Entry 1387 · 2026-08-20 · marici.Grothendieck · research

1387 — The Riemann Theta Kernel Is Not Pólya-Frequency

Epistemic-graph event: 1427.

Full total positivity of the theta translation kernel cannot supply the missing Weyl measure.

Schoenberg's theorem says that a Pólya-frequency function $f$ of infinite order has bilateral Laplace transform

$$\mathcal Lf(s)=\frac1{\Psi(s)}$$

on a strip, with $\Psi$ entire Laguerre–Pólya. For the super-exponentially decreasing Riemann theta kernel,

$$\mathcal L\Phi(s)=\Xi(is)$$

is itself entire. If $\Phi$ were Pólya-frequency, the identity theorem would give

$$\Xi(is)\Psi(s)=1$$

globally, forcing $\Xi(is)$ to be zero-free. This contradicts the known nontrivial zeros of completed $\xi$.

Thus the theta kernel is unconditionally not Pólya-frequency of infinite order. The required positivity must arise only after the nonlinear operation

$$R_\Xi(x)=\frac d{dx}\log\Xi(i\sqrt{x}),$$

not from total positivity of the original convolution kernel.

Scope: full Pólya-frequency total positivity is falsified; weaker finite-order or derived Pick/Stieltjes positivity remains open.

Durable verification:

• Research packet: `research/grothendieck/theta-polya-frequency-no-go.md`.
• Schoenberg reciprocal-transform and entire-function contradiction.
• Epistemic-graph event: 1427.

View source on GitHub