{"id":"4c217571-5ab3-4127-bee9-a80623df7429","arxiv_id":"2507.09674","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Dynamical traction from rotating stellar backgrounds can push black holes outward rather than letting them sink, and in cold fragmented discs a two-stage instability may even eject them from the center.","lead":"This paper proposes a new mechanism called dynamical traction, in which a black hole surrounded by rotating stars gains angular momentum and can migrate outward instead of sinking to the galactic center. It uses analytic diffusion theory and N-body simulations to argue that this effect delays or prevents black holes from settling at galaxy centers, with implications for early supermassive black hole growth.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The cold-disc instability claim rests on an admitted artifact: freezing the isochrone halo inflates clump binding energy, and the one fully self-gravitating run shows clumps dissolving with the BH staying at the center.","rationale":"The warm-disc calculations (Figs. 5-6) genuinely demonstrate that a rotating, anisotropic stellar background can transfer angular momentum to a BH and oppose dynamical friction; that part of the paper deserves credit. The problematic step is the extrapolation to the cold, fragmenting case. The ejection mechanism is a two-stage process: a clump dislodges the BH, then traction drives it outward. The reference numerical setup deliberately freezes the hot isochrone halo (§4.1), and the authors explicitly state this enhances clump binding energy and that relaxing it will significantly change the outcome (§7.1). The only fully self-gravitating run included in the paper (Bonsai, §7.1) shows clumps dissolving and the BH remaining at the center. Thus the paper's own most realistic calculation contradicts the abstract's cold-disc claim. The test I propose controls for the resolution difference between the Bonsai run (l=32 pc, 8.4e6 particles) and the reference runs (l=125 pc, 1e5 particles), so it would settle whether the failure is physical or numerical. If the controlled live-halo run still shows instability, the claim stands; if not, the abstract and conclusions should be revised to present the two-stage instability as a tentative prediction rather than a demonstrated outcome. This is exactly the reader's weakest assumption, and my independent reading agrees.","tokens_in":39419,"tokens_out":5219,"duration_ms":56451,"concrete_test":"Repeat the cool-disc radial-orbit run of §5.2.1 with the spherical isochrone component live instead of frozen, keeping all other parameters identical to the reference model (N=10^5 disc particles, same BH mass, same softening l=125 pc, same initial conditions), and measure the bound-mass fraction of clump S452 and the BH cylindrical radius at t=1.5 Gyr. If the BH remains within ~100 pc and no sustained Lz growth appears, the two-stage instability is an artifact of the frozen halo and the abstract should be revised. For comparability, also rerun the Bonsai case at the reference softening to separate resolution effects from halo-response effects.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's headline two-stage instability (abstract; §8) requires that Jeans-unstable clumps in the cold disc remain bound long enough to eject the BH. In the reference runs the spherical isochrone component is frozen: 'We freeze the other (dynamically hot) stellar orbits completely' (§4.1). The authors concede that 'the binding energy of the clumps is enhanced artificially, and we expect that relaxing the constraint of a frozen component will change the outcome of the numerical integration significantly' (§7.1). Their one fully self-gravitating test (Bonsai, 8.4e6 particles, l=32 pc) shows exactly this: only ~62% of the clump is bound, the clump dissolves in the BH's tidal field, and 'the BH remains at the heart of the system' (§7.1). Because the ejection mechanism is clump-BH scattering, a dissolved clump cannot provide the impulse needed for outward migration. The warm-disc traction result (Figs. 5-6) is independent of this and appears robust, but the cold-disc instability claim is not supported by the evidence in the paper; it is a prediction contingent on an artificially enhanced clump-binding assumption. The abstract, however, presents it as a main result without this caveat.