On the Architectural Boundaries of the Hilbert–Pólya Program:
Adèlic Self-Duality, Theta Correspondence, Scattering, and the Problem of Autonomous Dynamics

SRFP311T1 Collaboration

September 2026

Abstract

We identify structural and analytic boundaries within the Hilbert–Pólya program by distinguishing established harmonic-analytic structures from the open problem of constructing an autonomous arithmetic dynamics. We prove that individual Euler factors cannot serve as physical perturbation determinants of self-adjoint operator pairs on the physical spectral sheet. We further establish a divisor obstruction proving that finite Euler products cannot be completed by meromorphic factors whose zeros and poles are confined to finitely many horizontal lines (including all standard Archimedean Gamma factors) so as to reproduce the reflection symmetry of the completed zeta function.

We then establish two bounded geometric obstruction propositions: (i) compact cross-section flows with a closed 11-form ω\omega satisfying ω(X)≡1\omega(X)\equiv 1 cannot realize the prime logarithms as primitive periods due to the finite ℚ\mathbb{Q}-rank of H1(M,ℤ)H_1(M,\mathbb{Z}); and (ii) finite-area hyperbolic surfaces cannot possess closed geodesics with lengths log⁡p\log p because the traces of a finitely generated Fuchsian group generate a field extension of finite algebraic degree over ℚ\mathbb{Q}, contradicting the infinite degree of multiquadratic extensions ℚ({p})\mathbb{Q}(\{\sqrt{p}\}). We explicitly delineate the boundaries of these results: they rule out specific classical geometric mechanisms but do not exclude every finite-dimensional smooth flow.

We also analyze automorphic scattering. Scattering theory does not evade the physical-sheet self-adjointness obstruction: its non-real poles and zeros occur through meromorphic continuation to an unphysical sheet. Nevertheless, relative traces and scattering determinants on purely scalar L2L^2 spaces establish that a negative prime-power contribution does not logically necessitate a ℤ2\mathbb{Z}_2-graded super-trace. Automorphic scattering provides a rigorous analytic mechanism for signs and resonances, not a Hilbert–Pólya operator for ζ\zeta.

Finally, we analyze smooth projective curves over finite fields as a rigorous characteristic-pp consistency benchmark. The discrete normalized Frobenius operator has unitary spectrum, the Grothendieck–Lefschetz trace formula supplies the alternating signs, and the cohomological determinant recovers the zeta function. This confirms the algebraic coherence of the spectral/determinantal architecture where geometry is established, while leaving the construction of an autonomous characteristic-zero dynamics over ℚ\mathbb{Q} strictly open.

Introduction and Problem Formulation

The Hilbert–Pólya program seeks a self-adjoint operator whose spectrum encodes the imaginary parts of the non-trivial zeros of the Riemann zeta function. Let Λ(s)≔π−s/2Γ(s/2)ζ(s),ξ(s)≔12s(s−1)Λ(s),Ξ(E)≔ξ(12+iE).\begin{align} \Lambda(s) &\coloneqq \pi^{-s/2}\Gamma(s/2)\zeta(s),\\ \xi(s) &\coloneqq \frac12 s(s-1)\Lambda(s),\\ \Xi(E) &\coloneqq \xi\left(\frac12+iE\right). \end{align} A spectral realization would have the schematic form Ξ(E)=Cdet⁡reg(E−H),H=H*,\begin{equation} \Xi(E) = C\det\nolimits_{\mathrm{reg}}(E-H), \qquad H=H^*, \end{equation} with zeros of Ξ\Xi corresponding to spectral values of HH.

If such a realization exists with the determinant identity established independently of the zeros, then H=H*⇒Spec⁡(H)⊂ℝ,\begin{equation} H=H^* \quad\Longrightarrow\quad \operatorname{Spec}(H)\subset\mathbb{R}, \end{equation} and hence every non-trivial zero would have the form ρ=12+iγ,γ∈ℝ.\begin{equation} \rho=\frac12+i\gamma, \qquad \gamma\in\mathbb{R}. \end{equation} Thus self-adjointness supplies the final spectral implication required for the Riemann Hypothesis, but does not by itself construct the operator or establish the determinant identity.

In the mathematical-physics literature, related approaches have included local Schrödinger operators H=−d2dx2+V(x),\begin{equation} H=-\frac{d^2}{dx^2}+V(x), \end{equation} the semiclassical dilation Hamiltonian H=12(xp+px),\begin{equation} H=\frac12(xp+px), \end{equation} and hypothetical decompositions H=?⨁p∈ℙHp.\begin{equation} H\stackrel{?}{=}\bigoplus_{p\in\mathbb P}H_p. \end{equation} The purpose of this paper is not to rule out every conceivable Hilbert–Pólya construction, but to identify precise obstructions to simple local or finite reductions and to separate established harmonic architecture from the genuinely open dynamical step.

The architectural boundary: established harmonic analysis versus the open dynamical realization.

Local and Finite Obstructions

The Physical-Sheet Barrier

For a prime pp, write Dp(E)≔1−p−1/2−iE.\begin{equation} D_p(E) \coloneqq 1-p^{-1/2-iE}. \end{equation} Its zeros satisfy p−1/2−iE=1p^{-1/2-iE}=1, hence Ek=2πklog⁡p+i2,k∈ℤ.\begin{equation} E_k = \frac{2\pi k}{\log p} +\frac{i}{2}, \qquad k\in\mathbb{Z}. \end{equation} Thus all these zeros lie strictly in the upper half-plane.

Proposition 1 (Physical-Sheet Incompatibility). Let H0H_0 and H=H0+VH=H_0+V be self-adjoint operators on a Hilbert space ℋ\mathcal H, with VV trace class. The perturbation determinant Δ(z)≔det⁡(I+V(H0−z)−1)\begin{equation} \Delta(z) \coloneqq \det\!\left(I+V(H_0-z)^{-1}\right) \end{equation} cannot vanish for z∈ℂ\ℝz\in\mathbb{C}\setminus\mathbb{R}.

Consequently, Dp(E)D_p(E) cannot be the physical-sheet perturbation determinant of such a self-adjoint operator pair under the identification z=Ez=E.

Proof. For z∉ℝz\notin\mathbb{R}, the resolvent (H0−z)−1(H_0-z)^{-1} exists and is bounded. If Δ(z0)=0\Delta(z_0)=0, then −1-1 is an eigenvalue of V(H0−z0)−1V(H_0-z_0)^{-1}. Hence there exists a nonzero vector ϕ\phi satisfying (I+V(H0−z0)−1)ϕ=0\left(I+V(H_0-z_0)^{-1}\right)\phi=0. Setting ψ=(H0−z0)−1ϕ\psi=(H_0-z_0)^{-1}\phi, we have ψ≠0\psi\neq0 and (H0+V−z0)ψ=0(H_0+V-z_0)\psi=0. Thus z0z_0 is an eigenvalue of the self-adjoint operator HH, contradicting Spec⁡(H)⊂ℝ\operatorname{Spec}(H)\subset\mathbb{R}. ◻

The conclusion is deliberately limited to the physical perturbation determinant. It does not prohibit the meromorphic continuation of scattering quantities. Resonances are poles or zeros of analytically continued objects on non-physical sheets and therefore are not counterexamples to the proposition.

The Divisor Obstruction for Finite Euler Products

Let S={p1,…,pN}⊂ℙS=\{p_1,\ldots,p_N\}\subset\mathbb P be a non-empty finite set of primes, and define PS(s)≔∏p∈S(1−p−s)−1.\begin{equation} P_S(s) \coloneqq \prod_{p\in S}(1-p^{-s})^{-1}. \end{equation} Its reciprocal PS(s)−1=∏p∈S(1−p−s)P_S(s)^{-1}=\prod_{p\in S}(1-p^{-s}) is entire, so PS(s)P_S(s) has no zeros in ℂ\mathbb{C}. Its poles are located at sp,k=2πiklog⁡p,p∈S,k∈ℤ,\begin{equation} s_{p,k} = \frac{2\pi i k}{\log p}, \qquad p\in S,\quad k\in\mathbb{Z}, \end{equation} all lying strictly on the imaginary axis ℜ(s)=0\Re(s)=0.

