If the purity of science is a myth, then its central product must be reconsidered.
Models are typically treated as the point at which science touches reality:
- representations of underlying structures
- approximations of how the world is
- tools for prediction and explanation
Even when acknowledged as imperfect, they are still assumed to aim at something beyond themselves.
This assumption now has to go.
1. Against the mirror
The dominant image is simple:
the world exists independently, and models reflect it—more or less accurately.
But this image depends on a separation that cannot be sustained:
- a world “as it is”
- and a model that represents it
From the perspective already established, there is no access to an unconstrued world.
There are only:
phenomena actualised through construal.
A model does not stand apart from what it describes.
It participates in bringing it forth.
2. Construal, not representation
This is the decisive shift.
A scientific model is not:
- a picture of reality
- a symbolic proxy
- a passive mapping
It is a semiotic operation:
- selecting distinctions
- organising relations
- stabilising patterns
- enabling certain phenomena to appear
Different models do not represent the same thing differently.
They actualise different phenomena.
3. The discipline of constraint
This does not mean anything goes.
Scientific models are not arbitrary.
They are:
disciplined construals under constraint.
Constraints include:
- prior models and theoretical commitments
- experimental setups
- mathematical formalisms
- institutional expectations
These do not ensure truth.
They ensure stability and reproducibility of construal.
4. Precision without innocence
Scientific models are often extraordinarily precise.
They:
- predict outcomes
- generate technologies
- coordinate large-scale action
This precision is frequently taken as evidence that they must be:
tracking an underlying reality.
But precision does not imply innocence.
It indicates:
that the construal is highly stabilised under constraint.
5. Multiplicity without contradiction
Once representation is set aside, a familiar problem dissolves.
How can multiple, incompatible models coexist?
- wave vs particle descriptions
- competing formalisms in different domains
- alternative modelling frameworks within the same field
From a representational view, this is a crisis.
From a construal view, it is expected.
Different models do not compete to describe the same reality.
They operate within different conditions of actualisation.
6. The role of mathematics
Mathematics is often taken as the guarantor of objectivity:
- abstract
- formal
- independent of context
But within this framework, mathematics is:
a resource for construal.
It provides:
- structured ways of drawing distinctions
- stable relations that can be iterated
- constraints that discipline variation
It does not anchor models to reality.
It stabilises how phenomena are brought forth.
7. From truth to viability
If models do not represent, what are they evaluated for?
Not truth in the correspondence sense.
But:
- viability
- stability
- reproducibility
- scope of application
A model is successful if it:
- holds under specified constraints
- coordinates practice effectively
- continues to generate usable phenomena
8. The illusion of discovery
Scientific practice is often described as discovery:
- uncovering what was already there
- revealing hidden structures
- finding the laws of nature
But if phenomena are actualised through construal, this language misleads.
Science does not uncover a pre-given world.
It extends the range of what can be brought forth as phenomenon.
What is discovered is not “out there,” waiting.
9. The coupling reappears
At this point, the earlier analysis returns.
Because models do not operate alone.
They are always embedded within:
- practices
- institutions
- norms of validation
Which means:
their stability depends on value coordination.
10. No retreat to relativism
This position will again invite a familiar misreading:
“If models don’t represent reality, then anything is as good as anything else.”
No.
Because construal is always under constraint.
Not all models stabilise.
Not all generate reproducible phenomena.
Not all coordinate practice effectively.
The difference is not between truth and error.
It is between:
- viable and non-viable construals.
11. The cost of clarity
What is lost here is comforting:
- the idea that science tells us how the world really is
- the sense of standing on solid ontological ground
What is gained is sharper:
- an account of what models actually do
- an understanding of their limits and power
- a way to analyse scientific practice without illusion
12. The next step
If models are disciplined construals, and their stability depends on coordination,
then the next question is unavoidable:
What kind of system maintains this coordination while denying that it exists?
Next: Post 3 — Practice Without Neutrality
No comments:
Post a Comment