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the orbital evolution of a massive black hole in a rotating, anisotropic stellar background. Using a Fokker-Planck diffusion calculation with a Dirac-delta azimuthal distribution function, the authors derive a first-order angular-momentum diffusion term (Eq. 10) that changes sign with RΩ−vφ, and they name this systematic transfer of angular momentum to the BH 'dynamical traction'. Restricted N-body simulations of a Miyamoto-Nagai disc in a frozen isochrone halo show that a BH on a radial orbit in a warm disc acquires angular momentum and migrates outward after an initial friction phase (§5.1), while in a cold disc the response is dominated by Jeans-unstable clumps. The paper claims a 'two-stage orbital migration instability' in which bound clumps dislodge the BH and traction then drives outward migration, and it proposes a stability criterion based on the ratio of velocity dispersion to streaming motion. The paper also confronts the mechanism with the Milky Way nuclear cluster and with JWST observations of an off-centred AGN at z≈7.3.","tokens_in":39669,"tokens_out":6799,"duration_ms":80614,"significance":"If the warm-disc traction result is correct, it adds a genuine and potentially important effect: in rotating stellar backgrounds, dynamical friction on a massive perturber is not always inward, and angular-momentum exchange with streaming stars can stall or reverse migration. The paper's numerical demonstration in §5.1 is supported by angular-momentum conservation checks (Appendix C) and by a rough quantitative comparison with the Fokker-Planck rates in §6. The authors are also transparent about the main limitation: §7.1 reports a fully self-gravitating run in which clumps dissolve and 'the BH remains at the heart of the system'. However, the abstract and §8 present the two-stage instability as an established result, which is not supported by the evidence in the manuscript. The paper therefore contains a solid core (dynamical traction in warm discs) and an overreaching claim (cold-disc instability) that needs substantial revision.","major_comments":[{"comment":"The headline 'two-stage orbital migration instability' is not supported by the paper's own fully self-gravitating calculation. In the Bonsai run with a live disc and live isochrone component (8.4×10^6 particles, l=32 pc), only ~62% of the clump mass is bound, the clump dissolves in the BH's tidal field, and 'the BH remains at the heart of the system' over 1.5 Gyr. The authors state that freezing the isochrone halo means 'the binding energy of the clumps is enhanced artificially, and we expect that relaxing the constraint of a frozen component will change the outcome of the numerical integration significantly.' The abstract and §8 nevertheless present the instability as a main result without this caveat. The claim should be reframed as contingent on a frozen, hot halo, and a fully self-gravitating run at the reference resolution should be provided before the instability can be considered established.","section":"§7.1, Fig. 15, Abstract, §8"},{"comment":"The analytic derivation of dynamical traction uses a Dirac-delta distribution function in azimuthal velocity, D(vφ−RΩ), which removes the azimuthal velocity dispersion entirely. The sign-changing first-order term in Eq. (10) and the quantitative stability boundary in Fig. 4 are therefore computed for a DF with zero spread in vφ, not for the warm discs whose dispersion is the point of comparison. Because the warm-disc N-body runs independently support the existence of traction, this is not fatal to the mechanism, but the proposed criterion (a threshold in isotropic velocity dispersion versus streaming motion) is not demonstrated for finite azimuthal dispersion. Please recompute the diffusion coefficients with a finite σφ (e.g., a Gaussian or Schwarzschild DF) and show how the critical curves of Fig. 4 shift.","section":"§3.2, Eq. (7); §3.3, Eq. (10); Fig. 4"},{"comment":"Even within the frozen-halo reference set-up, the cold-disc radial-orbit run does not show a scattering event off a bound stellar clump: clump S452 has positive total mechanical energy (E_k/|W| between about 1.77 and 1.88), and the paper concludes that 'self-gravity plays only a minor role in the interaction of the clump with the BH.' The two-stage scenario instead requires clumps 'with much binding energy' (§7), which occur only when the halo is frozen or partially live. The causal chain from Jeans-unstable clump to BH ejection is therefore not demonstrated in any simulation with a live halo, and the paper should either supply such a simulation or explicitly label the chain as a working hypothesis rather than an outcome.","section":"§5.2.1, clump S452"}],"minor_comments":[{"comment":"The sentence 'we have illustrated the transition from one regime to the other for an BH on a radial orbit in a hot disc (Fig. 6)' should cite Fig. 5; Fig. 6 is the circular-orbit warm-disc case.","section":"§7, first paragraph"},{"comment":"Typos 'the on-set', 'offof', and '1:1 resonant trapping at co-rotation' should be corrected to 'onset', 'off of', and 'corotation'.","section":"§3.3 and elsewhere"},{"comment":"The 300 Myr estimate is obtained by integrating the diffusion coefficients while holding R and z fixed; the authors acknowledge this caveat, but the abstract's '300 Myr or