Quantum Action Invariance Principle

• Quantum Action Invariance Principle is a framework where quantum dynamics are encoded by an action functional, ensuring invariance under both continuous and discrete symmetry transformations.
• It provides a unifying approach to derive equations of motion and conservation laws, as demonstrated by Schwinger’s dynamical principle and the TCP/CPT invariance theorem.
• The concept extends to discrete systems like quantum walks, Husimi phase-space dynamics, and cellular automata, offering insights into bridging deterministic models and continuum quantum mechanics.

Searching arXiv for the cited works and closely related literature on quantum action principles and invariance. I’ll look up the specified arXiv papers and nearby work to ground the article in the current arXiv record. Quantum action invariance principle denotes a family of formulations in which quantum dynamics is encoded by an action functional and in which symmetry statements are obtained from the invariance of the action, or of its variation, including boundary terms. In Schwinger’s formulation, the basic dynamical law is the relation

$$\delta\langle \phi_1,\sigma_1 \mid \phi_2,\sigma_2\rangle = i\langle \phi_1,\sigma_1 \mid \delta S \mid \phi_2,\sigma_2\rangle,$$

so that whenever $\delta S$ vanishes or is mapped into itself, the corresponding amplitudes are invariant. In the arXiv literature, this logic appears in several technically distinct settings: interacting quantum field theory and the TCP/CPT theorem, discrete-time quantum walks and quantum automata, Husimi phase-space dynamics, and integer-valued cellular automata mapped to continuum quantum mechanics (Selover et al., 2013, Debbasch, 2018, Zhdanov et al., 2021, Elze, 2013).

1. Formal basis in Schwinger’s dynamical principle

In quantum field theory one defines an action functional

$S[\phi]=\int d^4x\,L(\phi(x),\partial_\mu\phi(x)),$

where $L$ is the operator Lagrangian density. Schwinger’s dynamical principle states that for any infinitesimal variation of the fields and of the end-point surfaces $\sigma_1,\sigma_2$, the change in the transition amplitude between two space-like surfaces is given by

$$\delta\langle \phi_1,\sigma_1 \mid \phi_2,\sigma_2\rangle = i\langle \phi_1,\sigma_1 \mid \delta S \mid \phi_2,\sigma_2\rangle.$$

If $\delta S$ vanishes on-shell, or for a symmetry transformation, the matrix element is invariant (Selover et al., 2013).

Starting from

$S[\phi]=\int d^4x\,L(\phi,\partial_\mu\phi),$

an infinitesimal variation $\phi(x)\to\phi(x)+\delta\phi(x)$ gives

$$\delta S=\int d^4x \left[ \frac{\partial L}{\partial \phi}\,\delta\phi + \frac{\partial L}{\partial(\partial_\mu\phi)}\,\delta(\partial_\mu\phi) \right].$$

After integration by parts,

$\delta S$0

The vanishing of the bulk contribution yields the Euler–Lagrange equations, while the surface terms encode the dependence on the initial and final boundary variations $\delta S$1, which Schwinger keeps explicitly (Selover et al., 2013).

The resulting principle is broader than a stationary-action statement for classical trajectories. It ties equations of motion, commutation relations, and scattering amplitudes to properties of the action functional. This suggests that, in the Schwinger framework, action invariance is not merely a compact reformulation of dynamics but the operative criterion for deriving both local field equations and global symmetry theorems (Selover et al., 2013).

2. Symmetry content: continuous, discrete, and discrete-time formulations

Whenever a continuous one-parameter family of transformations $\delta S$2 leaves the Lagrangian and measure strictly invariant, one has $\delta S$3; by Noether’s theorem this guarantees the existence of a conserved current. The same arXiv corpus extends the action-based logic beyond continuously connected symmetries, treating discrete antiunitary symmetries, discrete Lorentz boosts, and discrete conservation laws within explicit action formalisms (Selover et al., 2013, Debbasch, 2018, Zhdanov et al., 2021, Elze, 2013).

