Cuts That Make Worlds: Practising Relational Analysis — 5 Constraint Without Reality

There is a moment in every relational account where the uncomfortable question surfaces:

If there is no independently given reality, what prevents analysis from becoming arbitrary?

It is a reasonable question. And it is usually answered too quickly—by smuggling reality back in under another name: “structure,” “system,” “invariance,” “truth conditions,” “material constraint,” or something similarly reassuring.

But that move is a retreat.

If construal is constitutive of phenomenon, then constraint cannot be grounded in an external real that sits outside it. The challenge is sharper than that: to account for constraint without reinstalling an independent reality as its guarantor.

---

The False Choice

The default framing offers two options:

1. Either there is a mind-independent reality that constrains our descriptions
2. Or anything goes, and analysis collapses into free invention

This is a bad binary. It depends on assuming that constraint must come from outside the analytic process—either from an external world or from subjective will.

But if phenomena are not pre-given objects, and if analysis is an act of construal that actualises them, then constraint must be located elsewhere.

Not outside construal.

Not prior to it.

But within the structure of construal itself.

---

Constraint as Stability of Selection

Constraint, in a relational register, is not the resistance of an external world. It is the stabilisation of patterns of selection within semantic potential.

Once a particular cut is made, not all subsequent moves remain available.

For example:

• if a phenomenon is construed as institutional discourse, certain semantic selections become coherent (authority, obligation, asymmetry)
• others become unstable or require distortion to maintain coherence

This is not because reality vetoes them.

It is because the initial construal has organised semantic potential in such a way that some continuations are viable and others are not.

Constraint emerges from internal coherence conditions, not external enforcement.

---

Why “Anything Goes” Never Actually Happens

The fear of arbitrariness assumes that, without external constraint, analysis would dissolve into unrestricted choice.

But this misunderstands how construal operates.

Once a cut is made, it is not freely revisable at every step without cost. Each move:

• commits subsequent distinctions
• narrows viable continuations
• stabilises certain patterns while excluding others

Incoherence does not appear as a violation of reality. It appears as a breakdown within the construal itself.

Some sequences of selection simply do not hold together.

Not because they are forbidden by the world, but because they fail to sustain a stable organisation of meaning.

---

Constraint Without Correspondence

It is crucial not to confuse constraint with correspondence.

Correspondence theory would say:

statements are constrained by how the world is

Here, constraint is instead:

patterns of meaning are constrained by the requirement that they remain internally coherent under successive selections

This shifts the locus entirely.

The question is no longer:

does this match reality?

But:

does this construal sustain itself as a viable organisation of semantic potential?

A construal that collapses is not falsified. It is non-sustaining.

---

The Role of Recurrence

One of the strongest forms of constraint is recurrence.

Certain construals stabilise because they:

• can be re-applied across instances
• generate predictable patterns of selection
• sustain coherence under variation

This is not external validation. It is internal durability.

A construal that works once but cannot be repeated under slightly different conditions is weak—not because reality rejects it, but because it lacks structural robustness.

Constraint, then, is partly a matter of reconstructability.

---

Where “Reality” Reappears (and What It Is Doing)

At this point, it becomes tempting to say: “but surely this just is reality under another description.”

That temptation should be resisted.

What is often called “reality” in these contexts is not an external tribunal. It is the stabilised outcome of long-term, cross-contextual construal alignment.

In other words:

• what resists us is not an external thing-in-itself
• but a highly sedimented structure of constrained, repeatable construals across domains

“Reality” is not what constrains construal. It is what emerges from constrained construals becoming stable across instantiation scales.

This reverses the usual direction.

---

Material Resistance Without Ontological Independence

But what about resistance? Surely some things do not yield to reinterpretation.

A wall remains a wall. A system fails. A body breaks. A constraint bites.

Yes—but the mistake is to treat this as evidence of an independent ontological layer exerting force on representation.

