Paper I of a two-part study extending the no-boundary programme of A smooth beginning for spacetime to the black-hole topology. The preceding series argued that the no-boundary saddle of the S3 universe carries the complete thermodynamics of the de Sitter horizon in its own data, that the Feldbrugge-Lehners-Turok fluctuation catastrophe is the fingerprint of data placed on a branch point, and that the horizon's dissipation erases the resulting bias wherever its ledger can afford it. This paper begins the same programme on the black-hole topology: the no-boundary state of a universe with spatial sections S1 x S2. The anatomy transposes. The S1 x S2 closure saddles are smooth caps: each member of the conjugate pair is exactly a complex Schwarzschild-de Sitter geometry, of constant complex mass M(u,v) and complex Euclidean period, closing regularly on one of its two horizons—the closure datum p is the tip's opening angle, Θ/2π = -p identically, and the standard prescription p = -1 is the smooth cap. The cap's weight is the horizon boundary term: the member action is I = p S_tip - 16π²c1/s exactly, a horizon term plus a wall term, which at the Nariai slice reduces to I = -S_tip with S_tip = S_tot_Nariai/2, so the pair carries |Ψ|² = e^(S_tot_Nariai): the black-hole universe squares to the Nariai ledger, by the same smooth-horizon mechanism as the sphere squares to the de Sitter one. These statements are exact in a units dictionary derived and gated here, and they reproduce, as their classical shadow, the large-circle statements of Turiaci and Wu. The geometry contains one horizon; the second is virtual, behind the data surface, which plays the role of York's cavity wall. That is why the two-horizon first law cannot be posed on this or any related family: at fixed Λ both horizon areas are functions of the one mass—Birkhoff's theorem—and the three added-data candidates (the b=0 closure, the opening angle as data, horizon-anchored Dirichlet families) are each closed exactly: the empty static patch, the one-horizon Legendre pair (S_tip, Θ/2π) and the closed two-horizon geometry with I = -Σ(Θ_i/2π)S_i = P Vol and identically zero energy. A short final section states what the smooth cap does to its perturbations—circle-mode indices ±n/2, real and selective as on the sphere; mode weights An = ±n/2, member-antisymmetric and linear in n—and the coalescence exit's exact closure, which Paper II takes up. Version 1 of this paper reported a complex conical deficit at the tip, a drifting mass, and imaginary indices; all three were one sign error in the reconstruction of the saddle's geometry, invisible to every value-level gate. The error, its diagnosis and what it withdrew are retained in the record, with the nine earlier corrections.
No takes yet. Share an insight, caveat, or question.
James Laurence Williams (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: