Reconstruction of Quantum Field Equations from Multi-Branch Hamilton–Jacobi Structures

Abstract

We develop a framework in which quantum field equations are reconstructed from a multi-branch Hamilton–Jacobi structure supplemented by density functionals on configuration space. Each branch is defined by a solution of a functional Hamilton–Jacobi equation, while the associated density encodes measure-theoretic and dynamical information. Under suitable analytical assumptions, we show that the superposition of these branchwise contributions yields a wave functional satisfying the functional Schrödinger equation. The construction is established in a mathematically controlled setting, including finite-dimensional regularizations and a continuum limit under appropriate convergence hypotheses. Explicit examples demonstrate the exact reconstruction of both free and interacting scalar quantum field equations. Extensions to fermionic and gauge fields are outlined, highlighting the role of graded structures and constraints. Beyond the reconstruction result, the framework provides a structural reinterpretation of quantum theory. Classical dynamics is encoded in the space of Hamilton–Jacobi solutions, while quantum phenomena arise from the coherent superposition of these classical configurations through density and phase. From a geometric viewpoint, the construction is naturally related to multisymplectic structures and admits a cohomological interpretation. This approach establishes a direct bridge between classical variational principles and quantum field theory, offering a unified perspective on the emergence of quantum behavior from structured ensembles of classical configurations.

1. Introduction

The relation between classical and quantum physics remains one of the central structural questions in theoretical physics. Classical field theory admits a variational formulation under which the dynamics can be encoded in Hamilton–Jacobi structures [1,2], whereas quantum theory is formulated in terms of linear evolution equations on Hilbert spaces [3,4]. Although these two descriptions are formally different, several well-established approaches, including semiclassical methods, hydrodynamic formulations, and path-integral techniques [5,6], show that classical variational data continue to play a fundamental role in quantum theory.

A recurrent feature of these approaches is the appearance of a phase determined by an action functional together with an amplitude or density carrying information that is not contained in the classical Hamilton–Jacobi equation alone. In particular, the polar decomposition of a wave functional leads formally to a Hamilton–Jacobi equation supplemented by a quantum potential [7,8,9]. This suggests that quantum dynamics may be viewed, at least in suitable representations, as a coupling between classical variational structure and additional measure-theoretic or density data on configuration space.

The purpose of the present work is to develop this idea at the level of quantum field theory. We consider a family of Hamilton–Jacobi branches, each equipped with a density functional; more precisely, we introduce a finite or countable family of action functionals

$${\left\{S_k\left[\varphi \right]\right\}}_{k\in K},$$

together with a corresponding family of non-negative density functionals

$${\left\{{\rho }_k\left[\varphi \right]\right\}}_{k\in K}.$$

The associated wave functional is represented in the form

$$\mathsf{\Psi }\left[\varphi \right]=\sum _{k\in K}\sqrt{{\rho }_k\left[\varphi \right]}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k\left[\varphi \right]\right).$$

The main result of the paper is a reconstruction statement: under precise regularity, convergence, and domain assumptions, this multi-branch Hamilton–Jacobi–density representation yields a solution of the functional Schrödinger equation.

Specific Contribution of the Present Work

The novelty claimed here does not lie in the Hamilton–Jacobi formalism itself, nor in the polar decomposition of a wave function; these are classical and well-established tools. Rather, the specific contribution of the present work is to organize quantum field states as admissible multi-branch Hamilton–Jacobi–density data. In this representation, the branches are not arbitrary components of a formal expansion; each branch carries Hamilton–Jacobi phase data, each density is constrained by a continuity equation, and the reconstructed wave functional is required to belong to the domain of the quantum Hamiltonian.

Therefore, the reconstruction theorem is not a restatement of the ordinary Bohm–Madelung rewriting of a single wave functional. The Bohm–Madelung formalism rewrites one wave function in terms of one phase and one density, while the present construction instead addresses a family of branch components, each with its own polar Hamilton–Jacobi–density system, then reconstructs the full quantum field state from their linear superposition. The additional content is the admissibility problem for a multi-branch family: the branches must satisfy their own polar equations, the series must converge in the appropriate Hilbert space, and the reconstructed state must lie in the domain of the quantum Hamiltonian.

Equivalently, the theorem identifies conditions under which a structured family of local Hamilton–Jacobi data gives rise to a global solution of the functional Schrödinger equation. This formulation also makes explicit the finite-dimensional regularization required before passing to the continuum limit, and connects the resulting branch structure with multisymplectic geometry and cohomological gluing data.

It is important to emphasize from the outset that the branch decomposition is not claimed to be unique. A quantum state may admit different admissible Hamilton–Jacobi–density representations. The relevant point is not uniqueness, but admissibility. The branches considered here are constrained by Hamilton–Jacobi equations, by associated continuity equations for the densities, and by the requirement that their superposition belong to the domain of the quantum Hamiltonian. In this sense, the construction should be understood as a constrained geometric representation of quantum states rather than as a canonical decomposition.

The physical origin of the branches is the non-uniqueness of Hamilton–Jacobi solutions in field theory. Multiple branches may arise from distinct extremal configurations, caustics, nontrivial boundary conditions, gauge sectors, or global obstructions to defining a single-valued action functional. Geometrically, each branch may be interpreted as defining, at least locally, a Lagrangian submanifold in the corresponding covariant or multisymplectic phase space [10,11,12,13]. The density functionals then provide a measure-theoretic structure on the family of such branches.

The present construction is closely related in spirit to approaches in which quantum states are reconstructed from classical action data [14]. Its specific feature is that it is formulated at the level of field theory and incorporates density functionals as part of the reconstruction data. The action functionals encode admissible classical variational structures, while the densities determine how those structures are weighted and coherently combined in the quantum state.

A central point of the paper is the role of the quantum potential. Rather than introducing a quantum potential by applying a nonlinear operation to the full multi-branch superposition, we treat each branch component

$${\psi }_k\left[\varphi \right]=\sqrt{{\rho }_k\left[\varphi \right]}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k\left[\varphi \right]\right)$$

through its own polar decomposition. Therefore, the branchwise quantum potential is the standard polar quantum potential associated with the individual component ${\psi }_k$. Since the functional Schrödinger equation is linear, a convergent sum of such branchwise solutions is again a solution, provided that the relevant analytical operations are justified.

The infinite-dimensional character of quantum field theory requires particular care. For this reason, the reconstruction is first formulated in finite-dimensional regularizations, for example by lattice approximation, spectral cutoff, or projection onto finitely many modes. Only after this finite-dimensional construction has been established do we discuss the continuum limit. The continuum statement is conditional on the convergence of the regularized states, the convergence of the corresponding Hamiltonians, and the validity of the required interchange of limits, summation, differentiation, and functional differentiation. In interacting theories, this also involves the usual regularization and renormalization issues associated with products of functional derivatives at the same point.

The framework also admits a natural geometric and cohomological interpretation. In general, the Hamilton–Jacobi action may fail to exist globally as a single-valued functional. The multi-branch structure can then be viewed as a collection of local variational data subject to nontrivial gluing conditions, with the differences between local branches encoding cohomological information [15,16,17]. From this perspective, the wave functional is a global object reconstructed from local Hamilton–Jacobi data and density functionals.

Within this interpretation, interference arises from phase relations between different branches, while decoherence may be described as the effective loss of phase coherence due to interactions with additional degrees of freedom [18,19,20]. Wavefunction reduction may likewise be interpreted at a structural level as a conditioning of the measure on the space of admissible branches. These interpretations are not used as additional postulates in the reconstruction theorem, but do indicate how the formalism may be connected to standard foundational concepts.

The present paper establishes the reconstruction mechanism for scalar quantum field theory, then outlines how the same structure can be extended formally to gauge and fermionic fields. These extensions only indicate how constraints, gauge invariance, Ward or Slavnov–Taylor identities, BRST compatibility, and graded structures enter the multi-branch Hamilton–Jacobi–density framework, and are not presented here as complete rigorous constructions.

The gravitational extension is also treated as formal and conditional. In this setting, one must address the construction of measures on superspace, the implementation of diffeomorphism invariance, the Hamiltonian constraint and the problem of time, and the regularization of the gravitational Hamiltonian. These questions are not solved in the present paper, but are stated as necessary conditions for a fully rigorous version of the proposed framework.

The rest of the paper is organized as follows: in Section 2, we review the covariant Hamilton–Jacobi formalism for classical field theory [10,11]; in Section 3, we introduce the multi-branch Hamilton–Jacobi structure and the associated density functionals, including their admissibility and non-uniqueness; Section 4 develops the foundational interpretation of this structure; Section 5 formulates the reconstruction of quantum field equations, first in a regularized setting and then under suitable continuum-limit assumptions; Section 6 presents an illustrative free-scalar-field example and its equivalence with the standard formulation; Section 7 gives the geometric and multisymplectic interpretation; Section 8 discusses the extension toward quantum gravity and the associated cohomological structure; and Section 9 summarizes the conclusions.

2. Covariant Hamilton–Jacobi Formalism for Classical Field Theory

2.1. Lagrangian Field Theory

Let M be an n-dimensional spacetime manifold with local coordinates $x^{\mu }$, $\mu =1,\dots ,n$ and let $\varphi :M\to \mathcal{F}$ be a classical field taking values in a configuration space $\mathcal{F}$. The following Lagrangian density determines the dynamics of the field:

$$\mathcal{L}=\mathcal{L}(x^{\mu },\varphi ,{\partial }_{\mu }\varphi ).$$

The corresponding action functional is

$$\mathcal{S}\left[\varphi \right]={\int }_M\mathcal{L}(x^{\mu },\varphi ,{\partial }_{\mu }\varphi )d^{n}x.$$

The Euler–Lagrange equations follow from the stationary action principle, and are given by

$${\displaystyle \frac{\partial \mathcal{L}}{\partial \varphi }}-{\partial }_{\mu }\left({\displaystyle \frac{\partial \mathcal{L}}{\partial \left({\partial }_{\mu }\varphi \right)}}\right)=0.$$

2.2. De Donder–Weyl Hamiltonian Formulation

A covariant Hamiltonian formulation is obtained by introducing the polymomenta [13,21]

$${\pi }^{\mu }:={\displaystyle \frac{\partial \mathcal{L}}{\partial \left({\partial }_{\mu }\varphi \right)}}.$$

The De Donder–Weyl Hamiltonian density is defined by the covariant Legendre transform

$$\mathcal{H}(x^{\mu },\varphi ,{\pi }^{\mu })={\pi }^{\mu }{\partial }_{\mu }\varphi -\mathcal{L}.$$

The field equations can then be written in Hamiltonian form as

$${\partial }_{\mu }\varphi ={\displaystyle \frac{\partial \mathcal{H}}{\partial {\pi }^{\mu }}},{\partial }_{\mu }{\pi }^{\mu }=-{\displaystyle \frac{\partial \mathcal{H}}{\partial \varphi }}.$$

This formulation treats space and time variables on the same footing, and is naturally adapted to relativistic field theories.

2.3. Covariant Hamilton–Jacobi Equation

The covariant Hamilton–Jacobi formalism seeks functions $S^{\mu }(x,\varphi )$, often called Hamilton–Jacobi functions, such that the polymomenta are expressed as

$${\pi }^{\mu }={\displaystyle \frac{\partial S^{\mu }}{\partial \varphi }}.$$

Substitution into the De Donder–Weyl Hamiltonian gives the covariant Hamilton–Jacobi equation

$${\partial }_{\mu }{S}^{\mu }(x,\varphi )+\mathcal{H}\left(x^{\mu },\varphi ,{\displaystyle \frac{\partial S^{\mu }}{\partial \varphi }}\right)=0.$$

This equation generalizes the classical Hamilton–Jacobi equation to field theory and encodes the dynamics in terms of first-order partial differential equations [21,22].

