Chapter 22

Open Problems and Future Mathematical Directions of Social Quantum Field Theory

Corrections to Existing Equations

Correction to Section 22.2

The observation equation should be written as

\mathcal{H}_{\theta}\bigl(\Phi(t)\bigr)
+
\varepsilon(t).
]

Here:

  • (\mathcal{H}_{\theta}) is a parameter-dependent observation operator;
  • (\theta) is a vector of unknown parameters;
  • (\varepsilon(t)) represents measurement noise.

Structural identifiability requires

(\theta_2,\Phi_2).
\tag{22.1}
]

When gauge redundancy is present, equality should be interpreted modulo the relevant gauge group.

Correction to Section 22.3

The invariance of expectation values should be shown explicitly:

\langle U\Psi|
U\hat{O}U^{\dagger}
|U\Psi\rangle.
]

Equivalently,

\langle\Psi|
U^{\dagger}
\bigl(U\hat{O}U^{\dagger}\bigr)
U
|\Psi\rangle.
]

The quotient model space is

\mathcal{M}_{\mathrm{model}}/\mathcal{G}.
\tag{22.3}
]

Correction to Section 22.8

A time-dependent Hilbert-space formulation should read

\mathcal{H}(t).
\tag{22.9}
]

The state belongs to the instantaneous fiber:

[
|\Psi(t)\rangle
\in
\mathcal{H}(t).
]

A covariant time-evolution equation may be written as

H(t)|\Psi(t)\rangle.
\tag{22.10}
]

The operator (\nabla_t) defines the transport of states between different Hilbert spaces.

Correction to Section 22.13

The random metric should be written as

g_{\mu\nu}(\omega).
\tag{22.11}
]

If the action depends on this metric, then

S[\Phi;g(\omega)].
]

Observable predictions may require averaging over the geometric uncertainty:

[
\mathbb{E}{\omega}
\left[
\langle\hat{O}\rangle{g(\omega)}
\right].
]

Correction to Section 22.16

The nonequilibrium renormalization-group equation is

\beta_i(g,t).
\tag{22.13}
]

The explicit time dependence indicates that the scale flow and temporal evolution need not be separable.

Correction to Section 22.17

The coarse-graining and time-evolution operators generally satisfy

[
\mathcal{R}\mathcal{U}_t
\neq
\mathcal{U}_t\mathcal{R}.
\tag{22.14}
]

Their commutator is

\mathcal{R}\mathcal{U}_t

\mathcal{U}_t\mathcal{R}.
\tag{22.15}
]

The relevant analytical quantity is the operator norm

[
\bigl|
[\mathcal{R},\mathcal{U}_t]
\bigr|,
]

not the scalar absolute value.

Correction to Section 22.18

The coupled field-network equations are

F_i(\Phi,A),
\tag{22.16}
]

and

G_{ij}(A,\Phi).
\tag{22.17}
]

The first equation describes evolution on the network. The second describes evolution of the network itself.

Correction to Section 22.22

The sparse estimator should be written as

\operatorname*{arg,min}_{\theta}
\left[
|Y-\widehat{Y}(\theta)|^2
+
\lambda|\theta|_1
\right].
\tag{22.19}
]

The (\ell^1) term promotes sparse interaction structures.

Correction to Section 22.24

The robustness estimate should use norms on both sides:

[
|\rho_0(t)-\rho_{\theta}(t)|
\leq
C(t)
|\mathcal{L}0-\mathcal{L}{\theta}|.
\tag{22.21}
]

The choice of norm must be specified. Possible choices include the trace norm, Hilbert–Schmidt norm, or an induced operator norm.

Restored and Supplemented Sections

22.29 Noncommutativity and Classical Sufficiency

A central methodological question is whether the noncommutative structure of SQFT provides explanatory or predictive value beyond classical probability theory.

If all empirically relevant observables commute,

[
[O_i,O_j]=0
]

for every pair (i,j), then the observable algebra is commutative.

Under suitable spectral assumptions, a commutative observable algebra can be represented as an algebra of ordinary random variables on a classical probability space.

This means that a formally quantum-inspired model may nevertheless be empirically equivalent to a classical stochastic model.

Definition 22.4 — Classical Sufficiency

A classical probabilistic representation is sufficient for an SQFT model when there exists a probability space

[
(\Omega,\mathcal{F},P)
]

and a family of classical random variables

[
{X_O\in\mathfrak{A}}
]

such that all empirically accessible joint distributions and expectation values are reproduced.

In particular,

\mathbb{E}_{P}[X_O].
]

Open Problem 22.27

Develop empirical and mathematical criteria capable of distinguishing genuinely useful noncommutative models from classical latent-variable models.

A noncommutative formulation should be preferred only when it improves at least one of the following:

  • predictive accuracy;
  • parameter economy;
  • structural compression;
  • cross-context generalization;
  • causal interpretation;
  • representation of order effects.

The mere use of operators does not establish the necessity of a quantum-inspired framework.

22.30 Contextuality

Contextuality refers to the dependence of an outcome distribution on the measurement context in which an observable is evaluated.

Suppose observable (A) may be measured either alone or after observable (B). Context dependence is present when

[
P(A=a)
\neq
P(A=a\mid B\text{ measured first}).
\tag{22.22}
]

Such dependence may be represented by noncommuting operators:

[
[A,B]\neq 0.
]

However, order effects do not automatically imply nonclassical contextuality. They may also result from:

  • memory;
  • learning;
  • fatigue;
  • strategic adaptation;
  • information acquisition;
  • ordinary measurement disturbance.

Open Problem 22.28

Determine whether empirical order effects require a non-Kolmogorov probability representation or can be reproduced by a classical model with an expanded state space.

A valid contextuality test must rule out classical explanations before invoking noncommutative probability.

22.31 Entanglement-Like Correlations

For a composite system (A\cup B), the joint state is represented by

[
\rho_{AB}.
]

A product state satisfies

\rho_A\otimes\rho_B.
]

A correlated state satisfies

[
\rho_{AB}
\neq
\rho_A\otimes\rho_B.
\tag{22.23}
]

Nonfactorization alone does not establish quantum entanglement. Classical systems may also possess strongly correlated joint distributions.

A separable operator state has the form

\sum_i
p_i
\rho_A^{(i)}
\otimes
\rho_B^{(i)},
]

where

[
p_i\geq 0,
\qquad
\sum_i p_i=1.
]

A state that cannot be represented in this form is nonseparable relative to the chosen tensor-product decomposition.

Open Problem 22.29

Develop rigorous criteria that distinguish among:

  • classical statistical dependence;
  • separable operator correlation;
  • contextual dependence;
  • nonseparable operator states;
  • genuinely nonclassical correlation structures.

In social applications, the term “entanglement” should be used only when the subsystem decomposition and nonseparability criterion are explicitly defined.

22.32 Bell-Type Inequalities

Let (A_1,A_2) be bounded observables associated with subsystem (A), and let (B_1,B_2) be bounded observables associated with subsystem (B).

Define the CHSH-type quantity

\langle A_2B_2\rangle.
\tag{22.24}
]

Under the assumptions of a local classical hidden-variable model,

[
|\mathcal{B}|
\leq
2.
\tag{22.25}
]

An operator model may permit

[
|\mathcal{B}|
\leq
2\sqrt{2}.
]

However, social systems generally violate the isolation and no-communication assumptions used in physical Bell experiments.

Open Problem 22.30

Determine whether meaningful Bell-type protocols can be constructed for social systems while controlling for:

  • communication between participants;
  • shared historical causes;
  • sample-selection bias;
  • post-selection;
  • memory;
  • measurement dependence;
  • strategic coordination.

An apparent violation of a Bell-type inequality does not demonstrate nonclassicality unless these alternative mechanisms are excluded.

22.33 Privacy-Preserving Field Reconstruction

Empirical SQFT models may require high-dimensional behavioral, organizational, economic, or political data.

These data may contain sensitive information about individuals and institutions.

A randomized mechanism (M) satisfies ((\varepsilon,\delta))-differential privacy when

[
P\bigl(M(D)\in S\bigr)
\leq
e^{\varepsilon}
P\bigl(M(D')\in S\bigr)
+
\delta
\tag{22.26}
]

for every measurable output set (S) and every pair of neighboring datasets (D) and (D').

Open Problem 22.31

Develop privacy-preserving estimators for:

  • latent social fields;
  • interaction kernels;
  • Hamiltonian parameters;
  • Lindblad operators;
  • correlation functions;
  • network topology.

The central challenge is to protect individual records without destroying the collective structure required for reliable field reconstruction.

22.34 Interpretability of Learned Operators

A high-dimensional generator may achieve strong predictive performance while remaining difficult to interpret.

A learned operator may depend on an arbitrary choice of basis. Individual matrix entries therefore need not possess invariant meaning.

Interpretability should instead be based on quantities preserved under admissible transformations.

Possible invariant summaries include:

  • spectral gaps;
  • dominant eigenmodes;
  • invariant subspaces;
  • conserved quantities;
  • effective potentials;
  • interaction graphs;
  • entropy-production channels;
  • scale-dependent couplings.

Open Problem 22.32

Construct coordinate-independent and empirically meaningful summaries of estimated SQFT operators.

An interpretability map may be represented as

[
\mathfrak{I}:
\mathcal{L}
\longrightarrow
\mathcal{Q},
]

where (\mathcal{Q}) is a space of human-interpretable structural quantities.

The map should satisfy

\mathfrak{I}
\bigl(
U\mathcal{L}U^{-1}
\bigr)
]

for gauge-equivalent representations.

22.35 Reproducibility and Benchmarking

A complete empirical SQFT study should specify

(D,\mathcal{H},\Theta,\mathscr{L},\mathcal{A},\mathcal{V}),
\tag{22.27}
]

where:

  • (D) is the dataset;
  • (\mathcal{H}) is the observation model;
  • (\Theta) is the parameter space;
  • (\mathscr{L}) is the likelihood or loss function;
  • (\mathcal{A}) is the estimation algorithm;
  • (\mathcal{V}) is the validation protocol.

The symbol (\mathscr{L}) is used for the statistical loss to distinguish it from the dynamical generator (\mathcal{L}).

Open Problem 22.33

Develop standardized datasets and benchmark tasks for comparing SQFT with:

  • classical dynamical systems;
  • stochastic differential equations;
  • network models;
  • agent-based models;
  • Markov processes;
  • Bayesian state-space models;
  • classical latent-variable models.

Benchmarking should evaluate predictive accuracy, calibration, interpretability, computational cost, and robustness.

22.36 Effective Classicality

The operator framework allows off-diagonal density-matrix elements and noncommuting observables.

An effective classical regime emerges when off-diagonal components decay in a dynamically selected basis:

[
\rho_{ij}(t)
\longrightarrow
0,
\qquad
i\neq j.
]

The remaining diagonal components may then evolve as an ordinary probability vector.

Conjecture 22.1 — Effective Classicality

Let an open SQFT system evolve under a primitive Lindblad semigroup with a unique stationary state.

If environmental coupling suppresses off-diagonal components in a stable pointer basis, then sufficiently coarse observables evolve approximately according to a classical Markov process.

Symbolically,

[
\mathcal{L}{\mathrm{SQFT}}
\longrightarrow
\mathcal{L}{\mathrm{Markov}}
\tag{22.28}
]

under decoherence, long-time evolution, and coarse-graining.

A rigorous result would require an explicit error bound such as

e^{t\mathcal{L}_{\mathrm{Markov}}}
\Pi(\rho)
\right|_1
\leq
\varepsilon(t),
]

where (\Pi) projects the operator state onto the effective classical subalgebra.

22.37 Multiscale Universality

Consider two microscopic theories with different local actions:

[
S_1^{(0)}
\neq
S_2^{(0)}.
]

Let (\mathcal{R}) denote a renormalization or coarse-graining transformation:

\mathcal{R}
\bigl(
S_a^{(n)}
\bigr),
\qquad
a\in{1,2}.
]

Conjecture 22.2 — Institutional Universality

SQFT models with the same symmetry class, effective dimensionality, conservation structure, and relevant operators converge toward the same infrared fixed point:

[
S_1^{(n)}
\longrightarrow
S^{*},
]

and

[
S_2^{(n)}
\longrightarrow
S^{*}
\qquad
\text{as }
n\rightarrow\infty.
\tag{22.29}
]

If valid, this conjecture would justify the classification of social systems into universality classes whose large-scale behavior is insensitive to microscopic detail.

22.38 Generic Identifiability

Let

[
G^{(n)}(t_1,\ldots,t_n)
]

denote an (n)-time correlation function.

A sufficiently rich collection of such functions may determine the dynamical generator.

Conjecture 22.3 — Generic Identifiability

For a finite-dimensional SQFT model with analytic dynamics, sufficiently informative observables, and persistent excitation,

[
{G^{(n)}}_{n=1}^{N}
\Longrightarrow
[\mathcal{L}],
\tag{22.30}
]

where ([\mathcal{L}]) denotes the gauge-equivalence class of the generator.

The conjecture asserts identifiability only modulo transformations that leave every admissible observation unchanged.

22.39 Low-Complexity Effective States

Although the full Hilbert space may be extremely large, empirically stable social systems may occupy a much smaller effective subset.

Conjecture 22.4 — Low-Complexity Social States

For a broad class of empirically stable systems,

[
\rho
\approx
\rho_r,
]

where

r
\ll
\dim\mathcal{H}.
\tag{22.31}
]

This conjecture would justify:

  • low-rank estimation;
  • tensor-network representations;
  • reduced-order models;
  • latent-subspace reconstruction;
  • compressed sensing of operators.

A stronger formulation would seek an approximation bound:

[
|\rho-\rho_r|_1
\leq
\varepsilon(r),
]

with

[
\varepsilon(r)\longrightarrow 0
]

as (r) increases.

22.40 Research Priorities

The open problems of SQFT may be divided into four levels.

Level I — Foundational Problems

These include:

  • mathematical ontology;
  • gauge equivalence;
  • causal structure;
  • algebraic formulation;
  • category-theoretic formulation;
  • correspondence with classical probability;
  • empirical necessity of noncommutativity.

Level II — Analytical Problems

These include:

  • non-Markovian positivity;
  • fractional evolution;
  • nonlinear density-operator dynamics;
  • time-dependent Hilbert spaces;
  • structural stability;
  • topological transitions;
  • existence and uniqueness on evolving networks.

Level III — Computational Problems

These include:

  • tensor-network approximation;
  • sparse operator discovery;
  • computational complexity;
  • scalable inverse problems;
  • low-rank reconstruction;
  • multiscale numerical simulation.

Level IV — Empirical Problems

These include:

  • structural identifiability;
  • causal validation;
  • model comparison;
  • out-of-sample prediction;
  • reproducibility;
  • privacy;
  • interpretability;
  • falsifiability.

The foundational and empirical levels should receive priority. Greater mathematical complexity is not scientifically valuable unless its objects can be identified and tested.

22.41 Minimal Scientific Standard and Methodological Warning

Before an SQFT model is regarded as scientifically informative, it should satisfy the following minimum conditions.

First, the latent state must be explicitly defined.

Second, observables must be distinguished from inferred quantities.

Third, the observation operator must be specified.

Fourth, parameters must be identifiable, at least up to a known equivalence class.

Fifth, the model must generate out-of-sample predictions.

Sixth, it must be compared with simpler classical alternatives.

Seventh, the conditions under which it would be rejected must be stated in advance.

Eighth, every imported term from quantum theory must correspond to an explicitly defined mathematical structure.

These requirements may be summarized as

[
\text{Definition}
+
\text{Identification}
+
\text{Prediction}
+
\text{Comparison}
+
\text{Falsification}.
\tag{22.32}
]

The greatest danger to SQFT is uncontrolled metaphor.

Terms such as “quantum,” “entanglement,” “collapse,” “vacuum,” “gauge field,” “topological defect,” and “renormalization” possess precise meanings in mathematics and physics.

Their use should be classified into three distinct categories:

  1. literal physical claim;
  2. mathematical structural analogy;
  3. informal metaphor.

The present framework primarily belongs to the second category.

SQFT does not assert that societies literally obey microscopic quantum mechanics. It investigates whether structures from field theory, operator theory, open-system dynamics, information theory, and renormalization can provide coherent mathematical models of complex social systems.

22.42 Final Perspective

Social Quantum Field Theory remains an unfinished mathematical research program.

Its present contribution is to organize a broad range of mathematical tools within a unified framework for describing:

  • multiscale interaction;
  • uncertainty;
  • contextual dependence;
  • institutional emergence;
  • collective synchronization;
  • structural collapse;
  • historical memory;
  • nonlinear adaptation.

Its future value will depend on whether this framework can produce:

  • rigorous theorems;
  • identifiable empirical models;
  • computationally tractable algorithms;
  • reproducible predictions;
  • explicit failure conditions.

Some parts of the framework may prove indispensable. Others may reduce to simpler classical models. Still others may need to be replaced by mathematical structures not yet introduced.

Such revision would not constitute theoretical failure. It would represent the normal development of a formal scientific research program.

The long-term objective is not to preserve the vocabulary of quantum field theory at all costs. The objective is to discover the smallest mathematically coherent framework capable of representing the observed complexity of social systems.

The governing methodological principle is therefore:

[
\text{Use the richest formalism necessary,}
]

[
\text{but no richer than the evidence permits.}
\tag{22.33}
]

This principle requires mathematical ambition to remain subordinate to empirical discipline.


Appendix A

Mathematical Foundations of Social Quantum Field Theory

A.1 Purpose of This Appendix

The main chapters of this monograph employ tools from functional analysis, operator theory, probability, differential geometry, dynamical systems, and mathematical physics.

This appendix collects the principal definitions and assumptions required for those constructions.

Its purpose is not to reproduce a complete textbook in pure mathematics. Instead, it establishes a common technical foundation for the SQFT framework and clarifies the conditions under which the equations used throughout the monograph are mathematically meaningful.

The central mathematical objects are:

[
\mathcal{M},
\qquad
\mathcal{H},
\qquad
\mathfrak{A},
\qquad
\Phi,
\qquad
H,
\qquad
\rho,
\qquad
\mathcal{L}.
]

Here:

  • (\mathcal{M}) is the social manifold or configuration space;
  • (\mathcal{H}) is a Hilbert space;
  • (\mathfrak{A}) is an algebra of observables;
  • (\Phi) is a field;
  • (H) is a Hamiltonian or dynamical generator;
  • (\rho) is a density operator;
  • (\mathcal{L}) is an open-system generator.

A.2 Sets, Maps, and Relations

Let (X) and (Y) be sets.

A map

[
f\rightarrow Y
]

assigns to every (x\in X) a unique element (f(x)\in Y).

The image of a subset (A\subseteq X) is

{f(x)\in A}.
]

The preimage of a subset (B\subseteq Y) is

{x\in X(x)\in B}.
]

A map is injective when

[
f(x_1)=f(x_2)
\quad\Longrightarrow\quad
x_1=x_2.
]

It is surjective when

[
f(X)=Y.
]

It is bijective when it is both injective and surjective.

An equivalence relation (\sim) on (X) satisfies:

[
x\sim x,
]

[
x\sim y
\quad\Longrightarrow\quad
y\sim x,
]

and

[
x\sim y,\quad y\sim z
\quad\Longrightarrow\quad
x\sim z.
]

The equivalence class of (x) is

{y\in X\sim x}.
]

The quotient space is

[
X/{\sim}.
]

This construction is used in SQFT to identify gauge-equivalent models.

A.3 Metric Spaces

Definition A.1 — Metric

A metric on a set (X) is a function

[
d\times X\rightarrow[0,\infty)
]

such that, for all (x,y,z\in X),

[
d(x,y)\geq 0,
]

[
d(x,y)=0
\quad\Longleftrightarrow\quad
x=y,
]

[
d(x,y)=d(y,x),
]

and

[
d(x,z)
\leq
d(x,y)+d(y,z).
]

The pair ((X,d)) is called a metric space.

An open ball centered at (x) with radius (r>0) is

{y\in X(x,y)<r}.
]

A sequence ({x_n}) converges to (x) when

[
d(x_n,x)\rightarrow 0.
]

A sequence is Cauchy when, for every (\varepsilon>0), there exists (N) such that

[
d(x_n,x_m)<\varepsilon
]

whenever (n,m\geq N).

A metric space is complete when every Cauchy sequence converges to an element of the space.

Completeness is essential for fixed-point arguments and existence theorems.

A.4 Normed Vector Spaces

Let (V) be a vector space over (\mathbb{R}) or (\mathbb{C}).

Definition A.2 — Norm

A norm is a function

[
|\cdot|\rightarrow[0,\infty)
]

satisfying

[
|x|\geq 0,
]

[
|x|=0
\quad\Longleftrightarrow\quad
x=0,
]

|\alpha||x|,
]

and

[
|x+y|
\leq
|x|+|y|.
]

A norm induces a metric:

|x-y|.
]

A complete normed vector space is called a Banach space.

Examples include:

[
\mathbb{R}^n,
]

[
\mathbb{C}^n,
]

[
L^p(\Omega),
]

and

[
C([a,b]).
]

A.5 Inner-Product Spaces

Definition A.3 — Inner Product

An inner product on a complex vector space (V) is a map

[
\langle\cdot,\cdot\rangle:
V\times V
\rightarrow
\mathbb{C}
]

such that

\overline{\langle y,x\rangle},
]

\alpha\langle x,y\rangle
+
\beta\langle x,z\rangle,
]

and

[
\langle x,x\rangle
\geq 0,
]

with equality only when (x=0).

The induced norm is

\sqrt{\langle x,x\rangle}.
]

The Cauchy–Schwarz inequality states that

[
|\langle x,y\rangle|
\leq
|x||y|.
\tag{A.1}
]

Two vectors are orthogonal when

[
\langle x,y\rangle=0.
]

A.6 Hilbert Spaces

Definition A.4 — Hilbert Space

A Hilbert space is a complete inner-product space.

The SQFT state space is assumed to be a complex separable Hilbert space:

[
\mathcal{H}.
]

Separability means that (\mathcal{H}) contains a countable dense subset.

A countable orthonormal basis is denoted by

[
{|e_n\rangle}_{n=1}^{\infty}.
]

Every state (|\Psi\rangle\in\mathcal{H}) may then be expanded as

\sum_{n=1}^{\infty}
c_n|e_n\rangle,
]

where

[
\sum_{n=1}^{\infty}|c_n|^2
<
\infty.
]

The norm is

\langle\Psi|\Psi\rangle.
]

A normalized state satisfies


]

A.7 Orthogonal Projections

Let (M\subseteq\mathcal{H}) be a closed subspace.

The orthogonal projection onto (M) is an operator

[
P_M:\mathcal{H}\rightarrow M
]

satisfying

[
P_M^2=P_M,
]

and

[
P_M^{\dagger}=P_M.
]

Every vector (|\Psi\rangle\in\mathcal{H}) can be written uniquely as

P_M|\Psi\rangle
+
(I-P_M)|\Psi\rangle.
]

The two components are orthogonal:


]

Projection operators are used to represent measurement outcomes, constrained sectors, and effective subspaces.

A.8 Bounded Linear Operators

A linear operator

[
A:\mathcal{H}\rightarrow\mathcal{H}
]

is bounded when there exists (C\geq 0) such that

[
|A\Psi|
\leq
C|\Psi|
]

for all (|\Psi\rangle\in\mathcal{H}).

The operator norm is

\sup_{|\Psi|=1}
|A\Psi|.
]

The set of bounded linear operators on (\mathcal{H}) is denoted by

[
\mathcal{B}(\mathcal{H}).
]

With operator composition and the operator norm, (\mathcal{B}(\mathcal{H})) is a Banach algebra.

A.9 Adjoint Operators

For a bounded operator (A), the adjoint (A^{\dagger}) is defined by

\langle A^{\dagger}\Phi,\Psi\rangle
]

for all (\Phi,\Psi\in\mathcal{H}).

An operator is self-adjoint when

[
A=A^{\dagger}.
]

An operator is unitary when

UU^{\dagger}

I.
]

An operator is normal when

A^{\dagger}A.
]

Observables in SQFT are generally represented by self-adjoint operators.

A.10 Unbounded Operators and Domains

Important dynamical generators are often unbounded.

An unbounded operator is written as

[
A(A)\subseteq\mathcal{H}
\rightarrow
\mathcal{H},
]

where (D(A)) is the domain of (A).

The domain must be explicitly specified because (A\Psi) may not exist for every (\Psi\in\mathcal{H}).

An operator is densely defined when (D(A)) is dense in (\mathcal{H}).

It is closed when

[
\Psi_n\rightarrow\Psi
]

and

[
A\Psi_n\rightarrow\Phi
]

imply

[
\Psi\in D(A)
]

and

[
A\Psi=\Phi.
]

It is closable when it has a closed extension.

The Hamiltonian (H) is assumed to be densely defined and self-adjoint on its domain (D(H)).

A.11 Spectral Theory

Theorem A.1 — Spectral Theorem

Let (A) be a self-adjoint operator on (\mathcal{H}). Then there exists a projection-valued measure (E_A) such that

\int_{\sigma(A)}
\lambda,dE_A(\lambda),
\tag{A.2}
]

where (\sigma(A)) is the spectrum of (A).

For a suitable function (f),

\int_{\sigma(A)}
f(\lambda),dE_A(\lambda).
]

In particular,

\int_{\sigma(A)}
e^{-it\lambda},dE_A(\lambda).
]

This result provides the mathematical foundation for operator exponentials and unitary time evolution.

A.12 Commutators and Anticommutators

For operators (A) and (B), the commutator is

AB-BA.
]

The anticommutator is

AB+BA.
]

If

[
[A,B]=0,
]

the operators commute.

Commuting self-adjoint operators may often be simultaneously diagonalized, subject to appropriate spectral conditions.

The Jacobi identity is


\tag{A.3}
]

Commutators generate the algebraic structure underlying dynamical equations and symmetry transformations.

A.13 Trace-Class Operators

An operator (A) is trace class when

[
\operatorname{Tr}|A|
<
\infty,
]

where

\sqrt{A^{\dagger}A}.
]

The trace is defined by

\sum_n
\langle e_n|A|e_n\rangle,
]

and is independent of the orthonormal basis.

For suitable operators,

\operatorname{Tr}(BA).
\tag{A.4}
]

The trace norm is

\operatorname{Tr}|A|.
]

Trace-class operators form a Banach space denoted by

[
\mathcal{T}_1(\mathcal{H}).
]

A.14 Hilbert–Schmidt Operators

An operator (A) is Hilbert–Schmidt when

\operatorname{Tr}(A^{\dagger}A)
<
\infty.
]

The Hilbert–Schmidt inner product is

\operatorname{Tr}(A^{\dagger}B).
]

The induced norm is

\sqrt{\operatorname{Tr}(A^{\dagger}A)}.
]

This norm is frequently used in parameter estimation and numerical model comparison.

A.15 Density Operators

Definition A.5 — Density Operator

A density operator is a trace-class operator (\rho) satisfying

[
\rho\geq 0,
]

[
\rho=\rho^{\dagger},
]

and


]

A pure state has the form

|\Psi\rangle\langle\Psi|.
]

For a pure state,

[
\rho^2=\rho.
]

A mixed state may be written as

\sum_i
p_i
|\Psi_i\rangle
\langle\Psi_i|,
]

where

[
p_i\geq 0,
]

and

[
\sum_i p_i=1.
]

The expectation value of an observable (O) is

\operatorname{Tr}(\rho O).
\tag{A.5}
]

A.16 Entropy and Relative Entropy

The von Neumann entropy is


\operatorname{Tr}(\rho\ln\rho).
\tag{A.6}
]

For a pure state,

[
S(\rho)=0.
]

For a mixed state,

[
S(\rho)\geq 0.
]

The quantum relative entropy of (\rho) with respect to (\sigma) is

\operatorname{Tr}
\left[
\rho(\ln\rho-\ln\sigma)
\right],
\tag{A.7}
]

provided the support of (\rho) is contained in the support of (\sigma).

Relative entropy measures distinguishability between states.

In SQFT it may be interpreted as a divergence between competing latent-state representations.

A.17 Tensor Products

Let (\mathcal{H}_A) and (\mathcal{H}_B) be Hilbert spaces.

The composite space is

\mathcal{H}_A\otimes\mathcal{H}_B.
]

A product state has the form

|\Psi_A\rangle
\otimes
|\Psi_B\rangle.
]

A general state need not be factorized.

For density operators, a product state is

\rho_A\otimes\rho_B.
]

The reduced state of subsystem (A) is obtained through the partial trace:

\operatorname{Tr}B(\rho{AB}).
\tag{A.8}
]

Similarly,

\operatorname{Tr}A(\rho{AB}).
]

Tensor products are used to represent composite actors, institutions, sectors, or interacting subsystems.

A.18 Completely Positive Maps

A linear map

[
\mathcal{E}:
\mathcal{T}_1(\mathcal{H})
\rightarrow
\mathcal{T}_1(\mathcal{H})
]

is positive when

[
\rho\geq 0
\quad\Longrightarrow\quad
\mathcal{E}(\rho)\geq 0.
]

It is completely positive when

[
\mathcal{E}\otimes I_n
]

is positive for every positive integer (n).

It is trace preserving when

\operatorname{Tr}\rho.
]

A completely positive trace-preserving map is abbreviated CPTP.

