Cognition as Non-Commutative Geometry
There is a simple mathematical idea, non-commutativity, with consequences that are easy to underestimate:
$$ AB\neq BA $$The order of operations, in some cases, matter.
In quantum mechanics this is encoded directly in the algebra of observables. In operator algebras it is more general: non-commutativity means that the operations available to a system cannot, in general, be reduced to simultaneous classical coordinates.
The proposal here is that cognition should be treated in the same structural way.
Not because cognition is quantum.
Not because the brain is secretly a quantum computer.
And not because non-commutative geometry is an exotic metaphor for the mind.
The claim is more basic:
A sufficiently self-referential cognitive system should be modeled as a non-commutative algebra of operations together with a state on that algebra.
The geometry then need not be specified beforehand. It can emerge from the algebra and its state.
This gives a possible mathematical architecture for cognition:
$$ \boxed{ \text{cognition} \longrightarrow (\mathcal A,\omega) \longrightarrow \text{representation} \longrightarrow \text{geometry and flow} } $$where \$\mathcal A\$ is an algebra of cognitive operations and \$\omega\$ is the current cognitive state.
The interesting part is what happens after making this identification.
Subjective time can be associated with modular flow. Semantic distance can arise from a spectral triple. Learning can deform the algebra itself rather than merely move a state through a fixed space. Self-reference becomes a mathematical constraint rather than a philosophical afterthought.
And several old problems — measurement, the mind–matter relation, consciousness, and the unreasonable effectiveness of mathematics — acquire a common structural form.
This is a conjecture, not an established theory of cognition. The mathematics used below is established mathematics. The identification of that mathematics with cognition is the hypothesis.
1. Cognition is not a point in a state space
A conventional mathematical description of cognition starts with some state
$$ x\in X $$The system has a state space \$X\$, and cognition corresponds to trajectories
$$ x(t)\in X $$This is useful, but it misses something essential.
A cognitive system does not merely have states. It performs operations.
It attends.
It recalls.
It compares.
It asks questions.
It changes context.
It composes concepts.
It evaluates propositions.
And those operations act on one another.
Let \$\mathcal A\$ denote the algebra generated by the relevant cognitive operations. A state is then a positive normalized functional
$$ \omega:\mathcal A\rightarrow\mathbb C $$satisfying
$$ \omega(A^*A)\geq 0, \qquad \omega(1)=1 $$The basic object is therefore not simply \$x\$, but
$$ (\mathcal A,\omega) $$The algebra contains the possible operations; the state specifies the current configuration of the system.
This distinction is important because the operations need not commute.
Suppose \$A\$ represents attending to one concept and \$B\$ represents applying some contextual inference. There is no general reason for
$$ AB=BA $$In fact, cognition is full of examples where it should not.
Ask a question, then evaluate the evidence.
Evaluate the evidence, then ask the question.
These are different cognitive processes.
The second operation is performed in a state modified by the first.
Formally,
$$ AB|\psi\rangle \neq BA|\psi\rangle $$The important object is therefore not only the result of an operation but the composition law of operations.
That is exactly what an algebra records.
2. Non-commutativity comes from self-reference
The deepest part of the hypothesis is not actually non-commutativity.
It is self-reference.
Non-commutativity is the algebraic structure required to represent one important consequence of self-reference: operations performed by the system can change the conditions under which subsequent operations are performed.
Consider a system that contains a model of itself.
Let
$$ S $$denote the system and
$$ M(S) $$its internal model of itself.
The model is not external to \$S\$:
$$ M(S)\subseteq S $$Now let \$A\$ be an operation that interrogates the system’s own state.
Then
$$ S \xrightarrow{A} S' $$and the internal model must now describe \$S'\$, not \$S\$.
The act of acquiring information about the system has become part of the system’s dynamics.
This is the basic closure condition:
$$ \boxed{ \text{the observer belongs to the observed system} } $$Once this is taken seriously, the classical picture becomes inadequate.
A classical observer can be idealized as standing outside the system and assigning values to independent observables.
