This page holds a single chapter, for direct linking and citation. For the full reading experience — table of contents, dark/light mode, typography controls — see the full continuous version. Full continuous version →

#Chapter 5: The Mathematical Language of Consciousness — A Formal Construction of Information Dynamics

#5.0 Chapter Orientation

Chapter 4 accomplished a critical terminological foundation for us. We no longer use the term "information flow" to refer to vague psychological impressions, but have precisely translated it into measurable physical quantities: differential entropy quantifies the richness of content, transfer entropy characterizes directed causal coupling, and synergistic information captures the emergent moment when "the whole is greater than the sum of its parts."

Now, armed with these refined tools, we can return to the E-I-B-A dynamical map left by Chapter 3. Chapter 3 defined consciousness as a coupling process of four-dimensional information flow on the stage of the global electromagnetic field — this definition is precise, but it has not yet been translated into computable language. Without mathematical formalization, we cannot make predictions, simulations, or empirical tests. A theory of consciousness that remains solely at the conceptual level, no matter how elegant, cannot enter into dialogue with EEG/MEG data in the laboratory.

The task of this chapter is to transform the E-I-B-A model from a "philosophical conception" and "physical intuition" into a computable scientific model. To this end, we must assign rigorous mathematical definitions to each conceptual variable — assembling the terminology introduced in Chapter 4 into a complete set of dynamical equations.

Core Decision: Why Information Theory?

The global electromagnetic field is a continuous physical medium. In a continuous field, "the number of active patterns" is a poorly defined concept — like asking how many eddies are in a river. To measure the richness and organization of a continuous system, the most natural mathematical language is information theory:

Differentiation is measured by information entropy: How unpredictable and complex is the state of the field?
Integration is measured by transfer entropy and phase synchronization: How does information flow and coordinate between different regions of the field?

Therefore, all core variables of this chapter will be redefined using the differential entropy, transfer entropy, synergistic information, and Kuramoto order parameter already laid out in Chapter 4. They will no longer be metaphors, but variable symbols waiting to be substituted with empirical data.

Now, let us compose the first line of the score for that burning fire of relations.

#5.1 Information-Dynamic Definitions of Core Variables

The E-I-B-A model comprises four levels of variables: the state of a single network (first-order), the interaction of network pairs (second-order), the synergy of triadic contexts (third-order), and global unity (fourth-order). Each level has three corresponding variables:

Structural capacity $\phi$ : The theoretical information upper limit (potential) of that level under physiological constraints.
Active pattern $\phi^{\text{active}}(t)$ : The instantaneous information dynamics of that level at time t (current state).
Activity ratio $\phi^{\text{activeratio}}(t)$ : The ratio of the instantaneous value to the maximum value, characterizing the current operating efficiency of that level.

Rigorous mathematical definitions are given below one by one.

#5.1.1 First-Order $\phi_X$ : Information Complexity of a Single Network

Physical meaning: The information richness of the electromagnetic field state in the brain regions corresponding to network X ( $X \in \{E, I, B, A\}$ ).

Structural capacity $\phi_X$ : Defined as the maximum differential entropy of that network's continuous electromagnetic field under physiological limits:

$\phi_X \equiv H_{\max}(X)$

The maximum differential entropy is jointly determined by the number of neurons, synaptic diversity, and energy metabolism upper limit of that brain region. A brain region with denser neurons and richer synapses can exhibit higher potential complexity in its electromagnetic field. This variable captures theinnate physiological constraints on the richness of consciousness.

Active pattern $\phi_X^{\text{active}}(t)$ : Defined as the instantaneous differential entropy of that network at time t within a time window $\Delta t$ :

$\phi_X^{\text{active}}(t) \equiv H(\mathbf{s}_X(t))$

where $\mathbf{s}_X(t)$ is the continuous electromagnetic field time series vector of that brain region after source localization processing. For continuous random variables, differential entropy is defined as $H(\mathbf{s}) = -\int p(\mathbf{s}) \log p(\mathbf{s}) d\mathbf{s}$ . In a fixed reference frame, it quantifies the unpredictability of the signal: high entropy corresponds to rich, varied field dynamics; low entropy corresponds to monotonous, regular oscillations.

Activity ratio $\phi_X^{\text{activeratio}}(t)$ : Defined as the ratio of instantaneous entropy to maximum entropy:

$\phi_X^{\text{activeratio}}(t) = \frac{H(\mathbf{s}_X(t))}{H_{\max}(X)} \in [0, 1]$

This ratio can be understood as the current "utilization rate" of that network's physiological potential. In deep anesthesia, this value approaches zero; in the awake state, it maintains a moderate to high level.

#5.1.2 Second-Order $\phi_{X \to Y}$ : Causal Information Flow Between Networks

Physical meaning: The strength of directed causal interaction from network X to network Y. This is not mere correlation, but how much additional information X's past state provides for predicting Y's current state.

Structural capacity $\phi_{X \to Y}$ : Defined as the maximum transfer entropy from X to Y:

$\boxed{\phi_{X \to Y} \equiv TE_{\max}(X \to Y)}$

The maximum transfer entropy is determined by the physical bandwidth and impedance of the white matter fiber tracts connecting the two brain regions. It characterizes the physical limit of this anatomical pathway for information transmission.

Active pattern $\phi_{X \to Y}^{\text{active}}(t)$ : Defined as the instantaneous transfer entropy from X to Y at time t:

$\boxed{\phi_{X \to Y}^{\text{active}}(t) \equiv TE_{X \to Y}(t)}$

Transfer entropy is defined as:

$TE_{X \to Y} = I(Y_t ; X_{\text{past}} \mid Y_{\text{past}})$

That is: given the history of Y itself, how much additional information does the history of X provide for predicting the current state of Y? This variable directly captures the information-theoretic generalization of Granger causality in nonlinear systems. When $\phi_{E \to I}^{\text{active}}$ is high, "seeing the scene evokes the feeling" is occurring; when $\phi_{I \to E}^{\text{active}}$ is high, "what the mind contemplates, the ear hears" is occurring.