What is actually visible is:

• extremely robust, cross-contextually stabilised patterns of construal
• organised across many interacting systems of meaning and practice
• resistant to local reconfiguration because they are globally reinforced

Resistance, then, is not ontological veto. It is high-density constraint across coupled systems of construal.

The wall does not resist because it is “really there.” It resists because the network of construals that constitute it is extraordinarily stable.

---

Discipline Without External Guarantees

Once constraint is internalised in this way, analysis becomes more demanding, not less.

Because there is no external reality to appeal to as final arbiter, rigour must be maintained through:

• coherence across steps of construal
• stability under variation of cuts
• resistance to collapse under reorganisation of semantic selections
• and reproducibility of analytic effects under controlled shifts

This is discipline without transcendence.

No external guarantee. No final court of appeal.

Only the structural integrity of the construal itself.

---

What Constraint Actually Secures

Constraint, properly understood, does not secure access to reality.

It secures:

• the possibility of stable phenomena
• the repeatability of distinctions
• the persistence of analysable structure across instances

Without constraint, there is not freedom. There is simply no analysable field at all.

Constraint is not what limits meaning from outside.

It is what allows meaning to hold together long enough to be analysed at all.

---

Closing Shift

If we remove “reality” as an external guarantor, we do not remove constraint.

We relocate it.

From:

• world → representation

To:

• constrained construal → stabilised phenomenon

And once this shift is made, the question is no longer whether analysis corresponds to what is real.

It becomes:

what kinds of construal can sustain themselves as stable regions of meaning under variation, and what does that stability make visible?

That is where constraint lives.

Not outside the system.

But in the difficulty of keeping it together.

Cuts That Make Worlds: Practising Relational Analysis — 4 Misalignment as Data

There is a persistent temptation in analysis to treat disagreement as noise.

Two accounts diverge. Interpretations conflict. Readings fail to converge. The impulse is immediate: refine the framework, clarify the terms, resolve the ambiguity, eliminate the error.

Misalignment is taken as a problem to be solved.

But this presumes something quite specific: that there is a single object beneath the divergence, and that the task of analysis is to bring representations into alignment with it. Once that presumption is in place, misalignment can only ever appear as failure—either of perception, of method, or of articulation.

If, however, analysis begins from construal rather than representation, this entire framing shifts.

Misalignment is no longer a disturbance in the data.

It is data.

---

Divergence Is Not Deviation

When two analyses do not coincide, it is easy to assume they are competing attempts to describe the same phenomenon. One is right, the other less so; or both are partial, awaiting synthesis.

But this assumes a shared object that precedes both accounts.

If, instead, analysis is understood as the actualisation of a phenomenon through construal, then divergence cannot be reduced to partial access to a common ground.

It becomes something else entirely:

• different cuts through a field of potential
• different organisations of semantic resources
• different stabilisations of what counts as a unit in the first place

Misalignment, then, is not a failure to converge on a single object. It is the manifestation of multiple objects emerging under different construal conditions.

---

The Seduction of Resolution

The pressure to resolve misalignment is not accidental. It is built into most analytic habits.

Resolution promises:

• coherence
• closure
• communicability
• the comfort of a single account that “works”

But resolution often comes at a cost: the flattening of differences that were structurally informative.

When divergent analyses are forced into alignment, what is lost is not just nuance, but the visibility of the cuts that produced the divergence in the first place.

What disappears is the structure of construal itself.

---

Misalignment as a Diagnostic Surface

If misalignment is treated as data, it becomes a diagnostic surface for examining how construal operates.

Consider two analyses of the same interaction:

• one treats it as interpersonal negotiation
• another treats it as institutional enactment
• a third treats it as lexicogrammatical patterning

These are not simply perspectives layered over a shared object. They are different ways of stabilising what the object is.

Where they diverge most sharply is precisely where the analytic cut is doing the most work.

Misalignment therefore reveals:

• where boundaries have been drawn differently
• where semantic potential has been organised in incompatible ways
• where different registers have been implicitly enacted

It makes the structure of the analysis visible.

