I forgot to respond to some of Reid’s questions about system-references in my last post. I don’t know whether Reid has been following my posts for the last few months, but I argue that objects are essentially systems. Following Maturana and Varela (though my major points of reference are Bateson, von Foerster, and especially Luhmann), I distinguish between autopoietic systems and allopoietic systems. An autopoietic machine, Maturana and Varela argue,
is a machine organized (defined as a unity) as a network of processes of production (transformation and destruction) of components that produce the 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 a space in which they (the components) exist by specifying the topological domain of its realization as such a network. (Autopoiesis and Cognition, 78 – 79)
Translated into English, autopoietic systems are systems that produce their own components through their own components. These systems roughly compose the domain of the living and the social, though there might be other autopoietic systems as well. By contrast, allopoietic systems are systems that are produced by something else. These systems are roughly the domain of the inanimate.
The key feature of autopoietic systems is that they are operationally closed. Operational closure refers to two things: First, it refers to the manner in which the operations of an autopoietic system only ever refer to and relate to themselves. For example, communication only ever refers to other communications. Second, it refers to the way in which a system relates to an environment. Systems do not directly relate to their environment, they do not receive information from their environment, but rather they constitute their own openness to an environment. A system can be perturbed or irritated by its environment, but the information value that this perturbation takes on is not something that was already there in the environment, but is rather constituted by the organization of the system itself. Put crudely, what counts as information for a frog can count as nothing for me, and supposing that a frog and I are perturbed by the same something in the environment, we can nonetheless produce entirely different information out of that perturbation. As such, information is always system-specific. I differ markedly from Maturana and Varela and argue that this second sense of closure (selective relations to an environment), is not unique to autopoietic systems, but is true of all objects, whether autopoietic or allopoietic (though information functions in very different ways in each case). One of the most important points here is that information is not something transmitted or exchanged between systems (sometimes we think of communication as the transmission of information that remains the same for sender and receiver). Information is system-specific and does not exist independent of the system in which it occurs. Or as Lacan (and Luhmann) liked to say, all communication is miscommunication.
Hopefully this is enough to give Reid a sense of what I’m talking about when I talk about “system-specificity” or “system-references”. The point is, that whenever we make claims we need to specify the system to which these claims pertain. We can’t generalize across systems because each system has its own internal organization and therefore relates to the world in its own specific way. As I suggested above following Luhmann, societies are themselves autopoietic systems. If this is true, we can’t make the sort of universalistic, a priori claims that a lot of transcendental philosophy would like to make. Rather, we have to analyze social structures on a case by case basis to determine 1) how they are organized and thus how they produce meaning events, 2) the specific way in which they’re open to the environment, and 3) how they evolved or developed the particular distinctions that regulate their own internal processes and relation to the environment. If this is true, certain forms of transcendental philosophy have to be excluded a priori because they illicitly generalize over very different cognitive and social systems, working on the premise that they all share the same internal structure or organization. I see this as thoroughly consistent with Marx’s understanding of values. Marx was always careful 1) to analyze the emergence of specific values in terms of particular forms of social organization, and 2) to emphasize the historical situatedness of particular values in particular social organizations. In this regard, values for Marx aren’t merely “instantiated” in particular material conditions as Reid seems to suggest, but rather are products or inventions of particular social forms not unlike evolution is the invention of new species and forms of life.
In order to discuss systems we have to engage in second-order observation, observing how other systems observe their environment, rather than working naively from the premise that we observe the world in the same way. In a number of respects, this is precisely the problem with more traditional transcendental approaches. Although they attempt to self-referentially take the organization of the observer into account by analyzing the transcendental structure of mind, they nonetheless don’t take the additional self-referential step of recognizing that they observe differently than other systems and therefore end up illicitly generalizing one transcendental structure to all subjects, rather than recognizing that the world is populated by an infinite plurality of transcendental structures not unlike Leibniz’s monads. The problem, then, is that while there might be an “a priori” (note the square quotes), this “a priori” is always system-specific and can’t be generalized across systems. And since autopoietic systems are evolving systems that each have a contingent history and a contingent organization, we can’t generalize a priori structures across cognitive systems or social systems, but have to look at systems in their specificity like good Lacanians who recognize that there’s no general structure of mind or good neurologists who recognize that each brain develops differently.