Every finite-dimensional CPTP map admits a Kraus representation:

\sum_k
K_k\rho K_k^{\dagger},
\tag{A.9}
]

with

I.
\tag{A.10}
]

A.19 One-Parameter Semigroups

A family of bounded operators

[
{T(t)}_{t\geq 0}
]

is a strongly continuous semigroup when

[
T(0)=I,
]

T(t)T(s),
]

and

\Psi
]

for all (\Psi\in\mathcal{H}).

The generator (A) is defined by

\lim_{t\rightarrow 0^+}
\frac{T(t)\Psi-\Psi}{t},
\tag{A.11}
]

for all (\Psi) for which the limit exists.

Formally,

e^{tA}.
]

A.20 Stone’s Theorem

Theorem A.2 — Stone’s Theorem

A strongly continuous one-parameter unitary group

[
U(t)
]

has a unique self-adjoint generator (H) such that

e^{-itH}.
\tag{A.12}
]

Conversely, every self-adjoint operator generates a strongly continuous unitary group.

In SQFT units, the evolution operator is written as

e^{-itH/\kappa_S}.
\tag{A.13}
]

The state evolves according to

U(t)|\Psi(0)\rangle.
]

A.21 Schrödinger-Type Evolution

The abstract evolution equation is

H|\Psi(t)\rangle.
\tag{A.14}
]

When (H) is time independent,

e^{-itH/\kappa_S}
|\Psi(0)\rangle.
]

For a time-dependent Hamiltonian (H(t)), the formal solution is

\mathcal{T}
\exp
\left(
-\frac{i}{\kappa_S}
\int_0^t
H(s),ds
\right)
|\Psi(0)\rangle,
\tag{A.15}
]

where (\mathcal{T}) denotes time ordering.

A.22 Heisenberg Evolution

For a time-independent Hamiltonian, the Heisenberg operator is

U(t)^{\dagger}
A
U(t).
]

Its evolution satisfies

\frac{i}{\kappa_S}
[H,A_H(t)]
+
\left(
\frac{\partial A}{\partial t}
\right)_H.
\tag{A.16}
]

The Schrödinger and Heisenberg descriptions generate the same expectation values:

\langle\Psi(0)|A_H(t)|\Psi(0)\rangle.
]

A.23 Lindblad Generators

A Markovian open-system evolution is governed by

\mathcal{L}(\rho),
]

where

\frac{1}{2}
{L_k^{\dagger}L_k,\rho}
\right).
\tag{A.17}
]

The operators (L_k) are Lindblad or jump operators.

The solution is formally

e^{t\mathcal{L}}\rho(0).
\tag{A.18}
]

Under appropriate assumptions, the semigroup

[
e^{t\mathcal{L}}
]

is completely positive and trace preserving.

A.24 Trace Preservation of Lindblad Dynamics

Taking the trace of Equation (A.17) gives

\operatorname{Tr}\mathcal{L}(\rho).
]

For the Hamiltonian part,


]

For each dissipative term,

\operatorname{Tr}
\left(
L_k^{\dagger}L_k\rho
\right).
]

Therefore,


]

Hence,

\operatorname{Tr}\rho(0).
\tag{A.19}
]

If the initial state is normalized, then

[
\operatorname{Tr}\rho(t)=1.
]

A.25 Measure Spaces

A measure space is a triple

[
(\Omega,\Sigma,\mu),
]

where:

  • (\Omega) is a set;
  • (\Sigma) is a sigma-algebra;
  • (\mu) is a measure.

A sigma-algebra satisfies:

[
\Omega\in\Sigma,
]

[
A\in\Sigma
\quad\Longrightarrow\quad
\Omega\setminus A\in\Sigma,
]

and

[
A_1,A_2,\ldots\in\Sigma
\quad\Longrightarrow\quad
\bigcup_{n=1}^{\infty}A_n\in\Sigma.
]

A measure satisfies

[
\mu(A)\geq 0,
]

[
\mu(\varnothing)=0,
]

and countable additivity:

\sum_{n=1}^{\infty}
\mu(A_n)
]

for pairwise disjoint sets (A_n).

A.26 Probability Spaces

A probability space is a measure space satisfying

[
\mu(\Omega)=1.
]

It is commonly written as

[
(\Omega,\mathcal{F},P).
]

A random variable is a measurable function

[
X:\Omega\rightarrow\mathbb{R}.
]

The expectation value is

\int_{\Omega}
X(\omega),dP(\omega).
\tag{A.20}
]

The variance is

\mathbb{E}
\left[
(X-\mathbb{E}[X])^2
\right].
\tag{A.21}
]

The covariance of (X) and (Y) is

\mathbb{E}
\left[
(X-\mathbb{E}[X])
(Y-\mathbb{E}[Y])
\right].
\tag{A.22}
]

A.27 (L^p) Spaces

For (1\leq p<\infty), the space (L^p(\Omega)) consists of measurable functions satisfying

[
\int_{\Omega}
|f(x)|^p,d\mu(x)
<
\infty.
]

The norm is

\left(
\int_{\Omega}
|f(x)|^p,d\mu(x)
\right)^{1/p}.
\tag{A.23}
]

For (p=2),

[
L^2(\Omega)
]

is a Hilbert space with inner product

\int_{\Omega}
\overline{f(x)}g(x),d\mu(x).
\tag{A.24}
]

Many SQFT field spaces are modeled as (L^2) spaces or Sobolev spaces derived from them.

A.28 Sobolev Spaces

Let (\Omega\subseteq\mathbb{R}^n).

The Sobolev space (W^{k,p}(\Omega)) consists of functions whose weak derivatives up to order (k) belong to (L^p(\Omega)).

For (p=2), one writes

W^{k,2}(\Omega).
]

The norm is

\sum_{|\alpha|\leq k}
|D^{\alpha}f|_{L^2}^2.
\tag{A.25}
]

Sobolev spaces provide a natural framework for weak solutions of field equations.

A.29 Weak Derivatives

A function (g) is the weak derivative of (f) when


\int_{\Omega}
g(x)\varphi(x),dx
\tag{A.26}
]

for all smooth test functions (\varphi) with compact support.

Weak derivatives permit the treatment of fields that are not classically differentiable.

A.30 Distributions

A distribution is a continuous linear functional acting on a space of test functions.

If (\mathcal{D}(\Omega)) denotes the space of smooth compactly supported functions, then a distribution (T) maps

[
T:
\mathcal{D}(\Omega)
\rightarrow
\mathbb{C}.
]

The Dirac delta distribution satisfies

\varphi(x_0).
\tag{A.27}
]

Operator-valued fields are often distributional rather than ordinary pointwise-defined functions.

A.31 Manifolds

Definition A.6 — Smooth Manifold

An (n)-dimensional smooth manifold (\mathcal{M}) is a topological space locally equivalent to (\mathbb{R}^n), equipped with a compatible smooth atlas.

A chart is a pair

[
(U,\varphi),
]

where (U\subseteq\mathcal{M}) is open and

[
\varphi\rightarrow\mathbb{R}^n
]

is a homeomorphism onto its image.

Coordinates are written as

[
x^{\mu},
\qquad
\mu=1,\ldots,n.
]

The tangent space at (x\in\mathcal{M}) is denoted by

[
T_x\mathcal{M}.
]

The cotangent space is

[
T_x^{*}\mathcal{M}.
]

A.32 Vector Fields

A vector field assigns a tangent vector to every point:

[
X:
x
\longmapsto
X_x\in T_x\mathcal{M}.
]

In local coordinates,

X^{\mu}(x)
\frac{\partial}{\partial x^{\mu}}.
]

The Lie bracket of two vector fields is

XY-YX.
\tag{A.28}
]

The Lie bracket measures the noncommutativity of flows generated by (X) and (Y).

A.33 Differential Forms

A differential (p)-form is a smooth section of the exterior power

[
\Lambda^pT^{*}\mathcal{M}.
]

A one-form is written as

\omega_{\mu}dx^{\mu}.
]

The exterior derivative is

[
d:
\Omega^p(\mathcal{M})
\rightarrow
\Omega^{p+1}(\mathcal{M}),
]

with

[
d^2=0.
\tag{A.29}
]

A form is closed when

[
d\omega=0.
]

It is exact when

[
\omega=d\eta.
]

Every exact form is closed.

Topological information arises from closed forms that are not exact.

A.34 Riemannian Metrics

A Riemannian metric assigns an inner product to each tangent space.

In local coordinates,

g_{\mu\nu}(x)
dx^{\mu}dx^{\nu}.
\tag{A.30}
]

The metric tensor satisfies

g_{\nu\mu}.
]

For a Riemannian metric, the quadratic form is positive definite.

The volume element is

\sqrt{\det g},
d^nx.
\tag{A.31}
]

The Laplace–Beltrami operator is

\frac{1}{\sqrt{\det g}}
\partial_{\mu}
\left(
\sqrt{\det g}
g^{\mu\nu}
\partial_{\nu}f
\right).
\tag{A.32}
]

A.35 Connections and Covariant Derivatives

A connection defines differentiation of vector fields along vector fields.

The covariant derivative is written as

[
\nabla_XY.
]

In local coordinates,

\partial_{\mu}Y^{\nu}
+
\Gamma^{\nu}_{\mu\lambda}Y^{\lambda}.
\tag{A.33}
]

The coefficients (\Gamma^{\nu}_{\mu\lambda}) are connection coefficients.

For the Levi–Civita connection,

\partial_{\sigma}g_{\mu\nu}
\right).
\tag{A.34}
]

Connections are relevant when SQFT states or parameters vary over curved configuration spaces.

A.36 Curvature

The Riemann curvature tensor is defined by

\nabla_X\nabla_YZ

\nabla_Y\nabla_XZ

\nabla_{[X,Y]}Z.
\tag{A.35}
]

In coordinates,

\partial_{\mu}\Gamma^{\rho}_{\nu\sigma}

\Gamma^{\rho}{\nu\lambda}
\Gamma^{\lambda}{\mu\sigma}.
\tag{A.36}
]

The Ricci tensor is

R^{\lambda}_{\ \mu\lambda\nu}.
]

The scalar curvature is

g^{\mu\nu}R_{\mu\nu}.
]

In social modeling, curvature may encode path dependence, transition asymmetry, or geometric obstruction, but only after a specific metric interpretation is defined.

A.37 Lie Groups

A Lie group (G) is simultaneously a smooth manifold and a group, with smooth multiplication and inversion.

A one-parameter subgroup is

e^{tX},
]

where (X) belongs to the Lie algebra (\mathfrak{g}).

The Lie algebra bracket is written as

[
[X,Y].
]

A representation of (G) on (\mathcal{H}) is a homomorphism

[
U\rightarrow\mathcal{U}(\mathcal{H}),
]

where (\mathcal{U}(\mathcal{H})) denotes the unitary operators on (\mathcal{H}).

Symmetry transformations in SQFT are often modeled by Lie-group representations.

A.38 Noether-Type Structure

Consider an action

\int
\mathcal{L}
\bigl(
\Phi,
\partial_{\mu}\Phi
\bigr)
,d^nx.
]

Suppose the transformation

[
\Phi
\longrightarrow
\Phi+\varepsilon\delta\Phi
]

leaves the action invariant to first order.

Then a conserved current may exist:

[
\partial_{\mu}j^{\mu}=0.
\tag{A.37}
]

The conserved charge is

\int_{\Sigma}
j^0,d^{n-1}x.
\tag{A.38}
]

In SQFT, such conservation laws are mathematical consequences of model symmetry, not automatically empirical laws of society.

A.39 Variational Derivatives

Let

\int_{\Omega}
\mathcal{L}
\bigl(
\Phi,
\partial_{\mu}\Phi,
x
\bigr)
,d^nx.
]

Under the variation

[
\Phi
\longrightarrow
\Phi+\varepsilon\eta,
]

the first variation is

\left.
\frac{d}{d\varepsilon}
S[\Phi+\varepsilon\eta]
\right|_{\varepsilon=0}.
]

Stationarity requires

[
\delta S=0.
]

This leads to the Euler–Lagrange equation:


\tag{A.39}
]

A.40 Hamiltonian Formulation

Let (\Phi) be a field and define the conjugate momentum

\frac{\partial\mathcal{L}}
{\partial(\partial_t\Phi)}.
]

The Hamiltonian density is

\Pi\partial_t\Phi

\mathcal{L}.
\tag{A.40}
]

The Hamiltonian is

\int
\mathcal{H},d^{n-1}x.
\tag{A.41}
]

Hamilton’s equations are

\frac{\delta H}{\delta\Pi},
]

and


\frac{\delta H}{\delta\Phi}.
\tag{A.42}
]

A.41 Fixed-Point Theorems

Theorem A.3 — Banach Fixed-Point Theorem

Let ((X,d)) be a complete metric space and let

[
T\rightarrow X
]

be a contraction:

[
d(Tx,Ty)
\leq
q,d(x,y),
\qquad
0\leq q<1.
]

Then (T) possesses a unique fixed point (x^{*}) satisfying

[
T(x^{})=x^{}.
\tag{A.43}
]

Moreover, the iterative sequence

[
x_{n+1}=T(x_n)
]

converges to (x^{*}).

This theorem is frequently used to prove local existence and uniqueness.

A.42 Lipschitz Continuity

A map

[
F\rightarrow Y
]

between normed spaces is Lipschitz continuous when there exists (L\geq 0) such that

[
|F(x)-F(y)|
\leq
L|x-y|.
\tag{A.44}
]

It is locally Lipschitz when every point has a neighborhood on which such an estimate holds.

Lipschitz continuity is a standard sufficient condition for uniqueness of solutions to evolution equations.

A.43 Ordinary Differential Equations

Consider

F(t,x),
\qquad
x(t_0)=x_0.
\tag{A.45}
]

Theorem A.4 — Picard–Lindelöf

If (F) is continuous in (t) and locally Lipschitz in (x), then Equation (A.45) has a unique local solution.

Global existence generally requires additional growth bounds such as

[
|F(t,x)|
\leq
a(t)+b(t)|x|.
\tag{A.46}
]

A.44 Evolution Equations in Banach Spaces

Consider the abstract equation

Au+F(u),
\qquad
u(0)=u_0,
\tag{A.47}
]

where (A) generates a strongly continuous semigroup (T(t)).

A mild solution satisfies

T(t)u_0
+
\int_0^t
T(t-s)F(u(s)),ds.
\tag{A.48}
]

This formulation remains meaningful even when the classical derivative does not exist.

A.45 Grönwall’s Inequality

Suppose a nonnegative function (u(t)) satisfies

[
u(t)
\leq
a
+
\int_0^t
b(s)u(s),ds.
]

Then

[
u(t)
\leq
a
\exp
\left(
\int_0^t
b(s),ds
\right).
\tag{A.49}
]

Grönwall’s inequality is central to uniqueness, stability, and error estimates.

A.46 Stability

Let (x^{*}) be an equilibrium satisfying

[
F(x^{*})=0.
]

The equilibrium is stable if small perturbations remain small.

It is asymptotically stable if

[
x(t)\rightarrow x^{*}
]

for nearby initial conditions.

Linearization gives

DF(x^{*})\delta x.
\tag{A.50}
]

The eigenvalues of (DF(x^{*})) determine local stability.

If all eigenvalues have negative real parts, the equilibrium is locally asymptotically stable.

A.47 Lyapunov Functions

A Lyapunov function is a scalar function

[
V\rightarrow[0,\infty)
]

such that

[
V(x^{*})=0,
]

[
V(x)>0
]

for (x\neq x^{*}), and

[
\frac{dV}{dt}
\leq
0
]

along trajectories.

If

[
\frac{dV}{dt}<0
]

away from (x^{}), then (x^{}) is asymptotically stable under suitable conditions.

Entropy, free energy, and relative entropy may serve as Lyapunov-type quantities in open-system dynamics.

A.48 Compact Operators

A bounded operator (K\rightarrow Y) is compact when it maps bounded sets into relatively compact sets.

For compact self-adjoint operators on a Hilbert space, the spectrum consists of real eigenvalues with possible accumulation only at zero.

Compactness is frequently used in spectral decomposition and inverse-problem theory.

A.49 Weak and Strong Convergence

A sequence (x_n) converges strongly to (x) when

[
|x_n-x|
\rightarrow
0.
]

It converges weakly when

[
\langle y,x_n\rangle
\rightarrow
\langle y,x\rangle
]

for every (y\in\mathcal{H}).

Strong convergence implies weak convergence, but the converse is generally false.

Weak convergence is often sufficient for existence proofs in infinite-dimensional spaces.

A.50 Compactness Assumptions

Many existence arguments require one or more of the following:

  • boundedness;
  • weak compactness;
  • compact embedding;
  • lower semicontinuity;
  • coercivity.

A functional (J) is coercive when

[
J(x)\rightarrow\infty
]

as

[
|x|\rightarrow\infty.
]

A functional is weakly lower semicontinuous when

[
x_n\rightharpoonup x
]

implies

[
J(x)
\leq
\liminf_{n\rightarrow\infty}
J(x_n).
\tag{A.51}
]

These properties support the existence of minimizers.

A.51 Inverse Problems

An inverse problem seeks to reconstruct (x) from

\mathcal{H}(x)
+
\varepsilon.
]

The problem is well posed in the sense of Hadamard when:

  1. a solution exists;
  2. the solution is unique;
  3. the solution depends continuously on the data.

Many empirical reconstruction problems are ill posed.

Regularization introduces a penalty:

\operatorname*{arg,min}_{x}
\left[
|\mathcal{H}(x)-y|^2
+
\lambda R(x)
\right].
\tag{A.52}
]

The function (R(x)) encodes prior smoothness, sparsity, boundedness, or low complexity.

A.52 Tikhonov Regularization

A standard choice is

[
R(x)=|x|^2.
]

Then

\operatorname*{arg,min}_{x}
\left[
|\mathcal{H}(x)-y|^2
+
\lambda|x|^2
\right].
\tag{A.53}
]

For a linear operator (\mathcal{H}), the normal equation is

\mathcal{H}^{\dagger}y.
\tag{A.54}
]

A.53 Bayesian Inference

Let (\theta) be an unknown parameter and (D) observed data.

Bayes’ theorem gives

\frac{
P(D|\theta)P(\theta)
}{
P(D)
}.
\tag{A.55}
]

Here:

  • (P(\theta)) is the prior;
  • (P(D|\theta)) is the likelihood;
  • (P(\theta|D)) is the posterior.

The evidence is

\int
P(D|\theta)P(\theta),d\theta.
\tag{A.56}
]

The maximum a posteriori estimator is

\operatorname*{arg,max}_{\theta}
P(\theta|D).
\tag{A.57}
]

A.54 Fisher Information

For a parameterized probability density (p(y|\theta)), the Fisher information matrix is

\mathbb{E}_{\theta}
\left[
\frac{\partial\ln p(Y|\theta)}{\partial\theta_i}
\frac{\partial\ln p(Y|\theta)}{\partial\theta_j}
\right].
\tag{A.58}
]

Under regularity conditions, the Cramér–Rao inequality states that

[
\operatorname{Cov}(\hat{\theta})
\geq
I(\theta)^{-1}
\tag{A.59}
]

for unbiased estimators.

The Fisher matrix also defines a local information-geometric metric on parameter space.

A.55 Identifiability and Rank Conditions

For a parameterized map

[
F:\Theta\rightarrow\mathcal{Y},
]

local identifiability may be studied through the Jacobian

\frac{\partial F}{\partial\theta}.
]

If

\dim\Theta,
\tag{A.60}
]

then (F) is locally injective near (\theta) under suitable regularity assumptions.

Rank deficiency indicates local parameter redundancy.

A.56 Correlation Functions

For an operator-valued field (\Phi(x)), the two-point correlation function is

\langle
\Phi(x)\Phi(y)
\rangle.
\tag{A.61}
]

The (n)-point function is

\langle
\Phi(x_1)
\cdots
\Phi(x_n)
\rangle.
\tag{A.62}
]

Connected correlation functions isolate irreducible dependence.

For two variables,

G^{(2)}(x,y)

\langle\Phi(x)\rangle
\langle\Phi(y)\rangle.
\tag{A.63}
]

A.57 Generating Functionals

A generating functional may be written as

\int
\mathcal{D}\Phi,
\exp
\left(
\frac{i}{\kappa_S}
S[\Phi]
+
\frac{i}{\kappa_S}
\int
J(x)\Phi(x),dx
\right).
\tag{A.64}
]

Correlation functions are obtained formally by functional differentiation:

\left.
\frac{1}{i^n}
\frac{\delta^n Z[J]}
{\delta J(x_1)\cdots\delta J(x_n)}
\right|_{J=0},
\tag{A.65}
]

subject to normalization conventions.

In rigorous treatments, the functional integral may require regularization, discretization, or probabilistic reformulation.

A.58 Functional Derivatives

The functional derivative is defined through

F[\Phi]

\varepsilon
\int
\frac{\delta F}{\delta\Phi(x)}
\eta(x),dx
+
o(\varepsilon).
\tag{A.66}
]

For the functional

\int
V(\Phi(x)),dx,
]

one has

V'(\Phi(x)).
]

A.59 Fourier Transformation

For (f\in L^1(\mathbb{R}^n)), the Fourier transform is

\int_{\mathbb{R}^n}
f(x)e^{-ik\cdot x},dx.
\tag{A.67}
]

The inverse transform is

\frac{1}{(2\pi)^n}
\int_{\mathbb{R}^n}
\widehat{f}(k)e^{ik\cdot x},dk.
\tag{A.68}
]

Differentiation becomes multiplication:

ik_{\mu}\widehat{f}(k).
\tag{A.69}
]

Fourier methods are useful for spectral analysis, propagation, stability, and renormalization.

A.60 Green Functions

For a linear operator (L), a Green function (G(x,y)) satisfies

\delta(x-y).
\tag{A.70}
]

The solution of

J
]

may be written formally as

\int
G(x,y)J(y),dy.
\tag{A.71}
]

Green functions describe response propagation from sources to fields.

A.61 Boundary Conditions

A field equation requires boundary or initial conditions.

Common boundary conditions include:

Dirichlet condition

f.
\tag{A.72}
]

Neumann condition

g.
\tag{A.73}
]

Robin condition

c
\quad
\text{on }
\partial\Omega.
\tag{A.74}
]

In SQFT, boundary conditions may represent fixed institutional constraints, imposed external flows, or partial permeability between sectors.

A.62 Compact Summary of Standing Assumptions

Unless otherwise stated, the principal SQFT constructions assume the following.

Assumption A.1 — State Space

The state space (\mathcal{H}) is a complex separable Hilbert space.

Assumption A.2 — Observable Operators

Physical or empirical observables are represented by densely defined self-adjoint operators.

Assumption A.3 — Density Operators

States represented by density operators satisfy

[
\rho\geq 0,
\qquad
\rho=\rho^{\dagger},
\qquad
\operatorname{Tr}\rho=1.
]

Assumption A.4 — Hamiltonian

The Hamiltonian (H) is self-adjoint on a dense domain (D(H)).

Assumption A.5 — Closed-System Evolution

Closed-system evolution is generated by the unitary group

e^{-itH/\kappa_S}.
]

Assumption A.6 — Open-System Evolution

Markovian open-system evolution is represented by a completely positive trace-preserving semigroup.

Assumption A.7 — Regularity

Nonlinear maps are assumed locally Lipschitz whenever uniqueness is invoked.

Assumption A.8 — Measurability

All random variables, field observables, and parameter maps are measurable with respect to their stated sigma-algebras.

Assumption A.9 — Integrability

All expectation values and action functionals are assumed finite whenever they are used.

Assumption A.10 — Boundary Conditions

Every differential equation is accompanied by admissible initial and boundary conditions.

Assumption A.11 — Operator Domains

Expressions involving products or commutators of unbounded operators are used only on a common invariant dense domain.

Assumption A.12 — Empirical Interpretation

The mathematical structures of SQFT do not imply that social systems literally obey microscopic quantum mechanics.

A.63 Final Remarks

The formalism of SQFT draws from several areas of mathematics, but these structures should not be combined without attention to compatibility.

In particular:

  • not every symmetric operator is self-adjoint;
  • not every positive map is completely positive;
  • not every formal operator exponential defines a valid semigroup;
  • not every path integral exists as an ordinary measure;
  • not every nonlinear evolution preserves positivity;
  • not every observed correlation implies causal interaction;
  • not every quotient space is smooth;
  • not every inverse problem is identifiable.

Accordingly, every application should explicitly state:

[
\text{state space},
]

[
\text{operator domains},
]

[
\text{regularity assumptions},
]

[
\text{boundary conditions},
]

[
\text{observation model},
]

and

[
\text{equivalence relations}.
]

The mathematical discipline of SQFT depends less on the number of advanced concepts it employs than on the precision with which their assumptions are specified.


Appendix B

Core Proofs and Mathematical Derivations of Social Quantum Field Theory

B.1 Purpose and Scope

The main text frequently invokes operator evolution, density matrices, Lindblad generators, projection measurements, decoherence, and coarse-graining.

This appendix provides the principal derivations needed to support those constructions.

The objective is not to prove every theorem in complete functional-analytic generality. Instead, the derivations are presented under assumptions sufficient for the finite-dimensional or bounded-operator formulations most commonly used in SQFT.

Unless otherwise stated, assume that:

[
\mathcal{H}
]

is a finite-dimensional complex Hilbert space,

[
H=H^{\dagger}
]

is a self-adjoint Hamiltonian,

[
\rho\geq 0,
]

and

[
\operatorname{Tr}\rho=1.
]

The SQFT action scale is denoted by

[
\kappa_S.
]

It plays the formal role occupied by (\hbar) in ordinary quantum theory, but it is not assumed to be a physical Planck constant.

B.2 Conservation of Norm under Schrödinger-Type Evolution

Consider the state equation

H|\Psi(t)\rangle.
\tag{B.1}
]

Its adjoint equation is

\langle\Psi(t)|H.
\tag{B.2}
]

The time derivative of the norm is

\frac{d\langle\Psi(t)|}{dt}
|\Psi(t)\rangle
+
\langle\Psi(t)|
\frac{d|\Psi(t)\rangle}{dt}.
]

Using Equations (B.1) and (B.2),

\frac{i}{\kappa_S}
\langle\Psi|H|\Psi\rangle.
]

Therefore,


\tag{B.3}
]

Hence,

\langle\Psi(0)|\Psi(0)\rangle.
\tag{B.4}
]

A normalized state remains normalized.

B.3 Unitary Evolution Operator

For a time-independent Hamiltonian, define

\exp
\left(
-\frac{iHt}{\kappa_S}
\right).
\tag{B.5}
]

Because

[
H=H^{\dagger},
]

one has

\exp
\left(
\frac{iHt}{\kappa_S}
\right).
]

Thus,

\exp
\left(
\frac{iHt}{\kappa_S}
\right)
\exp
\left(
-\frac{iHt}{\kappa_S}
\right).
]

Since both exponentials are functions of the same operator (H), they commute. Therefore,

I.
\tag{B.6}
]

Similarly,

I.
]

Hence (U(t)) is unitary.

The solution of Equation (B.1) is

U(t)|\Psi(0)\rangle.
\tag{B.7}
]

B.4 Group Property of Closed-System Evolution

For time-independent (H),

\exp
\left(
-\frac{iH(t+s)}{\kappa_S}
\right).
]

Since (Ht) and (Hs) commute,

\exp
\left(
-\frac{iHt}{\kappa_S}
\right)
\exp
\left(
-\frac{iHs}{\kappa_S}
\right).
]

Therefore,

U(t)U(s).
\tag{B.8}
]

Also,

[
U(0)=I,
]

and

[
U(-t)=U(t)^{-1}.
]

Closed-system evolution therefore forms a one-parameter unitary group.

B.5 Density-Operator Evolution for a Closed System

For a pure state,

|\Psi(t)\rangle
\langle\Psi(t)|.
]

Using Equation (B.7),

U(t)\rho(0)U(t)^{\dagger}.
\tag{B.9}
]

Differentiating,

\frac{dU}{dt}
\rho(0)U^{\dagger}
+
U\rho(0)
\frac{dU^{\dagger}}{dt}.
]

Since

-\frac{i}{\kappa_S}HU,
]

and

\frac{i}{\kappa_S}U^{\dagger}H,
]

one obtains

-\frac{i}{\kappa_S}H\rho
+
\frac{i}{\kappa_S}\rho H.
]

Therefore,

-\frac{i}{\kappa_S}
[H,\rho].
\tag{B.10}
]

This is the von Neumann equation.

B.6 Conservation of Trace under Closed Evolution

Taking the trace of Equation (B.10),

-\frac{i}{\kappa_S}
\operatorname{Tr}[H,\rho].
]

Using cyclicity of the trace,

\operatorname{Tr}(\rho H).
]

Hence,


]

Therefore,


\tag{B.11}
]

B.7 Conservation of Purity under Unitary Evolution

The purity of a state is

\operatorname{Tr}(\rho^2).
\tag{B.12}
]

Differentiating,

2\operatorname{Tr}
\left(
\rho\frac{d\rho}{dt}
\right).
]

Using Equation (B.10),

-\frac{2i}{\kappa_S}
\operatorname{Tr}
\left(
\rho[H,\rho]
\right).
]

Expanding the commutator,

0
]

by cyclicity.

Thus,


\tag{B.13}
]

Unitary evolution preserves purity.

