We argue that the Riemann hypothesis is not a conjecture awaiting proof but a ground: a structural identity whose negation requires the identity it negates. The argument has three layers. The algebraic layer: the zeta function is defined simultaneously as the Dirichlet series (additive) and the Euler product (multiplicative); an off-line zero uses this identity to exist while asserting that the two descriptions are locally out of balance. The geometric layer: the Riemann sphere, with the icosahedron inscribed via Klein's 1884 integration, places the critical line at the exact logarithmic midpoint of the icosahedral vertex shells at |w| = φ and |w| = 1/φ; a seven-step derivation chain, each link proved, identifies the critical line as the unique equilibrium of the smooth witness (sphere) and the facetted witness (dodecahedron). The self-undermining layer: an off-line zero of ζ uses the identity De = Dφ to exist as a zero of the function so defined, and denies it by sitting where the two modes are not in balance — the same self-undermining structure as the denial of the Principle of Sufficient Reason. The argument is tautological, and the tautology is the content: a ground stands on itself because there is nothing beneath it. A theorem is derived from something external; a ground is the axiom that derives itself. The same self-undermining structure applies to three further Millennium Prize problems (Yang–Mills, BSD, Hodge), each admitting an additive and a multiplicative description whose identity is the content of the conjecture.
Gereon Kraemer (Mon,) studied this question.