Proposition 2 (Divisor Obstruction for Finite Euler Products). Let ℋm=⋃j=1m{s∈ℂ:ℑ(s)=cj}\mathcal H_m = \bigcup_{j=1}^m \{s\in\mathbb{C}: \Im(s)=c_j\} be a finite union of horizontal lines (cj∈ℝc_j\in\mathbb{R}). Let M(s)M(s) be any meromorphic function on ℂ\mathbb{C} whose divisor (set of zeros and poles) is contained in ℋm\mathcal H_m. Then the completed product ΦS(s)≔M(s)PS(s)\begin{equation} \Phi_S(s)\coloneqq M(s)P_S(s) \end{equation} cannot satisfy:

  1. an ordinary functional equation: ΦS(s)=wΦS(1−s),w∈ℂ×;\begin{equation} \Phi_S(s) = w\,\Phi_S(1-s), \qquad w\in\mathbb{C}^\times; \end{equation}

  2. nor a conjugate reflection relation: ΦS(s)=wΦS(1−s¯)¯,|w|=1.\begin{equation} \Phi_S(s) = w\,\overline{\Phi_S(1-\overline{s})}, \qquad |w|=1. \end{equation}

Proof. For any p∈Sp\in S, the set of poles {2πiklog⁡p:k∈ℤ}\{ \frac{2\pi i k}{\log p} : k\in\mathbb{Z}\} is infinite and has imaginary parts tending to ±∞\pm\infty. Since ℋm\mathcal H_m intersects the imaginary axis in at most mm points {icj:1≤j≤m}\{ic_j : 1\le j\le m\}, there exist infinitely many k∈ℤ\{0}k\in\mathbb{Z}\setminus\{0\} such that sp,k∉ℋmand−sp,k∉ℋm.\begin{equation} s_{p,k}\notin\mathcal H_m \quad\text{and}\quad -s_{p,k}\notin\mathcal H_m. \end{equation} Fix such a point s0=2πiklog⁡ps_0 = \frac{2\pi i k}{\log p} with k≠0k\neq0.

At s=s0s=s_0, M(s)M(s) has neither a zero nor a pole by hypothesis, so 0<|M(s0)|<∞0 < |M(s_0)| < \infty. Because PS(s)P_S(s) has a pole at s0s_0, ΦS(s)=M(s)PS(s)\Phi_S(s)=M(s)P_S(s) has a pole of order at least 11 at s=s0s=s_0.

Now evaluate the reflected expression ΦS(1−s)=M(1−s)PS(1−s)\Phi_S(1-s)=M(1-s)P_S(1-s) at s=s0s=s_0. The reflected argument is 1−s0=1−2πiklog⁡p,ℜ(1−s0)=1.\begin{equation} 1-s_0 = 1-\frac{2\pi i k}{\log p}, \qquad \Re(1-s_0)=1. \end{equation} The poles of PS(1−s)P_S(1-s) lie exclusively on ℜ(1−s)=0⇔ℜ(s)=1\Re(1-s)=0 \iff \Re(s)=1. Since ℜ(s0)=0\Re(s_0)=0, PS(1−s0)P_S(1-s_0) is holomorphic and non-vanishing. Furthermore, ℑ(1−s0)=−ℑ(s0)=−2πklog⁡p\Im(1-s_0) = -\Im(s_0) = -\frac{2\pi k}{\log p}. By our choice of kk, 1−s0∉ℋm1-s_0\notin\mathcal H_m, implying that M(1−s)M(1-s) has neither a zero nor a pole at s=s0s=s_0. Consequently, ΦS(1−s)\Phi_S(1-s) is holomorphic and non-vanishing at s=s0s=s_0, contradicting ΦS(s)=wΦS(1−s)\Phi_S(s)=w\,\Phi_S(1-s).

Similarly, consider the conjugate-reflected argument 1−s0¯=1+2πiklog⁡p1-\overline{s_0} = 1+\frac{2\pi i k}{\log p}, which also has ℜ(1−s0¯)=1\Re(1-\overline{s_0})=1. Thus PS(1−s0¯)P_S(1-\overline{s_0}) is holomorphic and non-vanishing. Its imaginary part is ℑ(1−s0¯)=2πklog⁡p=ℑ(s0)\Im(1-\overline{s_0}) = \frac{2\pi k}{\log p} = \Im(s_0). Since s0∉ℋms_0\notin\mathcal H_m, 1−s0¯∉ℋm1-\overline{s_0}\notin\mathcal H_m, so M(1−s0¯)M(1-\overline{s_0}) is holomorphic and non-vanishing. Thus ΦS(1−s0¯)¯\overline{\Phi_S(1-\overline{s_0})} is holomorphic and non-vanishing, contradicting ΦS(s)=wΦS(1−s¯)¯\Phi_S(s) = w\,\overline{\Phi_S(1-\overline{s})}. ◻

Remark 3 (Scope of the Divisor Obstruction). Every standard Archimedean Euler factor Γℝ(αs+β)=π−(αs+β)/2Γ(αs+β2),α>0,β∈ℝ,\begin{equation} \Gamma_{\mathbb{R}}(\alpha s+\beta) = \pi^{-(\alpha s+\beta)/2} \Gamma\left(\frac{\alpha s+\beta}{2}\right), \qquad \alpha>0,\quad \beta\in\mathbb{R}, \end{equation} has poles located exclusively on the real axis ℑ(s)=0\Im(s)=0 and has no zeros in ℂ\mathbb{C}. Finite products of such Gamma factors satisfy the horizontal-divisor hypothesis with m=1m=1 and c1=0c_1=0. The proposition proves that no such Archimedean factor can bridge the geometric mismatch between the vertical pole lattices on ℜ(s)=0\Re(s)=0 and ℜ(s)=1\Re(s)=1.

Primitive Arithmetic Lengths and Their Iterates

The Euler product imposes a distinction essential for any periodic-orbit interpretation. The primitive arithmetic datum is a prime pp, while the integer k≥1k\geq1 indexes repeated traversal of that primitive datum.

For ℜ(s)>1\Re(s)>1, log⁡ζ(s)=∑p∈ℙ∑k≥1p−ksk.\begin{equation} \log\zeta(s) = \sum_{p\in\mathbb P} \sum_{k\geq1} \frac{p^{-ks}}{k}. \end{equation} Differentiation gives −ζ′(s)ζ(s)=∑p∈ℙ∑k≥1(log⁡p)p−ks.\begin{equation} -\frac{\zeta'(s)}{\zeta(s)} = \sum_{p\in\mathbb P} \sum_{k\geq1} (\log p)p^{-ks}. \end{equation} This formula has the formal structure of a primitive-orbit expansion with primitive arithmetic length Lp=log⁡pL_p=\log p and iterated lengths Lp,k=kLp=klog⁡pL_{p,k}=kL_p=k\log p.

Definition 4 (Arithmetic Primitive Length). The quantity Lp=log⁡pL_p=\log p is called the primitive arithmetic length attached to the prime pp. The quantity Lp,k=klog⁡pL_{p,k}=k\log p is its kk-fold iterate.

