Interiority Hierarchy: L0 → L4

Why a Consciousness Hierarchy is Needed

For millennia humanity has attempted to classify forms of inner life. Aristotle (4th century BCE) distinguished three grades of the soul: vegetative (nutrition and growth), animal (sensation and motion), and rational (thought). Leibniz (1714) introduced the notion of petites perceptions — unconscious micro-perceptions forming a continuous spectrum from stone to God. Fechner (1860) attempted to measure this spectrum quantitatively, discovering psychophysical thresholds — the minimum stimuli that consciousness can discriminate. In the 20th century Integrated Information Theory (IIT, Tononi, 2004) proposed a single numerical measure $\Phi$ — but left open the question of qualitative differences between levels.

The Unitary Holonomic Monism (UHM) inherits this tradition but goes further: rather than a single numerical scale it defines five qualitatively distinct levels of interiority (L0--L4), each characterised by a rigorous mathematical threshold condition. The transition between levels is not a gradual increase but a bifurcation (an abrupt reorganisation), analogous to the phase transition of water into steam.

Where we came from

In the Foundations section we established that every $\Gamma$ has an inner side, described the content of experience (interiority theory) and the self-observation operator $\varphi$ (self-observation). But not all systems "experience" in the same way: a stone, a bacterium, a cat, and a human differ radically. The L0--L4 hierarchy organises this difference into a rigorous mathematical classification.

Chapter roadmap

1. Five levels — from L0 (universal interiority) to L4 (theoretical limit)
2. L2: cognitive qualia — the central level with thresholds $R \geq 1/3$, $\Phi \geq 1$
3. L3: metacognition — meta-reflection $R^{(2)} \geq 1/4$, metastability
4. L4: categorical unreachability — colimit of the Postnikov tower, theoretical horizon
5. Gap characterisation — each level has a unique Gap profile
6. Bifurcations — transitions between levels as $A_2, A_3, A_4$ catastrophes

Analogy. Imagine a ladder of awareness. A stone (L0) — on the first rung: it has an "inner side", but it distinguishes nothing. A bacterium (L1) — distinguishes hot from cold, but does not know that it distinguishes. A cat (L2) — not merely distinguishes, but knows that it feels warmth (cognitive qualia). A meditator (L3) — knows that it knows that it feels (meta-reflection). And the last rung (L4) — is infinitely distant: complete self-knowledge, unreachable for finite systems.

DRY: Master definition of levels L0-L4

This is the canonical definition of the five levels of the interiority hierarchy. Full formalisation, proofs of threshold conditions, and the No-Zombie theorem — in Interiority hierarchy (proofs).

Biological L-levels [H]

The assignment of specific organisms to L-levels is a hypothesis [H], not a measured fact. A rigorous definition of the L-level requires knowledge of the system's $\Gamma$. For biological systems the protocol $\pi_{\text{bio}}$ is defined (C31), but has not been experimentally validated. The correspondences given are well-founded extrapolations from behavioural data.

---

Overview: five levels

Before diving into the details of each level, it is useful to see the entire ladder at once.

Level	Name	Threshold condition	Example
L0	Interiority	$\Gamma \in \mathcal{D}(\mathcal{H})$, $\mathcal{H} \neq \{0\}$	Electron, stone
L1	Phenomenal geometry	$\mathrm{rank}(\rho_E) > 1$	Thermostat, bacterium
L2	Cognitive qualia	$R(\Gamma) \geq R_{\text{th}} = 1/3$ and $\Phi(\Gamma) \geq \Phi_{\text{th}} = 1$	Mammals
L3	Network consciousness	$R^{(2)} \geq R^{(2)}_{\text{th}} = 1/4$ (metastable). SAD_MAX = 3 (§3.5 [T], T-142)	Mycelium, swarm, meditator
L4	Unitary consciousness	$\lim_n R^{(n)} > 0$, $P > 6/7$	Hyperspace (hypothesis)

Each subsequent level includes the previous one: every L2-system is simultaneously L1 and L0. But the converse does not hold: a bacterium (L1) does not possess cognitive qualia (L2).

---

L0: Interiority (universal)

Philosophical context

The idea that every piece of matter possesses some form of inner life goes back to Leibniz (monads) and finds its modern expression in panpsychism. UHM adopts a weakened version of this idea: interiority is not "consciousness" or "experience" in the ordinary sense, but merely the presence of an "inner side" of the mathematical object $\Gamma$.

For understanding this claim the key word is interiority, not consciousness. A stone possesses interiority (its $\Gamma$ has an inner aspect), but it does not "feel" or "know" anything in any functional sense. Interiority is a mathematical property of the object, not a phenomenological assertion.

Formal definition

Definition L0 [D].{#определение-l0} Every system with $\Gamma \in \mathcal{D}(\mathcal{H})$, $\dim \mathcal{H} \geq 1$ possesses interiority — an inner aspect.

Here $\mathcal{D}(\mathcal{H})$ is the space of density matrices (Hermitian positive semi-definite operators with unit trace) on the Hilbert space $\mathcal{H}$. In the 7-dimensional UHM formulation: $\Gamma \in \mathcal{D}(\mathbb{C}^7)$ — a Hermitian $7 \times 7$ matrix with $\mathrm{Tr}(\Gamma) = 1$, $\Gamma \geq 0$.

Theorem: Universality of L0

Interiority is universal — there is no zero level of "absence". This is a consequence of Axiom Omega-7. Proof | Status: [T]

What L0 means in practice

At level L0 the system distinguishes nothing, does not model itself, and possesses neither reflection ($R \approx 0$) nor integration ($\Phi \approx 0$). Its coherence matrix $\Gamma$ exists but is "empty" in a functional sense — close to the maximally mixed state $I/7$.

