{"id":"67259c0c-2069-49e3-a1ea-36acd6a5ec7c","arxiv_id":"2501.18125","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A backward search in time from known gravitational wave detections could detect or constrain superluminal non-tensor polarizations, and the authors argue this is feasible with current detectors.","lead":"Gravitational waves predicted by some modified theories of gravity include extra polarizations that would travel faster than light and arrive at a detector before the usual tensor modes. This paper proposes scanning the data recorded before known detections to find or rule out such early-arriving signals.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Feasibility rests on an undemonstrated sensitivity claim: no injection or template-bank study shows that non-tensor polarizations at allowed amplitudes can be recovered from the searched pre-merger data.","rationale":"The reader's weakest-assumption analysis identified the same load-bearing gap: the search only works if non-tensor polarizations have amplitudes current detectors can see, and the paper does not demonstrate this. My stress-test confirms that this is the central unsupported link in the abstract's feasibility conclusion. The paper is honest about the gap, explicitly deferring the needed sensitivity and template-bank studies to future work in Section IV, but the abstract's claim that the test is 'feasible with current detectors' goes beyond what is shown. I considered whether the Appendix A energy-loss argument contains an internal inconsistency—the O(ζ) cross term in Eq. (A.11) might bias GR-derived source parameters at first order rather than second order—but the cross term involves a modification of the tensor polarization itself, not the additional non-tensor polarizations, so the paper's statement that energy carried by additional polarizations scales as ζ^2 is not directly contradicted. The more decisive and clearly load-bearing issue is the missing sensitivity demonstration. This does not invalidate the proposal; it means the paper should be accepted conditionally on the promised injection and template-bank studies, which is exactly the reader's verdict. No change to the verdict is therefore needed.","tokens_in":22696,"tokens_out":11152,"duration_ms":122758,"concrete_test":"Conduct an injection study on real O2/O3 data, e.g., for GW170817 and GW190425: inject scalar-breathing and vector-polarization chirp signals with source parameters fixed to the tensor-inferred values, placed at time offsets corresponding to δv = 10^-8, 10^-7, ..., 10^-5, with injected amplitudes spanning 1% to 100% of the tensor signal's amplitude. Run the proposed targeted search with the adapted Einstein-æther or generic non-tensor template bank, and require recovery with a false-alarm rate below 1 per decade at the designed detection threshold. Report the minimum detectable relative amplitude as a function of δv. If that minimum lies above the amplitudes allowed by current stochastic-background bounds or by specific theory constraints, the feasibility claim would be unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that the proposed test is feasible with current detectors. This requires that for at least some already-detected events, non-tensor polarizations emitted with amplitudes consistent with existing constraints would actually be detectable in the earlier data. The paper does not establish this. Section IIIC argues only that stochastic-background upper limits do not rule out detectable amplitudes, but those limits bound a background, not the single-event non-tensor amplitude from a particular source. Section IV explicitly defers to future work the construction and performance of the template bank and the question of whether source parameters derived from GR templates are sufficiently unbiased for the search. No injections are performed, no minimum detectable amplitude is computed, and no false-alarm rate or search-efficiency estimate is given. The kinematic framework and the treatment of observing-time gaps are useful, but the transition from 'physically well-motivated' to 'feasible with current detectors' is an extrapolation, not a demonstrated capability. The authors themselves flag this gap when they write that they leave these questions to future work.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a new, model-agnostic test for superluminal non-tensor gravitational-wave polarizations. Building on the facts that Cherenkov constraints require extra polarizations to propagate at or above the speed of light and that GW170817 pins the tensor speed to c, the authors argue that any non-tensor modes from a given source must arrive at the detector no later than the tensor modes. Starting from an already-detected merger event, one can therefore search backward through the earlier data for non-tensor polarization signals using source parameters inferred from the tensor analysis. The paper derives the kinematic relation between arrival-time difference and propagation-speed difference, analyzes how observing-time gaps affect the parameter space that a non-detection could exclude, discusses stacking of multiple events and changing detector sensitivity, and considers existing amplitude constraints. It concludes that the test is physically well motivated and feasible with current detectors, while noting that constructing the template bank and assessing its performance are left to future work.","tokens_in":22935,"tokens_out":11354,"duration_ms":123509,"significance":"If implemented, the proposed search would open a new window on non-tensor polarizations that arrive before the tensor signal, a region missed by