A self-referential observer cannot do this in general.
The operation
$$ A:\text{“inspect the system”} $$is itself an operation inside the system.
It can affect the state on which the next operation \$B\$ acts.
Hence the natural structure is
$$ AB\neq BA $$I would not state this as a mathematical theorem saying that every self-referential system must have a non-commutative algebra. That is too strong.
The conjecture is narrower and more useful:
For cognition, self-reference is primary, and non-commutativity is the natural algebraic representation of the resulting order-sensitive internal operations.
This reverses the usual intuition.
Non-commutativity is not being imported from quantum mechanics into cognition.
It is being proposed as the mathematical language of cognitive closure.
Quantum theory is one place where that language is already fundamental.
It need not be the only one.
3. From algebra to geometry
If \$\mathcal A\$ is non-commutative, there may be no underlying classical space whose points can simply be listed.
Non-commutative geometry begins precisely here.
Rather than starting with a manifold \$M\$ and its algebra of functions
$$ C^\infty(M), $$one starts with the algebra.
The classical case is recovered when the algebra is commutative:
$$ fg=gf $$The non-commutative case generalizes the entire construction.
For cognition this suggests something stronger than “mental states have a complicated geometry.”
Perhaps there is no fixed cognitive geometry at the fundamental level.
The geometry is generated by the algebra of available operations and by the current state.
Given \$(\mathcal A,\omega)\$, the Gelfand–Naimark–Segal construction produces a representation
$$ \pi_\omega:\mathcal A\rightarrow B(\mathcal H_\omega) $$and a cyclic vector
$$ \Omega_\omega\in\mathcal H_\omega $$The resulting Hilbert space depends on \$\omega\$.
So the cognitive state does not merely occupy a point in a pre-existing geometry.
It helps determine the representation in which the algebra is realized.
This gives a natural mathematical interpretation of context.
Two cognitive states can involve the same underlying operations while realizing them differently:
$$ (\mathcal A,\omega) \qquad\text{and}\qquad (\mathcal A,\varphi) $$The algebra is the same.
The effective representation is not necessarily the same.
This is much closer to contextual meaning than the usual picture of every concept being assigned a fixed coordinate.
A word does not have a single invariant semantic location.
Its effect depends on the state in which it is encountered.
The state changes the effective geometry.
4. Subjective time
The next step is the strongest mathematical reason to consider this framework.
Take a von Neumann algebra \$\mathcal M\$ with a suitable faithful normal state \$\omega\$.
Tomita–Takesaki theory produces a modular operator \$\Delta_\omega\$ and the modular automorphism group
$$ \boxed{ \sigma_t^\omega(A)
=
\Delta_\omega^{it} A \Delta_\omega^{-it} } $$
This is a genuine one-parameter automorphism group of the algebra.
The important fact is structural:
$$ (\mathcal M,\omega) \longrightarrow \sigma_t^\omega $$A flow is generated by the algebra-state pair.
This suggests the central cognitive identification:
$$ \boxed{ \text{subjective time} \sim \text{modular flow of the cognitive state}. } $$The symbol \$\sim\$ matters.
This is not a theorem of Tomita–Takesaki theory about human consciousness. It is the proposed identification.
The appeal is that time is now endogenous.
It is not an externally supplied parameter \$t\$ against which cognition passively evolves.
The current state determines a canonical internal flow.
In finite dimensions, the same structure is explicit. For a density matrix \$\rho\$,
$$ \sigma_t^\rho(A) \rho^{it}A\rho^{-it} $$That is just matrix evolution.
There is no fundamental obstruction to implementing it computationally.
The empirical question is whether a cognitive model endowed with such state-dependent flow captures temporal phenomena better than one with an externally imposed clock.
The hypothesis predicts that it should.
5. KMS equilibrium and cognitive dynamics
The modular flow is intimately related to the KMS condition. In the normalized convention \$\beta=1\$,
$$ \omega\left( \sigma_t^\omega(A)B \right)
=
\omega\left( B\sigma_{t+i}^\omega(A) \right) $$
In operator algebra and statistical mechanics this characterizes equilibrium relative to a flow.