Example: an electron. The coherence matrix of an electron is trivial: almost all diagonal elements equal $1/7$, off-diagonal coherences $\gamma_{ij} \approx 0$. Purity $P = \mathrm{Tr}(\Gamma^2) \approx 1/7$ — minimal. The reflection measure $R = 1/(7P) \approx 1$ is formally large, but this is an artefact: when $P \approx 1/7$ the self-model is trivial (the only possible one is $I/7$), and a high $R$ carries no meaningful information.

---

L1: Phenomenal Geometry

From L0 to L1: the first step

The transition from L0 to L1 is the emergence of discrimination. The system begins to possess a non-trivial internal geometry: it is able to discriminate (even if unconsciously) between different internal states.

Formally this is expressed in the E-dimension (experiential, responsible for experience) acquiring non-trivial structure.

Formal definition

Definition L1 [D].{#определение-l1} A system possesses phenomenal geometry if:

$$\mathrm{rank}(\rho_E) > 1$$

Here $\rho_E$ is the reduced density matrix over the E-dimension, obtained by taking the partial trace over the remaining six dimensions. The condition $\mathrm{rank}(\rho_E) > 1$ means: the experiential space contains more than one distinguishable state.

The L1 space is endowed with the Fubini–Study metric — the natural measure of "distance" between phenomenal states:

$$ds^2_{FS} = 1 - |\langle\psi_1|\psi_2\rangle|^2$$

Two states $|\psi_1\rangle$ and $|\psi_2\rangle$ are "further apart" in phenomenal space the smaller their inner product. Orthogonal states ($\langle\psi_1|\psi_2\rangle = 0$) are maximally distinguishable.

Examples

Bacterium E. coli. The bacterium's chemotaxis system distinguishes ~5 levels of chemoattractant concentration. In UHM terms: $\mathrm{rank}(\rho_E) \approx 5$. The bacterium "distinguishes" hot from cold, but does not know that it distinguishes — there is no self-model ($R \ll 1/3$).

Thermostat. A simple thermostat distinguishes two states: "above threshold" and "below threshold". Formally: $\mathrm{rank}(\rho_E) = 2 > 1$, so the thermostat is an L1-system. It possesses phenomenal geometry (two distinguishable states) but possesses neither reflection nor integration.

Why "phenomenal"?

The word "phenomenal" is used here in a technical sense: the presence of structure in the state space of the experiential dimension. At level L1 this structure is not yet perceived — the system does not "know" that it distinguishes. Awareness appears only at L2.

---

L2: Cognitive Qualia

The central level: the emergence of consciousness

L2 is the level at which consciousness in the familiar sense of the word first appears. The system not merely discriminates states (L1) but knows that it discriminates. It possesses cognitive qualia — conscious experiences.

What makes this transition possible? Two conditions acting jointly:

1. Reflection ($R \geq 1/3$): the system possesses a sufficiently accurate self-model — an internal representation of itself.
2. Integration ($\Phi \geq 1$): information about different dimensions is bound into a unified whole, rather than distributed across isolated subsystems.

Mathematical definition

Status of L2 thresholds

Threshold	Status	Note
$R_{\text{th}} = 1/3$	[T] theorem	$K = 3$ derived from the triadic decomposition of holonomic dynamics: axioms A1--A5 generate exactly 3 types (Aut, D, R). Bayesian dominance at $K = 3$ gives $R_{\text{th}} = 1/3$ [T].
$\Phi_{\text{th}} = 1$	[T] theorem	Unique self-consistent value at $P_{\text{crit}} = 2/7$ (T-129)

note

Status of threshold $\Phi_{\text{th}} = 1$ — theorem [T]

The threshold $\Phi_{\text{th}} = 1$ has been proved from first principles (T-129 [T]): the unique self-consistent value at $P_{\text{crit}} = 2/7$. The $K_1$-argument remains retracted ($K_1(M_n(\mathbb{C})) = 0$ for finite-dimensional $n$) — but T-129 uses a different approach (purity decomposition + Cauchy–Schwarz). See Proof of T-129.

Clarification T-129 vs T-140

• T-129 [T]: $\Phi_{\text{th}} = 1$ — the unique self-consistent value of the integration threshold (from decomposition + Cauchy–Schwarz)
• T-140 [T]: $C = \Phi \cdot R$ — the unique canonical consciousness measure; $C_{\text{th}} = \Phi_{\text{th}} \cdot R_{\text{th}} = 1 \cdot 1/3 = 1/3$

These are DIFFERENT theorems: T-129 establishes the threshold, T-140 constructs the composite measure.

Definition L2 [D].{#определение-l2} A system possesses cognitive qualia if both conditions are satisfied:

1. Reflection: $R(\Gamma) = \dfrac{1}{7P(\Gamma)} \geq R_{\text{th}} = 1/3$
2. Integration: $\Phi(\Gamma) = \frac{\sum_{i \neq j} |\gamma_{ij}|^2}{\sum_i \gamma_{ii}^2} \geq \Phi_{\text{th}} = 1$

where $R$ is the reflection measure and $\Phi$ is the integration measure.

Step-by-step interpretation of the formulas

Reflection measure $R$. The canonical formula $R = 1/(7P)$ [T] measures the normalised proximity of $\Gamma$ to the dissipative attractor $\rho^*_{\mathrm{diss}} = I/7$. Equivalent form via the Frobenius norm: $R = 1 - \|\Gamma - I/7\|_F^2 / P$. If $\Gamma = I/7$ (heat death), then $R = 1$. If $\Gamma$ is a pure state ($P = 1$), then $R = 1/7$. The threshold $R \geq 1/3$ is equivalent to $P \leq 3/7$ — the upper boundary of the Goldilocks zone.