standard coincident matched-filter searches and null-stream tests. The kinematic derivation in Sec. IIB is simple and correct, and the use of real events and duty-cycle data to illustrate gap filling is concrete and useful. The paper is also admirably transparent: it explicitly identifies the main open questions, such as template-bank construction and parameter bias, and it does not overclaim the reach of existing amplitude constraints. The main weakness is that the abstract's central claim of feasibility with current detectors is not supported by a quantitative sensitivity demonstration; no injections, minimum detectable amplitudes, or search-efficiency estimates are given. This gap is fixable and does not invalidate the underlying idea, but it is load-bearing for the paper's headline conclusion.","major_comments":[{"comment":"The central claim that the proposed test is 'feasible with current detectors' (abstract and Sec. IV) is not supported by a sensitivity demonstration. The paper presents no injections, no template-bank construction, no minimum detectable amplitude, and no false-alarm or search-efficiency estimate; Sec. IV explicitly says that these questions are left to future work. The GWB upper limits quoted in Eq. (3.2) bound the stochastic background, not the single-event non-tensor amplitude from a particular source, so they do not establish that detectable amplitudes are allowed. To support the feasibility claim, the authors should compute, for at least one existing event such as GW170817, the expected SNR of non-tensor polarizations as a function of their amplitude and speed, or perform an injection study into O3 data with a concrete search pipeline.","section":"Sec. IV / Sec. IIIC"},{"comment":"The search window preceding t_merger is not empty: it contains the tensor inspiral of the same source. The paper does not explain how the proposed search for non-tensor polarizations handles this known foreground—for example, by masking the final seconds or minutes of the window, subtracting the best-fit tensor waveform, or using templates that are exactly orthogonal to the tensor modes. Without such a procedure, the false-alarm behavior of the search and the interpretation of a non-detection are ambiguous, which is directly relevant to the feasibility claim.","section":"Sec. IIB, Fig. 1"},{"comment":"The regions in Figs. 7-9 are computed under the implicit assumption that any non-tensor signal with Δt inside a searched data segment would be detected. Since no detection efficiency is presented, these shaded regions are prospective upper limits on constraining power, not actual exclusions. The text should state this more prominently, particularly in the caption of Fig. 9, where 'the regions ... that could be excluded' could be misread as a demonstrated result.","section":"Sec. IIIA, Figs. 7-9"}],"minor_comments":[{"comment":"The statement that a non-detection places a 'lower bound on the speed' is correct only for polarizations with amplitude above the detection threshold. The text says this, but it would help to repeat it when interpreting Fig. 5, where the 'best-case' curves might be mistaken for unconditional constraints.","section":"Sec. IIB, Fig. 5"},{"comment":"The stacking argument assumes that the extra-polarization propagation speed is the same for all events. This assumption should be stated in the main text as well, since a theory with energy-dependent speeds would not allow the gap-filling shown in Figs. 7-8.","section":"Sec. IIIA, footnote 6"},{"comment":"The conclusion that parameter bias from energy loss through additional polarizations is negligible relies on the assumption that all non-GR deviations scale with a single coupling ζ and that O(ζ^2) terms vanish. This assumption is strong and should be flagged as such; it is not a general theorem.","section":"Appendix A"},{"comment":"The five events highlighted in the text as giving the best constraints are hard to identify in the dense figure; adding labels or markers would improve readability.","section":"Fig. 9"}],"recommendation":"major_revision","confidential_remarks":"The paper is transparent about its limitations and the kinematic framework is sound. My main concern is that the abstract's feasibility claim outruns the evidence; however, this is fixable with an injection study or a sensitivity estimate, so I recommend major revision rather than rejection. I would also encourage the editor to consider whether the paper's length and style fit a full research article if the sensitivity study is not added."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a genuine new idea, cleanly argued, and it deserves a proper referee—but the paper's headline claim 'feasible with current detectors' is an extrapolation, not a demonstration. The new step is simple and I think it works: take a detected tensor event, use its source parameters, and search the earlier data for scalar or vector polarizations that must have arrived no later than the tensor modes. That reframing—anchor to the tensor event and look backward—is the useful contribution. The kinematics, the Cherenkov/GW170817 logic, and the discussion of data gaps and event stacking are all sound and clearly written.\n\nThe paper is also honest in the right places. It flags that the search requires a non-tensor amplitude large enough to see, and that the authors do not yet know whether the template banks can deliver. It correctly notes that stochastic-background limits do not rule out detectable single-event amplitudes. And it does not oversell the Einstein-æther example: the coupling constants are already so tight that the new speed constraint adds little there.