Logic

Dot Theory as a representational programme and a game of naming

Stefaan Vossen
First publication: September 2024

A short orientation

This site is a research programme about representation.

For internal site links to:

External GitHub Repo: https://github.com/stefaanvossen-dot/Dot-theory

Logic and the role it plays in a large fraction of what humans understand about reality becomes actionable only after it has first been named, recorded, structured, and computed. The act of naming does not conjure new matter into existence. It does something more modest but still powerful: it binds attention, measurement, and inference to a stable symbol, and therefore makes an object available to reasoning and to institutions.

These pages combined explain the logical spine of Dot Theory in that sense. It is not a physics paper. It is a page that sets the programme’s representational stance, the discipline of its claims, and the mathematical structure of its “game” out.

If you want the technical formalism, it belongs on dedicated technical pages. This page is on the bridge between structures describing human intelligibility and formal work in reinterpreting how we formalise the terms that describe physical reality. There are various papers across the blog that flirted with various quasi-metaphysical approaches that demonstrate those extensions within my limited skill.

Abstract

Dot Theory proposes a representational hypothesis:

In some contexts, standard state representations discard or leave implicit information that is relevant to prediction and control. When that contextual structure can be formalised as metadata and carried through computation, model performance can improve in regimes where the metadata matters.

This is a claim about modelling and information, not a completed unification of physics. Any extension must be defined precisely, must reduce to the standard formalism in the appropriate limits, and must produce discriminable predictions to count as mathematical physics.

The “Name-and-Claim game” is used here as a logical model of how meaning, measurement, and agency enter representational systems. It is offered as a disciplined metaphor and a design constraint for formal modelling, not as a substitute for derivation.

1. The Name-and-Claim game

The Name-and-Claim game describes a basic operation that is common to science, computing, and everyday life.

Name: bind a pattern of observations to a symbol.
Claim: treat the symbol as an object within a model, and permit inferences and decisions based on it.

This is not mystical. It is how a vague “some” becomes a “thing” inside a ruled system of reasoning. Without the name, there is no stable handle. Without the claim, there is no operational consequence.

In practice, “naming” involves:

  • selecting what counts as data

  • selecting what counts as context

  • choosing a resolution or granularity

  • adopting a measurement convention

  • choosing what an error is relative to a standard

That selection is not arbitrary, but it is conditional. It depends on goals, constraints, and available instruments.

2. Why this matters for modelling

Most systems we care about are partially observable. The underlying state is not directly given. We infer.

In that setting, representation does work. It sets the space of possibilities you can even consider.

A practical modelling claim follows:

If a representation collapses distinct contexts into the same state description, then the model will predict as if those contexts are identical. In regimes where they are not identical, errors become systematic.

Dot Theory is a programme to formalise and test when this is happening.

3. A restrained core thesis

This site uses the word “dots” to refer to informational units, not metaphysical atoms.

Let D be a collection of data points, and let μ be contextual metadata. Dot Theory focuses on two questions:

  1. What does the model treat as data D?

  2. What does it omit or treat implicitly as metadata μ?

The programme claim is:

There exist contexts where carrying μ explicitly improves predictive fidelity, because the mapping D ↦ state collapses information that the dynamics depends on.

This is not a universal claim. It is a conditional claim. It must be shown case by case.

4. Three motives, one modelling warning

The site often discusses three motives because motive shapes what information is selected and how it is used.

A rough triad is:

  1. Self-improvement: optimise future outcomes under constraints

  2. Victory: optimise persuasion, status, or dominance within a social game

  3. Self-reduction: sabotage or refusal of optimisation

This is not a moral sermon. It is an accuracy-acquiring modelling warning structure.

When a system optimises for “victory”, it often rewards rhetorical coherence over empirical correction. When that happens, error-correction mechanisms degrade. That is one reason scientific practice privileges falsification and adversarial testing.

The Name-and-Claim game is therefore not only about meaning. It is also about the conditions for corrigibility and the opportunities that currently exist.

5. Where observer-conditioning actually enters, without metaphysics

This site generally uses “observer” in a strict sense:

  • measurement context

  • instrument configuration

  • sampling and selection effects

  • modelling priors and constraints

  • the agent’s goal and loss function

These are formal, and they can be represented in mathematical structures.

The site does not require claims about consciousness to make the basic representational thesis.

6. How Dot Theory relates to physics, in a programme-safe way.

  • Dot Theory is not a replacement for Quantum Mechanics or General Relativity.

  • It is a representational programme that asks whether certain extensions to state description are warranted in specific regimes.

  • Any physics-facing claim must be expressed as a defined extension, with symmetry constraints and falsifiable predictions.

Physics is not persuaded by analogy. It is persuaded by derivation, limit recovery, and measurable divergence.

7. A minimal formal scaffold

This section given for reference.

Let:

  • ℋ be a conventional state space

  • ψ ∈ ℋ be a standard state

  • ℳ be a metadata space

  • μ ∈ ℳ be contextual metadata

  • 𝒮 ≔ ℋ × ℳ be an extended state space

  • Ψ = (ψ, μ) ∈ 𝒮 be an extended state

Let π : 𝒮 → ℋ be the projection π(Ψ) = ψ.

The key representational question becomes:

When does π discard information required to predict observables of interest?

That can be formalised as an identifiability or observability problem:

Distinct extended states Ψ₁ ≠ Ψ₂ may project to the same ψ while producing different distributions over observables in the regimes we care about.

If that happens, modelling solely in ℋ is representationally lossy.

This claim is testable once ℳ and the dynamics on 𝒮 are defined.

8. What would count as success, and what would count as failure

Dot Theory is not validated by rhetoric or breadth. It is validated by a narrow checklist:

Success conditions

  • A precise definition of metadata space ℳ in a concrete domain

  • A coherent extension of dynamics on 𝒮

  • Recovery of the standard model as a limiting case in regimes where μ is irrelevant

  • A discriminable prediction ΔO that differs from the reduced model

  • A falsification condition stated in advance

Failure conditions

  • No definable ℳ that improves prediction under controlled tests

  • No coherent dynamics that preserves required symmetries

  • No discriminable predictions

  • Retrospective fitting without falsification criteria

This is the discipline required for a programme that wants to be taken seriously.

9. Closing

The Name-and-Claim game is a compact way to describe a serious point declared across the project:

Representation is an action. It selects what is counted, and creates what is ignored, and what is permitted to matter in that interaction.

Dot Theory is a programme for making that specific action explicit, formal, and testable.

If the programme is wrong, it should fail cleanly. If it is right in specific regimes, it should yield measurable improvements in prediction and control.

Either outcome is informative.

Stefaan Vossen

Appendix

A restrained “formal proof” statement

Rather than offering a long pseudo-proof with speculative physics constants, I will state the logical core as a lemma and an implication:

Lemma (Projection loss).
Let Ψ = (ψ, μ) ∈ 𝒮 = ℋ × ℳ and π(Ψ) = ψ. If there exist Ψ₁ ≠ Ψ₂ such that π(Ψ₁) = π(Ψ₂) but the induced distributions over an observable O differ, then a model defined solely on ℋ is representationally incomplete for predicting O in that regime.

Implication (Need for extension).
In any domain where the lemma holds for observables of interest, extending the state representation to include a formally defined μ is warranted as a research direction. Whether it improves prediction is an empirical question.