{"id":"55c2191c-6501-4de9-b030-be57e33cc8c1","arxiv_id":"1908.07199","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A sample of 16 black hole X-ray binaries yields a potential kick velocity distribution consistent with a unimodal Gaussian of mean 107±16 km/s, suggesting strong natal kicks are common.","lead":"Black hole X-ray binaries appear to receive strong natal kicks: three new VLBI proper motions plus archival data yield a population kick distribution centered near 107 km/s. The result bears on black hole formation pathways, globular cluster retention, and the spin-orbit misalignment seen in gravitational wave mergers.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The hierarchical fit's input for four systems without radial-velocity measurements is under-specified; the 107 km/s mean may depend on which of five ad hoc γ hypotheses is used.","rationale":"I read the paper as a careful, transparent population-level analysis of a newly defined proxy, PKV, with real new VLBI proper motions and public code. The reader's weakest assumption was the plane-crossing proxy itself; I agree that is a major caveat, and the paper explicitly acknowledges in Sections 6 and 7.2 that PKV is not the true natal kick for at least Cyg X-1 and other short-lived systems. However, I found a more concrete, internally testable gap: the hierarchical population fit in Section 7.1 uses one PKV distribution per system, but for four systems Table 5 provides five alternative PKV distributions with no explicit statement of which one enters the fit or how the five γ hypotheses are weighted. Since those four systems span a wide range of possible PKV medians, the reported unimodal mean of 107±16 km/s and the AICc-based model comparison are not fully reproducible from the text. This supports the reader's CONDITIONAL verdict without changing it: the statistical machinery and measurements are solid, but the headline number depends on an unspecified input-selection step for a quarter of the sample. The proposed test directly checks whether this ambiguity matters by rerunning the hierarchical analysis under all five γ hypotheses and under marginalization. I am not claiming the result is wrong; I am claiming the current manuscript does not yet pin down one of the load-bearing inputs to its central claim.","tokens_in":35705,"tokens_out":3473,"duration_ms":40000,"concrete_test":"Using the public code in https://github.com/pikkyatri/BHnatalkicks, rerun the Section 7.1 hierarchical fit for the four systems without measured systemic radial velocities in two ways: (a) entering each of the five PKV distributions from Table 5 separately; (b) marginalizing over the five γ hypotheses with equal prior weights, or with a continuous uniform prior on γ_offset in [-100, 100] km/s at σ=50 km/s. Compare the posterior mean, standard deviation, and AICc for the unimodal and bimodal models against Table 6. If the unimodal mean shifts by more than ~20 km/s, or if the unimodal/bimodal AICc ranking changes, the headline 107±16 km/s value is an artifact of the unspecified γ selection; if not, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The headline result, a unimodal Gaussian PKV distribution with mean 107±16 km/s, comes from the hierarchical fit in Section 7.1. For four of the sixteen systems (GRS 1716–249, Swift J1753.5–0127, MAXI J1820+070, VLA J2130+12), Section 5.2 defines five ad hoc systemic radial-velocity Gaussian hypotheses (γ̄, γ̄±50, γ̄±100 km/s, each with σ=50 km/s). Table 5 then reports five materially different PKV distributions per system; for example, the PKV median for MAXI J1820+070 ranges from 84 to 153 km/s and for GRS 1716–249 from 67 to 100 km/s depending on the chosen γ. Section 7.1 says the hierarchical model was run on 'all the BHXB systems' but never states which of the five PKV distributions was used for these four systems, nor whether the γ uncertainty was marginalized over in the population fit. If only the central γ hypothesis was entered for each of these four systems, the resulting population mean and AICc in Table 6 are not uniquely determined by the data. This is load-bearing not because the proxy PKV-to-natal-kick step is imperfect (the paper acknowledges that in Section 7.2), but because the input to the central statistical claim is ambiguous for 25% of the sample. A different, equally defensible choice of γ representative could shift the inferred unimodal mean by more than the quoted 16 km/s uncertainty.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports new VLBI proper-motion measurements for three black hole X-ray binaries (GX 339–4, GRS 1716–249, Swift J1753.5–0127), compiles astrometric, distance, and systemic radial-velocity constraints for 16 BHXBs, and integrates their Galactocentric orbits backward in time to derive probability distributions for the peculiar velocity at each Galactic-plane crossing, which the authors call the potential kick velocity (PKV). A Monte Carlo code propagates measurement uncertainties, and a Bayesian hierarchical fit is used to model the PKV distribution of the population. The headline result is a unimodal Gaussian PKV distribution with mean 107±16 km/s, along with the statement that 75% of the systems have median PKVs above 70 km/s, which is interpreted as evidence for strong natal kicks and spin-orbit misalignment. The paper also compares the PKV distribution with pulsar kick velocities, discusses implications for globular-cluster retention and BH-BH merger rates, and finds no significant correlation between BH mass and PKV.","tokens_in":36025,"tokens_out":3754,"duration_ms":39386,"significance":"If the central result holds, this is a valuable observational constraint on the natal kick distribution of stellar-mass black holes, with implications for BH formation pathways, spin-orbit misalignment in X-ray binaries, globular-cluster BH retention, and rates of BH-BH mergers. The paper's strengths are the careful VLBI astrometric reduction, the transparent Monte Carlo propagation of measurement uncertainties, the public code, and the explicit enumeration of caveats in Section 7.2. The hierarchical population analysis is a sensible way to combine heterogeneous posterior distributions, and the comparison with pulsar kick velocities is informative. The main concern is not the quality of the astrometry but whether the population-fit result is uniquely determined by the data, given the treatment of four systems with unmeasured systemic radial velocities.","major_comments":[{"comment":"The hierarchical population fit in Section 7.1 is run on 'all the BHXB systems' (16 systems), but the manuscript never states which of the five systemic radial-velocity hypotheses defined in Section 5.2 was used for the four systems without measured γ (GRS 1716–249, Swift J1753.5–0127, MAXI J1820+070, VLA J2130+12), nor whether the uncertainty in γ was marginalized over. Table 5 shows that the PKV distributions for these systems differ substantially across the five hypotheses; for example, the PKV median for MAXI J1820+070 ranges from 84 to 153 km/s and for GRS 1716–249 from 67 to 100 km/s. Since one-eighth to one-quarter of the sample is affected, the quoted population mean of 107±16 km/s is not uniquely determined by the reported input choices. Please specify exactly which PKV distribution was entered for each of these four systems, and ideally run the population fit with a prior over the five γ hypotheses to test whether the inferred mean and its uncertainty are robust.","section":"Section 7.1 and Table 5"},{"comment":"There is an internal inconsistency in the adopted proper motion for GRS 1716–249. Equation (2) reports µα cosδ = −3.83 ± 1.25 mas/yr, but Table 5 lists µα cosδ = −1.7 ± 1.25 mas/yr for the same source. The difference is more than 1.6σ and directly affects the computed PKV distribution and hence the population fit. This must be corrected or explained, because the table is the input used for the quantitative claims of the paper.","section":"Section 4.2 and Table 5"},{"comment":"The central interpretation throughout the abstract, Section 7.3, Section 7.7, and Section 7.9 is that the PKV distribution is a natal kick distribution, but Section 7.2 correctly notes that PKV is only a proxy and that the interpretation assumes birth in the Galactic plane, no globular-cluster origin, and no secular orbital evolution beyond the adopted Milky-Way potential. These caveats are acknowledged but not propagated into the headline claims; for instance, Cyg X-1 is included despite the statement that it has never crossed the Galactic plane in its lifetime, and a GC-ejection origin for objects such as XTE J1118+480 