If “text types” feel solid, it is because we are accustomed to looking from the pole of instance.
To correct this, we need to re-orient our perspective.
1. The Architecture We Already Have
In a Hallidayan model, language is stratified and clinally instantiated.
-
Language is a system of meaning potential.
-
A text is an instance — an event of meaning.
-
Between them lies the cline of instantiation.
system → subpotential / recurrent patterning → instance
Register sits within this gradient.
2. Register as Functional Variety
Canonically, register is variation according to:
-
Field — what is happening
-
Tenor — who is involved
-
Mode — what role language is playing
These are contextual variables.
They are realised in semantics, and probabilistically patterned in lexicogrammar.
Crucially:
Register is not a set of formal features.
When field, tenor, and mode align in particular ways, they generate a patterned region of semantic potential.
That region is what we call register.
From the pole of system, register is therefore:
-
A subpotential.
-
A clustering of probabilities.
-
A region within the semantic system.
It is not a box waiting to be filled.
3. Probability, Not Membership
When we describe a register, we are describing:
-
Tendencies in transitivity selections.
-
Preferences in mood and modality.
-
Characteristic thematic organisation.
-
Typical discourse semantic patterns.
But these are probabilistic.
If we speak in terms of “belonging to a type,” we have already shifted ontologically.
We have moved from:
patterned potential
to
classificatory membership.
That move is not trivial.
4. The Subtle Ontological Drift
When register is understood from the pole of potential:
-
It is gradient.
-
It is probabilistic.
-
It is relational.
-
It exists only within the system.
When register is reinterpreted from the pole of instance:
-
It appears as a class of texts.
-
It looks bounded.
-
It seems stable and nameable.
The shift is perspectival.
But once the instance perspective dominates, the subpotential disappears from view.
And with it, the cline.
5. Why This Matters
If register is a subpotential, then:
No text instantiates a register in the way a specimen instantiates a species.
A text is an event of actualisation within a field of probabilities.
All of this is normal — because instantiation is clinal.
Classification, by contrast, forces boundary decisions:
-
Is it a report or not?
-
Is it narrative or not?
-
Does it count as an explanation?
The cline asks a different question:
Where, and to what degree, is potential being actualised?
That is a different ontology of meaning.
6. Register on the Edge of Instance
We are now standing near the edge of instance.
From here, recurrence begins to look like type.
But we must resist the gravitational pull of classification.
Register is not something texts “have.”
It is a construal of contextual configuration realised as patterned probability within the system.
What appears as a “text type” is the afterimage of repeated actualisations across the cline.
If we hold firmly to potential, the ontology remains coherent.
If we slip toward classification, we begin to reify instance-patterns as essences.
And that is where confusion begins.
In the next post, we will examine how recurrence gradually becomes essence — how descriptive grouping drifts into ontological reification.
That drift is subtle.
And extraordinarily powerful.
No comments:
Post a Comment