---

Non-Convergence as Structure

Not all misalignment is resolvable.

Some divergences persist not because of insufficient refinement, but because the construals involved are structurally incommensurable. They do not fail to meet; they organise different phenomena.

In such cases, seeking resolution is a category error.

The task is not to reconcile the accounts, but to map:

• what each construal makes visible
• what each excludes in order to do so
• and how each stabilises a different region of potential as “the phenomenon”

Non-convergence is not an analytic deficit. It is a structural feature of working within a field where phenomena are brought into being through cuts.

---

The Analyst Within Misalignment

Once misalignment is treated as data, the position of the analyst shifts.

The analyst is no longer the arbiter who decides which account is correct. Nor are they the synthesiser who fuses partial perspectives into a unified view.

They become something more difficult to stabilise:

• the site where multiple construals are held in relation without being collapsed
• the observer of how different analytic cuts produce different ontological effects
• the one who tracks the conditions under which divergence emerges and persists

This requires resisting the reflex to resolve.

Not because resolution is wrong, but because it prematurely ends the analysis.

---

From Error to Structure

What is typically labelled “error” in interpretation is often just misalignment between construal regimes.

A reading fails not because it misrepresents an object, but because it:

• operates with a different notion of what counts as relevant
• organises semantic potential differently
• stabilises a different boundary for the phenomenon

Seen this way, error is not a deviation from correctness. It is a signal of alternative structuring.

The question is no longer:

“Which reading is right?”

But:

“What must be true of the construal for this divergence to appear?”

---

Holding Divergence

To work with misalignment as data is to develop a particular discipline: the ability to hold incompatible construals without collapsing them into equivalence or hierarchy.

This is not relativism, because the differences are not arbitrary. They are structured by distinct analytic cuts, each with its own internal constraints.

Nor is it synthesis, because nothing is being merged.

It is closer to cartography than adjudication:

• mapping the terrain of construal differences
• tracing where and how phenomena diverge
• and making explicit the conditions under which each emerges

---

What Becomes Visible

Once misalignment is no longer treated as noise, a different picture of analysis emerges.

Instead of a single line of progressively refined interpretation, we see:

• a field of competing stabilisations
• each actualising different aspects of semantic potential
• each revealing what the others necessarily exclude

The object of analysis is no longer what lies beneath divergence.

It is the structure of divergence itself.

And in that structure, something else becomes visible:

not the failure of analysis to converge, but the way in which analysis—when properly understood—is always already the production of multiple, coexisting worlds of meaning.

Cuts That Make Worlds: Practising Relational Analysis — 3 Register as Analytic Lever

If the analytic cut brings a phenomenon into view, register is one of the most powerful ways of moving that cut.

But this requires precision. Register is often treated as a classificatory device: a way of sorting texts into types based on field, tenor, and mode. In that usage, it functions as a label applied after the object has been secured.

Here, it cannot be that.

Register is not something we assign to a pre-existing text. It is a way of construing semantic potential such that a text emerges as this kind of instance rather than another.

Used analytically, register does not describe the object. It reconfigures it.

---

Register on the Cline of Instantiation

Within a Hallidayan stratification, register occupies a precise position—but only if the strata and the directions of realisation are kept distinct.

From the pole of potential, register is a semantic subpotential: a patterned configuration of meaning resources.

From the pole of instance, that same region appears as a text type: a recurrent form of semantic selection in particular instances.

These are not two things. They are the same variation seen from different points along the cline of instantiation—within the semantic stratum.

The relation to context is one of realisation:

• contextual configuration (field, tenor, mode) is realised by semantics
• register is the semantic organisation through which that realisation occurs

To treat register as a fixed category—something a text has—is to freeze this movement. It turns a dynamic relation into a static property.

Analytically, that is a loss.

---

Register as a Way of Cutting

Recall: there is no unit until a cut is made.

