Revisiting the Duff Argument in Cosmology: From Dimensional Quantities to Cross-Epoch Observables

arXiv:2609.35903 · gr-qc · Submitted 2026-09-28 · Read on arXiv

Listen

Transcript

Introduction to the show: ident: General Relativity and Quantum Cosmology. The day in general relativity and quantum cosmology.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Revisiting the Duff Argument in Cosmology".

Mira: Here's the gist of the paper. The paper reconsiders whether measurable content remains in relations between physical standards across different cosmological epochs when considering the unit dependence of dimensional quantities.

Kai: The central empirical question is whether the unit-rate realization NG = 1 is sufficient for cross-epoch cosmological timing or whether observations support a nontrivial NG(a) within the meVSL realization.

Mira: Temporal Framework and Covariance The authors introduce Generalized Cosmological Time (GCT) which retains an explicit common cosmological time parameter TG, defining the temporal-rate relation as dτ = NG dTG, where NG = (Uµ∇µTG)−1.

Kai: This formulation separates the temporal reparametrization freedom already present in Robertson–Walker spacetime from the phenomenological realization of a non-unit cosmological temporal rate.

Title and authors: Mira: Now, let's look at the summary section of "Revisiting the Duff Argument in Cosmology: From Dimensional Quantities to Cross-Epoch Observables." It boils down to how they tackle the fundamental issue of unit dependence in dimensional quantities across different cosmological epochs.

Kai: Basically, they state that a dimensional constant’s numerical value depends entirely on the units you choose, so looking at it in isolation doesn't tell us much about its actual physical evolution across time scales.

Lev: So if I was building a sensor, knowing that the numerical output changes just because I switched from meters to centimeters wouldn't help me understand the physics of what’s happening.

Mira: Right. The paper suggests we need a framework, like their GCT with NG equals (U mu nabla mu TG) minus one, to properly embed the temporal rate relation into the structure of spacetime itself.

Kai: This framework allows them to separate coordinate freedom from the physical rate of change, which is what they call NG. They are focusing on this specific factor because it dictates how proper time intervals relate to that common cosmological time parameter TG.

Lev: That separation is key for me because if we can isolate the physical rate factor, then I can figure out exactly how much overhead or correction that introduces when running a simulation or testing error correction codes.

Mira: They then develop the cosmological time dilation test, which relates the proper-time intervals delta tau o and delta tau e using the factor one plus z times NG e over NG o. This is their primary observational tool for testing if things scale as expected across epochs.

Kai: And they show that within their meVSL parametrization, this leads to a specific endpoint rate relation of RT (z) equals one plus z minus b over four. That’s the observable prediction derived from the underlying model assumptions.

Lev: The caveat here is that this entire derivation hinges on the assumption of the meVSL framework being correct, which is where I have my reservations—if we assume a different theory for how light travels, this whole scaling relationship might be off.

Mira: That's fair skepticism. But the paper’s contribution seems to be providing a rigorous way to test the unit-rate realization against this more general non-unit rate realization NG(a) by looking at those interval conversions in GCT.

Kai: So, essentially, they are providing the mathematical machinery to ask if observations support a specific scaling factor b being non-zero when we compare sources across different redshifts.

The paper's summary: Kai: Moving into their proposed improvements for the original Duff argument paper, they suggest focusing on cross-epoch observables rather than just dimensional quantities that are defined in isolation.

Mira: They argue that the improvement is to use a more robust framework, specifically incorporating Generalized Cosmological Time and the meVSL parametrization. This gives them NG(a) equals a b over four as their working model for the temporal rate.

Lev: From my perspective, incorporating this specific functional form of NG(a) is good because it gives us something concrete to compare against real-world constraints rather than just abstract theoretical possibilities.

Kai: They also introduce the source equivalence condition as a way to test this using actual astrophysical sources at different redshifts, which demands that the compared sources represent the same intrinsic timescale distribution.

Mira: This source equivalence condition is important because it provides a concrete way to compare two sets of observations—one at redshift e and one at redshift o—using either GCT intervals or proper-time intervals.

Lev: If we use proper-time intervals, the paper shows that if source equivalence is formulated for WG, equation nine gives WG o equals one plus z times NG e over NG o WG e. That’s a direct test of the scaling relationship based on what we measure in time.

Kai: So the improvement is moving from a purely dimensional analysis to a testable cosmological timing relation derived from this specific model parametrization of time evolution.

Mira: They also note that they are careful about how they handle lapse variation, making sure that by using the non-unit-lapse parametrization N(TG) equals a b over four after variation with respect to the lapse, you retain the Hamiltonian constraint explicitly.