For cognition, the useful interpretation is not “the brain is a thermal system.”
It is that a cognitive regime might have an internally preferred dynamical structure.
A state close to its own dynamically compatible regime would be stable.
Perturbing it produces deviations.
Learning changes the state.
Surprise changes the state.
Cognitive dissonance changes the state.
Attention can move the system toward another coherent regime.
This suggests a measurable quantity:
$$ \mathcal L_{\rm KMS}
\sim
\text{KMS violation} $$
Rather than declaring that a particular cognitive state is KMS equilibrium, one can test whether cognitive systems naturally exhibit approximate KMS structure and whether departures from it correlate with measurable behavioral or temporal properties.
It also gives a potential definition of intrinsic cognitive time that is independent of any particular neural implementation.
6. Geometry from a Dirac operator
Once there is an algebra and a state-dependent representation, the next question is geometry.
The natural non-commutative object is a spectral triple,
$$ (\mathcal A,\mathcal H,D) $$For cognition one could consider
$$ (\mathcal A_{\rm cog},\mathcal H_\omega,D_{\rm cog}) $$The operator \$D_{\rm cog}\$ is not assumed to be a literal physical Dirac operator.
It is the operator from which the cognitive geometry would be reconstructed.
Its relation to the algebra is encoded by
$$ [D_{\rm cog},A] $$The commutator measures how \$A\$ interacts with the proposed geometry.
If \$[D,A]\$ is small, \$A\$ changes the geometry only weakly.
If it is large, \$A\$ is geometrically significant.
Connes’ distance formula is
$$ d(\omega,\varphi)
=
\text{sup}\left{ |\omega(A)-\varphi(A)| : |[D,A]|\leq1 \right} $$
This gives a very different notion of cognitive distance.
Instead of saying that two mental states are close because their embedding vectors happen to be close, one asks how distinguishable they are under the algebra of admissible operations.
That is a genuinely operational geometry.
And because both the algebra and the state can change, the geometry can change too.
Learning is then potentially not just
$$ \omega\rightarrow\omega', $$but
$$ (\mathcal A,\omega) \rightarrow (\mathcal A',\omega') $$The system can change the set of operations it can perform.
That is a more natural mathematical description of structural learning than motion through a fixed manifold.
7. Cognition as an evolving algebra
This distinction becomes important once learning and development are considered.
There are two fundamentally different kinds of change:
$$ \text{state change} $$and
$$ \text{algebra change}. $$A state change keeps the available operations fixed:
$$ \omega\rightarrow\omega' $$An algebra change alters what operations are available in the first place:
$$ \mathcal A\rightarrow\mathcal A' $$Acquiring a new conceptual operation is of the second kind.
A child learning symbolic recursion is not merely assigning another value to an existing variable.
The space of possible transformations has changed.
An expert mathematician does not merely know more facts than a beginner.
The expert has acquired operators that reorganize enormous regions of the conceptual space.
This gives a possible mathematical language for abstraction, compositionality, and cognitive development:
$$ \boxed{ \text{intelligence is partly the ability to construct and compose useful operators.} } $$That is also where the hypothesis becomes relevant to artificial intelligence.
A system whose representation is primarily a vector can learn useful transformations.
A system whose primary object is an algebra can potentially learn transformations as first-class objects.
The distinction is architectural rather than cosmetic.
8. The observer problem in quantum gravity
There is an independent development in fundamental physics that points in a remarkably similar mathematical direction.
It is important to state immediately what this does not mean.
The cognitive hypothesis above does not come from quantum gravity.
It does not assume quantum mechanics.
It does not assume gravity.
Its motivation is self-reference and the structure of cognition.
The convergence is instead that quantum gravity has independently encountered the problem of the observer and the inadequacy of treating observables as belonging to a fixed external spacetime region.
In ordinary quantum field theory one can associate an algebra of observables with a region.
Gravity makes this problematic because the geometry itself is dynamical.