is mentioned as a plausible alternative. Please either soften the natal-kick language to 'potential kick' throughout the abstract and conclusions, or provide quantitative robustness tests (for example, re-running the hierarchical fit excluding Cyg X-1 and high-latitude systems that may be GC escapees) to show that the inferred mean of ~107 km/s is not driven by systems for which the proxy assumption is least secure.","section":"Sections 1.3, 6, and 7.2"}],"minor_comments":[{"comment":"There is a typo: 'median of 200 km s−1 snd' should read 'median of 200 km s−1 and'.","section":"Section 6.1.1"},{"comment":"The name 'Blauuw' in the first paragraph of Section 1 should be 'Blaauw' for the Blaauw kick.","section":"Section 1"},{"comment":"The source list 'GS 1352–64' appears to be a typo for 'GS 1354–64'.","section":"Section 7.1"},{"comment":"The x-axis label in Figure 8 reads 'Natal kick velocity', but the plotted quantity is the potential kick velocity; relabeling would avoid overclaiming.","section":"Figure 8"},{"comment":"The statement that '75% of our systems have potential kicks >70 km/s' is based on the median of each system's PKV distribution, not on a direct measurement; the wording should make this distinction explicit.","section":"Abstract and Section 8"}],"recommendation":"major_revision","confidential_remarks":"The central scientific question is timely and the astrometric work appears carefully done, but the population-fit ambiguity for the four systems without systemic radial velocities must be resolved before the 107±16 km/s result can be accepted. The proper-motion discrepancy for GRS 1716–249 is a straightforward fix but must not be overlooked. I do not see an irreparable flaw; the requested clarifications are within the scope of a revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper is the largest systematic compilation of BHXB kick constraints to date, and it's mostly good work. But two things need attention before trusting the headline 107 km/s mean: the hierarchical fit never says which of five assumed radial velocities was used for four systems, and Table 5 lists a proper motion for GRS 1716–249 that contradicts Section 4.2.\n\nWhat's genuinely new: first VLBI proper motions for GX 339–4, GRS 1716–249, and Swift J1753.5–0127; a 16-system catalog of potential kick velocity (PKV) distributions; and a hierarchical population fit following Mandel/Hogg. The MC code is on GitHub, the distance estimation uses a defined Milky Way prior, and the discussion of sample biases and the PKV proxy is honest.\n\nThe main result—a unimodal Gaussian with mean 107±16 km/s—is plausible. The 12-system subset with measured radial velocities gives 112±22, so it doesn't hinge on the four guessed-γ systems. But the omission matters: the paper never states which of the five γ hypotheses for GRS 1716–249, Swift J1753.5–0127, MAXI J1820+070, and VLA J2130+12 entered the fit. If only the central values were used, the AICc and mean are conditional in an undisclosed way. The abstract's \"75% of systems with PKV >70 km/s\" also shifts: with the central γ, GRS 1716–249 has median PKV 67 km/s, making the fraction 11/16, not 12/16.\n\nThe Table 5 issue is more concrete: µα cosδ is listed as −1.7±1.25 mas/yr for GRS 1716–249, but Section 4.2 reports the fitted value −3.83±1.25. That's a 2.1 mas/yr discrepancy, well outside errors, and it feeds into the PKV. It needs a correction or an explanation.\n\nThe GRS 1716–249 proper-motion fit also relies on discarding the first LBA epoch, justified partly by a K-band position with ~0.2 arcsec uncertainty—large relative to the astrometric signal. The sparse-uv story is plausible, but the supporting measurement is weak.\n\nThe PKV is explicitly not the true natal kick, which limits the birth-mechanism claims; the paper says this clearly. Small sample and selection effects are acknowledged. None of this is fatal; the soft spots are about transparency and one data entry.