Lev: That’s a nice piece of engineering—ensuring that you don't lose any physical degrees of freedom when you choose how you define your time evolution; keeping that constraint visible is crucial.

Kai: So, the improvement is providing a complete set of tools—GCT, the meVSL realization, and source equivalence—to probe whether NG equals one or if NG(a) is nontrivial.

The paper's improvements: Mira: To wrap up on "Revisiting the Duff Argument in Cosmology: From Dimensional Quantities to Cross-Epoch Observables," they summarize that the isolated numerical variation of a dimensional constant isn't an observable by itself.

Kai: They reiterate that the test isn't about C equals C0; it’s about whether physical standards associated with different cosmological epochs can enter unit-independent cross-epoch relations when the temporal parametrization is kept explicit.

Lev: From my side, it confirms that we need to build tests that look at the relationship between time intervals across cosmic history, not just the values of individual constants.

Mira: They conclude by stating that they provide a temporal framework for this question without assigning independent observable significance to the numerical evolution of any single dimensional constant.

Kai: The paper ends on a crucial note: whether we need a unit-rate realization NG equals one or if observations support a nontrivial NG(a) within the meVSL realization.

Lev: I just add that for practical work, it confirms that if you want to make these comparisons robust, you have to incorporate source equivalence conditions and control for those systematics.

Mira: So, the big picture is they've given us a way to probe whether we need NG equals one or not by focusing on the temporal evolution of intervals rather than just static dimensional values.

Conclusion: Kai: So, we just looked at "Revisiting the Duff Argument in Cosmology: From Dimensional Quantities to Cross-Epoch Observables." Essentially, they’re asking if you can really link physical standards between different eras when you think about how units change.

Mira: Right. The paper sets up this Generalized Cosmological Time thing, TG, where the temporal rate NG is defined by that common time parameter. It separates the freedom in spacetime from the actual physical scaling of things like c or.

Lev: From my side, it’s interesting how they manage to keep the Hamiltonian constraint explicit when they parameterize that lapse as N equals a b over four. It shows you don't lose a degree of freedom just by choosing a non-unit lapse.

Kai: The core test is the cosmological time dilation, where they show that the proper-time relation delta tau o divided by delta tau e equals one plus z times NG e over NG o. That’s their main observational tool here.

Mira: And they apply this to their meVSL model, which gives them a specific endpoint rate relation: RT (z) equals one plus z minus b over four. This 'b' parameter is what governs the correlated scaling of all those dimensional quantities like c and G.

Lev: If you want to test this on real data, you need that source equivalence condition. It means the sources you compare have to represent the same intrinsic timescale distribution at different redshifts.

Kai: And if you use proper-time intervals with that source equivalence, they show that WG o equals one plus z times NG e over NG o WG e. That’s a direct measurement of the scaling factor b.

Mira: So, what this means for us is that we don't just need to look at individual constants in isolation when we study cosmology; we have to look at how they scale together across time.

Lev: It tells us that if observations show a deviation from the unit-rate NG equals one, then the structure of our spacetime evolution itself has a measurable temporal rate component.

Kai: It confirms that the question boils down to whether we need NG equals one for cross-epoch timing or if the meVSL realization with NG(a) is what observations actually support.

Mira: That's it for this paper. It gives us the framework to ask that big question without getting bogged down in just looking at isolated values of c or.

Lev: Next up, we’re looking at some local obstructions to trapped-region escape in modified gravity. We’ll see if those constraints are as tight as they look on paper.

Institute of Basic Science, Sungkyunkwan University

gr-qc

Submitted: 2026-09-28

Updated: 2026-10-29

Comments: 17 pages, 1 table, v2: Substantially revised and expanded. Title changed; cross-epoch observability, temporal parametrization, and source-equivalence/source-evolution treatments have been reformulated and clarified. References updated

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 77/100

The gist: The paper reconsiders whether measurable content remains in relations between physical standards across different cosmological epochs when considering the unit dependence of dimensional quantities.