B.8 Heisenberg Equation of Motion

The Heisenberg operator is

U(t)^{\dagger}
A_S(t)
U(t).
\tag{B.14}
]

Differentiating,

\frac{dU^{\dagger}}{dt}
A_SU
+
U^{\dagger}
\frac{\partial A_S}{\partial t}
U
+
U^{\dagger}
A_S
\frac{dU}{dt}.
]

Substituting

-\frac{i}{\kappa_S}HU
]

and

\frac{i}{\kappa_S}U^{\dagger}H,
]

gives

\frac{i}{\kappa_S}
U^{\dagger}
(HA_S-A_SH)
U
+
U^{\dagger}
\frac{\partial A_S}{\partial t}
U.
]

Therefore,

\frac{i}{\kappa_S}
[H_H,A_H]
+
\left(
\frac{\partial A}{\partial t}
\right)_H.
\tag{B.15}
]

For time-independent (H),

[
H_H=H.
]

B.9 Equivalence of Schrödinger and Heisenberg Pictures

In the Schrödinger picture,

\langle\Psi(t)|
A_S
|\Psi(t)\rangle.
]

Using

U(t)|\Psi(0)\rangle,
]

one has

\langle\Psi(0)|
U(t)^{\dagger}
A_S
U(t)
|\Psi(0)\rangle.
]

Since

U(t)^{\dagger}A_SU(t),
]

it follows that

\langle\Psi(0)|
A_H(t)
|\Psi(0)\rangle.
\tag{B.16}
]

The two pictures therefore yield identical observable predictions.

B.10 General Lindblad Equation

The Markovian open-system evolution is

\frac{1}{2}
\rho L_k^{\dagger}L_k
\right).
\tag{B.17}
]

Equivalently,

\frac{1}{2}
{L_k^{\dagger}L_k,\rho}
\right).
\tag{B.18}
]

The first term represents reversible evolution.

The remaining terms represent irreversible environmental effects.

B.11 Trace Preservation of Lindblad Dynamics

Taking the trace of Equation (B.18),

\frac{1}{2}
\rho L_k^{\dagger}L_k
\right).
]

The Hamiltonian contribution vanishes:

[
\operatorname{Tr}[H,\rho]=0.
]

For the dissipative contribution,

\operatorname{Tr}
\left(
L_k^{\dagger}L_k\rho
\right).
]

Also,

\operatorname{Tr}
\left(
L_k^{\dagger}L_k\rho
\right).
]

Therefore, each summand becomes


]

Thus,


\tag{B.19}
]

A normalized density operator remains normalized.

B.12 Hermiticity Preservation

Assume

[
\rho=\rho^{\dagger}.
]

Taking the adjoint of the Hamiltonian term,

-\frac{i}{\kappa_S}[H,\rho].
]

For the dissipative term,

L_k\rho L_k^{\dagger},
]

and

{L_k^{\dagger}L_k,\rho}.
]

Therefore,

\mathcal{L}(\rho).
\tag{B.20}
]

Hence Hermiticity is preserved by Lindblad evolution.

B.13 Infinitesimal Kraus Representation

For a short time interval (\Delta t), define

I

\left(
\frac{iH}{\kappa_S}
+
\frac{1}{2}
\sum_k
L_k^{\dagger}L_k
\right)
\Delta t,
\tag{B.21}
]

and

\sqrt{\Delta t},L_k.
\tag{B.22}
]

The short-time state update is

K_0\rho(t)K_0^{\dagger}
+
\sum_k
K_k\rho(t)K_k^{\dagger}
+
O(\Delta t^2).
\tag{B.23}
]

Expanding the first term to first order,

\rho

\frac{i\Delta t}{\kappa_S}[H,\rho]

\frac{\Delta t}{2}
\sum_k
{L_k^{\dagger}L_k,\rho}
+
O(\Delta t^2).
]

The remaining terms give

\Delta t
\sum_k
L_k\rho L_k^{\dagger}.
]

Combining terms,

\Delta t,\mathcal{L}(\rho)
+
O(\Delta t^2).
]

Dividing by (\Delta t) and taking the limit gives the Lindblad equation.

B.14 Complete Positivity of the Short-Time Map

The map

[
\rho
\longmapsto
\sum_{\alpha}
K_{\alpha}\rho K_{\alpha}^{\dagger}
]

is completely positive because it has Kraus form.

Moreover,

I
+
O(\Delta t^2).
\tag{B.24}
]

Therefore, the map is trace preserving to first order.

Repeated composition of these infinitesimal completely positive maps yields the completely positive semigroup generated by the Lindblad operator, subject to the standard regularity assumptions.

B.15 Stationary States

A stationary state (\rho_*) satisfies


\tag{B.25}
]

Then

\rho_*
]

for all (t) whenever

[
\rho(0)=\rho_*.
]

In finite dimensions, the existence of at least one stationary state follows under broad conditions for a continuous CPTP semigroup acting on a compact convex state space.

Uniqueness requires stronger assumptions, such as primitivity or irreducibility.

B.16 Pure Dephasing Model

Consider a basis

[
{|n\rangle}
]

and a diagonal Lindblad operator

\sum_n
\ell_n
|n\rangle\langle n|.
\tag{B.26}
]

Assume the Hamiltonian is also diagonal:

\sum_n
E_n
|n\rangle\langle n|.
]

The matrix elements of the density operator satisfy

\frac{1}{2}
|\ell_m-\ell_n|^2
\rho_{mn}.
\tag{B.27}
]

The solution is

\rho_{mn}(0)
\exp
\left[
-\frac{i(E_m-E_n)t}{\kappa_S}
\right]
\exp
\left[
-\frac{1}{2}
|\ell_m-\ell_n|^2t
\right].
\tag{B.28}
]

For (m\neq n), the off-diagonal terms decay when

[
\ell_m\neq\ell_n.
]

The diagonal terms satisfy

[
\frac{d\rho_{nn}}{dt}=0.
]

This is pure dephasing: phase coherence is lost without changing the basis-state probabilities.

B.17 Interpretation of Decoherence in SQFT

Suppose the basis states represent mutually exclusive institutional configurations:

[
|1\rangle,
|2\rangle,
\ldots,
|N\rangle.
]

A coherent density operator may initially contain off-diagonal terms:

\sum_n
p_n|n\rangle\langle n|
+
\sum_{m\neq n}
c_{mn}|m\rangle\langle n|.
]

Environmental monitoring suppresses the coefficients (c_{mn}):

[
c_{mn}(t)
\longrightarrow
0.
]

The effective state becomes

\sum_n
p_n|n\rangle\langle n|.
\tag{B.29}
]

This state can be interpreted as a classical probability mixture over stable alternatives.

The mathematical statement concerns diagonalization relative to a selected basis. It does not imply literal microscopic quantum decoherence in society.

B.18 Amplitude-Damping Model

Consider a two-state system with basis

[
|0\rangle,
\qquad
|1\rangle.
]

Let

\sqrt{\gamma}
|0\rangle\langle 1|.
\tag{B.30}
]

The Lindblad equation becomes

\frac{1}{2}
{
|1\rangle\langle1|,
\rho
}
\right).
]

Writing

\begin{pmatrix}
\rho_{00} & \rho_{01} \
\rho_{10} & \rho_{11}
\end{pmatrix},
]

one obtains

-\gamma\rho_{11},
\tag{B.31}
]

\gamma\rho_{11},
\tag{B.32}
]

and

-\frac{\gamma}{2}\rho_{01}.
\tag{B.33}
]

Therefore,

\rho_{11}(0)e^{-\gamma t},
]

1-\rho_{11}(0)e^{-\gamma t},
]

and

\rho_{01}(0)e^{-\gamma t/2}.
]

This model represents irreversible relaxation from one effective state to another.

B.19 Projective Measurement

Let an observable have spectral decomposition

\sum_a
aP_a,
\tag{B.34}
]

where the projectors satisfy

\delta_{ab}P_a,
]

and

I.
]

For a state (\rho), the probability of outcome (a) is

\operatorname{Tr}(P_a\rho).
\tag{B.35}
]

After obtaining outcome (a), the conditional state is

\frac{
P_a\rho P_a
}{
\operatorname{Tr}(P_a\rho)
}.
\tag{B.36}
]

If the outcome is not recorded, the post-measurement state is

\sum_a
P_a\rho P_a.
\tag{B.37}
]

This map removes coherence between different measurement sectors.

B.20 Preservation of Trace under Projective Measurement

Using Equation (B.37),

\sum_a
\operatorname{Tr}
\left(
P_a\rho P_a
\right).
]

Since

[
P_a^2=P_a,
]

cyclicity gives

\operatorname{Tr}
\left(
P_a\rho
\right).
]

Therefore,

\operatorname{Tr}
\left[
\left(
\sum_aP_a
\right)
\rho
\right].
]

Because

[
\sum_aP_a=I,
]

one obtains

\operatorname{Tr}\rho


\tag{B.38}
]

B.21 Generalized Measurement

A generalized measurement is specified by measurement operators

[
{M_a}
]

satisfying

I.
\tag{B.39}
]

The probability of outcome (a) is

\operatorname{Tr}
\left(
M_a\rho M_a^{\dagger}
\right).
\tag{B.40}
]

The conditional state is

\frac{
M_a\rho M_a^{\dagger}
}{
p(a)
}.
\tag{B.41}
]

The associated positive-operator-valued measure elements are

M_a^{\dagger}M_a.
]

They satisfy

[
E_a\geq 0,
]

and

[
\sum_aE_a=I.
]

The outcome probability can also be written as

\operatorname{Tr}(E_a\rho).
\tag{B.42}
]

B.22 Repeated Projective Measurement

Let (P) project onto a selected subspace.

Suppose the system evolves freely for a short interval

\frac{t}{N},
]

and is then projected back into that subspace.

After (N) repetitions, the state is proportional to

[
\left[
P
e^{-iHt/(N\kappa_S)}
P
\right]^N
|\Psi(0)\rangle.
\tag{B.43}
]

If the initial state lies in the projected subspace,

|\Psi(0)\rangle,
]

then frequent projection may suppress transitions out of that subspace.

B.23 Short-Time Survival Probability

Let

e^{-iHt/\kappa_S}
|\Psi(0)\rangle.
]

The survival amplitude is

\langle\Psi(0)|\Psi(t)\rangle.
]

Expanding the exponential,

I

\frac{iHt}{\kappa_S}

\frac{H^2t^2}{2\kappa_S^2}
+
O(t^3).
]

Therefore,

1

\frac{i\langle H\rangle t}{\kappa_S}

\frac{\langle H^2\rangle t^2}{2\kappa_S^2}
+
O(t^3).
]

The survival probability is

|A(t)|^2.
]

Keeping terms through second order,

1

\frac{
(\Delta H)^2
}{
\kappa_S^2
}
t^2
+
O(t^3),
\tag{B.44}
]

where

\langle H^2\rangle

\langle H\rangle^2.
]

Define the characteristic time

\frac{\kappa_S}{\Delta H}.
\tag{B.45}
]

Then

1

\frac{t^2}{\tau_Z^2}
+
O(t^3).
]

B.24 Quantum-Zeno-Type Limit

If the system is measured (N) times over total duration (t), the interval is

\frac{t}{N}.
]

The approximate survival probability after each interval is

\frac{t^2}{N^2\tau_Z^2}.
]

After (N) repetitions,

\frac{t^2}{N^2\tau_Z^2}
\right)^N.
]

Taking logarithms,

\frac{t^2}{N^2\tau_Z^2}
\right).
]

For large (N),

\frac{t^2}{N\tau_Z^2}.
]

Therefore,


\tag{B.46}
]

The corresponding projected evolution is

[
\lim_{N\rightarrow\infty}
\left[
P
e^{-iHt/(N\kappa_S)}
P
\right]^N
|\Psi(0)\rangle.
\tag{B.47}
]

This is the mathematical basis of the quantum-Zeno-type analogy used in SQFT.

Repeated institutional verification, monitoring, or intervention may suppress transitions, but the analogy is formal unless a specific projection map and dynamical model are defined.

B.25 Effective Zeno Hamiltonian

Within the projected subspace, the effective generator is

PHP.
\tag{B.48}
]

Under suitable assumptions,

P
e^{-iH_Zt/\kappa_S}.
\tag{B.49}
]

Thus, frequent projection does not necessarily freeze all evolution. It confines evolution to the projected sector.

This distinction is important.

The Zeno effect suppresses transitions between sectors but may permit nontrivial dynamics within the selected sector.

B.26 Reduced Density Operators

For a composite system

\mathcal{H}_A
\otimes
\mathcal{H}_B,
]

the reduced state of subsystem (A) is

\operatorname{Tr}B(\rho{AB}).
\tag{B.50}
]

Let

[
{|b_j\rangle}
]

be an orthonormal basis of (\mathcal{H}_B).

Then

\sum_j
\langle b_j|
\rho_{AB}
|b_j\rangle.
\tag{B.51}
]

For every observable (O_A) acting only on subsystem (A),

\operatorname{Tr}_A
(\rho_AO_A).
\tag{B.52}
]

The reduced state therefore contains all information needed to predict local measurements on subsystem (A).

B.27 Nonfactorization and Correlation

A product state satisfies

\rho_A\otimes\rho_B.
]

For product observables,

\langle O_A\rangle
\langle O_B\rangle.
]

The connected correlation is

\langle O_AO_B\rangle

\langle O_A\rangle
\langle O_B\rangle.
\tag{B.53}
]

For a product state,

[
C_{AB}=0.
]

However, the converse need not hold for every chosen pair of observables. A state may be correlated even when one particular connected correlation vanishes.

B.28 Separable and Nonseparable States

A bipartite state is separable when it can be written as

\sum_i
p_i
\rho_A^{(i)}
\otimes
\rho_B^{(i)},
\tag{B.54}
]

where

[
p_i\geq 0,
]

and

[
\sum_i p_i=1.
]

A state that does not admit such a decomposition is nonseparable relative to the specified tensor-product structure.

In social applications, the subsystem decomposition must be explicitly justified before nonseparability is interpreted.

B.29 Von Neumann Entropy of a Reduced State

For a pure bipartite state,

|\Psi_{AB}\rangle
\langle\Psi_{AB}|,
]

the total entropy is

[
S(\rho_{AB})=0.
]

The reduced states may nevertheless be mixed:

\operatorname{Tr}B\rho{AB},
]

\operatorname{Tr}A\rho{AB}.
]

For a pure bipartite state,

S(\rho_B).
\tag{B.55}
]

This equality follows from the Schmidt decomposition.

B.30 Schmidt Decomposition

Any pure bipartite state in finite dimensions may be written as

\sum_{r=1}^{R}
\sqrt{\lambda_r}
|u_r\rangle
\otimes
|v_r\rangle,
\tag{B.56}
]

where

[
\lambda_r\geq 0,
]

[
\sum_r\lambda_r=1,
]

and (R) is the Schmidt rank.

The reduced states are

\sum_r
\lambda_r
|u_r\rangle\langle u_r|,
]

and

\sum_r
\lambda_r
|v_r\rangle\langle v_r|.
]

They have the same nonzero eigenvalues, so

S(\rho_B).
]

A pure product state has Schmidt rank one.

B.31 Relative Entropy under a CPTP Map

Let (\mathcal{E}) be a completely positive trace-preserving map.

The data-processing inequality states

[
D(\rho|\sigma)
\geq
D
\left(
\mathcal{E}(\rho)
|
\mathcal{E}(\sigma)
\right).
\tag{B.57}
]

This means that admissible processing cannot increase the distinguishability between two states.

In SQFT, coarse-graining, partial observation, or aggregation may therefore reduce the empirical distinguishability of competing latent states.

B.32 Coarse-Graining as a Channel

Let

[
\mathcal{R}
]

be a CPTP map representing coarse-graining.

Then

\mathcal{R}(\rho_{\mathrm{fine}}).
\tag{B.58}
]

If two fine-scale states are (\rho) and (\sigma), then

[
D(\rho|\sigma)
\geq
D
\left(
\mathcal{R}(\rho)
|
\mathcal{R}(\sigma)
\right).
\tag{B.59}
]

Thus, coarse-graining generally removes information.

It may map several microscopically distinct states to the same effective state.

B.33 Coarse-Graining and Time Evolution

Let

[
\mathcal{U}_t
]

denote fine-scale evolution and

[
\mathcal{R}
]

denote coarse-graining.

Two possible procedures are:

\mathcal{R}
\mathcal{U}_t(\rho_0),
]

and

\mathcal{U}^{\mathrm{eff}}_t
\mathcal{R}(\rho_0).
]

Exact scale consistency requires

\mathcal{U}^{\mathrm{eff}}_t
\mathcal{R}.
\tag{B.60}
]

The coarse-graining error is

\mathcal{R}\mathcal{U}_t(\rho_0)

\mathcal{U}^{\mathrm{eff}}_t
\mathcal{R}(\rho_0).
\tag{B.61}
]

A useful bound is

[
|\Delta_t(\rho_0)|_1
\leq
\varepsilon(t).
\tag{B.62}
]

If (\varepsilon(t)) is small over the relevant time interval, the effective model is dynamically reliable.

B.34 Projection-Based Coarse-Graining

Let (P) be a projection onto a reduced subspace.

A simple projection-based reduction is

[
\rho
\longmapsto
P\rho P.
]

However,

[
\operatorname{Tr}(P\rho P)
]

may be less than one.

A normalized conditional reduction is

\frac{
P\rho P
}{
\operatorname{Tr}(P\rho)
}.
\tag{B.63}
]

This is nonlinear because of the normalization denominator.

An unconditional trace-preserving projection channel requires a complete family of projectors:

\sum_a
P_a\rho P_a.
\tag{B.64}
]

B.35 Nakajima–Zwanzig Projection Identity

Let the full state satisfy

\mathcal{L}\rho.
]

Introduce a projection superoperator (\mathcal{P}) and its complement

I-\mathcal{P}.
]

Then

\mathcal{P}\mathcal{L}\mathcal{P}\rho
+
\mathcal{P}\mathcal{L}\mathcal{Q}\rho,
]

and

\mathcal{Q}\mathcal{L}\mathcal{P}\rho
+
\mathcal{Q}\mathcal{L}\mathcal{Q}\rho.
]

Solving the second equation formally and substituting it into the first yields

\mathcal{P}\mathcal{L}\mathcal{P}\rho(t)
+
\int_0^t
\mathcal{K}(t-s)
\mathcal{P}\rho(s),ds
+
\eta(t),
\tag{B.65}
]

where

\mathcal{P}\mathcal{L}
e^{t\mathcal{Q}\mathcal{L}}
\mathcal{Q}\mathcal{L}\mathcal{P}
\tag{B.66}
]

is the memory kernel, and

\mathcal{P}\mathcal{L}
e^{t\mathcal{Q}\mathcal{L}}
\mathcal{Q}\rho(0)
\tag{B.67}
]

depends on unresolved initial components.

This identity shows how memory can emerge from eliminating hidden degrees of freedom.

B.36 Markovian Approximation

The exact reduced equation contains a memory integral:

[
\int_0^t
\mathcal{K}(t-s)
\mathcal{P}\rho(s),ds.
]

If the memory kernel decays rapidly relative to the timescale of the reduced state, one may approximate

[
\mathcal{P}\rho(s)
\approx
\mathcal{P}\rho(t)
]

inside the integral.

Then

[
\int_0^t
\mathcal{K}(t-s)
\mathcal{P}\rho(s),ds
\approx
\left[
\int_0^{\infty}
\mathcal{K}(u),du
\right]
\mathcal{P}\rho(t).
\tag{B.68}
]

This produces a time-local effective generator.

The Markovian approximation is therefore not fundamental. It is a timescale-separation approximation.

B.37 Fractional Dynamics from Long Memory

Suppose the memory kernel decays as a power law:

[
\mathcal{K}(t)
\sim
t^{-\alpha},
\qquad
0<\alpha<1.
]

Such slow decay may lead to a fractional evolution equation.

The Caputo derivative is

\frac{1}{
\Gamma(1-\alpha)
}
\int_0^t
\frac{
f'(s)
}{
(t-s)^{\alpha}
}
,ds.
\tag{B.69}
]

A fractional SQFT equation may be written as

\mathcal{L}\rho(t).
\tag{B.70}
]

For (\alpha=1), this reduces to ordinary first-order evolution.

For (0<\alpha<1), the derivative depends on the full history of the state.

B.38 Solution of a Linear Fractional Equation

Consider

\lambda x(t),
]

with

[
x(0)=x_0.
]

The solution is

x_0
E_{\alpha}
(\lambda t^{\alpha}),
\tag{B.71}
]

where the Mittag–Leffler function is

\sum_{n=0}^{\infty}
\frac{
z^n
}{
\Gamma(\alpha n+1)
}.
\tag{B.72}
]

When

[
\alpha=1,
]

one has

[
E_1(z)=e^z.
]

Fractional evolution therefore replaces exponential relaxation with Mittag–Leffler relaxation.

B.39 Duhamel Formula

Consider the inhomogeneous equation

Au+f(t),
]

with

[
u(0)=u_0.
]

If (A) generates a semigroup (T(t)), then

T(t)u_0
+
\int_0^t
T(t-s)f(s),ds.
\tag{B.73}
]

This is the variation-of-constants or Duhamel formula.

For a nonlinear equation

Au+N(u),
]

the mild solution satisfies

T(t)u_0
+
\int_0^t
T(t-s)N(u(s)),ds.
\tag{B.74}
]

B.40 Local Existence by Contraction Mapping

Define

T(t)u_0
+
\int_0^t
T(t-s)N(u(s)),ds.
]

Suppose

[
|T(t)|
\leq
M
]

for

[
0\leq t\leq T,
]

and (N) is locally Lipschitz:

[
|N(u)-N(v)|
\leq
L|u-v|.
]

Then

[
|\mathcal{T}u-\mathcal{T}v|
\leq
MLT
|u-v|.
]

If

[
MLT<1,
]

then (\mathcal{T}) is a contraction.

By the Banach fixed-point theorem, a unique local mild solution exists.

B.41 Stability Estimate

Let (u(t)) and (v(t)) solve

Au+N(u),
]

and

Av+N(v).
]

Suppose

[
|T(t)|
\leq
Me^{\omega t}
]

and (N) is Lipschitz with constant (L).

Then

[
|u(t)-v(t)|
\leq
Me^{\omega t}
|u_0-v_0|
+
ML
\int_0^t
e^{\omega(t-s)}
|u(s)-v(s)|,ds.
]

Applying Grönwall’s inequality gives an estimate of the form

[
|u(t)-v(t)|
\leq
M
\exp
\left[
(\omega+ML)t
\right]
|u_0-v_0|.
\tag{B.75}
]

The solution depends continuously on its initial condition.

B.42 Generator-Perturbation Bound

Let

\mathcal{L}_0\rho_0,
]

and

\mathcal{L}{\theta}\rho{\theta}.
]

Assume both semigroups are contractive in the trace norm.

Using Duhamel’s formula,

e^{t\mathcal{L}_{\theta}}

\int_0^t
e^{(t-s)\mathcal{L}0}
(\mathcal{L}0-\mathcal{L}{\theta})
e^{s\mathcal{L}{\theta}}
,ds.
\tag{B.76}
]

Therefore,

e^{t\mathcal{L}_{\theta}}
\right|
\leq
t
|\mathcal{L}0-\mathcal{L}{\theta}|
\tag{B.77}
]

under unit contractivity.

For the same initial state (\rho(0)),

[
|\rho_0(t)-\rho_{\theta}(t)|_1
\leq
t
|\mathcal{L}0-\mathcal{L}{\theta}|.
\tag{B.78}
]

More general semigroup bounds yield

[
|\rho_0(t)-\rho_{\theta}(t)|_1
\leq
C(t)
|\mathcal{L}0-\mathcal{L}{\theta}|.
]

B.43 Gauge Invariance of Expectation Values

Let

U|\Psi\rangle,
]

and

UOU^{\dagger},
]

where (U) is unitary.

Then

\langle\Psi|
U^{\dagger}
UOU^{\dagger}
U
|\Psi\rangle.
]

Using

[
U^{\dagger}U=I,
]

one obtains

\langle\Psi|O|\Psi\rangle.
\tag{B.79}
]

For density operators,

U\rho U^{\dagger},
]

and

UOU^{\dagger}.
]

Then

\operatorname{Tr}(\rho O).
\tag{B.80}
]

Gauge-related representations therefore yield identical expectation values.

B.44 Invariance of the Spectrum

If

UAU^{\dagger},
]

then (A) and (A') have the same spectrum.

Suppose

\lambda|v\rangle.
]

Then

UAU^{\dagger}U|v\rangle

UA|v\rangle

\lambda U|v\rangle.
]

Therefore, (\lambda) is also an eigenvalue of (A').

Thus,

\sigma(A).
\tag{B.81}
]

Spectral quantities are natural candidates for gauge-invariant model summaries.

B.45 Preservation of Commutation Relations

Let

[
A'=UAU^{\dagger},
]

and

[
B'=UBU^{\dagger}.
]

Then

UAU^{\dagger}UBU^{\dagger}

UBU^{\dagger}UAU^{\dagger}.
]

Using

[
U^{\dagger}U=I,
]

one obtains

U[A,B]U^{\dagger}.
\tag{B.82}
]

Hence,

[
[A,B]=0
]

if and only if

[
[A',B']=0.
]

Commutativity is preserved under unitary changes of representation.

B.46 Classical Limit of Commuting Observables

Suppose the observable algebra is generated by mutually commuting self-adjoint operators:

[
[O_i,O_j]=0.
]

In finite dimensions, they can be simultaneously diagonalized.

There exists an orthonormal basis

[
{|n\rangle}
]

such that

\sum_n
o_i(n)
|n\rangle\langle n|.
]

For a density operator (\rho), define

\langle n|\rho|n\rangle.
]

Then

[
p_n\geq 0,
]

and

[
\sum_n p_n=1.
]

The expectation value becomes

\sum_n
p_n o_i(n).
\tag{B.83}
]

This is identical to the classical expectation of a random variable (o_i(n)) on a finite probability space.

Thus, when all relevant observables commute, a classical probabilistic representation is sufficient.

B.47 Emergence of a Classical Master Equation

Suppose (\rho) is diagonal in the basis ({|n\rangle}):

\sum_n
p_n
|n\rangle\langle n|.
]

Let the jump operators be

\sqrt{W_{mn}}
|m\rangle\langle n|,
]

where (W_{mn}\geq 0) is the transition rate from state (n) to state (m).

Substituting into the Lindblad equation and evaluating diagonal components yields

W_{nm}p_m
\right).
\tag{B.84}
]

This is a classical continuous-time master equation.

The first term represents inflow into state (m).

The second term represents outflow from state (m).

Equation (B.84) demonstrates how a classical Markov process can emerge as the diagonal sector of Lindblad dynamics.

B.48 Conservation of Total Probability in the Classical Limit

Summing Equation (B.84) over (m),

\sum_{m,n}
W_{nm}p_m.
]

Relabeling indices in the second sum shows that the two terms cancel.

Therefore,


\tag{B.85}
]

If

[
\sum_m p_m(0)=1,
]

then

[
\sum_m p_m(t)=1.
]

B.49 Detailed Balance

A stationary distribution (p^*) satisfies detailed balance when

W_{nm}p_m^*
\tag{B.86}
]

for every pair (m,n).

Under detailed balance, each transition flow is balanced by its reverse flow.

Substituting into Equation (B.84) gives


]

Detailed balance is stronger than stationarity. A stationary process may possess nonzero circulating probability currents and fail to satisfy detailed balance.

B.50 Entropy Production in a Classical Master Equation

For the probability distribution (p_n(t)), define Shannon entropy:


\sum_n
p_n(t)\ln p_n(t).
\tag{B.87}
]

For Markov dynamics, entropy production can be written in terms of probability currents.

Define

W_{mn}p_n

W_{nm}p_m.
]

A standard nonnegative entropy-production rate is

\frac{1}{2}
\sum_{m,n}
J_{mn}
\ln
\left(
\frac{
W_{mn}p_n
}{
W_{nm}p_m
}
\right).
\tag{B.88}
]

Under suitable positivity assumptions,

[
\sigma\geq 0.
]

This quantity vanishes under detailed balance.

B.51 BCH Expansion

For noncommuting operators (A) and (B),

[
e^Ae^B
\neq
e^{A+B}
]

in general.

The Baker–Campbell–Hausdorff formula begins as

A+B
+
\frac{1}{2}[A,B]
+
\frac{1}{12}[A,[A,B]]
+
\frac{1}{12}[B,[B,A]]
+\cdots.
\tag{B.89}
]

This expansion explains why the order of noncommuting interventions matters.

If

[
[A,B]=0,
]

then

e^{A+B}.
]

B.52 Trotter Product Formula

Let (A) and (B) be suitable generators.

The Trotter formula is

\lim_{N\rightarrow\infty}
\left(
e^{tA/N}
e^{tB/N}
\right)^N.
\tag{B.90}
]

For Hamiltonian evolution,

\lim_{N\rightarrow\infty}
\left(
e^{-itH_1/(N\kappa_S)}
e^{-itH_2/(N\kappa_S)}
\right)^N.
\tag{B.91}
]

This formula provides a basis for numerical operator splitting and multicomponent simulations.

B.53 First-Order Splitting Error

For small (\Delta t),

e^{\Delta t A}
e^{\Delta t B}
+
O(\Delta t^2).
]

The leading error depends on the commutator:

e^{
\Delta t(A+B)
+
\frac{\Delta t^2}{2}[A,B]
+
O(\Delta t^3)
}.
\tag{B.92}
]

Thus, splitting errors become larger when (A) and (B) strongly fail to commute.