\n\nThe soft spot is not a math error; it is that the central feasibility claim rests on unverified assumptions. There are no injections, no minimum detectable amplitude, no false-alarm estimate, and no template-bank study. The paper defers all of these to future work. That is acceptable for a proposal, but the conclusion 'feasible with current detectors' goes beyond what is shown. If the authors had framed it as 'we identify a promising search region and the obstacles to reaching it,' the claim would match the evidence. The expected constraints are also modest (parts in 10^9), which the paper admits, so the payoff is narrow but not zero.\n\nThe citation practice is fine; the heavy reliance on their own earlier work is natural since that is the same line of argument. I disagree with anyone who'd desk-reject this: it's a testable idea with a clear pipeline, and the missing pieces are exactly the kind of thing a good referee can push on. My recommendation is to send it to review, with a request that the authors either provide a detection-sensitivity estimate or soften the feasibility language.","headline":"A sound, genuinely new search proposal for superluminal non-tensor GW polarizations, but 'feasible with current detectors' overstates what is demonstrated.","tokens_in":23405,"tokens_out":2553,"would_cite":true,"duration_ms":27042,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["04.30.-w","04.80.Nn"],"model":"deepseek-v4-flash","headline":"A new test seeks gravitational-wave polarizations that arrive before the main signal, using data already collected around known detections.","keywords":["gravitational waves","gravitational wave polarizations","superluminal propagation","tests of general relativity","modified gravity","scalar and vector polarizations","matched-filter searches","Einstein-æther theory"],"falsifier":"Run the proposed template bank over real pre-event data for nearby catalog events with simulated scalar and vector modes injected at known speeds and amplitudes; if injections with amplitudes near the detector noise floor are not recovered, the feasibility claim fails. A non-tensor polarization observed arriving after the tensor modes would likewise falsify the underlying premise that all extra polarizations are superluminal.","tokens_in":22519,"feed_emoji":"🌊","tokens_out":8187,"duration_ms":79174,"temperature":0.7,"pith_summary":"In general relativity a gravitational wave carries only two transverse-traceless tensor polarizations, but many modified theories allow up to six polarizations that can each travel at a different speed, some faster than light. Existing observations force any such non-tensor polarizations to arrive at the detector no later than the tensor modes. The paper therefore proposes taking events that have already been detected and scanning the earlier stretch of data, with source parameters fixed from the tensor signal, for scalar or vector polarizations from the same source. If found, such early signals would be direct evidence of physics beyond general relativity; if not found, the absence still converts into a lower bound on propagation speed for detectable amplitudes and an upper bound on amplitude for detectable speeds. The authors argue that this test is feasible with data already collected.","feed_headline":"Search early data to catch extra gravitational-wave polarizations","feed_subtitle":"If extra polarizations outrun light, they arrive before the main signal—and a null search still tightens limits.","key_machinery":"The load-bearing object is the arrival-time lead $\\Delta t_N = t_T - t_N = D_L/c - D_L/v_N$ between the tensor and non-tensor polarizations of the same source, together with its re-expression as a fractional speed excess $\\delta v_N = v_N/c - 1 = c\\Delta t_N/(D_L - c\\Delta t_N)$. This identity turns a calendar of searched data into a two-dimensional exclusion region in the plane of propagation speed and amplitude: the searched time $T_{\\rm Obs}$ fixes the minimum measurable lead, the source distance $D_L$ fixes how that lead maps to speed, and the detector noise floor fixes the minimum detectable amplitude. The search is made tractable by using the tensor-signal analysis to fix sky location and intrinsic source parameters, reducing the template bank to the small set of modified-gravity waveforms that are sensitive to the extra polarizations.","core_discovery":"The central claim is that a practical, model-agnostic test for superluminal non-tensor polarizations is obtained by anchoring to a detected tensor-mode event and searching backward in time. Because gravitational Cherenkov constraints require $v_S,V \\ge c$ and the joint gravitational-wave/gamma-ray observation pins tensor modes to $v_T \\approx c$ within $O(10^{-15})$, any scalar or vector polarization from the same source must arrive with or before the tensor modes. The paper's key relation is $\\Delta t_N = D_L/c - D_L/v_N$, so a non-detection over an observing time $T_{\\rm Obs}$ translates, via $\\delta v_N = c\\Delta t_N/(D_L - c\\Delta t_N)$, into a lower bound on the fractional speed excess for polarizations with detectable amplitude, and an upper bound on amplitude for speeds that would have placed the signal inside the searched window. The authors then show that gaps in detector observing schedules can be filled by stacking events at different times and distances, that sensitivity changes across runs can be handled by sub-threshold searches or by restricting to nearby sources, and that existing upper limits on stochastic backgrounds and scalar-mode amplitudes do not rule out detectable extra polarizations.","pith_inferences":["The same search-back strategy