\n\nWho it's for: people working on BH formation, BBH merger population synthesis, or GC retention. It deserves a serious referee. I'd recommend acceptance with revisions: fix the GRS 1716–249 table entry, specify the γ handling in the hierarchical fit, and test sensitivity to the γ choice.","headline":"A useful population-level compilation of BHXB kick constraints, but the headline 107 km/s mean rests on an unstated choice for four systems' radial velocities, and one proper-motion entry disagrees with the body text.","tokens_in":36647,"tokens_out":9273,"would_cite":true,"duration_ms":84563,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Using VLBI and Gaia astrometry for 16 black hole X-ray binaries, the paper finds about 75% have potential kick velocities above 70 km/s and fits a unimodal Gaussian kick distribution with mean 107±16 km/s, implying most black holes…","keywords":["black hole X-ray binaries","natal kicks","proper motions","VLBI astrometry","Gaia DR2","potential kick velocity","Galactic orbits","spin-orbit misalignment"],"falsifier":"A calculation that reconstructs the true natal kick for a system such as GX 339-4 from its binary parameters and a supernova model, yielding a kick below 50 km/s while its PKV is about 200 km/s, would falsify the claim that PKV is a reliable proxy for natal kick.","tokens_in":35527,"feed_emoji":"🕳️","tokens_out":5041,"duration_ms":48383,"temperature":0.7,"pith_summary":"The paper tries to turn black hole X-ray binaries (BHXBs) into tracers of how black holes are born. Since a black hole's birth kick pushes the binary out of the Galactic plane, the authors use VLBI and Gaia proper motions, distances, and radial velocities for 16 systems to reconstruct each binary's past orbit and read off its velocity at every Galactic plane crossing, which they call the potential kick velocity (PKV). They find that about 75% of the systems have PKVs above 70 km/s, and that the population is well described by a unimodal Gaussian with mean 107±16 km/s. If correct, this implies strong natal kicks are common, favouring supernova or asymmetric-fallback birth over quiet direct collapse for most black holes, and it would make spin-orbit misalignment common in BHXBs.","feed_headline":"Most black hole binaries were born with kicks above 70 km/s","feed_subtitle":"New astrometry of 16 black hole X-ray binaries points to supernova births and frequent spin-orbit misalignment.","key_machinery":"The central object is the potential kick velocity (PKV): the peculiar velocity a binary has at the moment its Galactocentric orbit crosses the Galactic plane, computed by Monte Carlo realization of about 5000 orbits per system integrated backwards 10 Gyr in the galpy MWPotential2014 Galactic potential. Uncertainties in proper motion, distance (from Gaia DR2 parallaxes with a Milky Way prior or from literature estimates), and systemic radial velocity are propagated by drawing inputs from Gaussian or uniform distributions. The resulting PKV probability distributions for all 16 systems are then combined with a Bayesian hierarchical unimodal/bimodal Gaussian model, with model comparison via the corrected Akaike information criterion.","core_discovery":"The central claim is that black hole X-ray binaries as a population received substantial natal kicks at birth, much larger than the recoil expected from symmetric mass loss in a supernova, which is limited to a few tens of km/s. Using the best available astrometry for 16 systems, the paper estimates the peculiar velocity at Galactic plane crossing (PKV) for each and finds that 75% exceed 70 km/s. A Bayesian hierarchical fit to the population favours a unimodal Gaussian with mean 107±16 km/s and standard deviation about 56±14 km/s, while a bimodal model with peaks near 41 and 136 km/s is possible but not preferred. The paper also reports no significant correlation between black hole mass and PKV, and notes that the BHXB PKV distribution sits a factor of 3–4 below the pulsar kick velocity distribution, suggesting black holes receive weaker kicks than neutron stars.","pith_inferences":["If the same PKV distribution applies to the wider black hole population and not just to binaries that stayed bound, the true natal kick distribution may be even stronger, since binaries disrupted by very large kicks are missing from this sample.","The hint of bimodality in the PKV distribution could be sharpened with future Gaia data releases and additional VLBI astrometry of quiescent BHXBs; if confirmed, it would map directly onto the two birth channels discussed in the paper.","A direct test of the PKV-as-kick assumption would be to reconstruct the actual natal kick for a system like GX 