Gautschi’s conjecture

For every integer n1n\ge 1, every 1<α<β-1<\alpha<\beta, and every 0<c10<c\le 1, let πn\pi_n be the monic degree-nn polynomial orthogonal on [c,c][-c,c] with respect to the Jacobi weight (1x)α(1+x)β(1-x)^\alpha(1+x)^\beta. Then

[πn(c)πn(c)]2(1c1+c)βα<1.\left[\frac{\pi_n(-c)}{\pi_n(c)}\right]^2\left(\frac{1-c}{1+c}\right)^{\beta-\alpha}<1.

Progress summary

Partially solved

A new unrefereed manuscript proves large portions of the conjecture, but a negative-parameter region remains open.

Gautschi’s 2018 conjecture asks whether a strict inequality for monic Jacobi orthogonal polynomials holds for every degree when 1<α<β-1<\alpha<\beta and 0<c10<c\leq 1. The unresolved issue is uniform validity across all admissible parameters and degrees.

Known results

  • Gautschi’s recorded mixed-sign result and Milovanović’s published criterion together cover the full range β0\beta\geq 0.

August 2026 partial resolution

The manuscript claims all degrees for 0<α<β0<\alpha<\beta, and in the negative wedge claims all degrees when c23/(3+r)c^2\leq 3/(3+r), hence for every admissible pair when 0<c3/20<c\leq\sqrt{3}/2. It also claims degree one generally and eventual validity for each fixed (α,β,c)(\alpha,\beta,c). The region α<β<0\alpha<\beta<0 with c>3/2c>\sqrt{3}/2 remains open uniformly in degree.

Current status (as of August 2026): The conjecture has substantial claimed partial coverage, including the full β0\beta\geq 0 range, but the negative-wedge case with c>3/2c>\sqrt{3}/2 is not settled uniformly in degree.

Sources
Sources & referencesView supporting material

Primary source

arXiv

Additional references

Solutions 0

No solutions have been posted yet.