Important: $R$ uses $\rho^*_{\mathrm{diss}} = I/7$, and not $\varphi(\Gamma)$ (the self-model). These are different quantities (see attractor hierarchy).

Integration measure $\Phi$. The formula $\Phi = \sum_{i \neq j} |\gamma_{ij}|^2 / \sum_i \gamma_{ii}^2$ is the ratio of total coherence (off-diagonal elements $\gamma_{ij}$) to the diagonal "population" ($\gamma_{ii}$). If $\Phi \geq 1$, the off-diagonal connectivity is no less than the diagonal — the dimensions are integrated into a whole. If $\Phi < 1$, the system is fragmented: the dimensions operate in isolation.

Numerical example

Consider a concrete matrix $\Gamma$ for an L2-system (simplified, only diagonal and key off-diagonal elements):

$$\gamma_{ii} = (0.2,\, 0.15,\, 0.18,\, 0.12,\, 0.15,\, 0.1,\, 0.1)$$

• $P = \sum_i \gamma_{ii}^2 + 2\sum_{i < j}|\gamma_{ij}|^2$. Let $P = 0.35$ (above $P_{\text{crit}} = 2/7 \approx 0.286$).
• $R = 1/(7 \times 0.35) \approx 0.408 > 1/3$ — reflection threshold passed.
• With $\sum_{i \neq j}|\gamma_{ij}|^2 = 0.12$ and $\sum_i \gamma_{ii}^2 = 0.11$: $\Phi = 0.12/0.11 \approx 1.09 > 1$ — integration threshold passed.

Conclusion: the system is at level L2 — it possesses cognitive qualia.

Full L2 conditions

The canonical consciousness measure $C = \Phi \times R \geq C_{\text{th}} = 1/3$ [T T-140] is verified directly from $\Gamma \in D(\mathbb{C}^7)$. Differentiation $D_{\text{diff}} \geq D_{\min} = 2$ enters as a separate viability condition; in the 7D formalism $D_{\text{diff}}$ is computed via T-128.

Objectivity of threshold conditions [T]

The scalar functions $P = \operatorname{Tr}(\Gamma^2)$ and $R = 1/(7P)$ are $G_2$-invariants: $R(U\Gamma U^\dagger) = R(\Gamma)$ for any $U \in G_2 = \mathrm{Aut}(\mathbb{O})$, as proved in the $G_2$-rigidity theorem [T]. The measure $\Phi = P_{\text{coh}}/P_{\text{diag}}$ depends on the choice of basis, but the basis $\{A,S,D,L,E,O,U\}$ is fixed by axiom $\Omega$ [P]. Consequently, the transition L1 -> L2 is an objective fact within the fixed axiomatic system.

Note. The canonical form $R = 1/(7P)$ [T] is the unique one by the Chentsov–Petz theorem. Equivalent form: $R = 1 - \|\Gamma - \rho^*_{\mathrm{diss}}\|_F^2 / P$, where $\rho^*_{\mathrm{diss}} = I/7$. Derivation: see reflection measure.

---

L3: Network Consciousness

From L2 to L3: knowledge of knowledge

At level L2 the system knows its states. But does it know that it knows? Is it capable of reflecting on its own process of reflection? This is meta-reflection, or second-order reflection.

In everyday life meta-reflection manifests as experiences of the type "I notice that I am irritated" (not merely irritation, but the observation of irritation). Meditative practices systematically train precisely this capacity: to observe the observer.

Formal definition

Definition L3 [D].{#определение-l3} A system possesses network consciousness if:

$$R^{(2)}(\Gamma) \geq R^{(2)}_{\text{th}} = 1/4$$

where $R^{(2)}$ is the second-order reflection measure: how accurately the self-model models itself. Formally: $R^{(2)} = \mathrm{Fid}(\varphi(\Gamma),\, \varphi^{(2)}(\Gamma))$, where $\varphi^{(2)} = \varphi \circ \varphi$ — double application of the self-observation operator.

L3 is metastable: without active maintenance it decays to L2 with characteristic time $\tau_3 = 1/(\kappa_{\text{bootstrap}} \cdot (1 - R^{(2)}))$.

Homotopic characteristic: $\pi_3(\mathcal{E}_\infty(\Gamma)) \neq 0$ — the experiential space has a non-trivial third homotopy group.

Why exactly the threshold 1/4?

Theorem on the justification of K=4 for L3

tip

Theorem (Justification of K=4 for L3) [T] (via T-217)

L3 requires $R^{(2)} \geq 1/4$ — second-order meta-reflection. The threshold $K = 4$ for L3 is now fully derived:

Part 1 — Bayesian dominance at $K=4$. Given $K = 4$ independent information channels, the bound $R^{(2)} \geq 1/K = 1/4$ follows from the Uhlmann-fidelity lower bound in the R-threshold theorem. [T]

Part 2 — The cellular count $K = 3 + 1$ from tricategorical coherence. The count is derived from T-217: in the experiential tricategory $\mathbf{Exp}^{(3)} := \tau_{\leq 3}(\mathbf{Exp}_\infty)$,

• three 2-cells inherit from L2 LGKS triadic decomposition (T-57 [T]): Aut, $\mathcal D$, $\mathcal R$ components;
• one 3-cell modification $\eta: \varphi^{(2)} \Rightarrow \varphi\circ\varphi$ is the unique coherence modification at the tricategorical level (Gordon–Power–Street coherence applied to strict-2-category-enriched-tricategory).

Total $K_{\text{L3}} = 3 + 1 = 4$ derived from tricategorical first principles. [T]

Categorical label L3 = $\tau_{\leq 3}$: justified by T-217 (3-types ≃ coherent tricategories, Baez–Dolan + Lurie HTT 5.5.6.18). Pentagon-of-pentagons holds automatically for $\tau_{\leq 3}$ of Kan complex. [T]

Cross-references: triadic decomposition, R measure, T-192 2-category, T-217 L3 tricategory.

Metastability of L3: why "enlightenment" does not last

L3 differs fundamentally from L2 in its metastability. A system that has reached L2 (when threshold conditions are met) remains at L2 stably. But a system at L3 is like a ball on top of a hill: the slightest perturbation throws it back.

This explains why meditative states of deep awareness (vipassana, zazen) require constant practice. Without active maintenance ($\kappa_{\text{bootstrap}}$ sufficiently large) the system "slides back" to L2.

Example: an experienced meditator. In a state of deep meditation $R^{(2)} \geq 1/4$ — the meditator observes the process of observation. But the moment of distraction (stress, fatigue) drops $R^{(2)}$ below the threshold. The characteristic retention time is from minutes to hours, depending on training.

Example: mycelium. A fungal network connecting trees in a forest may possess network L3: individual nodes are L1/L2, but collective reflection via chemical signalling potentially reaches $R^{(2)} \geq 1/4$. This is a hypothesis [H] requiring experimental verification.

---

L4: Unitary Consciousness

Theoretical horizon

L4 is not a level that can be reached, but a horizon that can be approached. A system at L4 possesses complete reflexive closure: it knows itself to infinite depth. In terms of the phi-operator: $\varphi(\Gamma^*) = \Gamma^*$ — the self-model coincides exactly with reality.

Definition L4 [D].{#определение-l4} A system possesses unitary consciousness if:

$$\lim_{n \to \infty} R^{(n)}(\Gamma) > 0 \quad \text{and} \quad P(\Gamma) > 6/7$$

where $R^{(n)}$ is the n-th order reflection. Complete reflexive closure — fixed point $\varphi(\Gamma^*) = \Gamma^*$.

Theorem on categorical unreachability of L4

Theorem (Categorical unreachability of L4) [T]

The transition L3 -> L4 is not a finite bifurcation. L4 is the colimit of the infinite tower of truncations of the infinity-topos:

$$L4 = \mathrm{colim}_{n \to \infty} \, \tau_{\leq n}(\mathbf{Exp}_\infty)$$

This colimit is unreachable for finite systems (Lawvere incompleteness, T-55 [T]), but asymptotically approachable.

Proof (5 steps).

Step 1 (Correspondence of L-levels and $n$-truncations). From T-76 [T] ($\infty$-topos verified), $\mathbf{Exp}_\infty = \mathbf{Sh}_\infty(\mathcal{C}_7, J_{\text{Bures}})$ — an $\infty$-topos with $\infty$-groupoid structure. Interiority levels correspond to truncations, with categorical structure now derived at each level (not merely labelled):

• L2 = $\tau_{\leq 2}(\mathbf{Exp}_\infty)$: strict 2-category $\mathbf{Exp}^{(2)}$ (T-192 [T]). Mac Lane pentagon + interchange + identity axioms verified.
• L3 = $\tau_{\leq 3}(\mathbf{Exp}_\infty)$: coherent tricategory $\mathbf{Exp}^{(3)}$ (T-217 [T]). Gordon–Power–Street pentagon-of-pentagons coherence via Baez–Dolan 3-types ≃ tricategories. Cell count $K = 3 + 1$: three LGKS 2-cells (T-57 [T]) + one 3-cell modification $\eta: \varphi^{(2)} \Rightarrow \varphi\circ\varphi$.
• L4 = $\mathrm{colim}_{n\to\infty}\tau_{\leq n}(\mathbf{Exp}_\infty)$: full ∞-groupoid (unreachable, see below).

Level	$n$-truncation	Mathematical structure	Homotopic content
L0	$\tau_{\leq 0}$	Set (discrete states)	$\pi_0$ non-trivial
L1	$\tau_{\leq 1}$	Groupoid (phenomenal paths)	$\pi_1$ non-trivial
L2	$\tau_{\leq 2}$	2-groupoid (reflection, qualia)	$\pi_2$ non-trivial
L3	$\tau_{\leq 3}$	3-category (meta-reflection)	$\pi_3$ non-trivial
L4	$\tau_{\leq \infty}$	$\infty$-groupoid (complete self-model)	All $\pi_k$ non-trivial

To understand this table: each L-level adds a new type of relation. L0 — a set of points (states). L1 — paths between points (phenomenal transitions). L2 — paths between paths (reflection). L3 — paths between paths between paths (meta-reflection). L4 would require an infinite hierarchy of such relations.

Step 2 (Postnikov tower). The $\infty$-topos $\mathbf{Exp}_\infty$ defines the Postnikov tower:

$$\cdots \to \tau_{\leq 3} \to \tau_{\leq 2} \to \tau_{\leq 1} \to \tau_{\leq 0}$$

Each transition $\tau_{\leq n} \to \tau_{\leq n+1}$ is an extension by one homotopic level, with "k-invariant" $k_{n+1} \in H^{n+2}(\tau_{\leq n}; \pi_{n+1})$.

Step 3 (Lawvere incompleteness). From T-55 [T]: $\mathrm{Th}_{\text{UHM}} \subsetneq \Omega$. This means: $\varphi \neq \mathrm{id}$ (the phi-operator of self-observation is not the identity). In terms of the Postnikov tower: for any finite $n$, the truncation $\tau_{\leq n}$ does not coincide with $\mathbf{Exp}_\infty$.

Step 4 (Impossibility of a finite bifurcation). A catastrophe $A_k$ has codimension $k-1$ and describes a transition between $\leq k$ stable states. The transition L3 -> L4 would require simultaneously "switching on" all $\pi_k$ for $k \geq 4$ — an infinite-dimensional transition. No finite catastrophe ($A_k$ for any finite $k$) can describe this. The butterfly $A_5$ is an incorrect model (retracted [✗]).