Proposition 5 (Euler-Product Iterate Structure). The prime-power sum in the logarithmic derivative of ζ\zeta is exactly the iterate expansion associated with the primitive lengths log⁡p\log p: −ζ′(s)ζ(s)=∑p∑k≥1Lpe−sLp,k.\begin{equation} -\frac{\zeta'(s)}{\zeta(s)} = \sum_p\sum_{k\geq1} L_p\,e^{-sL_{p,k}}. \end{equation}

Proof. Since e−sLp,k=e−sklog⁡p=p−kse^{-sL_{p,k}} = e^{-sk\log p} = p^{-ks}, the assertion is precisely the differentiated Euler product. ◻

Warning 6 (Primitive Orbits Versus Prime Powers). The preceding proposition is an arithmetic identity and not a theorem that the scaling flow on the idele class group possesses geometrically closed orbits indexed by the primes. In particular, one must not describe every pkp^k as a distinct primitive orbit. The correct periodic-orbit grammar is: pprimitive,log⁡pprimitive length,klog⁡pk-fold iterate.\begin{equation} \boxed{ p\ \text{primitive}, \qquad \log p\ \text{primitive length}, \qquad k\log p\ \text{$k$-fold iterate}. } \end{equation} A genuine autonomous arithmetic-flow construction must still prove that these arithmetic lengths arise as geometric periods of its closed orbits.

Obstructions to Classical Geometric Prime-Orbit Realizations

Before considering infinite-dimensional or non-commutative frameworks, we investigate whether the primitive arithmetic lengths Lp=log⁡pL_p = \log p can be realized as the closed-orbit spectrum of an autonomous flow on a classical smooth geometric space.

Cross-Section and Suspension Flows

Proposition 7 (Homological Obstruction for Cross-Section Flows). Let MM be a smooth, connected, compact manifold, and let XX be a non-vanishing smooth vector field generating a flow ϕt\phi_t on MM. Suppose that the flow admits a smooth closed 11-form ω∈Ω1(M)\omega\in\Omega^1(M) satisfying ω(X)≡1.\begin{equation} \omega(X)\equiv 1. \end{equation} Then the set of primitive periods of closed orbits of ϕt\phi_t cannot contain the arithmetic prime logarithms {log⁡p:p∈ℙ}\{\log p : p\in\mathbb P\}.

Proof. Let γ\gamma be a closed orbit of ϕt\phi_t of period T(γ)T(\gamma), viewed as a smooth 1-cycle in MM. Integrating ω\omega along γ\gamma gives ∫γω=∫0T(γ)ω(γ̇(t))dt=∫0T(γ)ω(X(γ(t)))dt=∫0T(γ)1dt=T(γ).\begin{equation} \int_\gamma \omega = \int_0^{T(\gamma)} \omega(\dot{\gamma}(t))\,dt = \int_0^{T(\gamma)} \omega(X(\gamma(t)))\,dt = \int_0^{T(\gamma)} 1\,dt = T(\gamma). \end{equation} Since dω=0d\omega = 0, Stokes’ theorem implies that ∫γω=⟨[ω],[γ]⟩\int_\gamma \omega = \langle [\omega], [\gamma] \rangle, where [ω]∈HdR1(M,ℝ)[\omega]\in H^1_{\mathrm{dR}}(M,\mathbb{R}) and [γ]∈H1(M,ℤ)[\gamma]\in H_1(M,\mathbb{Z}).

Because MM is compact, its fundamental group π1(M)\pi_1(M) is finitely presented, and its abelianization H1(M,ℤ)H_1(M,\mathbb{Z}) is a finitely generated abelian group: H1(M,ℤ)≅ℤr⊕Tors⁡(H1(M,ℤ)),r=b1(M)<∞.\begin{equation} H_1(M,\mathbb{Z}) \cong \mathbb{Z}^r \oplus \operatorname{Tors}(H_1(M,\mathbb{Z})), \qquad r = b_1(M) < \infty. \end{equation} Torsion classes pair trivially with de Rham cohomology. Therefore, the evaluation homomorphism ⟨[ω],⋅⟩\langle [\omega], \cdot \rangle maps H1(M,ℤ)H_1(M,\mathbb{Z}) onto a finitely generated subgroup of ℝ\mathbb{R}: Im⁡(⟨[ω],⋅⟩)=ℤω1+…+ℤωr⊂ℝ.\begin{equation} \operatorname{Im}(\langle [\omega], \cdot \rangle) = \mathbb{Z}\omega_1 + \dots + \mathbb{Z}\omega_r \subset \mathbb{R}. \end{equation} In particular, the ℚ\mathbb{Q}-linear span of all closed-orbit periods satisfies dim⁡ℚspan⁡ℚ{T(γ):γ closed orbit}≤b1(M)<∞.\begin{equation} \dim_{\mathbb{Q}} \operatorname{span}_{\mathbb{Q}} \{T(\gamma) : \gamma \text{ closed orbit}\} \leq b_1(M) < \infty. \end{equation} By unique prime factorization in ℤ\mathbb{Z}, the set of prime logarithms {log⁡p:p∈ℙ}\{\log p : p\in\mathbb P\} is linearly independent over ℚ\mathbb{Q}: ∑i=1kcilog⁡pi=0(ci∈ℚ)⇒∏i:ci>0piai=∏j:cj<0pjbj⇒ci=0.\begin{equation} \sum_{i=1}^k c_i \log p_i = 0 \quad (c_i\in\mathbb{Q}) \quad\Longrightarrow\quad \prod_{i:\,c_i>0} p_i^{a_i} = \prod_{j:\,c_j<0} p_j^{b_j} \quad\Longrightarrow\quad c_i = 0. \end{equation} Hence dim⁡ℚspan⁡ℚ{log⁡p:p∈ℙ}=∞\dim_{\mathbb{Q}} \operatorname{span}_{\mathbb{Q}}\{\log p : p\in\mathbb P\} = \infty. Since ∞>b1(M)\infty > b_1(M), the primitive period spectrum of ϕt\phi_t cannot contain {log⁡p:p∈ℙ}\{\log p : p\in\mathbb P\}. ◻

Hyperbolic Geodesic Flows

For geodesic flows on the unit tangent bundle of a hyperbolic surface, periods are not homological and the preceding proposition does not apply. Nevertheless, an algebraic number-theoretic obstruction rules out finite-area hyperbolic surfaces.

Proposition 8 (Algebraic Obstruction for Finite-Area Hyperbolic Surfaces). Let Σ=Γ∖ℍ\Sigma = \Gamma \backslash \mathbb{H} be a hyperbolic surface of finite hyperbolic area, where Γ⊂PSL(2,ℝ)\Gamma\subset PSL(2,\mathbb{R}) is a Fuchsian group of the first kind. The set of lengths of closed geodesics on Σ\Sigma cannot contain the prime logarithms {log⁡p:p∈ℙ}\{\log p : p\in\mathbb P\}.

Proof. Because Σ\Sigma has finite area, Γ≅π1(Σ)\Gamma \cong \pi_1(\Sigma) is finitely generated. By Malcev’s theorem for linear groups, Γ\Gamma is conjugate to a subgroup of PSL(2,K)PSL(2,K) for some finitely generated field extension K/ℚK/\mathbb{Q}. Thus, for every hyperbolic element γ∈Γ\gamma\in\Gamma, its trace Tr⁡(γ)\operatorname{Tr}(\gamma) lies in KK.

The length ℓ(γ)\ell(\gamma) of the closed geodesic in the free homotopy class of γ\gamma satisfies 2cosh⁡(ℓ(γ)2)=|Tr⁡(γ)|.\begin{equation} 2 \cosh\left(\frac{\ell(\gamma)}{2}\right) = |\operatorname{Tr}(\gamma)|. \end{equation} Suppose there exists a closed geodesic γp\gamma_p with ℓ(γp)=log⁡p\ell(\gamma_p) = \log p for every p∈ℙp\in\mathbb P. Then |Tr⁡(γp)|=e12log⁡p+e−12log⁡p=p+1p=p+1p.\begin{equation} |\operatorname{Tr}(\gamma_p)| = e^{\frac12 \log p} + e^{-\frac12 \log p} = \sqrt{p} + \frac{1}{\sqrt{p}} = \frac{p+1}{\sqrt{p}}. \end{equation} Solving for p\sqrt{p} yields p=p+1|Tr⁡(γp)|∈K\sqrt{p} = \frac{p+1}{|\operatorname{Tr}(\gamma_p)|} \in K.