Register provides one way of making that cut—not by drawing boundaries around a chunk of language, but by organising what counts as relevant within it.

To construe a text in terms of register is to foreground patterns of semantic selection that are congruent with particular contextual configurations:

• field: what is going on
• tenor: who is involved
• mode: how meaning is being exchanged

But these are not variables to be filled in, nor are they properties of the text. They are dimensions of context, realised through semantic organisation.

Shifting register is therefore not a matter of redescribing the same text more precisely. It is a matter of re-cutting the semantic potential through which the phenomenon is actualised.

---

A Simple Shift

Take a fragment of language—a brief exchange, a paragraph, a statement.

One construal might organise its semantic resources in ways congruent with:

• everyday conversation: low interpersonal distance, spoken mode, a field of casual coordination

Another might organise them in ways congruent with:

• institutional discourse: asymmetrical tenor, written-like mode, a field of formal decision-making

Nothing in the wording alone forces one construal over the other. The difference lies in how semantic potential is being organised relative to context.

And once the register shifts, everything else follows:

• what counts as appropriate or marked
• how choices are interpreted
• what patterns become visible

The “same text” does not persist beneath these shifts. Each construal actualises a different instance.

---

Register Is Not Added On

A common analytic move is to begin with a text, analyse its grammar and semantics, and then “add” register as contextual explanation.

This sequence presumes that the text is already there, fully formed, awaiting contextualisation.

But register is already operative in the cut that made the text analysable in the first place.

To analyse transitivity, mood, or theme is already to be working within a particular organisation of semantic potential—one that is congruent with a construal of field, tenor, and mode, whether this is made explicit or not.

Bringing register to the foreground does not enrich an existing analysis. It reorganises it from the ground up.

---

Leveraging Register

If register is treated as an analytic lever, the task shifts.

Instead of asking:

“What register does this text belong to?”

We ask:

“What happens to the phenomenon if semantic potential is organised in this way rather than that one—and how is that organisation related to context?”

This is an experimental move.

One can:

• hold the wording relatively stable
• shift the organisation of semantic resources
• and observe how the phenomenon reorganises

Certain patterns will stabilise under one construal and dissolve under another. What appeared coherent may fragment; what was invisible may become salient.

This is not error. It is the analytic work.

---

Constraint and Discipline

Not every shift in register is equally viable.

Some construals will fail to cohere:

• the semantic patterns will not align
• the distinctions will not sustain themselves
• the analysis will collapse into inconsistency

This is where rigour re-enters—not as adherence to procedure, but as discipline in managing the cut.

A viable register construal must:

• align with the semantic resources it brings into play
• sustain coherence across the analysis
• remain accountable to the system of meanings within which it operates

The lever can be pulled in many directions. Not all of them hold.

---

Register in Motion

To work with register analytically is not to assign a text to a box. It is to move along the cline of instantiation—within the semantic stratum:

• from subpotential (a configuration of semantic resources)
• to instance (a particular text as semantic selection)
• and back again

Movement between context and semantics, by contrast, is a movement across strata, governed by realisation. These must not be conflated.

Each movement reconfigures the phenomenon:

• the instance is seen as a selection from semantic potential
• the potential is reshaped in light of the instance

Analysis becomes a controlled oscillation.

Not between text and context as separate domains, but between different construals of how semantic potential is organised and realised.

---

What This Enables

With register as an analytic lever, the question shifts again.

No longer:

“What is this text?”

But:

“How is semantic potential being organised here, how is that organisation related to context, and what follows if it were organised differently?”

The analyst is no longer classifying.

They are reconfiguring the conditions under which the phenomenon appears.

And in doing so, they begin to see something that method, in its procedural form, cannot show:

that the stability of any analysis depends not on the object it describes, but on the coherence of the construal that brings that object into being.

Cuts That Make Worlds: Practising Relational Analysis — 2 The Analytic Cut

If there is no pre-given object of analysis, then the first question is unavoidable:

what, exactly, is being analysed?