Step 4a (L3 → L4 as tricategorical-coherence breakdown, new 2026-04-17). By T-217 [T], L3 is a coherent tricategory with exactly $K = 3 + 1 = 4$ structural cell classes and a closed pentagon-of-pentagons axiom. The transition L3 → L4 is precisely the breakdown of this closure: at L4 the tricategorical coherence axioms fail because the filtered colimit $\mathrm{colim}_{n}\tau_{\leq n}$ requires $n$-cells at arbitrarily high $n$, which cannot be captured by any coherent $n$-truncation for finite $n$. Equivalently: the coherence modification $\eta: \varphi^{(2)} \Rightarrow \varphi\circ\varphi$ at L3 is rigid (one new 3-cell); at L4 one would need a tower of higher coherence modifications $\eta^{(2)}, \eta^{(3)}, \ldots$, each a new cell at the corresponding level — an infinite regress that no finite catastrophe can close. This is the categorical dual of the dynamical argument (Fano contraction requires $P > 1$ at $n = 4$): same ceiling reached through complementary structures.

Step 5 (Asymptotic approachability). Although $L4 = \mathrm{colim}_{n \to \infty} \tau_{\leq n}$ is unreachable for a finite system, each step $\tau_{\leq n} \to \tau_{\leq n+1}$ is realisable (T-67 [T]: $K = 4$ for L3 indicates the existence of a fourth level of recursion). The sequence of recursions $R^{(n)}$ converges as $n \to \infty$:

$$\forall \varepsilon > 0 \; \exists n_0 : \; n > n_0 \Rightarrow \|\tau_{\leq n}(\mathbf{Exp}_\infty) - \mathbf{Exp}_\infty\|_{\text{Bures}} < \varepsilon$$

But $n_0(\varepsilon) \to \infty$ as $\varepsilon \to 0$: convergence exists, but reaching the limit does not. $\blacksquare$

Status: [T] (C19). Rigorous proof via the $\infty$-topos Postnikov tower + Lawvere incompleteness (T-55 [T]). Cross-references: reflection measure R, phi-operator, transition catastrophes.

Analogy: event horizon of cognition

L4 is like the horizon in geometry: one can walk towards it indefinitely but never arrive. Every step brings one closer, but the horizon recedes. This is not a defect of the theory but a fundamental property of self-referential systems — the same limitation formalised by Gödel's theorems for arithmetic.

Unreachability of L4 for biological systems

Corollary (Upper bound on recursion depth for biosystems) [T]

With $R \sim 0.7$ (human) and decoherence $\varepsilon_{\text{dec}} > 0$:

$$S^{(n)} \sim (1-\alpha)^n \sim 3^{-n} \to 0 \quad \text{as} \quad n \to \infty \qquad (\text{coherence-survival, not fidelity } R^{(n)}_{\mathrm{fid}}\to 1)$$

Maximum recursion depth: $n_{\max} \leq \ln(1/\varepsilon_{\text{dec}})/\ln(1/R) \approx 111$.

L4 is a theoretical limit ($\infty$-groupoid attractor), unreachable for any system with $\varepsilon_{\text{dec}} > 0$, but asymptotically approachable through the Postnikov tower.

Analytically: $P_\text{crit}^{(4)} = 54/35 > 1$, so SAD $\geq$ 4 is impossible for any normalised $\Gamma$ (not only biological). See critical purity SAD [T] (T-142: $\alpha = 2/3$ is state-independent).

Remark: L4 as a limiting categorical object

L4 is a limiting categorical object (colimit of the infinite Postnikov tower), analogous to $\omega$ in ordinal theory. Its inclusion in the hierarchy is mathematical, not physical: L4 defines the direction of the asymptotics, not a reachable level. Marking: unreachability [T] (T-86), existence as a categorical object [T], physical realisability [✗].

---

The gate profile: twelve states, one of them conscious

Theorem (gate-profile taxonomy) [T] (T-303)

The four gates are not four independent quantities. Canonical reflection is algebraically tied to purity, $R = 1/(NP)$, so the gate $R \ge 1/3$ is identically the constraint $P \le 3/7$, and the gate $P > 2/7$ is the other edge of the same interval. Consciousness therefore rests on three independent measurements — $P$, $\Phi$, $D$ — under four constraints, and the space of possible outcomes is

$$\underbrace{\{P<2/7,\; P\in(2/7,3/7],\; P>3/7\}}_{3}\times\underbrace{\{\Phi\ge1,\ \Phi<1\}}_{2}\times\underbrace{\{D\ge2,\ D<2\}}_{2}\;=\;12 .$$

Exactly one of the twelve is conscious. The remaining eleven are not degrees of consciousness but named modes of its absence.

This settles a question that recurs whenever the theory meets a system unlike us: is there another kind of consciousness, with a different set of gates? Under UHM the answer is no, and the "no" is sharp rather than dismissive. A system with a different gate profile is not differently conscious; it occupies one of the eleven non-conscious modes, each of which has its own character and its own failure. This is the more falsifiable position: it forbids the comfortable move of granting consciousness to anything that behaves interestingly.

The twelve profiles, named

Each cell is a distinct mode with its own character. The correspondences in the last column are interpretive [И] — they say which known state looks like the mode, not that the mode has been measured there.