Since each p\sqrt{p} is algebraic over ℚ\mathbb{Q}, all such square roots must lie in the algebraic closure of ℚ\mathbb{Q} inside KK, denoted Kalg=K∩ℚ¯K_{\mathrm{alg}} = K \cap \overline{\mathbb{Q}}. Because KK is finitely generated as a field extension of ℚ\mathbb{Q}, its algebraic subfield KalgK_{\mathrm{alg}} is a finite algebraic number field: [Kalg:ℚ]=d<∞.\begin{equation} [K_{\mathrm{alg}} : \mathbb{Q}] = d < \infty. \end{equation} However, the multiquadratic extension generated by the square roots of NN distinct primes has degree [ℚ(p1,…,pN):ℚ]=2N.\begin{equation} [\mathbb{Q}(\sqrt{p_1}, \dots, \sqrt{p_N}) : \mathbb{Q}] = 2^N. \end{equation} If p∈Kalg\sqrt{p}\in K_{\mathrm{alg}} for all primes pp, we would have 2N≤d2^N \leq d for all N∈ℕN\in\mathbb N, which is impossible. ◻

Remark 9 (Scope and Boundaries of the Geometric Obstructions). The two propositions above rule out two distinguished classical settings: (i) suspension flows with constant return time, and (ii) geodesic flows on finite-area hyperbolic Riemann surfaces. They do not exclude general Anosov flows, general Axiom-A flows with non-constant roof functions, or flows on infinite-dimensional or non-commutative spaces. The question of whether a smooth autonomous flow on some smooth manifold can possess the prime spectrum remains an open problem.

The Kinematic Foundation: Adelic Harmonic Analysis

Tate’s Thesis

The adelic ring is 𝔸ℚ=ℝ×∏p′ℚp\mathbb{A}_\mathbb{Q}= \mathbb{R}\times\prod_p'\mathbb{Q}_p. For the standard factorized Schwartz–Bruhat function f=f∞⊗⨂pfpf=f_\infty\otimes\bigotimes_p f_p with f∞(x)=e−πx2f_\infty(x)=e^{-\pi x^2} and fp=𝟏ℤpf_p=\mathbf 1_{\mathbb{Z}_p}, the local Tate zeta integrals are Z∞(f∞,s)=∫ℝ×e−πx2|x|sd×x=π−s/2Γ(s/2),Zp(fp,s)=∫ℚp×𝟏ℤp(x)|x|psd×x=11−p−s.\begin{align} Z_\infty(f_\infty,s) &= \int_{\mathbb{R}^\times} e^{-\pi x^2}|x|^s\,d^\times x = \pi^{-s/2}\Gamma(s/2), \\ Z_p(f_p,s) &= \int_{\mathbb{Q}_p^\times} \mathbf 1_{\mathbb{Z}_p}(x)|x|_p^s\,d^\times x = \frac{1}{1-p^{-s}}. \end{align} Consequently, Z(f,s)=π−s/2Γ(s/2)ζ(s)=Λ(s)Z(f,s) = \pi^{-s/2}\Gamma(s/2)\zeta(s) = \Lambda(s).

Poisson summation and Fourier self-duality yield the functional equation Λ(s)=Λ(1−s)\Lambda(s)=\Lambda(1-s). The entire function ξ(s)\xi(s) is obtained by removing the simple poles at s=0s=0 and s=1s=1 via 12s(s−1)\frac12s(s-1). This is an established harmonic-analytic realization of the functional equation; it is not, by itself, a construction of a Hilbert–Pólya Hamiltonian.

The Weil Representation and Theta Correspondence

A second established structure is supplied by the Weil representation. Let VV be a non-degenerate rational quadratic space. The associated metaplectic representation provides theta kernels through the reductive dual-pair framework (SL2,O(V))(SL_2,O(V)).

Selected theta-series realizations within the Weil-theoretic framework. The table records parallel harmonic structures rather than an approximation sequence converging to ζ(s)\zeta(s).
Dimension Lattice / Space Theta Series Weight Dirichlet Structure
d=1d=1 ℤ\mathbb{Z} θ(τ)=∑n∈ℤqn2\theta(\tau)=\sum_{n\in\mathbb{Z}}q^{n^2} 1/21/2 Theta-type LL-structure
d=4d=4 D4D_4 ΘD4(τ)\Theta_{D_4}(\tau) 22 Divisor-sum / Eisenstein
d=8d=8 E8E_8 ΘE8=E4\Theta_{E_8}=E_4 44 240ζ(s)ζ(s−3)240\,\zeta(s)\zeta(s-3)
d=24d=24 Λ24\Lambda_{24} E12−65520691ΔE_{12}-\frac{65520}{691}\Delta 1212 Eisenstein plus cusp form

Remark 10 (Parallel Realizations, Not Convergence). The sequence D4→E8→Λ24D_4\longrightarrow E_8\longrightarrow\Lambda_{24} is not an approximation sequence converging to ζ(s)\zeta(s). The lattice theta series and the adelic zeta integral are different outputs of closely related self-dual harmonic structures.

The Motivic Layer: Periods and Cyclotomic Expansions

Odd zeta values and special formulas such as BBP belong to a motivic and period-theoretic layer. In the theory of mixed Tate motives over ℤ\mathbb{Z}, the motivic Galois structure organizes odd zeta values through the weight grading. The period map relates motivic objects to numerical periods ζ(3),ζ(5),…\zeta(3),\zeta(5),\ldots

BBP as a Coordinate Expansion of a Period

The Bailey–Borwein–Plouffe formula π=∑n=0∞116n(48n+1−28n+4−18n+5−18n+6)\begin{equation} \pi = \sum_{n=0}^\infty \frac{1}{16^n} \left( \frac{4}{8n+1} - \frac{2}{8n+4} - \frac{1}{8n+5} - \frac{1}{8n+6} \right) \end{equation} is an efficient base-1616 expansion of a period expressible through polylogarithmic integrals at cyclotomic points. It serves as an arithmetic coordinate projection of a period, but there is no theorem identifying BBP with a spectral determinant or with the D4/E8D_4/E_8 theta correspondence.

The Explicit Formula as the Dynamical Target

Normalization of the Dynamical Generator

To compare Deninger’s skew-adjoint formulation with Hilbert–Pólya’s self-adjoint formulation, introduce a formal generator Θ\Theta satisfying Θψρ=(ρ−12)ψρ\Theta\psi_\rho = \left(\rho-\frac12\right)\psi_\rho. If ρ=12+iγ\rho=\frac12+i\gamma, then Θψρ=iγψρ\Theta\psi_\rho=i\gamma\psi_\rho. Thus the desired operator relation is Θ*=−Θ\Theta^*=-\Theta. Equivalently, H≔−iΘH\coloneqq-i\Theta satisfies H*=HH^*=H.

The Weil Explicit Formula

Let hh be a regular test function. The Weil explicit formula has the schematic form ∑ρh(ρ−12i)=Archimedean contribution−∑p∈ℙ∑k≥1log⁡ppk/2ĥ(klog⁡p2π)+pole/trivial-zero terms.\begin{equation} \begin{aligned} \sum_\rho h\!\left(\frac{\rho-\frac12}{i}\right) ={}& \text{Archimedean contribution} \\ &- \sum_{p\in\mathbb P} \sum_{k\geq1} \frac{\log p}{p^{k/2}} \widehat h \left( \frac{k\log p}{2\pi} \right) + \text{pole/trivial-zero terms}. \end{aligned} \end{equation} This formula constitutes the design specification for any proposed autonomous arithmetic dynamics: primitive arithmetic lengths Lp=log⁡pL_p=\log p, iterates klog⁡pk\log p, and a relative minus sign on the arithmetic side.

Automorphic Scattering and the Unphysical-Sheet Complement

The physical-sheet obstruction of Proposition 2.1 is unconditional. Scattering theory does not invalidate it; rather, scattering theory provides a distinct analytic mechanism in which a self-adjoint operator has a real physical spectrum while its meromorphically continued scattering data possess complex poles and zeros.

For the modular surface X=SL(2,ℤ)∖ℍX=SL(2,\mathbb{Z})\backslash\mathbb{H}, the Eisenstein scattering coefficient is ϕ(s)=πc(s)c(1−s)=πΓ(s−12)ζ(2s−1)Γ(s)ζ(2s).\begin{equation} \phi(s) = \sqrt{\pi}\, \frac{c(s)}{c(1-s)} = \sqrt{\pi}\, \frac{\Gamma(s-\frac12)\zeta(2s-1)} {\Gamma(s)\zeta(2s)}. \end{equation} Its logarithmic derivative contains ϕ′(s)ϕ(s)=Γ′Γ(s−12)−Γ′Γ(s)+2ζ′ζ(2s−1)−2ζ′ζ(2s).\begin{equation} \frac{\phi'(s)}{\phi(s)} = \frac{\Gamma'}{\Gamma}\left(s-\frac12\right) - \frac{\Gamma'}{\Gamma}(s) + 2\frac{\zeta'}{\zeta}(2s-1) - 2\frac{\zeta'}{\zeta}(2s). \end{equation}

The minus sign in the second zeta logarithmic derivative is already present in a genuine relative/scattering quantity. However, the poles and zeros of the meromorphically continued scattering coefficient are not physical-sheet eigenvalues; they arise after analytic continuation across the continuous-spectrum boundary. Consequently: physical-sheet self-adjointness≠meromorphic scattering continuation.\begin{equation} \boxed{ \text{physical-sheet self-adjointness} \neq \text{meromorphic scattering continuation}. } \end{equation}

Proposition 11 (Scattering Does Not Bypass the Physical-Sheet Barrier). Let HH be a self-adjoint Laplace-type operator with continuous spectrum, and suppose its scattering matrix admits meromorphic continuation. Non-real poles of the continued scattering matrix do not contradict the physical-sheet incompatibility proposition, because they are singularities of the analytically continued scattering object rather than non-real eigenvalues of HH on the physical sheet.

Proof. Self-adjointness implies that the physical spectral parameter has real spectrum. The scattering matrix is initially defined on the physical continuous spectrum and analytically continued to another sheet. A pole of this continuation is therefore not an eigenvalue of the self-adjoint operator on its physical spectral sheet. ◻

Candidate Sign Mechanisms and the Non-Necessity of Super-Trace

The negative prime contribution in the explicit formula is incompatible with the simplest interpretation as an ordinary positive scalar periodic-orbit trace. However, the explicit formula does not prove that a super-trace is mathematically necessary.

Automorphic Scattering as a Non-Supertrace Sign Mechanism

In the Selberg trace formula for SL(2,ℤ)∖ℍSL(2,\mathbb{Z})\backslash\mathbb{H}, the continuous spectrum contribution to the relative trace is governed by the scattering phase shift: Tr⁡rel(h(Δ)−h(Δ0))=14π∫−∞∞h(r)(−ϕ′ϕ(12+ir))dr.\begin{equation} \operatorname{Tr}_{\mathrm{rel}}(h(\Delta) - h(\Delta_0)) = \frac{1}{4\pi} \int_{-\infty}^\infty h(r) \left( -\frac{\phi'}{\phi}\left(\frac12+ir\right) \right) dr. \end{equation} Evaluating −ϕ′ϕ(s)-\frac{\phi'}{\phi}(s) yields the term +2ζ′ζ(2s)+2\frac{\zeta'}{\zeta}(2s). In established automorphic trace theory (e.g., Venkov , Iwaniec ), moving the integration contour into the half-plane of absolute convergence ℜ(2s)>1\Re(2s)>1 and inserting the Dirichlet series ζ′(w)ζ(w)=−∑p∈ℙ∑k≥1(log⁡p)p−kw\begin{equation} \frac{\zeta'(w)}{\zeta(w)} = -\sum_{p\in\mathbb P}\sum_{k\ge 1}(\log p)p^{-kw} \end{equation} produces prime-power contributions carrying an overall negative sign.

This demonstrates an essential structural principle:

  1. The underlying Hilbert space L2(SL(2,ℤ)∖ℍ)L^2(SL(2,\mathbb{Z})\backslash\mathbb{H}) is purely scalar, with no ℤ2\mathbb{Z}_2-grading.

  2. The operator Δ\Delta is standard non-negative self-adjoint.

  3. The negative arithmetic sign arises from the fact that ζ(2s)\zeta(2s) resides in the denominator of the scattering coefficient ϕ(s)\phi(s), which combines with the orientation sign of the scattering phase shift.

Comparison of Candidate Mechanisms

Several mechanisms can in principle generate the required sign:

  1. Relative Trace / Scattering Phase: Tr⁡rel(etΘ−etΘ0)\operatorname{Tr}_{\mathrm{rel}}(e^{t\Theta} - e^{t\Theta_0}) where arithmetic LL-factors reside in the denominator of a scattering matrix.

  2. Super-Trace: A ℤ2\mathbb{Z}_2-graded space ℋ=ℋ0⊕ℋ1\mathcal H = \mathcal H^0 \oplus \mathcal H^1 with STr⁡(etΘ)=Tr⁡ℋ0(etΘ)−Tr⁡ℋ1(etΘ)\operatorname{STr}(e^{t\Theta}) = \operatorname{Tr}_{\mathcal H^0}(e^{t\Theta}) - \operatorname{Tr}_{\mathcal H^1}(e^{t\Theta}).

  3. Alternating Determinants: An alternating determinant det⁡reg(s−Θ∣H0)−1det⁡reg(s−Θ∣H1)\det\nolimits_{\mathrm{reg}}(s-\Theta\mid H^0)^{-1}\det\nolimits_{\mathrm{reg}}(s-\Theta\mid H^1), as in the finite-field Lefschetz formula.

Remark 12 (The Necessity Question). The three mechanisms above are logically distinct. The explicit formula demands the correct relative signs in the final identity, but it does not prove that characteristic-zero constructions must utilize a super-trace. Consequently, while super-trace is viable, it is not logically forced by the explicit formula alone.

The Analytic Domain Problem

A formal vector field or derivation is not enough for a Hilbert–Pólya construction. A genuine generator must be defined on a Hilbert space ℋ\mathcal H with a specified dense domain 𝒟(Θ)⊂ℋ\mathcal D(\Theta)\subset\mathcal H and must satisfy essential skew-adjointness: Θ¯=Θ*\overline{\Theta}=\Theta^*. One must specify: (1) the underlying space; (2) its measure structure; (3) the domain 𝒟(Θ)\mathcal D(\Theta); (4) leaf boundary behavior; (5) self-adjointness/skew-adjointness proof; and (6) trace and determinant regularization schemes.

The Non-Circularity Criterion

Criterion 13 (Non-Circular Spectral Realization). A proposed Hilbert–Pólya construction is non-circular only if the operator domain and infinitesimal action of Θ\Theta are specified independently of the non-trivial zeros of ζ(s)\zeta(s), and the identity det⁡reg(s−12−Θ)=Cξ(s)\begin{equation} \det\nolimits_{\mathrm{reg}} \left( s-\frac12-\Theta \right) = C\,\xi(s) \end{equation} is then derived from an independently established trace, index, or determinant theorem.