Activity ratio $\phi_{X \to Y}^{\text{activeratio}}(t)$ : Defined as Normalized Transfer Entropy (NTE):

$\boxed{\phi_{X \to Y}^{\text{activeratio}}(t) = \text{NTE}_{X \to Y}(t) = \frac{TE_{X \to Y}(t)}{TE_{\max}(X \to Y)} \in [0, 1]}$

#5.1.3 Third-Order $\phi_{X,Y,Z}$ : Synergistic Emergence of Triadic Contexts

Physical meaning: When three networks (e.g., E, I, B) are coupled simultaneously, the information that emerges as "the whole is greater than the sum of its parts" — the residual information that cannot be explained by any combination of pairwise interactions.

Structural capacity $\phi_{X,Y,Z}$ : Defined as the maximum synergistic information of that triad under physiological limits:

$\boxed{\phi_{X,Y,Z} \equiv Syn_{\max}(X, Y, Z)}$

Active pattern $\phi_{X,Y,Z}^{\text{active}}(t)$ : Defined as the instantaneous synergistic information extracted from data via Partial Information Decomposition (PID):

$\boxed{\phi_{X,Y,Z}^{\text{active}}(t) \equiv Syn(X, Y, Z; t)}$

Synergistic information is that portion of the total mutual information that can be attributed neither to any single source nor to any combination of pairwise sources. It is the precise mathematical characterization of "the whole is greater than the sum of its parts."

Activity ratio $\phi_{X,Y,Z}^{\text{activeratio}}(t)$ : Defined as normalized synergistic information:

$\boxed{\phi_{X,Y,Z}^{\text{activeratio}}(t) = \frac{Syn(X, Y, Z; t)}{Syn_{\max}(X, Y, Z)} \in [0, 1]}$

This definition completely replaces the arbitrary parameter $\kappa_{XYZ}$ that required manual fitting in the old version, making the computation of third-order integration strictly data-driven.

#5.1.4 Fourth-Order $\phi_{EIBA}$ : Global Unified Presence

Physical meaning: The degree to which the electromagnetic field oscillations of the four networks E, I, B, A achieve global phase locking as an indivisible whole. This corresponds to a pure "sense of presence" or "unity experience."

Structural capacity $\phi_{EIBA}$ : In the Kuramoto model, the theoretical limit of complete phase synchronization is a physical constant:

$\boxed{\phi_{EIBA} \equiv R_{\max} = 1}$

This means the potential for whole-brain phase synchronization is absolutely bounded — just as synchronization in any physical system has an upper limit.

Active pattern $\phi_{EIBA}^{\text{active}}(t)$ : Defined as the global phase synchronization order parameter at time t. Let $\theta_X(t)$ be the instantaneous phase of network X in the frequency band of interest (e.g., high $\gamma$ band 30-80 Hz):

$\boxed{\phi_{EIBA}^{\text{active}}(t) \equiv R_{EIBA}(t) = \left\| \frac{1}{4} \sum_{X \in \{E, I, B, A\}} e^{i\theta_X(t)} \right\|}$

This is a geometric measure on the complex plane, averaging the lengths of the four networks' phase vectors added together.

Activity ratio $\phi_{EIBA}^{\text{activeratio}}(t)$ : Since the denominator is 1, the active pattern itself is the ratio:

$\boxed{\phi_{EIBA}^{\text{activeratio}}(t) = R_{EIBA}(t) \in [0, 1]}$

When $R \to 1$ : the oscillations of all four dimensions are perfectly locked (as in the "oneness" of deep meditation). When $R \to 0$ : the four are completely desynchronized (as in deep anesthesia or extreme confusion).

#5.2 Matrix Description of System State

To compactly characterize the complete state of the entire system at time t, define three $4 \times 4$ matrices.

#5.2.1 Structural Capacity Matrix $\mathbf{\Phi}$

The "hardware blueprint" of the system — the theoretical upper limit of all information channels. The diagonal entries are $H_{\max}$ for each network, the off-diagonal entries are $TE_{\max}$ for each directed pathway.

\begin{aligned} \mathbf{\Phi}(t) &= \begin{bmatrix} \phi_{E}(t) & \phi_{E\to I}(t) & \phi_{E\to B}(t) & \phi_{E\to A}(t)\\ \phi_{I\to E}(t) & \phi_{I}(t) & \phi_{I\to B}(t) & \phi_{I\to A}(t)\\ \phi_{B\to E}(t) & \phi_{B\to I}(t) & \phi_{B}(t) & \phi_{B\to A}(t)\\ \phi_{A\to E}(t) & \phi_{A\to I}(t) & \phi_{A\to B}(t) & \phi_{A}(t) \end{bmatrix}\\ &= \begin{bmatrix} H_{\max}(E) & TE_{\max}(E \to I) & TE_{\max}(E \to B) & TE_{\max}(E \to A)\\ TE_{\max}(I \to E) & H_{\max}(I) & TE_{\max}(I \to B) & TE_{\max}(I \to A)\\ TE_{\max}(B \to E) & TE_{\max}(B \to I) & H_{\max}(B) & TE_{\max}(B \to A)\\ TE_{\max}(A \to E) & TE_{\max}(A \to I) & TE_{\max}(A \to B) & H_{\max}(A) \end{bmatrix} \end{aligned}

$\phi_{X \to Y} \neq \phi_{Y \to X}$

#5.2.2 Activity Ratio Matrix $\mathbf{\Phi}^{\text{activeratio}}(t)$

\mathbf{\Phi}^{\text{activeratio}}(t) = \begin{bmatrix} \frac{H_E(t)}{H_{\max}(E)} & \text{NTE}_{E \to I}(t) & \text{NTE}_{E \to B}(t) & \text{NTE}_{E \to A}(t)\\ \text{NTE}_{I \to E}(t) & \frac{H_I(t)}{H_{\max}(I)} & \text{NTE}_{I \to B}(t) & \text{NTE}_{I \to A}(t)\\ \text{NTE}_{B \to E}(t) & \text{NTE}_{B \to I}(t) & \frac{H_B(t)}{H_{\max}(B)} & \text{NTE}_{B \to A}(t)\\ \text{NTE}_{A \to E}(t) & \text{NTE}_{A \to I}(t) & \text{NTE}_{A \to B}(t) & \frac{H_A(t)}{H_{\max}(A)} \end{bmatrix}

The "instantaneous efficiency map" of the system — the utilization rate of each channel at the current moment. Diagonal entries are first-order activity ratios, off-diagonal entries are normalized transfer entropies. All elements naturally fall within the $[0, 1]$ interval.