QG researchers proposed replacing the algebra associated with a spacetime region by an algebra of observables along the timelike worldline of an observer. The observer is therefore built into the definition of the algebra rather than appended as an external measurement device. Witten developed this explicitly as a candidate background-independent algebra for quantum gravity.
The conceptual structure is:
$$ \boxed{ \text{observer} \longrightarrow \text{worldline} \longrightarrow \text{observable algebra} } $$This is a significant convergence with the cognitive proposal.
In cognition:
$$ \text{observer} \longrightarrow \text{self-reference} \longrightarrow \text{non-commutative cognitive algebra}. $$In quantum gravity:
$$ \text{observer} \longrightarrow \text{relational localization} \longrightarrow \text{observer-associated algebra} $$These are not the same theory.
They should not be merged.
But they point toward the same methodological lesson:
the observer may belong inside the fundamental algebraic description rather than being an external classical object that is introduced only at the end.
This line of thought has continued in more recent work on quantum-gravity observables and relational constructions. Giddings, for example, has argued that gravitationally dressed observables and related structures can significantly alter the algebraic structure familiar from local quantum field theory, including connections to crossed-product and type-II structures in special cases.
The important point for the present conjecture is not that quantum gravity validates a theory of cognition.
It does not.
The point is that once observers, localization, and self-contained description become fundamental, algebra ceases to be bookkeeping.
It becomes part of the physics.
That makes the observer-centric algebraic formulation of cognition considerably less alien to the broader trajectory of modern mathematical physics.
9. Measurement without an external observer
This leads directly to the measurement problem.
The traditional formulation divides evolution into unitary dynamics and measurement updates:
$$ \rho \rightarrow U\rho U^\dagger $$versus
$$ \rho \rightarrow \frac{M\rho M^\dagger} {\operatorname{Tr}(M\rho M^\dagger)} $$The problem is not that the second equation is mathematically mysterious.
It is that the theory does not fundamentally specify where the measurement boundary is.
Who counts as the observer?
Where does the apparatus end?
Why should one physical interaction be described by a different rule merely because it has acquired the semantic status of “measurement”?
An observer-centered algebraic formulation removes the need to put a classical observer outside the system.
Write instead
$$ \mathcal A_{\rm total}
=
\mathcal A_{\rm system} \otimes \mathcal A_{\rm apparatus} \otimes \mathcal A_{\rm observer} $$
The measurement is an interaction inside the larger system.
State updating can be represented using the usual completely positive instrument/Kraus formalism; in appropriate algebraic settings, conditional expectations give the corresponding projection onto a subalgebra of accessible information.
The conceptual shift is:
$$ \text{collapse as a special fundamental event} $$becomes
$$ \text{state update relative to an observer and an accessible algebra} $$This does not by itself prove that the measurement problem has been solved.
It changes the question.
Instead of asking when the universe stops evolving quantum mechanically, one asks how an observer’s state and accessible subalgebra are produced by interaction.
The observer is no longer outside the formalism.
That is the point.
10. Consciousness and self-description
The consciousness problem is harder.
A non-commutative cognitive model can explain order effects, contextuality, state-dependent geometry, modular flow, and limits on self-description.
None of those statements, by themselves, derive phenomenal experience.
The strongest defensible claim is therefore not
$$ \text{NCG}=\text{consciousness} $$It is:
$$ \boxed{ \text{self-referential cognition may have structural limits on complete self-description.} } $$A cognitive system capable of representing itself contains the representation inside the represented system.
Let
$$ R:\mathcal S\rightarrow\mathcal S $$be a self-modeling operation.
Then applying \$R\$ changes the state on which subsequent applications of \$R\$ operate.
There is no final external vantage point.
The sequence does not terminate by producing a perfectly external description of itself.
This is not identical to Gödel incompleteness.
Gödel’s theorems concern sufficiently expressive formal systems under precise conditions.
A cognitive algebra is not automatically such a formal system.
The connection is therefore structural rather than theorem-for-theorem.
