencing sociality or an intersubjective reality, we come to experience our constructions as socially shared to the extent that they appear to be (and, for all functional purposes, can be treated as if they are) also held by others”.

Each one of us construct an experiential reality of the external world. This external world includes other people in it. Our ongoing interaction with these people enhances and updates our own experiential world. We come to see the external world as a social construction. Our personal construction gets triggered in a social setting resulting in a social version of that construction. The more frequent and diverse interactions we get, the more viable this construction becomes. The other people are part of this experiential reality and thereby become cocreators of the social reality. In many regards, what we construct are not representations of the external world, but more a domain of constraints and possibilities. Making sense of the external world is a question about viability. If it does not affect viability, one may very well believe in a God or think that the world is flat. The moment, the viability is impacted, the constructions of the reality will have to adjusted/modified.

The image I have chosen for the post is an artwork by the Japanese Zen master, Nakahara Nantenbō (1839 – 1925). The artwork is a depiction of ensō (circle). The caption reads:

Born within the ensō (circle) of the world, the human heart must also become an ensō (circle).

Please maintain social distance and wear masks. Please take vaccination, if able. Stay safe and Always keep on learning…

In case you missed it, my last post was The Ghost in the System:

This post is also available as a podcast –


  1. An Introduction to Cybernetics, Ross Ashby (1956)
  2. An introductory perturbation: what is constructivism and is there a future in it?, Raskin, Jonathan D. (2015)

7 thoughts on “The Reality of Informationally Closed Entities:

  1. Couple of items (what’s new 🙂 I wrote about and think might be interesting for further discussion:
    1. Ashby’s statement “that the machine or the system under observation selects its actions from a set of possible actions” looks mightily similar to Shannon’s, that in communication “the actual message is one selected from a set of possible messages”. Don’t understand why Ashby thought that Shannon’s work has no place in his cybernetics. I use his description for a “transducer with memory” as a model for dynamical systems and I believe I have shown how it is in line and can provide more insight into Ashby’s work.
    2. “Informationally tight” means also that the system cannot be controlled from the environment. Why is Ashby then using an outside “parameter” to change the behaviour of a system?
    3. Raskin is mistaken. A toaster or washing machine are not “informationally closed” and even less “autopoietic” systems (cannot maintain themselves). They are not even “active” systems (as per Rosenblueth/Wiener/Bigelow classification) because they can’t store energy for an “independent” action.
    4. I propose the “living system” is both allo- and auto-poietic. The allopoietic part is harvesting resources from the environment needed for the maintenance of the “praxis of living” and is producing a behaviour + waste that end up changing the environment. The “neural (sub)system” may be said to be autopoietic though. In this view, the “structural coupling” is also internal, between the auto and allopoietic parts of the system, and not between the “living system” as a whole and its environment.


    • Thank you Yosip for the thoughtful comment. My thoughts are below:
      Do keep in mind that Ashby’s law of requisite variety is based on Shannon’s tenth theorem. He also uses variety to mean the statistical entropy, again possibly a reference to Shannon’s informational entropy. We have discussed the Shannon’s transducer in our PMs on Twitter. 👍
      I believe Ashby is talking about trivial machines here. The set of possibilities is much smaller here and a parameter or an input to a trivial machine results in specific outputs. Unless, the machine is broken, they have to follow their structure. The user is able to use parameters to get specific outputs. I do not see a confusion here. To expand further – the user can introduce constraints so that user-specific outcomes results. For example, this could be an appropriate setting on the toaster. This would be Ashby’s LRV.
      A machine follows its structure, unless it is broken. Ashby said – when a machine breaks, it changes its mind. I think, for the purpose as an example, the toaster and washing machine provide clarity regarding being information tight. Ashby noted that if information is to pass from outside to inside, then it will be a significant event. This could be upgrading a toaster to also play songs while it toasted – a change in the structure.
      It is interesting to see behavior as a product being produced (allopoeitic).


      • Sorry for the delay. Didn’t check the “notify” box 😦
        The main problem I have with Ashby’s reasoning is that he does not allow for a system to change its behaviour in any other way except through a parameter supplied by another “coupled” machine. The original machine can go through a predefined state trajectory by itself (without any input), but if you want another trajectory (behaviour) you must supply a parameter. He is using such reasoning even when discussing self-organization, so I think this is his general stance, not only for simple (trivial) systems. And for me, you can’t use a parameter to control a “information tight” system.
        Shannon instead, in his description of the transducer (with memory), which I use as a generic model for dynamical systems, defines the next state of the system as a function of *two* variables (the input and the current state). Ashby explicitly discards this idea, and I still have a hard time to understand why, because, as I think I’ve show in some experiments with this model, when you combine Shannon’s transducer with Ashby’s notion of “transformation” you end with very “rich” behaviours even in systems with a limited number of variables.
        I wrote extensively about this in various places and a decent overview can be found here:
        I would appreciate any comments from you, or anyone who might have some ideas how to solve this conundrum.