Key concepts

Generalized Cosmological Time (GCT): This framework defines temporal evolution using dτ = NG dTG, where TG is a common cosmological time parameter. This separates the freedom in spacetime parametrization from the physical realization of a non-unit temporal rate, allowing for a clearer test of cosmological timing.
Cosmological Time Dilation Test (CTD): This test shows that for null-linked intervals, the proper-time relation is ∆τo = (1 + z)∆τe. When using GCT, this translates to a ratio involving NG: ∆TG,o/∆TG,e = (1 + z)NG,e/NG,o.
Minimally Extended Varying Speed of Light (meVSL): This is a phenomenological model where the temporal rate is given by NG(a) = a b/4. The parameter 'b' dictates how dimensional quantities like the speed of light (c), gravitational constant (G), and Planck's constant ($\hbar$) scale across different epochs.
Source Equivalence Condition: To compare observations from different redshifts, sources must be equivalent in their intrinsic timescale distribution. This condition allows researchers to use GCT or proper-time intervals to test the CTD relation robustly.

Terminology

Generalized Cosmological Time — 8× in this paper
Homogeneity — 2× in this paper · explained in 2 episodes
Core — 1× in this paper · explained in 2 episodes
Initial Data — 1× in this paper · explained in 2 episodes
National Research Foundation of Korea — 1× in this paper
Parameterization — 1× in this paper · explained in 2 episodes

Summary

The paper reconsiders whether measurable content remains in relations between physical standards across different cosmological epochs when considering the unit dependence of dimensional quantities.

The central empirical question is whether the unit-rate realization NG = 1 is sufficient for cross-epoch cosmological timing or whether observations support a nontrivial NG(a) within the meVSL realization.

Temporal Framework and Covariance

The authors introduce Generalized Cosmological Time (GCT) which retains an explicit common cosmological time parameter TG, defining the temporal-rate relation as dτ = NG dTG, where NG = (Uµ∇µTG)−1. This formulation separates the temporal reparametrization freedom already present in Robertson–Walker spacetime from the phenomenological realization of a non-unit cosmological temporal rate. Covariance fixes how this relation is represented under coordinate changes but does not select a preferred temporal parameter.

Cosmological Time Dilation Test

The paper develops the cosmological time dilation (CTD) test, showing that for null-linked emission and observation intervals, the proper-time relation remains ∆τo = (1 + z)∆τe. When expressed in terms of TG, this yields ∆TG,o/∆TG,e = (1 + z)NG,e/NG,o. The normalized timing ratio RT (z) is defined as the ratio of these intervals.

Phenomenological Model Realization

The minimally extended varying speed of light (meVSL) framework provides a phenomenological realization where the temporal rate is given by NG(a) = a b/4, with ao = 1 and NG,o = 1 fixing the present normalization. This parameter b governs correlated cross-epoch scalings of dimensional quantities such as c, G, ħ, and corresponding matter and electromagnetic quantities. The endpoint-rate relation in this parametrization is RT (z) = RG(e, o) = (1 + z)−b/4.

Observable Scope and Conclusion

The paper concludes that the observable content must be derived from the complete combination of quantities entering a specified measurement, as isolated relations like ce = Nc0 are not by themselves coordinate-independent observables. CTD constrains the temporal relation in Eqs. (9)–(15), not the separate evolution of c, G, ħ, or another dimensional quantity. The central empirical question is whether the unit-rate realization NG = 1 is sufficient for cross-epoch cosmological timing or whether observations support a nontrivial NG(a) within the meVSL realization.

Relation to Dimensional Constants

Dimensional quantities are assigned correlated slice-wise realizations so that local equations and defining relations retain their standard form on each homogeneous hypersurface. Within meVSL, the factors Ni = NG,i on a homogeneous hypersurface Σi are given by Ni = a b/4 i. The paper notes that Gi/c4 i, ħici, and mic2 i exhibit how corresponding local relations are preserved while the photon-energy relation retains the physical factor a−1i. GCT/meVSL does not assign unit-independent meaning to the numerical variation of c, G, ħ, or m taken separately.

Source Equivalence Condition

To test the CTD relation using distinct astrophysical sources at different redshifts, an additional source-equivalence condition is required: the compared sources must represent the same intrinsic timescale distribution. This condition can be formulated in GCT intervals pG(WG θ, z) = pG(WG θ) or directly in proper-time intervals pτ (Wτ θ, z) = pτ (Wτ θ). If source equivalence is formulated for WG, Eq. (9) gives WG,o = (1 + z)NG,e/NG,o WG,e. A robustly inferred b ≠ 0 would indicate a non-unit endpoint-rate scaling within this parametrization after source evolution and observational systematics have been controlled.

