The open question
Research note · the testable core, in half a page · Sep 2026 (rev. 3)
The claim
T≡M holds that time is identical to motion, with cosmological expansion as the root tick. Version 2.2 adds the condition that gives the identity its content: only a motion that closes on itself constitutes proper time (postulate T≡M-4). Its first consequence — no closure, no proper time — corresponds to the zero proper time along null trajectories in special relativity.
The weak, testable branch is a differential drift ε_R between clock species. A universal shift cancels in frequency ratios, and a constant per-species offset is absorbed into calibration, so only a drift that differs between clock types can be seen. Existing optical-clock comparisons bound it today at |Δε_R| ≲ 10⁻⁸.
ε_R should not be read as expansion “stretching” clocks. A bound system does not undergo a continuous stretching proportional to H: its response is all-or-nothing after transients (Price & Romano 2012), the known local effect of a cosmological constant is a static offset (Carrera & Giulini 2010), and the conclusion holds under quantum treatment as well (Lemos 2014). ε_R is the parameter of a residual coupling of unspecified origin.
The problem
Operationally, a differential drift is the same class of signal as a temporal variation of fundamental constants. A detection could not be attributed to T≡M without a rule for how ε_R depends on the type of transition — one that differs from the ordering by sensitivity coefficients to α, μ, and the rest.
The observable has the two-factor form that is standard in the phenomenology of locally coupled light scalars: a universal cosmological rate times a per-species coefficient (Trivedi 2025). In that literature the per-species coefficient has no independent experimental determination and is estimated from sensitivity to constants. That is the gap where a rule of its own would have to live.
What was tried
The closure condition suggests such a rule: systems with different types of closure might couple differently to the root tick. Five orderings were explored — closure compactness, binding strength, hierarchical level, rhythm affinity, and autonomy (the fraction of what sustains a closure that comes from its own internal relations). Autonomy went furthest: it is the only one whose direction follows from the framework rather than being chosen.
The linking hypothesis was therefore withdrawn. This is a result about the present formalism, not a proof of impossibility. The full report is in the Companion Paper v1.2.
The question
A related, more concrete one: the relation between clock drift and the evolving background is symmetric. A per-species coefficient different from the constants-sensitivity estimate would also bound the dark-energy equation of state differently, so clocks and cosmology would test the same hypothesis from two sides.
What is missing
A way to turn a relational property — closure, autonomy — into a physical object that enters the dynamics. Until that exists, T≡M remains an ontological framework with a bounded parameter and no signature of its own, and it says so. Every version, including this negative result, is archived with a DOI — see the revision history.
References
- Price, R. H. & Romano, J. D. (2012). In an expanding universe, what doesn’t expand? Am. J. Phys. 80, 376–381.
- Carrera, M. & Giulini, D. (2010). Influence of global cosmological expansion on local dynamics and kinematics. Rev. Mod. Phys. 82, 169.
- Lemos, N. A. (2014). On what does not expand in an expanding universe: a very simple model. Braz. J. Phys. 44, 91–94.
- Sherrill, N. et al. (2023). Analysis of atomic-clock data to constrain variations of fundamental constants. New J. Phys. 25, 093012.
- Trivedi, O. (2025). Future dark energy constraints from atomic clocks. arXiv:2512.16804 (preprint).
Framework sources: T≡M v2.2 (10.5281/zenodo.22901969) · Companion Paper v1.2 (10.5281/zenodo.22923299)