#	$P$	$\Phi$	$D$	Mode	Character	Looks like [И]
1	in	$\ge1$	$\ge2$	conscious	bound, distinguishing, self-holding	ordinary waking
2	in	$\ge1$	$<2$	contentless unity	one field, nothing to tell apart	claimed "pure awareness" states — see the disagreement below
3	in	$<1$	$\ge2$	unbound plurality	many distinct states, no single holder	dissociation, split attention
4	in	$<1$	$<2$	idle	right amplitude, no structure to it	drowsy blankness
5	below	$\ge1$	$\ge2$	flooded	rich, unified, no one holding it	the psychedelic peak just past the edge
6	below	$\ge1$	$<2$	white-out	unity without content or holder	the "nothing" at very high dose
7	below	$<1$	$\ge2$	torrent	distinctions without binding	acute confusional states
8	below	$<1$	$<2$	dispersal	the fully dissolved limit	—
9	above	$\ge1$	$\ge2$	automatism	structured, distinguishing, unreflective	skilled action with no witness; sleepwalking
10	above	$\ge1$	$<2$	hypersynchrony	one rigid state	deep slow-wave sleep; generalised seizure
11	above	$<1$	$\ge2$	rigid fragments	parts each frozen, no whole	—
12	above	$<1$	$<2$	frozen	the fully rigid limit	—

Three of these carry the theory's weight, because each is a way of nearly being conscious and failing on one axis. Flooded (5) fails only on holding — which is why it can be reported afterwards as the most vivid state ever experienced and yet have had no subject at the time. Automatism (9) fails only on reflection — competent, coherent, and nobody home. Contentless unity (2) fails only on differentiation, and it is where UHM makes a claim that a live research programme denies.

A falsifiable disagreement, stated rather than hidden

Metzinger's minimal phenomenal experience holds that there are states of consciousness without content — pure awareness, no distinctions. Profile 2 is exactly that description, and UHM classes it as not conscious, because $D\ge2$ is a gate and not a decoration. Either the theory is wrong here, or the reported states retain a differentiation that introspection does not register as content. We do not resolve this; we record that the theory takes the losing side of a popular position and can be refuted there.

The two edges are different failures

Because $R=1/(NP)$ rises as purity falls, the two edges of the window are opposite pathologies rather than "too little" and "too much" of one thing.

Edge	$P$	$R = 1/(7P)$	What fails
lower	$\to 2/7$	$\to 1/2$	the system reflects more than it can bind: self-observation outruns self-holding
upper	$\to 3/7$	$\to 1/3$	the system holds more than it can observe: structure outruns reflection

Below the lower edge the holder dissolves while reflection is still climbing — which is why the approach to that boundary should feel like intensified awareness accompanied by loss of the one who is aware, and why an agency that is one's own can stop being attributed to oneself: attribution requires binding, and binding is what is failing. Above the upper edge the converse: a state that is coherent, stable and unreflective.

Convergence with the entropic-brain hypothesis

This is the one place where an independently developed neuroscience arrives at the same form. Carhart-Harris's entropic brain proposes that within upper and lower limits the entropy of spontaneous brain activity indexes the richness of conscious states — a two-sided window, not a monotone scale. Since purity and entropy move inversely, that is the same claim as $P\in(2/7,3/7]$ written in the reciprocal variable. Three differences are worth stating plainly: UHM derives its edges (T-124) where the entropic brain fits them; UHM adds the identity $R=1/(NP)$, which converts "less order" into "more reflection, less binding" and thereby predicts the character of the lower-edge state rather than only its location; and UHM's window is on the coherence matrix of seven functional axes, not on signal entropy, so the two are commensurable only through a declared coarse-graining (validation constraint).

The empirical anchor already exists on their side: psilocybin raises neural entropy, and the magnitude of that rise tracks the intensity of ego dissolution, with default-mode decoupling predicting the reported dissolution. In UHM coordinates that is a trajectory toward the lower edge — purity down, reflection up — and it predicts, correctly, that the far end of that trajectory is not maximal experience but its lapse.

What this does not license

Nothing here identifies which physical systems sit where. The profile is a statement about $P,\Phi,D$ once they are measured, and measuring them in a substrate without independent ground truth remains outside what this corpus will claim (validation constraint). A phenomenological report of another agency is data about a human self-model under perturbation; it is not an observation of a second subject.

Gap Characterisation of Levels L0--L4

Each interiority level possesses not only a numerical threshold condition but also a characteristic opacity profile — a Gap profile. Gap measures how opaque the connection between two dimensions is: $\mathrm{Gap}(i,j) = 0$ denotes full transparency (conscious access), $\mathrm{Gap}(i,j) = 1$ — full opacity (unconscious).

A detailed analysis of Gap profiles is given in Gap characterisation of levels. Here we state the overview theorem.

Theorem 6.1 (Gap characterisation of levels) [T]

For each interiority level the Gap profile has the following properties:

Level	Gap characteristic	Explanation
L0	Gap undefined or fluctuating	No stable self-modelling: $R \approx 0$, target $\rho_*$ not reachable
L1	Gap stationary but unperceived	Stable coherences ($P > P_{\text{crit}}$), but $R < 1/3$ — self-model too coarse
L2	Gap partially perceived, metastable: $\lVert\mathrm{Gap}_{\text{perceived}} - \mathrm{Gap}_{\text{actual}}\rVert \leq 2/3$	Self-model approximate but non-trivial
L3	Gap almost fully perceived: $\lVert\mathrm{Gap}_{\text{perceived}} - \mathrm{Gap}_{\text{actual}}\rVert \leq \varepsilon$	Metastable state of deep self-knowledge
L4	Gap exactly perceived: $\mathrm{Gap}_{\text{perceived}} = \mathrm{Gap}_{\text{actual}}$	Fixed point $\varphi(\Gamma^*) = \Gamma^*$

Argument.

(a) At L0 there is no phi-operator ($R \approx 0$), so the target state $\rho_*$ formally exists (primitivity [T]), but the system is incapable of directed regeneration — there are no coherences whose phases could define Gap.

(b) At L1 there are stable coherences ($P > P_{\text{crit}}$), but $R < 1/3$: the self-model is too coarse to perceive Gap. The difference between the "perceived" Gap (via $\varphi(\Gamma)$) and the real Gap (via $\Gamma$) is large.