I’ve written about autopoiesis quite a bit lately, and have been writing about Luhmann for years. In my view, one of the major failings of contemporary social and political thought is that it fails to take into account the operational closure of systems and therefore doesn’t even raise the question of how to communicate with a social system to change it when that system is closed by virtue of being organized by its own distinctions. This question has been one of the oldest and longest running themes on this blog. Moreover, I’m surprised that more Lacanians haven’t raised similar questions given Lacan’s theory of interpretation and the challenges facing interpretation or the analytic act when dealing with an autopoietically closed analysand. At any rate, if Reid is interested he can read more about systems here, here, here, and here, or he can do a search for Luhmann on this blog.
June 25, 2010 at 8:32 pm
Great post Levi. I have a question though. Many ecological theorists such as Gibson talk about animals “picking up” information from the environment e.g. picking up information about the ground texture that is relevant to locomotion or picking up information about a female’s ovulation. I am wondering how to reconcile this notion of “information pickup” with the idea of operational closure wherein information is not “transmitted” from environment to organism as if across a telephone line.
But perhaps I already have the answer in mind. Gibson sometimes substituted “resonance” for “pickup”, implying that organisms don’t “process” or “receive” information but rather, the nervous system “resonates” to it in terms of life-sustaining behaviors. Do you think Varela and Matura’s notion of “perturbation” is compatible with the idea of “information resonance”? In other words, how do you reconcile the idea of operational closure with the idea that “ecological” information exists in the environment and is indeed picked up by animals to secure their survival?
June 25, 2010 at 8:50 pm
Thanks Gary,
I’m not familiar with Gibson’s work so it’s hard for me to say. A lot of this stuff is really close to von Uexkull’s innenwelt/umwelt schema. Keeping in mind that I’m not familiar with Gibson’s work, we could say that he’s speaking loosely when he talks about systems picking up information from the environment. Systems can be perturbed by the environment, but there’s no information in the environment. Rather, perturbations have to be transformed into information by the organization of the system in question. The problem is that the term “environment” is ambiguous in these discussions. When we talk about the environment are we talking about other entities actually out there in the world, or about the umwelt of a system? The two aren’t the same. An umwelt is not “out there”, but is on the interior of a system or organism. Maybe one way to articulate this difference would be to hijack Heidegger’s distinction between world and earth (and here I’m not sure whether this is what he had in mind by the distinction). The world would be the umwelt of a system, whereas the earth would be actually existing entities (other systems) apart from a system regardless of whether a system relates to them or not. Does that make any sense?
Luhmann goes in a similar direction to what you’re talking about with “resonance” (in Ecological Communication). Alternatively, we could talk in terms of Maturana and Varela’s “structural coupling”. In all these cases, what you get are mutually closed systems related to one another in such a way that they mutually perturb one another while producing very different information internal to themselves as a result of these perturbations. In this way, you can begin to build up all sorts of ecological dependencies where very different closed systems co-evolve together without determining one another. In the chapter I’m working on now, I want to pick up on concepts like resonance and structural coupling and expand on them to articulate the role of material constraints in the development of systems (autopoietic theory is often criticized for ignoring these). So even though, for example, I’m independent of other objects in that I can be separated from them or cease to relate to them, materially I am nonetheless coupled to all sorts of other objects that play an important role in how I develop. For example, my relationship to the planet Earth played in important role in the development of my height. It didn’t determine my height (because resonance is non-linear and doesn’t allow for a one cause one effect model), but it did, and continues to, play an important role in my height. Had I been born on Mars I would have developed in a very different way because the gravity there is about half that of the Earth’s. Likewise, I think similar points can be made about lung capacities of Peruvians that live in the Andes. From what I understand they have impressive stamina despite living at these altitudes, such that they can put NBA players to shame in their ability to move about when highly athletic people from lower altitudes come to visit. To what degree is this Peruvian lung capacity in the Andes the result of resonance or structural coupling resulting from development in that particular high altitude environment (i.e., to what degree is it not a matter of genetics)? Hopefully some of this makes some sense.
