In enactivist theory, one of the central explanatory notions is structural coupling.
Associated with the work of Francisco Varela and further developed by Evan Thompson and Alva Noë, structural coupling is typically invoked to describe the ongoing history of mutual influence between an organism and its environment.
At a glance, the idea appears straightforward:
- an organism and its environment repeatedly interact,
- each perturbing the other,
- resulting in coordinated, co-evolving structures.
This “coupling” is often presented as the mechanism by which cognition becomes embodied and situated.
However, when placed under constraint, the assumptions supporting this mechanism require closer examination.
1. What Structural Coupling Presupposes
In its standard formulation, structural coupling involves:
- distinct systems (organism and environment),
- a shared domain in which interaction occurs,
- temporal continuity allowing interactions to accumulate as a history,
- and reciprocal perturbation between systems over time.
Even when enactivism explicitly rejects internal/external dualism, it retains:
- a distinction between coupled systems,
- and a medium in which their coupling is realised.
Structural coupling is thus not merely a descriptive term; it carries implicit ontological commitments.
2. The Constraint: No Shared Domain
Under a relational ontology organised around the logic of the cut, the following constraints apply:
- There is no pre-given domain within which entities interact.
- There is no substrate that hosts interaction.
- There is no temporally unfolding process that connects independent systems.
- Distinctions between “organism” and “environment” are not assumed in advance.
This does not eliminate distinction.
It reframes it:
distinctions are constituted under conditions of constraint, rather than presupposed as elements within a shared field.
3. Reinterpreting “Coupling”
If there is no shared domain, then “coupling” cannot denote:
- interaction within a medium,
- causal influence between independently existing systems,
- or a process linking entities across time.
Instead, the term must be re-evaluated.
What is described as coupling must be understood as:
the non-independence of determinations that are distinguishable under the cut.
This shifts the role of coupling from:
- a mechanism of interaction,
to:
- a description of relational dependence between distinctions.
Coupling does not connect pre-existing entities.
It expresses that what is distinguished as “organism” and “environment” are not independently grounded.
4. The Problem of the Interaction Medium
Structural coupling, in its standard usage, presupposes a medium:
- a physical, biological, or environmental substrate in which interactions occur.
Examples typically include:
- chemical exchanges,
- sensory-motor coordination,
- physical perturbations.
Under constraint, the notion of a shared medium becomes problematic.
A medium implies:
- something that contains both interacting entities,
- and across which interaction is transmitted.
But if no such containing domain is available, then:
interaction cannot be located within a shared environment.
This does not deny that correlations, coordination, or dependencies can be described.
It denies that these are grounded in a pre-existing field that hosts interaction.
5. Temporal Accumulation and the History of Coupling
Structural coupling is often characterised as historical:
- interactions accumulate over time,
- leading to stable patterns of coordination,
- which in turn shape future interactions.
This introduces a temporal narrative in which coupling is something that develops.
Under constraint, temporal continuity cannot be treated as a primitive explanatory layer.
Instead:
- “history” is a way of construing stability across distinctions,
- not a process that underwrites their formation.
Thus, structural coupling cannot rely on:
- a sequence of past interactions building toward present organisation.
Rather:
what is described as a “history of coupling” must itself be reinterpreted as a pattern of co-variation observed under a given construal.
6. From Interaction to Co-Determination
If we remove:
- shared domain,
- substrate,
- and temporal process as explanatory primitives,
then structural coupling can no longer be understood as interaction between independent systems.
What remains is a more minimal formulation:
distinctions such as “organism” and “environment” are not independently grounded, but are co-determined under the conditions that allow them to be distinguished at all.
In this sense:
- “organism” and “environment” are not two entities that become coupled,
- but two poles of a distinction that only holds under constraint.
Coupling, then, does not join them.
It expresses their non-independence.
7. The Residual Import: Hidden Domain Assumptions
Even when enactivist accounts reject Cartesian dualism, they often retain a weaker assumption:
- that organism and environment are already embedded within a common ontological field.
This assumption is rarely stated explicitly, but it becomes visible in phrases such as:
- “organism–environment interaction”
- “embedded in an environment”
- “situated in a world”
Each of these expressions subtly reintroduces a shared domain in which interaction is meaningful.
Under constraint, this domain cannot be taken as given.
It must itself be understood as:
a product of the conditions under which distinctions are made, rather than a pre-existing container for those distinctions.
8. Reframing Structural Coupling
We can now restate structural coupling in a way that remains coherent under constraint:
- It does not describe interaction within a shared environment.
- It does not describe a process unfolding between independent systems.
- It does not presuppose a medium in which coordination occurs.
Instead:
structural coupling describes the relational dependence between distinctions that are co-constituted under the same conditions of constraint.
This reframing preserves the intuition that organism and environment are not independent.
But it removes the idea that their relationship is mediated by a common domain.
9. What Has Been Removed—and Why It Matters
Under this reading, three key components of standard enactivist explanation are no longer available as primitives:
- The shared domain in which coupling occurs
- The temporal process through which coupling unfolds
- The independent entities that enter into coupling
What remains is not interaction, but co-determination under constraint.
This is not a weakening of the account.
It is a reparameterisation of what counts as explanation.
Closing Orientation
Structural coupling, as ordinarily understood, is a mechanism that explains how organism and environment influence one another within a shared world.
Under constraint, that mechanism cannot be retained in its original form.
What survives is not a model of interaction, but a description of dependence:
the distinction between organism and environment does not precede their relation; rather, it is sustained through the very conditions that allow such a distinction to be made.
In the next post, we can take the same approach to autopoiesis, which will force an even sharper question:
what does “self-production” mean when there is no substrate in which a system produces itself?
No comments:
Post a Comment