{"id":"348b7382-1d11-4c3a-8e8d-23292d41e9ee","arxiv_id":"2505.07200","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Using OH 18 cm absorption lines toward PKS 1830-211, the authors place 1-sigma limits of about 6e-5 on changes in alpha times g_p^0.27, 2e-3 on the fine structure constant, and 8e-3 on the proton g-factor over 7.3 Gyr, all consistent with no evolution.","lead":"This paper re-analyzes radio observations of hydroxyl molecules in a distant galaxy to test whether the constants of physics were different 7.3 billion years ago. The measured changes are all consistent with zero, adding a new data point to the search for drifting constants.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The quoted alpha and g_p limits are only as secure as the assumption that OH excitation produces no frequency offsets between the 1665/1667 and 1612/1720 lines; the KS/AD test does not bound that offset.","rationale":"The paper is an honest, internally consistent re-analysis of published MeerKAT data. The algebra in Section 4 checks out, the parameter count is small, and the authors do not overclaim: they explicitly state that the conjugate OH satellite behavior is a crucial assumption and that their constraints are weaker than the existing mu constraint. The most load-bearing point is not the mu sign convention or the low precision; it is that the reported 1-sigma limits on alpha and g_p depend on the absence of excitation-induced frequency shifts between OH lines. Because the relevant observable combinations are small differences of large frequencies, a few km/s systematic centroid offset would shift the derived limits by order 10^-3, comparable to their quoted uncertainties. The KS/AD tests in Section 3 only establish that the residual sum is consistent with noise; they do not constrain the size of a non-conjugate frequency shift, since the tied centroids in the fit can absorb such a shift. The reader's CONDITIONAL verdict is therefore appropriate: the paper should not be rejected, but the systematic floor should be quantified with an unconstrained refit (or a non-LTE radiative transfer calculation) before these limits are used as a reference data point.","tokens_in":9992,"tokens_out":11627,"duration_ms":116200,"concrete_test":"Refit the published MeerKAT spectra of Combes et al. (2021) with the centers of the OH 1665 and 1667 G1/G2 components, and the 1612 and 1720 G3/G4 components, allowed to vary independently while keeping amplitudes and FWHMs tied by the LTE/conjugate model. Compare the unconstrained relative velocities Delta v(1665-1667) and Delta v(1612-1720) for each component to zero. If the offsets are consistent with zero within about 1-3 km/s, the current limits stand; if they scatter by several km/s, add that dispersion as a systematic floor to the uncertainties in Eq. (5) and recompute Delta alpha/alpha and Delta g_p/g_p. This directly bounds the size of the excitation-induced offset that the KS/AD test cannot.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central limits in Eq. (5) rest on the assumption that the frequency offsets FO1 and FO2 fitted in Section 3 are purely due to varying fundamental constants. The observable combinations nu1667-nu1665 and nu1720-nu1612 are small differences of large frequencies, so a tiny astrophysical centroid shift is amplified into a 10^-3 fractional shift in Delta(alpha g_p^0.27), Delta alpha/alpha, and Delta g_p/g_p. For example, the 7.9x10^-3 uncertainty in Eq. (5a) corresponds to ~16 kHz in the 2 MHz 1665-1667 difference, equivalent to a few km/s. The line fitting ties the redshifts of the thermal G1/G2 pairs and the conjugate G3/G4 pairs, so any excitation or radiative-transfer asymmetry is either absorbed into the fitted centroids or hidden in the residuals. The KS and AD p-values (0.64, 0.46) are computed on the summed residual after subtracting the conjugate-tied model; they test Gaussianity of the residual noise, not the magnitude of a possible non-conjugate frequency shift. The paper itself calls the conjugate behavior 'a crucial assumption' but never bounds the size of an excitation-induced offset. With line widths of 30-70 km/s, centroid shifts of a few km/s from opacity or pumping asymmetries are physically plausible and would move the derived limits by an amount comparable to their stated 1-sigma uncertainty.