2.4. Geometric Interpretation

The covariant Hamilton–Jacobi structure admits a natural geometric interpretation within multisymplectic geometry [10,11,12,13]. The polymomenta ${\pi }^{\mu }$ define a section of the polymomentum phase space, while the functions $S^{\mu }$ generate, at least locally, a Lagrangian submanifold in this space. The Hamilton–Jacobi equation expresses the condition that this submanifold is compatible with the covariant Hamiltonian flow.

In this setting, solutions of the field equations correspond to integral manifolds with tangent spaces determined by the derivatives of the Hamilton–Jacobi functions. Therefore, classical dynamics can be viewed as arising from generating functions $S^{\mu }$ which encode the extremal properties of the action.

2.5. Limitations of the Single-Branch Structure

In the standard formulation, the Hamilton–Jacobi functions $S^{\mu }$ are often assumed to be single-valued and sufficiently smooth. This assumption corresponds to the local classical picture in which a prescribed set of boundary or initial data determines a single extremal configuration of the field.

However, this single-branch description is insufficient in many physically relevant situations. Multiple extremal configurations may coexist, for example in the presence of constraints, caustics, nontrivial boundary conditions, or global topological obstructions. From the variational point of view, this means that the classical action may naturally be multi-valued or defined only locally on configuration space.

This observation becomes particularly relevant when one attempts to relate the Hamilton–Jacobi formalism to quantum theory. Quantum phenomena such as interference and tunneling indicate that several classical configurations may contribute to physical amplitudes simultaneously. In the path-integral formulation, this feature appears as a summation over histories [5,6].

Motivated by these considerations, the next section generalizes the covariant Hamilton–Jacobi framework to a multi-branch structure. In that setting, a family of action functionals together with their associated density functionals provides the data from which the quantum field equations are reconstructed under suitable analytical assumptions.

3. Multi-Branch Hamilton–Jacobi Structure and Density Functionals

3.1. Configuration Space and Functional Setting

Let M be a smooth n-dimensional spacetime manifold and let $\mathcal{F}$ denote the space of field configurations. In order to give a precise mathematical meaning to functionals on $\mathcal{F}$, we assume that $\mathcal{F}$ is modeled as a separable Banach or Hilbert space. Typical examples are

$$\mathcal{F}=L^2\left(M\right)\mathrm{or}\mathcal{F}=H^s\left(M\right),$$

depending on the regularity required by the theory [23,24].

Observables and state functionals are defined on $\mathcal{F}$. Derivatives with respect to the field $\varphi$ are understood in the sense of Fréchet or Gâteaux derivatives whenever these derivatives exist. In infinite-dimensional field theory, all differentiability statements should be interpreted either on a specified functional domain or through a finite-dimensional regularization.

3.2. Multi-Branch Hamilton–Jacobi Structure

We introduce a finite or countable family of Hamilton–Jacobi functionals

$${\left\{S_k\left[\varphi \right]\right\}}_{k\in \mathcal{K}},$$

where each

$$S_k:\mathcal{F}⟶\mathbb{R}$$

is assumed to be sufficiently regular. Each branch is required to satisfy a functional Hamilton–Jacobi equation of the form

$${\displaystyle \frac{\partial S_k}{\partial t}}+H\left[\varphi ,{\displaystyle \frac{\delta S_k}{\delta \varphi }}\right]=0.$$

This equation may be viewed as an infinite-dimensional extension of the classical Hamilton–Jacobi equation [1,2]. In the present framework, the index k labels distinct admissible Hamilton–Jacobi branches. These branches may arise from multiple extremal configurations, caustics, boundary conditions, gauge sectors, or global obstructions to the existence of a single-valued action functional.

It is important to emphasize that the branch decomposition is not assumed to be unique. The relevant requirement is admissibility: each branch must satisfy the Hamilton–Jacobi equation, and the full family must satisfy the analytical and compatibility conditions stated below.

3.3. Existence, Non-Uniqueness, and Physical Meaning of Branch Decompositions

The existence of a multi-branch Hamilton–Jacobi representation is not automatic. In finite-dimensional mechanics, Hamilton–Jacobi branches may be constructed locally by the method of characteristics, at least before the formation of caustics. In field theory, the situation is more delicate because the configuration space is infinite-dimensional and because the Hamiltonian functional may require regularization. For this reason, the existence of the branch family should first be understood in a regularized setting, for example after a lattice approximation, a spectral cutoff, or a projection onto finitely many modes.

In such a regularized setting, one may obtain branches from local solutions of the Hamilton–Jacobi equation, from different boundary conditions, from different extremal configurations, or from distinct sectors of the configuration space. Passing to the continuum requires additional assumptions: convergence of the branch data, convergence of the reconstructed wave functional, and stability of the relevant operator domains. These assumptions are part of the reconstruction framework, not automatic consequences of the formalism.

The decomposition is also not unique. A given wave functional may admit several admissible representations of the form

$$\mathsf{\Psi }\left[\varphi \right]=\sum _{k\in \mathcal{K}}\sqrt{{\rho }_k\left[\varphi \right]}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k\left[\varphi \right]\right).$$

Thus, the aim is not to construct a canonical or unique expansion of $\mathsf{\Psi }$ but to identify conditions under which a proposed family of Hamilton–Jacobi phases and densities is admissible, meaning that it reconstructs a wave functional satisfying the required quantum evolution equation.

Hence, the physical meaning of the branches is constrained rather than arbitrary: each branch represents a possible local variational structure, for example a classical extremal history, a caustic branch, a sector determined by boundary conditions, or a local action functional defined on a chart of configuration space. The density ${\rho }_k$ specifies how this branch contributes to the reconstructed quantum state. In this sense, the multi-branch representation is a structured Hamilton–Jacobi–density representation, not an arbitrary decomposition of a wave functional.

3.4. Measures on Configuration Space

In order to rigorously define density functionals, we introduce a reference measure $\mu$ on $\mathcal{F}$. In infinite-dimensional spaces, there is in general, no nontrivial translation-invariant measure. Therefore, the reference measure must either be specified as a Gaussian measure or obtained through cylindrical or finite-dimensional approximations [24,25].

A typical formal expression for a Gaussian measure associated with a free-field theory is

$$d\mu \left(\varphi \right)=Z^{-1}\exp \left(-⟨\varphi ,A\varphi ⟩\right)\mathcal{D}\varphi ,$$

where A is a positive self-adjoint operator, for instance $A=-\Delta +m^2$. More generally, $\mu$ is assumed to be a $\sigma$-finite measure on $(\mathcal{F},\mathcal{B})$, where $\mathcal{B}$ denotes the Borel $\sigma$-algebra.

3.5. Density Functionals as Radon–Nikodym Derivatives

For each branch, we define a probability measure ${\mu }_k$ on $\mathcal{F}$ such that

$${\mu }_k\ll \mu ,$$

that is, ${\mu }_k$ is absolutely continuous with respect to $\mu$. The density functional ${\rho }_k$ is then defined as the Radon–Nikodym derivative

$${\rho }_k\left[\varphi \right]={\displaystyle \frac{d{\mu }_k}{d\mu }}\left(\varphi \right).$$

The normalization condition is

$$\sum _{k\in \mathcal{K}}{\int }_{\mathcal{F}}{\rho }_k\left[\varphi \right]d\mu \left(\varphi \right)=1.$$

This normalization may also be imposed after reconstruction of the total state, depending on whether the branch densities are treated as individually normalized components or as weights in a global decomposition.

3.6. Continuity Equation in Infinite Dimensions

The evolution of the densities is governed formally by a continuity equation on $\mathcal{F}$. Let $v_k\left[\varphi \right]$ denote the functional velocity field associated with the branch $S_k$:

$$v_k\left[\varphi \right]={\displaystyle \frac{\delta S_k}{\delta \varphi }}.$$

Then, the density ${\rho }_k$ is required to satisfy

$${\displaystyle \frac{\partial {\rho }_k}{\partial t}}+{\mathrm{div}}_{\mu }\left({\rho }_k{v}_k\right)=0.$$

This equation generalizes the continuity equation appearing in hydrodynamic formulations of quantum mechanics [7,8,9].

In the Gaussian setting, the divergence ${\mathrm{div}}_{\mu }$ can be defined by integration by parts. For suitable test functionals F, one has

$${\int }_{\mathcal{F}}⟨\nabla F,v_k⟩d\mu =-{\int }_{\mathcal{F}}F{\mathrm{div}}_{\mu }\left(v_k\right)d\mu .$$

This identity should be understood in the domain on which the corresponding integration-by-parts formula is valid.

3.7. Reconstruction of the Quantum State Functional

The quantum state functional is represented as

$$\mathsf{\Psi }\left[\varphi \right]=\sum _{k\in \mathcal{K}}\sqrt{{\rho }_k\left[\varphi \right]}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k\left[\varphi \right]\right).$$

Under appropriate integrability and convergence assumptions, $\mathsf{\Psi }$ belongs to $L^2(\mathcal{F},d\mu )$, meaning that it defines a quantum state in the Hilbert space of square-integrable functionals [24,25].

The reconstruction formula should be read as an admissibility condition. It is not assumed that every formal choice of $(S_k,{\rho }_k)$ defines a physical state. The branch data define a valid representation only when the sum converges, when the required differentiations are justified, and when the reconstructed functional satisfies the relevant quantum evolution equation.

3.8. Analytical Conditions

To ensure that the preceding construction is well-defined, we impose the following conditions:

• $S_k\in C^1\left(\mathcal{F}\right)$, with $\delta S_k/\delta \varphi$ belonging to the appropriate $L^2$-space with respect to $\mu$.
• ${\rho }_k\in L^1(\mathcal{F},\mu )$ and ${\rho }_k\ge 0$.
• $\sqrt{{\rho }_k}\in L^2(\mathcal{F},\mu )$.
• The series defining $\mathsf{\Psi }$ converges in $L^2(\mathcal{F},\mu )$.

These conditions ensure that the multi-branch construction defines a well-behaved element of the quantum Hilbert space. Stronger assumptions will be required in the reconstruction theorem, where termwise differentiation and the action of the Hamiltonian operator must also be justified.

In the continuum theory, these assumptions must be supplemented by regularization and renormalization conditions. In particular, products of functional derivatives at the same point are generally ill-defined without a regulator. This makes the continuum reconstruction conditional on the existence of a limiting Hilbert space, a limiting Hamiltonian, and a domain on which the branchwise equations and reconstructed Schrödinger equation are meaningful.

3.9. Interpretation

In this framework, a quantum state is represented as a superposition of Hamilton–Jacobi branches weighted by density functionals on configuration space. The action functionals $S_k$ encode admissible classical variational structures, while the densities ${\rho }_k$ specify how these structures are distributed and coherently combined.

This construction should not be interpreted as replacing the full path integral by an arbitrary finite or countable sum; rather, it provides a constrained Hamilton–Jacobi–density representation of quantum states. The branch data are admissible only when they satisfy the branchwise equations and when the reconstructed wave functional satisfies the required analytical conditions.

In the next section, we formulate the reconstruction theorem. We first explain the relation with the polar decomposition of a single wave functional, then show that under suitable assumptions, a convergent sum of branchwise solutions yields a solution of the functional Schrödinger equation associated with the Hamiltonian H.

4. Foundational Interpretation of the Multi-Branch Hamilton–Jacobi–Density Structure

The multi-branch Hamilton–Jacobi–density construction developed in this work is not just a technical reconstruction device; it also suggests a conceptual framework in which quantum dynamics is represented by a structured family of classical variational data supplemented by density or measure-theoretic information. In this section, we clarify the mathematical status and physical interpretation of the fundamental objects $(S_k,{\rho }_k)$ and address several conceptual points concerning their role.