Setting	Action statement	Invariance consequence
Interacting QFT	$\delta S$4	symmetry of amplitudes; TCP invariance
DTQW / quantum automata	$\delta S$5 yields $\delta S$6	conserved charge, energy, momentum; discrete Lorentz boosts
Husimi phase space	$\delta S$7	generalized Hamilton equations; quantum Liouville equation
Integer-valued CA	$\delta S$8	discrete conservation laws; modified Schrödinger equation

A recurrent misconception is that action invariance is restricted to smooth spacetime symmetries. The cited work shows otherwise. In the Schwinger–Sudarshan derivation, TCP is represented by the discrete antiunitary operator $\delta S$9, not by a continuously connected Lie group. In the DTQW setting, the action is exactly invariant under discrete Lorentz boosts generated by the Spin–Operator Lorentz Transformation law. In the Husimi formulation, the action is invariant under the addition of a Skodje flux-gauge term that re-routes phase-space fluxes without changing the quantum Liouville equation. In the cellular-automaton setting, invariance under global unitary transformations generated by matrices commuting with the Hamiltonian yields discrete analogues of norm and energy conservation (Selover et al., 2013, Debbasch, 2018, Zhdanov et al., 2021, Elze, 2013).

3. TCP/CPT as an action-invariance theorem

The paper “Derivation of the TCP Theorem using Action Principles” reformulates the proof of TCP/CPT invariance by applying the Schwinger dynamical relation directly to the variation of the action rather than to free-field mode expansions. The starting point is

$S[\phi]=\int d^4x\,L(\phi(x),\partial_\mu\phi(x)),$0

Applying the discrete map $S[\phi]=\int d^4x\,L(\phi(x),\partial_\mu\phi(x)),$1 to the bra-ket and to $S[\phi]=\int d^4x\,L(\phi(x),\partial_\mu\phi(x)),$2 produces three simultaneous effects: $S[\phi]=\int d^4x\,L(\phi(x),\partial_\mu\phi(x)),$3 reverses the time-ordering of the boundary surfaces $S[\phi]=\int d^4x\,L(\phi(x),\partial_\mu\phi(x)),$4 and flips the overall sign of the action integral; $S[\phi]=\int d^4x\,L(\phi(x),\partial_\mu\phi(x)),$5 flips the sign introduced by $S[\phi]=\int d^4x\,L(\phi(x),\partial_\mu\phi(x)),$6 and sends $S[\phi]=\int d^4x\,L(\phi(x),\partial_\mu\phi(x)),$7; and $S[\phi]=\int d^4x\,L(\phi(x),\partial_\mu\phi(x)),$8 reorders spatial arguments. The sign and argument changes then conspire to return the right-hand side to its original form (Selover et al., 2013).

For a field of spinor or tensor character, the discrete operator acts schematically as

$S[\phi]=\int d^4x\,L(\phi(x),\partial_\mu\phi(x)),$9

where $L$0 implements the spin-parity map, complex conjugation encodes charge conjugation, and $L$1 is a possible phase. If the Lagrangian density and the boundary data are invariant under $L$2, the resulting statement is

$L$3

This is the equality of amplitudes before and after TCP reversal, and therefore the proof of TCP invariance of the full theory (Selover et al., 2013).

The significance of this derivation lies in its scope. The paper states that it extends the TCP theorem beyond the original proofs by Pauli-Luders and Jost, which were constrained by limits of free quantum fields and the asymptotic condition. The new proof is based on invariance of the variation of action with interactions included and no free field or asymptotic condition on the quantum fields; it is therefore stated to apply to more complicated quantum field systems that include $L$4-particle bound states and unstable states (Selover et al., 2013).