#5.2.3 Absolute Activity Matrix $\mathbf{\Phi}^{\text{active}}(t)$

\begin{aligned} \mathbf{\Phi}^{\text{active}}(t) &= \mathbf{\Phi}(t) \odot \mathbf{\Phi}^{\text{activeratio}}(t)\\ &= \begin{bmatrix} \phi^{\text{active}}_{E}(t) & \phi^{\text{active}}_{E\to I}(t) & \phi^{\text{active}}_{E\to B}(t) & \phi^{\text{active}}_{E\to A}(t)\\ \phi^{\text{active}}_{I\to E}(t) & \phi^{\text{active}}_{I}(t) & \phi^{\text{active}}_{I\to B}(t) & \phi^{\text{active}}_{I\to A}(t)\\ \phi^{\text{active}}_{B\to E}(t) & \phi^{\text{active}}_{B\to I}(t) & \phi^{\text{active}}_{B}(t) & \phi^{\text{active}}_{B\to A}(t)\\ \phi^{\text{active}}_{A\to E}(t) & \phi^{\text{active}}_{A\to I}(t) & \phi^{\text{active}}_{A\to B}(t) & \phi^{\text{active}}_{A}(t) \end{bmatrix} \end{aligned}

The "actual information flow map" of the system — the amount of information genuinely flowing at this moment, obtained by multiplying corresponding elements of the previous two matrices.

#5.3 Consciousness Intensity $\bigstar(t)$ : The Richness of Differentiation and the Unity of Integration

We need a single scalar metric to quantify the overall level of consciousness. According to the core insight of this theory — consciousness requires simultaneously rich content (differentiation) and unified structure (integration) — consciousness intensity is defined as the product of the two:

$\boxed{\bigstar(t) = D(t) \times I(t)}$

#5.3.1 Normalized Differentiation $D(t)$

Defined as the weighted average of the first-order activity ratios of the four networks:

$D(t) = \sum_{X \in \{E, I, B, A\}} W_X(t) \cdot \frac{H(\mathbf{s}_X(t))}{H_{\max}(X)}$

where $W_X(t)$ are the dynamic power weights of each network, and $\sum W_X(t) = 1$ . This formula strictly outputs values in $[0, 1]$ .

#5.3.2 Normalized Integration $I(t)$

Defined as the weighted sum of second-order, third-order, and fourth-order integration components:

$\boxed{I(t) = \omega_2 I_2(t) + \omega_3 I_3(t) + \omega_4 I_4(t)}$

Constraints: $\omega_2 + \omega_3 + \omega_4 = 1$ , and $\omega_4 \geq \omega_3 \geq \omega_2 > 0$ , reflecting that higher-order integration contributes more fundamentally to the unity of consciousness.

Second-order interactive integration $I_2(t)$ : The average normalized transfer entropy across all 12 directed pairs. Corresponds to basic cross-modal information binding.

Third-order synergistic integration $I_3(t)$ : The weighted combination of normalized synergistic information across all triads. Corresponds to the generative capacity for complex experiences such as "context" or "atmosphere."

Fourth-order global integration $I_4(t)$ : Directly equals $R_{EIBA}(t)$ . Corresponds to pure "sense of presence" — the highest level of unity in consciousness.

#5.3.3 Explanatory Power

This model uniformly explains the continuous spectrum of consciousness states:

Wakeful consciousness: D is high, $I_2, I_3, I_4$ are all high, hence $\bigstar(t)$ is extremely high.
Deep dreamless sleep: D is extremely low, $\bigstar(t)$ approaches zero.
Epileptic seizure: D may not be low (whole-brain synchronized discharge), but the integration components collapse due to lack of differentiated interaction, $\bigstar(t)$ is extremely low.
Deep meditation: D and $I_2$ are actively suppressed, but $I_4$ is significantly elevated. $\bigstar(t)$ remains at a moderate level, but the experiential texture is special ("empty clarity," "oneness").
Schizophrenia: D may be normal, but $I_3$ (contextual integration) and $I_4$ (self-unity) are severely impaired, resulting in fragmented experience.

#5.4 Consciousness Measurement Units: White-Cabbage-Rice

#——An Information-Benchmark-Based Standardized Measurement System

To transform abstract information quantities into practical tools that can be intuitively understood and compared across states, we establish a standardized measurement system centered on the imagery of "White-Cabbage-Rice." This system is not a direct measurement of absolute physical quantities (such as field strength), but rather a ratio obtained by standardizing the key dynamical indicators of consciousness defined in Section 5.3 (differentiation IDI, integration III, consciousness intensity $\bigstar$ ) relative to a baseline state.

#5.4.1 Setting the Baseline

Any ratio system requires a zero point and a reference point. We set the following two frames of reference:

Theoretical zero point: When the system's information entropy and integration both approach zero (e.g., brain death or burst suppression under deep anesthesia), the consciousness measurement value is defined as 0.
Reference point (1 unit): Defined as the average value across a large sample of healthy adults in a wakeful, resting, task-free state. This is a statistical constant calibratable through large-sample experimental data.

#5.4.2 Definition of the Three Basic Units

"White" — Standardized Integration Measures the unity of conscious experience. It quantifies the strength of the current whole-brain causal interaction and phase synchronization relative to the baseline state. $\boxed{\text{W}(t) = \frac{I(t)}{\langle I \rangle_{\text{benchmark}}}}$
- 1 White: The current information integration equals the baseline state.
"Cabbage" — Standardized Differentiation Measures the richness of conscious content. It quantifies whether the current whole-brain electromagnetic field's information entropy is higher or lower than the baseline state. $\boxed{\text{C}(t) = \frac{D(t)}{\langle D \rangle_{\text{benchmark}}}}$
- 1 Cabbage: The current information richness equals the baseline state.
"Rice" — Standardized Consciousness Intensity Measures the overall intensity of consciousness. This is the most core composite indicator of this theory, defined as the product of the above two. $\boxed{\text{R}(t) = \frac{\bigstar(t)}{\langle \bigstar \rangle_{\text{benchmark}}} = \frac{D(t) \times I(t)}{\langle D \times I \rangle_{\text{benchmark}}}}$
- 1 Rice: Represents the standard level of consciousness for a healthy adult in a wakeful resting state.
- Numerically, since the baseline values are independently measured statistical averages, 1 White × 1 Cabbage ≈ 1 Rice serves as an excellent approximate estimate.