The construction is based on a family of action functionals $S_k$ and associated density functionals ${\rho }_k$. The quantum state is represented as

$$\mathsf{\Psi }\left[\varphi \right]=\sum _{k\in \mathcal{K}}\sqrt{{\rho }_k\left[\varphi \right]}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k\left[\varphi \right]\right).$$

This expression should not be interpreted as a dynamical process in which a physical system evaluates or selects among alternative trajectories but as defining a representation of the wave functional in terms of Hamilton–Jacobi branches and density data. As such, the multi-branch decomposition is structural, not algorithmic or teleological.

Each functional $S_k$ is an admissible solution of a Hamilton–Jacobi equation of the form

$${\partial }_t{S}_k+H\left[\varphi ,{\displaystyle \frac{\delta S_k}{\delta \varphi }}\right]=0,$$

or of its quantum-corrected version when the reconstruction theorem is formulated branchwise. The multiplicity of branches reflects the non-uniqueness of Hamilton–Jacobi solutions. Because of distinct extremal configurations, caustics, constraints, or global topological features of configuration space, several solutions may exist for given boundary or initial data.

The density functionals ${\rho }_k$ are constrained by a continuity equation

$${\partial }_t{\rho }_k+{\mathrm{div}}_{\mu }\left({\rho }_k{v}_k\right)=0,v_k={\displaystyle \frac{\delta S_k}{\delta \varphi }}$$

that expresses the transport of density along the flow generated by $S_k$, with the divergence understood with respect to the reference measure $\mu$. However, this continuity equation alone does not uniquely determine ${\rho }_k$; a decisive additional condition is the global compatibility requirement that the reconstructed wave functional $\mathsf{\Psi }$ satisfy the functional Schrödinger equation. Thus, the densities are not arbitrary weights but form part of an admissible family of data for which the superposition must reproduce the quantum dynamics.

A useful way to organize the branch data is through a variational principle defined on the pairs $(S_k,{\rho }_k)$. Formally, one may consider an action of the type

$${\mathcal{A}}_{\mathrm{MB}}=\sum _{k\in \mathcal{K}}\int dt{\int }_{\mathcal{F}}{\rho }_k\left({\partial }_t{S}_k+H\left[\varphi ,{\displaystyle \frac{\delta S_k}{\delta \varphi }}\right]\right)d\mu +{\displaystyle \frac{{\hslash }^2}{8}}\sum _{k\in \mathcal{K}}\int dt{\int }_{\mathcal{F}}{\displaystyle \frac{{\left|{\nabla }_{\mathcal{F}}{\rho }_k\right|}^2}{{\rho }_k}}d\mu ,$$

subject to positivity and normalization conditions on the densities. The second term is of Fisher-information type, and in the polar formulation is responsible for the appearance of a quantum-potential contribution.

Variation with respect to $S_k$ gives the continuity equation. Variation with respect to ${\rho }_k$, in the appropriate formal setting, gives a Hamilton–Jacobi equation corrected by a term with the following form:

$$Q\left[{\rho }_k\right]=-{\displaystyle \frac{{\hslash }^2}{2R_k}}\int {\displaystyle \frac{{\delta }^2{R}_k}{\delta {\varphi }^2}},R_k=\sqrt{{\rho }_k}.$$

In field-theoretic applications, this expression must be understood in a regularized sense. More explicitly, for a scalar field on a spatial hypersurface $\sum$, the corresponding formal quantum potential is

$$Q\left[{\rho }_k\right]=-{\displaystyle \frac{{\hslash }^2}{2R_k}}{\int }_{\sum }{\displaystyle \frac{{\delta }^2{R}_k}{\delta \varphi {\left(x\right)}^2}}{d}^{d}x,R_k=\sqrt{{\rho }_k}.$$

This is the usual quantum potential appearing in polar, hydrodynamic, and Bohmian formulations of quantum theory [7,8,9]. In the present framework, it is attached to each individual branch component, not to the full multi-branch superposition as a nonlinear operation.

The phase factor associated with a branch,

$$\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k\right),$$

is determined by the corresponding action functional. It encodes the accumulated variational information associated with that branch. Interference arises from the relative phases between different branches. Formally, for a finite sum, one obtains

$$|\mathsf{\Psi }{|}^2=\sum _k{\rho }_k+2\sum _{k<\ell }\sqrt{{\rho }_k{\rho }_{\mathcal{l}}}\cos \left({\displaystyle \frac{S_k-S_{\mathcal{l}}}{\hslash }}\right),$$

which explicitly displays the interference terms produced by phase differences between branches.

An explicit two-branch illustration of this interference mechanism is given in Section 5.10.

The framework admits a natural geometric interpretation. Each solution $S_k$ defines, at least locally, a Lagrangian submanifold in the corresponding multisymplectic or covariant Hamiltonian phase space. The densities ${\rho }_k$ define a measure-theoretic structure on the family of such branches. A quantum field state is represented as a measure-supported family of classical variational geometries.

Within this interpretation, interference arises from phase relations between branches, while decoherence corresponds to an effective loss of phase coherence induced by interactions with additional degrees of freedom [18,19,20]. Wavefunction reduction may be interpreted at a structural level as a conditioning of the measure on the space of admissible branches. These statements are interpretative consequences of the representation, not additional assumptions in the reconstruction theorem.

The classical limit is recovered when the relevant action scales are large compared with ℏ. In that regime, rapid phase oscillations suppress interference terms except near stationary configurations. A restricted set of branches then dominates the effective dynamics, and classical behavior emerges as an approximation.

Foundational Intuition

The respective roles of the action functionals $S_k$ and density functionals ${\rho }_k$ can be summarized as follows: the family $\left\{S_k\right\}$ encodes admissible classical variational structures, while the family $\left\{{\rho }_k\right\}$ defines how these structures are weighted, transported, and coherently combined.

Quantum theory does not simply select one classical configuration; rather, in this representation, it equips the space of admissible classical Hamilton–Jacobi data with an additional density structure. The action functionals determine which classical variational evolutions are admissible, while the density functionals determine how these possibilities contribute to the reconstructed quantum state.

This interpretation also admits a cohomological formulation. In general, the Hamilton–Jacobi action need not be globally definable as a single-valued functional. Instead, it may exist only locally. The collection of local branches $\left\{S_k\right\}$ can then be viewed as co-chain data, while the differences $S_k-S_{\mathcal{l}}$ encode transition information between local descriptions. Nontriviality of these gluing data reflects an obstruction to a global single-branch classical description [15,16].

The density functionals ${\rho }_k$ provide a measure-theoretic organization of these local variational data and contribute to the reconstruction of the global wave functional. In this sense, $\mathsf{\Psi }$ appears as a global object reconstructed from local Hamilton–Jacobi branches and density functionals, subject to compatibility conditions.

This viewpoint suggests that the passage from classical to quantum theory can be understood as the enrichment of classical Hamilton–Jacobi dynamics by nontrivial measure and gluing structures on configuration space. Classical dynamics corresponds to the regime in which a single branch or a globally coherent branch gives an adequate description, while quantum dynamics corresponds to the regime in which the multi-branch structure and its associated density data become essential.

5. Reconstruction of Quantum Field Equations

5.1. Functional Schrödinger Representation

Let $\mathcal{F}$ be the configuration space of fields, equipped with a reference measure $\mu$, and let

$${\mathsf{\Psi }}_t\left[\varphi \right]\in L^2(\mathcal{F},d\mu )$$

be a time-dependent wave functional. The functional Schrödinger representation is the canonical representation in which the field $\varphi \left(x\right)$ acts by multiplication, and its conjugate momentum is represented formally by a functional derivative [3,4,25,26,27].

For a scalar field on a spatial hypersurface $\sum$, consider the classical Hamiltonian functional

$$H[\varphi ,\pi ]={\int }_{\sum }\left({\displaystyle \frac{1}{2}}\pi {\left(x\right)}^2+{\displaystyle \frac{1}{2}}{|\nabla \varphi \left(x\right)|}^2+V\left(\varphi \left(x\right)\right)\right)d^{d}x.$$

Canonical quantization formally gives

$$\pi \left(x\right)⟼-i\hslash {\displaystyle \frac{\delta }{\delta \varphi \left(x\right)}}.$$

Thus, the functional Schrödinger equation takes the form

$$i\hslash {\displaystyle \frac{\partial {\mathsf{\Psi }}_t\left[\varphi \right]}{\partial t}}=\widehat{H}{\mathsf{\Psi }}_t\left[\varphi \right],$$

where

$$\widehat{H}={\int }_{\sum }\left[-{\displaystyle \frac{{\hslash }^2}{2}}{\displaystyle \frac{{\delta }^2}{\delta \varphi {\left(x\right)}^2}}+{\displaystyle \frac{1}{2}}{|\nabla \varphi \left(x\right)|}^2+V\left(\varphi \left(x\right)\right)\right]d^{d}x.$$

This expression is formal in the continuum. Products of functional derivatives at the same point require regularization; in interacting theories, they generally require renormalization [23,25,28]. For this reason, the reconstruction theorem below should first be understood in a finite-dimensional regularization, and only afterwards in the continuum limit under explicit convergence assumptions.

5.2. Polar Decomposition of a Single Wave Functional

We first recall the standard single-branch polar decomposition. Suppose that a nonzero wave functional can be written as

$${\mathsf{\Psi }}_t\left[\varphi \right]=R_t\left[\varphi \right]\exp \left({\displaystyle \frac{i}{\hslash }}{S}_t\left[\varphi \right]\right),R_t\left[\varphi \right]=\sqrt{{\rho }_t\left[\varphi \right]}.$$

Substitution into the functional Schrödinger equation and separation of real and imaginary parts gives the functional continuity equation

$${\displaystyle \frac{\partial {\rho }_t}{\partial t}}+{\int }_{\sum }{\displaystyle \frac{\delta }{\delta \varphi \left(x\right)}}\left({\rho }_t{\displaystyle \frac{\delta S_t}{\delta \varphi \left(x\right)}}\right)d^{d}x=0$$

and the quantum Hamilton–Jacobi equation

$${\displaystyle \frac{\partial S_t}{\partial t}}+H\left[\varphi ,{\displaystyle \frac{\delta S_t}{\delta \varphi }}\right]+Q\left[{\rho }_t\right]=0.$$

The quantum potential is

$$Q\left[{\rho }_t\right]=-{\displaystyle \frac{{\hslash }^2}{2R_t}}{\int }_{\sum }{\displaystyle \frac{{\delta }^2{R}_t}{\delta \varphi {\left(x\right)}^2}}{d}^{d}x.$$

Thus, at least formally, the Schrödinger equation is equivalent to a continuity equation coupled to a Hamilton–Jacobi equation corrected by the quantum potential [7,8,9].

5.3. Branchwise Quantum Hamilton–Jacobi System

We now consider a finite or countable family of branches. For each $k\in \mathcal{K}$, let

$$S_k[\varphi ,t]\mathrm{and}{\rho }_k[\varphi ,t]\ge 0$$

be sufficiently regular branch data and define

$${\psi }_k[\varphi ,t]=\sqrt{{\rho }_k[\varphi ,t]}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k[\varphi ,t]\right).$$

The full wave functional is represented as

$$\mathsf{\Psi }[\varphi ,t]=\sum _{k\in \mathcal{K}}{\psi }_k[\varphi ,t].$$

For each branch, we impose the polar system

$${\displaystyle \frac{\partial {\rho }_k}{\partial t}}+{\int }_{\sum }{\displaystyle \frac{\delta }{\delta \varphi \left(x\right)}}\left({\rho }_k{\displaystyle \frac{\delta S_k}{\delta \varphi \left(x\right)}}\right)d^{d}x=0$$

and

$${\displaystyle \frac{\partial S_k}{\partial t}}+H\left[\varphi ,{\displaystyle \frac{\delta S_k}{\delta \varphi }}\right]+Q\left[{\rho }_k\right]=0,$$

where

$$Q\left[{\rho }_k\right]=-{\displaystyle \frac{{\hslash }^2}{2R_k}}{\int }_{\sum }{\displaystyle \frac{{\delta }^2{R}_k}{\delta \varphi {\left(x\right)}^2}}{d}^{d}x,R_k=\sqrt{{\rho }_k}.$$

The first equation transports the density along the velocity field

$$v_k(x;\varphi )={\displaystyle \frac{\delta S_k}{\delta \varphi \left(x\right)}}.$$

The second equation is a quantum-corrected Hamilton–Jacobi equation for the individual branch.

