Wigner-Style Rigidity Theorem

• Wigner-style rigidity is a classification result showing that preserving a basic invariant, like transition probability, forces a transformation to be unitary or antiunitary.
• The theorem extends Wigner's original result to rank-n projections, orthogonality-preserving maps, and infinite-dimensional as well as combinatorial and algebraic settings.
• In random matrix theory, rigidity results yield precise eigenvalue bounds using semicircle laws and resolvent estimates, underpinning spectral universality.

Wigner-style rigidity theorems encompass powerful classification results for symmetry transformations in quantum mechanics and random matrix theory, extending Wigner’s seminal work on rank-one projections to higher rank, infinite-dimensional settings, and more general algebraic structures. At their core, these theorems show that very mild invariance properties—typically, the preservation of a single scalar invariant such as transition probability or a related spectral function—force a transformation to be implemented by one of a small set of canonical symmetries: unitary or antiunitary operators (or, in algebraic and combinatorial settings, their analogues).

1. Rigidity for Rank-$n$ Projections and Grassmann Spaces

Let $H$ be a complex Hilbert space, and fix $n\in\mathbb N$. The Grassmann space $G_n(H)$ consists of all rank-$n$ orthogonal projections $P$ on $H$, naturally identified with $n$-dimensional subspaces via $P\mapsto \mathrm{Im}\,P$.

The central invariant is the transition probability between two projections, given by

$$\mathrm{tr}(P Q) = \sum_{k=1}^n \cos^2(\theta_k(P,Q)),$$

where the $\theta_k(P,Q)$ are the Jordan principal angles defined by the singular values of the operator $P Q$ acting between $\mathrm{Im}\,Q$ and $\mathrm{Im}\,P$.

Main rigidity theorem: If $\phi: G_n(H)\to G_n(H)$ (not necessarily bijective) preserves $\mathrm{tr}(P Q)$ for every pair $P,Q\in G_n(H)$, then $\phi$ is implemented either by a unitary or antiunitary operator $U$ via $\phi(P)=U P U^*$, or by $\phi(P)=U(I-P)U^*$ if $\dim H = 2n$ (Gehér, 2017). The proof extends $\phi$ to a Hilbert-Schmidt isometry on the space of finite-rank self-adjoint operators and leverages a combination of projection orthogonality, adjacency relations, and Chow’s theorem to force linear or conjugate-linear isometries.

This rigid classification result generalizes both Wigner's original theorem for rank-one projections and Molnár’s principal angle characterizations.

2. Rank-One Transition Probability and Orthogonality Rigidity

For the set $\mathcal{P}(H)$ of rank-one projections (projective space), Wigner’s original theorem classifies bijections preserving transition probability. Recent developments have relaxed bijectivity and strengthened the connection with orthogonality.

Given $\phi:\mathcal{P}(H)\to \mathcal{P}(H)$ that preserves orthogonality in one direction and maps a single complete orthonormal system (COSP) to another COSP, Šemrl’s optimal rigidity theorem ensures that $\phi$ is implemented by a unitary or antiunitary operator (Semrl, 2021). The proof employs Gleason’s theorem, which ties additive frame functions to density operators, and proceeds by showing that the preservation of transition probability suffices to force the symmetry to be a Wigner symmetry. This result is optimal: further weakening (dropping even the COSP condition) allows "wild" transformations not induced by unitaries or antiunitaries.

In finite dimensions, any orthogonality-preserving transformation (not just bijections) is likewise induced by a unitary or antiunitary map (Pankov et al., 2020). The projective-geometric method lifts orthogonality-preserving maps to semilinear isometries, with the underlying field automorphism forced to be either the identity or complex conjugation.

3. Infinite-Dimensional, Combinatorial, and Algebraic Generalizations

Infinite-dimensional extensions involve sophisticated adjacency relations. In the infinite Grassmannian $\mathcal{G}_\infty(H)$ (closed subspaces of infinite dimension and codimension), ortho-adjacency (compatibility plus adjacency) replaces ordinary orthogonality. Pankov–Tyc show that any ortho-adjacency-preserving bijection restricts on each maximal "A-component" to a map implemented by a unitary or antiunitary operator (possibly followed by orthocomplementation) (Pankov et al., 2023).

In the context of orthomodular lattices and infinite sets, Harding demonstrates that the automorphism group of the poset of direct product decompositions, $\mathrm{Fact}(X)$, for an infinite set $X$, is precisely the permutation group of $X$ (Harding, 2016). This combinatorial rigidity mirrors the quantum case, with automorphisms corresponding to underlying set symmetries.

For C$^*$-algebras, Landsman–Rang clarify that the sharpest Wigner-style theorem is the correspondence between pure-state transition probability preserving bijections (Wigner symmetries) and Jordan $^*$-automorphisms, rather than unitary or antiunitary implementability. In general, there is no intrinsic extension to unitary-implemented automorphisms beyond type I factors (e.g. $\mathcal{B}(H)$) (Landsman et al., 2019).

4. Random Matrix Theory: Eigenvalue Rigidity for Wigner Ensembles

In random matrix theory, "Wigner-style rigidity theorem" refers to precise probabilistic bounds on the deviation of eigenvalues from their classical locations in Wigner matrices and deformations thereof.

For generalized Wigner matrices, eigenvalues $\lambda_j$ are tightly concentrated near their quantiles $\gamma_j$ under the semicircle law,

$$|\lambda_j - \gamma_j| \lesssim (\log N)^L \left[\min(j,N-j+1)\right]^{-1/3} N^{-2/3}$$

with overwhelming probability, achieving the optimal Tracy–Widom edge scaling and $N^{-1}$ bulk rigidity (Erdos et al., 2010). Recent results improve this to optimal $(\log N)/N$ bulk rigidity with explicit Gaussian tails (Bourgade et al., 2023). Proofs utilize strong local semicircle laws, resolvent estimates, and coupling and homogenization via Dyson Brownian motion, together with Gaussian multiplicative chaos techniques.

For deformed Wigner matrices $H = W + \lambda V$, with $V$ diagonal and possessing fast edge decay, the rigidity theorem adapts to the density’s edge behavior: $$|\lambda_i - \gamma_i| \lesssim N^{-\varepsilon'}\,i^{-1/(b+1)},$$ where $b$ governs the edge power-law decay rate of the spectral density (Lee et al., 2021). Rigidity scales interpolate between bulk and edge regimes, driven by the free-convolution law.

5. Projective and Contextual Variants: Frame and Tuple Rigidity

Extensions to frame bundles and line tuples are encapsulated in the rigidity theorem for symmetric, contextually constrained maps between manifolds of ordered lines and frames. Maps from $\mathbb{F}^\perp(\mathbb{C}^n)$ (ordered orthogonal $n$-frames) to $\mathbb{F}(\mathbb{C}^n)$ (ordered linearly independent $n$-frames) that preserve $\pi$-linkage for all partitions and are symmetric are exactly those induced by semilinear injections, and thus by linear or conjugate-linear maps when measurability is imposed (Chirvasitu, 16 Jan 2026). The only other possibility, in a more general context, is a "contextual-global symmetry" mapping each line $(\ell_i)$ to the orthogonal complements of the span of the other lines, which fundamentally depends on the collective configuration.

6. Conceptual Synthesis and Physical Significance

Wigner-style rigidity reflects the phenomenon that quantum symmetries, geometric or probabilistic, are determined by the preservation of numerically or structurally minimal invariants—transition probability, orthogonality, adjacency, or spectral position. Once such a property is fixed for all pairs, the allowed transformations must coincide globally with the canonical symmetry group (unitary/antiunitary operators for Hilbert spaces; permutations for sets; Jordan automorphisms for C$^*$-algebras). In physical terms, this underlies why quantum symmetry operations are fundamentally constrained and why universality phenomena arise in the spectrum of large random matrices.

The synthesis shows that these rigidity phenomena persist in finite, infinite, algebraic, combinatorial, and probabilistic contexts, sometimes with novel exceptions in higher-level structures (contextual symmetries, purely algebraic automorphisms). The foundational implications extend across quantum mechanics, functional analysis, and random matrix theory, confirming the central role of symmetry and invariance principles in determining structure and statistics.