B.54 Symmetric Strang Splitting

A second-order approximation is

e^{\Delta t A/2}
e^{\Delta t B}
e^{\Delta t A/2}
+
O(\Delta t^3).
\tag{B.93}
]

This method is often more accurate and preserves structural properties better than first-order splitting.

For SQFT computation, one may separate:

  • Hamiltonian evolution;
  • dissipative evolution;
  • network evolution;
  • external forcing.

B.55 Path-Integral Stationary-Phase Approximation

Consider the formal generating amplitude

\int
\mathcal{D}\Phi
\exp
\left(
\frac{i}{\kappa_S}
S[\Phi]
\right).
\tag{B.94}
]

Suppose the dominant contribution comes from a stationary configuration (\Phi_{\mathrm{cl}}) satisfying


\tag{B.95}
]

Write

\Phi_{\mathrm{cl}}
+
\eta.
]

Expanding the action,

S[\Phi_{\mathrm{cl}}]
+
\frac{1}{2}
\langle
\eta,
S^{(2)}[\Phi_{\mathrm{cl}}]\eta
\rangle
+
O(\eta^3).
\tag{B.96}
]

The linear term vanishes because (\Phi_{\mathrm{cl}}) is stationary.

The leading approximation is therefore

[
Z
\approx
e^{iS[\Phi_{\mathrm{cl}}]/\kappa_S}
\int
\mathcal{D}\eta
\exp
\left[
\frac{i}{2\kappa_S}
\langle
\eta,
S^{(2)}\eta
\rangle
\right].
\tag{B.97}
]

The dominant path is the solution of the Euler–Lagrange equation.

B.56 Classical Limit by Stationary Phase

When (\kappa_S) is formally small, rapidly oscillating phases cancel except near stationary points of the action.

Thus,

[
\kappa_S\rightarrow 0
]

selects configurations satisfying

[
\frac{\delta S}{\delta\Phi}=0.
\tag{B.98}
]

This is the formal sense in which classical field equations emerge from a path-integral description.

In SQFT, this limit represents the dominance of sharply defined effective trajectories over broad distributions of alternatives.

It should not be interpreted as a literal physical limit unless (\kappa_S) has been empirically defined.

B.57 Euler–Lagrange Field Equation

Let

\int
\mathcal{L}
\left(
\Phi,
\partial_{\mu}\Phi,
x
\right)
d^nx.
]

Under

[
\Phi
\longrightarrow
\Phi+\varepsilon\eta,
]

the first variation is

\int
\left[
\frac{\partial\mathcal{L}}{\partial\Phi}\eta
+
\frac{\partial\mathcal{L}}
{\partial(\partial_{\mu}\Phi)}
\partial_{\mu}\eta
\right]
d^nx.
]

Integrating the second term by parts gives

\partial_{\mu}
\left(
\frac{\partial\mathcal{L}}
{\partial(\partial_{\mu}\Phi)}
\right)
\right]
\eta,d^nx
+
\text{boundary term}.
]

If (\eta) vanishes on the boundary, stationarity implies


\tag{B.99}
]

B.58 Example: Nonlinear Scalar Social Field

Consider the Lagrangian density

V(\Phi),
]

with potential

J\Phi.
\tag{B.100}
]

The Euler–Lagrange equation is

J


\tag{B.101}
]

In a nonrelativistic diffusion-type formulation, one may instead use

D\nabla^2\Phi

m^2\Phi

\lambda\Phi^3
+
J.
\tag{B.102}
]

The parameters may be interpreted structurally as:

  • (D): propagation or diffusion strength;
  • (m^2): restoring or suppressing tendency;
  • (\lambda): nonlinear self-interaction;
  • (J): external forcing.

B.59 Symmetry Breaking in a Scalar Potential

Consider

-\frac{\mu^2}{2}\Phi^2
+
\frac{\lambda}{4}\Phi^4,
\qquad
\mu^2>0,
\quad
\lambda>0.
\tag{B.103}
]

Stationary points satisfy


]

Thus,


]

The solutions are

[
\Phi=0,
]

and

\pm
\frac{\mu}{\sqrt{\lambda}}.
\tag{B.104}
]

The second derivative is

-\mu^2+3\lambda\Phi^2.
]

At (\Phi=0),

-\mu^2<0,
]

so the origin is unstable.

At

\pm
\frac{\mu}{\sqrt{\lambda}},
]

one has

2\mu^2>0.
]

These are stable minima.

The equations are symmetric under

[
\Phi\longrightarrow-\Phi,
]

but each stable solution selects one branch.

B.60 Linear Stability around an Equilibrium

Consider

F(\Phi).
]

Let (\Phi_*) satisfy

[
F(\Phi_*)=0.
]

Write

\Phi_*
+
\delta\Phi.
]

Expanding,

F(\Phi_)
+
DF(\Phi_)\delta\Phi
+
O(|\delta\Phi|^2).
]

Since (F(\Phi_*)=0),

DF(\Phi_*)\delta\Phi.
\tag{B.105}
]

If all eigenvalues of (DF(\Phi_*)) have negative real parts, the equilibrium is locally asymptotically stable.

If at least one eigenvalue has positive real part, the equilibrium is unstable.

B.61 Lyapunov Stability for Gradient Flow

Suppose


\nabla V(\Phi).
\tag{B.106}
]

Then

\nabla V
\cdot
\frac{d\Phi}{dt}.
]

Substituting Equation (B.106),


|\nabla V|^2
\leq
0.
\tag{B.107}
]

Thus, (V) is a Lyapunov function.

The system moves toward lower values of the potential, and stationary points satisfy

[
\nabla V=0.
]

B.62 Free-Energy Dissipation

Consider a probability density (p(x,t)) satisfying a Fokker–Planck equation:

\nabla\cdot
\left(
p\nabla V
+
D\nabla p
\right).
\tag{B.108}
]

Define the free-energy functional

\int
p(x)V(x),dx
+
D
\int
p(x)\ln p(x),dx.
\tag{B.109}
]

Under suitable boundary conditions,

[
\frac{d\mathcal{F}}{dt}
\leq
0.
\tag{B.110}
]

The equilibrium distribution is

\frac{1}{Z}
e^{-V(x)/D}.
\tag{B.111}
]

This structure connects potential dynamics, diffusion, entropy, and equilibrium.

B.63 Continuity Equation

Let (\rho(x,t)) be a conserved density and (j(x,t)) the associated current.

Local conservation is expressed by


\tag{B.112}
]

Integrating over a region (\Omega),


\int_{\Omega}
\nabla\cdot j,dx.
]

Using the divergence theorem,


\int_{\partial\Omega}
j\cdot n,dS.
\tag{B.113}
]

The change inside the region equals the net flux across its boundary.

B.64 Noether-Type Conservation

Suppose the action is invariant under a continuous transformation

[
\Phi
\longrightarrow
\Phi+\varepsilon\delta\Phi.
]

If the Lagrangian changes by a total divergence,

\partial_{\mu}K^{\mu},
]

then the Noether current is

K^{\mu}.
\tag{B.114}
]

On solutions of the Euler–Lagrange equation,


\tag{B.115}
]

The conserved charge is

\int_{\Sigma}
j^0,d^{n-1}x.
\tag{B.116}
]

B.65 Symmetry and Generator Commutation

Let (Q) generate a continuous unitary symmetry:

e^{-i\varepsilon Q/\kappa_S}.
]

If the Hamiltonian is invariant,

H,
]

then differentiating at (\varepsilon=0) gives

[
[Q,H]=0.
\tag{B.117}
]

In the Heisenberg picture,

\frac{i}{\kappa_S}[H,Q].
]

Therefore,

[
[H,Q]=0
\quad\Longrightarrow\quad
\frac{dQ}{dt}=0.
\tag{B.118}
]

A continuous symmetry produces a conserved generator.

B.66 Renormalization Linearization near a Fixed Point

Let the running couplings satisfy

\beta_i(g),
\tag{B.119}
]

where (\ell) is logarithmic scale.

A fixed point (g^*) satisfies

[
\beta_i(g^*)=0.
]

Write

g_i^*
+
\delta g_i.
]

Linearizing,

\sum_j
B_{ij}
\delta g_j,
\tag{B.120}
]

where

\left.
\frac{\partial\beta_i}{\partial g_j}
\right|_{g=g^*}.
]

If (v^{(a)}) is an eigenvector of (B) with eigenvalue (y_a), then

e^{y_a\ell}
\delta g^{(a)}(0).
\tag{B.121}
]

A direction is:

  • relevant if (y_a>0);
  • irrelevant if (y_a<0);
  • marginal if (y_a=0).

B.67 Universality near an Infrared Fixed Point

Suppose two microscopic models differ only in irrelevant couplings.

Under repeated coarse-graining,

e^{y_{\mathrm{irr}}\ell}
\delta g_{\mathrm{irr}}(0),
]

with

[
y_{\mathrm{irr}}<0.
]

As

[
\ell\rightarrow\infty,
]

one has

[
\delta g_{\mathrm{irr}}(\ell)
\longrightarrow
0.
]

Therefore, microscopic differences disappear at large scales.

The long-distance behavior depends primarily on:

  • relevant couplings;
  • marginal couplings;
  • symmetry;
  • effective dimension;
  • conservation laws.

This is the mathematical basis of universality.

B.68 Information Loss under Model Reduction

Let the full parameter be (\theta), and let the reduced representation be

R(\theta).
]

If (R) is not injective, then distinct microscopic parameters may satisfy

R(\theta_2).
]

The reduced model cannot distinguish (\theta_1) from (\theta_2).

The fibers

{\theta(\theta)=\phi}
\tag{B.122}
]

represent equivalence classes of microscopic descriptions with the same macroscopic representation.

B.69 Local Identifiability by the Jacobian

Let

[
F:\Theta\rightarrow\mathcal{Y}
]

map parameters to observable predictions.

The Jacobian is

\frac{\partial F}{\partial\theta}.
\tag{B.123}
]

If

\dim\Theta,
\tag{B.124}
]

then (F) is locally injective near (\theta_0) under the inverse-function theorem.

If the Jacobian is rank deficient, there exists a nonzero vector (v) such that

[
J(\theta_0)v=0.
]

Perturbations in direction (v) are locally invisible to the observations.

B.70 Fisher Information and Local Identifiability

For likelihood

[
p(Y|\theta),
]

the Fisher information matrix is

\mathbb{E}_{\theta}
\left[
\frac{\partial\ln p(Y|\theta)}
{\partial\theta_i}
\frac{\partial\ln p(Y|\theta)}
{\partial\theta_j}
\right].
\tag{B.125}
]

If (I(\theta)) is singular, some parameter directions cannot be estimated with finite asymptotic variance.

A nonsingular Fisher matrix is therefore a standard local identifiability condition.

Gauge redundancy often appears as zero eigenvalues of the Fisher information matrix.

B.71 Sparse Recovery Condition

Suppose

[
Y=X\theta+\varepsilon
]

and (\theta) is sparse.

The LASSO estimator is

\operatorname*{arg,min}_{\theta}
\left[
\frac{1}{2}
|Y-X\theta|_2^2
+
\lambda|\theta|_1
\right].
\tag{B.126}
]

Exact or stable sparse recovery requires structural conditions on the design matrix, such as:

  • restricted isometry;
  • restricted eigenvalue conditions;
  • mutual incoherence;
  • sufficient sample size.

The operator-discovery problem in SQFT is an extension in which columns of (X) correspond to candidate dynamical operators.

B.72 Low-Rank Reconstruction

Suppose the state (\rho) is approximately low rank.

A nuclear-norm estimator may be written as

\operatorname*{arg,min}_{\rho}
\left[
|Y-\mathcal{H}(\rho)|2^2
+
\lambda|\rho|*
\right],
\tag{B.127}
]

subject to

[
\rho\geq 0,
]

and

[
\operatorname{Tr}\rho=1.
]

The nuclear norm is

\operatorname{Tr}
\sqrt{\rho^{\dagger}\rho}.
]

For a positive operator,

\operatorname{Tr}\rho.
]

Because the trace is fixed, low-rank estimation may instead require entropy penalties, factorized parameterizations, or constraints on effective dimension.

B.73 Model Comparison

Suppose two models (M_1) and (M_2) generate predictive distributions

[
p(Y_{\mathrm{test}}|D_{\mathrm{train}},M_1)
]

and

[
p(Y_{\mathrm{test}}|D_{\mathrm{train}},M_2).
]

A model is empirically preferable when it provides better out-of-sample predictive performance under a prespecified scoring rule.

The expected log score is

\mathbb{E}
\left[
\ln
p(Y_{\mathrm{test}}|D_{\mathrm{train}},M)
\right].
\tag{B.128}
]

SQFT should be compared against simpler alternatives using the same data splits, observables, and validation criteria.

B.74 Falsifiability Condition

Let a model predict an admissible set of distributions

[
\mathcal{P}_M.
]

The model is empirically falsifiable only if

[
\mathcal{P}M
\neq
\mathcal{P}{\mathrm{all}},
]

where (\mathcal{P}_{\mathrm{all}}) is the set of every possible observable distribution.

A model that can reproduce every dataset imposes no empirical restriction.

A practical falsification rule may take the form

[
T(Y)>c_{\alpha},
\tag{B.129}
]

where (T) is a prespecified test statistic and (c_{\alpha}) is a rejection threshold.

B.75 Summary of Proven Structural Properties

Under the assumptions stated in this appendix, the following results hold.

Closed-System Dynamics

If

[
H=H^{\dagger},
]

then evolution generated by

e^{-itH/\kappa_S}
]

is unitary.

Consequently:

|\Psi(0)|,
]

1,
]

and

\operatorname{Tr}(\rho(0)^2).
]

Markovian Open-System Dynamics

If the generator has Lindblad form, then the evolution is:

  • trace preserving;
  • Hermiticity preserving;
  • completely positive;
  • compatible with probabilistic interpretation.

Measurement

Projective and generalized measurements preserve total probability when their completeness relations are satisfied.

Decoherence

Diagonal Lindblad operators suppress off-diagonal matrix elements in their preferred basis.

Classical Limit

If relevant observables commute and the density operator is effectively diagonal, the model reduces to a classical probability process.

Coarse-Graining

A CPTP coarse-graining map cannot increase relative-entropy distinguishability.

Stability

Local Lipschitz continuity and semigroup bounds provide local existence, uniqueness, and continuous dependence on initial conditions.

B.76 Limitations of the Derivations

The derivations above do not automatically extend to every infinite-dimensional SQFT model.

Additional issues include:

  • unbounded operators;
  • operator-domain compatibility;
  • nonclosable generators;
  • ultraviolet divergences;
  • nonexistence of path-integral measures;
  • non-Markovian loss of complete positivity;
  • time-dependent Hilbert spaces;
  • nonlinear density-operator equations;
  • changing state-space dimension.

Therefore, formal calculations must not be mistaken for complete proofs in the absence of explicit analytical assumptions.

B.77 Final Mathematical Principle

The formal structure of SQFT should obey the following rule:

[
\text{Every dynamical equation must preserve}
]

[
\text{the mathematical conditions required by its interpretation.}
]

For a density operator, this means at minimum:

[
\rho(t)\geq 0,
]

[
\rho(t)=\rho(t)^{\dagger},
]

and

[
\operatorname{Tr}\rho(t)=1.
\tag{B.130}
]

For an empirical model, it additionally means:

[
\text{identifiability},
]

[
\text{testability},
]

[
\text{comparability},
]

and

[
\text{reproducibility}.
]

The mathematical legitimacy of SQFT depends not on borrowing advanced notation, but on proving that its transformations preserve the structures the notation is intended to represent.


Appendix C

Notation, Symbol Conventions, and Model Specification Standards

C.1 Purpose of This Appendix

A formal theory becomes difficult to verify when the same symbol is used for several unrelated objects or when closely related objects are denoted inconsistently across chapters.

Social Quantum Field Theory uses notation drawn from:

  • quantum theory;
  • field theory;
  • open-system dynamics;
  • differential geometry;
  • probability theory;
  • statistical inference;
  • network science;
  • renormalization theory.

This appendix establishes a unified notation system for the entire manuscript.

The objectives are:

  1. to distinguish states from observables;
  2. to distinguish fields from operators;
  3. to distinguish dynamical generators from statistical loss functions;
  4. to distinguish physical analogy from empirical implementation;
  5. to define the minimum specification required for every SQFT model.

C.2 General Typographic Conventions

The following typographic conventions are adopted.

Scalars

Ordinary scalar quantities are written in italic form:

[
a,
\qquad
b,
\qquad
t,
\qquad
\lambda,
\qquad
\mu.
]

Vectors

Finite-dimensional vectors are written in bold lowercase form:

[
\mathbf{x},
\qquad
\mathbf{y},
\qquad
\boldsymbol{\theta}.
]

When Dirac notation is used, state vectors are written as

[
|\Psi\rangle.
]

Matrices

Finite-dimensional matrices are written in uppercase italic or bold uppercase form:

[
A,
\qquad
B,
\qquad
\mathbf{K}.
]

Operators

Linear operators on Hilbert space are written with uppercase letters, optionally with hats when emphasis is required:

[
H,
\qquad
O,
\qquad
\hat{H},
\qquad
\hat{O}.
]

The manuscript should not alternate between (H) and (\hat{H}) without reason.

The preferred convention is:

[
H
]

for the Hamiltonian in formal equations, and

[
\hat{H}
]

only when distinguishing the operator from a scalar Hamiltonian function.

Superoperators

Maps acting on operators are written in calligraphic form:

[
\mathcal{L},
\qquad
\mathcal{E},
\qquad
\mathcal{R},
\qquad
\mathcal{U}_t.
]

Functionals

Functionals acting on fields or distributions are written in script or calligraphic form:

[
S[\Phi],
\qquad
\mathcal{F}[\rho],
\qquad
\Gamma[\Phi].
]

Sets and Spaces

General sets are written as

[
X,
\qquad
Y,
\qquad
\Omega.
]

State spaces and manifolds are written as

[
\mathcal{H},
\qquad
\mathcal{M},
\qquad
\mathcal{N}.
]

Operator algebras are written as

[
\mathfrak{A},
\qquad
\mathfrak{g}.
]

C.3 Core State Variables

The fundamental SQFT state variables are distinguished as follows.

State Vector

A pure state is written as

[
|\Psi(t)\rangle
\in
\mathcal{H}.
]

Its normalization condition is


]

Density Operator

A general state is written as

[
\rho(t).
]

It must satisfy

[
\rho(t)\geq 0,
]

[
\rho(t)=\rho(t)^{\dagger},
]

and

[
\operatorname{Tr}\rho(t)=1.
]

Classical Probability Vector

A classical probability distribution over discrete states is written as

\bigl(
p_1(t),\ldots,p_N(t)
\bigr).
]

It satisfies

[
p_i(t)\geq 0,
]

and

[
\sum_i p_i(t)=1.
]

Social Field

A social field is written as

[
\Phi(x,t),
]

where (x) may denote:

  • geographic position;
  • institutional position;
  • network coordinate;
  • configuration-space coordinate;
  • information-geometric coordinate.

The field may be scalar, vector-valued, tensor-valued, or operator-valued.

C.4 Distinguishing the Field from the State

The symbols (\Phi) and (\rho) must not be treated as interchangeable.

A field configuration is written as

[
\Phi.
]

A statistical or operator state over possible field configurations is written as

[
\rho.
]

For example, (\rho) may encode uncertainty over different field configurations:

\sum_i
p_i
|\Phi_i\rangle
\langle\Phi_i|.
]

Thus:

  • (\Phi_i) denotes a possible configuration;
  • (p_i) denotes its weight;
  • (\rho) denotes the complete state description.

In a classical field model, the probability functional may instead be written as

[
P[\Phi].
]

C.5 Configuration Space and Social Manifold

The social manifold is denoted by

[
\mathcal{M}.
]

A point

[
x\in\mathcal{M}
]

represents a local social, institutional, informational, or relational configuration.

The field is a map

[
\Phi:
\mathcal{M}
\rightarrow
\mathcal{N},
]

where (\mathcal{N}) is the target space.

If time is included explicitly,

[
\Phi:
\mathcal{M}\times\mathbb{R}
\rightarrow
\mathcal{N}.
]

The symbol (M) should not be used simultaneously for the manifold and for a measurement operator.

Measurement operators should be denoted by

[
M_a.
]

The manifold should always be denoted by

[
\mathcal{M}.
]

C.6 Time Variables

Physical or model time is denoted by

[
t.
]

A second time variable may be denoted by

[
s.
]

Initial time is written as

[
t_0.
]

A discrete time index is written as

[
n.
]

Renormalization-group scale time is denoted by

[
\ell.
]

It is commonly defined as

\ln
\left(
\frac{\mu_0}{\mu}
\right).
]

The scale variable (\mu) must not be confused with a probability measure.

When both occur in the same section, the measure should be denoted by

[
\nu
]

or

[
dP.
]

C.7 The SQFT Action Scale

The formal SQFT action scale is denoted by

[
\kappa_S.
]

The Schrödinger-type equation is

H|\Psi(t)\rangle.
]

The closed density-operator equation is

-\frac{i}{\kappa_S}
[H,\rho].
]

The symbol (\kappa_S) is preferred to (\hbar) in social applications because it prevents accidental interpretation as the physical Planck constant.

The manuscript should explicitly state:

[
\kappa_S
\neq
\hbar
]

unless a literal physical model is intended.

The value and units of (\kappa_S) must be specified in any empirical implementation.

C.8 Hamiltonians and Generators

The Hamiltonian is denoted by

[
H.
]

A time-dependent Hamiltonian is

[
H(t).
]

A field-dependent Hamiltonian is

[
H[\Phi].
]

The open-system generator is denoted by

[
\mathcal{L}.
]

The Lindblad equation is

\mathcal{L}(\rho).
]

Its decomposition is

\mathcal{L}_{H}
+
\mathcal{D},
]

where

-\frac{i}{\kappa_S}
[H,\rho]
]

and

\frac{1}{2}
{L_k^{\dagger}L_k,\rho}
\right).
]

The symbol (\mathcal{L}) should not also be used for a statistical loss function.

A statistical loss function should be denoted by

[
\mathscr{L},
]

or

[
\mathcal{J}.
]

C.9 Lindblad Operators

Lindblad or jump operators are denoted by

[
L_k.
]

The index (k) labels distinct environmental channels.

Examples include:

Dephasing Channel

\sqrt{\gamma_k}
P_k.
]

Transition Channel

\sqrt{W_{mn}}
|m\rangle\langle n|.
]

Collective Channel

\sum_i
c_iO_i.
]

Every use of (L_k) should specify:

  • its domain;
  • its empirical interpretation;
  • its associated rate;
  • whether it is local or collective;
  • whether it is estimated or imposed.

C.10 Observables

An observable is denoted by

[
O.
]

A family of observables is written as

[
{O_a}.
]

The expected value is

\operatorname{Tr}(\rho O).
]

The variance is

\operatorname{Tr}(\rho O^2)

\left[
\operatorname{Tr}(\rho O)
\right]^2.
]

A local observable at (x) is written as

[
O(x).
]

An observable must be distinguished from a raw data variable.

Observed data should be denoted by

[
Y.
]

The observation equation is

\mathcal{H}_{\theta}(\rho)
+
\varepsilon,
]

or

\mathcal{H}_{\theta}(\Phi)
+
\varepsilon.
]

C.11 Observation Operators

The observation map is denoted by

[
\mathcal{H}_{\theta}.
]

This symbol must not be confused with the Hilbert space (\mathcal{H}).

When both appear repeatedly, the observation operator may instead be denoted by

[
\mathcal{O}_{\theta}
]

or

[
\mathscr{H}_{\theta}.
]

The recommended form is

\mathscr{H}_{\theta}
\bigl(
\rho(t)
\bigr)
+
\varepsilon(t).
\tag{C.1}
]

For field data,

\mathscr{H}_{\theta}
\bigl(
\Phi(x,t)
\bigr)
+
\varepsilon(x,t).
\tag{C.2}
]

This convention reduces ambiguity between the observation map and the Hilbert space.

C.12 Measurement Operators

A generalized measurement operator is denoted by

[
M_a.
]

The associated POVM element is

M_a^{\dagger}M_a.
]

The completeness condition is

[
\sum_aE_a=I.
]

The outcome probability is

\operatorname{Tr}(E_a\rho).
]

Projection operators are denoted by

[
P_a.
]

They satisfy

[
P_a^2=P_a,
]

[
P_a^{\dagger}=P_a,
]

and

\delta_{ab}P_a.
]

C.13 Interaction Kernels

A pairwise interaction kernel is denoted by

[
K(x,y).
]

A time-dependent kernel is

[
K(x,y;t).
]

A memory kernel is denoted by

[
\mathcal{K}(t-s).
]

These must not be confused.

The pairwise spatial or relational interaction term may be written as

[
\int_{\mathcal{M}}
K(x,y)
\Phi(y,t),dy.
]

A non-Markovian memory equation is

\int_0^t
\mathcal{K}(t-s)\rho(s),ds.
]

Thus:

  • (K(x,y)) acts across configuration space;
  • (\mathcal{K}(t-s)) acts across time.

C.14 Coupling Constants

Generic coupling constants are denoted by

[
g_i.
]

A scale-dependent coupling is

[
g_i(\mu).
]

A scale- and time-dependent coupling is

[
g_i(\mu,t).
]

The beta function is

\mu
\frac{\partial g_i}{\partial\mu}.
]

In nonequilibrium models,

\beta_i(g,t).
]

The symbol (\lambda) is reserved for:

  • a scalar interaction strength;
  • an eigenvalue;
  • a regularization parameter.

Its meaning must be stated locally.

To avoid ambiguity, a regularization parameter may be written as

[
\lambda_{\mathrm{reg}}.
]

C.15 Network Notation

The adjacency matrix is denoted by

[A_{ij}(t)].
]

The network Laplacian is

D-A,
]

where (D) is the degree matrix.

Because (L_k) is already used for Lindblad operators, the graph Laplacian should not be denoted merely by (L).

The preferred notation is

[
\Delta_G
]

or

[
L_G.
]

A node field is

[
\Phi_i(t).
]

A coupled field-network system is

F_i(\Phi,A),
]

and

G_{ij}(A,\Phi).
]

C.16 Hypergraph and Higher-Order Notation

A pairwise coupling tensor is

[
J_{ij}.
]

A three-body interaction tensor is

[
\mathcal{I}_{ijk}.
]

An (n)-body interaction tensor is

[
\mathcal{I}_{i_1\cdots i_n}.
]

The associated interaction term is

[
\sum_{i_1,\ldots,i_n}
\mathcal{I}{i_1\cdots i_n}
\Phi{i_1}
\cdots
\Phi_{i_n}.
]

The order of interaction should always be stated.

C.17 Geometric Notation

The metric tensor is

[
g_{\mu\nu}.
]

Its inverse is

[
g^{\mu\nu}.
]

The determinant is written as

\det(g_{\mu\nu}).
]

To avoid ambiguity between the determinant and a coupling constant, the determinant may be written as

[
\det g.
]

The line element is

g_{\mu\nu}
dx^{\mu}dx^{\nu}.
]

The covariant derivative is

[
\nabla_{\mu}.
]

The connection coefficients are

[
\Gamma^{\rho}_{\mu\nu}.
]

The Riemann tensor is

[
R^{\rho}_{\ \sigma\mu\nu}.
]

The Ricci tensor is

[
R_{\mu\nu}.
]

The scalar curvature is

[
R.
]

A random metric is written as

[
g_{\mu\nu}(\omega).
]

C.18 Topological Notation

The (n)-th homotopy group is

[
\pi_n(\mathcal{N}).
]

A topological invariant is denoted by

[
T[\Phi].
]

A winding number may be written as

[
\nu[\Phi].
]

A topological charge is

[
Q_{\mathrm{top}}.
]

The notation must distinguish topological invariants from statistical test statistics.

A statistical test statistic should be written as

[
T_{\mathrm{stat}}.
]

C.19 Probability and Statistical Notation

A probability space is

[
(\Omega,\mathcal{F},P).
]

A random variable is

[
X:\Omega\rightarrow\mathbb{R}.
]

Expectation is

[
\mathbb{E}[X].
]

Conditional expectation is

[
\mathbb{E}[X\mid\mathcal{G}].
]

Variance is

[
\operatorname{Var}(X).
]

Covariance is

[
\operatorname{Cov}(X,Y).
]

A probability density is

[
p(y\mid\theta).
]

A posterior distribution is

[
p(\theta\mid D).
]

A likelihood function is

[
\mathcal{L}_{\mathrm{like}}(\theta;D),
]

or preferably

[
\mathscr{L}_{\mathrm{like}}(\theta;D)
]

when (\mathcal{L}) already denotes the Lindblad generator.

C.20 Parameter Notation

The parameter vector is

[
\boldsymbol{\theta}.
]

The true parameter is

[
\boldsymbol{\theta}_0.
]

An estimator is

[
\hat{\boldsymbol{\theta}}.
]

A sequence of estimators is

[
\hat{\boldsymbol{\theta}}_n.
]

The parameter space is

[
\Theta.
]

A gauge-equivalence class is

[
[\boldsymbol{\theta}].
]

When parameters are non-identifiable individually, the empirically meaningful object may be the equivalence class rather than a unique vector.

C.21 Norms

The Euclidean norm is

[
|\mathbf{x}|_2.
]

The (\ell^1) norm is

[
|\mathbf{x}|_1.
]

The operator norm is

[
|A|.
]

The trace norm is

\operatorname{Tr}|A|.
]

The Hilbert–Schmidt norm is

\sqrt{
\operatorname{Tr}(A^{\dagger}A)
}.
]

Because (|\cdot|_1) and (|\cdot|_2) can refer to vector norms or operator norms, the space should be stated whenever ambiguity is possible.