It is important that the quantum potential is introduced branchwise only after applying the polar decomposition to each component ${\psi }_k$, not by applying a nonlinear polar decomposition to the full sum $\mathsf{\Psi }={\sum }_k{\psi }_k$.

5.4. Finite-Dimensional Regularization

To make the reconstruction precise, we first introduce a finite-dimensional regularization. Let ${\mathcal{F}}_N$ be a finite-dimensional approximation of $\mathcal{F}$, obtained, for example, by a lattice regularization, spectral cutoff, or projection onto finitely many modes. Let

$${\widehat{H}}_N$$

be the corresponding regularized Hamiltonian acting on

$$L^2({\mathcal{F}}_N,d{\mu }_N).$$

For each N, define branch data

$${\{S_{k,N},{\rho }_{k,N}\}}_{k\in {\mathcal{K}}_N}$$

and the regularized wave functional

$${\mathsf{\Psi }}_N=\sum _{k\in {\mathcal{K}}_N}\sqrt{{\rho }_{k,N}}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_{k,N}\right).$$

At this level, all derivatives are ordinary derivatives on a finite-dimensional space, and the polar decomposition gives a well-defined system of equations provided that the amplitudes and phases are sufficiently regular.

The continuum statement is obtained only under the additional assumptions

$${\mathsf{\Psi }}_N⟶\mathsf{\Psi },{\widehat{H}}_N{\mathsf{\Psi }}_N⟶\widehat{H}\mathsf{\Psi }$$

in the relevant Hilbert or distributional topology and with the time derivative commuting with the limit. These assumptions are part of the reconstruction statement, not automatic.

5.5. Main Reconstruction Theorem

Theorem 1

(Multi-Branch Reconstruction). Let $\mathcal{K}$ be finite or countable. For each $k\in \mathcal{K}$, define

$${\psi }_k[\varphi ,t]=R_k[\varphi ,t]\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k[\varphi ,t]\right),R_k[\varphi ,t]=\sqrt{{\rho }_k[\varphi ,t]}.$$

Assume that each pair $({\rho }_k,S_k)$ satisfies the branchwise continuity equation and the branchwise quantum Hamilton–Jacobi equation associated with the Hamiltonian $\widehat{H}$. Assume moreover that

$$\sum _{k\in \mathcal{K}}{\psi }_k[\varphi ,t]$$

converges in the domain of $\widehat{H}$, that the series may be differentiated termwise in time, and that $\widehat{H}$ may be applied termwise to the series. Then, the reconstructed wave functional

$$\mathsf{\Psi }[\varphi ,t]=\sum _{k\in \mathcal{K}}\sqrt{{\rho }_k[\varphi ,t]}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k[\varphi ,t]\right)$$

satisfies the functional Schrödinger equation

$$i\hslash {\displaystyle \frac{\partial \mathsf{\Psi }}{\partial t}}=\widehat{H}\mathsf{\Psi }.$$

Proof. 

The proof is conditional on the stated regularity, convergence, and domain assumptions. For a fixed branch k, write

$${\psi }_k=R_k{e}^{i{S}_k/\hslash },R_k=\sqrt{{\rho }_k}.$$

Substituting this polar form into the branchwise functional Schrödinger equation

$$i\hslash {\partial }_t{\psi }_k=\widehat{H}{\psi }_k$$

and separating the real and imaginary parts gives precisely the branchwise quantum Hamilton–Jacobi equation and the branchwise continuity equation. Conversely, if these two real equations hold and the required differentiability assumptions are satisfied, then recombining the real and imaginary parts gives

$$i\hslash {\partial }_t{\psi }_k=\widehat{H}{\psi }_k$$

for the individual branch component.

Thus, each admissible branch component ${\psi }_k$ satisfies the same linear functional Schrödinger equation. Let

$$\mathsf{\Psi }=\sum _{k\in \mathcal{K}}{\psi }_k.$$

By the assumed termwise differentiability,

$$i\hslash {\partial }_t\mathsf{\Psi }=i\hslash {\partial }_t\sum _{k\in \mathcal{K}}{\psi }_k=\sum _{k\in \mathcal{K}}i\hslash {\partial }_t{\psi }_k.$$

Using the branchwise Schrödinger equation for each term, this becomes

$$i\hslash {\partial }_t\mathsf{\Psi }=\sum _{k\in \mathcal{K}}\widehat{H}{\psi }_k.$$

Because the series converges in the domain of $\widehat{H}$ and since termwise application of $\widehat{H}$ is assumed to be justified, one has

$$\sum _{k\in \mathcal{K}}\widehat{H}{\psi }_k=\widehat{H}\sum _{k\in \mathcal{K}}{\psi }_k=\widehat{H}\mathsf{\Psi }.$$

Therefore,

$$i\hslash {\partial }_t\mathsf{\Psi }=\widehat{H}\mathsf{\Psi }.$$

This proves the reconstruction statement under the stated hypotheses. The theorem does not assert that the convergence, domain, or continuum-limit assumptions hold automatically; these are part of the admissibility conditions for a multi-branch representation. □

5.6. Scalar Quantum Field Equations

The reconstruction theorem immediately gives the functional Schrödinger equations for scalar quantum fields.

For a real scalar field with potential V, one obtains

$$i\hslash {\displaystyle \frac{\partial \mathsf{\Psi }[\varphi ,t]}{\partial t}}={\int }_{\sum }\left[-{\displaystyle \frac{{\hslash }^2}{2}}{\displaystyle \frac{{\delta }^2}{\delta \varphi {\left(x\right)}^2}}+{\displaystyle \frac{1}{2}}{|\nabla \varphi \left(x\right)|}^2+V\left(\varphi \left(x\right)\right)\right]d^{d}x\mathsf{\Psi }[\varphi ,t].$$

For the free massive scalar field

$$V\left(\varphi \right)={\displaystyle \frac{1}{2}}{m}^2{\varphi }^2,$$

this becomes

$$i\hslash {\displaystyle \frac{\partial \mathsf{\Psi }[\varphi ,t]}{\partial t}}={\displaystyle \frac{1}{2}}{\int }_{\sum }\left[-{\hslash }^2{\displaystyle \frac{{\delta }^2}{\delta \varphi {\left(x\right)}^2}}+{|\nabla \varphi \left(x\right)|}^2+m^2\varphi {\left(x\right)}^2\right]d^{d}x\mathsf{\Psi }[\varphi ,t],$$

which is the functional Schrödinger representation associated with the Klein–Gordon field [26,27].

For an interacting scalar field with

$$V\left(\varphi \right)={\displaystyle \frac{1}{2}}{m}^2{\varphi }^2+{\displaystyle \frac{\lambda }{4!}}{\varphi }^4,$$

the reconstructed equation is

$$i\hslash {\displaystyle \frac{\partial \mathsf{\Psi }[\varphi ,t]}{\partial t}}={\int }_{\sum }\left[-{\displaystyle \frac{{\hslash }^2}{2}}{\displaystyle \frac{{\delta }^2}{\delta \varphi {\left(x\right)}^2}}+{\displaystyle \frac{1}{2}}{|\nabla \varphi \left(x\right)|}^2+{\displaystyle \frac{1}{2}}{m}^2\varphi {\left(x\right)}^2+{\displaystyle \frac{\lambda }{4!}}\varphi {\left(x\right)}^4\right]d^{d}x\mathsf{\Psi }[\varphi ,t].$$

This is the functional Schrödinger equation for the interacting scalar quantum field [25,26,29]. In the continuum, the equation is understood through the chosen regularization and renormalization scheme.

5.7. Electromagnetic Field: Gauge-Constrained Reconstruction

The same mechanism extends formally to gauge fields provided that the constraints are imposed branchwise and the chosen regularization preserves the gauge symmetry [30]. For the electromagnetic field, let $A_i\left(x\right)$ be the spatial gauge potential and $E_i\left(x\right)$ its conjugate momentum. The classical Hamiltonian is

$$H[A,E]={\displaystyle \frac{1}{2}}{\int }_{\sum }\left(E_i\left(x\right)E_i\left(x\right)+B_i\left(x\right)B_i\left(x\right)\right)d^{3}x,$$

where

$$B_i\left(x\right)={ϵ}_{ijk}{\partial }_j{A}_k\left(x\right).$$

The Gauss constraint is

$$\mathcal{G}\left(x\right)={\partial }_i{E}_i\left(x\right)=0.$$

Canonical quantization gives

$$E_i\left(x\right)⟼-i\hslash {\displaystyle \frac{\delta }{\delta A_i\left(x\right)}}.$$

Therefore, the functional Schrödinger equation is

$$i\hslash {\displaystyle \frac{\partial \mathsf{\Psi }[A,t]}{\partial t}}={\displaystyle \frac{1}{2}}{\int }_{\sum }\left[-{\hslash }^2{\displaystyle \frac{{\delta }^2}{\delta A_i{\left(x\right)}^2}}+B_i{\left(A\right)}^2\right]d^{3}x\mathsf{\Psi }[A,t],$$

together with the quantum Gauss constraint [30,31]:

$${\partial }_i{\displaystyle \frac{\delta \mathsf{\Psi }[A,t]}{\delta A_i\left(x\right)}}=0.$$

We introduce the multi-branch ansatz

$$\mathsf{\Psi }[A,t]=\sum _{k\in \mathcal{K}}\sqrt{{\rho }_k[A,t]}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k[A,t]\right).$$

For each branch,

$$E_{k,i}\left(x\right)={\displaystyle \frac{\delta S_k}{\delta A_i\left(x\right)}}.$$

The branchwise Hamilton–Jacobi equation is

$${\displaystyle \frac{\partial S_k}{\partial t}}+{\displaystyle \frac{1}{2}}{\int }_{\sum }\left[{\displaystyle \frac{\delta S_k}{\delta A_i\left(x\right)}}{\displaystyle \frac{\delta S_k}{\delta A_i\left(x\right)}}+B_i{\left(A\right)}^2\right]d^{3}x+Q_k\left[A\right]=0,$$

where

$$Q_k\left[A\right]=-{\displaystyle \frac{{\hslash }^2}{2R_k\left[A\right]}}{\int }_{\sum }{\displaystyle \frac{{\delta }^2{R}_k\left[A\right]}{\delta A_i{\left(x\right)}^2}}{d}^{3}x,R_k\left[A\right]=\sqrt{{\rho }_k\left[A\right]}.$$

The continuity equation is

$${\displaystyle \frac{\partial {\rho }_k}{\partial t}}+{\int }_{\sum }{\displaystyle \frac{\delta }{\delta A_i\left(x\right)}}\left({\rho }_k{\displaystyle \frac{\delta S_k}{\delta A_i\left(x\right)}}\right)d^{3}x=0.$$

Gauge invariance is imposed branchwise by requiring

$${\partial }_i{\displaystyle \frac{\delta S_k}{\delta A_i\left(x\right)}}=0,{\partial }_i{\displaystyle \frac{\delta {\rho }_k}{\delta A_i\left(x\right)}}=0.$$

These conditions state that both the branch phase and the branch density are constant along infinitesimal gauge orbits. Under the same convergence assumptions as before, the reconstructed functional satisfies both of the following:

$$i\hslash {\displaystyle \frac{\partial \mathsf{\Psi }}{\partial t}}=\widehat{H}\mathsf{\Psi },{\partial }_i{\displaystyle \frac{\delta \mathsf{\Psi }}{\delta A_i}}=0.$$

The physical configuration space is the quotient

$$\mathcal{A}/\mathcal{G},$$

where $\mathcal{G}$ is the group of gauge transformations. Thus, the densities must be interpreted as gauge-invariant densities on $\mathcal{A}$, or equivalently as densities on the space of gauge orbits [31]. This is the basic mechanism by which the multi-branch reconstruction preserves Abelian gauge symmetry at the level of physical wave functionals.