Both involve the failure of a sufficiently self-referential system to collapse its internal perspective into a completely external description.
This suggests a more restrained formulation of the qualia problem:
Perhaps the apparent impossibility of deriving the first-person perspective entirely from a third-person description is not simply an explanatory failure.
Perhaps it reflects a structural asymmetry between:
$$ \text{description of a system from outside} $$and
$$ \text{state of a system participating in its own description}. $$That is not a solution to the hard problem.
It is a candidate explanation for why the hard problem has the form it does.
11. The unreasonable effectiveness of mathematics
There is then another loop.
The universe produces physical systems.
Some of those systems evolve cognition.
Cognition constructs mathematics.
Mathematics is then used to describe the universe.
So:
$$ \text{universe} \rightarrow \text{cognition} \rightarrow \text{mathematics} \rightarrow \text{universe}. $$The usual formulation of Wigner’s problem implicitly treats mathematics and reality as independent objects whose agreement requires explanation.
But if cognition is a subsystem of the same world it is trying to model, the independence is weaker.
The mathematics discovered by cognitive systems is constrained by the structure of those systems.
If the same physical world that generated cognition also constrains the algebraic structure of fundamental physics, then mathematical compatibility is no longer entirely mysterious.
The conjecture is not that human minds literally contain the mathematics of physics.
It is that the space of mathematical structures humans find natural may itself be biased toward structures that are compatible with the world in which those minds evolved.
In that picture, the emergence of non-commutative geometry in both mathematical physics and a theory of cognition would not be a coincidence of vocabulary.
It would be a structural recurrence.
The same kind of algebra appears in the observer’s description of the world and in the mathematics used to describe the observer.
12. A possible theory of the self
There is another consequence.
If cognition is fundamentally represented by \$(\mathcal A,\omega)\$, the self need not correspond to a particular elementary object.
It can correspond to a stable structure of the pair.
This is particularly interesting in von Neumann algebra theory.
For a Type III algebra \$\mathcal M\$ with a modular action \$\sigma^\phi\$, one can form the crossed product
$$ \mathcal N
=
\mathcal M\rtimes_{\sigma^\phi}\mathbb R $$
The resulting algebra has a much richer trace structure than the original Type III algebra.
Mathematically, this is a precise construction.
The cognitive interpretation is speculative:
perhaps the persistence we call a “self” is not a substance but an invariant structure emerging from the interaction of algebra and flow.
Then personal identity would not mean
$$ \text{same microscopic state} $$It would mean something closer to
$$ \text{same persistent structure under transformation} $$The distinction matters.
A human cognitive state changes continuously.
Memories change.
Neural configurations change.
Beliefs change.
The body changes.
Yet some structural identity persists.
An algebraic theory of cognition has a natural vocabulary for distinguishing the changing state from the invariant structure.
The trace should not simply be declared “the self.” That is too strong.
The more defensible conjecture is that trace-like invariants may provide mathematical candidates for what persistence of cognitive identity means.
13. Where pathology might enter
The same framework suggests that a cognitive pathology could be represented as a structural deformation rather than only a collection of symptoms.
One could imagine failures of:
$$ \text{representation coherence}, $$ $$ \text{modular stability}, $$ $$ \text{state-dependent geometry}, $$or
$$ \text{operator composition} $$For example, a system could become trapped near an unusually stable dynamical regime, exhibit unstable modular flow, or support incompatible internal representations.
These should initially be treated as mathematical hypotheses, not clinical identifications.
The useful question is whether quantities such as
$$ |\text{KMS violation}|, $$spectral distances, representation changes, or modular-flow statistics correlate with independently measurable cognitive variables.
The advantage of the algebraic framework is precisely that it creates quantities that can in principle be computed.
It turns philosophical vocabulary into candidate observables.
14. Artificial cognition
The same point applies to artificial intelligence.
Suppose an architecture is built around
$$ (\mathcal A,\rho) $$rather than only a hidden vector \$h\$.
Let \$\mathcal A\$ contain trainable operators.