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper re-analyzes the MeerKAT OH 18 cm absorption spectra toward PKS 1830–211 at z≈0.8858, fitting the four lines with a model consisting of two thermal (LTE) Gaussian components and two conjugate satellite components, with global frequency offsets FO1 and FO2 between the 1665/1612 lines and the 1667 line. Using the sensitivity equations of Chengalur & Kanekar (2003) and adopting the independent methanol-based Δμ/μ = (-1.8±1.2)×10^-7 (Muller et al. 2021), the authors derive Δ(αg_p^0.27)/(αg_p^0.27) = (-3.2±5.7)×10^-5, Δα/α = (-3.3±23.0)×10^-4, and Δg_p/g_p = (1.1±7.9)×10^-3, which they interpret as upper limits consistent with no evolution over 7.3 Gyr.","tokens_in":10293,"tokens_out":15171,"duration_ms":139569,"significance":"If the underlying assumptions hold, this is a useful new null measurement of fundamental-constant evolution at z=0.89, and one of the very few cosmological constraints on the proton g-factor. The use of same-species lines avoids inter-species velocity offsets, and the simultaneous fit with tied strengths and widths is a reasonable approach. The statistical checks (reduced χ²=1.17, KS and AD p-values of 0.64 and 0.46) are reported transparently, and the paper is honest about the key assumptions. The main value is as a new data point for the evolution of α and especially g_p, though the precision is modest compared to other probes of α.","major_comments":[{"comment":"The paper states that 'a crucial assumption is the existence of conjugate OH satellite lines,' but the KS and AD tests (p = 0.64 and 0.46) are performed on the summed residual after the conjugate-tied model is subtracted; they test whether that residual is consistent with Gaussian noise, not whether the data exclude a non-conjugate component of the size relevant here. A centroid offset of ~1–2 km/s between the 1612 and 1720 conjugate components (or between the 1665 and 1667 main-line components) would produce residual amplitudes of order A×Δv/FWHM ≈ 2×10^-5, below the per-channel rms of ~2×10^-4, so the test cannot bound such offsets. Since the derived limits on Δα/α and Δg_p/g_p in Eq. (5) correspond to frequency offsets of order 1–10 kHz (roughly 0.2–2 km/s), an excitation-induced offset of this size would shift the central values by an amount comparable to their stated uncertainties. The paper should either constrain the size of possible excitation-induced frequency offsets from OH radiative-transfer/excitation models, or add a systematic error term to the limits that accounts for this. As written, the quoted constraints in Eq. (5) and the Summary rest entirely on an assumption whose associated uncertainty is not bounded.","section":"Section 3, conjugate-line assumption"},{"comment":"The combination of Eqs. (1a),(2a) and (1b),(2b),(3),(4) uses inverse-variance weighting, which assumes the individual component measurements are statistically independent. However, G1–G4 are fitted simultaneously with global offsets FO1 and FO2, so the parameter estimates are correlated. Without the covariance matrix (or a justification of independence), the quoted uncertainties of 7.9×10^-3 and 2.1×10^-4 in Eq. (5) may be underestimated. Please propagate the full covariance of the fit into the combined constraints, or explain why the components can be treated as independent.","section":"Section 4, Eq. (5)"},{"comment":"The fitted values and uncertainties of the global frequency offsets FO1 and FO2, which are the direct observables entering Eqs. (1)–(5), are not reported anywhere. The reader cannot verify the error propagation from the line centroids to the final constant limits. The paper should list FO1, FO2 (and their covariance), and state explicitly how the uncertainties on the right-hand sides of Eqs. (1)–(4) are derived from the fit.","section":"Section 3, Table 1"}],"minor_comments":[{"comment":"The reported '1σ limits' are actually the 1σ uncertainties of the fitted constants rather than upper limits computed from the full posterior: for Δg_p/g_p = (1.1±7.9)×10^-3, the 1σ upper limit is 9.0×10^-3, not 7.9×10^-3. Please clarify how the limits are defined and whether the central values are being included.","section":"Section 6"},{"comment":"The text refers to 'comparisons of the redshifts between ν1667+ν1665 and ν1667-ν1665,' but the equations appear to use frequency combinations with a different normalization. Please rewrite this sentence to match the actual equations and