#5.4.3 From Mathematics to Measurement

It must be emphasized that White-Cabbage-Rice are currently theoretical constructs. Their practical application depends on the implementation roadmap in Section 5.7: we must first extract $H(\mathbf{s}_X)$ , $\text{NTE}_{X \to Y}$ , and $R_{EIBA}$ from EEG/MEG data, assemble IDI and III, and then standardize them relative to large-sample norms from healthy adults. Once this step is completed, we can attach a precise consciousness measurement label to each instant of subjective experience.

#5.5 The Regulatory Equations of the A Network: The Director of Information Flow

The A network plays the role of "director" in the consciousness system — it does not directly contribute stable content, but dynamically regulates the real-time information flow directed to the three major networks E, I, and B. This regulatory function is mathematically implemented through the interaction of the following three vectors.

#5.5.1 Demand Vector $\mathbf{O}(t)$

Quantifies the competitive intensity of external perception, internal thoughts, and bodily sensations for attentional resources, defined as the normalized transfer entropy flowing from non-A networks to A:

\mathbf{O}(t) = \begin{pmatrix} \text{NTE}_{E \to A}(t)\\ \text{NTE}_{I \to A}(t)\\ \text{NTE}_{B \to A}(t) \end{pmatrix}^T

It represents the unregulated, bottom-up informational "voices."

#5.5.2 Suppression Vector $\mathbf{S}(t)$ : The Information Dynamics of Resource Competition

The suppression vector represents the top-down regulatory output of the A network. It is not a static mask, but a dynamic filter applied to different sensory channels based on the current whole-brain information occupancy state.

To this end, we define a general resource competition instantiation formula for each perceptual channel $X \in \{E, I, B\}$ . Let $\mathcal{Y}_X$ be the set of the two networks excluding X and A. Then the suppression strength of A on input from X is defined as:

S_{X \to A}(t) = \underbrace{\left( 1 - \text{NTE}_{X \to A}(t) \right)}_{\text{Demand protection factor}} \cdot \underbrace{\left( \frac{1}{2} \sum_{Y \in \mathcal{Y}_X} \text{NTE}_{Y \to X}(t) \right)}_{\text{Noise defense factor}} \cdot \underbrace{S_{\text{base}_X}(t)}_{\text{Prior target factor}}

This formula perfectly balances three forces in the consciousness field:

Demand protection: When the A network is extremely dependent on a particular channel (i.e., $\text{NTE}_{X \to A} \to 1$ ), the first term approaches 0, forcibly suppressing the total suppression and ensuring that the information flow required for the core task remains unobstructed.
Noise defense: When X is heavily occupied by other networks (e.g., perception is severely distorted by internal thoughts and bodily emotions), the average NTE inside the parentheses increases, driving up suppression and preventing consciousness from being hijacked by unreliable "noise."
Prior guidance: $S_{\text{base}_X}(t)$ preserves the intervention space for high-level cognitive goals (e.g., "I am meditating and must ignore external distractions").

Thus, the complete suppression vector can be concisely expressed as:

\mathbf{S}(t) = \begin{pmatrix} S_{E \to A}(t)\\ S_{I \to A}(t)\\ S_{B \to A}(t) \end{pmatrix}^T

Why adopt this model?

#5.5.3 Consciousness Position Vector $\mathbf{C}(t)$

The final regulatory output of the A network — the "coordinates" of consciousness at this moment in the three-dimensional E-I-B state space, defined as the difference between the demand vector and the suppression vector:

\mathbf{C}(t) = \mathbf{O}(t) - \mathbf{S}(t) = \begin{pmatrix} \text{NTE}_{A \to E}(t)\\ \text{NTE}_{A \to I}(t)\\ \text{NTE}_{A \to B}(t) \end{pmatrix}^T

Geometric interpretation: The tip of $\mathbf{C}(t)$ is the instantaneous position of consciousness within the E-I-B cube. It precisely quantifies "whether my attention is currently occupied by the external world, internal thoughts, or bodily sensations."

#5.6 Dynamical Evolution Equations (Sections 5.6.1–5.6.3 Formulas Are Imprecise — Read with Caution)

The preceding variables describe "the state at this moment," but consciousness is a continuously flowing process. Physical intuition drives us to inquire: how do these macroscopic variables evolve over time? Below, the system's time-evolution differential equations are given.

Core principle: The evolution of the consciousness system is aimed at achieving a dynamic trade-off between the two competing goals of maximizing information entropy (differentiation) and maximizing information integration (integration).

#5.6.1 Evolution of Single-Network Information Entropy

\frac{dH_X(t)}{dt} = -\lambda_X H_X(t) + \left( \sum_{Y \neq X} TE_{Y \to X}(t) \right) \cdot \left( 1 - \frac{H_X(t)}{H_{\max}(X)} \right)

Where: $-\lambda_X H_X(t)$ is the natural decay term — in the absence of sustained input, the spontaneous activity complexity of neural populations naturally decays; $\sum TE_{Y \to X}(t)$ is the total information influx from other networks; $(1 - \frac{H_X}{H_{\max}})$ is the capacity saturation constraint, ensuring the system remains in a bounded state.

#5.6.2 Evolution of Inter-Network Transfer Entropy

\frac{dTE_{X \to Y}(t)}{dt} = -\lambda_{XY} TE_{X \to Y}(t) + \eta_{XY} \cdot H_X(t) \cdot H_Y(t) \cdot \left( 1 - \text{NTE}_{X \to Y}(t) \right)

Where: $-\lambda_{XY} TE$ is the interaction decay term; $\eta_{XY} \cdot H_X \cdot H_Y$ is the autocatalytic generation term — when two networks are simultaneously in a high-complexity state, the information channel between them is nonlinearly enhanced; $(1 - \text{NTE})$ is the channel capacity constraint, confining transfer entropy within the upper limit determined by anatomical structure.

