Quantum Mechanics in Configuration Space

• Quantum mechanics in configuration space is a formulation where a system’s state is represented on its configuration manifold, emphasizing positions over abstract phase space.
• This approach employs methods like coherent states, bundle geometry, and Schrödinger mapping to capture many-particle dynamics and relational formulations.
• Alternative formulations include noncommutative, curved, and discrete configuration spaces, which influence quantum dynamics, classical limits, and underlying symmetries.

Searching arXiv for recent and foundational papers on quantum mechanics in configuration space to ground the article. Quantum mechanics in configuration space is the formulation in which the state of a system is represented on the space of its possible configurations rather than on classical phase space. For $N$ structureless point particles, the basic configuration space is $Q=\mathbb{R}^{3N}$, and the standard nonrelativistic wavefunction is a complex field $\psi(q,t)$ on $Q\times \mathbb{R}$. In this setting, configuration space is not merely a calculational device: across the literature it is treated as the spectral space of commuting position operators, as a homogeneous space of kinematical symmetry, as the base of bundle-geometric and relational constructions, as the arena for density and trajectory formulations without an ontic wavefunction, and, in deformed models, as a compact, curved, polymer, or noncommutative manifold (Kong, 2017, François et al., 3 Jan 2025, Siebersma et al., 4 Jun 2026, Gneiting et al., 2013).

1. Standard configuration-space formulation

In the conventional formulation, a single spinless particle is described by

$H=L^2(\mathbb{R}^3),$

with kets $|\mathbf{x}\rangle$ diagonalizing the three commuting position operators $\hat{\mathbf{x}}=(\hat{x}_1,\hat{x}_2,\hat{x}_3)$. They satisfy

$$\langle \mathbf{x}|\mathbf{x}'\rangle=\delta^{(3)}(\mathbf{x}-\mathbf{x}'), \qquad \int_{\mathbb{R}^3} d^3x\, |\mathbf{x}\rangle\langle \mathbf{x}|=\mathbb{I},$$

and any state is represented by $\psi(\mathbf{x})=\langle \mathbf{x}|\psi\rangle$. In configuration space, $\hat{\mathbf{x}}$ acts by multiplication and $Q=\mathbb{R}^{3N}$0, with

$Q=\mathbb{R}^{3N}$1

For a free particle,

$Q=\mathbb{R}^{3N}$2

with plane-wave momentum eigenstates, Gaussian wave-packet evolution, the free propagator, and the Born density $Q=\mathbb{R}^{3N}$3 all formulated directly on configuration space (Kong, 2017).

For many-particle systems, the configuration space is $Q=\mathbb{R}^{3N}$4, a configuration being $Q=\mathbb{R}^{3N}$5 with $Q=\mathbb{R}^{3N}$6. In the standard global trivialization $Q=\mathbb{R}^{3N}$7, the Schrödinger equation is

$Q=\mathbb{R}^{3N}$8

and the inner product is the fiber-wise integration over $Q=\mathbb{R}^{3N}$9 at fixed time,

$\psi(q,t)$0

This many-coordinate version is the baseline for subsequent relational, geometric, and hydrodynamic reformulations (François et al., 3 Jan 2025).

A broader geometric statement also appears in statistical derivations of quantization: the classical kinetic energy defines an $\psi(q,t)$1-dimensional metric configuration space with metric $\psi(q,t)$2, and canonical quantization becomes the replacement of the classical conjugate momentum vector by $\psi(q,t)$3 times the gradient in that metric space. The kinetic operator is then the Laplace–Beltrami operator, and the continuity equation uses the metric volume form $\psi(q,t)$4 (Goedecke, 2014).

2. Configuration space as physical space from symmetry

A central line of work identifies physical space itself with the configuration space of a free particle. In this view, classical Newtonian space is a homogeneous space of the relativity symmetry, and its quantum analogue is obtained from the unitary projective representation of the centrally extended Galilei group. The relevant quantum kinematical symmetry is the Heisenberg–Weyl group with rotations, $\psi(q,t)$5, with central generator $\psi(q,t)$6 and

$\psi(q,t)$7

In an irreducible representation, $\psi(q,t)$8 acts as $\psi(q,t)$9, reproducing the canonical commutator (Kong, 2017).

From this symmetry viewpoint, the commuting position operators admit a joint spectral resolution, and their spectrum is identified with $Q\times \mathbb{R}$0. The orbit of the translation subgroup generated by $Q\times \mathbb{R}$1 acting on a fiducial position eigenstate $Q\times \mathbb{R}$2 generates the position basis,

$Q\times \mathbb{R}$3

The standard Hilbert space $Q\times \mathbb{R}$4 thus appears as the irreducible representation in which the spectrum of $Q\times \mathbb{R}$5 labels points of physical space (Kong, 2017).

This construction is paired with a quantum phase-space picture. The phase-space coset $Q\times \mathbb{R}$6 is realized by canonical coherent states

$Q\times \mathbb{R}$7

which furnish an overcomplete phase-space labeling and reproduce minimal-uncertainty Gaussian wavefunctions centered at $Q\times \mathbb{R}$8. The contrast is explicit: quantum noncommutativity resides in $Q\times \mathbb{R}$9, not in $H=L^2(\mathbb{R}^3),$0, which remain zero in the nonrelativistic model (Kong, 2017).

The classical limit is implemented as an Inönü–Wigner contraction. With $H=L^2(\mathbb{R}^3),$1, $H=L^2(\mathbb{R}^3),$2, and $H=L^2(\mathbb{R}^3),$3,

$H=L^2(\mathbb{R}^3),$4

In this limit, the central extension decouples, operator commutators divided by $H=L^2(\mathbb{R}^3),$5 reduce to Poisson brackets, coherent states become mutually orthogonal, projective Hilbert space reduces to six-dimensional classical phase space, and the configuration-space submanifold reduces to Newtonian $H=L^2(\mathbb{R}^3),$6 with Hamilton’s equations (Kong, 2017). A related presentation emphasizes that the quantum configuration space is not fundamentally a finite-dimensional manifold but a representation-theoretic construction whose classical manifold interpretation emerges only after contraction (Chew et al., 2016).

3. Bundle geometry, Weyl geometry, and relational formulations

A different reconstruction starts from bundle geometry. In the relational bundle formulation of nonrelativistic many-particle mechanics, configuration space-time is a principal bundle $H=L^2(\mathbb{R}^3),$7 with base the Newtonian time line and fiber $H=L^2(\mathbb{R}^3),$8, with structure group $H=L^2(\mathbb{R}^3),$9 acting by translations on the fibers. The wavefunction is a $|\mathbf{x}\rangle$0-valued cocyclic tensorial $|\mathbf{x}\rangle$1-form, the action functional defines a flat cocyclic connection

$|\mathbf{x}\rangle$2

and the cocyclic covariant constancy condition $|\mathbf{x}\rangle$3 yields both the operator identification $|\mathbf{x}\rangle$4 and the many-particle Schrödinger equation. The Dirac–Feynman path integral then appears as the holonomy or parallel transport kernel of this flat cocyclic connection (François et al., 3 Jan 2025).

The same framework gives a relational reformulation through the dressing field method. Choosing a particle position as dressing field, $|\mathbf{x}\rangle$5, removes the unphysical diagonal translation sector and produces a relational configuration bundle $|\mathbf{x}\rangle$6. The dressed wavefunction $|\mathbf{x}\rangle$7 becomes a basic cocyclic $|\mathbf{x}\rangle$8-form on the quotient bundle, and the relational Schrödinger equation is written in coordinates $|\mathbf{x}\rangle$9. Different choices of dressing field are related by unitary $\hat{\mathbf{x}}=(\hat{x}_1,\hat{x}_2,\hat{x}_3)$0-valued cocycles, interpreted as covariance under changes of physical reference frame (François et al., 3 Jan 2025).

Trajectory–Weyl theory places the quantum problem on configuration space equipped with a Weyl geometry $\hat{\mathbf{x}}=(\hat{x}_1,\hat{x}_2,\hat{x}_3)$1. The action augments a many-systems classical ensemble action by a curvature term $\hat{\mathbf{x}}=(\hat{x}_1,\hat{x}_2,\hat{x}_3)$2. Variation with respect to the Weyl connection yields

$\hat{\mathbf{x}}=(\hat{x}_1,\hat{x}_2,\hat{x}_3)$3

so the geometry becomes integrable and determined by the ensemble density $\hat{\mathbf{x}}=(\hat{x}_1,\hat{x}_2,\hat{x}_3)$4. The resulting Hamilton–Jacobi equation contains

$\hat{\mathbf{x}}=(\hat{x}_1,\hat{x}_2,\hat{x}_3)$5

which reproduces the Bohm quantum potential once $\hat{\mathbf{x}}=(\hat{x}_1,\hat{x}_2,\hat{x}_3)$6. To address Wallstrom’s objection, the theory extends the total derivative in the action by an $\hat{\mathbf{x}}=(\hat{x}_1,\hat{x}_2,\hat{x}_3)$7-valued contribution $\hat{\mathbf{x}}=(\hat{x}_1,\hat{x}_2,\hat{x}_3)$8, yielding quantized circulation

$\hat{\mathbf{x}}=(\hat{x}_1,\hat{x}_2,\hat{x}_3)$9

without postulating a physical wavefunction (Roser, 2015).

A related geometric formulation treats quantum mechanics as a time-dependent map $$\langle \mathbf{x}|\mathbf{x}'\rangle=\delta^{(3)}(\mathbf{x}-\mathbf{x}'), \qquad \int_{\mathbb{R}^3} d^3x\, |\mathbf{x}\rangle\langle \mathbf{x}|=\mathbb{I},$$0, where configuration space is the source manifold and the target is a two-dimensional flat manifold. The area function on the target defines a density on configuration space, the pull-back of the rotational Killing one-form defines the current, and the guidance law is

$$\langle \mathbf{x}|\mathbf{x}'\rangle=\delta^{(3)}(\mathbf{x}-\mathbf{x}'), \qquad \int_{\mathbb{R}^3} d^3x\, |\mathbf{x}\rangle\langle \mathbf{x}|=\mathbb{I},$$1

For the specific choice $$\langle \mathbf{x}|\mathbf{x}'\rangle=\delta^{(3)}(\mathbf{x}-\mathbf{x}'), \qquad \int_{\mathbb{R}^3} d^3x\, |\mathbf{x}\rangle\langle \mathbf{x}|=\mathbb{I},$$2, the map equation becomes linear in Cartesian target coordinates and reduces to the Schrödinger equation; for $$\langle \mathbf{x}|\mathbf{x}'\rangle=\delta^{(3)}(\mathbf{x}-\mathbf{x}'), \qquad \int_{\mathbb{R}^3} d^3x\, |\mathbf{x}\rangle\langle \mathbf{x}|=\mathbb{I},$$3, it reduces to a first-order classical Hamilton–Jacobi system (Goulart et al., 2019).

4. Density formulations, trajectory ensembles, and hidden variables

Configuration space is also the arena for formulations that attempt to dispense with an ontic wavefunction. Configuration-space density frameworks take as fundamental a positive density $$\langle \mathbf{x}|\mathbf{x}'\rangle=\delta^{(3)}(\mathbf{x}-\mathbf{x}'), \qquad \int_{\mathbb{R}^3} d^3x\, |\mathbf{x}\rangle\langle \mathbf{x}|=\mathbb{I},$$4, a velocity field $$\langle \mathbf{x}|\mathbf{x}'\rangle=\delta^{(3)}(\mathbf{x}-\mathbf{x}'), \qquad \int_{\mathbb{R}^3} d^3x\, |\mathbf{x}\rangle\langle \mathbf{x}|=\mathbb{I},$$5, and a current $$\langle \mathbf{x}|\mathbf{x}'\rangle=\delta^{(3)}(\mathbf{x}-\mathbf{x}'), \qquad \int_{\mathbb{R}^3} d^3x\, |\mathbf{x}\rangle\langle \mathbf{x}|=\mathbb{I},$$6 on $$\langle \mathbf{x}|\mathbf{x}'\rangle=\delta^{(3)}(\mathbf{x}-\mathbf{x}'), \qquad \int_{\mathbb{R}^3} d^3x\, |\mathbf{x}\rangle\langle \mathbf{x}|=\mathbb{I},$$7, obeying

$$\langle \mathbf{x}|\mathbf{x}'\rangle=\delta^{(3)}(\mathbf{x}-\mathbf{x}'), \qquad \int_{\mathbb{R}^3} d^3x\, |\mathbf{x}\rangle\langle \mathbf{x}|=\mathbb{I},$$8

When $$\langle \mathbf{x}|\mathbf{x}'\rangle=\delta^{(3)}(\mathbf{x}-\mathbf{x}'), \qquad \int_{\mathbb{R}^3} d^3x\, |\mathbf{x}\rangle\langle \mathbf{x}|=\mathbb{I},$$9 is written as gradients of a phase field $\psi(\mathbf{x})=\langle \mathbf{x}|\psi\rangle$0, these equations reproduce the Madelung form of Schrödinger dynamics and the Bohmian guiding law. However, standard quantum predictions are recovered only if one imposes the Wallstrom condition

$\psi(\mathbf{x})=\langle \mathbf{x}|\psi\rangle$1

When that condition is relaxed, the framework admits non-quantized circulation around nodes, continuous families of stationary states, and non-quantized angular-momentum-like behavior. The paper’s central numerical result is that the non-quantumness parameter $\psi(\mathbf{x})=\langle \mathbf{x}|\psi\rangle$2 remains equal to its initial value $\psi(\mathbf{x})=\langle \mathbf{x}|\psi\rangle$3 in the tested evolutions, which suggests that a purely dynamical justification of the Wallstrom condition is unlikely (Roser et al., 2023).

A distinct wavefunction-free program represents a quantum state by an $\psi(\mathbf{x})=\langle \mathbf{x}|\psi\rangle$4-parameter family of real trajectories $\psi(\mathbf{x})=\langle \mathbf{x}|\psi\rangle$5 on configuration space, with the density determined by the Jacobian of the map $\psi(\mathbf{x})=\langle \mathbf{x}|\psi\rangle$6. In one dimension, the time-dependent dynamics becomes a perturbed Newton equation in which the quantum term depends on derivatives with respect to the trajectory label $\psi(\mathbf{x})=\langle \mathbf{x}|\psi\rangle$7; in many dimensions and on general Riemannian configuration spaces, the action can be written covariantly in terms of the Jacobian and the metric $\psi(\mathbf{x})=\langle \mathbf{x}|\psi\rangle$8. The formalism reproduces the continuity equation and the Madelung quantum term when recast in terms of $\psi(\mathbf{x})=\langle \mathbf{x}|\psi\rangle$9 and $\hat{\mathbf{x}}$0, but it is formulated without wavefunctions or complex amplitudes (Schiff et al., 2012).

Quantum Analytical Mechanics, by contrast, explicitly supplements Hilbert-space quantum mechanics with hidden configuration variables and stochastic trajectories. For particle systems the configuration space is $\hat{\mathbf{x}}$1, or $\hat{\mathbf{x}}$2 when orientation variables are included. The forward and backward Itô equations are driven by the current velocity $\hat{\mathbf{x}}$3 and the osmotic velocity $\hat{\mathbf{x}}$4, while the ensemble dynamics reproduces the Schrödinger equation and preserves $\hat{\mathbf{x}}$5 by equivariance. This makes measurement a dynamical process in configuration variables—positions and orientations—rather than an operation on abstract Hilbert space alone (Paul, 26 Nov 2025).

5. Noncommutative configuration spaces

The nonrelativistic model in which physical space is identified with configuration space retains commuting spatial coordinates, $\hat{\mathbf{x}}$6, and places noncommutativity in the position–momentum sector. A substantial literature nonetheless studies deformations in which configuration space itself is noncommutative (Kong, 2017).

One such model quantizes two harmonic oscillators on a noncommutative plane with

$\hat{\mathbf{x}}$7

Using anti-Wick quantization, the analysis shows that removing configuration-space noncommutativity, $\hat{\mathbf{x}}$8, and taking the classical limit, $\hat{\mathbf{x}}$9, are not commuting operations. If $Q=\mathbb{R}^{3N}$00 is taken first, one recovers the ordinary phase-space description on $Q=\mathbb{R}^{3N}$01; if $Q=\mathbb{R}^{3N}$02 is taken first, the effective theory collapses to a configuration-space description on $Q=\mathbb{R}^{3N}$03, and the subsequent $Q=\mathbb{R}^{3N}$04 limit cannot reconstruct the lost phase-space structure (Benatti et al., 2013).

A different route derives noncommutativity in configuration space from a conjugate magnetic field in momentum space. With a momentum-space vector potential $Q=\mathbb{R}^{3N}$05 and field strength $Q=\mathbb{R}^{3N}$06, quantization yields

$Q=\mathbb{R}^{3N}$07

For constant nondegenerate $Q=\mathbb{R}^{3N}$08, the coordinates satisfy $Q=\mathbb{R}^{3N}$09, and the coordinate algebra can be organized into Fock pairs. In that setting, canonical coherent states associated with the noncommutative coordinates are the localized analogues of sharp points, and the Hilbert space can be identified with the enveloping algebra of the position operators (Govaerts, 2024).

Space-time noncommutativity has also been treated in $Q=\mathbb{R}^{3N}$10-dimensional Moyal space-time with

$Q=\mathbb{R}^{3N}$11

States are represented as Hilbert–Schmidt operators on an auxiliary Fock space, and coherent-state symbols lead to an effective commutative Schrödinger equation

$Q=\mathbb{R}^{3N}$12

where the Voros star product is required to obtain a positive probability density $Q=\mathbb{R}^{3N}$13 and a consistent induced inner product on equal-time slices (Nandi et al., 2017).

For the Coulomb problem, noncommutative configuration space can be realized by fuzzy $Q=\mathbb{R}^{3N}$14, with

$Q=\mathbb{R}^{3N}$15

The resulting noncommutative Schrödinger-like equation for hydrogen is exactly solvable for bound states and scattering. The attractive Coulomb branch yields $Q=\mathbb{R}^{3N}$16-deformed hydrogenic energies that reduce to the usual spectrum as $Q=\mathbb{R}^{3N}$17, while a second branch produces bound states for repulsive potential at ultra-high energies. Those states leave the Hilbert space in the commutative limit (Gáliková et al., 2015).

6. Curved, compact, intrinsic, and alternative configuration manifolds

Configuration space need not be Euclidean or noncompact. On a general Riemannian configuration space $Q=\mathbb{R}^{3N}$18 with metric $Q=\mathbb{R}^{3N}$19, the coordinate basis resolves the identity with the measure $Q=\mathbb{R}^{3N}$20, the canonical momenta are

$Q=\mathbb{R}^{3N}$21

and the Wigner–Weyl–Moyal formalism can be extended to the phase space $Q=\mathbb{R}^{3N}$22. The resulting Wigner function satisfies the expected marginals, any Weyl-ordered operator acquires a phase-space symbol, and the quantum Liouville equation reduces to the classical Liouville equation on $Q=\mathbb{R}^{3N}$23 in the semiclassical limit (Gneiting et al., 2013).

Compact configuration spaces admit still different quantizations. Polymer quantum mechanics imposes a discrete topology on configuration space and produces a non-regular representation of the Weyl algebra. On a ring, the polymer configuration space is $Q=\mathbb{R}^{3N}$24 and the Hilbert space reduces, for a regular $Q=\mathbb{R}^{3N}$25-site lattice, to the finite-dimensional graph space $Q=\mathbb{R}^{3N}$26. The momentum operator does not exist because the shift operators are not weakly continuous, so dynamics is expressed in terms of holonomies and a discrete Laplacian. Exact energy eigenvalues and eigenfunctions can be obtained for the ring and the box, and under graph refinement the eigenfunctions converge uniformly to their continuum Schrödinger counterparts (Siebersma et al., 4 Jun 2026).

An even more radical alternative takes the configuration space to be a compact Lie group such as $Q=\mathbb{R}^{3N}$27 or $Q=\mathbb{R}^{3N}$28. In “intrinsic quantum mechanics,” the Hilbert space is $Q=\mathbb{R}^{3N}$29 with Haar measure, the kinetic operator is the Laplace–Beltrami or Casimir operator on $Q=\mathbb{R}^{3N}$30, and left-invariance in intrinsic space is interpreted as local gauge invariance in laboratory space. The framework uses periodic intrinsic potentials and toroidal coordinates on the maximal torus, and it is applied to proton structure, baryon spectra, electroweak mixing, and related observables (Trinhammer, 2017).

Alternative computational organizations of standard configuration-space quantum mechanics also exist. One formulation expands $Q=\mathbb{R}^{3N}$31 in a complete square-integrable basis $Q=\mathbb{R}^{3N}$32,

$Q=\mathbb{R}^{3N}$33

where the coefficients $Q=\mathbb{R}^{3N}$34 are orthogonal polynomials in energy. The tridiagonal representation of the Hamiltonian implies a three-term recurrence relation, and scattering amplitudes, phase shifts, and bound-state spectra are extracted from the asymptotics and orthogonality measure of the relevant polynomial family. This enlarges the class of exactly solvable nonrelativistic problems beyond those solved directly in coordinate space (Alhaidari, 2017).

7. Classical continuity, reconstructed space-time, and conceptual disputes

Recent work has also proposed replacing the standard position basis by a basis of pairwise distinguishable position–velocity states $Q=\mathbb{R}^{3N}$35. In this “quantum mechanics in configuration space” based on a physically motivated quantization of Newtonian mechanics, $Q=\mathbb{R}^{3N}$36 forms the Hilbert-space basis, $Q=\mathbb{R}^{3N}$37 and $Q=\mathbb{R}^{3N}$38 commute, and the independent generators are $Q=\mathbb{R}^{3N}$39 and the “acceleratum” $Q=\mathbb{R}^{3N}$40, with

$Q=\mathbb{R}^{3N}$41

For the free particle, the dynamical generator is $Q=\mathbb{R}^{3N}$42, so

$Q=\mathbb{R}^{3N}$43

and the wavefunction obeys the transport equation $Q=\mathbb{R}^{3N}$44. The stated aim is to increase the continuity between quantum and classical mechanics and to avoid the conceptual inconsistency that standard canonical quantization attaches to the word “momentum,” namely its conflation with both spatial-translation generator and mass times velocity (Bukhari et al., 16 Jun 2026).

A different conceptual program attempts a reconstruction of space-time itself from Hilbert-space data. In that account, internal observers have access only to detector clicks, and configuration-space labels are operational products of synchronization and alignment procedures. The core transcription is measurement-theoretic: a state $Q=\mathbb{R}^{3N}$45 is mapped to a probability measure on a label space $Q=\mathbb{R}^{3N}$46 by

$Q=\mathbb{R}^{3N}$47

with $Q=\mathbb{R}^{3N}$48 a POVM defined by the operationally constructed detector network. The paper argues that separable states can operationalize Poincaré–Einstein radar synchronization, whereas entangled shares, despite satisfying parameter independence, cannot by themselves define classical space-time scales because their correlations are outcome-dependent and relational. This suggests that configuration space, on that view, is not presupposed but reconstructed from Hilbert-space structure through “it-from-click imaging” (Svozil, 2023).

Taken together, these lines of research show that “quantum mechanics in configuration space” is not a single doctrine. In the standard theory it is the familiar $Q=\mathbb{R}^{3N}$49 formulation of nonrelativistic dynamics. In symmetry-based approaches it becomes the homogeneous space singled out by the quantum relativity group. In geometric approaches it is the base of cocyclic bundles, Weyl-curved ensembles, or Schrödinger maps. In hydrodynamic and hidden-variable approaches it carries densities, velocities, stochastic drifts, or trajectory ensembles. In deformed theories it can be discrete, compact, curved, Lie-group-valued, or noncommutative. The common feature is that the physically relevant variables are organized on configuration space itself, while the principal disputes concern whether that space is fundamental, emergent, relational, or only one representation among several mathematically equivalent descriptions.