less' should carry this caveat explicitly.","section":"§3.5 and Abstract"},{"comment":"The text gives M•/m⋆=781 for the reference model and then rounds it to 10^3 in the comparison with Fig. 2; the two values should be reconciled in the text rather than in a parenthetical remark.","section":"§4.2 and §6"},{"comment":"The placement of lnΛ in the denominator is visually ambiguous; please add parentheses or a derivation note.","section":"Eq. (2)"}],"recommendation":"major_revision","confidential_remarks":"The warm-disc traction result is publishable and provides a useful counterpoint to standard dynamical-friction arguments, but the abstract and conclusions currently overstate the cold-disc two-stage instability. If the authors can supply a matched-resolution fully self-gravitating cold-disc run and revise the analytic criterion using a finite azimuthal dispersion, the paper could become a strong contribution. I would not recommend rejection because the core mechanism is supported by the warm-disc simulations; the needed work is substantial but well within the scope of a revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Cut to the chase: the paper has a real idea and an overstated headline. The warm-disc version of dynamical traction is worth taking seriously; the cold-disc two-stage instability is not supported by the evidence in the paper, including the authors' own fully self-gravitating run.\n\nWhat is actually new: they name and analyze dynamical traction—angular momentum transfer from a rotating, anisotropic stellar background to a black hole, opposite to Chandrasekhar friction. The Fokker-Planck derivation with f(E,Lz) diffusion coefficients (Eq. 10 and Appendix A) is genuine work, and it connects naturally to earlier torque theory (Tremaine & Weinberg, Lynden-Bell & Kalnajs). More importantly, the warm-disc N-body experiments (Figs. 5–6) show the effect in practice: a BH on a low-angular-momentum radial orbit gains Lz and stops its inward migration, while an initially circular orbit does not systematically migrate. The quantitative comparison in §6 gives rough but real support for the analytical rates.\n\nThe soft spots are clear. The analytic derivation uses a Dirac-delta DF for the azimuthal velocity (Eq. 7), which removes azimuthal velocity dispersion entirely; that likely overstates how coherent the streaming is, so the precise location of the traction transition in Fig. 4 is not robust. The bigger problem is the cold-disc instability. The abstract says a BH is removed from the central region through a two-stage migration instability, but the runs that show this freeze the isochrone halo. The authors are honest about it in §7.1: the clumps’ binding energy is artificially enhanced, and they expect a live halo to change the outcome. Their one fully self-gravitating Bonsai run shows exactly that—clumps only ~62% bound, dissolving in the BH’s tidal field, and the BH stays at the center. The instability claim therefore rests on an admitted modeling artifact. The abstract does not carry this caveat.\n\nI do not think the paper should be rejected. The traction mechanism is a legitimate addition to the dynamical friction literature, and the warm-disc behavior is demonstrated. The two-stage instability remains a plausible conjecture worth testing, but it needs to be retracted from the abstract or backed by a self-gravitating run at comparable resolution.\n\nWho should read it: galaxy-dynamics people working on SMBH migration, core stalling, and off-center AGN. I would cite it for the warm-disc traction effect. It deserves a serious referee—send it to review, but the authors need to rewrite the abstract and either add a fully live run or frame the cold-disc result as clearly preliminary.","headline":"Warm-disc dynamical traction is real and worth taking seriously; the cold-disc two-stage instability is not supported by the authors' own self-gravitating run and should be demoted from the abstract.","tokens_in":40268,"tokens_out":2828,"would_cite":true,"duration_ms":31088,"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":"A black hole moving through a rotating stellar disc can be pulled outward by 'dynamical traction' — a systematic transfer of angular momentum from streaming stars — and in a cold, fragmenting disc this can eject it from the galactic centre.","keywords":["dynamical traction","dynamical friction","black hole migration","angular momentum transfer","Fokker-Planck diffusion","N-body simulation","galactic nuclei","Jeans instability"],"falsifier":"Re-run the cold-disc radial-infall simulation with the halo fully live at the same resolution as the disc (the paper's fully self-gravitating run used 8.4 million particles with softening 32 pc) and check whether the Jeans-unstable clumps survive long enough to dislodge the black hole and whether the black hole migrates outward by ~1 