5.8. Yang–Mills Fields

For Yang–Mills theory, let G be a compact Lie group with Lie algebra $\mathfrak{g}$. Let $A_i^a\left(x\right)$ be a gauge connection on $\sum$ and let $E_i^a\left(x\right)$ be the conjugate electric field. The classical Hamiltonian is

$$H[A,E]={\displaystyle \frac{1}{2}}{\int }_{\sum }\left(E_i^a{E}_i^a+B_i^a{B}_i^a\right)d^{d}x,$$

with

$$B_i^a={\displaystyle \frac{1}{2}}{ϵ}_{ijk}{F}_{jk}^a,F_{ij}^a={\partial }_i{A}_j^a-{\partial }_j{A}_i^a+f^{abc}{A}_i^b{A}_j^c.$$

The Gauss constraint is

$${\mathcal{G}}^a\left(x\right)=D_i^{ab}{E}_i^b\left(x\right)=0,$$

where

$$D_i^{ab}={\delta }^{ab}{\partial }_i+f^{acb}{A}_i^c.$$

Canonical quantization gives

$$E_i^a\left(x\right)⟼-i\hslash {\displaystyle \frac{\delta }{\delta A_i^a\left(x\right)}}.$$

The quantum Hamiltonian is formally

$$\widehat{H}={\displaystyle \frac{1}{2}}{\int }_{\sum }\left[-{\hslash }^2{\displaystyle \frac{{\delta }^2}{\delta A_i^a\left(x\right)\delta A_i^a\left(x\right)}}+B_i^a\left(A\right)B_i^a\left(A\right)\right]d^{d}x.$$

The physical wave functionals satisfy

$$i\hslash {\displaystyle \frac{\partial \mathsf{\Psi }[A,t]}{\partial t}}=\widehat{H}\mathsf{\Psi }[A,t]$$

and

$${\widehat{\mathcal{G}}}^a\left(x\right)\mathsf{\Psi }[A,t]=-i\hslash D_i^{ab}{\displaystyle \frac{\delta \mathsf{\Psi }[A,t]}{\delta A_i^b\left(x\right)}}=0.$$

The multi-branch ansatz is

$$\mathsf{\Psi }[A,t]=\sum _{k\in \mathcal{K}}\sqrt{{\rho }_k[A,t]}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k[A,t]\right).$$

For each branch,

$$E_{k,i}^a\left(x\right)={\displaystyle \frac{\delta S_k}{\delta A_i^a\left(x\right)}}.$$

The Yang–Mills branchwise Hamilton–Jacobi equation is

$${\partial }_t{S}_k+{\displaystyle \frac{1}{2}}{\int }_{\sum }\left[{\displaystyle \frac{\delta S_k}{\delta A_i^a\left(x\right)}}{\displaystyle \frac{\delta S_k}{\delta A_i^a\left(x\right)}}+B_i^a\left(A\right)B_i^a\left(A\right)\right]d^{d}x+Q_k\left[A\right]=0,$$

with

$$Q_k\left[A\right]=-{\displaystyle \frac{{\hslash }^2}{2R_k\left[A\right]}}{\int }_{\sum }{\displaystyle \frac{{\delta }^2{R}_k\left[A\right]}{\delta A_i^a\left(x\right)\delta A_i^a\left(x\right)}}{d}^{d}x,R_k\left[A\right]=\sqrt{{\rho }_k\left[A\right]}.$$

The continuity equation is

$${\partial }_t{\rho }_k+{\int }_{\sum }{\displaystyle \frac{\delta }{\delta A_i^a\left(x\right)}}\left({\rho }_k{\displaystyle \frac{\delta S_k}{\delta A_i^a\left(x\right)}}\right)d^{d}x=0.$$

Gauge invariance is imposed through the branchwise conditions

$$D_i^{ab}{\displaystyle \frac{\delta S_k}{\delta A_i^b\left(x\right)}}=0,D_i^{ab}{\displaystyle \frac{\delta {\rho }_k}{\delta A_i^b\left(x\right)}}=0.$$

These equations mean that each branch is constant along non-Abelian gauge orbits in the appropriate covariant sense. Under the convergence and domain assumptions of the reconstruction theorem, and provided that the regularization preserves the gauge constraint, the reconstructed wave functional satisfies both the functional Schrödinger equation and the Gauss constraint.

5.9. Gauge Invariance, Ward Identities, and BRST Compatibility

The gauge constraints above provide the canonical expression of gauge invariance. They ensure that the reconstructed state is defined on the physical quotient space of gauge orbits, at least formally. However, in quantum gauge field theory, gauge invariance must also be compatible with regularization, renormalization, and the identities satisfied by correlation functions. These consistency conditions are usually expressed through Ward identities in Abelian theories and through Slavnov–Taylor identities in non-Abelian theories [32,33,34].

In the present framework, these identities should not be viewed as being automatically produced by an arbitrary multi-branch decomposition. They are consistency requirements on admissible branch data and on the regularization scheme. If the branch phases $S_k$ and densities ${\rho }_k$ satisfy the Gauss constraints and if the regulator and renormalization prescription preserve the gauge symmetry, then the reconstructed wave functional satisfies the same gauge constraints as in the standard canonical formulation.

In the Abelian case, the infinitesimal gauge condition can be written schematically as

$$\widehat{\mathcal{G}}\left(x\right)\mathsf{\Psi }=0.$$

If the Hamiltonian commutes with the constraint

$$[\widehat{H},\widehat{\mathcal{G}}\left(x\right)]=0,$$

then gauge invariance is preserved by the reconstructed Schrödinger evolution. This is the canonical counterpart of the Ward identities. In the non-Abelian case, the corresponding statement is more delicate because the constraints close as an algebra and the quantum theory may require ghosts and gauge fixing. The preservation of the constraint algebra is the canonical counterpart of the Slavnov–Taylor identities.

A fully covariant treatment may be formulated by extending the configuration space to include ghosts and auxiliary fields:

$$(A,c,\overline{c},b)$$

leading to a BRST formulation [29,31,35]. In that setting, the branch data become

$$S_k[A,c,\overline{c},b],{\rho }_k[A,c,\overline{c},b]$$

and the physical states satisfy

$$Q_{\mathrm{BRST}}\mathsf{\Psi }=0,\mathsf{\Psi }\sim \mathsf{\Psi }+Q_{\mathrm{BRST}}\mathsf{\Lambda }.$$

The multi-branch structure is compatible with BRST cohomology when each branch component is BRST-closed or when BRST-exact changes of the branch data leave the reconstructed physical state unchanged. In schematic form, admissible branch components satisfy

$$Q_{\mathrm{BRST}}{\psi }_k=0,{\psi }_k=\sqrt{{\rho }_k}{e}^{i{S}_k/\hslash }.$$

Then, by linearity,

$$Q_{\mathrm{BRST}}\mathsf{\Psi }=\sum _{k\in \mathcal{K}}{Q}_{\mathrm{BRST}}{\psi }_k=0.$$

Thus, gauge invariance is preserved by the reconstruction when the branch data are chosen in the BRST cohomology and when the regularization respects the BRST symmetry.

This discussion should be understood as a consistency statement rather than a complete derivation of Ward or Slavnov–Taylor identities from the multi-branch formalism. A full proof would require specifying the gauge fixing, ghost measure, regularization scheme, and renormalized BRST operator. The present construction identifies the conditions under which the multi-branch Hamilton–Jacobi–density representation is compatible with the standard gauge-theoretic consistency requirements.

5.10. Fermionic Sector and Multi-Branch Dirac Reconstruction

The fermionic sector is formulated in a graded configuration space. Let

$${\mathcal{F}}_{\mathrm{tot}}={\mathcal{F}}_{\mathrm{B}}\oplus \mathsf{\Pi }{\mathcal{F}}_{\mathrm{F}},{\mathsf{\Phi }}^A=({\varphi }^a,{\psi }^{\alpha }),$$

where ${\varphi }^a$ are even bosonic variables and ${\psi }^{\alpha }$ are odd Grassmann variables carrying spinor indices [16,29,36]. The graded Hamiltonian decomposes as

$$H[\mathsf{\Phi },\mathsf{\Pi }]=H_{\mathrm{B}}[\varphi ,\pi ]+H_{\mathrm{F}}[\psi ,\overline{\psi },{\mathsf{\Pi }}_{\overline{\psi }}].$$

For the fermionic part, take

$$H_{\mathrm{F}}={\int }_{\sum }{\overline{\psi }}_{\alpha }\left(x\right){\mathcal{D}}_{\alpha \beta }{\mathsf{\Pi }}_{{\overline{\psi }}_{\beta }}\left(x\right)d^{3}x,$$

where

$${\mathcal{D}}_{\alpha \beta }=-i{\gamma }_{\alpha \beta }^i{\partial }_i+m{\delta }_{\alpha \beta }.$$

Upon quantization,

$${\mathsf{\Pi }}_{{\overline{\psi }}_{\beta }}\left(x\right)⟼-i\hslash {\displaystyle \frac{\delta }{\delta {\overline{\psi }}_{\beta }\left(x\right)}},$$

so that

$${\widehat{H}}_{\mathrm{F}}=-i\hslash {\int }_{\sum }{\overline{\psi }}_{\alpha }\left(x\right){\mathcal{D}}_{\alpha \beta }{\displaystyle \frac{\delta }{\delta {\overline{\psi }}_{\beta }\left(x\right)}}{d}^{3}x.$$

The full functional Schrödinger equation is

$$i\hslash {\partial }_t\mathsf{\Psi }=\left({\widehat{H}}_{\mathrm{B}}+{\widehat{H}}_{\mathrm{F}}\right)\mathsf{\Psi }.$$

We use the graded multi-branch ansatz

$$\mathsf{\Psi }\left[\mathsf{\Phi }\right]=\sum _{k\in \mathcal{K}}{\rho }_k[\mathsf{\Phi }{]}^{1/2}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k\left[\mathsf{\Phi }\right]\right),$$

where $S_k$ is even and ${\rho }_k^{1/2}$ is a graded amplitude. In the fermionic sector, the amplitude may take values in the relevant Clifford module, so that spin degrees of freedom are encoded algebraically in the density structure.

For one branch,

$${\mathsf{\Psi }}_k={\rho }_k^{1/2}{e}^{i{S}_k/\hslash }.$$

Acting with the fermionic Hamiltonian gives

$${\widehat{H}}_{\mathrm{F}}{\mathsf{\Psi }}_k=H_{\mathrm{F}}\left[\mathsf{\Phi },{\displaystyle \frac{\delta S_k}{\delta \mathsf{\Phi }}}\right]{\mathsf{\Psi }}_k-i\hslash {\mathcal{L}}_{\mathrm{F}}\left({\rho }_k^{1/2}\right)e^{i{S}_k/\hslash },$$

where

$$H_{\mathrm{F}}\left[\mathsf{\Phi },{\displaystyle \frac{\delta S_k}{\delta \mathsf{\Phi }}}\right]={\int }_{\sum }{\overline{\psi }}_{\alpha }{\mathcal{D}}_{\alpha \beta }{\displaystyle \frac{\delta S_k}{\delta {\overline{\psi }}_{\beta }}}{d}^{3}x$$

and

$${\mathcal{L}}_{\mathrm{F}}\left({\rho }_k^{1/2}\right)={\int }_{\sum }{\overline{\psi }}_{\alpha }{\mathcal{D}}_{\alpha \beta }{\displaystyle \frac{\delta {\rho }_k^{1/2}}{\delta {\overline{\psi }}_{\beta }}}{d}^{3}x.$$

Combining the bosonic and fermionic sectors yields the branchwise graded Hamilton–Jacobi equation

$${\partial }_t{S}_k+H_{\mathrm{B}}\left[\varphi ,{\displaystyle \frac{\delta S_k}{\delta \varphi }}\right]+H_{\mathrm{F}}\left[\mathsf{\Phi },{\displaystyle \frac{\delta S_k}{\delta \mathsf{\Phi }}}\right]+Q_{\mathrm{B}}\left[{\rho }_k\right]=0.$$

The corresponding graded continuity equation is

$${\partial }_t{\rho }_k+{\mathrm{Div}}_{\mu }^{\mathrm{gr}}\left({\rho }_k{V}_k\right)=0,V_k^A={\displaystyle \frac{\delta S_k}{\delta {\mathsf{\Phi }}^A}}.$$

There is no additional fermionic quantum potential of the same form as in the bosonic sector, since the fermionic Hamiltonian is first-order in functional derivatives. The fermionic contribution is instead carried by graded transport and Clifford-algebraic structure.