#5.6.3 Evolution of Global Phase Synchronization

\frac{dR_{EIBA}(t)}{dt} = -\lambda_{EIBA} R_{EIBA}(t) + \gamma \cdot \bigstar(t) \cdot \left( 1 - R_{EIBA}(t) \right)

Where: $\gamma \cdot \bigstar(t)$ is the consciousness intensity positive feedback term — the higher the overall consciousness intensity of the system, the stronger the driving force for maintaining global synchronization, forming a positive feedback loop at the macroscopic level.

#5.6.4 Structural Issues with These Equations

The above equations constitute a formally closed system: given initial conditions, the evolutionary trajectories of $H_X(t)$ , $TE_{X\to Y}(t)$ , and $R_{EIBA}(t)$ can be solved. However, they suffer from several structural issues that need to be explicitly flagged here.

First issue: Closure assumption of macroscopic variables.

Information entropy H, transfer entropy $TE$ , phase synchronization order parameter R, consciousness intensity $\bigstar$ — these quantities are readouts, not drivers. They are statistical summaries of the firing patterns of millions of underlying neurons. In physics, the Navier-Stokes equations close because the causal chain between pressure and velocity is closed. But the brain's macroscopic information variables are constantly perturbed by underlying details (thalamic rhythms, neuromodulators, sensory stimuli). The relationships between macroscopic variables are statistical, not purely dynamical.

Second issue: Inversion of causal direction.

Equation 5.6.3 uses $\bigstar(t)$ to drive $R_{EIBA}$ . But the definition of $\bigstar(t)$ already includes $R_{EIBA}$ . Using a quantity that contains R to predict the change of R is a self-loop, not a causal relationship. Consciousness intensity is a readout indicator; it cannot act back on the system like a physical force.

Third issue: Asymmetry of information flow.

The equations treat all transfer entropy as positive contributions. But the A network's regulation of B and I is typically inhibitory. Conflating inhibitory and excitatory information flows robs the equations of functional discriminative power.

Fourth issue: Confusion of time scales.

The equations fail to distinguish between millisecond-level neural activity fluctuations and day-level structural remodeling (such as myelination). This renders the system's time constants $\lambda$ and $\eta$ physically ambiguous.

Conclusion: This set of equations is a formalization attempt at consciousness dynamics, but they are likely untenable.

#5.6.5 Incomplete Prediction Does Not Mean Powerlessness

Acknowledging the unattainability of closed dynamics does not mean abandoning scientific research. Consciousness science can make effective but incomplete predictions:

Diagnose the current state: Measuring blood pressure does not require knowing its differential equation; calculating $\bigstar(t)$ suffices to diagnose the level of consciousness.
Track state transitions: Using the "trajectory tracing method" of Chapter 7 to identify precursors of phase transitions.
Identify pathological fingerprints: Establishing diagnoses from excessive locking of $NTE$ or collapse of $R_{EIBA}$ .
Simulate microscopic generation: Using neural population models (such as Wilson-Cowan equations) for microscopic simulation, letting macroscopic patterns naturally emerge.

#5.6.6 Positioning of This Section

The equations in 5.6.1–5.6.3 mark the shape of an attempt. They are probes toward the unknown; the core strength of the theory lies in its diagnostic framework, not its predictive framework.

#5.7 Multi-Scale Information Dynamics of the A Network

Section 5.5 defined the instantaneous logic of A network regulation — how the current consciousness position $\mathbf{C}(t)$ arises from the competition between demand and suppression. However, consciousness is not merely a static snapshot; it is a river flowing through time.

To capture the urgency and slowness and turning points of this river, we need to perform calculus analysis on $\mathbf{O}(t)$ , $\mathbf{S}(t)$ , and $\mathbf{C}(t)$ . This analytical framework grants us the ability to track the "acceleration" and "inertia" of the stream of consciousness.

#5.7.1 First Derivative: Tracking the Instantaneous Velocity of Consciousness

Taking the first derivative of these vectors yields their instantaneous rate of change, reflecting the system's response speed and dynamic trends.

Rate of change of the demand vector:

The components of the demand vector are directly measured NTE values, so their differential is the instantaneous slope of NTE over time:

\frac{d\mathbf{O}(t)}{dt} = \begin{pmatrix}\frac{d}{dt}\text{NTE}_{E \to A}(t) \\frac{d}{dt}\text{NTE}_{I \to A}(t) \\frac{d}{dt}\text{NTE}_{B \to A}(t)\end{pmatrix}^{T}

Rate of change of the suppression vector:

Each component $S_{X \to A}(t)$ ( $X \in \{E, I, B\}$ ) of the suppression vector is the product of three factors. Its differential follows the product rule:

\frac{d}{dt} S_{X \to A}(t) = \frac{d}{dt} \left[ \left( 1 - \text{NTE}_{X \to A} \right) \cdot \left( \frac{1}{2} \sum_{Y \in \mathcal{Y}_X} \text{NTE}_{Y \to X} \right) \cdot S_{\text{base}_X} \right]

Expanding, $\frac{d\mathbf{S}}{dt}$ captures the competition of two forces:

Loosening of demand protection: When $\text{NTE}_{X \to A}$ begins to surge, the $(1 - \text{NTE})$ term drops sharply, causing suppression to collapse — the attentional gate is forced open.
Tightening of noise defense: When $\text{NTE}_{Y \to X}$ increases, the noise factor rises, pushing up suppression — the system begins to filter out contaminated inputs.

Rate of change of the consciousness position:

\boxed{\frac{d\mathbf{C}(t)}{dt} = \frac{d\mathbf{O}(t)}{dt} - \frac{d\mathbf{S}(t)}{dt}}

This directly corresponds to reportable subjective experience:

Large magnitude of $\frac{d\mathbf{C}}{dt}$ : Attention is switching rapidly (such as jumps during multitasking).
Stable direction and moderate magnitude of $\frac{d\mathbf{C}}{dt}$ : Represents smooth focus (such as a flow state).

#5.7.2 Second Derivative: Capturing Inflection Points of Consciousness

The second derivative reveals changes in the rate of change (acceleration), helping us identify critical dynamical transitions.