definitions used.","section":"Section 4"},{"comment":"There are encoding artifacts in the displayed text (e.g., '/uni03BC/' and 'Normali ed flux den ity'), which should be fixed in the final version.","section":"Figure 1 and Table 1"},{"comment":"The abstract quotes limits with the symbol '≲' but the body gives central values and uncertainties; please be consistent about whether the constraints are 1σ upper limits or 1σ uncertainties, and specify the confidence level (1σ, 2σ, etc.).","section":"Abstract and Summary"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of A&A. The central issue is whether the systematic uncertainty from the conjugate-line assumption can be adequately bounded. I would encourage the authors to perform a sensitivity test (e.g., refit with the 1612 and 1720 centroids allowed to differ by a free parameter and report the resulting change in limits) and to provide the fit covariance. If they can do that, the paper could become acceptable. The lack of reported FO1/FO2 values is also a reproducibility concern."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nHere's my read of arXiv:2505.07200. The paper does one thing and does it honestly: it re-fits the MeerKAT OH 18 cm absorption toward PKS 1830-211 and extracts constraints on alpha, mu, and g_p at z~0.89, using the established Chengalur & Kanekar sensitivity coefficients. That is a genuinely new measurement—the tightest combination Delta(alpha g_p^0.27) at this absorber, and one of the very few constraints on g_p at any cosmological redshift. The fitting is careful: reduced chi-squared of 1.17, residuals tested, and the authors are upfront that the constraints are weak and consistent with no evolution.\n\nThe algebra in Eqs. (1)-(5) checks out, and the input Delta mu/mu from methanol is an independent prior, so there is no circularity. The paper is well aware of the low precision and why it is low (the nu1720-nu1612 and nu1665-nu1667 differences are small, so redshift uncertainties dominate). That is all to the good.\n\nWhere I would push back is the thing the stress-test flags: the conjugate satellite line assumption. The 1612/1720 lines are assumed to be exactly conjugate, with any asymmetry absorbed by the extra Gaussians. But the KS/AD tests on the summed residual only show the noise after subtracting the conjugate-tied model is Gaussian; they do not bound a non-conjugate frequency shift. A few km/s centroid offset from excitation or opacity effects would move the derived alpha and g_p limits by a good fraction of their 1-sigma uncertainties. The paper calls this a 'crucial assumption' but never quantifies its possible size. That is the main soft spot, and it is a real one, though it does not obviously overturn the 'consistent with no evolution' conclusion—the limits are so weak that a few km/s would not flip the sign.\n\nSecondary nits: no systematic error budget for baseline/continuum subtraction (the authors argue variability is negligible, which is plausible but not quantified), and the mu sign/convention could confuse readers given the electron-proton vs proton-electron switch between papers. Minor.\n\nSo who is this for? Someone compiling a census of fundamental-constant measurements at high redshift. It is a legitimate new data point, but I would treat the error bars as optimistic until the conjugate offset is bounded. It deserves a serious referee, mostly to force that quantification. I would send it out rather than desk-reject.","headline":"A new but weak data point on alpha and g_p at z=0.89 that hangs on an unquantified conjugate-line assumption; fine as a null result, needs systematics before being used as a reference.","tokens_in":10877,"tokens_out":2258,"would_cite":true,"duration_ms":22948,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"New limits from hydroxyl absorption lines at z = 0.89 bound changes in the fine-structure constant, the proton-electron mass ratio, and the proton g-factor over 7.3 Gyr, all consistent with no evolution.","keywords":["fundamental constants","fine structure constant","proton g-factor","OH 18 cm lines","conjugate satellite lines","absorption spectroscopy","PKS 1830-211","cosmological evolution of constants"],"falsifier":"Fitting the same spectra while allowing the 1612 and 1720 MHz rest frequencies to float independently, or modeling the OH excitation with a radiative-transfer code that predicts an intrinsic $1612{-}1720$ offset at the $\\sim 10^{-5}$ fractional level, would settle whether the conjugate assumption is safe; a still cleaner check is to observe a local ($z=0$) OH absorber with similar physical conditions and test whether the satellite lines are exactly conjugate when the constants are known to be unchanged.","tokens_in":9806,"feed_emoji":"🔭","tokens_out":10514,"duration_ms":86114,"temperature":0.7,"pith_summary":"This paper tries to establish that the fine-structure constant $\\alpha$, the proton-electron mass ratio $\\mu$, and the proton g-factor $g_p$ have not changed detectably over the past 7.3 billion years. The evidence comes from a re-analysis of the four hydroxyl (OH) 18 cm absorption lines toward the lensed quasar PKS 1830-211 at $z = 0.89$, where all four lines arise in the same gas and any frequency mismatch between them carries information about constant evolution. Combining a methanol-based prior of $\\Delta\\mu/\\mu = (-1.8\\pm1.2)\\times10^{-7}$ with the fitted OH line offsets yields upper limits of $\\Delta(\\alpha g_p^{0.27})/(\\alpha g_p^{0.27}) \\lesssim 5.7\\times10^{-5}$, $\\Delta\\alpha/\\alpha \\lesssim 2.3\\times10^{-3}$, and $\\Delta g_p/g_p \\lesssim 7.9\\times10^{-3}$. The result matters because unified theories predict slow drift in these constants, and radio absorption lines offer the cleanest astronomical probe over cosmological timescales.","feed_headline":"Hydroxyl lines cap drift in physics constants at 7.3 Gyr","feed_subtitle":"18 cm hydroxyl absorption at z=0.89 finds no drift in α or g_p over 7.3 billion years.","key_machinery":"The load-bearing object is the frequency pattern of the four OH 18 cm lines. In the rest frame the frequencies satisfy $\\nu_{1612} + \\nu_{1720} = \\nu_{1665} + \\nu_{1667}$, and each transition has a different sensitivity coefficient to $\\alpha$, $\\mu$, and $g_p$. Because the lines come from the same gas, physical velocity offsets cancel; only differential shifts remain. The fitting procedure adds two free offsets, one between the 1665 and 1667 MHz lines and one between the 1612 and 1667 MHz lines, models the 1612/1720 pair as conjugate absorption/emission, and converts the fitted offsets into the linear equations that constrain the constant variations.","core_discovery":"The central discovery is a null measurement of cosmological constant evolution in the $z = 0.8858$ absorber toward PKS 1830-211. By simultaneously fitting two thermalized OH main lines (1665 and 1667 MHz) and two conjugate OH satellite lines (1612 absorption and 1720 emission), the paper obtains two linear combinations relating $\\Delta\\mu/\\mu$, $\\Delta\\alpha/\\alpha$, and $\\Delta g_p/g_p$. Substituting the prior $\\Delta\\mu/\\mu = (-1.8\\pm1.2)\\times10^{-7}$ gives $\\Delta(\\alpha g_p^{0.27})/(\\alpha g_p^{0.27}) = (-3.2\\pm5.7)\\times10^{-5}$, $\\Delta\\alpha/\\alpha = (-3.3\\pm23.0)\\times10^{-4}$, and $\\Delta g_p/g_p = (1.1\\pm7.9)\\times10^{-3}$, which the paper quotes as 1$\\sigma$ upper limits $5.7\\times10^{-5}$, $2.3\\times10^{-3}$, and $7.9\\times10^{-3}$. It concludes that the constants are consistent with no evolution over a look-back time of 7.3 Gyr.","pith_inferences":["Editorial inference: if a second OH 18 cm absorber is found at a different redshift, the same pipeline would turn this single null result into a two-epoch measurement, which is the minimum needed to separate a real drift from a constant systematic offset.","Editorial inference: the $g_p$ and $\\alpha$ limits are dominated by the satellite-line frequency combination $\\nu_{1720}-\\nu_{1612}$, so deeper integrations on the 1720 MHz line alone could tighten both limits substantially without new instruments.","Editorial inference: the conjugate-line assumption is only validated statistically on noise-dominated residuals; a velocity-resolved comparison of the 1612/1720 line ratio against the fitted offsets would test whether an excitation effect is masquerading as a constant change."],"forward_implications":["If the limits are correct, any model predicting fractional changes larger than roughly $10^{-5}$ in $\\alpha g_p^{0.27}$, $10^{-3}$ in $\\alpha$, or $10^{-2}$ in $g_p$ by $z < 1$ is ruled out.","The measurement adds a second, independent constraint at $z = 0.89$ alongside the much tighter methanol-based $\\mu$ bound, and it is one of the few direct bounds on $g_p$ evolution.","The null result extends