For example,

[
|\rho|_{\mathrm{tr}}
]

may be used for the trace norm, and

[
|\rho|_{\mathrm{HS}}
]

for the Hilbert–Schmidt norm.

C.22 Convergence Notation

Strong convergence is written as

[
x_n\rightarrow x.
]

Weak convergence in a Hilbert or Banach space is written as

[
x_n\rightharpoonup x.
]

Convergence in probability is

[
X_n\xrightarrow{P}X.
]

Convergence in distribution is

[
X_n\xrightarrow{d}X.
]

Almost-sure convergence is

[
X_n\xrightarrow{\mathrm{a.s.}}X.
]

Convergence in (L^p) is

[
X_n\xrightarrow{L^p}X.
]

Each convergence statement must specify the relevant topology or probability mode.

C.23 Derivative Notation

An ordinary time derivative is

[
\frac{d}{dt}.
]

A partial derivative is

[
\frac{\partial}{\partial t}.
]

A functional derivative is

[
\frac{\delta}{\delta\Phi(x)}.
]

A covariant derivative is

[
\nabla_{\mu}.
]

The Caputo fractional derivative is

[
{}^{C}D_t^{\alpha}.
]

The manuscript should not write only

[
D_t^{\alpha}
]

unless the chosen fractional derivative has already been defined.

Different fractional derivatives generate different initial-value problems.

C.24 Commutators and Anticommutators

The commutator is

AB-BA.
]

The anticommutator is

AB+BA.
]

The superoperator commutator is

\mathcal{R}\mathcal{U}_t

\mathcal{U}_t\mathcal{R}.
]

The norm of a commutator is written as

[
|[A,B]|.
]

A scalar absolute value,

[
|[A,B]|,
]

should not be used for operator-valued commutators.

C.25 Expectation-Value Notation

For a pure state,

\langle\Psi|O|\Psi\rangle.
]

For a density operator,

\operatorname{Tr}(\rho O).
]

For a classical probability distribution,

\mathbb{E}_{P}[f].
]

The brackets must not conceal which state or measure is being used.

When more than one state is under discussion, the subscript is mandatory.

C.26 Correlation Functions

The one-point function is

\langle\Phi(x)\rangle.
]

The two-point function is

\langle
\Phi(x)\Phi(y)
\rangle.
]

The connected two-point function is

G^{(2)}(x,y)

G^{(1)}(x)G^{(1)}(y).
]

The (n)-point function is

\left\langle
\prod_{j=1}^{n}
\Phi(x_j)
\right\rangle.
]

For noncommuting fields, operator ordering must be specified.

Possible conventions include:

[
\mathcal{T}
]

for time ordering and

[
\mathcal{N}
]

for normal ordering.

C.27 Action and Lagrangian Notation

The action is

[
S[\Phi].
]

The Lagrangian density is

[
\mathscr{L}
\bigl(
\Phi,
\partial_{\mu}\Phi,
x
\bigr).
]

The use of (\mathscr{L}) is preferred because (\mathcal{L}) is reserved for the open-system generator.

The action is

\int_{\mathcal{M}}
\mathscr{L}
\bigl(
\Phi,
\partial_{\mu}\Phi,
x
\bigr)
,dV_g.
]

The Hamiltonian density is

[
\mathscr{H}.
]

The total Hamiltonian is

\int_{\Sigma}
\mathscr{H},d\Sigma.
]

C.28 Effective Action

The generating functional is

[
Z[J].
]

The connected generating functional is

[
W[J].
]

A standard convention is

-i\kappa_S\ln Z[J].
]

The effective field is

\frac{\delta W[J]}
{\delta J(x)}.
]

The effective action is

W[J]

\int
J(x)\Phi_c(x),dx.
]

The subscript (c) indicates a classical or expectation field and must not be confused with a connected correlation function.

C.29 Entropy Notation

The von Neumann entropy is


\operatorname{Tr}(\rho\ln\rho).
]

The Shannon entropy is


\sum_i
p_i\ln p_i.
]

Relative entropy is

[
D(\rho|\sigma).
]

Mutual information is

[
I(A).
]

Thermodynamic entropy, information entropy, and social heterogeneity measures should not all be denoted simply by (S) without qualification.

C.30 Gauge Notation

The gauge group is

[
\mathcal{G}.
]

A gauge transformation is

[
g\in\mathcal{G}.
]

A unitary representation is

[
U(g).
]

The state transformation is

[
\rho
\longrightarrow
U(g)\rho U(g)^{\dagger}.
]

An observable transforms as

[
O
\longrightarrow
U(g)OU(g)^{\dagger}.
]

The gauge-equivalence class of a model (\mathfrak{S}) is

[
[\mathfrak{S}].
]

The empirically distinguishable model space is

\mathcal{M}_{\mathrm{model}}/\mathcal{G}.
]

C.31 Model Objects and Empirical Objects

SQFT must distinguish formal model objects from measured quantities.

Formal Objects

These include:

[
\rho,
\qquad
\Phi,
\qquad
H,
\qquad
L_k,
\qquad
\mathcal{L}.
]

Empirical Objects

These include:

[
Y,
\qquad
D,
\qquad
\widehat{G}^{(n)},
\qquad
\hat{\boldsymbol{\theta}}.
]

A formal state is not directly observed unless the observation operator is the identity.

In general,

[
Y
\neq
\rho,
]

and

[
Y
\neq
\Phi.
]

Instead,

\mathscr{H}_{\theta}(\rho)
+
\varepsilon.
]

C.32 Literal, Structural, and Metaphorical Usage

Every imported physical term should be assigned one of three labels.

Type I — Literal Physical Usage

The object is claimed to represent an actual physical quantum process.

This requires physical units, physical degrees of freedom, and experimentally justified quantum dynamics.

Type II — Mathematical Structural Usage

A mathematical structure from quantum theory or field theory is used to model a nonphysical system.

Examples include:

  • density operators as generalized state representations;
  • Lindblad equations as open-system dynamics;
  • gauge equivalence as representational redundancy;
  • renormalization as multiscale model reduction.

Type III — Informal Metaphorical Usage

A physical term is used rhetorically without a corresponding mathematical object.

Type III language should not appear in formal derivations unless explicitly marked as analogy.

The main SQFT framework belongs primarily to Type II.

C.33 Mandatory Model Specification

Every formal SQFT model should specify the following tuple:

\left(
\mathcal{M},
\mathcal{H},
\mathfrak{A},
\rho_0,
H,
{L_k},
\mathscr{H}_{\theta},
\Theta,
\mathcal{B},
\mathcal{V}
\right).
\tag{C.3}
]

Here:

  • (\mathcal{M}) is the configuration space or social manifold;
  • (\mathcal{H}) is the state space;
  • (\mathfrak{A}) is the observable algebra;
  • (\rho_0) is the initial state;
  • (H) is the reversible generator;
  • ({L_k}) are dissipative channels;
  • (\mathscr{H}_{\theta}) is the observation map;
  • (\Theta) is the parameter space;
  • (\mathcal{B}) denotes boundary and initial conditions;
  • (\mathcal{V}) is the validation protocol.

A model lacking one or more of these elements is incomplete.

C.34 Extended Model Specification

For nonlinear, geometric, or multiscale models, the specification should be extended to

\left(
\mathfrak{S},
g_{\mu\nu},
\mathcal{G},
\mathcal{R},
K,
\mathcal{N},
\Pi,
\mathcal{D}
\right).
\tag{C.4}
]

Here:

  • (g_{\mu\nu}) is the metric or geometric structure;
  • (\mathcal{G}) is the gauge group;
  • (\mathcal{R}) is the coarse-graining map;
  • (K) is the interaction kernel;
  • (\mathcal{N}) is a nonlinear term;
  • (\Pi) is a projection or reduction map;
  • (\mathcal{D}) is the dataset or data-generating structure.

C.35 Minimum Dynamical Specification

A dynamical SQFT model must provide:

State Equation

\mathcal{L}_{\theta}(\rho,t).
\tag{C.5}
]

Initial Condition

\rho_0.
\tag{C.6}
]

Admissible State Space

[
\rho(t)
\in
\mathcal{S}(\mathcal{H}),
\tag{C.7}
]

where

\left{
\rho:
\rho\geq 0,
;
\rho=\rho^{\dagger},
;
\operatorname{Tr}\rho=1
\right}.
]

Observation Equation

\mathscr{H}_{\theta}
\bigl(
\rho(t)
\bigr)
+
\varepsilon(t).
\tag{C.8}
]

Parameter Domain

[
\theta
\in
\Theta.
\tag{C.9}
]

Rejection Criterion

[
T(D,\mathfrak{S})


c_{\alpha}.
\tag{C.10}
]

Without these components, the model cannot be fully simulated, estimated, or falsified.

C.36 Minimum Field-Theoretic Specification

A field-theoretic model should specify:

Field Space

[
\Phi
\in
\mathcal{X}.
]

Action

[
S[\Phi].
]

Boundary Conditions

[
\mathcal{B}[\Phi]=0.
]

Equation of Motion


]

Symmetry Group

[
\mathcal{G}.
]

Observable Functionals

[
O_a[\Phi].
]

Probability or State Structure

[
P[\Phi]
]

or

[
\rho.
]

Scale Definition

[
\mu
]

or

[
\ell.
]

The phrase “field theory” should not be used formally unless a field space and a dynamical or variational law have been specified.

C.37 Minimum Open-System Specification

An open-system model should state:

  1. the system Hilbert space;
  2. the environment or effective environmental channels;
  3. the Hamiltonian (H);
  4. the Lindblad operators (L_k);
  5. the initial density operator;
  6. the time interval;
  7. the basis in which decoherence is interpreted;
  8. the observable map;
  9. the approximation under which Markovianity is assumed.

The complete equation is

\frac{1}{2}
{L_k^{\dagger}L_k,\rho}
\right).
\tag{C.11}
]

Every term must be empirically or structurally motivated.

C.38 Minimum Non-Markovian Specification

A non-Markovian model should specify:

  • the memory kernel;
  • the initial-history interval;
  • the state space;
  • positivity conditions;
  • normalization conditions;
  • whether the equation is time convolution or time local.

A memory-kernel equation is

\int_0^t
\mathcal{K}(t,s)
\rho(s),ds
+
\eta(t).
\tag{C.12}
]

If the kernel is translation invariant,

\mathcal{K}(t-s).
]

The inhomogeneous term (\eta(t)) should not be omitted when unresolved initial correlations are present.

C.39 Minimum Fractional-Dynamics Specification

A fractional model must identify:

  • the type of fractional derivative;
  • its order;
  • the initial conditions;
  • the function space;
  • the units of the fractional coefficient.

For the Caputo derivative,

F(\Phi(t)),
\qquad
0<\alpha<1.
\tag{C.13}
]

The notation (D_t^{\alpha}) alone is insufficient unless the derivative convention has already been declared.

C.40 Minimum Network Specification

A network SQFT model must state:

  • the node set;
  • the edge set;
  • whether the graph is directed;
  • whether it is weighted;
  • whether topology is static or dynamic;
  • the meaning of edge weights;
  • the field attached to each node or edge.

For a dynamic network,

F_i(\Phi,A),
]

and

G_{ij}(A,\Phi).
\tag{C.14}
]

The model should state whether (A_{ij}) is continuous, binary, probabilistic, or constrained.

C.41 Minimum Renormalization Specification

A renormalization analysis must define:

  • the microscopic variables;
  • the blocking or coarse-graining rule;
  • the rescaling operation;
  • the effective parameters;
  • the beta functions;
  • the fixed points;
  • the observable quantities expected to remain invariant.

The scale flow is

\beta_i(g).
\tag{C.15}
]

A fixed point satisfies

[
\beta_i(g^*)=0.
]

The classification of relevant and irrelevant directions must be based on the eigenvalues of the linearized flow.

C.42 Minimum Empirical Specification

An empirical SQFT study must specify:

Dataset

{Y_1,\ldots,Y_n}.
]

Sampling Design

The sampling population, time range, and missing-data mechanism must be described.

Observation Model

[
p(D\mid\rho,\theta).
]

Estimator

\mathcal{A}(D).
]

Validation Set

[
D_{\mathrm{test}}.
]

Comparator Models

[
M_1,\ldots,M_K.
]

Evaluation Rule

[
\mathcal{S}
\bigl(
M_k;
D_{\mathrm{test}}
\bigr).
]

Falsification Threshold

[
c_{\alpha}.
]

The model should not be evaluated only by in-sample fit.

C.43 Identifiability Declaration

Every empirical model should include an identifiability declaration.

The declaration should state whether the model is:

  • globally identifiable;
  • locally identifiable;
  • identifiable only modulo gauge transformations;
  • partially identifiable;
  • non-identifiable.

A local rank condition may be written as

\dim\Theta.
\tag{C.16}
]

If the rank is deficient, the non-identifiable directions should be reported.

C.44 Gauge Declaration

When gauge redundancy is present, the model must specify:

  • the gauge group (\mathcal{G});
  • the group action;
  • the invariant observables;
  • the chosen gauge-fixing condition;
  • the residual gauge freedom.

A gauge condition may be written as

[
F[A]=0.
]

The model must distinguish:

[
\text{parameter non-identifiability}
]

from

[
\text{gauge redundancy}.
]

Gauge redundancy is a known equivalence of representations. Non-identifiability may reflect inadequate data or model structure.

C.45 Causal Declaration

A causal SQFT model must identify:

  • causal variables;
  • intervention operators;
  • temporal ordering;
  • confounders;
  • mediation channels;
  • causal assumptions.

Correlation functions alone are insufficient.

An intervention may be written as

[
\operatorname{do}
\bigl(
\Phi(x,t)=\phi
\bigr),
]

or as an intervention operator

[
\mathcal{I}_{x,t}.
]

The causal response is then compared under two intervention regimes.

C.46 Approximation Declaration

Every approximation should be explicitly labeled.

Examples include:

  • weak-coupling approximation;
  • mean-field approximation;
  • Markov approximation;
  • adiabatic approximation;
  • low-rank approximation;
  • Gaussian approximation;
  • semiclassical approximation;
  • finite-dimensional truncation;
  • perturbative expansion.

A generic approximation statement should take the form

\mathcal{M}_{\mathrm{approx}}
+
O(\varepsilon^p).
\tag{C.17}
]

The approximation parameter (\varepsilon) and regime of validity must be stated.

C.47 Error and Uncertainty Declaration

A complete result should distinguish:

  • measurement error;
  • sampling uncertainty;
  • parameter uncertainty;
  • model misspecification;
  • numerical error;
  • approximation error;
  • structural uncertainty.

The total predictive error may be decomposed formally as

E_{\mathrm{data}}
+
E_{\mathrm{parameter}}
+
E_{\mathrm{model}}
+
E_{\mathrm{numerical}}
+
E_{\mathrm{approx}}.
\tag{C.18}
]

This decomposition need not be exactly additive in every application, but the sources should be conceptually separated.

C.48 Units and Dimensional Consistency

Every empirical equation should be dimensionally consistent.

In

H|\Psi\rangle,
]

the dimensions must satisfy

[H][t].
]

For a decay rate (\gamma),

[t]^{-1}.
]

For a diffusion coefficient (D),

[x]^2[t]^{-1}.
]

A purely dimensionless model should state that all quantities have been nondimensionalized.

C.49 Nondimensionalization

Suppose the original variables are

[
x,
\qquad
t,
\qquad
\Phi.
]

Introduce reference scales

[
x_0,
\qquad
t_0,
\qquad
\Phi_0.
]

Define dimensionless variables:

\frac{x}{x_0},
]

\frac{t}{t_0},
]

and

\frac{\Phi}{\Phi_0}.
]

The resulting dimensionless groups determine the actual dynamical regimes.

Nondimensionalization should precede claims about the relative magnitude of parameters.

C.50 Index Conventions

Repeated geometric indices are summed according to the Einstein convention:

\sum_{\mu}
A_{\mu}B^{\mu}.
]

Discrete actor or node indices are not automatically summed unless explicitly stated.

For example,

[
J_{ij}\Phi_j
]

may imply summation over (j) if the convention has been declared.

To avoid ambiguity, sums may be written explicitly:

[
\sum_j
J_{ij}\Phi_j.
]

Greek indices are reserved for geometric coordinates:

[
\mu,\nu,\rho,\sigma.
]

Latin indices are reserved for discrete components:

[
i,j,k,m,n.
]

C.51 Adjoint, Complex Conjugate, and Transpose

The adjoint of an operator is

[
A^{\dagger}.
]

The complex conjugate of a scalar is

[
z^*.
]

The transpose of a real matrix is

[
A^{\mathsf{T}}.
]

The conjugate transpose is

[
A^{\dagger}.
]

These operations must not be treated as interchangeable.

For a self-adjoint operator,

[
A=A^{\dagger}.
]

For a real symmetric matrix,

[
A=A^{\mathsf{T}}.
]

C.52 Identity Operators

The identity operator on (\mathcal{H}) is written as

[
I_{\mathcal{H}},
]

or simply

[
I
]

when the space is clear.

For a subsystem,

[
I_A,
\qquad
I_B.
]

The identity superoperator is

[
\mathcal{I}.
]

This distinction is important in expressions such as

[
\mathcal{E}\otimes\mathcal{I}.
]

C.53 Zero Objects

The scalar zero is

[
0.
]

The zero vector is

[
\mathbf{0}.
]

The zero operator is

[
0_{\mathcal{H}}.
]

The zero superoperator is

[
\mathcal{0}.
]

The context usually makes the distinction clear, but formal proofs should avoid ambiguous zero objects.

C.54 Equation Numbering

Equations in chapters are numbered by chapter:

[
(22.1),
\qquad
(22.2).
]

Equations in appendices are numbered by appendix:

[
(A.1),
\qquad
(B.1),
\qquad
(C.1).
]

When an equation is revised, later equation numbers should be updated consistently.

The manuscript should not retain two unrelated equations with the same number.

C.55 Definitions, Propositions, and Conjectures

Formal statements should follow a consistent hierarchy.

Definition

A definition introduces a term and does not require proof.

Assumption

An assumption specifies a condition under which a model or theorem is valid.

Lemma

A lemma is an auxiliary proved result.

Proposition

A proposition is a proved result of limited or local scope.

Theorem

A theorem is a central proved result.

Corollary

A corollary follows directly from a preceding result.

Conjecture

A conjecture is a mathematically precise but unproved claim.

Open Problem

An open problem states a question for future research.

A conjecture should not be presented as an established theorem.

C.56 Recommended Model Header

Each applied SQFT model should begin with a standardized header.

Model Name

Provide a unique model title.

Social System

Specify the empirical system being modeled.

State Space

\cdots
]

State Variable

\cdots
]

or

\cdots
]

Generator

\cdots
]

Observables

[
O_1,\ldots,O_m.
]

Observation Map

\mathscr{H}_{\theta}(\rho)
+
\varepsilon.
]

Parameters

[
\theta\in\Theta.
]

Initial and Boundary Conditions

State all required conditions.

Estimation Method

Describe the algorithm or likelihood.

Comparator Models

List classical alternatives.

Rejection Criterion

Specify what empirical result would count against the model.

C.57 Recommended Equation Presentation

For reliable rendering, formulas should use simple single-level display equations.

Preferred form:

\frac{1}{2}
{L_k^{\dagger}L_k,\rho}
\right).
]

Avoid unnecessary nested environments such as:

  • multiple aligned blocks;
  • nested boxed structures;
  • text-heavy expressions inside equations;
  • manual line-break commands;
  • nonstandard macros.

Long equations should be divided into two separately numbered equations when necessary.

C.58 Unicode Backup Convention

When LaTeX rendering is unavailable, the following Unicode forms may be used.

Closed Evolution

-\frac{i}{\kappa_S}[H,\rho]
]

Unicode:

dρ/dt = −(i/κₛ)[H,ρ]

Lindblad Evolution

Unicode:

dρ/dt = −(i/κₛ)[H,ρ] + Σₖ(LₖρLₖ† − ½{Lₖ†Lₖ,ρ})

Heisenberg Evolution

Unicode:

dAᴴ/dt = (i/κₛ)[H,Aᴴ] + (∂A/∂t)ᴴ

Trace Condition

Unicode:

Tr(ρ) = 1

Positivity

Unicode:

ρ ≥ 0

The Unicode form is a backup representation and should not replace properly typeset equations in the final manuscript.

C.59 Prohibited Ambiguities

The following notation practices should be avoided.

Ambiguity 1

Using

[
\mathcal{H}
]

for both Hilbert space and observation operator.

Ambiguity 2

Using

[
\mathcal{L}
]

for both Lindblad generator and statistical loss.

Ambiguity 3

Using

[
L
]

for Lindblad operator, graph Laplacian, and Lagrangian.

Ambiguity 4

Using

[
S
]

for action, entropy, social system, and statistical score in the same section.

Ambiguity 5

Using

[
M
]

for manifold, model, and measurement operator.

Ambiguity 6

Using scalar absolute-value bars for operator norms.

Incorrect:

[
|A|.
]

Preferred:

[
|A|.
]

Ambiguity 7

Writing a field equation without specifying the field domain.

Ambiguity 8

Writing a density matrix without positivity and normalization conditions.

C.60 Recommended Symbol Replacements

To maintain consistency, the following replacements are recommended throughout the manuscript.

Ambiguous SymbolPreferred ReplacementMeaning
(\mathcal{H}_{\theta})(\mathscr{H}_{\theta})Observation map
(\mathcal{L}) for loss(\mathscr{L})Statistical loss
(L) for graph Laplacian(L_G) or (\Delta_G)Graph Laplacian
(S) for entropy(S_{\mathrm{vN}}) or (S_{\mathrm{Sh}})Entropy
(M) for manifold(\mathcal{M})Manifold
(D_t^{\alpha})({}^{C}D_t^{\alpha})Caputo derivative
([A,B])(|[A,B]|)Operator norm
(\hbar)(\kappa_S)SQFT action scale

C.61 Canonical SQFT Equation Set

The following equations constitute the canonical notation set of the manuscript.

Pure-State Evolution

H|\Psi(t)\rangle.
\tag{C.19}
]

Closed Density Evolution

-\frac{i}{\kappa_S}
[H,\rho].
\tag{C.20}
]

Open Density Evolution

\frac{1}{2}
{L_k^{\dagger}L_k,\rho}
\right).
\tag{C.21}
]

Heisenberg Evolution

\frac{i}{\kappa_S}
[H,A_H]
+
\left(
\frac{\partial A}{\partial t}
\right)_H.
\tag{C.22}
]

Observation Equation

\mathscr{H}_{\theta}
\bigl(
\rho(t)
\bigr)
+
\varepsilon(t).
\tag{C.23}
]

Non-Markovian Evolution

\int_0^t
\mathcal{K}(t-s)\rho(s),ds.
\tag{C.24}
]

Fractional Evolution

\mathcal{L}\rho(t),
\qquad
0<\alpha<1.
\tag{C.25}
]

Coupled Field-Network Evolution

F_i(\Phi,A),
\tag{C.26}
]

G_{ij}(A,\Phi).
\tag{C.27}
]

Renormalization Flow

\beta_i(g).
\tag{C.28}
]

Coarse-Graining Consistency

\mathcal{U}^{\mathrm{eff}}_t
\mathcal{R}.
\tag{C.29}
]

C.62 Canonical State Conditions

Every density-operator model must satisfy

[
\rho(t)\geq 0,
\tag{C.30}
]

[
\rho(t)=\rho(t)^{\dagger},
\tag{C.31}
]

and

[
\operatorname{Tr}\rho(t)=1.
\tag{C.32}
]

Every pure-state model must satisfy

[
\langle\Psi(t)|\Psi(t)\rangle=1.
\tag{C.33}
]

Every classical probability model must satisfy

[
p_i(t)\geq 0,
\tag{C.34}
]

and

[
\sum_i p_i(t)=1.
\tag{C.35}
]

These conditions are not optional stylistic choices. They define the admissible state space.

C.63 Canonical Empirical Conditions

A scientifically interpretable SQFT model should satisfy:

[
\text{Explicit state definition},
]

[
\text{Explicit observation map},
]

[
\text{Known parameter domain},
]

[
\text{Identifiability statement},
]

[
\text{Comparator model},
]

[
\text{Out-of-sample validation},
]

and

[
\text{Falsification criterion}.
]

These conditions may be summarized as

\mathfrak{S}{\mathrm{formal}}
+
\mathfrak{S}{\mathrm{empirical}}.
\tag{C.36}
]

A formal model without empirical specification remains a mathematical construction rather than a tested social theory.

C.64 Final Notational Principle

The notation of SQFT should satisfy three requirements:

[
\text{consistency},
]

[
\text{distinguishability},
]

and

[
\text{interpretability}.
]

Consistency means that one symbol preserves one principal meaning throughout the manuscript.

Distinguishability means that mathematically different objects are not represented by visually identical notation.

Interpretability means that every formal symbol can be connected to either:

  • a mathematical definition;
  • an empirical quantity;
  • a declared structural analogy.

The governing principle is therefore:

[
\text{No symbol should carry more meaning}
]

[
\text{than the model has explicitly defined.}
\tag{C.37}
]


Appendix D

Axiomatic Foundation of Social Quantum Field Theory

D.1 Motivation

Every mature mathematical theory begins with a small collection of explicit axioms.

Examples include:

  • Euclidean geometry
  • Kolmogorov probability
  • Hilbert-space quantum mechanics
  • General relativity
  • Quantum field theory

Similarly, Social Quantum Field Theory (SQFT) should not merely consist of isolated equations.

Instead, every theorem in the main text should ultimately be derivable from a finite collection of structural assumptions.

The purpose of this appendix is therefore to formulate SQFT as an axiomatic theory.

D.2 Primitive Objects

The theory assumes the existence of the following primitive mathematical objects.

(1) Configuration Space

A social configuration manifold

M.\mathcal{M}.

(2) State Space

A Hilbert space

H.\mathcal H.

or an equivalent admissible state space.

(3) State

A state is represented by

ρ∈S(H),\rho \in \mathcal S(\mathcal H),

where

S(H)={ρ≥0,ρ†=ρ,Tr⁡ρ=1}.\mathcal S(\mathcal H) = \{ \rho\ge0, \rho^\dagger=\rho, \operatorname{Tr}\rho=1 \}.

(4) Observable Algebra

A set

A\mathfrak A

of admissible observables.

(5) Generator

A dynamical generator

L.\mathcal L.

(6) Observation Map

An empirical observation operator

Hθ.\mathscr H_\theta.

(7) Data Space

Observed data

D.D.

These objects are regarded as primitive.

All later constructions depend upon them.

D.3 Axiom 1 — State Existence

At every admissible time,

there exists a state

ρ(t)∈S(H).\rho(t) \in \mathcal S(\mathcal H).

The state completely specifies every probabilistic prediction of the theory.

D.4 Axiom 2 — State Admissibility

Every admissible state satisfies

ρ≥0,\rho\ge0, ρ†=ρ,\rho^\dagger=\rho,

and

Tr⁡ρ=1.\operatorname{Tr}\rho=1.

These conditions remain invariant under the admissible dynamics.

D.5 Axiom 3 — Dynamical Evolution

The evolution of the state is governed by

dρdt=L(ρ,t).\frac{d\rho}{dt} = \mathcal L(\rho,t).

The generator may be

  • Hamiltonian
  • Lindblad
  • Non-Markovian
  • Fractional
  • Hybrid

provided the admissibility conditions remain satisfied.

D.6 Axiom 4 — Observation Principle

No state is directly observable.

Instead,

every observation satisfies

Y=Hθ(ρ)+ε.Y = \mathscr H_\theta(\rho) + \varepsilon.

Thus,

empirical data are always images of the hidden state.

D.7 Axiom 5 — Representation Invariance

Observable predictions are invariant under admissible gauge transformations.

If

ρ′=UρU†,\rho' = U\rho U^\dagger,

and

O′=UOU†,O' = UOU^\dagger,

then

Tr⁡(ρO)=Tr⁡(ρ′O′).\operatorname{Tr}(\rho O) = \operatorname{Tr}(\rho' O').

Hence,

only gauge-invariant quantities possess empirical meaning.

D.8 Axiom 6 — Coarse-Graining

Every macroscopic description is obtained from

R:S(H)→S(Heff).\mathcal R: \mathcal S(\mathcal H) \rightarrow \mathcal S(\mathcal H_{\rm eff}).

Different microscopic states may share the same effective description.

D.9 Axiom 7 — Scale Dependence

Model parameters evolve with scale,

dgidℓ=βi(g).\frac{dg_i}{d\ell} = \beta_i(g).

Scale therefore constitutes an intrinsic component of the theory.

D.10 Axiom 8 — Empirical Testability

Every SQFT model must specify

  • observables,
  • estimation procedure,
  • validation protocol,
  • falsification criterion.

Otherwise,

it is regarded only as a mathematical framework,

not a scientific theory.

D.11 Axiom 9 — Structural Correspondence

Every mathematical object appearing in the theory must belong to one of three categories.

Type I

Literal physical object.

Type II

Mathematical structural analogue.

Type III

Metaphorical description.

The type must always be declared explicitly.

This prevents mathematical formalism from being mistaken for empirical evidence.

D.12 Axiom 10 — Internal Consistency

No theorem of SQFT may violate

  • positivity,
  • normalization,
  • gauge consistency,
  • dimensional consistency,
  • or logical consistency.