The habitual answer—the text, the discourse, the interaction—only postpones the problem. For each of these already presumes that something has been delimited, stabilised, and rendered available as an object. The question is not what we call it. The question is how it came to appear as this rather than that.

This is the work of the analytic cut.

---

There Is No Unit Until There Is One

Analysis typically begins by identifying its unit:

a sentence, a clause, a conversation, a genre, a concept.

But units do not precede analysis. They are its first achievement.

What counts as “the same text” is not given by the world. It depends on where the boundaries are drawn:

• does a text include its title?
• its surrounding commentary?
• its circulation across contexts?
• its uptake by different readers?

Each decision produces a different object. Not a different view of the same object, but a different object altogether.

The analytic cut is precisely this decision—not as an arbitrary choice, but as a construal that actualises a particular configuration of potential.

Until the cut is made, there is no unit. There is only the possibility of one.

---

Cutting Is Not Carving

It is tempting to imagine the cut as a kind of carving: the analyst carefully tracing the natural joints of an already-formed entity.

But this metaphor smuggles in what must be refused—that the joints are already there.

The analytic cut does not reveal structure; it enacts it.

This does not mean that “anything can be cut anywhere.” Cuts are not free. They are constrained by the system of meanings within which they operate. Some cuts stabilise; others disintegrate. Some open up coherent lines of analysis; others produce noise.

But constraint is not the same as pre-existence. The viability of a cut does not derive from its correspondence to an underlying object. It derives from its capacity to sustain a coherent construal.

---

The Consequences of a Cut

Once a cut is made, it does more than delimit a unit. It organises what can count as relevant within that unit.

Consider a simple case: an exchange between two speakers.

One cut might construe it as:

• a sequence of clauses, analysable in terms of transitivity and mood

Another might construe it as:

• a negotiation of interpersonal positioning across turns

Another still:

• an instance of a broader register, shaped by field, tenor, and mode

These are not layers applied to the same object. They are different objects, each brought forth by a different cut.

And each cut carries consequences:

• what distinctions become available
• what counts as evidence
• what can be said to follow

To change the cut is not to shift perspective on a stable phenomenon. It is to reconstitute the phenomenon itself.

---

Competing Cuts

Disagreement in analysis is often framed as a dispute over interpretation:

which account better captures what is really going on?

But from the perspective of the analytic cut, the disagreement runs deeper.

Analysts are not merely offering different readings. They may be operating with incompatible cuts—each actualising a different phenomenon.

This is why some debates never resolve. They are not disagreements about conclusions drawn from shared premises. They are divergences in what has been taken as the object in the first place.

To engage such a disagreement is not to argue for a better interpretation. It is to expose the cut:

• what has been included and excluded
• what has been stabilised as the unit
• what potential has been actualised, and what has been left unrealised

Only then can the relation between the analyses be understood.

---

Making the Cut Visible

Most analysis proceeds as if the cut were self-evident.

The text is presented. The analysis follows. The object appears to have been there all along.

This is not an innocent omission. It is what allows method to present itself as procedure: if the object is given, then the only question is how best to analyse it.

To foreground the analytic cut is to undo this illusion.

It is to show that:

• the object is an outcome, not a starting point
• the boundaries of analysis are decisions, not facts
• every analysis carries within it the trace of its own conditions of possibility

This does not paralyse analysis. It disciplines it.

For once the cut is visible, it becomes available for scrutiny:

• could it have been drawn differently?
• what would follow if it were?
• what does this cut enable—and what does it foreclose?

---

From Cut to Practice

If analysis begins with a cut, then analytic practice is not the application of a method, but the management of cuts.

Not in the sense of choosing arbitrarily, but in the sense of:

• making them deliberately
• sustaining them coherently
• and, where necessary, shifting them to explore the space of possible construals

The analyst does not stand outside the phenomenon, applying tools.

The analyst operates within a field of potential, drawing distinctions that bring phenomena into being.