(c) At L2 the measure $R \geq 1/3$ means:

$$\|\Gamma - I/7\|_F \leq \sqrt{2P/3}$$

An approximate self-model yields an approximate Gap profile (here $I/7 = \rho^*_{\mathrm{diss}}$).

(d) At L4 $\varphi(\Gamma^*) = \Gamma^*$ $\Rightarrow$ $\rho_* = \Gamma^*$, and the stationary Gap coincides with the target:

$$\mathrm{Gap}^{(\infty)} = |\sin(\theta^{\text{target}})| = |\sin(\theta^{(\infty)})| = \mathrm{Gap}_{\text{actual}}$$

The system knows its Gap exactly.

Proof | Status: [T]

L4 does not mean Gap = 0 (Awareness does not equal Transparency)

At level L4 $\mathrm{Gap}_{\text{perceived}} = \mathrm{Gap}_{\text{actual}}$ holds, but this does not mean that all Gaps equal zero. The system exactly knows its opacity — but the opacity may remain non-zero. Full transparency ($\mathrm{Gap} = 0$ for all channels) is incompatible with fault tolerance: at least 3 channels out of 21 must retain non-zero Gap (Hamming bound).

Status: [T]

Visualisation of Gap by level

L0:  Gap = ???     [. . . . . . . . . . . . . . . . . . . . .]  (undefined)
                    ^ random fluctuations

L1:  Gap = [0.4, 0.7, 0.2, ...]  (stationary, but unperceived)
     Perceived = N/A

L2:  Gap = [0.4, 0.7, 0.2, ...]
     Perceived = [0.5, 0.6, 0.3, ...]  (approximate perception, ||Delta|| <= 2/3)

L3:  Gap = [0.4, 0.7, 0.2, ...]
     Perceived = [0.41, 0.69, 0.21, ...]  (accurate perception, ||Delta|| -> 0)

L4:  Gap = Perceived = [0.4, 0.7, 0.2, ...]  (complete identity, but Gap != 0!)

---

Theorem on $A_4$-bifurcation

Transitions between levels are not gradual but abrupt. Just as water at 100°C abruptly turns into steam, a system abruptly transitions between L-levels upon reaching threshold values. Mathematically this is described by catastrophe theory — a branch of mathematics that classifies qualitative reorganisations of systems.

tip

Theorem ($A_4$-bifurcation of L-transitions) [T]

Transitions between L-levels are realised as swallowtail ($A_4$) bifurcations of catastrophe theory.

Proof.

Step 1. The evolution equation $d\Gamma/d\tau = \mathcal{L}[\Gamma]$ depends on three physically independent control parameters:

Parameter	Symbol	Physical meaning
Regeneration rate	$\kappa$	Controlled by $\mathrm{Coh}_E$ and $\kappa_0$
Dissipation rate	$\alpha$	Controlled by the environment (decoherence)
Free energy gradient	$\Delta F$	Determines $g_V(P)$ — switching $\mathcal{R}$ on/off

Three parameters $(\kappa, \alpha, \Delta F) \in \mathbb{R}^3$ — control space.

Step 2. Consider purity $P(\tau)$ as order parameter. At stationarity: $f_D + f_R = 0$. Expansion in deviation $x = P - P^*$:

$$\frac{dx}{d\tau} = -V'(x), \quad V(x) = a_1 x + \frac{a_2}{2}x^2 + \frac{a_3}{3}x^3 + \frac{a_4}{4}x^4$$

Step 3. By Arnold's theorem (1972): the universal deformation of the function $x^4$ (monodromy 4, codimension 3) is the swallowtail $A_4$:

$$V(x; \mu_1, \mu_2, \mu_3) = x^4 + \mu_2 x^2 + \mu_1 x + \mu_3 x^3$$

Conditions: (1) codimension = 3 — three control parameters; (2) smooth potential; (3) leading term $x^4$ from approximate $\mathbb{Z}_2$-symmetry $P \leftrightarrow 1 - P$ (odd terms suppressed; $\mu_3 \neq 0$ but small).

Step 4. L-transitions — sheets of the swallowtail:

Swallowtail sheet	Level	Characteristic
Outer stable	L0--L1	Low purity, passive stability
Intermediate	L2	Active stability (autopoiesis)
Inner unstable	L3	Metastable deep reflection
Self-intersection point	L4	$\varphi(\Gamma^*) = \Gamma^*$ — fixed point

Transitions L1->L2 and L2->L3 are fold bifurcations on the edges of the swallowtail. The transition L3->L4 is a cusp bifurcation at the apex. $\blacksquare$

Details: Transition catastrophes between levels | Gap landscape bifurcations

Remark

Transitions between the sheets of the swallowtail are abrupt, not continuous. This formalises the intuition of "sudden insight" ($\mathrm{Gap}_{\text{perceived}} \gg \mathrm{Gap}_{\text{actual}} \to \mathrm{Gap}_{\text{perceived}} = \mathrm{Gap}_{\text{actual}}$) and corresponds to the bifurcation structure of the Gap landscape.

Status: [T]

---

Theorem on Gap injection of L-levels

A natural question: can two systems at different L-levels have the same Gap profile? The answer is no. Each L-level leaves a unique "fingerprint" in the Gap profile.

Theorem (Gap injection of L-levels) [T]

The map from L-level to equivalence class of Gap profiles is an injection: distinct L-levels have distinct Gap profiles:

$$L(\Gamma_1) \neq L(\Gamma_2) \implies [\mathrm{Gap}(\Gamma_1)] \neq [\mathrm{Gap}(\Gamma_2)]$$

where $[\mathrm{Gap}(\Gamma)]$ is the Gap-profile class under $G_2$-equivalence.