\frac{d^2\mathbf{O}(t)}{dt^2}, \quad \frac{d^2\mathbf{S}(t)}{dt^2}, \quad \frac{d^2\mathbf{C}(t)}{dt^2}

Theoretical predictions (inflection point detection):

Attention saturation warning: When $\frac{d^2}{dt^2}\text{NTE}_{E \to A}$ transitions from positive to negative, it means that although external demand is still increasing, the acceleration has slowed. This presages that the current focus of attention is approaching saturation and may soon shift.
Cognitive control mutation: If $\frac{d^2 S_{E \to A}}{dt^2}$ shows a sharp spike, it means the A network's suppression strategy has undergone a "hard switch" in an extremely short time. This corresponds to the instantaneous redistribution of brain resources when a sudden event (such as a loud noise) interrupts deep thought.

#5.7.3 Integration: Measuring the Cumulative Load of Consciousness

Integrating these vectors over a time interval $[t_1, t_2]$ yields cumulative quantities, reflecting long-term resource allocation and cognitive costs:

\int_{t_1}^{t_2} \mathbf{O}(t) dt, \quad \int_{t_1}^{t_2} \mathbf{S}(t) dt, \quad \int_{t_1}^{t_2} \mathbf{C}(t) dt

$\int \mathbf{O} dt$ : Measures the total demand pressure exerted on the A network by the external world, internal thoughts, and bodily signals over a period of time.
$\int \mathbf{S} dt$ : Measures the total suppressive effort (cognitive control cost) expended by the A network during this period. A high integral represents prolonged self-restraint and can predict cognitive fatigue.
$\int \mathbf{C} dt$ : Measures the actual cumulative flow of attentional resources. This is the mathematical characterization of "what I've mainly been paying attention to over the past hour."

#5.7.4 Double Integration: The Inertia and Habits of Consciousness

Double integration introduces a time-squared weight, emphasizing the impact of persistence:

\iint_{t_1}^{t_2} \mathbf{C}(t) dt^2

High-inertia pattern: If $\iint \mathbf{C}$ consistently accumulates high scores along a certain dimension (such as the I axis), it represents not just momentary mind-wandering, but a stable introspective habit or professional expertise.
Clinical perspective: In depression, the double integral would show $\mathbf{C}(t)$ chronically trapped in the negative region of the B-I plane, forming extremely deep cumulative inertia. Correcting it requires applying a similarly long-term sustained reverse-direction intervention.

#5.7.5 Summary: A Universal Multi-Scale Analytical Paradigm

What has been established here for the A network is not merely a set of formulas, but a universal language for dynamical analysis. Not only $\mathbf{C}(t)$ , but all core state variables in this theory — consciousness intensity $\bigstar(t)$ , differentiation $D(t)$ , integration $I(t)$ , and even any NTE channel — are continuous functions of time. They can all be decomposed through "first derivative (capturing velocity), second derivative (identifying inflection points), integration (measuring accumulation)."

This multi-scale analytical framework provides a unified mathematical scalpel for quantifying consciousness phenomena of interwoven short and long cycles — such as "sudden inspiration," "precursors of depressive episodes," and "habit formation" — in subsequent chapters.

#5.8 Implementation Roadmap: From Mathematics to Measurement

To ensure that $\bigstar(t)$ of the Rice Consciousness theory is not merely a philosophical symbol, we must build a computational bridge from the "continuous electromagnetic field" to the "digital consciousness intensity." This section provides operational definitions and engineering implementation formulas for each core variable, clearly distinguishing which are direct measurements and which require calibration against large-sample norms.

#5.8.1 Data Preprocessing: Taking a Snapshot of the Consciousness Field

First, we need to obtain the pure electromagnetic field activity of network $X \in \{E, I, B, A\}$ at time t. This includes the following standard steps:

Signal acquisition: Record brainwaves using high-density EEG (≥64 channels) or whole-head MEG at a sampling rate of at least 500 Hz.
Source reconstruction: Using eLORETA or beamforming, back-project the scalp signals to the target brain regions, obtaining the local field potential (LFP) time series for each brain region.
Time windowing: At time t, weextract a sliding window of length $W = 1$ second, obtaining the array $s_X = [x_1, x_2, ..., x_{W \times fs}]$ .

All subsequent variables will be computed based on these one-dimensional time series arrays.

#5.8.2 First-Order Variables: Measuring Network Differentiation

Operational definition: $\phi_X^{\text{active}}(t)$ no longer pursues theoretical differential entropy, but uses robust spectral complexity to capture "how many different oscillatory patterns are simultaneously active in this brain region at this moment." We provide two complementary practical algorithms.

Algorithm A: Normalized Spectral Entropy (Recommended)

This most directly reflects the "dimensional richness" of the electromagnetic field.

Compute the power spectral density (PSD) of the signal (using Welch's method), obtaining the power $P_i$ of $N_f$ frequency components.
Normalize the power into a probability distribution $p_i = \frac{P_i}{ \sum P_i}$ .
Compute the Shannon entropy and divide by the theoretical maximum $\log_2 N_f$ , naturally normalizing it to $[0,1]$ :\ $\boxed{H_{\text{spec}}[X](t) = -\frac{\sum_{i=1}^{N_f} p_i \log_2 p_i}{\log_2 N_f}}$ \ Physical intuition: When this brain region produces only a single frequency (such as the delta waves of deep sleep), $H_{\text{spec}} \approx 0$ ; when it simultaneously produces diverse frequencies such as θ, α, β, γ, $H_{\text{spec}} \to 1$ .

Algorithm B: Sample Entropy (for Cross-Validation)

Used to capture the nonlinear unpredictability of the signal as a supplement. Embedding dimension $m=2$ , tolerance $r=0.2 \times \text{SD}$ . Can directly call the open-source package antropy.sample_entropy.

Calibration of Maximum Entropy $H_{\max}(X)$

Since the biological brain cannot reach the physical limit of uniformly distributed frequencies across all bands, we adopt the large-sample norm calibration method:

$H_{\max}(X) = P_{99}\big( H_{\text{spec}}[X] \big)$

That is, from thousands of healthy adults under extreme limit states such as high cognitive load, TMS perturbation, and REM dream states, calculate the 99th percentile of spectral entropy for that brain region as its physiological capacity upper limit. This ensures that $\phi_X^{\text{activeratio}}(t) = H_{\text{spec}}[X](t) / H_{\max}(X)$ strictly falls in $[0,1]$ , and can distinguish "normal richness" from "abnormal overload."