4. Discrete quantum automata and discrete-time quantum walks

For a general quantum automaton with Hilbert space $L$5, a time-dependent unitary update operator $L$6 defines the evolution

$L$7

An action whose extremum yields this update is

$L$8

For discrete-time quantum walks on a one-dimensional lattice $L$9, one takes $\sigma_1,\sigma_2$0, with $\sigma_1,\sigma_2$1 the shift operator and $\sigma_1,\sigma_2$2 the local coin operator, so that the DTQW action becomes

$\sigma_1,\sigma_2$3

Independent variations in $\sigma_1,\sigma_2$4 and $\sigma_1,\sigma_2$5 yield the single discrete Euler–Lagrange equation

$\sigma_1,\sigma_2$6

that is, exactly the DTQW update (Debbasch, 2018).

The formulation is then extended by introducing two real scalar fields on the grid,

$\sigma_1,\sigma_2$7

as independent dynamical variables. With discrete derivatives

$\sigma_1,\sigma_2$8

the extended action

$\sigma_1,\sigma_2$9

is constructed so that variation with respect to the discrete gradients of the coordinates yields, on-shell, the local energy and momentum densities $$\delta\langle \phi_1,\sigma_1 \mid \phi_2,\sigma_2\rangle = i\langle \phi_1,\sigma_1 \mid \delta S \mid \phi_2,\sigma_2\rangle.$$0, $$\delta\langle \phi_1,\sigma_1 \mid \phi_2,\sigma_2\rangle = i\langle \phi_1,\sigma_1 \mid \delta S \mid \phi_2,\sigma_2\rangle.$$1 and their currents $$\delta\langle \phi_1,\sigma_1 \mid \phi_2,\sigma_2\rangle = i\langle \phi_1,\sigma_1 \mid \delta S \mid \phi_2,\sigma_2\rangle.$$2, $$\delta\langle \phi_1,\sigma_1 \mid \phi_2,\sigma_2\rangle = i\langle \phi_1,\sigma_1 \mid \delta S \mid \phi_2,\sigma_2\rangle.$$3. The vanishing of the relevant functional derivatives yields the discrete conservation laws

$$\delta\langle \phi_1,\sigma_1 \mid \phi_2,\sigma_2\rangle = i\langle \phi_1,\sigma_1 \mid \delta S \mid \phi_2,\sigma_2\rangle.$$4

(Debbasch, 2018).

The same paper rewrites the extended action in a manifestly covariant form by introducing a discrete 2-bein $$\delta\langle \phi_1,\sigma_1 \mid \phi_2,\sigma_2\rangle = i\langle \phi_1,\sigma_1 \mid \delta S \mid \phi_2,\sigma_2\rangle.$$5 and discrete analogues of Dirac derivatives. The discrete stress-energy tensor is defined by functional differentiation with respect to the gradients of the coordinates viewed as functions of the discrete grid points. In the grid frame $$\delta\langle \phi_1,\sigma_1 \mid \phi_2,\sigma_2\rangle = i\langle \phi_1,\sigma_1 \mid \delta S \mid \phi_2,\sigma_2\rangle.$$6, its components reproduce $$\delta\langle \phi_1,\sigma_1 \mid \phi_2,\sigma_2\rangle = i\langle \phi_1,\sigma_1 \mid \delta S \mid \phi_2,\sigma_2\rangle.$$7, $$\delta\langle \phi_1,\sigma_1 \mid \phi_2,\sigma_2\rangle = i\langle \phi_1,\sigma_1 \mid \delta S \mid \phi_2,\sigma_2\rangle.$$8, $$\delta\langle \phi_1,\sigma_1 \mid \phi_2,\sigma_2\rangle = i\langle \phi_1,\sigma_1 \mid \delta S \mid \phi_2,\sigma_2\rangle.$$9, and $\delta S$0, and by construction it obeys the covariant conservation law

$\delta S$1

The action also admits a global phase invariance $\delta S$2, yielding exactly conserved charge; discrete Lorentz boosts, parametrized by $\delta S$3, via the Spin–Operator Lorentz Transformation law; and space-time translations in $\delta S$4. In the continuous limit $\delta S$5, one recovers the usual $\delta S$6-dimensional Dirac action in Minkowski space, the continuous Dirac equation, and the standard Lorentz group (Debbasch, 2018).