The required logical direction is: geometry / representation / dynamics⇒trace identity⇒ξ(s).\begin{equation} \boxed{ \text{geometry / representation / dynamics} \Longrightarrow \text{trace identity} \Longrightarrow \xi(s). } \end{equation}

Remark 14 (Prescribed Zeros vs. Autonomous Realization). It is well understood in regularized product theory that given the sequence of non-trivial zeros ρ=12+iγ\rho = \frac12 + i\gamma as an extrinsic input, the ζ\zeta-regularized product associated with the sequence can be rigorously evaluated to reproduce ξ(s)\xi(s) up to explicit elementary factors (cf. Deninger , Illies ). The unresolved problem in Condition (E) is therefore not the absence of a regularization scheme for a prescribed set of zeros, but rather the construction of an independent geometric operator Θ\Theta whose spectrum coincides with those zeros without imposing them by definition.

The Function-Field Model

The strongest rigorous test of the architectural requirements occurs over finite fields. Let X/𝔽qX/\mathbb{F}_q be a smooth, projective, geometrically connected curve of genus gg. Let Vℓ=Hét1(X𝔽q¯,ℚℓ)V_\ell = H^1_{\acute{e}t}(X_{\overline{\mathbb{F}_q}},\mathbb{Q}_\ell) and let F=Frob⁡q*F=\operatorname{Frob}_q^* denote the chosen Frobenius action. The Grothendieck–Lefschetz trace formula yields ZX(T)=det⁡(1−TF∣Vℓ)(1−T)(1−qT).\begin{equation} Z_X(T) = \frac{\det(1-TF\mid V_\ell)}{(1-T)(1-qT)}. \end{equation}

Normalized Frobenius

The Weil bounds give |αj|=q1/2|\alpha_j|=q^{1/2} for each eigenvalue of FF on VℓV_\ell. Define the normalized Frobenius U≔q−1/2F.\begin{equation} U \coloneqq q^{-1/2}F. \end{equation} Its eigenvalues satisfy |λj|=1|\lambda_j|=1. Equipped with an appropriate Hermitian inner product, UU is unitary: U*U=I,Spec⁡(U)⊂S1.\begin{equation} U^*U=I, \qquad \operatorname{Spec}(U)\subset S^1. \end{equation}

Warning 15 (Discrete Frobenius, No Artificial Continuous Logarithm). The finite-field model operates with the discrete unitary operator UU itself. Writing a continuous generator Θ=log⁡Flog⁡q−12I\Theta = \frac{\log F}{\log q} - \frac12 I requires an arbitrary choice of logarithm branch and embedding ℚℓ↪ℂ\mathbb{Q}_\ell\hookrightarrow\mathbb{C}, destroying canonicity. The statement Spec⁡(U)⊂S1\operatorname{Spec}(U)\subset S^1 is canonical; a continuous generator is an unnecessary and non-canonical choice.

Alternating Sign and the Odd Cohomological Sector

The middle cohomological degree carries the negative sign: #X(𝔽qn)=Tr⁡(Fn∣H0)−Tr⁡(Fn∣H1)+Tr⁡(Fn∣H2).\begin{equation} \#X(\mathbb{F}_{q^n}) = \operatorname{Tr}(F^n\mid H^0) - \operatorname{Tr}(F^n\mid H^1) + \operatorname{Tr}(F^n\mid H^2). \end{equation} This is the Euler–Poincaré sign (−1)i(-1)^i built into the Lefschetz trace formula. Equivalently, the zeta function is the alternating determinant: ZX(T)=∏i=02det⁡(1−TF∣Hi)(−1)i+1.\begin{equation} Z_X(T) = \prod_{i=0}^2 \det(1-TF\mid H^i)^{(-1)^{i+1}}. \end{equation}

Theorem 16 (Characteristic-pp Consistency of the Spectral Architecture). For a smooth, projective, geometrically connected curve X/𝔽qX/\mathbb{F}_q, the discrete normalized Frobenius operator U=q−1/2FU=q^{-1/2}F provides a finite-dimensional unitary spectral model, while the Grothendieck–Lefschetz trace formula supplies the alternating signs, and the cohomological determinant recovers ZX(T)Z_X(T). normalized unitary spectrum Spec⁡(U)⊂S1⇓Frobenius eigenvalues on H1⇓Lefschetz alternating trace⇓ZX(T)=∏i=02det⁡(1−TF∣Hi)(−1)i+1.\begin{equation} \boxed{ \begin{array}{c} \text{normalized unitary spectrum } \operatorname{Spec}(U)\subset S^1 \\[1mm] \Downarrow \\[-1mm] \text{Frobenius eigenvalues on }H^1 \\[2mm] \Downarrow \\[-1mm] \text{Lefschetz alternating trace} \\[2mm] \Downarrow \\[-1mm] Z_X(T) = \displaystyle\prod_{i=0}^2 \det(1-TF\mid H^i)^{(-1)^{i+1}}. \end{array}} \end{equation}

Proof. The unit-modulus spectrum of UU follows from Deligne’s theorem |αj|=q1/2|\alpha_j|=q^{1/2}. The alternating trace formula is the Grothendieck–Lefschetz formula. Exponentiating the resulting point count identity yields the cohomological determinant expression for ZX(T)Z_X(T). ◻

Remark 17 (What the Function-Field Model Does and Does Not Prove). The theorem proves the internal consistency of the spectral, alternating-sign, and determinant architecture in characteristic pp. It does not produce a characteristic-zero operator with spectrum equal to the Riemann zeros. The model serves as an algebraic benchmark for Conditions (A)–(E), not as a proof of the Riemann Hypothesis over ℚ\mathbb{Q}.

The Missing Arrow

The established structures and the open Dynamical Arrow are summarized: $$\begin{equation} \boxed{ \begin{array}{ccccc} \text{Adelic Fourier Duality} & \longrightarrow & \text{Tate }L\text{-function} & \longrightarrow & \Lambda(s) \\[2mm] \downarrow && && \downarrow \\[2mm] \text{Weil Representation} & \longrightarrow & \text{Theta Correspondence} & \longrightarrow & \Theta_L \\[3mm] && \Downarrow\ \text{\bf OPEN} \\[-1mm] && \text{Autonomous Arithmetic Flow }\Theta \\[2mm] && \Downarrow\ \text{\bf OPEN} \\[-1mm] && \text{Relative / Index / Scattering Trace} \\[2mm] && \Downarrow\ \text{\bf OPEN} \\[-1mm] && \xi(s). \end{array}} \end{equation}$$

The Five Conditions Recast Precisely

A candidate autonomous arithmetic system should satisfy: $$\begin{align} &\Theta^*=-\Theta, \tag{A}\\ &\operatorname{Spec}(\Theta) \text{ encodes } \rho-\frac12, \tag{B}\\ &\text{primitive arithmetic periods are } L_p=\log p, \text{ with iterates }L_{p,k}=k\log p, \tag{C$_{\rm arith}$}\\ &\operatorname{Tr}_{\mathrm{rel}}, \text{ an index, super-trace, or scattering determinant recovers the explicit formula}, \tag{D}\\ &\det\nolimits_{\mathrm{reg}} \left(s-\frac12-\Theta\right) \propto \xi(s). \tag{E} \end{align}$$

Historical Paradigms

  1. Berry–Keating and Local Hamiltonians: The semiclassical Hamiltonian H=xpH=xp reproduces the smooth average counting law N¯(E)∼E2π(logE2π−1)\overline{N}(E) \sim \frac{E}{2\pi}\left(\log\frac{E}{2\pi}-1\right), but lacks an autonomous mechanism for arithmetic prime fluctuations. While standard one-dimensional Schrödinger operators impose rigid asymptotic spectral constraints, no accepted independent differential realization is known that non-circularly produces the zeros together with their arithmetic structure.