constant-drift tests to a look-back time of 7.3 Gyr, beyond the reach of terrestrial atomic-clock and Oklo reactor limits.","With only one OH absorber known above $z = 0.1$, this is currently the highest-redshift anchor for simultaneous $\\alpha$, $\\mu$, and $g_p$ constraints from a single species."],"supporting_citations":[{"why":"Supplies the MeerKAT OH 18 cm spectra of the z=0.8858 absorber and the conjugate satellite-line interpretation that this paper re-fits.","marker":"Combes et al. 2021"},{"why":"Provides the prior $\\Delta\\mu/\\mu = (-1.8\\pm1.2)\\times10^{-7}$ from methanol lines used to separate $\\alpha$ and $g_p$.","marker":"Muller et al. 2021"},{"why":"Gives Eqs. (12) and (13), the linear relations connecting OH line redshifts to $\\Delta\\alpha/\\alpha$, $\\Delta\\mu/\\mu$, and $\\Delta g_p/g_p$.","marker":"Chengalur & Kanekar 2003"},{"why":"Establishes the OH 18 cm method and the relative precision of the frequency combinations used here.","marker":"Kanekar & Chengalur 2004"},{"why":"Originates the idea that the four OH 18 cm lines from the same gas can simultaneously constrain $\\alpha$, $\\mu$, and $g_p$.","marker":"Darling 2004"},{"why":"Demonstrates the use of conjugate OH satellite lines for fundamental-constant constraints, supporting the assumption adopted in the fit.","marker":"Kanekar et al. 2018"}],"fun_headline_variants":["No constant drift over 7.3 Gyr from OH lines","7.3 Gyr back, physics constants unchanged","Hydroxyl absorption limits constant evolution","α and g_p stable for 7.3 billion years","MeerKAT OH lines freeze constants at z=0.89"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result collapses if the four OH 18 cm lines do not trace exactly the same gas, or if excitation and radiative-transfer effects produce intrinsic frequency shifts between the 1612 and 1720 MHz satellite lines, because the fitting would then attribute those shifts to changes in the constants.","fun_headline_variants_meta":{"raw":{"variants":["No constant drift over 7.3 Gyr from OH lines","7.3 Gyr back, physics constants unchanged","Hydroxyl absorption limits constant evolution","α and g_p stable for 7.3 billion years","MeerKAT OH lines freeze constants at z=0.89"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000572,"raw_usage":{"total_tokens":2760,"prompt_tokens":1060,"completion_tokens":1700,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":676,"completion_tokens_details":{"reasoning_tokens":1618}},"tokens_in":676,"tokens_out":1700,"duration_ms":12395,"temperature":1.0,"reasoning_tokens":1618,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:23:43.197649+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fitting the same spectra while allowing the 1612 and 1720 MHz rest frequencies to float independently, or modeling the OH excitation with a radiative-transfer code that predicts an intrinsic $1612{-}1720$ offset at the $\\sim 10^{-5}$ fractional level, would settle whether the conjugate assumption is safe; a still cleaner check is to observe a local ($z=0$) OH absorber with similar physical conditions and test whether the satellite lines are exactly conjugate when the constants are known to be unchanged.","supporting_citations":[{"cited_title":"2021, A&A, 648, A116","cited_arxiv_id":null,"evidence_quote":"Supplies the MeerKAT OH 18 cm spectra of the z=0.8858 absorber and the conjugate satellite-line interpretation that this paper re-fits."},{"cited_title":"M., Henkel, C., & Kanekar, N","cited_arxiv_id":null,"evidence_quote":"Provides the prior $\\Delta\\mu/\\mu = (-1.8\\pm1.2)\\times10^{-7}$ from methanol lines used to separate $\\alpha$ and $g_p$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives Eqs. (12) and (13), the linear relations connecting OH line redshifts to $\\Delta\\alpha/\\alpha$, $\\Delta\\mu/\\mu$, and $\\Delta g_p/g_p$."},{"cited_title":"& Chengalur, J","cited_arxiv_id":null,"evidence_quote":"Establishes the OH 18 cm method and the relative precision of the frequency combinations used here."},{"cited_title":"2004, ApJ, 612, 58","cited_arxiv_id":null,"evidence_quote":"Originates the idea that the four OH 18 cm lines from the same gas can simultaneously constrain $\\alpha$, $\\mu$, and $g_p$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates the use of conjugate OH satellite lines for fundamental-constant constraints, supporting the assumption adopted in the fit."}],"review_version":1}