The Riemann Hypothesis has been open for 167 years. It asserts that every non-trivial zero of the Riemann zeta function ζ(s) lies on the critical line Re(s) = ½. The functional equation ξ(s) = ξ(1−s) tells you that zeros come in pairs reflected across this line. What it does not tell you — and what nobody has proved — is that zeros cannot appear off the line in symmetric pairs with Re(s) ≠ ½.
Holding the Line addresses this from the direction of Haar measure. The argument starts not with the zeta function but with the scale group (ℝ+, ×), whose Haar measure is dμ(x) = dx/x. This measure has one fundamental property:
Under the Mellin transform, inversion x ↦ x−1 acts on the spectral variable as s ↦ 1−s. The functional equation of ξ(s) is precisely this Haar inversion restricted to the arithmetic spectrum — a fact made precise through Tate's 1950 thesis, which reformulates ξ(s) as a Fourier transform on the adèle ring. The functional equation is not a property of the zeta function that happens to hold. It is the arithmetic expression of a measure-theoretic identity.
The Plancherel Constraint
The Mellin transform on L²(ℝ+, dx) is unitary precisely on the spectral line Re(s) = ½. This is the half-density Plancherel theorem for the scale group, and it has nothing to do with primes or zeros: it is a theorem of harmonic analysis that holds for any function in the relevant Hilbert space. Under the celestial normalization Δ = 2s used throughout the Shadow Framework, this is the principal series Re(Δ) = 1 — the unitarity locus of the celestial conformal field theory.
The compact complement of the idèle class group, K = 𝔸×/(ℚ× × ℝ+×), is compact. Peter–Weyl theory for compact groups forces its spectrum to be discrete. Combined with the Plancherel normalization for the noncompact ℝ+ factor, the full adelic spectral structure selects Re(s) = ½ as the unique admissible locus for square-integrable characters.
What Is and Is Not Proved
The paper establishes the admissible spectral support. The unitary axis is Re(s) = ½, determined by Haar measure before any arithmetic is invoked. The functional equation of ξ(s) is the restriction of this measure-theoretic identity to the arithmetic spectrum. These results are unconditional.
What they do not yet prove is that the non-trivial zeros of ζ are spectral data of the scale Hilbert space — that is, the Hilbert–Pólya admissibility condition. If they are, Re(s) = ½ for all of them follows immediately from the self-adjointness of the dilation generator. That admissibility step is the remaining gap, and it is stated precisely rather than papered over. The Haar mechanism is not a gap; the connection from that mechanism to the specific zeros is.
The line Re(s) = ½ is not a geometric accident of the critical strip. It is the unitary axis of the scale group, the unique locus preserved by Haar inversion, and the fixed set of time reversal acting on the celestial principal series.
Holding the Line was the first paper in the GPP to establish this chain of reasoning cleanly, and its results are incorporated in full into the Riemann Hypothesis page and the main monograph.