If each branch satisfies the graded Hamilton–Jacobi and continuity equations, then

$$i\hslash {\partial }_t{\mathsf{\Psi }}_k=\left({\widehat{H}}_{\mathrm{B}}+{\widehat{H}}_{\mathrm{F}}\right){\mathsf{\Psi }}_k.$$

By linearity,

$$i\hslash {\partial }_t\mathsf{\Psi }=\left({\widehat{H}}_{\mathrm{B}}+{\widehat{H}}_{\mathrm{F}}\right)\mathsf{\Psi }.$$

This gives the formal multi-branch reconstruction of Dirac dynamics in the graded Hamiltonian setting.

5.11. Interference, Decoherence, and Measure Localization

Although each branch ${\psi }_k$ satisfies the same linear quantum equation, probabilities are not obtained by summing the densities alone. For a finite sum,

$$|\mathsf{\Psi }{|}^2={\left|\sum _k\sqrt{{\rho }_k}{e}^{i{S}_k/\hslash }\right|}^2.$$

Hence,

$$|\mathsf{\Psi }{|}^2=\sum _k{\rho }_k+2\sum _{k<\mathcal{l}}\sqrt{{\rho }_k{\rho }_{\mathcal{l}}}\cos \left({\displaystyle \frac{S_k-S_{\mathcal{l}}}{\hslash }}\right).$$

The interference pattern is controlled by the relative phases between Hamilton–Jacobi branches [5,6].

A simple two-branch example makes this mechanism explicit. Suppose that only two admissible branches contribute:

$$\mathsf{\Psi }=\sqrt{{\rho }_1}{e}^{i{S}_1/\hslash }+\sqrt{{\rho }_2}{e}^{i{S}_2/\hslash }.$$

Then

$$|\mathsf{\Psi }{|}^2={\rho }_1+{\rho }_2+2\sqrt{{\rho }_1{\rho }_2}\cos \left({\displaystyle \frac{S_1-S_2}{\hslash }}\right).$$

The first two terms are the classical-looking branch weights, while the last term is the genuinely coherent contribution. If the phase difference $S_1-S_2$ is approximately constant on a region of configuration space, then the two branches interfere constructively or destructively according to the sign of the cosine. If $S_1-S_2$ varies rapidly compared with ℏ, then the oscillatory term is suppressed after coarse-graining or integration over unresolved degrees of freedom. This is the multi-branch Hamilton–Jacobi version of the usual stationary-phase intuition.

For example, if ${\rho }_1={\rho }_2=\rho /2$, then

$$|\mathsf{\Psi }{|}^2=\rho \left[1+\cos \left({\displaystyle \frac{S_1-S_2}{\hslash }}\right)\right].$$

Thus, the same total incoherent density ${\rho }_1+{\rho }_2=\rho$ can give very different probabilities depending on the relative action. This illustrates why the density data alone do not determine the observable probability distribution: the phase relations between branches are essential.

Decoherence may be described by enlarging the configuration space to include environmental variables. A typical branch decomposition then takes the form

$$\mathsf{\Psi }[\varphi ,E]=\sum _k\sqrt{{\rho }_k\left[\varphi \right]}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k\left[\varphi \right]\right){\chi }_k\left[E\right],$$

where ${\chi }_k\left[E\right]$ are environment states correlated with different branches. The reduced probability density for the field variables contains overlap factors of the form

$$⟨{\chi }_{\mathcal{l}},{\chi }_k⟩.$$

For two branches, one schematically obtains

$$|{\mathsf{\Psi }}_{\mathrm{red}}{\left[\varphi \right]|}^2={\rho }_1\left[\varphi \right]+{\rho }_2\left[\varphi \right]+2\sqrt{{\rho }_1\left[\varphi \right]{\rho }_2\left[\varphi \right]}\mathrm{Re}\left[⟨{\chi }_2,{\chi }_1⟩\exp \left({\displaystyle \frac{i}{\hslash }}(S_1\left[\varphi \right]-S_2\left[\varphi \right])\right)\right].$$

If

$$⟨{\chi }_k,{\chi }_{\mathcal{l}}⟩\approx 0\mathrm{for}k\ne \mathcal{l},$$

then the interference terms in the reduced state become dynamically suppressed [18,19,20]. In measure-theoretic language, this corresponds to an effective transition from a coherent branch measure to an approximately diagonal mixture.

If a measurement outcome corresponds to a subset ${\mathcal{K}}_{\alpha }\subset \mathcal{K}$, then the associated conditional measure may be written as

$$d{\mu }_{\alpha }={\displaystyle \frac{1}{P\left(\alpha \right)}}\sum _{k\in {\mathcal{K}}_{\alpha }}{\rho }_{k}d{\mu }_0,$$

where

$$P\left(\alpha \right)=\sum _{k\in {\mathcal{K}}_{\alpha }}{\int }_{\mathcal{F}}{\rho }_{k}d{\mu }_0.$$

Wavefunction reduction is then interpreted as an effective localization or conditioning of the branch measure relative to a measurement context, rather than as an additional fundamental postulate. This interpretation is not used in the reconstruction theorem itself; it is a structural consequence of representing the quantum state by phase-related Hamilton–Jacobi branches and density data.

5.12. Status of the Reconstruction

The reconstruction theorem shows that quantum field equations may be represented by a multi-branch Hamilton–Jacobi–density system. This does not reduce quantum field theory to ordinary classical field theory. Instead, it identifies an intermediate structure:

$$\mathrm{quantum}\mathrm{field}⟷\mathrm{multi-branch}\mathrm{Hamilton–Jacobi}\mathrm{geometry}+\mathrm{density}\mathrm{functional}.$$

The action functionals $S_k$ encode variational information, while the density functionals ${\rho }_k$ encode amplitude, probability, and the quantum-potential contribution. In finite-dimensional regularizations, the construction is rigorous under the stated smoothness and convergence assumptions. In the continuum, it remains conditional on the existence of a suitable limiting Hamiltonian, the convergence of the regularized states, and the control of the required renormalization procedures [23,25,28].

6. Illustrative Example: Free Scalar Field and Equivalence with the Standard Formulation

We now illustrate the reconstruction framework in the case of the free real scalar field. This example is useful because the functional Schrödinger equation is exactly solvable and the vacuum wave functional is explicitly known [25,26,27]. The purpose of the example is not to obtain a new spectrum but to show how the standard canonical result is represented in Hamilton–Jacobi–density form.

Let $\varphi \left(x\right)$ be a real scalar field on the spatial hypersurface $\sum ={\mathbb{R}}^3$. The classical Hamiltonian is

$$H[\varphi ,\pi ]={\displaystyle \frac{1}{2}}{\int }_{\sum }\left(\pi {\left(x\right)}^2+{|\nabla \varphi \left(x\right)|}^2+m^2\varphi {\left(x\right)}^2\right)d^{3}x.$$

Canonical quantization formally replaces

$$\pi \left(x\right)⟼-i\hslash {\displaystyle \frac{\delta }{\delta \varphi \left(x\right)}}.$$

The corresponding functional Schrödinger equation is

$$i\hslash {\displaystyle \frac{\partial \mathsf{\Psi }[\varphi ,t]}{\partial t}}={\displaystyle \frac{1}{2}}{\int }_{\sum }\left(-{\hslash }^2{\displaystyle \frac{{\delta }^2}{\delta \varphi {\left(x\right)}^2}}+{|\nabla \varphi \left(x\right)|}^2+m^2\varphi {\left(x\right)}^2\right)d^{3}x\mathsf{\Psi }[\varphi ,t].$$

In the standard formulation, the vacuum state is a Gaussian functional:

$${\mathsf{\Psi }}_0\left[\varphi \right]=\mathcal{N}\exp \left[-{\displaystyle \frac{1}{2\hslash }}{\int }_{\sum }{d}^{3}x{d}^{3}y\varphi \left(x\right)\mathsf{\Omega }(x,y)\varphi \left(y\right)\right]$$

where $\mathsf{\Omega }$ denotes the positive operator

$$\mathsf{\Omega }=\sqrt{-\Delta +m^2}.$$

Equivalently, in Fourier variables,

$${\mathsf{\Psi }}_0\left[\varphi \right]=\mathcal{N}\exp \left[-{\displaystyle \frac{1}{2\hslash }}\int {\displaystyle \frac{d^{3}k}{{\left(2\pi \right)}^3}}{\omega }_k{\varphi }_k{\varphi }_{-k}\right],{\omega }_k=\sqrt{{\left|k\right|}^2+m^2}.$$

We now rewrite the same vacuum state in Hamilton–Jacobi–density form. Since the vacuum is stationary, its phase is time-dependent only through the vacuum energy:

$$S_0[\varphi ,t]=-E_{0}t.$$

Consequently,

$${\displaystyle \frac{\delta S_0}{\delta \varphi \left(x\right)}}=0.$$

The corresponding density is

$${\rho }_0\left[\varphi \right]=|{\mathsf{\Psi }}_0{\left[\varphi \right]|}^2={\left|\mathcal{N}\right|}^2\exp \left[-{\displaystyle \frac{1}{\hslash }}{\int }_{\sum }{d}^{3}x{d}^{3}y\varphi \left(x\right)\mathsf{\Omega }(x,y)\varphi \left(y\right)\right].$$

Thus,

$$R_0\left[\varphi \right]=\sqrt{{\rho }_0\left[\varphi \right]}=\left|\mathcal{N}\right|\exp \left[-{\displaystyle \frac{1}{2\hslash }}{\int }_{\sum }{d}^{3}x{d}^{3}y\varphi \left(x\right)\mathsf{\Omega }(x,y)\varphi \left(y\right)\right].$$

The corresponding branch wave functional is

$${\psi }_0[\varphi ,t]=R_0\left[\varphi \right]\exp \left(-{\displaystyle \frac{i}{\hslash }}{E}_{0}t\right),$$

which is precisely the standard vacuum wave functional evolved by its vacuum energy.

The quantum Hamilton–Jacobi equation for this stationary branch is

$${\partial }_t{S}_0+H\left[\varphi ,{\displaystyle \frac{\delta S_0}{\delta \varphi }}\right]+Q\left[{\rho }_0\right]=0.$$

Since $\delta S_0/\delta \varphi =0$, this becomes

$$-E_0+{\displaystyle \frac{1}{2}}{\int }_{\sum }\left({|\nabla \varphi |}^2+m^2{\varphi }^2\right)d^{3}x+Q\left[{\rho }_0\right]=0.$$

Therefore,

$$Q\left[{\rho }_0\right]=E_0-{\displaystyle \frac{1}{2}}{\int }_{\sum }\left({|\nabla \varphi |}^2+m^2{\varphi }^2\right)d^{3}x.$$

This identity shows explicitly how the quantum potential generated by the Gaussian density compensates the classical field potential up to the vacuum energy. In this way, the density functional contains precisely the quantum correction required to reproduce the standard vacuum state.

