The ontological gap between entities and interaction

Another theme which feels important to me in Pickering’s superb The Cybernetic Brain is the ontological gap between entities and interaction. If we imagine the world as composed of discrete entities with defined characteristics, it invites an approach to knowledge in which we merely place them into a taxonomy in a manner which leaves them in principle knowable in full.

If I understand the cybernetic impulse correctly, it rests on the aforementioned ontological gap. Even if we learn stuff through this approach, it tells us little about the interaction of these entities and obscures much that is important about the world in which this interaction occurs. What matters is how these discrete elements enter into interaction with each other, in a manner which is inherently unpredictable and cannot be discerned through the decompositional and representational approach to knowledge previously described. This is why we should start with the performance i.e. the reality of their interaction.

It’s possible I’m translating this too much into the conceptual idiom of morphogenetic theory and perhaps missing something of cybernetics in the process. But I’m finding it a stimulating activity nonetheless, something which I think the cyberneticians would have approved of.

Posted In:

6 Comments

  1. The cybernetic view does challenge the discrete object or entity – it doesn’t reject these ideas entirely, it treats the entity as ‘outputs’ of a dynamic system. What cybernetics opens up is the possibility of exploring the behaviour of dynamics systems and the interaction between different systems, rather than just looking at the possible relationships between outputs. I guess post-structuralism was of a similar mind, but was a critique rather than an alternative philosophical project.

  2. a dynamic system is defined by its capacity to equilibrate or achieve equilibrium. That involves interaction with other systems. But an entity is the characterisation of the outputs of a system in any moment or over a period of time when it is in equilibrium. The life phases of a butterfly or the phases of matter (gas, solid, liquid). What we sense of the butterfly is its output (of itself as a dynamic system) and that’s what make it a thing, an entity.

  3. I think you’re onto something here that has been explored in different ways. For example Von Foerster’s second-order cybernetics, the notion of applying cybernetics to itself.
    https://en.wikipedia.org/wiki/Second-order_cybernetics

    In my opinion, though, this “go meta” approach complicates the structural model without actually helping the “ontological gap” problem. Instead it reifies the construct of the “observer” as separate from the “model”, reproducing the gap and leaving agency and interaction as metaphysically mysterious as ever.

    The crux of the problem lies not in modeling, but in metaphysics, or should I say, the metaphysics of modeling. Cybernetics is essentially a theory of modeling the world descriptively, but in a way that the modeled parameters are dynamic, but not the model itself. Modeling is construed via a paradigm of description (vs observation) that has the “gap” baked into it from the beginning. Adding another observer level that can alter the model does not change the quality of descriptivism to which the model approach is committed.

    There are two metaphysical assumptions in the cybernetic approach which are mostly unexamined, and that replicate broader western cultural metaphysics. The first is the notion of static ontology, the idea that the values might change but the ontology is static, replicating the theological construct of the “eternal”. Thus in cybernetics the construct of a “system” is defined by equilibrium, the posited capacity to stay the same. There’s no (or not much) theorizing or modeling process and change, because the static version seems adequate given the metaphysical ethos. The second is the separate transcendental position of the “observer”, which even in modernity and cybernetics is derived from the construct of the transcendent universal observer (in one version, the “god” that created everything and now stands back as a mostly-passive observer). This same construct will be encountered in any metaphysically-informed social ontology of agency and interaction. And maybe I don’t need to point out that this “gap” is also the explanatory gap of the so-called “hard problem” of consciousness.

    If there is a way for the fly to escape the metaphorical / metaphysical bottle, it might be found in process metaphysics, which are unfortunately so counter to the broad cultural ethos that it has not been retrieved from the more mystical aspects of Whitehead’s thought and brought into a naturalized metaphysics. But there is some action in this area, for example, Isabelle Stengers _Thinking with Whitehead_
    https://books.google.com/books/about/Thinking_with_Whitehead.html?id=w_KRoAEACAAJ&source=kp_book_description

    Another more practical site of activity for this “leading edge” concerns the current critique and effort to understand the limitations and cultural effects (and ultimately, metaphysics) of our use of artificial “intelligence” and the nature of its statistical modeling approaches. Philosophically they are in some ways a step backward (e.g. away from rule-based modeling), unless one is prepared to go “all in” with pragmatism and dispense with the rule and “law” constructs in favor of data-driven models. But one could argue that doing so simply conceals the causal forces and regularities of the human, agentic, interacting, relational world behind a veneer of measured and sampled behaviour, which itself is born within the human world of agentic and metaphysically-saturated interaction.

    One representative example in this area is Sridhar Mahadevan’s article “Imagination Machines: A New Challenge for Artificial Intelligence. He does not, however, attempt to engage metaphysics but mainly seeks to improve the mathematics of modeling without questioning its basis. That is, the tacit attempt in this and related work (e.g. Gardenfor’s Conceptual Spaces) is a somewhat philosophically naive attempt to “ground” the modeling in better mathematical constructs. This carries the risk of creating more powerful pseudo-reality models whose obscuring power is proportional to their power to create compelling simulacra, reasonable-seeming concepts, and practical tools that reproduce specific cultural practices. I think you (Mark) alluded to something similar using the term “ontological veiling”.

    https://markcarrigan.net/2019/03/22/ontological-veiling/

    https://people.cs.umass.edu/~mahadeva/papers/aaai2018-imagination.pdf

    https://books.google.com/books?id=FSLFjw1EcBwC&printsec=frontcover#v=onepage&q

    1. Links are very interesting thanks. In terms of cybernetics though, surely what you’re saying a true of first wave but not of subsequent approaches?

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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

This site uses Akismet to reduce spam. Learn how your comment data is processed.