  2. Actually, I think living beings – which includes you and me – do have accurate representations of what we’re calling reality (out there). I’ve got several reasonable explanations for this:

    1. Because we’re using any given (this is why we’re calling data – Latin for given – “given”) reality as maps or models of that (our) reality. Your body shape (or form), following the Law of the Requisite Variety, Ashby’s Law., has been and is being shaped (or in-formed, see why we’re using the word information) by that reality.
    Take for instance the way your sitting on your chair. A chair has been shaped to accommodate your sitting body and your body adapts to the shape or form of the chair (or the gate, when you’re a-sitting-on-a-gate

    2. Using one’s real reality as a constructed reality, you don’t need to “remember” or memorize maps or models to refer to that reality. It save memory space. Remember that Ashby himself wrote that we use “memory” as a explanatory principle for observed behaviour.
    Our brains don’t work like a computer or machine. We don’t store data. There is no “place” in our brain that remembers; no such number, no such zone, address unknown (

    3. As reality is always present, it’s always “ready-at-hand”, one doesn’t need to refer – literally, “re-make” – to reality and it can be used tacitly. (Didn’t I say reality is given, like a present,

    4. As all systems are systems within systems, the very concept of “open” and “closed” or “inside” and “outside” can be considered a matter of choice. Making a distinction is making a choice. These concepts can be used as verbs: closing and opening require making distinctions between inside and outside. You cannot open something without closing and vice versa.

    Also, I treat open and closed as complementary concepts that do not necessarily deny each other. They’re no absolutes, but relatives. Open doesn’t need to imply “not closed”. Like you can be helped in a shop, while it’s “closed”.

    5. One can only construct (subjective) reality when one has been constructed in (objective) reality. As I once read, we need reality to distinguish between dreaming and being awake. Reality, in a way, realizes itself (I’m aware that I’m stretching the meaning of words to their limit) and reality realizes (y)our-self.

    We’re models modelling models of reality. What we’re calling time, prevents these thinging (making things in thinking, sometimes one has to invent words) from happening all at once. The word time we derived from “di”, as in “di-vide”.

    6. Structural coupling (con)structurally couples constructions. Constructing a reality, one needs structures to construct. Self-structuring structures structure themselves, while constructing “reality”. Organisms organically organize organs. The word “organ” has been derived from the Greek “erg” or “work”. En-ergy is at-work. Coupling couples couple couples. Models modelling models.

    “Language inexorably forces us to present everything as a sequence” – Von Glaserfeld. In using language, one uses order in constructing meaningful sentences. Language contains an implicit request (or order) to comply, to comply with “the group”. Language has been forged to control the behaviour of each other.

    It’s our use of language that induces in us the feeling that we – if we understand reality – can control reality. We’re still using words as (magical) spells. Because I say so.

    This order however can not be attributed to reality: reality structures itself “unordered”. And I mean, by not being ordered, but by itself.

    Our machines have been structured in a certain order by us, with instructions. Our bodies instruct themselves. Our DNA doesn’t instruct anything. Proteins construct each other without instructions. Cells don’t instruct each other.

    7. Our sensing actively reconstructs reality from it’s “disturbances”. Moving creatures need perceiving (sensing and intuiting) and judging (feeling and thinking) to prevent bumping into walls, finding food and shelter. You perceive what you need to perceive and – I’m still working on this idea – not the objects themselves, but their usefulness.

    8. I’m a radical constructivist, because constructing is the root (radix) of life. Constructing through autopoiesis. (See “The Invented Reality, edited by Paul Watzlawick). I also recommend the chapter by Paul on Components of Ideological “Realities” (quotes by Paul).

    Actually, by closing one’s mind, one looses contact with reality and reverts to one’s own inaccurate representation of reality. The next step being to impose this on others. Then one constructs hell for other people. (


    • Thank you Janlelie. My thinking is that we only have representations that work or fit the reality. We only cope with the reality. We lack the variety to construct an accurate reality. I have noted here before that our organizational viability depends upon consistency and not completeness. The structural coupling is only to certain facets of reality and not the entire reality as a whole.

      I am not sure if by accuracy, you are meaning a true replica of the external world. As a constructivist, I would imagine this is not possible in your worldview. Can you clarify?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s