The same conclusion holds mode-by-mode for excited states. In Fourier space, the free field decomposes into independent harmonic oscillators labeled by momentum k [26,27]. For a single mode, the Schrödinger equation is

$$i\hslash {\displaystyle \frac{\partial {\mathsf{\Psi }}_k}{\partial t}}={\displaystyle \frac{1}{2}}\left(-{\hslash }^2{\displaystyle \frac{{\partial }^2}{\partial {\varphi }_k^2}}+{\omega }_k^2{\varphi }_k^2\right){\mathsf{\Psi }}_k.$$

The stationary eigenstates are

$${\mathsf{\Psi }}_{n_k}\left({\varphi }_k\right)=N_{n_k}{H}_{n_k}\left(\sqrt{{\displaystyle \frac{{\omega }_k}{\hslash }}}{\varphi }_k\right)\exp \left(-{\displaystyle \frac{{\omega }_k}{2\hslash }}{\varphi }_k^2\right),$$

with energies

$$E_{n_k}=\hslash {\omega }_k\left(n_k+{\displaystyle \frac{1}{2}}\right).$$

In the Hamilton–Jacobi–density representation, each stationary mode can be written as

$${\mathsf{\Psi }}_{n_k}({\varphi }_k,t)=\sqrt{{\rho }_{n_k}\left({\varphi }_k\right)}\exp \left(-{\displaystyle \frac{i}{\hslash }}{E}_{n_k}t\right),$$

where

$${\rho }_{n_k}\left({\varphi }_k\right)={\left|N_{n_k}\right|}^2{H}_{n_k}^2\left(\sqrt{{\displaystyle \frac{{\omega }_k}{\hslash }}}{\varphi }_k\right)\exp \left(-{\displaystyle \frac{{\omega }_k}{\hslash }}{\varphi }_k^2\right).$$

For each oscillator mode, the corresponding quantum potential reproduces the standard harmonic oscillator energy eigenvalue. Thus, the Hamilton–Jacobi–density representation recovers the usual Fock-space spectrum of the free scalar field,

$$E_{\left\{n_k\right\}}=\sum _k\hslash {\omega }_k\left(n_k+{\displaystyle \frac{1}{2}}\right).$$

This example shows that the multi-branch Hamilton–Jacobi–density formulation is not an alternative approximation to quantum field theory. For the free scalar field, it reproduces the standard vacuum functional, the stationary excited states, and the energy spectrum obtained by canonical quantization. The difference lies in the structural interpretation: the quantum state is represented by Hamilton–Jacobi phase data together with a density functional for which the configuration-space curvature gives the quantum-potential contribution.

7. Geometric and Multisymplectic Interpretation

7.1. Multisymplectic Phase Space

Let M be spacetime and let $\mathcal{F}$ be the configuration space of fields. The covariant Hamiltonian formulation is naturally expressed on a multisymplectic phase space $\mathcal{P}$ [10,11,12,13]. In local coordinates, one writes

$$(x^{\mu },\varphi ,{\pi }^{\mu }),$$

where ${\pi }^{\mu }$ are the polymomenta conjugate to the derivatives ${\partial }_{\mu }\varphi$.

Locally, the multisymplectic form can be written as

$$\omega =d{\pi }^{\mu }\wedge d\varphi \wedge d^{n-1}{x}_{\mu },$$

where $d^{n-1}{x}_{\mu }$ denotes the contraction of the spacetime volume form with ${\partial }_{\mu }$. A Hamiltonian density $\mathcal{H}$ encodes the dynamics, and the field equations correspond to integral submanifolds satisfying the multisymplectic Hamilton equations [12,13].

7.2. Hamilton–Jacobi Submanifolds

A solution $S_k^{\mu }(x,\varphi )$ of the covariant Hamilton–Jacobi equation defines a section of the polymomentum phase space by

$${\pi }_k^{\mu }={\displaystyle \frac{\partial S_k^{\mu }}{\partial \varphi }}.$$

This section determines a submanifold

$${\mathcal{L}}_k\subset \mathcal{P}.$$

Under the usual regularity and isotropy conditions, ${\mathcal{L}}_k$ is Lagrangian, or more precisely Hamiltonian/Lagrangian in the appropriate multisymplectic sense. Formally, this is expressed by the vanishing of the pullback of the multisymplectic form

$${\omega |}_{{\mathcal{L}}_k}=0.$$

Thus, each Hamilton–Jacobi branch corresponds geometrically to a classical variational submanifold in multisymplectic phase space [10,11].

7.3. Multi-Branch Structure as a Family of Lagrangian Submanifolds

The multi-branch construction introduced above can be interpreted geometrically as a family

$${\left\{{\mathcal{L}}_k\right\}}_{k\in \mathcal{K}}$$

of Lagrangian submanifolds in $\mathcal{P}$. Instead of selecting a single classical branch, the quantum state is represented by a structured collection of classical variational geometries. Each ${\mathcal{L}}_k$ encodes one admissible Hamilton–Jacobi branch, associated for example with a distinct extremal configuration, boundary condition, or local action functional.

This interpretation also clarifies the non-uniqueness of the branch decomposition. The family $\left\{{\mathcal{L}}_k\right\}$ should not be understood as a unique canonical object but as an admissible geometric representation constrained by the Hamilton–Jacobi equations, the density evolution equations, and the reconstruction condition.

7.4. Density Functionals as Measures on Lagrangian Families

The density functionals ${\rho }_k\left[\varphi \right]$ admit a geometric interpretation as weights, or Radon–Nikodym densities, associated with the different Hamilton–Jacobi branches. More precisely, for each branch, there can be an associated measure ${\mu }_k$ that is absolutely continuous with respect to a chosen reference measure on configuration space; formally, one can write

$$d{\mu }_k={\rho }_{k}d\mu .$$

The total branch measure is then

$${\mu }_{\mathsf{\Psi }}=\sum _{k\in \mathcal{K}}{\mu }_k.$$

In this sense, the quantum state is represented not by a single classical geometry but by a measure-supported family of classical geometries. This viewpoint is related in spirit to the path-integral picture [5,6], but is formulated in terms of Hamilton–Jacobi branches and density functionals rather than an unconstrained sum over arbitrary paths.

7.5. Reconstruction of the Wave Functional

The wave functional is reconstructed as

$$\mathsf{\Psi }\left[\varphi \right]=\sum _{k\in \mathcal{K}}\sqrt{{\rho }_k\left[\varphi \right]}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k\left[\varphi \right]\right).$$

Each phase factor

$$\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k\right)$$

encodes the action associated with the branch ${\mathcal{L}}_k$, while the amplitude $\sqrt{{\rho }_k}$ gives the weight of that branch in the reconstructed quantum state.

The square root of the density should not be interpreted as an arbitrary amplitude; it is part of the branch data and is constrained by the continuity equation, the regularity assumptions, and the requirement that the reconstructed wave functional satisfy the functional Schrödinger equation.

7.6. Interference as Geometric Phase Interaction

Interference effects arise from the relative phases between different branches. For a finite sum, one obtains

$$|\mathsf{\Psi }{|}^2=\sum _k{\rho }_k+2\sum _{k<\mathcal{l}}\sqrt{{\rho }_k{\rho }_{\mathcal{l}}}\cos \left({\displaystyle \frac{S_k-S_{\mathcal{l}}}{\hslash }}\right).$$

Geometrically, the phase differences $S_k-S_{\mathcal{l}}$ measure the relative action between Hamilton–Jacobi branches. Therefore, the resulting interference pattern reflects the global organization of the family

$${\left\{{\mathcal{L}}_k\right\}}_{k\in \mathcal{K}}.$$

This is the sense in which interference can be viewed as a geometric phase interaction between distinct variational submanifolds.

7.7. Quantum Fields as Measure-Supported Lagrangian Geometry

The preceding discussion leads to the following interpretation: a quantum field state may be represented as a measure-supported family of Hamilton–Jacobi, or Lagrangian, submanifolds in multisymplectic phase space. In this picture:

• Classical field theory corresponds locally to a single Hamilton–Jacobi branch.
• Quantum field theory is represented by a family of admissible branches equipped with density data.
• Interference arises from phase relations between branches.
• The quantum potential arises from the density contribution in the polar decomposition of each branch.

This interpretation is structural rather than ontological. It does not assert that the quantum field literally consists of a set of classical submanifolds; rather, it shows that, under the hypotheses of the reconstruction theorem, solutions of the functional Schrödinger equation can be represented by Hamilton–Jacobi branches supplemented by density functionals.

8. Toward Quantum Gravity and Cohomological Structure

In a gravitational setting, the configuration space must be enlarged to include spatial geometries. Let $\mathcal{G}$ denote the space of Riemannian metrics on a spatial hypersurface $\sum$ and let $\mathcal{F}$ denote the matter-field configuration space. Therefore, the relevant configuration variables are pairs

$$(g,\varphi )\in \mathcal{G}\times \mathcal{F}.$$

At the classical level, the Einstein–Hilbert action coupled to matter governs the dynamics:

$$S[g,\varphi ]={\displaystyle \frac{1}{16\pi G}}{\int }_{M}R\left(g\right)\sqrt{-g}{d}^{4}x+S_{\mathrm{matter}}[g,\varphi ]$$

where $R\left(g\right)$ is the scalar curvature of the spacetime metric g [37].

In canonical ADM form, the phase-space variables are $(g_{ij}\left(x\right),{\pi }^{ij}\left(x\right))$ and the dynamics is encoded in constraints [31,38]. The Hamiltonian is formally written as

$$H={\int }_{\sum }\left(N\left(x\right)\mathcal{H}\left(x\right)+N^i\left(x\right){\mathcal{H}}_i\left(x\right)\right)d^{3}x,$$

where N is the lapse function, $N^i$ is the shift vector, $\mathcal{H}\left(x\right)$ is the Hamiltonian constraint, and ${\mathcal{H}}_i\left(x\right)$ is the momentum, or spatial diffeomorphism, constraint. The Hamiltonian constraint has the schematic form

$$\mathcal{H}\left(x\right)=G_{ijkl}\left(g\right){\pi }^{ij}{\pi }^{kl}-\sqrt{g}(R-2\mathsf{\Lambda })+{\mathcal{H}}_{\mathrm{matter}}\left(x\right),$$

where $G_{ijkl}$ denotes the DeWitt supermetric and $\mathsf{\Lambda }$ is the cosmological constant [38]. Numerical factors and factor-ordering conventions are not essential for the structural discussion below.

8.1. Multi-Branch Hamilton–Jacobi Structure on Superspace

We introduce a finite or countable family of action functionals on the gravitational configuration space,

$${\left\{S_k[g,\varphi ]\right\}}_{k\in \mathcal{K}}.$$

Each branch is required to satisfy the gravitational Hamilton–Jacobi constraint

$$\mathcal{H}\left[g,\varphi ;{\displaystyle \frac{\delta S_k}{\delta g}},{\displaystyle \frac{\delta S_k}{\delta \varphi }}\right]=0$$

together with the momentum constraint

$${\mathcal{H}}_i\left[g,\varphi ;{\displaystyle \frac{\delta S_k}{\delta g}},{\displaystyle \frac{\delta S_k}{\delta \varphi }}\right]=0.$$

Each functional $S_k$ describes, at least formally, a classical spatial geometry together with a matter configuration satisfying the canonical constraints. The multi-branch structure corresponds to a family of Hamilton–Jacobi branches, or local Lagrangian submanifolds, in the gravitational phase space [10,11].