5. Husimi phase-space Hamilton variation and flux-gauge freedom

The Husimi formulation begins with the density operator $\delta S$7 of an $\delta S$8-dimensional quantum system and defines the Husimi function

$\delta S$9

where $S[\phi]=\int d^4x\,L(\phi,\partial_\mu\phi),$0 is a minimum-uncertainty packet. By construction $S[\phi]=\int d^4x\,L(\phi,\partial_\mu\phi),$1, so $S[\phi]=\int d^4x\,L(\phi,\partial_\mu\phi),$2 plays the rôle of a genuine probability density on phase space. The formalism uses the Husimi-product $S[\phi]=\int d^4x\,L(\phi,\partial_\mu\phi),$3, a Moyal-type nonlocal associative product, and the Husimi generator

$S[\phi]=\int d^4x\,L(\phi,\partial_\mu\phi),$4

whose lowest-order approximation is the classical Hamiltonian $S[\phi]=\int d^4x\,L(\phi,\partial_\mu\phi),$5 (Zhdanov et al., 2021).

With an arbitrary Lagrangian label $S[\phi]=\int d^4x\,L(\phi,\partial_\mu\phi),$6 attached to each fluid parcel, the action is

$S[\phi]=\int d^4x\,L(\phi,\partial_\mu\phi),$7

Under admissible variations of the collective trajectories that preserve the purity constraint $S[\phi]=\int d^4x\,L(\phi,\partial_\mu\phi),$8, stationarity yields the generalized Hamilton equations

$S[\phi]=\int d^4x\,L(\phi,\partial_\mu\phi),$9

Substitution into the continuity equation for $\phi(x)\to\phi(x)+\delta\phi(x)$0 gives

$\phi(x)\to\phi(x)+\delta\phi(x)$1

which is exactly the quantum Liouville equation in the Husimi representation (Zhdanov et al., 2021).

A central structural feature is the Skodje flux-gauge freedom. The action may be extended by

$\phi(x)\to\phi(x)+\delta\phi(x)$2

where $\phi(x)\to\phi(x)+\delta\phi(x)$3 is an arbitrary scalar gauge potential and $\phi(x)\to\phi(x)+\delta\phi(x)$4 for an antisymmetric flux-gauge tensor $\phi(x)\to\phi(x)+\delta\phi(x)$5. The new stationary-action conditions modify the phase-space velocities by divergenceless terms, but because $\phi(x)\to\phi(x)+\delta\phi(x)$6 identically, the continuity equation and therefore the quantum Liouville equation are unchanged. The paper characterizes this as a Husimi-phase-space analogue of electromagnetic gauge invariance: the added term re-routes the fluid streams in phase space while leaving the physical evolution invariant (Zhdanov et al., 2021).

The same construction applies to a classical probability distribution convolved with the same Gaussian kernel, yielding a “classical Husimi” action

$\phi(x)\to\phi(x)+\delta\phi(x)$7

The paper states that both classical and quantum dynamics arise from exactly the same Hamilton variation principle, with the differences residing in the generator $\phi(x)\to\phi(x)+\delta\phi(x)$8 versus $\phi(x)\to\phi(x)+\delta\phi(x)$9 and in the purity condition on the state. Applications discussed include constructing semiclassical approximations and hybrid classical-quantum theories (Zhdanov et al., 2021).

