[Paper Review] Moyal's Characteristic Function, the Density Matrix and von Neumann's Idempotent

B. J. Hiley|arXiv (Cornell University)|Aug 25, 2014

Quantum Mechanics and Applications14 references3 citations

TL;DR

This paper demonstrates that Moyal’s characteristic function in the Wigner-Moyal phase-space formulation of quantum mechanics is mathematically equivalent to von Neumann’s primitive idempotent operator, revealing that the Wigner quasi-probability distribution is not a classical statistical distribution but a representation of the quantum density matrix. The equivalence confirms the non-classical, non-commutative foundation of the Wigner-Moyal formalism and explains why the Wigner function can take negative values.

ABSTRACT

In the Wigner-Moyal approach to quantum mechanics, we show that Moyal's starting point, the characteristic function $M(τ,θ)=\int ψ^{*}(x)e^{i(τ{\hat p}+θ{\hat x})}ψ(x)dx$, is essentially the primitive idempotent used by von Neumann in his classic paper "Die Eindeutigkeit der Schrödingerschen Operatoren". This paper provides the original proof of the Stone-von Neumann equation. Thus the mathematical structure Moyal develops is simply a re-expression of what is at the heart of quantum mechanics and reproduces exactly the results of the quantum formalism. The "distribution function" $F(X,P,t)$ is simply the quantum mechanical density matrix expressed in an $( X,P)$-representation, where $X$ and $P$ are the mean co-ordinates of a cell structure in phase space. The whole approach therefore clearly has little to do with classical statistical theories but is a consequence of a non-commutative nature of the theory.

Motivation & Objective

• To clarify the foundational mathematical structure underlying the Wigner-Moyal phase-space formulation of quantum mechanics.
• To resolve the apparent contradiction between the classical-statistical appearance of Moyal’s characteristic function and the inherently quantum mechanical origin of the Wigner function.
• To establish that the Wigner function F(x,p,t) is a representation of the quantum density operator, not a classical probability distribution.
• To unify the Wigner-Moyal and density matrix approaches by showing they stem from the same core quantum structure — the primitive idempotent.
• To reinforce the non-classical nature of phase-space quantum mechanics by linking it to the algebraic foundations of quantum theory, particularly the Heisenberg algebra and symplectic geometry.

Proposed method

• Derive Moyal’s characteristic function M(τ,θ) as the expectation value of the operator e^{i(τp̂ + θx̂)} in the position representation.
• Express Moyal’s characteristic function in the form M(τ,θ) = ∫ψ*(x−τ/2)e^{iθx}ψ(x+τ/2)dx, showing its dependence on the wavefunction and non-commutative phase-space structure.
• Reconstruct the Wigner function F(x,p,t) as the inverse Fourier transform of M(τ,θ), demonstrating it as a phase-space representation of the density matrix.
• Identify von Neumann’s primitive idempotent operator as the fundamental object underlying Moyal’s characteristic function, using the projection onto a pure state in Hilbert space.
• Show that the matrix element ⟨x|N(α,β)|x⟩, when restricted to a single state, reproduces Moyal’s characteristic function up to a phase factor.
• Establish that the Wigner function F(x,p,t) is equivalent to the density matrix in the (X,P)-representation, where X and P are mean co-ordinates of a phase-space cell.

Experimental results

Research questions

• RQ1 Is Moyal’s characteristic function in the Wigner-Moyal formalism fundamentally classical or quantum in origin?
• RQ2 How is the Wigner function F(x,p,t) related to the quantum mechanical density operator ρ̂?
• RQ3 Can the Wigner-Moyal approach be derived from the foundational work of von Neumann on the uniqueness of the Schrödinger representation?
• RQ4 Why does the Wigner function take negative values, and what does this imply about its interpretation?
• RQ5 What is the geometric and algebraic structure of phase space in quantum mechanics, and how does it differ from classical phase space?

Key findings

• Moyal’s characteristic function M(τ,θ) is mathematically identical to the primitive idempotent used by von Neumann in his proof of the Stone-von Neumann theorem.
• The Wigner function F(x,p,t) is not a classical probability distribution but a representation of the quantum density matrix in the (X,P)-phase-space representation.
• The negative values of the Wigner function are consistent with the density matrix formalism and do not violate quantum mechanics, as the density matrix need not be positive-definite in all representations.
• The phase-space structure underlying the Wigner-Moyal formalism is inherently non-classical, with a cellular structure defined by conjugate pairs of points (x,p) and (x′,p′), reflecting the non-commutativity of quantum mechanics.
• The Wigner function arises from a non-commutative algebraic structure — specifically, the Heisenberg algebra — and is formally linked to symplectic geometry and Gromov’s 'no squeezing' theorem.
• The underlying quantum phase space is constructed from the simultaneous eigenvalues of mean position and momentum operators X̂ and P̂, which commute, but whose construction relies on non-commuting x̂ and p̂, confirming the non-classical origin of the formalism.