#5.8.3 Second-Order Variables: Measuring Causal Information Flow Between Networks

Operational definition: $\phi_{X \to Y}^{\text{active}}(t) \equiv TE_{X \to Y}(t)$ , i.e., the transfer entropy from source X to target Y. It quantifies "how much the past state of X can reduce the uncertainty of Y's current state, where this information cannot be explained by Y's own history."

$TE_{X \to Y} = \sum p(Y_t, Y_{past}, X_{past}) \log_2 \frac{p(Y_t | Y_{past}, X_{past})}{p(Y_t | Y_{past})}$

Implementation method: In a Python environment, use the IDTxl package with the KSG nearest-neighbor estimator to handle the high-dimensional probability density problem of continuous signals. Since transfer entropy can produce spurious positive values due to noise, permutation testing must be performed: shuffle the time axis of X multiple times and compare whether the real $TE$ is significantly higher than the null distribution.

Structural Constraint of Maximum Transfer Entropy $TE_{\max}$

$TE_{\max}$ is determined by the bandwidth of the anatomical structure. We employ a dual calibration:

Diffusion Tensor Imaging (DTI) constraint: Measure the fractional anisotropy (FA) of that pathway, establishing a mapping function $TE_{\max} \propto g(\text{FA}, \text{Volume})$ .
Behavioral extreme calibration: Also extract the 99th percentile from large samples under extreme load, using the FA value as a regularization condition, ensuring the upper limit does not exceed the physical channel capacity.\ $\text{NTE}_{X \to Y}(t) = \frac{TE_{X \to Y}(t)}{TE_{\max}(X \to Y)} \in [0, 1]$

#5.8.4 Third-Order Variables: Approximating Emergent Synergistic Information

Challenge: True information decomposition is computationally intensive and sensitive to model selection.

Operational definition: We adopt O-information as a robust proxy indicator for triadic synergistic integration. O-information is the difference between total mutual information and all pairwise mutual informations, capturing the net amount by which "the whole is greater than the sum of its parts":

$\Omega(X, Y, Z) = I(X; Y; Z) - \big[ I(X;Y) + I(Y;Z) + I(Z;X) \big]$

When $\Omega > 0$ , the system is redundancy-dominated; when $\Omega < 0$ , it represents synergistic emergence (Synergy). We take $\phi_{X,Y,Z}^{\text{active}}(t) = \max(0, -\Omega)$ to measure the current synergistic strength.

Maximum synergistic information $Syn_{\max}$ : In practice, we set it as the historical maximum of $-\Omega$ observed under extreme (such as deep flow states, intense psychedelic states). This is the most pragmatic approximation of the physiological limit.

#5.8.5 Fourth-Order Variables: Measuring Global Unity

Operational definition: This is the only variable with an absolute physical limit. It is computed directly via the Kuramoto order parameter.

Narrowband filtering: From the clean time series of the four networks E, I, B, A, extract the $\gamma$ band of interest (30–80 Hz), obtaining $x_\gamma(t)$ .
Instantaneous phase extraction: Apply the Hilbert transform: $\theta_X(t) = \arg\left( x_\gamma(t) + i \mathcal{H}\{x_\gamma(t)\} \right)$
Geometric synchronization evaluation: Treat the four phases as unit vectors on the complex plane, compute the average length: $\boxed{\phi_{EIBA}^{\text{active}}(t) = R_{EIBA}(t) = \left\| \frac{1}{4} \sum_{X \in \{E, I, B, A\}} e^{i\theta_X(t)} \right\|}$ Since the theoretical upper bound $R_{\max}=1$ , this value itself is the activity ratio. $R \to 1$ corresponds to a deep sense of oneness, $R \to 0$ corresponds to conscious dissociation.

#5.8.6 Assembling the Final Dashboard

Once all foundational variables have been computed, the final consciousness intensity engine need only execute the following composition

Differentiation $D(t) = \sum_{X} W_X \cdot \left( H_{\text{spec}}[X](t) / H_{\max}(X) \right)$
Integration $I(t) = \omega_2 \cdot \overline{\text{NTE}}(t) + \omega_3 \cdot \overline{\|\Omega\|}(t) + \omega_4 \cdot R_{EIBA}(t)$
Consciousness Intensity $\bigstar(t) = D(t) \times I(t)$
Unit conversion: Dividing $D, I, \bigstar$ respectively by the norm averages of the wakeful resting state yields the specific readings for "Cabbage (C)," "White (W)," and "Rice (R)."

#5.8.7 Recommended Software Tool Stack

Step	Function	Recommended Tool
Preprocessing	EEG/MEG source reconstruction	MNE-Python, Brainstorm
First-order (Differentiation)	Spectral entropy, sample entropy	scipy.signal, antropy
Second-order (Integration)	KSG transfer entropy	IDTxl, PyIF
Third-order (Synergy)	O-information	dit package
Fourth-order (Unity)	Hilbert transform, order parameter	scipy.signal.hilbert

#5.8.8 From Blueprint to Evidence: The Honest Boundaries of Theory

It must be frankly pointed out that the precise computational workflow described in Sections 5.7.2 through 5.7.6 constitute a measurement blueprint that the Rice Consciousness theory draws for consciousness science. They are the "how it should be measured" derived from theoretical axioms, representing the ultimate expectations of this theory for empirical technology. However, from blueprint to daily laboratory work, we still face several real technical gaps:

Establishing norms for $H_{\max}$ and $TE_{\max}$ : These physiological limits require ultra-large-scale high-density EEG/MEG and DTI joint databases. Currently, such multimodal, large-sample open data are still under construction and not yet widely available.
Robustness of transfer entropy: The KSG nearest-neighbor estimator is extremely sensitive to data length and noise levels. In real magnetoencephalographic recordings, achieving satisfactory reliability and validity for $\text{NTE}_{X \to Y}$ often requires longer time windows than 1 second, creating a trade-off tension with the goal of "instantaneous tracking."
Computational nightmare of third-order synergistic information: Even using the O-information proxy, information decomposition of high-dimensional neural signals is still computationally expensive and sensitive to model selection. Although our proposed "limit experience calibration method" is pragmatic, the scientific community has not yet reached a consensus on what constitutes a true "physiological limit."
Uncertainty in source reconstruction: Back-projecting intracranial source signals (eLORETA) from scalp EEG is an inverse problem whose solution always carries spatial ambiguity.