Let \$\rho\$ represent the internal state.
Then internal temporal evolution can be defined by
$$ \sigma_t^\rho(A)
=
\rho^{it}A\rho^{-it} $$
Inputs can act through completely positive maps.
Learning can modify \$\rho\$.
Structural learning can modify \$\mathcal A\$.
A geometry can be encoded by \$D\$.
Distance can be computed through
$$ d(\omega,\varphi)
=
\sup_{|[D,A]|\leq1} |\omega(A)-\varphi(A)|. $$
The architecture therefore has three different levels of plasticity:
$$ \rho \quad\text{changes the state}, $$ $$ \mathcal A \quad\text{changes the available operations}, $$and
$$ D \quad\text{changes the geometry in which those operations are organized} $$This is a substantially richer model of cognitive architecture than a fixed latent vector.
It also gives direct experiments.
Compare a commutative and non-commutative model under matched parameter count and compute.
Test order-sensitive reasoning.
Test contextual semantics.
Test continual learning.
Measure the stability and recovery of intrinsic flows.
Measure whether state-dependent geometry predicts generalization.
The conjecture should rise or fall on these comparisons.
15. What would count as evidence?
A theory this broad needs unusually aggressive evidence:
There are several relatively clean tests.
Non-commutativity should improve prediction
Find tasks with experimentally established order effects and compare:
$$ AB=BA $$against
$$ AB\neq BA. $$The non-commutative model should win out of sample, not merely fit the training distribution.
Geometry should depend on state
For two states \$\omega\$ and \$\varphi\$,
$$ d_\omega(A,B) \neq d_\varphi(A,B) $$should be empirically useful.
Context should alter the effective geometry.
Modular flow should capture temporal structure
A model with
$$ \sigma_t^\omega $$should predict aspects of subjective or cognitive timing that an equivalent architecture with an externally imposed clock cannot.
Algebraic plasticity should matter
Allowing
$$ \mathcal A_t\neq\mathcal A_{t+1} $$should produce advantages on tasks requiring genuine conceptual restructuring.
Self-reference should have measurable structural signatures
A system modeling itself should exhibit identifiable changes in operator composition, representation structure, or modular dynamics.
If these predictions fail, the hypothesis should be abandoned or reduced.
The point is not to make every cognitive phenomenon compatible with NCG after the fact.
The point is to derive predictions that competing models do not naturally make.
16. What the hypothesis does and does not claim
The cleanest way to state the program is probably this.
It does not claim
$$ \text{cognition}=\text{quantum mechanics}. $$It does not require quantum superposition.
It does not require microscopic quantum effects in the brain.
It does not assume that consciousness is a quantum phenomenon.
It does not claim that every theorem of operator algebra has an immediate psychological interpretation.
And it does not currently solve the hard problem.
It does claim
$$ \boxed{ \text{cognition may be fundamentally better represented as } (\mathcal A,\omega) \text{ than as a point in a fixed state space.} } $$From that starting point:
$$ (\mathcal A,\omega) \rightarrow \mathcal H_\omega $$gives a state-dependent representation;
$$ (\mathcal A,\omega) \rightarrow \sigma_t^\omega $$gives intrinsic modular dynamics;
$$ (\mathcal A,\mathcal H,D) \rightarrow d $$gives non-commutative geometry;
and self-reference imposes structural constraints on what the system can represent about itself.
The strongest version of the conjecture is therefore not that non-commutative geometry is a convenient mathematical analogy for cognition.
It is that non-commutativity may be the algebraic signature of cognitive closure.
17. The convergence
There are now two apparently different problems in which the observer refuses to disappear.
In cognition:
$$ \text{self-reference} \rightarrow \text{observer inside system} \rightarrow \text{non-commutative cognitive algebra}. $$In quantum gravity:
$$ \text{dynamical geometry} \rightarrow \text{localization becomes relational} \rightarrow \text{observer/worldline} \rightarrow \text{observer-associated algebra}. $$The second line is an independent development in quantum gravity, including Witten’s proposal for a background-independent algebra associated with an observer’s worldline.