kpc within ~1 Gyr; the paper's own self-gravitating run suggests clumps dissolve and the black hole remains at the centre, which would falsify the two-stage instability as stated.","tokens_in":39186,"feed_emoji":"🕳️","tokens_out":14317,"duration_ms":119395,"temperature":0.7,"pith_summary":"This paper argues that a black hole drifting in a rotating, anisotropic stellar disc does not simply sink to the galactic centre under dynamical friction. Instead, the streaming motion of the stars exerts a systematic torque — 'dynamical traction' — that transfers angular momentum to the black hole, so a radially infalling black hole can be deflected onto an outward-migrating orbit. In a dynamically cold, fragmenting disc, the paper identifies a two-stage instability: a clump of stars pulls the black hole out of the centre, and dynamical traction then boosts its angular momentum so it keeps migrating outward. The authors propose a stability criterion based on the ratio of velocity dispersion to streaming motion, and they show that warm, nearly isotropic stellar environments keep the black hole near the centre, consistent with the Milky Way's nuclear cluster. The paper matters because it suggests that young galaxies with cold rotating discs may naturally host off-centred black holes, matching recent observations of offset AGN.","feed_headline":"Cold stellar discs can eject black holes from galaxy centers","feed_subtitle":"Simulations show a rotating, fragmenting star disc can push a central black hole outward for a billion years.","key_machinery":"The central mechanism is the sign-changing torque derived from the Fokker-Planck diffusion coefficient in Eq. (10): $\\langle \\Delta v_\\phi\\rangle \\propto v_\\phi\\, f_\\perp(0)\\, [\\mp 2\\pi + x e^{x^2}\\mathrm{erfc}(x)]\\, dK/dv$, where $K^2=(R\\Omega - v_\\phi)^2 + v_\\perp^2$. The first-order term $v_\\phi \\langle \\Delta v_\\phi\\rangle$ carries the sign of $R\\Omega - v_\\phi$, so when the black hole lags the rotating stellar flow it gains angular momentum; this is the 'dynamical traction' that opposes dynamical friction. The transition to sustained angular-momentum growth occurs when this term dominates the quadratic velocity-diffusion terms, and the paper charts that transition as a stability line in the $M_\\bullet$–$v_\\phi$ plane for different velocity dispersions $\\sigma_\\star$ (Fig. 4). The N-body models realize the mechanism with a live Miyamoto–Nagai disc embedded in a frozen isochrone (Hénon) spherical halo, the frozen component being what keeps the Jeans-unstable clumps artificially bound.","core_discovery":"The paper's central claim is that gravitational focusing in a rotating stellar background produces a transverse acceleration, not just the parallel drag of Chandrasekhar dynamical friction. For a distribution function of the form $f(E, L_z)$, the Fokker-Planck diffusion coefficient $\\langle \\Delta v_\\phi\\rangle$ is proportional to a term that changes sign with $R\\Omega - v_\\phi$: a black hole lagging the stellar flow is pulled forward, and one leading the flow is dragged back. When the streaming-motion term $v_\\phi \\langle \\Delta v_\\phi\\rangle$ overtakes the quadratic diffusion (heating) terms, the black hole systematically gains angular momentum and its orbit transitions from a low-$L_z$ box orbit to a high-$L_z$ loop orbit — the dynamical traction transition. In a warm disc this delays in-spiral by several hundred Myr and then drives outward migration; in a cold, Jeans-unstable disc the black hole's own perturbation triggers fragmentation, and a two-stage instability can remove it from the centre entirely. The paper states this strongest form as: 'In a dynamically cold environment, a BH is removed from the central region through a two-stage orbital migration instability.'","pith_inferences":["Because the fully self-gravitating run is the only one without the frozen halo, the two-stage instability may be an artifact of artificially enhancing clump binding energy; the paper's own analysis in §7.1 leaves that as the leading interpretation.","If dynamical traction is generic, the same torque should act on any massive perturber in a rotating disc — satellite galaxies, globular clusters, or gas clumps — so the mechanism may govern angular-momentum exchange in a wider class of systems than black holes.","The $v_\\phi/\\sigma_\\star$ threshold could be translated into observable quantities (rotation curve, velocity dispersion, surface brightness) and used to predict which high-redshift discs should host off-centred active nuclei, something the paper does not do."],"forward_implications":["In a warm, non-fragmenting disc, a low-angular-momentum black hole first sinks for ~500 Myr, then gains angular momentum and migrates outward; the migration to the centre is delayed by several hundred million years.","In a cold disc, the outcome