could be extended to next-generation detector networks, where longer observation baselines and access to closer, louder sources would make the speed lower bounds several orders of magnitude tighter than the $O(10^{-9})$ example quoted here.","If a non-tensor burst is ever seen without an associated tensor trigger, the multi-detector timing argument in Appendix B could be applied to place an upper bound on its speed, complementing the lower bounds from the search-back method.","The amplitude-speed degeneracy in the exclusion region could be partially broken by comparing multiple events at different distances, since both tensor and non-tensor amplitudes fall as $1/D_L$ while the speed-dependent arrival lead grows with $D_L$.","Before claiming a detection, the search would need noise-background calibration on long stretches of real data, using time-shifted or injected signals to establish the false-alarm rate over the same observation periods."],"forward_implications":["A non-detection of extra polarizations in one year of data before a source at 50 Mpc would place a lower bound $v > (1 + 6\\times10^{-9})c$, enough to make tensor and non-tensor arrivals from distant sources differ by years.","Existing polarization-content studies that assumed equal propagation speeds for all modes do not rule out faster non-tensor polarizations, so this search covers parameter space those analyses missed.","Stacking events that arrive at different times or from different distances fills gaps left by detector downtime, so the test does not require a continuous observing record.","A direct detection of early non-tensor polarizations would be immediate evidence of physics beyond general relativity.","In Einstein-æther theory, speed constraints translate only weakly to coupling constants because the relevant couplings are already small, so the main payoff of the test is model-agnostic rather than theory-specific."],"supporting_citations":[{"why":"Establishes that small differences in polarization propagation speeds cause non-tensor modes to fall outside standard search windows and supplies the arrival-time relation used here.","marker":"[3]"},{"why":"Gives the gravitational Cherenkov lower bound that forces any extra polarizations to travel at or faster than light.","marker":"[11]"},{"why":"Provides the cosmic-ray based constraint on modified dispersion relations that supports the same lower bound.","marker":"[12]"},{"why":"Pins the tensor polarization speed to the speed of light to within one part in $10^{15}$ through a coincident gravitational-wave and gamma-ray observation, so extra modes must arrive no later than tensor modes.","marker":"[13]"},{"why":"Supplies an Einstein-æther waveform template and current coupling-constant constraints used to illustrate when speed bounds do or do not constrain amplitudes.","marker":"[27]"},{"why":"Provides the catalog of detected events that serve as anchors for the backward-time search.","marker":"[41]"},{"why":"Shows that existing tests of polarization content assumed equal speeds and therefore do not exclude faster non-tensor modes.","marker":"[43]"},{"why":"Places upper limits on scalar, vector, and tensor stochastic backgrounds that are used to argue detectable extra polarizations are not already ruled out.","marker":"[54]"}],"fun_headline_variants":["New test hunts for early-arriving gravitational polarizations","Search early data for faster-than-light gravity waves","A practical test for superluminal polarizations from gravity","Look backward in time to test gravitational wave speed"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The test only works if non-tensor polarizations can have amplitudes large enough for current detectors to see, and the paper does not prove that any available or proposed search actually reaches that sensitivity.","fun_headline_variants_meta":{"raw":{"variants":["New test hunts for early-arriving gravitational polarizations","Search early data for faster-than-light gravity waves","A practical test for superluminal polarizations from gravity","Look backward in time to test gravitational wave speed"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00075,"raw_usage":{"total_tokens":3354,"prompt_tokens":978,"completion_tokens":2376,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":594,"completion_tokens_details":{"reasoning_tokens":2312}},"tokens_in":594,"tokens_out":2376,"duration_ms":18147,"temperature":1.0,"reasoning_tokens":2312,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T00:33:23.764077+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the proposed template bank over real pre-event data for nearby catalog events with simulated scalar and vector modes injected at known speeds and amplitudes; if injections with amplitudes near the detector noise floor are not recovered, the feasibility claim fails. A non-tensor polarization observed arriving after the tensor modes would likewise falsify the underlying premise that all extra polarizations are superluminal.","supporting_citations":[{"cited_title":"Constraint on modified dispersion relations for gravitational waves from gravitational Cherenkov radiation","cited_arxiv_id":"1509.00610","evidence_quote":"Provides the cosmic-ray based constraint on modified dispersion relations that supports the same lower bound."},{"cited_title":"stacking","cited_arxiv_id":null,"evidence_quote":"Provides the catalog of detected events that serve as anchors for the backward-time search."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows that existing tests of polarization content assumed equal speeds and therefore do not exclude faster non-tensor modes."}],"review_version":1}