Whenever two formulations produce contradictory empirical predictions,

at least one formulation is regarded as inadmissible.

D.13 Derived Principles

From the ten axioms above, the following principles follow naturally.

Conservation Principle

Admissible evolution preserves the state space.

Representation Principle

Different mathematical representations may describe the same empirical system.

Emergence Principle

Macroscopic structure emerges through coarse-graining.

Renormalization Principle

Effective theories depend upon observational scale.

Information Principle

Observed data never determine the hidden state uniquely without additional assumptions.

Falsifiability Principle

Every admissible SQFT model must permit empirical rejection.

D.14 Meta-Theorem

Taken together, the preceding axioms define SQFT as a mathematical theory of state evolution under observation, coarse-graining, and multiscale transformation.

They do not imply that social systems are literally quantum mechanical.

Instead, they specify the conditions under which quantum-theoretic mathematical structures may be used as a rigorous modeling language for complex social dynamics.


Appendix E

Mathematical Glossary and Index of Symbols

E.1 Purpose of This Appendix

Social Quantum Field Theory combines concepts from several mathematical disciplines.

These include:

  • functional analysis;
  • operator theory;
  • probability;
  • open-system dynamics;
  • differential geometry;
  • topology;
  • network theory;
  • inverse problems;
  • renormalization;
  • statistical inference.

Because many terms possess different meanings in different fields, a unified glossary is necessary.

This appendix provides:

  1. definitions of the principal mathematical terms;
  2. standardized English terminology;
  3. consistent symbol usage;
  4. distinctions between closely related concepts;
  5. an alphabetical index of the most important notation.

The definitions below follow the conventions established in Appendices A–D.

Part I

Core Mathematical Glossary

E.2 Action

An action is a functional of a field:

[
S[\Phi].
]

It is commonly defined by

\int_{\mathcal{M}}
\mathscr{L}
\bigl(
\Phi,
\partial_{\mu}\Phi,
x
\bigr)
,dV_g.
]

A stationary field satisfies


]

In SQFT, the action defines a compact representation of the dynamical structure.

It should not be interpreted as literal physical action unless units and physical variables are explicitly specified.

E.3 Adjoint Operator

For an operator (A), the adjoint (A^{\dagger}) satisfies

\langle A^{\dagger}\Phi,\Psi\rangle.
]

An operator is self-adjoint when

[
A=A^{\dagger}.
]

Self-adjoint operators are used to represent observables and Hamiltonians.

E.4 Algebra of Observables

The observable algebra is denoted by

[
\mathfrak{A}.
]

It is a collection of observables closed under appropriate algebraic operations such as:

[
A+B,
]

[
AB,
]

and

[
A^{\dagger}.
]

If all observables commute, the algebra is commutative.

A noncommutative observable algebra permits order-dependent transformations and contextual structure.

E.5 Anticommutator

The anticommutator of (A) and (B) is

AB+BA.
]

It appears in the dissipative part of the Lindblad equation:

[
-\frac{1}{2}
{L_k^{\dagger}L_k,\rho}.
]

The anticommutator should not be confused with the commutator.

E.6 Asymptotic Stability

An equilibrium (x^*) is asymptotically stable if:

  1. nearby trajectories remain near (x^*);
  2. nearby trajectories converge to (x^*).

Formally,

[
x(t)\rightarrow x^*
]

as

[
t\rightarrow\infty.
]

E.7 Banach Space

A Banach space is a complete normed vector space.

Completeness means that every Cauchy sequence converges to an element of the space.

Examples include:

[
L^p(\Omega),
]

[
C([a,b]),
]

and trace-class operator spaces.

E.8 Beta Function

A beta function describes the change of a coupling parameter under scale transformation:

\frac{dg_i}{d\ell}.
]

Equivalently,

\mu
\frac{\partial g_i}{\partial\mu}.
]

A fixed point satisfies

[
\beta_i(g^*)=0.
]

E.9 Boundary Condition

A boundary condition specifies field behavior on the boundary of a domain.

Examples include:

Dirichlet

f.
]

Neumann

g.
]

Robin

c.
]

Boundary conditions are part of the model definition rather than optional numerical details.

E.10 Causal Kernel

A causal kernel describes directed influence between configurations or events.

It may be written as

[
K(x,y;t,s).
]

A causal kernel must distinguish temporal precedence and should not be inferred from correlation alone.

E.11 Classical Limit

A classical limit is a regime in which the operator model becomes equivalent to a classical probability or deterministic model.

Typical mechanisms include:

  • decoherence;
  • commuting observables;
  • diagonal density operators;
  • coarse-graining;
  • small fluctuation limits;
  • stationary-phase approximation.

A symbolic correspondence is

[
\mathcal{L}{\mathrm{SQFT}}
\longrightarrow
\mathcal{L}{\mathrm{classical}}.
]

E.12 Coarse-Graining

Coarse-graining is a map

[
\mathcal{R}:
\mathcal{S}(\mathcal{H})
\rightarrow
\mathcal{S}(\mathcal{H}_{\mathrm{eff}}).
]

It removes or aggregates fine-scale information.

Different microscopic states may satisfy

\mathcal{R}(\rho_2).
]

Coarse-graining is therefore generally non-injective.

E.13 Commutator

The commutator of (A) and (B) is

AB-BA.
]

If

[
[A,B]=0,
]

the operators commute.

Nonzero commutators represent order dependence or incompatibility within the operator formalism.

E.14 Complete Positivity

A map (\mathcal{E}) is completely positive when

[
\mathcal{E}\otimes I_n
]

is positive for every finite dimension (n).

Complete positivity ensures that the map remains physically and probabilistically admissible when extended to a larger composite system.

E.15 Configuration Space

The configuration space is the set of all admissible system configurations.

In SQFT, it may be written as

[
\mathcal{M}.
]

A point

[
x\in\mathcal{M}
]

may represent a geographic, institutional, informational, or relational position.

E.16 Connected Correlation Function

The connected two-point function is

\langle\Phi(x)\Phi(y)\rangle

\langle\Phi(x)\rangle
\langle\Phi(y)\rangle.
]

It removes the contribution generated by independent means.

Connected correlations measure dependence but do not by themselves establish causality.

E.17 Contextuality

Contextuality refers to dependence of outcomes on the measurement or decision context.

Formally, an outcome distribution may satisfy

[
P(A=a\mid C_1)
\neq
P(A=a\mid C_2).
]

In an operator model, contextuality may be associated with noncommuting observables.

However, order effects, memory, learning, and information acquisition may also generate context dependence in classical models.

E.18 Contraction Map

A map (T) is a contraction when

[
d(Tx,Ty)
\leq
q,d(x,y),
\qquad
0\leq q<1.
]

By the Banach fixed-point theorem, a contraction on a complete metric space has a unique fixed point.

E.19 Correlation Function

An (n)-point correlation function is

\left\langle
\Phi(x_1)\cdots\Phi(x_n)
\right\rangle.
]

For noncommuting fields, operator ordering must be specified.

E.20 Covariant Derivative

A covariant derivative permits differentiation on curved spaces or vector bundles.

It is written as

[
\nabla_{\mu}.
]

For a vector field (V^{\nu}),

\partial_{\mu}V^{\nu}
+
\Gamma^{\nu}_{\mu\lambda}V^{\lambda}.
]

E.21 CPTP Map

A CPTP map is a completely positive trace-preserving map.

It is written as

[
\mathcal{E}.
]

In finite dimensions, it has a Kraus representation:

\sum_k
K_k\rho K_k^{\dagger},
]

with

I.
]

E.22 Curvature

Curvature describes the failure of parallel transport to be path independent.

The Riemann curvature tensor is

[
R^{\rho}_{\ \sigma\mu\nu}.
]

In SQFT, curvature may represent path dependence or geometric obstruction only when a specific metric and connection have been defined.

E.23 Data-Processing Inequality

For a CPTP map (\mathcal{E}),

[
D(\rho|\sigma)
\geq
D
\left(
\mathcal{E}(\rho)
|
\mathcal{E}(\sigma)
\right).
]

Processing, aggregation, or observation cannot increase distinguishability.

E.24 Decoherence

Decoherence is the suppression of off-diagonal density-matrix elements in a selected basis.

A typical form is

[
\rho_{mn}(t)
\longrightarrow
0,
\qquad
m\neq n.
]

The basis in which this occurs is called the pointer basis.

In SQFT, decoherence is primarily a structural analogy for the loss of cross-alternative coherence.

E.25 Density Operator

A density operator (\rho) satisfies

[
\rho\geq 0,
]

[
\rho=\rho^{\dagger},
]

and

[
\operatorname{Tr}\rho=1.
]

A pure state satisfies

[
\rho^2=\rho.
]

A mixed state satisfies, in general,

[
\rho^2\neq\rho.
]

E.26 Differential Privacy

A mechanism (M) satisfies ((\varepsilon,\delta))-differential privacy when

[
P(M(D)\in S)
\leq
e^{\varepsilon}
P(M(D')\in S)
+
\delta
]

for neighboring datasets (D) and (D').

E.27 Dissipator

The dissipative part of a Lindblad generator is

\frac{1}{2}
{L_k^{\dagger}L_k,\rho}
\right).
]

It represents irreversible environmental effects.

E.28 Effective Theory

An effective theory describes a system only at a chosen scale or resolution.

Its parameters need not equal the parameters of the microscopic model.

Effective theories are connected through coarse-graining and renormalization.

E.29 Eigenvalue and Eigenvector

An eigenvector (|v\rangle) of an operator (A) satisfies

\lambda|v\rangle.
]

The scalar (\lambda) is the corresponding eigenvalue.

E.30 Entanglement-Like Correlation

A composite operator state is nonfactorized when

[
\rho_{AB}
\neq
\rho_A\otimes\rho_B.
]

It is nonseparable when no decomposition of the form

\sum_i
p_i
\rho_A^{(i)}
\otimes
\rho_B^{(i)}
]

exists.

In social applications, “entanglement” should be used only when the tensor-product structure and nonseparability criterion are explicitly defined.

E.31 Entropy

The von Neumann entropy is


\operatorname{Tr}(\rho\ln\rho).
]

The Shannon entropy is


\sum_i
p_i\ln p_i.
]

These quantities should not be treated as identical without a specific correspondence.

E.32 Equilibrium

An equilibrium state is invariant under the dynamics.

For an ordinary differential equation,

[
F(x^*)=0.
]

For a generator,

[
\mathcal{L}(\rho_*)=0.
]

E.33 Euler–Lagrange Equation

For an action

\int
\mathscr{L}
\bigl(
\Phi,
\partial_{\mu}\Phi
\bigr)
,d^nx,
]

the Euler–Lagrange equation is


]

E.34 Expectation Value

For a density operator,

\operatorname{Tr}(\rho O).
]

For a pure state,

\langle\Psi|O|\Psi\rangle.
]

For a classical probability distribution,

\mathbb{E}_P[X].
]

E.35 Field

A field is a map from a base space to a target space:

[
\Phi:
\mathcal{M}
\rightarrow
\mathcal{N}.
]

A time-dependent field is

[
\Phi(x,t).
]

A field may be scalar, vector, tensor, or operator valued.

E.36 Fisher Information

The Fisher information matrix is

\mathbb{E}_{\theta}
\left[
\frac{\partial\ln p(Y|\theta)}
{\partial\theta_i}
\frac{\partial\ln p(Y|\theta)}
{\partial\theta_j}
\right].
]

A singular Fisher matrix indicates locally unidentifiable parameter directions.

E.37 Fixed Point

A fixed point of a map (T) satisfies

[
T(x^)=x^.
]

A renormalization-group fixed point satisfies

[
\beta_i(g^*)=0.
]

E.38 Fractional Derivative

The Caputo derivative of order (0<\alpha<1) is

\frac{1}{
\Gamma(1-\alpha)
}
\int_0^t
\frac{f'(s)}
{(t-s)^{\alpha}}
,ds.
]

Fractional derivatives are used to represent long memory.

E.39 Functional Derivative

The functional derivative satisfies

F[\Phi]

\varepsilon
\int
\frac{\delta F}
{\delta\Phi(x)}
\eta(x),dx
+
o(\varepsilon).
]

E.40 Gauge Equivalence

Two models are gauge equivalent when they differ only by a transformation that leaves all empirical predictions invariant.

If

U\rho U^{\dagger},
]

and

UOU^{\dagger},
]

then

\operatorname{Tr}(\rho'O').
]

The empirical model space is therefore a quotient:

\mathcal{M}_{\mathrm{model}}/\mathcal{G}.
]

E.41 Gauge Fixing

Gauge fixing selects one representative from each gauge-equivalence class.

A gauge condition may be written as

[
F[A]=0.
]

Gauge fixing does not change observable predictions.

E.42 Generating Functional

A generating functional is written as

[
Z[J].
]

Correlation functions are obtained by functional differentiation with respect to the source (J).

A formal expression is

\int
\mathcal{D}\Phi
\exp
\left[
\frac{i}{\kappa_S}
\left(
S[\Phi]
+
\int J\Phi
\right)
\right].
]

E.43 Generator

A generator determines the infinitesimal evolution of a dynamical system.

For a semigroup,

e^{tA}.
]

The operator (A) is the generator.

For open SQFT dynamics, the generator is denoted by

[
\mathcal{L}.
]

E.44 Green Function

A Green function satisfies

\delta(x-y).
]

It represents the response at (x) to a source at (y).

E.45 Hamiltonian

The Hamiltonian is denoted by

[
H.
]

For a closed system,

H|\Psi\rangle.
]

A valid Hamiltonian is generally assumed to be self-adjoint.

E.46 Hermiticity

An operator (A) is Hermitian or self-adjoint when

[
A=A^{\dagger},
]

subject to the appropriate domain condition.

Hermiticity ensures real expectation values for observables.

E.47 Hilbert Space

A Hilbert space is a complete inner-product space.

The SQFT state space is denoted by

[
\mathcal{H}.
]

E.48 Hilbert–Schmidt Norm

The Hilbert–Schmidt norm is

\sqrt{
\operatorname{Tr}(A^{\dagger}A)
}.
]

E.49 Hypergraph

A hypergraph permits interactions among more than two nodes.

A three-body interaction may be represented by

[
\mathcal{I}_{ijk}
\Phi_i\Phi_j\Phi_k.
]

E.50 Identifiability

A model is identifiable when distinct parameter values generate distinct observable distributions.

Global identifiability requires

P_{\theta_2}(Y)
\quad\Longrightarrow\quad
\theta_1=\theta_2.
]

Gauge-invariant identifiability requires equality only up to a known equivalence class.

E.51 Information Geometry

Information geometry equips a statistical model with a metric derived from the Fisher information:

I_{ij}(\theta).
]

It describes local distinguishability in parameter space.

E.52 Interaction Kernel

An interaction kernel is written as

[
K(x,y).
]

It specifies the strength and structure of influence from (y) to (x).

E.53 Kraus Operator

A Kraus representation of a CPTP map is

\sum_k
K_k\rho K_k^{\dagger}.
]

The completeness condition is

I.
]

E.54 Lie Algebra

A Lie algebra is a vector space with a bracket operation satisfying bilinearity, antisymmetry, and the Jacobi identity.

The bracket is written as

[
[X,Y].
]

E.55 Lie Group

A Lie group is both a smooth manifold and a group.

Continuous symmetries are represented by Lie groups.

E.56 Lindblad Equation

The canonical Markovian open-system equation is

\frac{1}{2}
{L_k^{\dagger}L_k,\rho}
\right).
]

E.57 Lindblad Operator

A Lindblad operator (L_k) represents an effective dissipative, transition, or decoherence channel.

Its empirical meaning must be specified in each model.

E.58 Lipschitz Continuity

A map (F) is Lipschitz continuous if

[
|F(x)-F(y)|
\leq
L|x-y|.
]

Local Lipschitz continuity is a standard uniqueness condition.

E.59 Local Excitation

A local excitation is a localized deviation of a field from its background configuration.

Formally,

\Phi_0(x)
+
\delta\Phi(x),
]

where (\delta\Phi) is concentrated near a particular region or node.

In SQFT, an actor may be represented as a local excitation of a broader social field.

E.60 Markov Property

A process is Markovian when its future conditional distribution depends only on the present state:

P(X_{t+\Delta t}\mid X_t).
]

E.61 Measurement Operator

A measurement operator is denoted by

[
M_a.
]

The outcome probability is

\operatorname{Tr}
\left(
M_a\rho M_a^{\dagger}
\right).
]

E.62 Memory Kernel

A memory kernel is denoted by

[
\mathcal{K}(t-s).
]

It appears in non-Markovian equations:

\int_0^t
\mathcal{K}(t-s)\rho(s),ds.
]

E.63 Metric

A metric is a distance function satisfying positivity, symmetry, and the triangle inequality.

On a manifold, a metric tensor is written as

[
g_{\mu\nu}.
]

E.64 Mild Solution

For

Au+F(u),
]

a mild solution satisfies

T(t)u_0
+
\int_0^t
T(t-s)F(u(s)),ds.
]

E.65 Mixed State

A mixed state has the form

\sum_i
p_i
|\Psi_i\rangle
\langle\Psi_i|.
]

It represents probabilistic uncertainty or reduced information.

E.66 Model Misspecification

A model is misspecified when the true data-generating process does not belong to the assumed model family.

A robustness bound may take the form

[
|\rho_0(t)-\rho_{\theta}(t)|
\leq
C(t)
|\mathcal{L}0-\mathcal{L}{\theta}|.
]

E.67 Mutual Information

The mutual information between subsystems (A) and (B) is

S(\rho_{AB}).
]

It measures total dependence, including classical and nonclassical components.

E.68 Noncommutativity

Noncommutativity means

[
[A,B]\neq 0.
]

The order of operations then affects the result.

Noncommutativity may represent contextual or sequential dependence.

E.69 Non-Markovian Dynamics

Non-Markovian dynamics depend on historical states.

A general form is

\int_0^t
\mathcal{K}(t-s)\rho(s),ds
+
\eta(t).
]

E.70 Norm

A norm measures vector or operator magnitude.

It satisfies:

[
|x|\geq 0,
]

|\alpha||x|,
]

and

[
|x+y|
\leq
|x|+|y|.
]

E.71 Observable

An observable is a measurable model quantity represented by a self-adjoint operator or a classical random variable.

Its expectation is

\operatorname{Tr}(\rho O).
]

E.72 Observation Map

The observation map connects the latent state to empirical data:

\mathscr{H}_{\theta}
\bigl(
\rho(t)
\bigr)
+
\varepsilon(t).
]

E.73 Open System

An open system exchanges information, influence, or resources with an external environment.

Its reduced dynamics may be irreversible and nonunitary.

E.74 Operator Norm

The operator norm is

\sup_{|\Psi|=1}
|A\Psi|.
]

E.75 Partial Trace

For a composite state (\rho_{AB}), the reduced state of (A) is

\operatorname{Tr}B(\rho{AB}).
]

E.76 Path Integral

A formal path integral is

\int
\mathcal{D}\Phi
\exp
\left(
\frac{i}{\kappa_S}
S[\Phi]
\right).
]

Its rigorous existence must be established or approximated through discretization or probabilistic construction.

E.77 Pointer Basis

The pointer basis is the basis in which environmental interaction suppresses off-diagonal density-matrix elements.

E.78 Positivity

An operator (\rho) is positive when

[
\langle\Psi|\rho|\Psi\rangle
\geq
0
]

for every (|\Psi\rangle).

E.79 Probability Space

A probability space is

[
(\Omega,\mathcal{F},P).
]

It consists of a sample space, sigma-algebra, and probability measure.

E.80 Projection Operator

A projection operator satisfies

[
P^2=P,
]

and

[
P^{\dagger}=P.
]

E.81 Pure State

A pure state is represented by

|\Psi\rangle\langle\Psi|.
]

It satisfies

[
\rho^2=\rho.
]

E.82 Quantum-Zeno-Type Effect

Repeated projection onto a subspace may suppress transitions out of that subspace:

[
\lim_{N\rightarrow\infty}
\left[
P
e^{-iHt/(N\kappa_S)}
P
\right]^N.
]

In SQFT, this is a structural analogy for repeated monitoring or intervention that constrains system transitions.

E.83 Reduced State

A reduced state describes a subsystem after tracing out the remaining degrees of freedom:

\operatorname{Tr}B(\rho{AB}).
]

E.84 Regularization

Regularization stabilizes an inverse problem by penalizing undesirable solutions.

A general estimator is

\operatorname*{arg,min}{\theta}
\left[
\mathscr{L}(\theta)
+
\lambda{\mathrm{reg}}R(\theta)
\right].
]

E.85 Relative Entropy

The relative entropy is

\operatorname{Tr}
\left[
\rho(\ln\rho-\ln\sigma)
\right].
]

It measures distinguishability between states.

E.86 Renormalization

Renormalization describes the transformation of model parameters under changes of scale.

It is represented by

[
g_i
\longrightarrow
g_i(\ell).
]

E.87 Renormalization Group

A renormalization group is a family of scale transformations.

Despite the name, the transformations may form a semigroup rather than a mathematical group because information lost under coarse-graining may not be recoverable.

E.88 Self-Adjoint Operator

An operator is self-adjoint when

[
A=A^{\dagger}
]

and the domains of (A) and (A^{\dagger}) coincide.

This is stronger than formal symmetry.

E.89 Semigroup

A semigroup satisfies

T(t)T(s),
]

and

[
T(0)=I
]

for (t,s\geq 0).

Unlike a group, inverse evolution need not exist.

E.90 Separable State

A bipartite state is separable when

\sum_i
p_i
\rho_A^{(i)}
\otimes
\rho_B^{(i)}.
]

E.91 Social Field

A social field is a mathematical function or operator-valued distribution representing a quantity defined across a social configuration space.

Examples include fields of:

  • influence;
  • institutional constraint;
  • expectation;
  • capital;
  • information;
  • risk;
  • coordination.

E.92 Social Manifold

The social manifold is the base space

[
\mathcal{M}
]

on which social fields are defined.

It need not be ordinary geographic space.

E.93 Sparse Operator Discovery

Sparse operator discovery estimates a dynamical law from a restricted set of active terms.

A typical estimator is

\operatorname*{arg,min}{\theta}
\left[
|Y-\widehat{Y}(\theta)|^2
+
\lambda{\mathrm{reg}}|\theta|_1
\right].
]

E.94 Spectral Gap

The spectral gap is the separation between dominant eigenvalues of a generator or operator.

For a Markov or Lindblad generator, the gap often determines the relaxation rate.

E.95 State Space

The state space is the set of all admissible states.

For density operators,

\left{
\rho:
\rho\geq 0,
;
\rho=\rho^{\dagger},
;
\operatorname{Tr}\rho=1
\right}.
]

E.96 Stationary State

A stationary state satisfies

[
\mathcal{L}(\rho_*)=0.
]

E.97 Stochastic Process

A stochastic process is a family of random variables

[
{X_t}_{t\in T}.
]

E.98 Structural Identifiability

Structural identifiability asks whether ideal noiseless observations uniquely determine model parameters.

It is a property of the model structure rather than the observed sample size.

E.99 Structural Stability

A dynamical system is structurally stable when sufficiently small perturbations do not change its qualitative behavior.

E.100 Superoperator

A superoperator is a map acting on operators:

[
\mathcal{L}:
\rho
\longmapsto
\mathcal{L}(\rho).
]

E.101 Symmetry

A symmetry is a transformation that leaves specified mathematical or empirical quantities invariant.

E.102 Tensor Network

A tensor network represents a high-dimensional state through contracted lower-order tensors.

A generic representation is

[
|\Psi\rangle
\approx
\sum
A_1A_2\cdots A_N
|i_1,\ldots,i_N\rangle.
]

E.103 Tensor Product

The composite state space is

\mathcal{H}_A
\otimes
\mathcal{H}_B.
]

E.104 Topological Defect

A topological defect is a field configuration that cannot be continuously deformed into a trivial configuration without crossing an inadmissible or singular region.

E.105 Topological Invariant

A topological invariant remains unchanged under continuous deformation.

It is written generically as

[
T[\Phi].
]

E.106 Trace

The trace of an operator is

\sum_n
\langle e_n|A|e_n\rangle.
]

E.107 Trace Class

An operator is trace class when

[
\operatorname{Tr}|A|
<
\infty.
]

Density operators are trace-class operators.

E.108 Trace Norm

The trace norm is

\operatorname{Tr}|A|.
]

E.109 Unitary Operator

An operator (U) is unitary when

UU^{\dagger}

I.
]

E.110 Universality

Universality means that different microscopic models converge toward the same effective large-scale behavior.

E.111 Variational Principle

A variational principle states that the physical or effective trajectory makes an action or functional stationary.

E.112 von Neumann Equation

Closed density-operator evolution is governed by

-\frac{i}{\kappa_S}
[H,\rho].
]

E.113 Weak Convergence

Weak convergence is written as

[
x_n\rightharpoonup x.
]

It means convergence against all continuous linear functionals.

E.114 Weak Solution

A weak solution satisfies an integrated form of a differential equation rather than possessing all classical derivatives.

Part II

Index of Core Symbols

E.115 Latin Symbols

(A)

Generic matrix or linear operator.

It may also denote an adjacency matrix when written as

[
A_{ij}.
]

(A_H(t))

Observable in the Heisenberg picture.

U(t)^{\dagger}A_SU(t).
]

(B)

Generic operator, matrix, or second observable.

(\mathcal{B})

Boundary and initial-condition specification.

It may also denote a Bell-type correlation expression when defined locally.

(C(t))

Time-dependent stability or perturbation-bound coefficient.

(D)

Dataset.

In a differential equation, (D) may also denote a diffusion coefficient.

The meaning must be stated locally.

(D(\rho|\sigma))

Relative entropy.

(E_a)

POVM element:

M_a^{\dagger}M_a.
]

(F)

Generic nonlinear map or dynamical vector field.

(F(x,v))

Finsler metric function.

(G^{(n)})

(n)-point correlation function.

(G_c^{(n)})

Connected (n)-point correlation function.

(G_{ij}(A,\Phi))

Network-evolution function.

(H)

Hamiltonian.

(H_Z)

Effective Zeno Hamiltonian:

PHP.
]

(I)

Identity operator.

(I(\theta))

Fisher information matrix.

(I(A))

Mutual information between subsystems (A) and (B).

(J)

External source field.

(J_{ij})

Pairwise coupling matrix.

(K(x,y))

Spatial, social, or relational interaction kernel.

(K_k)

Kraus operator.

(L_k)

Lindblad operator.

(L_G)

Graph Laplacian.

(M_a)

Measurement operator.

(N)

Number of repeated steps, measurements, or discretization intervals.

(\mathcal{N})

Target space of a field.

It may also denote a nonlinear operator when defined locally.

(O)

Observable.

(P)

Projection operator or probability measure, depending on context.

(P_a)

Projection associated with outcome (a).

(Q)

Conserved charge or symmetry generator.

(Q_{\mathrm{top}})

Topological charge.

(R)

Scalar curvature.

(\mathcal{R})

Coarse-graining or renormalization map.

(S[\Phi])

Action functional.

(S_{\mathrm{vN}}(\rho))

von Neumann entropy.

(S_{\mathrm{Sh}}(P))

Shannon entropy.

(T(t))

Strongly continuous semigroup.

(T[\Phi])

Topological invariant.

(U(t))

Unitary evolution operator.

(U(g))

Unitary representation of a gauge or symmetry transformation.

(V(\Phi))

Potential function.

(W[J])

Connected generating functional.

(W_{mn})

Classical transition rate from state (n) to state (m).

(X)

Random variable or general state-space element.

(Y)

Observed data.

(Z[J])

Generating functional.

E.116 Greek Symbols

(\alpha)

Fractional derivative order or generic parameter.

(\beta_i(g))

Renormalization-group beta function.

(\gamma)

Decay, dissipation, or transition rate.

(\Gamma[\Phi])

Effective action.

(\Gamma^{\rho}_{\mu\nu})

Connection coefficient.

(\delta)

Variation, Dirac delta, or privacy parameter.

(\Delta_G)

Graph Laplacian.

(\varepsilon)

Measurement noise, perturbation parameter, or privacy parameter.

(\eta(t))

Inhomogeneous term arising from unresolved initial conditions.

(\theta)

Parameter.

(\boldsymbol{\theta})

Parameter vector.

(\hat{\boldsymbol{\theta}})

Estimated parameter vector.

(\kappa_S)

SQFT action scale.

(\lambda)

Eigenvalue, interaction coefficient, or regularization parameter.

A regularization coefficient should preferably be written as

[
\lambda_{\mathrm{reg}}.
]

(\mu)

Renormalization scale.

(\nu[\Phi])

Winding number or topological index.

(\rho)

Density operator.

(\rho_*)

Stationary density operator.

(\sigma)

Reference density operator, standard deviation, or entropy-production rate.

The meaning must be stated locally.

(\tau_Z)

Zeno timescale.

(\Phi)

Social field.

(\Phi_0)

Reference or background field.

(\Psi)

State vector.

(\omega)

Sample-space element or random-geometry realization.

E.117 Calligraphic Symbols

(\mathcal{A})

Algorithm or admissible class, depending on context.

(\mathcal{D})

Dissipator or dataset structure.

(\mathcal{E})

CPTP map or general channel.

(\mathcal{F})

Sigma-algebra or functional.

(\mathcal{G})

Gauge group.

(\mathcal{H})

Hilbert space.

(\mathscr{H}_{\theta})

Observation operator.

(\mathcal{I})

Identity superoperator.

(\mathcal{K})

Memory kernel.

(\mathcal{L})

Dynamical generator.

(\mathscr{L})

Lagrangian density or statistical loss, depending on subscript and context.

(\mathcal{M})

Social manifold or model space.

(\mathcal{P})

Projection superoperator.

(\mathcal{Q})

Complementary projection superoperator.

(\mathcal{R})

Coarse-graining map.

(\mathcal{S}(\mathcal{H}))

Set of admissible density operators.

(\mathcal{T})

Time-ordering operator or transition map.

(\mathcal{U}_t)

Evolution superoperator.

(\mathcal{V})

Validation protocol.

E.118 Fraktur Symbols

(\mathfrak{A})

Observable algebra.

(\mathfrak{g})

Lie algebra.

(\mathfrak{S})

Complete SQFT model specification.

Part III

Abbreviations

E.119 Mathematical and Physical Abbreviations

BCH

Baker–Campbell–Hausdorff formula.

CHSH

Clauser–Horne–Shimony–Holt inequality.

CPTP

Completely Positive Trace-Preserving.

GKSL

Gorini–Kossakowski–Sudarshan–Lindblad.

The GKSL theorem characterizes finite-dimensional Markovian quantum dynamical generators.

HS

Hilbert–Schmidt.

MAP

Maximum A Posteriori.

MLE

Maximum Likelihood Estimation.

POVM

Positive-Operator-Valued Measure.

QFT

Quantum Field Theory.

QZE

Quantum Zeno Effect.

RG

Renormalization Group.

SQFT

Social Quantum Field Theory.

SDE

Stochastic Differential Equation.

PDE

Partial Differential Equation.

ODE

Ordinary Differential Equation.

CPT

Completely Positive and Trace Preserving.

The preferred abbreviation is CPTP.

E.120 Statistical and Computational Abbreviations

AIC

Akaike Information Criterion.

BIC

Bayesian Information Criterion.

LASSO

Least Absolute Shrinkage and Selection Operator.

MCMC

Markov Chain Monte Carlo.

PCA

Principal Component Analysis.

SVD

Singular Value Decomposition.

VI

Variational Inference.

Part IV

Frequently Confused Terms

E.121 State and Observation

A state is latent:

[
\rho(t).
]

An observation is empirical:

[
Y(t).
]

They are connected through

\mathscr{H}_{\theta}
\bigl(
\rho(t)
\bigr)
+
\varepsilon(t).
]

The state should not be equated directly with raw data.

E.122 Field and Density Operator

A field is a configuration:

[
\Phi(x,t).
]

A density operator represents a probabilistic or operator state:

[
\rho(t).
]

They may be related, but they are not identical objects.

E.123 Correlation and Causation

Correlation is a statistical relation:

[
G_c^{(2)}(x,y)\neq 0.
]

Causation requires an intervention, directed model, or causal assumption.

E.124 Gauge Redundancy and Non-Identifiability

Gauge redundancy is a known equivalence between representations.

Non-identifiability means data cannot determine parameters uniquely.

Gauge redundancy may cause non-identifiability, but the two concepts are not synonymous.

E.125 Decoherence and Collapse

Decoherence is continuous or dynamical suppression of off-diagonal terms.

Measurement collapse is a conditional state update associated with a selected outcome:

[
\rho
\longrightarrow
\frac{
P_a\rho P_a
}{
\operatorname{Tr}(P_a\rho)
}.
]

These should not be treated as the same process.

E.126 Mixed State and Classical Mixture

A mixed density operator may admit several ensemble decompositions.

A classical mixture refers to uncertainty over predefined classical alternatives.

The two are formally related only after a preferred basis or commutative observable algebra has been specified.

E.127 Nonfactorization and Entanglement

Nonfactorization means

[
\rho_{AB}
\neq
\rho_A\otimes\rho_B.
]

Entanglement requires nonseparability.

All entangled states are nonfactorized, but not all nonfactorized states are entangled.

E.128 Symmetric and Self-Adjoint Operators

A symmetric operator satisfies

\langle A\Phi,\Psi\rangle
]

on its domain.

A self-adjoint operator additionally satisfies equality of operator and adjoint domains.

In infinite dimensions, the distinction is essential.

E.129 Markovian and Memoryless

“Memoryless” is an intuitive description of the Markov property.

A process may appear memoryless only after hidden variables have been included in an enlarged state.

Thus, Markovianity depends on the chosen state representation.

E.130 Renormalization and Rescaling

Rescaling changes units or coordinates.

Renormalization includes coarse-graining, parameter transformation, and identification of effective laws.

Simple rescaling is not by itself renormalization.

E.131 Topology and Geometry

Topology concerns properties preserved under continuous deformation.

Geometry concerns distance, angle, curvature, and local metric structure.

Topological invariants do not generally depend on a metric.

E.132 Structural Analogy and Literal Physics

A structural analogy uses mathematical form without claiming identical physical substance.

For example, using a density operator to represent social uncertainty does not imply that human agents are microscopic quantum particles.

Part V

Canonical Translation Table

E.133 English–Chinese Terminology

English TermRecommended Traditional Chinese
Action作用量
Adjoint Operator伴隨算子
Algebra of Observables可觀測量代數
Anticommutator反交換子
Asymptotic Stability漸近穩定性
Banach Space巴拿赫空間
Beta Function貝塔函數
Boundary Condition邊界條件
Causal Kernel因果核
Classical Limit古典極限
Coarse-Graining粗粒化
Commutator交換子
Complete Positivity完全正性
Configuration Space組態空間
Connected Correlation連通關聯
Contextuality脈絡性
Contraction Map壓縮映射
Correlation Function關聯函數
Covariant Derivative協變導數
Curvature曲率
Data-Processing Inequality資料處理不等式
Decoherence退相干
Density Operator密度算子
Differential Privacy差分隱私
Dissipator耗散項
Effective Theory有效理論
Eigenvalue特徵值
Eigenvector特徵向量
Entanglement糾纏
Entropy熵
Equilibrium平衡態
Euler–Lagrange Equation歐拉–拉格朗日方程
Expectation Value期望值
Field場
Fisher Information費雪資訊
Fixed Point固定點
Fractional Derivative分數階導數
Functional Derivative泛函導數
Gauge Equivalence規範等價
Gauge Fixing規範固定
Generating Functional生成泛函
Generator生成元
Green Function格林函數
Hamiltonian哈密頓量
Hermiticity厄米性
Hilbert Space希爾伯特空間
Interaction Kernel交互作用核
Kraus Operator克勞斯算子
Lie Algebra李代數
Lie Group李群
Lindblad Equation林德布拉德方程
Local Excitation局域激發
Markov Property馬可夫性
Measurement Operator測量算子
Memory Kernel記憶核
Metric度量
Mild Solution溫和解
Mixed State混合態
Model Misspecification模型錯置
Mutual Information互資訊
Noncommutativity非交換性
Non-Markovian Dynamics非馬可夫動力學
Observable可觀測量
Observation Map觀測映射
Open System開放系統
Operator Norm算子範數
Partial Trace偏跡
Path Integral路徑積分
Pointer Basis指標基底
Positivity正性
Projection Operator投影算子
Pure State純態
Quantum Zeno Effect量子芝諾效應
Reduced State約化態
Regularization正則化
Relative Entropy相對熵
Renormalization重整化
Renormalization Group重整化群
Self-Adjoint Operator自伴算子
Semigroup半群
Separable State可分態
Social Field社會場
Social Manifold社會流形
Sparse Operator Discovery稀疏算子發現
Spectral Gap譜隙
State Space狀態空間
Stationary State定態
Stochastic Process隨機過程
Structural Identifiability結構可識別性
Structural Stability結構穩定性
Superoperator超算子
Symmetry對稱性
Tensor Network張量網路
Tensor Product張量積
Topological Defect拓撲缺陷
Topological Invariant拓撲不變量
Trace跡
Trace Norm跡範數
Unitary Operator酉算子
Universality普適性
Variational Principle變分原理
von Neumann Equation馮紐曼方程
Weak Convergence弱收斂
Weak Solution弱解

E.134 Preferred Chinese Terminological Distinctions

The following distinctions are recommended in the Chinese edition.

State

Use:

狀態

for a general state.

Use:

態

in established technical expressions such as:

  • 純態;
  • 混合態;
  • 定態;
  • 糾纏態.

Operator

Use:

算子

rather than “操作符” in mathematical contexts.

Field

Use:

場

for the mathematical field.

Use:

場域

only when referring specifically to Bourdieu’s sociological field or an institutional domain.

This distinction is essential in the manuscript.

Collapse

Use:

坍縮

for measurement-state collapse.

Use:

崩解

for institutional, organizational, or structural breakdown.

Entanglement

Use:

糾纏

only when the formal state structure is specified.

Use:

高度關聯

or

結構耦合

for ordinary social dependence.

Gauge

Use:

規範

in mathematical physics.

Do not translate it merely as “尺度” or “標準”.

Part VI

Final Glossary Principle

E.135 Terminological Discipline

The legitimacy of an interdisciplinary mathematical theory depends partly on terminological discipline.

A term imported from another field should be used only when at least one of the following is provided:

  1. a precise mathematical definition;
  2. an explicit structural correspondence;
  3. an operational empirical interpretation.

The following implication must be avoided:

[
\text{similar vocabulary}
\quad\Longrightarrow\quad
\text{identical phenomenon}.
]

Instead, the correct principle is

[
\text{shared mathematical structure}
\quad\Longrightarrow\quad
\text{possible analytical correspondence}.
]

The glossary therefore serves not merely as a language reference, but as a boundary condition on theoretical interpretation.

No term should carry more empirical force than its mathematical definition and supporting evidence permit.


Appendix F

Research Roadmap, Conjectures, and Testable Programs of Social Quantum Field Theory

F.1 Purpose and Scope

Social Quantum Field Theory is not complete merely because it possesses a formal vocabulary, a set of dynamical equations, and an axiomatic structure.

A mature research program must also specify:

  • which problems should be solved first;
  • which conjectures are mathematically precise;
  • which models can be tested empirically;
  • which observations would count against the theory;
  • which parts of the framework may reduce to simpler classical models.

This appendix organizes the future development of SQFT into a sequence of research programs.

The roadmap is divided into four major layers:

  1. foundational mathematics;
  2. analytical and computational development;
  3. empirical identification and testing;
  4. comparative and falsification studies.

The objective is not to preserve every current component of SQFT.

The objective is to determine which components survive rigorous mathematical and empirical examination.

Part I

Foundational Research Programs

F.2 Program I — State-Space Selection

The first foundational question concerns the appropriate state space.

Possible choices include:

[
\mathcal{H},
]

a Hilbert space;

[
\mathcal{S}(\mathcal{H}),
]

the space of density operators;

[
L^p(\mathcal{M}),
]

a classical function space;

[
\mathcal{P}(\mathcal{M}),
]

a space of probability measures;

or a hybrid structure combining several of these.

Research Question F.1

Under what empirical and structural conditions is a Hilbert-space representation preferable to an ordinary classical state space?

Testable Criterion

A Hilbert-space model should demonstrate at least one of the following:

[
\text{better prediction},
]

[
\text{lower effective complexity},
]

[
\text{more stable parameter recovery},
]

or

[
\text{representation of order effects not reproduced classically}.
]

Failure Condition

If a classical latent-state model reproduces all observables with equal or lower complexity, the Hilbert-space formulation is not empirically necessary.

F.3 Program II — Observable Algebra Reconstruction

The theory assumes an observable algebra

[
\mathfrak{A}.
]

However, in most empirical applications, the algebra is not known in advance.

Research Question F.2

Can the commutation structure of observables be inferred from sequential data?

For observables (O_i) and (O_j), one seeks to estimate whether

0
]

or

[
[O_i,O_j]
\neq
0.
]

Proposed Method

Compare the distributions generated by two intervention orders:

[
O_i
\rightarrow
O_j,
]

and

[
O_j
\rightarrow
O_i.
]

Define an empirical order-effect statistic:

d
\left(
P_{i\rightarrow j},
P_{j\rightarrow i}
\right),
\tag{F.1}
]

where (d) is a statistical distance.

Open Problem F.1

Determine when nonzero (\Delta_{ij}) requires a noncommutative operator model rather than a classical model with memory.

F.4 Program III — Gauge Structure

The existence of multiple formal representations with identical empirical predictions suggests a gauge structure.

Let

[
\mathcal{G}
]

act on model parameters and states.

Conjecture F.1 — Gauge-Reduced Identifiability

For a regular SQFT model, the observable distribution determines the model uniquely up to the action of (\mathcal{G}):

g\cdot\theta_1
]

for some

[
g\in\mathcal{G}.
\tag{F.2}
]

Research Tasks

  1. Define the gauge group for specific SQFT classes.
  2. Construct gauge-invariant observables.
  3. Identify gauge directions in the Fisher information matrix.
  4. Develop numerically stable gauge-fixing procedures.

Expected Result

The empirically meaningful parameter space should be

\Theta/\mathcal{G}.
\tag{F.3}
]

F.5 Program IV — Causal Structure

Correlation functions do not automatically determine causal influence.

A causal SQFT requires additional structure.

Let

[
\mathcal{I}_{x,t}
]

denote an intervention at location (x) and time (t).

Research Question F.3

Can causal response functions be defined by

\frac{
\delta
\langle O(x,t)\rangle
}{
\delta J(y,s)
},
\tag{F.4}
]

where (J) is an external intervention source?

Causal Requirement

A causal response kernel should satisfy

0
]

whenever

[
s>t.
\tag{F.5}
]

Open Problem F.2

Construct causal identification conditions for SQFT models with:

  • hidden common causes;
  • delayed influence;
  • feedback;
  • endogenous observation;
  • changing network topology.

Part II

Analytical Research Programs

F.6 Program V — Existence and Uniqueness

Consider the nonlinear evolution equation

\mathcal{L}(\rho)
+
\mathcal{N}(\rho).
\tag{F.6}
]

Conjecture F.2 — Local Well-Posedness

Suppose:

[
\mathcal{L}
]

generates a strongly continuous contraction semigroup, and

[
\mathcal{N}
]

is locally Lipschitz on the admissible state space.

Then a unique local mild solution exists.

Research Tasks

  • prove preservation of Hermiticity;
  • prove preservation of trace;
  • derive positivity conditions;
  • characterize blow-up or global existence;
  • extend the result to infinite-dimensional spaces.

Principal Difficulty

Ordinary local existence does not guarantee that

[
\rho(t)\geq 0
]

remains valid.

The nonlinear term must preserve the convex cone of positive operators.

F.7 Program VI — Nonlinear Positivity Preservation

A nonlinear map

[
\mathcal{N}(\rho)
]

may destroy the probabilistic interpretation.

Research Question F.4

Which nonlinear operators preserve

[
\rho\geq 0,
]

[
\rho=\rho^{\dagger},
]

and

[
\operatorname{Tr}\rho=1?
]

Candidate Structure

A nonlinear evolution may be written as

\Lambda(\rho).
\tag{F.7}
]

The correction term (\Lambda(\rho)) must enforce trace preservation.

Open Problem F.3

Derive a nonlinear analogue of the GKSL characterization theorem.

F.8 Program VII — Non-Markovian Dynamics

A general memory equation is

\int_0^t
\mathcal{K}(t-s)\rho(s),ds
+
\eta(t).
\tag{F.8}
]

Conjecture F.3 — Completely Positive Memory Evolution

There exists a nontrivial class of memory kernels (\mathcal{K}) such that the resulting evolution map

[
\Lambda_t:
\rho(0)
\longmapsto
\rho(t)
]

is CPTP for every (t\geq 0).

Research Tasks

  • characterize admissible memory kernels;
  • relate kernels to hidden-state embeddings;
  • derive time-local representations;
  • distinguish genuine memory from omitted-variable effects.

Empirical Test

Compare:

[
M_{\mathrm{Markov}},
]

[
M_{\mathrm{hidden\ Markov}},
]

and

[
M_{\mathrm{non-Markov}}
]

using identical validation data.

F.9 Program VIII — Fractional Evolution

Long memory may be represented by

\mathcal{L}\rho(t),
\qquad
0<\alpha<1.
\tag{F.9}
]

Conjecture F.4 — Fractional Positivity

If (\mathcal{L}) generates a CPTP semigroup, then the subordinated fractional evolution preserves positivity and trace.

Candidate Representation

A fractional evolution may admit the subordination formula

\int_0^{\infty}
f_{\alpha}(s,t)
e^{s\mathcal{L}}
\rho_0
,ds,
\tag{F.10}
]

where

[
f_{\alpha}(s,t)\geq 0,
]

and


]

Research Question F.5

Can the fractional order (\alpha) be identified reliably from finite social time series?

F.10 Program IX — Time-Dependent Hilbert Spaces

In changing social systems, the effective number of degrees of freedom may vary.

The state space becomes

[
\mathcal{H}(t).
]

Research Question F.6

How should states at different times be compared when they belong to different spaces?

Candidate Structure

Introduce transport maps

[
\mathcal{T}_{t_1\rightarrow t_2}:
\mathcal{H}(t_1)
\rightarrow
\mathcal{H}(t_2).
\tag{F.11}
]

A covariant evolution equation is

H(t)|\Psi(t)\rangle.
\tag{F.12}
]

Open Problem F.4

Construct a bundle-theoretic formulation in which:

  • the fibers are Hilbert spaces;
  • transport is probabilistically admissible;
  • inner products evolve consistently;
  • state dimension may change.

F.11 Program X — Structural Stability

Consider

\mathcal{L}_{\theta}(\rho).
]

A perturbation of the generator produces

[
\mathcal{L}{\theta}
\longrightarrow
\mathcal{L}{\theta}
+
\delta\mathcal{L}.
]

Conjecture F.5 — Spectral Stability

If the generator possesses an isolated stationary state and a strictly positive spectral gap, then sufficiently small perturbations preserve the qualitative relaxation structure.

Research Tasks

  • estimate movement of stationary states;
  • derive perturbation bounds for spectral gaps;
  • classify bifurcations;
  • identify regime transitions.

Empirical Interpretation

A structural transition should correspond to a qualitative change in observable dynamics, not merely a large numerical fluctuation.

Part III

Geometric and Topological Programs

F.12 Program XI — Geometry of Social State Space

Let

[
\mathcal{M}
]

be a social manifold with metric

[
g_{\mu\nu}.
]

Research Question F.7

Can observed transition costs determine the metric?

Suppose the cost of moving from (x) to (x+dx) is

g_{\mu\nu}(x)
dx^{\mu}dx^{\nu}.
\tag{F.13}
]

Proposed Empirical Strategy

Estimate local transition probabilities:

[
P(x+dx\mid x).
]

Infer the metric from the local quadratic form of transition cost.

Open Problem F.5

Determine conditions under which (g_{\mu\nu}) is identifiable from finite transition data.

F.13 Program XII — Information Geometry

Let a statistical model be

[
p(Y\mid\theta).
]

The Fisher metric is

I_{ij}(\theta).
\tag{F.14}
]

Research Question F.8

Can institutional or ideological distances be defined as geodesic distances in parameter space?

The geodesic distance is

\inf_{\gamma}
\int_0^1
\sqrt{
g_{ij}(\gamma)
\dot{\gamma}^{i}
\dot{\gamma}^{j}
}
,ds.
\tag{F.15}
]

Testable Use

Compare whether geodesic distance predicts transition difficulty better than Euclidean parameter distance.

F.14 Program XIII — Topological Defects

Let

[
\Phi:
\mathcal{M}
\rightarrow
\mathcal{N}.
]

A nontrivial homotopy class may represent a persistent structural obstruction.

Conjecture F.6 — Persistent Institutional Defects

Certain long-lived institutional divisions correspond to nontrivial elements of

[
\pi_n(\mathcal{N}).
\tag{F.16}
]

Required Evidence

A valid application must specify:

  • the base space;
  • the target space;
  • the boundary conditions;
  • the relevant homotopy group;
  • the observable consequence of the topological class.

Failure Condition

If the classification changes under arbitrarily small perturbations of representation, it is not a robust topological description.

F.15 Program XIV — Topological Transitions

Let

[
T[\Phi]
]

be a topological invariant.

A transition occurs when

[
T[\Phi_-]
\neq
T[\Phi_+].
\tag{F.17}
]

Research Question F.9

Can institutional restructuring be represented as a change in a robust topological invariant?

Candidate Data

  • communication networks;
  • organizational hierarchies;
  • coalition structures;
  • migration pathways;
  • supply-chain connectivity.

Open Problem F.6

Develop statistical tests for changes in topological invariants under noisy network observation.

Part IV

Multiscale and Renormalization Programs

F.16 Program XV — Scale-Dependent Parameters

Let

[
g_i(\ell)
]

denote effective couplings at scale (\ell).

Their flow is

\beta_i(g).
\tag{F.18}
]

Research Question F.10

Which parameters are:

  • relevant;
  • irrelevant;
  • marginal?

Proposed Procedure

  1. estimate models at multiple aggregation scales;
  2. track parameter changes;
  3. fit beta functions;
  4. identify fixed points;
  5. test whether large-scale observables collapse onto common curves.

F.17 Program XVI — Universality Classes

Conjecture F.7 — Social Universality

Systems with different microscopic structures but identical:

  • symmetry;
  • effective dimension;
  • conservation law;
  • interaction range;
  • relevant operators;

may converge to the same macroscopic scaling behavior.

Formally,

[
\mathcal{R}^n(S_1)
\longrightarrow
S^*,
]

and

[
\mathcal{R}^n(S_2)
\longrightarrow
S^*.
\tag{F.19}
]

Empirical Test

Estimate critical exponents or scaling relations across distinct social systems.

Universality is supported only if the same exponents appear within uncertainty bounds.

F.18 Program XVII — Noncommuting Coarse-Graining and Evolution

In general,

[
\mathcal{R}\mathcal{U}_t
\neq
\mathcal{U}^{\mathrm{eff}}_t\mathcal{R}.
\tag{F.20}
]

Define the scale-consistency error:

\mathcal{R}\mathcal{U}_t

\mathcal{U}^{\mathrm{eff}}_t\mathcal{R}.
\tag{F.21}
]

Research Question F.11

Can one construct effective dynamics satisfying

[
|\Delta_t|
\leq
\varepsilon
]

over a prescribed time interval?

Practical Importance

This determines whether a model estimated at one aggregation scale remains valid at another.

F.19 Program XVIII — Dynamic Networks

Let the node field satisfy

F_i(\Phi,A),
]

and the network satisfy

G_{ij}(A,\Phi).
\tag{F.22}
]

Research Questions

  • When does the coupled system possess a unique solution?
  • When does synchronization occur?
  • When do network edges collapse or reorganize?
  • Can field instability cause topological transition?
  • Can topology stabilize an otherwise unstable field?

Conjecture F.8 — Coevolutionary Criticality

There exist parameter regimes in which field instability and network restructuring reinforce each other, producing abrupt regime change.

F.20 Program XIX — Higher-Order Interaction

Pairwise interactions may be insufficient.

A higher-order model includes

[
\sum_{i,j,k}
\mathcal{I}_{ijk}
\Phi_i\Phi_j\Phi_k.
\tag{F.23}
]

Research Question F.12

Can higher-order operators be identified from data without severe overfitting?

Proposed Estimator

\operatorname*{arg,min}_{\theta}
\left[
|Y-\widehat{Y}(\theta)|^2
+
\lambda_1|\theta|1
+
\lambda_2R{\mathrm{hier}}(\theta)
\right].
\tag{F.24}
]

The hierarchical penalty should ensure that higher-order terms are selected only when lower-order structure is insufficient.

Part V

Computational Programs

F.21 Program XX — Tensor-Network Approximation

The full state space may grow exponentially with subsystem number.

A tensor-network approximation is

[
|\Psi\rangle
\approx
\sum
A_1A_2\cdots A_N
|i_1,\ldots,i_N\rangle.
\tag{F.25}
]

Conjecture F.9 — Low-Entanglement Social States

Empirically stable SQFT states possess sufficiently limited correlation complexity to admit efficient tensor-network approximation.

Testable Quantity

The approximation error is

|\Psi_{\chi}\rangle
|,
\tag{F.26}
]

where (\chi) is the bond dimension.

Research Goal

Determine how (\chi) must scale with system size for realistic social networks.

F.22 Program XXI — Sparse Generator Discovery

Suppose the generator is expanded in a dictionary:

\sum_{j=1}^{p}
\theta_j
\mathcal{B}_j.
\tag{F.27}
]

A sparse estimator is

\operatorname*{arg,min}{\theta}
\left[
\mathscr{L}{\mathrm{data}}(\theta)
+
\lambda_{\mathrm{reg}}
|\theta|_1
\right].
\tag{F.28}
]

Research Questions

  • How many observations are required?
  • Which operator dictionaries are identifiable?
  • How should gauge redundancy be removed?
  • How should positivity constraints be enforced?
  • Can hidden environmental channels be recovered?

F.23 Program XXII — Low-Rank State Reconstruction

Assume

r
\ll
\dim\mathcal{H}.
]

Proposed Factorization

XX^{\dagger},
\tag{F.29}
]

where

[
X
\in
\mathbb{C}^{d\times r}.
]

Normalization requires


]

Research Question F.13

Can low-rank SQFT states be reconstructed from incomplete observables with provable error bounds?

F.24 Program XXIII — Numerical Positivity Preservation

Naive integration may produce states violating

[
\rho\geq 0.
]

Research Goal

Develop integrators that preserve:

[
\rho\geq 0,
]

[
\rho=\rho^{\dagger},
]

and

[
\operatorname{Tr}\rho=1.
]

Candidate Methods

  • exponential integrators;
  • Kraus-form updates;
  • operator splitting;
  • positivity-constrained optimization;
  • manifold-based integration.

Validation Criterion

At every numerical step,

[
\lambda_{\min}(\rho_n)
\geq
-\varepsilon_{\mathrm{num}}.
\tag{F.30}
]

F.25 Program XXIV — Computational Complexity

Research Question F.14

What is the complexity of:

  • exact SQFT simulation;
  • generator estimation;
  • low-rank reconstruction;
  • tensor-network contraction;
  • topological classification;
  • gauge reduction?

Classification Goal

Determine whether central problems are:

  • polynomial-time solvable;
  • NP-hard;
  • statistically intractable;
  • approximable under structural assumptions.

A theory that cannot be computed or approximated cannot support practical empirical testing.

Part VI

Empirical Research Programs

F.26 Program XXV — Order Effects

Sequential questions, decisions, or interventions may exhibit order dependence.

Let

[
P_{AB}
]

denote outcomes when (A) precedes (B), and

[
P_{BA}
]

the reversed order.

Define

d(P_{AB},P_{BA}).
\tag{F.31}
]

Competing Models

  1. classical memory model;
  2. Bayesian updating model;
  3. hidden-state Markov model;
  4. noncommutative SQFT model.

Test

Use held-out sequential data to determine whether the operator model provides superior prediction.

F.27 Program XXVI — Collective Synchronization

Let actors be represented by local fields

[
\Phi_i(t).
]

Define the collective order parameter

\left|
\frac{1}{N}
\sum_{j=1}^{N}
e^{i\phi_j(t)}
\right|.
\tag{F.32}
]

Research Question F.15

Can synchronization transitions be predicted from the interaction spectrum?

Candidate Systems

  • financial herding;
  • electoral coordination;
  • online attention;
  • organizational consensus;
  • protest mobilization.

F.28 Program XXVII — Institutional Collapse

Let

[
\rho(t)
]

represent a distribution over institutional regimes.

A collapse event may be modeled as a rapid transition from a metastable region

[
\mathcal{A}
]

to another region

[
\mathcal{B}.
]

Observable Indicators

  • decline of spectral gap;
  • increase in variance;
  • critical slowing down;
  • rising cross-correlation;
  • increased escape probability.

Research Question F.16

Can SQFT indicators predict regime transition earlier than classical early-warning models?

F.29 Program XXVIII — Environmental Decoherence Analogy

Suppose off-diagonal terms represent unresolved cross-alternative coherence.

A dephasing model gives

\rho_{mn}(0)e^{-\Gamma_{mn}t}.
\tag{F.33}
]

Empirical Program

Estimate whether repeated public observation, regulation, or institutional monitoring suppresses ambiguity between competing states.

Methodological Limitation

Observed convergence may also arise from:

  • conformity;
  • information diffusion;
  • strategic adaptation;
  • selection effects.

The SQFT interpretation must be compared against these alternatives.

F.30 Program XXIX — Zeno-Type Institutional Locking

Repeated intervention may inhibit transition.

Let (P) denote a constrained institutional sector.

The effective evolution is

[
\left[
P
e^{-iHt/(N\kappa_S)}
P
\right]^N.
\tag{F.34}
]

Research Question F.17

Do frequent audits, approvals, reviews, or veto points reduce transition rates?

Classical Comparator

A classical hazard model with intervention-dependent rate:

\lambda_0(t)
f(N).
\tag{F.35}
]

The Zeno-type model is useful only if it outperforms simpler hazard-based explanations.

F.31 Program XXX — Capital and Resource Fields

Let

[
\Phi_C(x,t)
]

represent an effective field of capital, resources, or influence.

A diffusion-interaction model is

D_C\nabla^2\Phi_C

\frac{dV}{d\Phi_C}
+
J_C.
\tag{F.36}
]

Research Questions

  • Does capital propagate diffusively?
  • Are there nonlinear thresholds?
  • Do institutional boundaries act as potential barriers?
  • Can resource concentration generate symmetry breaking?