Lapse Variation and Constraints

The standard Einstein–Hilbert gravitational action leads to the nondynamical character of the RW lapse, as its conjugate momentum vanishes pN = ∂Lg/∂N˙ = 0. The reduced Lagrangian contains no time derivative of the lapse, and variation with respect to N yields the Hamiltonian/Friedmann constraint rather than an independent propagation equation for N. This nondynamical nature is emphasized when imposing a temporal parametrization such as N(TG) = a b/4 after variation with respect to the lapse, so that the Hamiltonian constraint is retained explicitly. The non-unit-lapse parametrization does not introduce a new propagating lapse degree of freedom.

Summary

The isolated numerical variation of a dimensional quantity is unit dependent and is not, by itself, a measurable statement. The question addressed here is narrower: whether physical standards associated with different cosmological epochs can never enter unit-independent cross-epoch relations when the temporal parametrization of RW spacetime is kept explicit. This paper provides the temporal framework for posing that question without assigning independent observable significance to the numerical evolution of any single dimensional constant.

Acknowledgments

SL is supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Science and ICT under Grant No. NRF-2022R1A2C1005050.

References

[1] J. P. Uzan, “The Fundamental Constants and Their Variation: Observational Status and Theoretical Motivations,” Rev. Mod. Phys. 75, 403 (2003) doi:10.1103/RevModPhys.75.403 [arXiv:hep-ph/0205340 [hep-ph]]

[2] J. Magueijo, “New varying speed of light theories,” Rept. Prog. Phys. 66, 2025 (2003) doi:10.1088/0034-4885/66/11/R04 [arXiv:astro-ph/0305457 [astro-ph]]

[3] M. J. Duff, L. B. Okun and G. Veneziano, “Trialogue on the number of fundamental constants,” JHEP 03, 023 (2002) doi:10.1088/1126-6708/2002/03/023 [arXiv:physics/0110457 [physics]]

[4] M. J. Duff, “Comment on time variation of fundamental constants,” [arXiv:hep-th/0208093 [hep-th]]

[5] G. F. R. Ellis and H. van Elst, “Cosmological models: Cargese lectures 1998,” NATO Sci. Ser. C 541, 1-116 (1999) doi:10.1007/978-94-011-4455-l 2 [arXiv:gr-qc/9832046 [gr-qc]]

[6] J.

Improvements for AI systems

Improved AI can perform cross-epoch timing tests by analyzing null-linked intervals expressed in Generalized Cosmological Time (TG), such as those satisfying ∆TG,o/∆TG,e = (1 + z)NG,e/NG,o.
The system can distinguish between the unit-rate realization NG = 1 and a non-unit rate NG(a) within the meVSL framework by comparing the proper-time relation ∆τo = (1 + z)∆τe with its TG counterpart.
The AI can infer endpoint rates in Generalized Cosmological Time using NG,o = 1 gives a dilation factor (1 + z) / 1 - b/4 if the intrinsic timescale is defined in TG, contrasting it with the usual (1 + z) law recovered in proper time.
The system can apply source equivalence conditions, such as pG(WG θ, z) = pG(WG θ), to compare distinct astrophysical sources at different redshifts using either GCT intervals or proper-time intervals.
The AI can model correlated dimensional scalings in the meVSL framework by utilizing the relationship NG(a) = a b/4, which governs the correlated cross-epoch scalings of dimensional quantities such as c, G, ħ.

Abstract

The standard Duff-type argument states that the numerical variation of a dimensional constant is unit-dependent when considered in isolation from the physical standards used to compare it, and is therefore not by itself an observable. We ask whether this conclusion is sufficient to determine the observational status of complete relations connecting different cosmological epochs. In a Robertson--Walker spacetime, the cosmological principle and Weyl's postulate fix the homogeneous geometry and fundamental timelike congruence without uniquely fixing the temporal rate along it. The Hamiltonian constraint likewise does not dynamically select that rate. Consequently, Robertson--Walker spacetime admits a family of homogeneous temporal parametrizations. Generalized Cosmological Time retains a specified common member, T G, related to proper time by d tau = N G d T G, with (tau, 1) as the unit-rate special member of the same class. In the minimally extended varying speed of light framework, length--mass--time quantities obey correlated slice-wise scalings chosen to preserve the same local physical laws on every homogeneous hypersurface. These scalings form a unit-rescaling structure, so purely local dimensionless observables are insensitive to N G. Cross-epoch observables require more than such local scalings. They also require local reference standards, a propagation or comparison map, and, when distinct source populations are compared, a source-equivalence prescription. Cosmological time dilation illustrates this distinction: its interpretation depends on source standardization and residual source evolution. We distinguish the unit dependence of individual dimensional quantities from the observational content of complete cross-epoch relations. The former does not by itself determine the latter.