  2. Connes and Meyer (Adèle Class Space): Connes’ noncommutative geometry on 𝔸ℚ/ℚ×\mathbb{A}_\mathbb{Q}/\mathbb{Q}^\times recovers the Weil explicit formula as an absorption-spectrum trace formula. Meyer realized this within a virtual representation of the idèle class group whose distribution character recovers the explicit formula. This provides a rigorous representation-theoretic framework, but it remains a virtual representation rather than a conventional discrete self-adjoint Hilbert–Pólya Hamiltonian.

  3. Deninger’s Cohomological Program: Deninger proposes a foliated dynamical space where primes correspond to closed orbits and an arithmetic Lefschetz trace formula yields the explicit formula. In the formulation of this program, the prospective cohomological spaces carrying the infinitesimal generator Θ\Theta are intrinsically infinite-dimensional real pre-Hilbert spaces. The construction of the underlying space, its foliated measure, and its closed operator domain remain open.

  4. de Branges’ Positivity Route and Conrey–Li: De Branges proposed an entire-function Hilbert-space framework ℋ(E)\mathcal{H}(E) wherein specific positivity conditions would force the zeros of LL-functions to lie on the critical line. Conrey and Li proved that the specific positivity condition required by this version of the de Branges program fails for the Riemann zeta function and Dirichlet LL-functions.

  5. Automorphic Scattering: Scalar automorphic scattering (SL(2,ℤ)∖ℍSL(2,\mathbb{Z})\backslash\mathbb{H}) demonstrates that relative traces and scattering phase shifts can rigorously generate negative prime-power contributions without requiring a ℤ2\mathbb{Z}_2-graded super-trace. Non-real scattering poles occur as resonances on analytically continued unphysical sheets, preserving self-adjointness of the physical Laplacian while leaving the Hilbert–Pólya operator problem open.

Scope and Limitations

  1. No new proof of the Riemann Hypothesis is claimed.

  2. No explicit Hilbert–Pólya operator is constructed.

  3. Tate’s thesis provides the functional equation, not a discrete spectrum.

  4. Weil representation does not identify ξ(s)\xi(s) with lattice theta series.

  5. Euler products supply arithmetic lengths klog⁡pk\log p, but an autonomous flow realizing them as geometric periods remains unconstructed.

  6. Geometric obstructions rule out cross-section flows with ω(X)≡1\omega(X)\equiv 1 and finite-area hyperbolic surfaces, but do not rule out all smooth flows.

  7. 1D local Schrödinger models face strong structural spectral constraints; no accepted independent realization reproducing the zeros is known.

  8. Automorphic scattering zeros are resonances, not physical eigenvalues.

  9. A super-trace is not logically necessary; scalar scattering yields negative signs.

  10. Characteristic-pp Frobenius verifies consistency, not characteristic-zero dynamics.

  11. Any proposed generator requires a specified dense Hilbert domain.

Falsification Boundaries and Residual Open Problems

The obstruction results permit a sharper classification of the architectural questions surrounding a Hilbert–Pólya realization. It is vital to distinguish:

  1. a proposed construction is impossible under specified hypotheses;

  2. a particular mechanism cannot establish the desired spectral interpretation;

  3. the underlying Hilbert–Pólya problem itself is impossible.

Only the first two conclusions are established by the obstruction results.

Classical Flows: Precise Obstruction Classes

The prime Euler product determines the primitive arithmetic lengths Lp=log⁡pL_p = \log p, with iterates Lp,k=klog⁡pL_{p,k}=k\log p. The homological obstruction of Section 4 proves that a compact cross-section flow with ω(X)≡1\omega(X)\equiv 1, dω=0d\omega=0, cannot have all log⁡p\log p as primitive periods because all periods lie in a subgroup of finite ℚ\mathbb{Q}-rank, whereas {log⁡p}p∈ℙ\{\log p\}_{p\in\mathbb P} has infinite ℚ\mathbb{Q}-rank.

Likewise, the algebraic obstruction rules out finite-area hyperbolic surfaces. These are class-specific no-go theorems, not universal theorems about all smooth flows.

Remark 18 (Boundary of the Flow Obstruction). The arguments do not exclude arbitrary Anosov or Axiom-A flows. For symbolic codings with a finite-valued or locally constant roof function, one obtains a finite-rank period group and hence an obstruction. For a general non-constant roof function, however, periods need not lie in a finitely generated subgroup of ℝ\mathbb{R}.

Similarly, Weyl asymptotics for an elliptic differential operator cannot be applied directly to an arbitrary flow’s periodic-orbit spectrum. Any argument using a Weyl-law obstruction must first establish a specific elliptic spectral realization and a precise correspondence between its eigenvalue counting function and the proposed arithmetic data.

Locally Symmetric Spaces

Proposition 19 (Algebraic Length Obstruction Criterion). Let Γ\Gamma be a finitely generated subgroup of a linear Lie group and suppose that the length of every relevant hyperbolic conjugacy class is determined by an algebraic expression in its eigenvalues. Assume further that all eigenvalues lie in a field KK finitely generated over ℚ\mathbb{Q}, and that p∈Kalg\begin{equation} \sqrt p\in K_{\mathrm{alg}} \end{equation} whenever log⁡p\log p occurs as one of the relevant primitive lengths. Then the set {log⁡p:p∈ℙ}\{\log p:p\in\mathbb P\} cannot be contained in the primitive length spectrum.

Proof. Since KK is finitely generated over ℚ\mathbb{Q}, Kalg/ℚK_{\mathrm{alg}}/\mathbb{Q} is a finite extension with [Kalg:ℚ]=d<∞[K_{\mathrm{alg}}:\mathbb{Q}]=d<\infty. If p∈Kalg\sqrt{p}\in K_{\mathrm{alg}} for all primes, then for every NN, [ℚ(p1,…,pN):ℚ]=2N≤d[\mathbb{Q}(\sqrt{p_1},\dots,\sqrt{p_N}):\mathbb{Q}] = 2^N \le d, which is impossible for 2N>d2^N > d. ◻

Remark 20. This is an algebraic criterion, not a blanket theorem about every locally symmetric space. It must be checked representation by representation.

Scattering Does Not Evade the Physical-Sheet Obstruction

Automorphic scattering on L2(SL(2,ℤ)∖ℍ)L^2(SL(2,\mathbb{Z})\backslash\mathbb{H}) operates on the physical Hilbert space with real spectrum Spec⁡(Δ)⊂ℝ≥0\operatorname{Spec}(\Delta)\subset\mathbb{R}_{\ge 0}. The resonances (zeros and poles of ϕ(s)\phi(s)) occur on the unphysical sheet via meromorphic continuation. Scattering demonstrates that a negative arithmetic contribution in a relative trace does not require a ℤ2\mathbb{Z}_2-graded super-trace, but it does not construct a physical-sheet Hilbert–Pólya operator.

The Determinant Problem

An identity det⁡reg(s−12−Θ)=Cξ(s)\det\nolimits_{\mathrm{reg}}\left(s-\frac12-\Theta\right) = C\xi(s) is meaningful only after the regularization genus, operator domain, and multiplicities are specified. While regularized product formulas for the non-trivial zero set are established (Illies ), finding an autonomous operator Θ\Theta remains the central open hurdle.

Conditions (A)–(E) and the Status of RH

If an operator Θ\Theta non-circularly satisfies (A) Θ*=−Θ\Theta^*=-\Theta and (B) Spec⁡(Θ)\operatorname{Spec}(\Theta) encodes ρ−12\rho-\frac12, it would imply ℜ(ρ)=12\Re(\rho)=\frac12 (RH). Conversely, the discovery of a single off-line zero ρ0\rho_0 with ℜ(ρ0)≠1/2\Re(\rho_0)\neq 1/2 would definitively rule out the simultaneous realization of (A) and (B) in this form.