Required Caution

The field representation must be derived from measurable variables rather than treated as a rhetorical substitute for ordinary economic quantities.

F.32 Program XXXI — Sports Coordination Case Study

A sports team provides a bounded and observable environment for testing coordination models.

Let each player be a local subsystem:

\bigotimes_{i=1}^{N}
\mathcal{H}_i.
]

The coach and tactical system are represented by global constraints or control operators:

H_{\mathrm{player}}
+
H_{\mathrm{interaction}}
+
H_{\mathrm{tactic}}.
\tag{F.37}
]

Observable Data

  • player positions;
  • passing networks;
  • reaction times;
  • tactical formations;
  • possession transitions;
  • pressing synchronization.

Research Question F.18

Does an operator-based coordination model predict tactical transitions better than classical network and state-space models?

Interpretation

Players are local excitations.

The coach modifies global constraints.

Tactics reshape the interaction field.

The model remains structural rather than literally quantum mechanical.

Part VII

Comparative Research Programs

F.33 Program XXXII — Classical Benchmarking

Every SQFT model should be compared against at least one simpler model.

Possible comparators include:

  • linear regression;
  • vector autoregression;
  • hidden Markov models;
  • stochastic differential equations;
  • agent-based models;
  • network diffusion;
  • Bayesian state-space models;
  • neural dynamical systems.

Evaluation Criteria

[
\text{predictive accuracy},
]

[
\text{calibration},
]

[
\text{parameter count},
]

[
\text{computational cost},
]

[
\text{interpretability}.
]

Rule

SQFT should not be preferred solely because it is mathematically richer.

F.34 Program XXXIII — Ablation Studies

An ablation study removes one formal component at a time.

For example:

[
M_{\mathrm{full}},
]

[
M_{\mathrm{no\ commutator}},
]

[
M_{\mathrm{no\ memory}},
]

[
M_{\mathrm{no\ topology}},
]

[
M_{\mathrm{classical}}.
]

Research Goal

Determine which components contribute genuine predictive value.

Failure Condition

If removing a component does not reduce performance or explanatory adequacy, that component is not empirically justified.

F.35 Program XXXIV — Out-of-Sample Prediction

A valid empirical model must be evaluated on unseen data.

Let

D_{\mathrm{train}}
\cup
D_{\mathrm{test}}.
]

The training data determine

[
\hat{\theta}.
]

Prediction is evaluated using

[
\mathcal{S}
\left(
\hat{\theta};
D_{\mathrm{test}}
\right).
\tag{F.38}
]

Prohibited Practice

The same data should not be used simultaneously to:

  • construct the model;
  • select operators;
  • estimate parameters;
  • report final performance.

F.36 Program XXXV — Cross-System Generalization

A strong theory should transfer across systems.

Train the model on system (A):

[
D_A.
]

Test it on system (B):

[
D_B.
]

Research Question F.19

Which SQFT quantities are system specific, and which are universal?

Candidate Transferable Objects

  • symmetry class;
  • spectral ratios;
  • scaling exponents;
  • normalized interaction kernels;
  • topological signatures.

F.37 Program XXXVI — Counterfactual Prediction

A causal SQFT model should predict outcomes under interventions.

Let

[
\mathcal{I}_a
]

and

[
\mathcal{I}_b
]

be two intervention maps.

The counterfactual contrast is

\operatorname{Tr}
\left[
O\mathcal{I}_b(\rho)
\right].
\tag{F.39}
]

Validation

Counterfactual predictions should be tested using:

  • randomized interventions;
  • natural experiments;
  • policy discontinuities;
  • synthetic controls;
  • sequential intervention data.

Part VIII

Falsification Programs

F.38 Program XXXVII — Classical Sufficiency Test

Null Hypothesis

A classical model is sufficient:

[
H_0:
M_{\mathrm{classical}}
]

reproduces every relevant observable.

Alternative Hypothesis

A noncommutative SQFT model is required:

[
H_1:
M_{\mathrm{SQFT}}.
]

Rejection Standard

The SQFT model must improve out-of-sample performance after penalizing complexity.

A possible criterion is

\mathrm{Score}_{\mathrm{SQFT}}

\mathrm{Score}_{\mathrm{classical}}.
\tag{F.40}
]

Support for SQFT requires

[
\Delta\mathrm{Score}>0
]

with uncertainty excluding zero.

F.39 Program XXXVIII — Positivity Failure Test

A proposed generator is inadmissible if it produces

[
\rho(t)\not\geq 0.
]

Define

\min
\sigma(\rho(t)).
]

The model fails when

[
\lambda_{\min}(t)
<
-\varepsilon
\tag{F.41}
]

for a tolerance exceeding numerical error.

F.40 Program XXXIX — Identifiability Failure Test

Let

[
J(\theta)
]

be the Jacobian of observable predictions.

If

[
\operatorname{rank}J(\theta)
<
\dim\Theta-\dim\mathcal{G},
\tag{F.42}
]

then the model contains unidentified nongauge directions.

Such parameters should not be given substantive interpretation.

F.41 Program XL — Predictive Collapse Test

A theory fails empirically when it performs well only in sample.

Define

[
E_{\mathrm{train}}
]

and

[
E_{\mathrm{test}}.
]

The generalization gap is

E_{\mathrm{test}}

E_{\mathrm{train}}.
\tag{F.43}
]

A large positive (G) indicates overfitting.

F.42 Program XLI — Metaphor Detection Test

Every claimed SQFT object should answer the following questions:

  1. What is its mathematical domain?
  2. What is its codomain?
  3. How is it measured?
  4. What equation governs it?
  5. What evidence would reject it?

If these questions cannot be answered, the object remains metaphorical rather than formal.

Part IX

Priority Schedule

F.43 Priority Level I — Immediate Foundations

The following problems should be addressed first:

  1. state-space justification;
  2. observation-map specification;
  3. gauge-reduced identifiability;
  4. classical benchmark construction;
  5. positivity-preserving dynamics;
  6. reproducible model specification.

These problems determine whether later sophistication is scientifically meaningful.

F.44 Priority Level II — Analytical Development

The second stage should address:

  • nonlinear well-posedness;
  • memory-kernel positivity;
  • fractional dynamics;
  • dynamic Hilbert spaces;
  • perturbation theory;
  • spectral stability.

F.45 Priority Level III — Computational Development

The third stage should develop:

  • sparse generator discovery;
  • low-rank reconstruction;
  • tensor-network simulation;
  • positivity-preserving solvers;
  • scalable model comparison;
  • gauge-aware optimization.

F.46 Priority Level IV — Empirical Case Studies

The fourth stage should focus on bounded, data-rich systems.

Recommended initial domains include:

  • sports coordination;
  • online information diffusion;
  • organizational decision sequences;
  • financial synchronization;
  • coalition-network evolution.

These domains possess clear temporal data and measurable interaction structures.

F.47 Priority Level V — Large-Scale Social Systems

Only after successful bounded-case validation should SQFT be extended to:

  • national political systems;
  • global financial networks;
  • long-term institutional history;
  • geopolitical transitions;
  • civilizational-scale dynamics.

Large systems contain severe confounding, incomplete observation, and changing measurement regimes.

Part X

Reproducibility Standard

F.48 Minimum Reproducible Package

Every empirical SQFT publication should include:

\left(
D,
C,
\Theta,
\mathcal{A},
S,
V
\right),
\tag{F.44}
]

where:

  • (D) is the data or a legally shareable substitute;
  • (C) is the source code;
  • (\Theta) is the parameter specification;
  • (\mathcal{A}) is the algorithm;
  • (S) is the random seed structure;
  • (V) is the validation protocol.

Required Reporting

  • preprocessing steps;
  • missing-data treatment;
  • initialization;
  • convergence criteria;
  • uncertainty estimates;
  • comparator models;
  • failed model variants.

F.49 Model Card

Each model should include a compact model card.

Model Name

Intended Use

State Representation

Observation Map

Data Requirements

Assumptions

Known Limitations

Comparator Models

Validation Results

Failure Conditions

Ethical Risks

This allows formal complexity to remain transparent to non-specialist users.

Part XI

Ethical Research Programs

F.50 Program XLII — Privacy

High-dimensional social-state reconstruction may reveal sensitive information.

A private estimator should satisfy

[
P(M(D)\in S)
\leq
e^{\varepsilon}
P(M(D')\in S)
+
\delta.
\tag{F.45}
]

Research Question F.20

How much predictive structure is lost under privacy-preserving noise?

Required Trade-Off Analysis

[
\text{privacy}
\leftrightarrow
\text{identifiability}
\leftrightarrow
\text{prediction}.
]

F.51 Program XLIII — Interpretability

A learned operator should not be interpreted through arbitrary matrix entries.

Interpretation should focus on invariant quantities such as:

  • eigenvalues;
  • spectral gaps;
  • invariant subspaces;
  • conserved quantities;
  • interaction graphs;
  • dominant channels.

Research Question F.21

Can an invariant interpretation map

[
\mathfrak{I}(\mathcal{L})
]

be constructed such that

\mathfrak{I}
\left(
U\mathcal{L}U^{-1}
\right)?
\tag{F.46}
]

F.52 Program XLIV — Intervention Risk

A model capable of predicting social transitions may also be used to manipulate them.

Research should distinguish:

  • descriptive use;
  • predictive use;
  • policy simulation;
  • direct behavioral intervention.

A high-impact intervention model should include:

  • uncertainty bounds;
  • distributional effects;
  • failure modes;
  • affected populations;
  • reversible testing stages.

Part XII

Consolidated Conjecture List

F.53 Conjecture F.1 — Gauge-Reduced Identifiability

Observable data identify the model uniquely up to gauge equivalence.

F.54 Conjecture F.2 — Local Well-Posedness

Locally Lipschitz nonlinear SQFT equations generated around a contraction semigroup admit unique local mild solutions.

F.55 Conjecture F.3 — Completely Positive Memory Evolution

A broad class of non-Markovian kernels generates CPTP evolution.

F.56 Conjecture F.4 — Fractional Positivity

Subordinated fractional evolution preserves positivity and normalization.

F.57 Conjecture F.5 — Spectral Stability

A stationary state separated by a positive spectral gap remains structurally stable under sufficiently small generator perturbations.

F.58 Conjecture F.6 — Persistent Institutional Defects

Certain durable institutional divisions correspond to robust topological classes.

F.59 Conjecture F.7 — Social Universality

Distinct microscopic systems with matching structural invariants converge toward the same macroscopic scaling laws.

F.60 Conjecture F.8 — Coevolutionary Criticality

Field instability and network restructuring may jointly generate abrupt transitions.

F.61 Conjecture F.9 — Low-Entanglement Social States

Empirically stable SQFT states admit efficient low-complexity tensor-network representation.

F.62 Conjecture F.10 — Effective Classicality

Under decoherence, coarse-graining, and restricted observability, SQFT dynamics converge toward a classical stochastic process:

[
\mathcal{L}{\mathrm{SQFT}}
\longrightarrow
\mathcal{L}{\mathrm{classical}}.
\tag{F.47}
]

F.63 Conjecture F.11 — Finite Correlation Reconstruction

For finite-dimensional analytic SQFT systems, a sufficiently rich finite collection of multi-time correlations identifies the generator modulo gauge equivalence.

F.64 Conjecture F.12 — Low-Rank Effective State

For stable observable regimes,

[
\rho
\approx
\rho_r,
]

with

[
r
\ll
\dim\mathcal{H}.
\tag{F.48}
]

Part XIII

Criteria for Theoretical Progress

F.65 Mathematical Progress

A mathematical result constitutes progress when it:

  • proves existence or uniqueness;
  • establishes positivity;
  • derives a stability bound;
  • characterizes a gauge class;
  • identifies an invariant;
  • proves a classical correspondence;
  • establishes computational complexity.

F.66 Computational Progress

A computational result constitutes progress when it:

  • reduces simulation cost;
  • preserves admissibility;
  • improves parameter recovery;
  • provides certified error bounds;
  • scales to larger systems;
  • enables reproducible comparison.

F.67 Empirical Progress

An empirical result constitutes progress when it:

  • predicts unseen data;
  • identifies stable parameters;
  • survives comparator testing;
  • generalizes across datasets;
  • produces calibrated uncertainty;
  • survives falsification attempts.

F.68 Negative Results

Negative results are also scientifically valuable.

Examples include:

  • failure to identify noncommutativity;
  • reduction to a classical model;
  • non-identifiable parameters;
  • absence of universal scaling;
  • instability of topological classification;
  • failure of tensor-network compression.

A rigorous negative result narrows the domain of the theory and improves its scientific credibility.

Part XIV

Final Research Principle

F.69 Long-Term Objective

The long-term objective of SQFT is not to show that every social phenomenon is quantum-like.

It is to determine whether a unified mathematical framework can rigorously describe systems possessing:

  • many interacting degrees of freedom;
  • contextual observation;
  • nonlinear feedback;
  • historical memory;
  • changing topology;
  • multiscale organization;
  • abrupt structural transition.

The theory succeeds only when its formal machinery produces results that simpler frameworks cannot provide with equal clarity and efficiency.

F.70 Final Roadmap Principle

The development of SQFT should proceed in the following order:

[
\text{define},
]

[
\text{prove},
]

[
\text{compute},
]

[
\text{estimate},
]

[
\text{compare},
]

[
\text{attempt to falsify}.
]

This order may be summarized as

[
\text{Formal coherence}
\rightarrow
\text{mathematical validity}
\rightarrow
\text{empirical necessity}.
\tag{F.49}
]

The governing research principle is therefore:

[
\text{Do not ask whether SQFT can describe everything.}
]

[
\text{Ask precisely where it describes more than simpler theories can.}
\tag{F.50}
]


Preface

Toward a Mathematical Language for Social Fields

Every scientific revolution begins with a change in language.

Classical mechanics introduced the language of differential equations. Electromagnetism unified electricity and magnetism through field equations. General relativity replaced gravitational force with the geometry of spacetime, while quantum mechanics reconstructed physical reality in terms of state spaces, operators, and probability amplitudes. Each major advance in science was accompanied not only by new discoveries, but also by the creation of a new mathematical framework capable of expressing previously inaccessible phenomena.

The social sciences have followed a different path.

Over the past century, sociology, economics, political science, psychology, and network science have produced numerous influential theories describing institutions, organizations, markets, collective behavior, and social interaction. Yet these theories often employ different conceptual languages and mathematical formalisms. Narrative explanation, statistical modeling, agent-based simulation, game theory, and network analysis each illuminate particular aspects of social systems, but no single mathematical framework has emerged that is capable of integrating them into a unified description.

This book is an attempt to explore such a possibility.

The central idea of Social Quantum Field Theory (SQFT) is not that society is literally governed by microscopic quantum mechanics, nor that human beings are quantum particles. Rather, it proposes that many complex social systems may be described more coherently by borrowing the mathematical structures developed in quantum field theory, open-system dynamics, operator theory, geometry, topology, and multiscale analysis.

Within this framework, individual actors are treated not as isolated and independent entities, but as localized excitations embedded within evolving social fields. Institutions, norms, information, and capital are interpreted as structured fields that constrain and shape collective dynamics. Observation itself is regarded as part of the dynamical process, while large-scale institutional transformations are represented as changes in the global state rather than merely the accumulation of independent individual decisions.

The purpose of this work is therefore not to replace existing social theories, but to provide them with a more rigorous mathematical language whenever such a language proves useful.

Throughout this book, mathematical precision takes precedence over metaphor. Whenever concepts originating from physics—such as field, state, entanglement, decoherence, collapse, topology, or renormalization—are introduced, their meanings are explicitly defined within the mathematical framework of SQFT. Similar terminology does not imply identical physical mechanisms. Structural correspondence should never be confused with physical identity.

For this reason, the theory proposed here should be understood as a program of mathematical modeling rather than as a claim regarding the microscopic constitution of society.

The reader will also notice that this book deliberately distinguishes between established results and open questions. Some chapters develop formal definitions and derive mathematical consequences from explicitly stated axioms. Other chapters propose conjectures, research programs, and empirical tests that remain to be explored. Such distinctions are essential. Scientific progress depends as much upon clearly identifying what is not yet known as upon presenting what is already understood.

The development of Social Quantum Field Theory should therefore be viewed as an ongoing research program rather than a completed doctrine. Like any new theoretical framework, its value will ultimately be determined neither by philosophical appeal nor by mathematical elegance alone, but by its ability to organize knowledge, generate testable predictions, and withstand empirical scrutiny.

The objective of this volume is modest, yet ambitious.

It is modest because it does not claim to provide a final theory of society.

It is ambitious because it seeks to establish the mathematical foundations upon which such a theory might eventually be constructed.

If the framework developed in the following chapters encourages clearer thinking, more rigorous modeling, and deeper dialogue between mathematics, physics, and the social sciences, then the purpose of this book will have been fulfilled.

Acknowledgements

The development of Social Quantum Field Theory (SQFT) has been a long interdisciplinary journey spanning mathematics, theoretical physics, sociology, complexity science, information theory, and philosophy of science.

This work has benefited from more than a century of intellectual developments in these disciplines. Although the framework proposed in this volume is original in its formulation, it is deeply indebted to the ideas developed by generations of researchers whose work has shaped our understanding of fields, dynamical systems, probability, geometry, topology, and social structure.

The author wishes to acknowledge the contributions of the broader scientific community whose published research has provided the intellectual foundation upon which this work has been built.

The author also gratefully recognizes the value of contemporary computational tools and artificial intelligence systems, which have greatly assisted in literature organization, mathematical typesetting, language refinement, and manuscript preparation. Such tools have served as research assistants in the process of scholarly writing; however, all theoretical constructions, interpretations, mathematical formulations, and conclusions presented in this book remain the sole responsibility of the author.

Finally, the author expresses sincere appreciation to all readers, researchers, reviewers, and future collaborators who may examine, criticize, refine, or extend the ideas proposed in this volume. Scientific knowledge advances through open discussion, careful verification, and continual revision. It is the author's hope that this work will contribute, even in a small way, to the ongoing dialogue between mathematics, physics, and the social sciences.

Any remaining errors, omissions, or misunderstandings are entirely the responsibility of the author.



Manifesto of Social Quantum Field Theory


Every scientific discipline begins with a question that existing languages cannot adequately answer.

Classical mechanics emerged because geometry alone could not describe motion. Electromagnetism arose because electricity and magnetism required a common field description. General relativity was born when gravity could no longer be understood merely as force. Quantum mechanics appeared because classical trajectories failed to explain the microscopic world.

The history of science is therefore not merely a history of new discoveries; it is a history of new mathematical languages.

The social sciences now face a similar moment.

Modern society consists of billions of interacting individuals, institutions, technologies, information flows, and adaptive networks. These systems evolve simultaneously across multiple temporal and spatial scales. They display emergence, collective synchronization, abrupt transitions, historical memory, contextual dependence, and continual structural reorganization.

Although many successful theories have been developed to explain particular aspects of these phenomena, no single mathematical framework has yet provided a unified description of their collective dynamics.

Social Quantum Field Theory begins with a simple proposition:

The fundamental object of social analysis is not the isolated individual, but the evolving field of relationships within which individuals exist and act.

Individuals are indispensable, yet they are not mathematically primary.

Just as particles are understood in modern physics as excitations of underlying quantum fields, social actors may be represented as localized excitations embedded within continuously evolving social fields. Institutions, organizations, norms, information, and capital are not external backgrounds; they participate in the dynamics of the field itself.

This shift of perspective—from object-centered description to field-centered description—is the central idea of this book.

However, the language of quantum field theory is adopted here as a mathematical framework rather than as a literal ontological claim.

Social Quantum Field Theory does not assert that societies obey microscopic quantum mechanics, nor that human consciousness is governed by quantum effects.

Instead, it argues that many mathematical structures developed in modern theoretical physics—operator algebras, open-system dynamics, multiscale renormalization, information geometry, topological invariants, and stochastic evolution—provide powerful tools for describing complex adaptive systems whenever their structural properties are shared.

Structural similarity should never be mistaken for physical identity.

Throughout this volume, every concept introduced from physics is therefore redefined within the mathematical framework of SQFT before being applied to social systems.

The goal is not metaphor.

The goal is formalization.

Scientific theories should ultimately be judged by four standards:

  • conceptual clarity;
  • mathematical consistency;
  • empirical testability;
  • explanatory power.

Novel terminology alone carries no scientific value.

Likewise, mathematical sophistication without observational relevance is insufficient.

For this reason, the present work deliberately distinguishes between established mathematical results, working hypotheses, conjectures, and future research programs.

Where rigorous proofs are available, they are presented.

Where only plausible hypotheses exist, they are explicitly identified as conjectures.

Where empirical validation remains incomplete, testable research programs are proposed rather than definitive conclusions asserted.

The reader should therefore regard this book not as the final form of a completed theory, but as the first systematic attempt to establish a coherent mathematical language for social fields.

Every new scientific framework must eventually answer three questions:

Can it explain existing observations?

Can it predict phenomena that previous theories cannot?

Can it survive attempts at falsification?

Only future research can answer these questions.

The purpose of this book is more limited, yet no less important.

It is to define the mathematical foundations upon which those future answers may be built.




留言

張貼留言

這個網誌中的熱門文章

社會量子場論（SQFT）超越布迪厄的場域理論，建構關係社會學的形式化 1~10

圖片
  預印本 社會量子場論與場論的形式化 周易先 獨立研究者 · 版本 2.0（數學版） · 2026年7月 摘要 本文提出社會量子場論（SQFT）作為用以表示複雜社會系統中關係動態的形式數學框架。SQFT並不主張社會體系遵守量子力學法則。它採用量子場論、量子資訊理論、微分幾何、拓撲學、變分動力學及對稱理論中選定的結構，以形式化關係社會學，其中社會行為者被視為演化關係場的局部顯現。此框架建立在皮埃爾·布迪厄的場、資本與慣習概念之上，解決了持續存在的形式問題：共享的關係配置如何在多個社會行為者間產生協調、非線性及結構轉換效應，而無需將這些效應簡化為孤立的個人間因果關係。八個公理建立了概念架構，並透過協調足球比賽的工作模型說明關係不可分離性及可測量的協調。該框架進一步透過社會希爾伯特空間、場算子、關係度量與曲率、有效社會作用與拉格朗日量、規範轉換、對稱破缺以及諾特定理類型的平衡定律加以擴展。操作途徑包括網絡分析、追蹤數據、事件研究、基於代理的建模、資訊幾何及拓撲數據分析。SQFT作為可證偽的建模語言與研究計劃提出，而非作為社會中存在物理量子行為的證據。 關鍵詞：關係社會學；場論；皮埃爾·布迪厄；量子形式主義；複雜系統；社會協調；開放系統動力學；拓撲學 中央論旨與研究目的 中央論旨（Central Thesis） 社會量子場論（SQFT）提出了一套正式的數學框架，用來描述複雜社會系統中的關係動力學。 其核心論旨是： 社會行動者並非獨立實體，僅僅在場域內相互作用； \text{社會行動者並非獨立實體，僅僅在場域內相互作用；} 他們是演化中的關係場域的局部顯現。 \text{他們是演化中的關係場域的局部顯現。} 因此，SQFT 的基本分析單位不是孤立的個別行動者，而是那個讓行動者、位置、身份、制度與各種資本獲得社會意義的 關係組態 。 這個框架延續了皮耶・布迪厄的場域理論，並引入量子場論、量子資訊理論、開放系統動力學、微分幾何、拓樸學以及複雜系統理論中經過挑選的數學結構。 轉移這些結構的目的 不是 主張人類社會在物理上服從量子力學。SQFT 是將用來分析關係系統的數學結構，借用來描述以下社會現象的特徵： 結構相互依賴 集體協調 非線性回饋 制度涌現 歷史路徑依賴 不連續危機 拓樸轉型 開放系統演化 理論的方法論邊界因此明確...
閱讀完整內容

量子之影：台灣QNF-3量子導航系統的崛起與其地緣政治影響

圖片
  量子之影：台灣QNF-3量子導航系統的崛起與其地緣政治影響 摘要 在2035年，全球局勢緊張之際，台灣的量子導航技術成為重塑國際權力格局的關鍵力量。本篇科幻論文式敘事，記錄了國家中山科學研究院（NCSIST）一個小型團隊秘密研發QNF-3系統的歷程。該系統整合原子干涉、量子隨機亂數生成（QRNG）與量子密鑰分發（QKD），超越傳統GPS依賴，在無GPS環境下實現厘米級定位精度。在「去中化」政策推動下，QNF-3促進技術自主，脫離中國大陸北斗導航網的影響，卻意外加劇美國對太平洋權力轉移的恐懼。通過歷史回顧與未來推測，本文探討技術細節、社會政治影響與倫理困境，揭示量子導航霸權的潛力。 引言：量子自主的誕生 台灣量子導航的開端可追溯至2020年代初，受美中貿易戰與兩岸緊張局勢驅動，台灣轉向技術自立。「去中化」（de-Sinicization）政策旨在減少中國大陸的政治、文化與經濟影響，強化本土認同與國際連結。該政策涵蓋教育改革（強調台灣史而非漢族歷史）與經濟脫鉤（降低對中供應鏈依賴），為量子研究奠定基礎。 2025年，聯合國宣布「國際量子科學與技術年」，台灣中央研究院在台北101舉辦量子燈光秀，象徵其量子領域崛起。然而，在台中龍井的中科院實驗室，一場秘密行動悄然展開。六人小組在廢棄倉庫中啟動「量子導航框架」（QNF）計畫，旨在解決GPS易受干擾的弱點。受英國2024年量子PNT試驗（5米/小時精度）的啟發，團隊以原子干涉為核心，採「務實工程導向」，從簡陋環境邁向技術突破。 方法：技術基礎與整合 原子干涉：QNF-3的核心 QNF-3的核心是基於Mach-Zehnder干涉儀（MZI）的原子干涉技術。利用磁光阱冷卻銣-87原子至微開爾文級，系統以π/2雷射脈衝分裂原子波包，沿不同路徑演化後再合併，產生干涉圖案。圖案反映環境變化，如加速度（相位差Δφ ∝ gT²，g為重力加速度，T為演化時間）或旋轉（薩格納克效應，Δφ = 4πΩA/λv，Ω為角速度，A為封閉面積，λ為德布羅意波長，v為原子速度）。 早期QNF-1原型受環境干擾（如颱風振動）影響，精度有限。至2028年，QNF-2採用多迴路設計，有效面積放大100倍，實現米級精度。量子陀螺儀測量角速度達10^{-10} rad/s，結合加速度計形成完整慣性導航，無懼電磁戰。 QRNG與QKD融...
閱讀完整內容

量子化學範式轉變對社會科學的啟示

圖片
當量子化學不再靠做實驗：它如何啟發社會學避開愚蠢實驗，走向計算與演繹模擬 摘要 量子化學的發展標誌著化學研究從傳統歸納法向演繹法的轉變，透過薛丁格方程式從基本原理出發進行計算預測。本文探討此轉變對社會學的啟示，強調計算模擬方法如代理人基礎模型可強化演繹面向，並進一步論述此方法可減少人類在現實中進行愚蠢且不道德的社會實驗，轉而於虛擬環境中測試政策與理論，從而提升倫理標準與預測效能。 上圖顯示薛丁格方程式，其時間獨立形式為量子化學的核心工具，用於演繹分子能量與結構。 引言：量子化學研究範式的轉變 量子化學以薛丁格方程式為基礎，發展從頭計算（ab initio）方法，從物理常數與近似演繹分子性質，避免大量依賴經驗試驗。此演繹轉變不僅提升預測精度，亦減少潛在危險的物理實驗。 上圖為分子軌道圖，展示量子化學如何從原子軌道演繹分子電子結構。 社會學傳統方法與愚蠢社會實驗的隱憂 傳統社會學多採歸納與實地實驗，然歷史上多項大規模社會政策或實驗因缺乏充分預測與倫理考量，導致嚴重後果。例如圖斯基吉梅毒研究（Tuskegee Syphilis Study）違反知情同意與治療原則，造成參與者長期傷害；中國文化大革命（Cultural Revolution）作為激進社會改造，引發廣泛動亂與人力損失；一胎化政策（One-Child Policy）雖旨在控制人口，卻造成性別失衡、人口老化與家庭結構扭曲；嚴格零容忍防疫政策（Zero-COVID Policy）下的長期封控措施，帶來經濟停滯、心理創傷與次生傷害；此外，武漢病毒研究所洩漏假說（Wuhan lab leak theory）相關討論，亦凸顯高風險實驗潛在的不可逆後果。 上圖對比歸納法與演繹法，傳統社會學多從觀察歸納，但演繹模擬可避免現實風險。 上圖為歷史不道德社會實驗的相關圖像，突顯現實實驗的倫理困境。 啟示：計算演繹模擬減少現實愚蠢實驗的必要性 借鑒量子化學的演繹計算，社會學可廣泛採用代理人基礎模型（ABM）與計算社會科學，從基本社會原理（如個體行為規則）出發，於虛擬環境模擬宏觀現象。此方法允許研究者測試極端情境、政策介入或社會變革，而無需在現實中實施潛在危險的實驗。 人類是否可以不用再做愚蠢的社會實驗？計算模擬提供肯定答案：在虛擬空間中重複運行數千次模擬，可預測政策後果、識別風險，並優化方案，從而避免歷史上那些因無知或疏忽而導致的悲劇...
閱讀完整內容