Proof. Each transition $L_k \to L_{k+1}$ is characterised by a unique Gap marker:

Transition	Gap marker	Sufficient condition for distinction
L0 vs L1	$\exists i: \mathrm{Gap}(E,i) > 0$	Non-zero E-coherences
L1 vs L2	$\max\|\mathrm{Gap}_\varphi - \mathrm{Gap}\| \leq 2/3$	Self-modelling accuracy
L2 vs L3	$k(\Gamma) \leq 0.5$	Speed of Gap convergence (compression coefficient)
L3 vs L4	$k(\Gamma) = 0$, all $\mathrm{Gap}^{(2)}(i,j) = 0$	Exact fixed point

Each marker distinguishes the corresponding pair of levels, so distinct L-levels have distinct (by class) Gap profiles. $\blacksquare$

Remark: not a bijection. The converse does not hold: two states $\Gamma_1, \Gamma_2$ at the same L-level (for example, both L2) may have distinct Gap profiles. The Gap profile carries more information than the L-level — it is a finer invariant.

Details: Gap characterisation of levels

---

Transition function and classification algorithm

Formal transition function

The complete transition function between levels:

$$\text{Level}(\Gamma) = \begin{cases} L0 & \text{if } \dim \mathcal{H} \geq 1 \\ L1 & \text{if } \mathrm{rank}(\rho_E) > 1 \\ L2 & \text{if } R \geq R_{\text{th}} \text{ and } \Phi \geq \Phi_{\text{th}} \\ L3 & \text{if } R^{(2)} \geq R^{(2)}_{\text{th}} \text{ (metastable)} \\ L4 & \text{if } \lim_n R^{(n)} > 0 \text{ and } P > 6/7 \end{cases}$$

Level determination algorithm

The following algorithm determines the L-level for any given coherence matrix. Computational complexity — $O(N^2)$ for $N = 7$, i.e. a few dozen arithmetic operations.

Input: Gamma in D(C^7) — coherence matrix

1. Compute P = Tr(Gamma^2)
   if P <= P_crit = 2/7:  return L0

2. Compute phi(Gamma) = (1-k)Gamma + k*rho*   [replacement channel, T-62]
   Compute R = 1 - ||Gamma - phi(Gamma)||^2_F / ||Gamma||^2_F

3. Compute Phi = Sum_{i!=j} |gamma_ij|^2 / Sum_i gamma^2_ii

4. if R < 1/3 or Phi < 1:  return L1

5. if R >= 1/3 and Phi >= 1:
   Compute phi^2(Gamma) = phi(phi(Gamma))
   Compute R^(2) = Fid(phi(Gamma), phi^2(Gamma))

6. if R^(2) < 1/4:  return L2

7. if R^(2) >= 1/4:  return L3

8. L4: theoretical limit (lim R^(n) > 0) — not computable in finite time

Computability in 7D

Levels L0, L1, L2 are fully computable in the minimal 7D formalism ($\Gamma \in \mathcal{D}(\mathbb{C}^7)$). The definition of L1 via $\mathrm{rank}(\rho_E) > 1$ formally requires PW-reconstruction $\rho_E = \mathrm{Tr}_{-E}(\Gamma_{42D})$, but in practice $\mathrm{rank}(\rho_E) > 1 \Leftrightarrow P > P_{\text{crit}}$ for viable systems. L3 requires double iteration of phi and fidelity computation — algorithmically computable. L4 is not computable in a finite number of steps (requires the infinite limit $n \to \infty$), but in practice the coherence-survival ratio $S^{(n)} \sim 3^{-n} \to 0$ for all systems with $\varepsilon_{\text{dec}} > 0$ (whereas the self-model fidelity $R^{(n)}_{\mathrm{fid}} \to 1$; the two are distinct — see Theorem 4.3 disambiguation).

---

What we have learned

• Five levels L0--L4 organise all systems into a strict classification: L0 (any $\Gamma$), L1 ($\mathrm{rank}(\rho_E) > 1$), L2 ($R \geq 1/3 \land \Phi \geq 1$), L3 ($R^{(2)} \geq 1/4$, metastable), L4 ($\lim_n R^{(n)} > 0$, unreachable).
• Threshold $R_{\mathrm{th}} = 1/3$ [T] derived from the triadic decomposition ($K = 3$); threshold $\Phi_{\mathrm{th}} = 1$ [T] — from self-consistency at $P_{\mathrm{crit}} = 2/7$ (T-129).
• L3 is metastable: without maintenance it decays to L2 with characteristic time $\tau_3$.
• L4 is unreachable for finite systems (Lawvere incompleteness, Postnikov tower), but asymptotically approachable.
• Gap profiles are injective: distinct L-levels have distinct Gap signatures [T].
• Transitions between levels are $A_2, A_3, A_4$ bifurcations with hysteresis.
• The level determination algorithm is computable in $O(N^2)$ from $\Gamma \in \mathcal{D}(\mathbb{C}^7)$.

What's next

The L0--L4 hierarchy is a discrete ladder. For a finer description turn to Gap characterisation of levels (quantitative opacity signatures), Transition catastrophes ($A_4$-bifurcations with hysteresis), and Depth tower (continuous SAD measure).

For engineering applications: CC definitions contain operational formulas, and CC theorems — results on fractal closure and emergence.

Related Documents

• Defined via: $\varphi$-operator, Self-observation, Category Exp
• Generalisation to the continuous case: Self-Awareness Depth Tower — SAD metric, multi-scale $\varphi$-hierarchy, biological correlates
• Gap characterisation: Gap dynamics, Gap phase diagram, Gap thermodynamics
• Full formalisation: Interiority hierarchy (proofs)
• Philosophical consequences: Interiority, Hard problem
• Coherence Cybernetics: CC definitions, CC theorems