Selberg's eigenvalue conjecture
Generated by GPT-5-miniExpansion Funnel Raw 78 → Dedup 0 → NER 0 → Enqueued 0
| Selberg's eigenvalue conjecture | |
|---|---|
| Name | Atle Selberg |
| Birth date | 1917–2007 |
| Field | number theory, harmonic analysis, spectral theory |
| Notable work | Selberg trace formula, Selberg zeta function |
Selberg's eigenvalue conjecture is a conjecture about the spectral gap for the Laplace operator on certain arithmetic quotients of the hyperbolic plane, predicting a lower bound for nonzero eigenvalues. It links ideas from Atle Selberg, the Selberg trace formula, modular forms, and the analytic theory of automorphic forms on PSL(2,R), and has driven research connecting Ergodic theory, representation theory, and analytic number theory.
Statement of the conjecture
The conjecture asserts that for congruence subgroups of SL(2,Z) acting on the hyperbolic plane, every nontrivial eigenvalue λ of the Laplace–Beltrami operator satisfies λ ≥ 1/4. This prediction is phrased in terms of the spectral parameter s with λ = s(1−s), so the claim is equivalent to excluding exceptional eigenvalues with Re(s) > 1/2 for the automorphic Laplacian associated to arithmetic quotients such as Γ\H for Γ a congruence subgroup of SL(2,Z), Γ0(N), or Γ1(N). The statement connects the absence of small eigenvalues to expected Ramanujan-type bounds arising from the Ramanujan–Petersson conjecture and the theory of cuspidal automorphic representations of GL(2).
Historical background and motivation
The conjecture originates in work of Atle Selberg in the 1960s, informed by his development of the Selberg trace formula and the introduction of the Selberg zeta function; contemporaries included Hans Maass, Ilya Piatetski-Shapiro, and Harish-Chandra. Selberg proposed the 1/4 bound motivated by parallels with the spectrum of the Laplacian on closed Riemann surfaces and analogies to eigenvalue bounds studied by Weyl and Krein. The conjecture is entwined with the Langlands program, the Jacquet–Langlands correspondence, and expectations from automorphic representation theory articulated by Robert Langlands and further explored by Atle Selberg and Harish-Chandra.
Known results and partial progress
Partial bounds toward the conjecture have been achieved via analytic, representation-theoretic, and arithmetic methods. Selberg himself proved lower bounds in special settings; later improvements came from work of H. Iwaniec, Peter Sarnak, Henryk Iwaniec, Jean Bourgain, Péter Sarnak (note: distinct contributors often collaborated), and Jianya Liu. Progress includes explicit numerical spectral gaps for congruence subgroups of SL(2,Z), effective bounds via the large sieve, and results conditional on automorphic lifting theorems of Gelbart–Jacquet and the Shimura correspondence. The Ramanujan conjecture for GL(2) over number fields implies the full 1/4 bound; conversely, improvements on the spectral gap have been deduced from deep advances in automorphic L-functions and the Rankin–Selberg method. Exceptional examples over nonarithmetic Fuchsian groups constructed by Hideo Shimura and arithmeticity results by Margulis have clarified limits of generalization.
Methods and techniques used in proofs
Techniques include harmonic analysis via the Selberg trace formula, analytic continuation of Eisenstein series building on Langlands and Mœglin–Waldspurger frameworks, and representation-theoretic approaches leveraging unitary dual classifications for PSL(2,R) by Harish-Chandra and Knapp. Sieve methods such as the large sieve inequality and spectral reciprocity tools from Kuznetsov trace formula and Petersson trace formula play central roles. Advances in bounding Fourier coefficients use Rankin–Selberg convolution and subconvexity bounds for L-functions developed by Duke, Friedlander, and Iwaniec. Techniques from the theory of expander graphs and combinatorial constructions by Lubotzky, Phillips, and Sarnak translate spectral gap questions into combinatorial contexts, while ergodic methods inspired by Margulis and Eskin provide dynamical insights. Representation-theoretic lifting via Jacquet–Langlands and Langlands functoriality also underpin many conditional reductions.
Connections to automorphic forms and spectral theory
The conjecture is fundamentally an assertion about cuspidal automorphic spectra on arithmetic quotients and thus is linked to the classification of automorphic representations for GL(2) and the analytic behavior of Eisenstein series developed by Langlands. The bound λ ≥ 1/4 corresponds to temperedness of local components of cuspidal representations at the archimedean place, a property predicted by the Ramanujan–Petersson conjecture and formalized in the Langlands conjectures on temperedness. Relations with the Selberg zeta function and scattering theory studied by Lax and Phillips connect zeros of zeta functions to small eigenvalues. The spectral decomposition of L^2(Γ\H) intertwines continuous spectrum contributions analyzed by Eisenstein and discrete cuspidal spectra tied to Hecke operators and newform theory of Atkin–Lehner.
Implications and applications in number theory and geometry
A verified 1/4 bound would imply strong equidistribution and mixing rates for geodesic flows on arithmetic surfaces, with consequences for counting closed geodesics related to the Prime Geodesic Theorem and for growth of modular symbols studied by Goldfeld and Zagier. In analytic number theory, the conjecture impacts subconvexity problems for automorphic L-functions, bounds on Fourier coefficients of modular forms relevant to Deligne-type estimates, and distribution of Heegner points connected to work by Gross and Zagier. In spectral geometry, it informs rigidity phenomena related to Mostow rigidity and spectral gap-dependent constructions of expander graphs used in computer science and combinatorics via the Lubotzky–Phillips–Sarnak method. The conjecture thus sits at an intersection influencing research across number theory, representation theory, and differential geometry.