6. Integer-valued cellular automata, discrete conservation laws, and the Schrödinger limit

A distinct discrete realization appears in the action principle for integer-valued cellular automata. The basic variables are integer-valued coordinates $$\delta S=\int d^4x \left[ \frac{\partial L}{\partial \phi}\,\delta\phi + \frac{\partial L}{\partial(\partial_\mu\phi)}\,\delta(\partial_\mu\phi) \right].$$0, conjugate momenta $$\delta S=\int d^4x \left[ \frac{\partial L}{\partial \phi}\,\delta\phi + \frac{\partial L}{\partial(\partial_\mu\phi)}\,\delta(\partial_\mu\phi) \right].$$1, a dynamical time variable $$\delta S=\int d^4x \left[ \frac{\partial L}{\partial \phi}\,\delta\phi + \frac{\partial L}{\partial(\partial_\mu\phi)}\,\delta(\partial_\mu\phi) \right].$$2, and its conjugate $$\delta S=\int d^4x \left[ \frac{\partial L}{\partial \phi}\,\delta\phi + \frac{\partial L}{\partial(\partial_\mu\phi)}\,\delta(\partial_\mu\phi) \right].$$3. With finite differences

$$\delta S=\int d^4x \left[ \frac{\partial L}{\partial \phi}\,\delta\phi + \frac{\partial L}{\partial(\partial_\mu\phi)}\,\delta(\partial_\mu\phi) \right].$$4

and

$$\delta S=\int d^4x \left[ \frac{\partial L}{\partial \phi}\,\delta\phi + \frac{\partial L}{\partial(\partial_\mu\phi)}\,\delta(\partial_\mu\phi) \right].$$5

the action is

$$\delta S=\int d^4x \left[ \frac{\partial L}{\partial \phi}\,\delta\phi + \frac{\partial L}{\partial(\partial_\mu\phi)}\,\delta(\partial_\mu\phi) \right].$$6

The dynamical law is the discrete variational principle $$\delta S=\int d^4x \left[ \frac{\partial L}{\partial \phi}\,\delta\phi + \frac{\partial L}{\partial(\partial_\mu\phi)}\,\delta(\partial_\mu\phi) \right].$$7 for arbitrary integer-valued variations, with symmetric discrete variation

$$\delta S=\int d^4x \left[ \frac{\partial L}{\partial \phi}\,\delta\phi + \frac{\partial L}{\partial(\partial_\mu\phi)}\,\delta(\partial_\mu\phi) \right].$$8

(Elze, 2013).

Applying $$\delta S=\int d^4x \left[ \frac{\partial L}{\partial \phi}\,\delta\phi + \frac{\partial L}{\partial(\partial_\mu\phi)}\,\delta(\partial_\mu\phi) \right].$$9 independently to $\delta S$00 yields discrete analogues of Hamilton’s equations. Introducing the complex state-vector components

$\delta S$01

and the self-adjoint Hamiltonian matrix $\delta S$02, the two real equations combine into

$\delta S$03

If $\delta S$04 is any constant matrix commuting with $\delta S$05, then

$\delta S$06

is independent of $\delta S$07. These are discrete analogues of norm and energy conservation, arising from invariance of the action under unitary transformations

$\delta S$08

(Elze, 2013).

The bridge to continuum quantum mechanics is supplied by Shannon’s sampling theorem. Each integer-labeled sequence $\delta S$09 is interpolated into a continuous function $\delta S$10 with

$\delta S$11

Discrete shifts become $\delta S$12, hence

$\delta S$13

For the choice $\delta S$14, the discrete evolution becomes

$\delta S$15

the modified Schrödinger equation. Expanding $\delta S$16 in powers of $\delta S$17 and defining $\delta S$18, the limit $\delta S$19 yields the linear Schrödinger equation

$\delta S$20

(Elze, 2013).

The same mapping yields a spectral bound

$\delta S$21

and converts each discrete Noether law into the standard continuum continuity equation. The paper’s stated conclusion is that the linearity of quantum mechanics is related to the action principle of such cellular automata and its conservation laws to discrete ones. This suggests a specific sense in which action invariance can mediate between integer-valued deterministic dynamics and the standard linear quantum formalism (Elze, 2013).