Work

Time from reconstruction: a timeless flat four-space, one advancing hypersurface, and the Lorentzian world it reconstructs

Year: 2026
Type: preprint
Abstract: We take space to be flat Euclidean four-space with no time in it, and matter to lie on its own proper-time level sets, which are hypersurfaces of it. We show that this one statement supplies the hypotheses of the Osterwalder–Schrader reconstruction, so that the Lorentzian signature, the light cone and the Lorentz group are outputs rather than inputs, and that the leaf's advance along its normal is forced rather than assumed: - the reflection axis is the leaf's normal, a physical direction rather than a choice; - reflection positivity follows from a single positive generator along it; - the rotation group is that of the Euclidean measure; - the advance is monotone, and it runs from small leaves to large: parallel copies of a convex leaf exist outward at every distance and focus inward at a finite one, so a stack has one end, and the soap bubble, the leaf's entropy and the Alexandrov–Fenchel equality all run the same way. Positivity holds for the scalar and Dirac fields and for Yang–Mills on the walled leaf, which is finite and gapped on every leaf. Complete monotonicity of each observable's correlator along the advance (Bernstein) gives the directional arrow. Pointwise, the unit tangent of a worldline gives a four-speed budget, $u_w = 1/\gamma$. That picture is known, and the reconstruction says how it meets special relativity: the worldline's tilt angle is the Gudermannian of the reconstructed rapidity. One rotation in the plane of the advance gives the mass shell, the lapse, every horizon (at the quarter turn), the Lorentzian minus sign ($i^2$) and de Sitter's static lapse $\cos\chi = \sqrt{1 - R^2/\ell^2}$. Mass is reach along the advance, $u_w = m/E$. The photon has no clock because Maxwell's measure is blind to scale in four dimensions, so its modes continue to $E = |\mathbf p|$. The number of boosts the reconstruction returns is a coset count: three for parallel leaves, which gives Poincaré, and none for concentric spheres, which gives the six-dimensional symmetry of closed Friedmann–Robertson–Walker. The verification bundle carries the paper's source, its verification scripts (test 612) and their run log; Appendix V of the paper indexes every displayed result against them. The author's archive: concept DOI 10.5281/zenodo.21264118.
Language: English
Open Access status: green
OpenAlex ID: W7235259604
Cites: 4
Cited by: 3
Related to: 10