The question is no longer:

“What is this an instance of?”

It becomes:

“What must be distinguished for this to appear as an instance of anything at all?”

Cuts That Make Worlds: Practising Relational Analysis — 1 Against Method: Why There Is No Procedure

There is a familiar expectation when one turns to theory: that it will eventually yield a method. A set of steps. A procedure that, if followed with sufficient care, will produce results—analyses that are correct, or at least defensible in a recognisable way.

This expectation will not be satisfied here.

Not because rigour is being abandoned, but because the expectation itself rests on a misunderstanding of what analysis is. It presumes that there is something there, already formed, awaiting examination; and that the analyst stands apart from it, equipped with tools adequate to the task. Method, in this sense, is a technology of access: a way of getting at what is already the case.

But if meaning is not something that exists independently of construal—if there is no phenomenon prior to the distinctions through which it is brought forth—then this entire picture begins to unravel.

There is no pre-given object of analysis.

There is no neutral vantage point from which such an object could be observed.

And there is no procedure that could guarantee that different analysts, following the same steps, would arrive at the same thing—because there is no “same thing” to arrive at prior to the act of analysis itself.

What, then, becomes of method?

---

Analysis as Actualisation

To analyse is not to uncover what is already there. It is to actualise a particular configuration of potential.

This is not a temporal claim. Analysis does not take something latent and bring it into being over time. Rather, it is a shift in perspective: a cut through a field of possible distinctions, in which a phenomenon emerges as this rather than that.

The “text” one analyses is not an object that precedes this cut. It is the outcome of it.

Different analyses do not offer competing views of the same phenomenon. They enact different phenomena.

This is the point at which the language of method begins to fail. For method presumes repeatability: that the same steps, applied to the same object, will yield the same result. But if the object itself is constituted in and through the analytic act, then repeatability cannot mean reproduction of a result. At most, it can mean the possibility of reconstrual under sufficiently aligned conditions.

---

The Illusion of Procedure

The appeal of procedure is obvious. It promises stability.

Follow these steps:

1. Identify the object
2. Apply the framework
3. Produce the analysis

Each step appears to rest on something given: the object is there; the framework is available; the outcome is a function of correct application.

But each step conceals a decision.

• What counts as the object is not given; it is cut out from a continuum of potential.
• What counts as the framework is not neutral; it privileges certain distinctions over others.
• What counts as the result is not discovered; it is the stabilisation of one construal among many that might have been actualised.

Procedure obscures these decisions by presenting them as steps. It converts acts of construal into operations on an object.

The cost of this conversion is high: the analyst disappears from the analysis, and with them, the very conditions under which the phenomenon has been constituted.

---

Rigour Without Method

If method, understood as procedure, is untenable, does anything like rigour remain?

Yes—but it must be relocated.

Rigour does not consist in following steps correctly. It consists in maintaining consistency of construal across the analysis.

This means:

• the distinctions one makes must cohere with one another
• the movement between levels (from potential to instance, across strata) must be disciplined
• the analysis must remain accountable to the system of meanings within which it operates

What it cannot mean is correspondence to a reality independent of construal. There is no such tribunal to which the analysis could appeal.

This is not relativism. Not anything goes. On the contrary: the constraints are internal, systemic, and unforgiving. An incoherent construal collapses under its own weight. A disciplined one holds—not because it matches the world, but because it sustains itself as a configuration of meaning.

---

What Follows

If there is no procedure, there is still practice.

But practice must be understood differently:

not as the application of a method, but as the cultivation of analytic cuts.

The question is no longer:

“What steps should I follow to analyse this?”

It becomes:

“What distinctions must be drawn for this phenomenon to emerge—and what follows from drawing them this way rather than another?”

The posts that follow will not provide a toolkit. They will stage a series of analyses, each making its cuts explicit, each showing how different construals actualise different phenomena.

The aim is not to teach a method.

It is to make visible what method conceals.