June 25, 2010 at 8:56 pm
Gary,
As a further follow up point to your final question, I think it’s crucially important to emphasize that the environment as such does not exist. What I mean is that we shouldn’t think of the environment as a container that organisms must adapt to, to survive. The reason for this would be that the environment consists of an infinity of differences. No organism relates to all of these differences. So there’s a very real sense in which organisms constitute or produce their environment by selecting those differences which, for them, will make a difference or be capable of making a difference. In other word, the information isn’t already there in the environment, but gets constituted as information. I wrote about this recently in relation to Timothy Morton’s work. Maybe the better way to think about the system/environment relation is in terms of experiments, where some of these experiments work out well and allow the organism to persist, and some are pretty flimsy.
June 25, 2010 at 10:15 pm
Levi, you need to do yourself a big favor and check out Gibson’s work. You might be interested in a edited collection of his essays entitled “Reasons for Realism”. His other two big works are The Senses Considered as Perceptual Systems and The Ecological Approach to Visual Perception. Both are landmark texts in “ecological” realism. In a way, you could say that Gibson was object-oriented because in his books on perception he spends half of his time talking about the objective layout and structure of the Earth as an umvelt. And for a psychologist, he was philosophically astute. He championed for direct nonnaive realism in epistemology and a realism in metaphysics. In fact, he thought that they both complimented each other.
When Gibson says that there is information “in” the environment, he is referring to ecological information rather than physical information. He makes a distinction between the “ecological” world perceived by organisms and the world as studied by the physical sciences. This is in line with the classic ecological and phenomenological work of the early 20th century.
Accordingly, Gibson was concerned with the ambient light that settles into the environment through the medium of air and reflects invariant features of the Earth which are relevant to the internal needs of the organism. Gibson points out that traditional approaches to perception on based on the idea of inadequate information on the retinal “image”, thus necessitating the internal construction of a subjective percept that “fills in the gaps” for the conscious mind. Gibson then argues that these theories are mistaken because they fail to consider the amount of information that is available by means of moving through the ambient array of light. Because the ambient light is nomothetically related to invariant structural determinations of the Earth, if we are able to move through that array, the pattern of transformation across the retina as a result of locomotion will be directly related to the actual invariants of the Earth and thus “specify” the Earth, obviating the need for a re-presentational construction. Accordingly, Gibson thinks the “problem” of depth perception and the “problem” of the retinal image are based on misconceptions concerning the nature of perceptual stimuli.
As for Heidegger’s distinction between the world and the Earth, I think you are perfectly right in your interpretation. I really don’t know how else to interpret that distinction in a consistent manner.
So there’s a very real sense in which organisms constitute or produce their environment by selecting those differences which, for them, will make a difference or be capable of making a difference.
I’ve heard of this concept referred to as “ecological niche construction”. You are of course right then to note that information is entity-specific insofar perturbation is always cashed out in terms of particular structural determinations changing in response to an interaction with the Earth. This point is particularly relevant to recent research on default mode activity during “resting states”. Brain researchers are starting to realize the basic autopoietic insight that brains are much more self-involved that previously believed. The old paradigm of task-specific attentional mechanisms is now being updated in light of our knowledge of intrinsic dynamic “resonant cell assemblies” and global functional connectivity. The lateral geniculate nucleus, for example, has 80% of its inputs coming in from other parts of the brain rather than the optic nerves. This paints a very different picture of the brain, one that is much more in line with Maturana and Varela’s notion of “operational closure”.
June 26, 2010 at 1:21 am
Fascinating! I’m looking forward to seeing this chapter! I did feel a bit in the wilderness when I was trying to write about this stuff.
Just a couple of things (I am travelling and don’t have references). Just thoughts, not winges.
I’m not sure if it makes much diff., but Mat wrote specifically on the ‘misuse of the notion of information in biological systems.’
It’s not just a matter of constituting what counts as info. – the term is misplaced.
Ap systems do not deal in ‘information’. The term starts to get used in the late 1950’s to describe bio systems
Also I think Maturana did make the distinction btwn a niche and an ‘environment’.
