In the previous post, relational ontology was positioned against more familiar scientific and philosophical frameworks by shifting a single assumption:
stability is not primitive; it is an effect that must be accounted for.
That move is easy to state and harder to sustain. If objects are not taken as given, and relations are not arbitrary, then a question immediately follows:
what does the work of stabilisation?
Without an answer, relational ontology risks dissolving into a general claim that “everything is relational,” which explains very little. What is needed is not a return to substance, but a more precise account of constraint.
1. Constraint is not added; it is constitutive
A common way of thinking about constraint is as something imposed on an otherwise free system. On that view, relations come first, and constraints limit their possible configurations.
Relational ontology cannot take that route. If constraint were external in this sense, we would already have presupposed a substrate capable of unconstrained variation. The problem would simply have been displaced.
Instead, constraint must be understood as intrinsic to relational differentiation itself. There is no prior field of unconstrained relations. There are only relations as they are differentiated under specific conditions.
This leads to a tighter formulation:
to relate is already to be constrained.
Constraint is not what restricts relation; it is what makes relation determinate enough to be anything at all.
2. Stability as sustained constraint-pattern
If constraint is intrinsic, then stability can be re-described without invoking substances or underlying carriers.
An “entity” holds—not because it possesses an invariant essence—but because certain patterns of constraint are sustained across variation.
This shifts the explanatory burden:
- from identifying what a thing is,
- to tracking how a pattern of relations remains sufficiently invariant to be treated as the “same” across different instances.
Identity, in this sense, is not a fixed property. It is a continuity of constraint-pattern.
This also explains why stability is always partial:
- constraint-patterns can weaken, shift, or fail
- identity can blur, fragment, or dissolve
An instance persists only so long as its constraints continue to hold.
3. Stratification: constraint is not uniform
Constraint does not operate at a single level.
One of the advantages of drawing on a stratified model of systems (as in systemic functional linguistics) is that it allows us to recognise that:
- different strata impose different kinds of constraint,
- and stability often depends on coordination across them.
For example, what counts as a stable “entity” in a biological system is not reducible to:
- its material composition alone,
- nor to its functional organisation alone,
- nor to its semiotic description alone.
It emerges from the alignment of constraints across multiple strata.
This matters because it prevents a collapse into flat relationalism. Not all relations are equivalent, and not all constraints operate in the same way.
Stability is therefore not local. It is:
a cross-stratal achievement.
4. Construal and instantiation
At this point, construal can be reintroduced without collapsing into subjectivism.
Construal is not the imposition of meaning onto a pre-given reality. It is one of the ways in which relational differentiation is actualised under constraint.
An instance is always:
- a configuration of relations,
- under constraint,
- as construed within a system.
This does not make stability arbitrary. On the contrary, construal itself is constrained:
- by the system within which it operates,
- by the resources it can deploy,
- and by the patterns it must sustain to remain intelligible.
So while construal is constitutive, it is not free.
5. Failure conditions
An account of stability is incomplete without an account of its breakdown.
If an entity is a sustained constraint-pattern, then failure occurs when:
- the relevant constraints can no longer be maintained,
- or when their coordination across strata collapses.
This can take different forms:
- gradual drift (loss of coherence over time),
- abrupt disruption (constraint breakdown under stress),
- or reconfiguration (transition to a different stabilised pattern).
What matters is that:
identity does not simply disappear; it ceases to be sustained.
This provides a way of distinguishing between:
- superficial variation (which preserves the pattern), and
- structural change (which transforms or dissolves it).
6. No return to substance
At this point, it may be tempting to reintroduce some form of underlying substrate to “carry” these patterns of constraint.
Relational ontology resists this move—not because substrates are meaningless, but because they are themselves better understood as stabilised patterns within a broader relational field.
What appears as a substance at one level of description is often:
- an effect of constraint at another.
So the aim is not to deny the usefulness of substance-like descriptions, but to avoid treating them as ontologically primitive.
Closing
Relational ontology does not eliminate stability; it relocates it.
What appears as an entity is a pattern of relations that has become sufficiently constrained, and sufficiently sustained, to function as a unit of identity within a system.
The question is no longer:
what is this thing?
But:
what constraints must hold for this to continue to count as the same thing?
That shift replaces substance with stabilisation, and essence with sustained constraint.
Whether that replacement is adequate is not decided in advance. It is tested wherever patterns must hold under pressure.
No comments:
Post a Comment