As in the non-gravitational case, the family $\left\{S_k\right\}$ is not assumed to be unique; it is an admissible representation of the gravitational wave functional only when the branch constraints, density equations, and global reconstruction conditions are satisfied. In the gravitational case, admissibility also requires compatibility with spatial diffeomorphism invariance and with the Hamiltonian constraint.

8.2. Density Functionals and Quantum Corrections

We associate a density functional ${\rho }_k[g,\varphi ]$ with each branch, interpreted formally as a density on the gravitational configuration space. More precisely, because physical gravitational configurations are spatial geometries modulo diffeomorphisms, the density should ultimately be understood on superspace,

$$\mathrm{Riem}(\sum )/\mathrm{Diff}(\sum ),$$

or on a suitable regularized approximation of this quotient. The construction of such measures is a nontrivial problem. In the present work, the density is treated as part of a formal or regularized reconstruction scheme.

Formally, the densities satisfy a continuity equation of the form

$${\mathrm{div}}_{\mu }\left({\rho }_k{\displaystyle \frac{\delta S_k}{\delta g}}\right)+{\mathrm{div}}_{\mu }\left({\rho }_k{\displaystyle \frac{\delta S_k}{\delta \varphi }}\right)=0,$$

where the divergence is understood with respect to a chosen reference measure $\mu$ on the regularized configuration space.

Writing

$$R_k[g,\varphi ]=\sqrt{{\rho }_k[g,\varphi ]},$$

the density contribution produces a formal quantum correction to the Hamilton–Jacobi constraint. The branchwise quantum Hamilton–Jacobi equation has the schematic form

$$\mathcal{H}\left[g,\varphi ;{\displaystyle \frac{\delta S_k}{\delta g}},{\displaystyle \frac{\delta S_k}{\delta \varphi }}\right]+Q_k[g,\varphi ]=0.$$

The gravitational quantum potential is formally given by

$$Q_k[g,\varphi ]=-{\displaystyle \frac{{\hslash }^2}{2R_k}}{\int }_{\sum }{G}_{ijkl}\left(g\left(x\right)\right){\displaystyle \frac{{\delta }^2{R}_k}{\delta g_{ij}\left(x\right)\delta g_{kl}\left(x\right)}}{d}^{3}x+\cdots ,$$

where the ellipsis denotes matter contributions, factor-ordering terms, gauge-fixing contributions if present, and regularization-dependent corrections. This expression is formal and must be understood in a regularized setting.

8.3. Multi-Branch Reconstruction of the Wheeler–DeWitt Equation

Canonical quantization promotes the momenta to functional derivatives,

$${\pi }^{ij}\left(x\right)⟼-i\hslash {\displaystyle \frac{\delta }{\delta g_{ij}\left(x\right)}}.$$

This leads formally to the Wheeler–DeWitt constraints [38]

$$\widehat{\mathcal{H}}\left(x\right)\mathsf{\Psi }[g,\varphi ]=0,{\widehat{\mathcal{H}}}_i\left(x\right)\mathsf{\Psi }[g,\varphi ]=0.$$

We consider the multi-branch decomposition

$$\mathsf{\Psi }[g,\varphi ]=\sum _{k\in \mathcal{K}}\sqrt{{\rho }_k[g,\varphi ]}\exp \left({\displaystyle \frac{i}{\hslash }}{S}_k[g,\varphi ]\right)$$

and write

$${\psi }_k[g,\varphi ]=\sqrt{{\rho }_k[g,\varphi ]}{e}^{i{S}_k[g,\varphi ]/\hslash }.$$

Assume that the following conditions hold:

• Each pair $(S_k,{\rho }_k)$ satisfies the branchwise quantum Hamilton–Jacobi system and the momentum constraint.
• The branch densities are compatible with spatial diffeomorphism invariance, either on configuration space before quotienting or directly on a regularized version of superspace.
• The series defining $\mathsf{\Psi }$ converges in the domain of the regularized constraint operators.
• Functional differentiation and summation may be interchanged.
• The continuum limit of the regularized constraints exists in the chosen topology.

Then, in direct analogy with the non-gravitational reconstruction theorem, each branch satisfies

$$\widehat{\mathcal{H}}{\psi }_k=0,{\widehat{\mathcal{H}}}_i{\psi }_k=0,$$

and by linearity and the assumed convergence conditions,

$$\widehat{\mathcal{H}}\mathsf{\Psi }=0,{\widehat{\mathcal{H}}}_i\mathsf{\Psi }=0.$$

Thus, the Wheeler–DeWitt constraints can be represented, formally and under the stated assumptions, by a multi-branch Hamilton–Jacobi–density structure on superspace. This statement should be read as a conditional reconstruction result, not as a complete construction of a quantum theory of gravity.

8.4. Geometric and Cohomological Interpretation

In this formulation, a quantum gravitational state is represented by a measure-supported family of classical gravitational Hamilton–Jacobi branches. The action functionals $\left\{S_k\right\}$ encode local classical solutions of the Einstein–matter Hamilton–Jacobi constraints, while the densities $\left\{{\rho }_k\right\}$ determine how these geometries are weighted and coherently combined.

In general, the Hamilton–Jacobi functional need not be globally definable on superspace [39]. The collection $\left\{S_k\right\}$ may be interpreted as a system of local representatives of a cohomological structure. On overlaps of local domains, the differences

$$S_k-S_{\mathcal{l}}$$

encode transition data between local Hamilton–Jacobi descriptions. The multi-branch structure then reflects the obstruction to defining a single global action functional.

The density functionals provide a measure-theoretic organization of these local variational data. The wave functional $\mathsf{\Psi }$ appears as a global object reconstructed from local Hamilton–Jacobi branches, density data, and compatibility conditions. From this perspective, the Wheeler–DeWitt equation is not replaced but reinterpreted as a global compatibility condition imposed on the local Hamilton–Jacobi–density data.

8.5. Formal Status and Limitations of the Gravitational Extension

The gravitational extension presented here is formal and conditional. It is included to indicate how the multi-branch Hamilton–Jacobi–density framework may be adapted to the canonical variables of gravity.

Several limitations must be made explicit. First, the construction of measures on superspace is a difficult open problem. Since superspace is a quotient by the diffeomorphism group and has singular strata associated with metrics possessing nontrivial symmetries, the existence of a suitable reference measure is not automatic. In the present approach, the density functionals should be understood either in a finite-dimensional approximation, in a gauge-fixed setting, or as formal objects for which a rigorous definition remains to be supplied.

Second, spatial diffeomorphism invariance must be imposed consistently. The momentum constraint requires the wave functional to be invariant under infinitesimal spatial diffeomorphisms. In the multi-branch representation, this means that the branch phases and densities must be compatible with the diffeomorphism orbits. Equivalently, the reconstructed state must descend from the space of metrics to the quotient space of geometries. This is an additional admissibility condition on the branch data.

Third, the Hamiltonian constraint is tied to the problem of time. The Wheeler–DeWitt equation is not an ordinary time-dependent Schrödinger equation; it is a constraint equation. Consequently, the reconstruction should not be interpreted as an evolution in an external time parameter but as giving a formal compatibility condition for branch data satisfying the Hamiltonian constraint. Any interpretation involving relational time, semiclassical time, or clock variables would require additional structure that is not developed here.

Fourth, the gravitational Hamiltonian operator requires regularization and may be affected by factor ordering and renormalization ambiguities. Products of functional derivatives with respect to the metric at the same spatial point are formal. A complete treatment would require a specific regularization, control of the constraint algebra after quantization, and a demonstration that the continuum limit is well-defined. These issues are beyond the scope of the present work.

For these reasons, the gravitational part of the paper should be understood as an outlook and structural extension of the multi-branch formalism. Its role is to show how the same Hamilton–Jacobi–density logic can be written at the level of gravitational constraints while clearly separating this formal reconstruction from the unresolved analytical and conceptual problems of canonical quantum gravity.

8.6. Interpretation and Outlook

The construction in this paper suggests that quantum gravity may be represented, at least formally, as a theory of classical geometries organized by density and gluing data on superspace. The passage from classical to quantum theory is then described not as a simple modification of the classical equations of motion but as enrichment of the space of Hamilton–Jacobi solutions by measure-theoretic and cohomological structures.

This interpretation remains conditional; a fully rigorous formulation would require addressing the construction of suitable measures on superspace, the implementation of diffeomorphism invariance, the regularization and renormalization of the Wheeler–DeWitt operator, the factor-ordering problem, the preservation of the quantum constraint algebra, and the problem of time [40,41]. These issues are beyond the scope of the present work, but indicate a possible route toward connecting quantum gravity, geometry, and cohomological structure within the multi-branch Hamilton–Jacobi–density framework.

9. Conclusions

We have developed a multi-branch Hamilton–Jacobi–density framework for representing solutions of quantum field equations. In this formulation, each branch carries Hamilton–Jacobi phase data, while the associated density functional supplies the amplitude information and the quantum-potential contribution required in the polar form of the Schrödinger equation.

The specific contribution of the work is not the Hamilton–Jacobi formalism itself, nor the polar decomposition of wave functionals, both of which are well-established; rather, the contribution is the organization of quantum field states as admissible multi-branch Hamilton–Jacobi–density data. Each branch is constrained by Hamilton–Jacobi and continuity equations, and the reconstructed wave functional is required to satisfy the functional Schrödinger equation under explicit convergence and domain assumptions.

For bosonic fields, we have shown that a superposition of branchwise contributions yields the functional Schrödinger equation under explicit regularity, convergence, and domain assumptions [4]. The quantum effects associated with each branch arise from the density functional through the quantum potential, which contains second-order functional derivatives. This is consistent with the hydrodynamic and Bohmian formulations of quantum theory [7,8,9].

The construction also admits a formal extension to fermionic fields within a graded Hamiltonian setting. In that case, the density is understood as a graded object, possibly carrying Clifford-module structure, and we reconstruct the Dirac dynamics through the first-order fermionic Hamiltonian [3,26,27]. Unlike the bosonic case, there is no quantum potential of the same scalar second-order form; the fermionic contribution is instead encoded through graded transport and Clifford-algebraic structure.

Gauge theories can be incorporated by imposing the relevant constraints branchwise. For electromagnetic and Yang–Mills fields, the Hamilton–Jacobi branches and the densities must satisfy Gauss-type constraints, so that the reconstructed wave functional descends to the physical configuration space of gauge orbits. Similarly, the extension to gravitational systems suggests a formal multi-branch representation of the Wheeler–DeWitt constraints on superspace [38,39,41]. In this setting, the Wheeler–DeWitt equation may be viewed as a global compatibility condition on Hamilton–Jacobi–density data, although a rigorous treatment would require substantial additional work.

From a conceptual point of view, our framework suggests that quantum field theory can be represented as classical Hamilton–Jacobi dynamics enriched by additional structure on configuration space. This additional structure is measure-theoretic and geometric in the bosonic sector, while in the fermionic sector it is graded and algebraic. The multi-branch representation also provides a natural language for describing interference, decoherence, and effective wavefunction reduction in terms of phase relations and density localization across branches [18,19].

The cohomological interpretation gives a further structural perspective. In general, a Hamilton–Jacobi action functional need not be globally single-valued on configuration space. The multi-branch structure may then be understood as a collection of local variational data for which the transition relations encode global information. In this sense, the wave functional is reconstructed as a global object from local Hamilton–Jacobi branches and density data subject to compatibility conditions [15,16,36].

Several open problems remain, including the rigorous construction of measures on infinite-dimensional configuration spaces [24,25], the control of convergence and renormalization in interacting field theories [29], the treatment of gauge constraints and BRST structure at a fully rigorous level [31,35], and the definition of the corresponding structures on superspace in quantum gravity. Addressing these issues is necessary before the present framework can be regarded as a complete mathematical reconstruction of quantum field theory. Nevertheless, the results obtained here indicate a possible geometric and measure-theoretic organization of quantum dynamics in terms of Hamilton–Jacobi branches and density functionals.