Again, as I you know autopoietic systems are thermodynamically open – and informationally closed.
Ultimately, they are ‘conservative/homeostatic’ systems: they change their structure to maintain their autopoietic organization.
I guess you could apply this to ‘capitalism’ and social systems?
Lastly, I see the umwelt as being a ‘relational interface’ rather than ‘on the interior’.
Deely is good on this in his valuable bk ‘New Beginnings’.
As you know Uexkull interpreted the umwelt in a kantian way as ‘subjective appearance’ (‘all reality is subjective appearance’), whereas semiotically it is suprasubjective and potentially intersubjective.
As Deely noted, any genuine understanding of the bees’ dance exits from the Kantian horizon. To really know ‘something’ of another umwelt…which brings us back to wholes and parts…
June 26, 2010 at 2:23 pm
Hello Levi,
*Intra*-systematic wise, is there communication understood as process of translation or, alternatively, something tantamount to communication understood as reception, or, can we say that there is no communication per say?
Will
June 26, 2010 at 2:29 pm
Will,
I’m not sure I understand your question. When you refer to intra-system relations are you referring to what takes place in a system?
June 26, 2010 at 2:40 pm
Apologies, yes I do mean that.
Will
June 26, 2010 at 2:45 pm
Will,
What are you asking about when you ask about reception? I ask because one of the key claims of Luhmann’s social systems theory is that social systems are composed entirely of communications. Put differently, human minds are in the environment (i.e., outside) social systems. Social systems aren’t composed of humans, nor are human minds elements of social systems. What communication in a social system produces is thus not reception but more communications. Insofar as an autopoietic system only exists by continuing its operations (it ceases to exist if these dissolve), the issue for social systems is how to produce subsequent communications.
June 26, 2010 at 3:22 pm
Signifier leads to other signifier etc…
However a certain pattern emerges i.e. a topography. You state in your post…
“As such, information is always system-specific.”
And then further….
“One of the most important points here is that information is not something transmitted or exchanged between systems (sometimes we think of communication as the transmission of information that remains the same for sender and receiver).”
Go back to Maturana and Varela…
“through their interactions and transformations continuously regenerate and realize the network of processes (relations) that produced them”
This is suggestive of an internal translation, my question stems from ascertaining the nature of the difference between internal and external translation as it relates to information. As external translation is dependent on the integrity of the different systems, i.e. that there must be at least two systems, when the perspective changes to an internal focus of the system obviously this does not hold true. So is communication in this last instance qualitatively different from the former?
Will
June 26, 2010 at 4:29 pm
I think I see what you’re getting at. Translation takes place as a result of the “distinctions” that different systems employ in relating to their environment. Communications, by contrast, are based on the same distinctions. There can, however, be somewhat intra-systemic translations. Systems can have subsystems that are organized around their own system/environment distinction and codes. For example, in social systems you get legal systems, political systems, economic systems, etc. All of these systems belong to our contemporary social system, but each one of them is in the environment of the other subsystems. The economic system can only communicate economically (according to the profit/loss code or distinction) and the legal system can only communicate legally (according to the legal/illegal code). The economic system can perturb or irritate the legal subsystem (and vice versa), but when the legal subsystem transforms the perturbation of the economic system into information it registers it in terms of the legal/illegal code, not the profit/loss code. It’s the codes/distinction/organization that determines whether or not something is translation, not the properties produced in the translation. Communication events are properties produced.
June 26, 2010 at 6:19 pm
Thank you,
The communication that is produced bears the mark of the quality of its production, but which is lost in translation as it is processed into information via a different quality.
A ‘lifeless’ substance corresponds to the withdrawn object, secondly, actualizing qualities which are defined and limited by the substance. Are subsystems exclusively arise as a result of contact with other systems?
Will
June 27, 2010 at 3:13 pm
[…] When we reject the centrality of the human within being, treating the human and social as two system-references among others rather than the ground of all others, what consequences follow for how we think the […]
July 10, 2010 at 9:28 am
[…] I think this is evident in the most recent exchange between Reid (here) and Levi (here and here), over how to interpret Marx’s philosophy, where it strikes me that Levi has missed the point […]