339-4 from its binary parameters and a supernova model; if the reconstructed kick is far below its ~200 km/s PKV, the proxy would be called into question."],"forward_implications":["Most BHXBs would have been born in supernovae or supernova-fallback events rather than by direct collapse, with only a minority (Cyg X-1, 1A 0620-00, V404 Cyg, GRS 1915+105) looking like low-kick systems.","Strong kicks imply spin-orbit misalignment should be common in BHXBs, which is consistent with the prevalence of low-frequency quasi-periodic oscillations, and suggests caution when assuming spin-orbit alignment in black hole spin measurements.","If such kicks are typical, many black holes would be ejected from globular clusters, lowering the retention fraction and changing predictions for BH-BH merger rates in dense stellar environments.","The lack of correlation between black hole mass and PKV supports formation models in which there is no clean mass threshold separating direct collapse from supernova birth.","Gravitational-wave mergers of black hole binaries should frequently show misaligned spins and be displaced from their birth sites if the natal kick distribution extends to these binaries."],"supporting_citations":[{"why":"Provides Gaia DR2 proper motions and parallaxes used as astrometric inputs for 11 of the 16 BHXBs.","marker":"Gaia Collaboration et al. (2018)"},{"why":"Supplies the galpy package and the MWPotential2014 Galactic potential used to integrate the orbits and compute plane-crossing velocities.","marker":"Bovy (2014)"},{"why":"Provides the LMXB spatial density model used as the Milky Way prior for converting parallax to distance.","marker":"Grimm et al. (2002)"},{"why":"Provides the Bayesian likelihood framework for converting parallax measurements into distance distributions.","marker":"Bailer-Jones (2015)"},{"why":"Outlines the Bayesian hierarchical methodology used to fit the population-level PKV distribution.","marker":"Mandel (2010)"},{"why":"Supplies the pulsar kick velocity distribution that the BHXB PKV distribution is compared against.","marker":"Verbunt et al. (2017)"},{"why":"Gives the formula for the maximum recoil velocity from symmetric supernova mass loss, the baseline that high PKVs must exceed.","marker":"Nelemans et al. (1999)"},{"why":"Earlier population synthesis claiming black holes and neutron stars receive equally strong kicks, which the paper's results contrast with.","marker":"Repetto et al. (2012)"}],"fun_headline_variants":["Most black hole X-ray binaries got natal kicks above 70 km/s","Black hole binaries: 75% born with kicks exceeding 70 km/s","New astrometry shows black holes often receive strong natal kicks","Black hole X-ray binaries commonly experience kicks beyond 70 km/s","Natal kicks in black hole binaries: 75% exceed 70 km/s, study says"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire analysis assumes that the velocity a binary has when it crosses the Galactic plane is a faithful stand-in for the kick it received at birth, which requires that the black hole formed near the plane (not in a globular cluster or the bulge) and that 10 Gyr of Galactic orbits have not erased the kick signal.","fun_headline_variants_meta":{"raw":{"variants":["Most black hole X-ray binaries got natal kicks above 70 km/s","Black hole binaries: 75% born with kicks exceeding 70 km/s","New astrometry shows black holes often receive strong natal kicks","Black hole X-ray binaries commonly experience kicks beyond 70 km/s","Natal kicks in black hole binaries: 75% exceed 70 km/s, study says"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000233,"raw_usage":{"total_tokens":1543,"prompt_tokens":1043,"completion_tokens":500,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":659,"completion_tokens_details":{"reasoning_tokens":402}},"tokens_in":659,"tokens_out":500,"duration_ms":5182,"temperature":1.0,"reasoning_tokens":402,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:22:57.666569+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A calculation that reconstructs the true natal kick for a system such as GX 339-4 from its binary parameters and a supernova model, yielding a kick below 50 km/s while its PKV is about 200 km/s, would falsify the claim that PKV is a reliable proxy for natal kick.","supporting_citations":[],"review_version":1}