The two programs should not be conflated.
The cognitive conjecture did not need quantum gravity.
Quantum gravity does not provide evidence that cognition is non-commutative.
But their convergence is conceptually important.
Both force the same methodological move:
stop pretending that the observer can always be placed outside the mathematical description.
Once the observer is internal, algebra becomes relational.
What can be observed depends on the state.
What can be distinguished depends on the algebra.
What counts as an accessible operation depends on the observer.
And, potentially, the geometry itself becomes state-dependent.
That is exactly the structure required by the cognitive hypothesis.
18. The broader implication
The usual hierarchy is
$$ \text{state} \rightarrow \text{dynamics} \rightarrow \text{geometry}. $$The non-commutative picture suggests something closer to
$$ \boxed{ (\mathcal A,\omega) \rightarrow \begin{cases} \mathcal H_\omega & \text{representation}\ \sigma_t^\omega & \text{time}\ D & \text{geometry}\ \text{invariants} & \text{identity}\ \text{self-reference} & \text{limits of description} \end{cases} } $$The state and algebra generate the structures that a classical theory would normally take as primitive.
Time is no longer necessarily an external parameter.
Geometry is no longer necessarily a fixed stage.
The observer is no longer necessarily outside.
The self is no longer necessarily an object.
And consciousness is no longer forced to be described as a mysterious substance added to an otherwise complete physical state.
That does not mean the problems disappear.
It means they may have been posed at the wrong level.
19. What this would touch
If the conjecture survives its empirical tests, it would have consequences far beyond one model of cognition.
Measurement. The observer can be incorporated into the algebra, replacing an externally imposed classical/quantum boundary with a relational state-update problem.
Consciousness. Self-reference becomes a structural feature of the cognitive dynamics, potentially explaining why complete third-person self-description is problematic without claiming that this alone derives phenomenal experience.
The mind–matter problem. The question becomes whether neural and cognitive organization instantiate related algebraic structures, rather than how an entirely different ontological substance emerges from matter.
The unreasonable effectiveness of mathematics. The mathematics produced by cognition may be structurally constrained by the same world whose physical regularities it describes.
Time. Subjective temporal structure may arise from the modular flow of a state rather than from an external clock.
Identity. Personal identity may correspond to invariants of an evolving algebra rather than to persistence of a particular microscopic state.
Learning. Structural learning becomes deformation of the algebra itself:
$$ \mathcal A_t\rightarrow\mathcal A_{t+1}. $$Intelligence. Intelligence becomes, at least partly, the ability to construct, compose, transform, and select operators in a changing state-dependent geometry.
Artificial consciousness. The question becomes whether an artificial system merely simulates these structures or physically instantiates the relevant self-referential dynamics.
None of these are established conclusions.
They are the research surface opened by the conjecture.
20. The actual claim
The claim can finally be made very compact.
Let cognition be described by an algebra \$\mathcal A\$ of operations and a state \$\omega\$.
Then propose:
$$ \boxed{ \text{self-reference} \Rightarrow \text{order-sensitive cognitive operations} \Rightarrow [\mathcal A,\mathcal A]\neq0 } $$and therefore
$$ \boxed{ (\mathcal A,\omega) \Rightarrow \text{state-dependent representation, geometry, and intrinsic flow}. } $$The first arrow is the central conjectural step.
The rest can be developed using established mathematics.
If this program is correct, then cognition is not fundamentally a point moving through a pre-existing psychological manifold.
It is a self-referential algebra whose state determines the geometry in which its own operations occur.
The mind would not merely occupy a geometry.
It would generate one.
And the deepest consequence is not that the brain is quantum.
It is that the mathematical structure required to describe a system that is inside its own description may be qualitatively different from the mathematics of a system observed from nowhere.
That is where non-commutativity enters.
Not as decoration.
Not because quantum mechanics made it fashionable.
Because closure itself may have an algebra.
And if cognition is a sufficiently powerful closed system, that algebra may be the geometry of thought.