These methods are measurement pathways that can potentially be realized. They prove that the core of the Rice Consciousness theory is not elusive metaphysical speculation, but points toward a series of specific neuro-computational problems that can be solved through future technological iteration. As high-density dry electrodes,next-generation magnetoencephalography (OPM-MEG), and more precise non-invasive source reconstruction algorithms mature, the roadmap described in this section will gradually move from "possible" to "feasible."

This roadmap proves that on the dashboard of the Rice Consciousness theory, every needle has a clear physical signal and code behind it — even if the magnifying glass we currently use to read these needles still needs to be polished a bit clearer.

#5.9 Restating the "Definition of Consciousness" in the Language of Entropy

In Chapter 3 (Section 3.4.1), we gave a definition of paramount importance in the history of consciousness studies:

Consciousness is not a "substance" or "attribute" of any kind. Consciousness is a continuous, self-referential, aperiodic dynamical process emergent from the coupling of four-dimensional information flow — External Projection (E), Internal Recurrence (I), Body Undertone (B), and Active Regulation (A) — within biological neural systems that meet specific complexity conditions. It is directly realized in the global electromagnetic field cooperatively excited by neural population electrical activity, and its subjective manifestation is the unique dynamical texture of each instant.

Now, having completed the meticulous forging of "entropy," "transfer entropy," and "synergistic information" in Chapter 4, we can restate this same definition using this physical language no longer reliant on metaphor — this time, every word has a corresponding measurement.

Step 1: Redefining "Four-Dimensional Information Flow"

"External Projection (E)": The local perceptual field excited by the neural population activity of the sensory cortex. The connotation of its "information flow" is physically defined as the differential entropy $H(\mathbf{s}_E(t))$ of this local field — it quantifies how many distinguishable oscillatory patterns are simultaneously active in the perceptual field at this moment. High entropy means a refined, diverse perceptual terrain; low entropy means monotonous or absent sensory input.
"Internal Recurrence (I)": The self-referential field centered on the default mode network. The connotation of its "information flow" is the internal differential entropy $H(\mathbf{s}_I(t))$ exhibited by this network when recursively processing the global field state, and the self-referential transfer entropy — the directed information quantity with which the I network takes its own past state as input to predict its own current state.
"Body Undertone (B)": The interoceptive field tone injected by the activity of the insula, anterior cingulate cortex, and subcortical structures. The connotation of its "information flow" is the differential entropy $H(\mathbf{s}_B(t))$ caused by this set of bodily signals in the global field, and the transfer entropy $\text{TE}(B \to I)$ , $\text{TE}(B \to A)$ it continuously outputs to the I field and A field.
"Active Regulation (A)": The modulatory gradient field established by the prefrontal-parietal executive control network. Its "information flow" is not stable content entropy, but manifests as directed transfer entropy outputs to the three information channels E, I, and B: $\text{TE}(A \to E)$ , $\text{TE}(A \to I)$ , $\text{TE}(A \to B)$ .

Step 2: Redefining "Coupling Emergence"

"The coupling and emergence of four-dimensional information flow" — this proposition was translated in Chapter 4 as:

In the continuous medium of the global electromagnetic field, the local entropy fields of E, I, and B are continuously generated and interact with one another. Their interactions are not vague "mutual influences," but twelve concrete directed transfer entropies (such as $\text{TE}(E \to I)$ , $\text{TE}(B \to I)$ , $\text{TE}(A \to B)$ , etc.).

The "present content" of consciousness is the set of instantaneous equilibrium solutions reached in the global field by these twelve causal information flows and the modulatory output of the A network. And those textures within such equilibrium solutions where "the whole is greater than the sum of its parts" — such as "situational atmosphere" and "intuition" — are physically defined as synergistic information: the residual information that cannot be explained by any pairwise interaction when three networks are simultaneously coupled.

Step 3: Redefining "Dynamical Texture" and "Subjective Manifestation"

Chapter 3 stated that "its subjective manifestation is the unique dynamical texture of each instant."

In the language of Chapter 4, this "texture" is no longer a literary metaphor. It is precisely decomposed into two measurable dimensions:

Differentiation $D(t)$ : The weighted sum of the local differential entropies of the four networks E, I, B, A at this instant. It quantifies the "richness" of conscious content at this moment — how many simultaneously existing feelings, thoughts, and perceptions you can distinguish. This is the "entropy" of the field.
Integration $I(t)$ : The weighted sum of the above twelve transfer entropies plus higher-order synergistic information and global phase synchronization ( $R_{EIBA}$ ) at this instant. It quantifies the "unity" of conscious content at this moment — the extent to which these different feelings, thoughts, and perceptions are bound into a seamless whole. This is the "causal structure" of the field.

And ultimately, consciousness intensity $\bigstar(t) = D(t) \times I(t)$ . This product is the complete physical characterization of that "dynamical process" from Chapter 3. It tells us: consciousness is not a static attribute, but an ongoing, entropy-and-information-flow-quantifiable "fire of relations" — how many distinguishable patterns are simultaneously burning in the global field at this very second, and how much directed causal information flows among them.

#5.10 Chapter Conclusion

We have transformed the E-I-B-A model of consciousness from a philosophical conception into a rigorous quantitative system composed of information entropy, transfer entropy, synergistic information, and phase synchronization order parameters.

Structural capacity $\mathbf{\Phi}$ defines the informationcarrying upper limit of the global electromagnetic field.
Activity ratio $\mathbf{\Phi}^{\text{activeratio}}$ and absolute activity $\mathbf{\Phi}^{\text{active}}$ characterize the information dynamics currently unfolding.
Consciousness intensity $\bigstar = D \times I$ provides a computable measure of overall consciousness level.
The A network equations reveal the information-flow nature of attention and free will.
The evolution equations describe how the system self-organizes in the tension between differentiation and integration.

This mathematical framework enables us to simulate, predict, and ultimately directly estimate the state and intensity of consciousness from neural data. In the next chapter, we will anchor this system of variables to real brain structures — finding the neural correlates for each $\phi$ .

[End of Chapter 5: The obsidian floor transitions from semi-matte to semi-gloss, the mist halves in density, and the undulations of the terrain begin to emerge.]