The Life & Mind Seminar Network

Autopoiesis and closure

Posted in General by Tom Froese on February 21, 2007

This is just a comment on something that occurred to me after today’s seminar. While I was of the opinion then (and previously) that operational closure and organizational closure are probably just two different concepts describing exactly the same phenomena, I’ve now changed my mind. I now see the crucial concepts as follows:

autopoiesis: a network of processes that produces itself in the physical domain

organizational closure: a network of processes that produces itself in any domain

operational closure: a network of processes that is characterized by self-reference

If we ignore the additional requirement that autopoiesis is limited to the physical domain, then it turns out that actually autopoiesis and organizational closure are one and the same. This could explain why the latter notion has hardly been used in recent literature, especially since it is not at all clear what the role of physicality in all of this exactly is.

Accordingly, the notion of operational closure appears more of a particular kind of non-representationalist epistemological stance which an observer can choose to adopt when studying a system’s behavior.

This is the case, for example, when we want to explain the operations of a system that is characterized by organizational closure (what Varela calls an “autonomous system”), and we do not want to make reference to any knowledge of the observer (because only the observer has epistemic access to both the system and the environment to which it is structurally coupled), but only to the system itself.

Of course, this is not to say that we cannot also study such a system in representationalist terms by distinguishing it as having inputs and outputs which somehow correspond to its environment, but we have to be clear that in such a case we make use of distinctions which do not pertain to the cognitive domain of the system itself.

Does this seem to be a reasonable way of interpreting these notions?



4 Responses

Subscribe to comments with RSS.

  1. Mike Beaton said, on February 22, 2007 at 1:28 pm

    I know that my discussions these days reflect some input from Tom. Perhaps Tom’s reflect (some) input from me.

    So now, if we are going to be a clean representationalist (i.e. being clear that the distinctions we use in such talk are ours, not the system’s) then what is still to be gained by refusing to use this talk?

    Don’t we make ourselves literally unable to answer the questions we want to answer (e.g. how does the system know to build its nest there, not there? how does it know to decorate its nest with blue, shiny things? how flexible is it, in carrying out these ‘plans’?) unless we are prepared to include terms from our own knowledge (understood as such) in our explanation?


  2. tomfroese said, on February 23, 2007 at 10:24 am

    I think that avoiding to make reference to any kind of contextual knowledge in our explanations, even though we have access to it as observers external to the system under study, certainly makes it more difficult to answer those questions – but I do not think that it makes it impossible in principle.

    Functional notions such as representations do have their uses, of course. In many cases the inclusion of such contextual knowledge in our explanations can be seen as a useful heuristic and a temporary shortcut to a fully fleshed out operational description.

    But once we accept that a system is autonomous (in Varela’s sense), and we want to give an explanation of, for example, how that system *as an autonomous system* is able to distinguish blue, shiny things, then it is just not enough to refer to the fact that we know that there are blue, shiny things in its environment.

    This is because an autonomous system is intrinsically unable to distinguish whether a particular perturbation has been triggered by an external cause or has been generated by internal activity.


  3. Mike Beaton said, on February 23, 2007 at 12:47 pm

    My point is that we can’t even *ask the question* if we refuse to use the terms from our knowledge.

    As I think you know, I am broadly sympathetic to your approach to getting the right answer – as I understand it.


  4. tomfroese said, on February 23, 2007 at 1:29 pm

    Sure, we need to be able to refer to both an organism and its environment in order to be able to ask how these two systems are coupled together. And we also agree that this correspondence link is provided by the observer and cannot feature in an operational explanation of the autonomous system’s behavior.


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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

%d bloggers like this: