A conjectural statistical statement about the local spacing distribution of zeros of the Riemann zeta function on the critical line: roughly, the scaled pair correlation of high zeros is predicted to match that of eigenvalues of large random Hermitian matrices from the Gaussian Unitary Ensemble (GUE).

Principle

At high heights, the fine-scale statistics of zeta zeros resemble universal random-matrix statistics because both systems reflect a common spectral repulsion mechanism; the pair correlation captures two-point spacing information after normalization by average density.

Demonstration

Numerical computations of zeros of the zeta function at very large heights show histograms of normalized spacings whose two-point correlation function closely follows the GUE formula 1 - (sin(pi x)/(pi x))^2, providing empirical support for the conjecture.

Misapplication

Treating Montgomery's pair correlation as an exact identity for finite ranges or for arbitrary L-functions without accounting for symmetry class, conductor, or lower-order terms can lead to erroneous inferences about individual zero positions or prime correlations.

Consequence

If true, it gives strong probabilistic predictions about zeros spacing (hence about moments and value distribution of L-functions) and supports heuristics linking primes, zero statistics, and random matrix models used throughout analytic number theory.

Reversal

The opposite extreme is Poisson (uncorrelated) spacing where zeros behave like independent random points; contrasting Poisson with GUE underscores the presence or absence of spectral repulsion and long-range rigidity.

Boundary

The conjecture is an asymptotic statement about high imaginary parts of zeros, typically conditional on the Riemann Hypothesis and appropriate averaging; it does not assert precise finite-height identities nor does it apply identically across different symmetry types of L-functions.

Semantic Tension

Sits in tension with alternative models such as Poisson statistics or arithmetic-induced anomalies at small heights; it competes conceptually with deterministic approaches to zero distribution and with results that emphasize arithmetic structure over universality.

Synthesis

Montgomery Pair Correlation posits that the two-point spacing statistics of high zeta zeros, after normalization by mean density, follow the same universal law as eigenvalues of large GUE matrices, thereby connecting analytic number theory with random matrix universality and providing a probabilistic framework for zero spacing phenomena.