Authoritative Epistemic Classification

Table 1 summarizes the controlling epistemic status of candidate mechanisms and open problems.

Authoritative epistemic ledger of established structures, bounded obstructions, and open problems.
Candidate Mechanism / Question Defensible Epistemic Status
Cross-section flows (dω=0d\omega=0, ω(X)≡1\omega(X)\equiv 1) Proved impossible (finite ℚ\mathbb{Q}-rank of H1H_1)
Finite-area hyperbolic surfaces Proved impossible ([ℚ({p}):ℚ]=∞[\mathbb{Q}(\{\sqrt{p}\}):\mathbb{Q}]=\infty)
Finite-valued / local-roof symbolic models Proved impossible under stated roof hypothesis
1D local Schrödinger operators No accepted independent realization; blanket no-go not established
General Anosov / Axiom-A flows Open
General compact smooth autonomous flows Open
Deninger cohomological / Lefschetz model Intrinsically infinite-dimensional in proposed formulation
Connes–Meyer adèlic representation Virtual representation; not a conventional HP Hamiltonian
de Branges positivity framework Specific positivity criterion ruled out (Conrey–Li)
Physical-sheet non-real eigenvalues Impossible for self-adjoint pairs (Proposition 2.1)
Automorphic scattering Valid scalar resonance/sign mechanism; not HP realization
Super-trace requirement Not logically necessary in principle
Regularized determinant of zero sequence Regularization theory established; independent operator open
Non-circular Θ\Theta satisfying (A)–(E) over ℚ\mathbb{Q} Open
Characteristic-pp Frobenius analogue Established algebraic benchmark
Riemann Hypothesis Open

Remark 21 (Central Architectural Conclusion). The obstruction theory does not prove that every finite-dimensional smooth realization is impossible. What it establishes is more precise: several natural classical mechanisms are rigorously excluded, while the remaining problem is concentrated in constructions whose geometry is not captured by those mechanisms.

Accordingly, a viable characteristic-zero Hilbert–Pólya construction must still supply, without encoding the zeros by definition, a precise operator domain, a skew-adjointness theorem, an arithmetic orbit or scattering interpretation, a valid sign mechanism, and a fully specified regularized determinant. None of these requirements is discharged merely by the existence of a formal analogy with Frobenius, scattering theory, or the adèle class space.

Conclusion

The architectural boundary of the Hilbert–Pólya program can now be stated with precision:

First, a substantial portion of the harmonic-analytic kinematics is established. Adelic Fourier duality gives the Tate zeta integral and the functional equation, while the Weil representation and theta correspondence organize a broad family of lattice theta series including those of D4D_4, E8E_8, and the Leech lattice. These structures represent parallel manifestations of self-dual harmonic analysis rather than an approximation sequence terminating in ζ(s)\zeta(s).

Second, simple local and finite reductions encounter genuine analytic obstructions. Individual Euler factors cannot be physical perturbation determinants of self-adjoint pairs on the physical sheet. Furthermore, under the horizontal-divisor hypothesis (which includes all standard finite-degree Archimedean Gamma factors), finite Euler products cannot reproduce the reflection symmetry of the completed zeta function due to the polar mismatch between ℜ(s)=0\Re(s)=0 and ℜ(s)=1\Re(s)=1.

Third, classical geometric realizations of the prime lengths log⁡p\log p are obstructed in specific natural settings. A compact flow admitting a closed transverse 1-form ω(X)≡1\omega(X)\equiv 1 has a period group of finite ℚ\mathbb{Q}-rank, which cannot contain the ℚ\mathbb{Q}-linearly independent prime logarithms. Similarly, closed geodesics on finite-area hyperbolic surfaces cannot have lengths log⁡p\log p because their traces would force a finitely generated field to contain multiquadratic extensions of arbitrarily large degree. These results rigorously motivate the necessity of looking beyond simple suspension geometries or modular geodesics, while leaving open the question for general smooth flows.

Fourth, automorphic scattering clarifies the sign problem. The relative trace of the self-adjoint Laplacian on L2(SL(2,ℤ)∖ℍ)L^2(SL(2,\mathbb{Z})\backslash\mathbb{H}) generates arithmetic logarithmic derivatives with a negative sign because ζ(2s)\zeta(2s) sits in the denominator of the scattering matrix. This disproves the claim that a super-trace is logically necessary, while respecting the physical-sheet barrier: scattering zeros appear as resonances on the unphysical sheet, rather than as physical-sheet eigenvalues.

Fifth, the finite-field model confirms that the overall spectral/determinantal/sign architecture is internally consistent. For a curve over 𝔽q\mathbb{F}_q, normalized Frobenius has unit-modulus spectrum, the Lefschetz trace formula supplies the alternating sign, and the cohomological determinant reproduces the zeta function.

The unresolved problem in characteristic zero is therefore isolated to Conditions (A)–(E): constructing an autonomous dynamics whose operator, domain, trace formula, and regularized determinant are defined independently of the non-trivial zeros and simultaneously recover the spectral and prime-power sides of the explicit formula. Until such an object is constructed, the Hilbert–Pólya program remains open in precisely this dynamical sense.

99

M. V. Berry and J. P. Keating, The Riemann zeros and eigenvalue asymptotics, SIAM Review 41 (1999), no. 2, 236–266.

F. Brown, Mixed Tate motives over ℤ\mathbb Z, Annals of Mathematics 175 (2012), no. 2, 949–976.

A. Connes, Trace formula in noncommutative geometry and the zeros of the Riemann zeta function, Selecta Mathematica (N.S.) 5 (1999), no. 1, 29–106.

J. B. Conrey and X.-J. Li, A note on some conditions of de Branges related to the Riemann hypothesis, Journal of Number Theory 84 (2000), no. 2, 263–272.

C. Deninger, Local LL-factors of motives and regularized determinants, Inventiones Mathematicae 107 (1992), 135–150.

C. Deninger, Motivic LL-functions and regularized determinants, in Motives, Proc. Sympos. Pure Math. 55, Amer. Math. Soc., 1994, 707–743.

P. Deligne, La conjecture de Weil. I, Publications Mathématiques de l’IHÉS 43 (1974), 273–307.

A. Grothendieck, Formule de Lefschetz et rationalité des fonctions LL, Séminaire Bourbaki, Exp. No. 279, 1965.

R. Howe, θ\theta-series and invariant theory, in Automorphic Forms, Representations and LL-functions, Proc. Sympos. Pure Math. 33, Amer. Math. Soc., 1979, 275–285.

G. Illies, Regularized products of zeros of the Riemann zeta function, Communications in Mathematical Physics 197 (1998), no. 3, 697–723.

H. Iwaniec, Spectral Methods of Automorphic Forms, 2nd ed., Graduate Studies in Mathematics 53, American Mathematical Society, 2002.

R. Meyer, On a representation of the idele class group related to the primes and zeros of LL-functions, Duke Mathematical Journal 127 (2005), no. 3, 519–595.

A. Selberg, Collected Papers, Springer-Verlag, Berlin, 1989.

J. T. Tate, Fourier analysis in number fields and Hecke’s zeta-functions, Ph.D. thesis, Princeton University, 1950; published in Algebraic Number Theory, J. W. S. Cassels and A. Fröhlich, eds., Academic Press, 1967, 305–347.

The Stacks Project Authors, The Stacks Project, Sections on L-functions, Lefschetz trace formulas, and cohomological interpretations over finite fields.

A. Weil, Sur certains groupes d’opérateurs unitaires, Acta Mathematica 111 (1964), 143–211.