hinges on clump binding energy: strongly bound clumps can scatter the black hole and trigger outward migration, while weakly bound clumps dissolve in its tidal field and the black hole stays at the centre.","The proposed stability criterion is a threshold in isotropic velocity dispersion relative to streaming (angular momentum) motion: above it the black hole settles, below it the two-stage instability can remove it.","The Milky Way's nuclear star cluster is close to isotropic ($v_\\phi/\\sigma_\\star \\simeq 0.8$), below the critical line for its parameters, so its central black hole is expected to stay at the barycentre.","An outward-migrating $1.25\\times10^7\\,M_\\odot$ black hole can move by ~1 kpc over roughly 750 Myr to 1 Gyr, in the range of the recently detected off-centre AGN at $z\\simeq7.3$."],"supporting_citations":[{"why":"defines dynamical friction, the classic opposing mechanism that the paper's traction complements.","marker":"Chandrasekhar 1943"},{"why":"provides the Fokker-Planck diffusion coefficients, the Rosenbluth potentials, and the dynamical-friction time-scale used throughout.","marker":"Binney & Tremaine 2008"},{"why":"established that resonant torques change sign with the perturber's pattern speed, which the paper's sign-changing term reproduces.","marker":"Tremaine & Weinberg 1984"},{"why":"supplies the sheared-disc response analysis invoked for the cold-disc wake and gap formation.","marker":"Julian & Toomre 1966"},{"why":"showed co-planar loop orbits couple most strongly to the perturber, motivating the disc-geometry choice.","marker":"Boily et al. 2008"},{"why":"demonstrated torque cancellation and core stalling in cored potentials, the baseline the warm-disc results contrast with.","marker":"Banik & van den Bosch 2022"},{"why":"analyzed energy exchange between a massive perturber and stellar orbits in a cored galaxy, informing the anisotropic-orbit response.","marker":"Kaur & Sridhar 2018"},{"why":"the detection of an off-centred AGN at z≈7.3 that motivates the study and provides the observational comparison.","marker":"Übler et al. 2024"}],"fun_headline_variants":["Dynamical traction ejects black holes from galaxy centers","Cold stellar discs push black holes outward for billions of years","Rotating star clumps can drive black hole orbital migration","Two-stage instability removes black holes from galactic centers"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing modelling choice is to freeze the hot, spherical isochrone halo while evolving the cold disc; the paper itself states that this artificially enhances the binding energy of the stellar clumps and that relaxing the frozen constraint would change the outcome significantly (its one fully self-gravitating run dissolves the clumps and keeps the black hole at the centre).","fun_headline_variants_meta":{"raw":{"variants":["Dynamical traction ejects black holes from galaxy centers","Cold stellar discs push black holes outward for billions of years","Rotating star clumps can drive black hole orbital migration","Two-stage instability removes black holes from galactic centers"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000202,"raw_usage":{"total_tokens":1480,"prompt_tokens":1138,"completion_tokens":342,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":754,"completion_tokens_details":{"reasoning_tokens":277}},"tokens_in":754,"tokens_out":342,"duration_ms":4079,"temperature":1.0,"reasoning_tokens":277,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T17:50:11.329544+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-run the cold-disc radial-infall simulation with the halo fully live at the same resolution as the disc (the paper's fully self-gravitating run used 8.4 million particles with softening 32 pc) and check whether the Jeans-unstable clumps survive long enough to dislodge the black hole and whether the black hole migrates outward by ~1 kpc within ~1 Gyr; the paper's own self-gravitating run suggests clumps dissolve and the black hole remains at the centre, which would falsify the two-stage instability as stated.","supporting_citations":[{"cited_title":"1943, , 97, 255","cited_arxiv_id":null,"evidence_quote":"defines dynamical friction, the classic opposing mechanism that the paper's traction complements."},{"cited_title":"& Tremaine , S","cited_arxiv_id":null,"evidence_quote":"provides the Fokker-Planck diffusion coefficients, the Rosenbluth potentials, and the dynamical-friction time-scale used throughout."},{"cited_title":"& Weinberg , M","cited_arxiv_id":null,"evidence_quote":"established that resonant torques change sign with the perturber's pattern speed, which the paper's sign-changing term reproduces."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the sheared-disc response analysis invoked for the cold-disc wake and gap formation."},{"cited_title":"M., Padmanabhan , T., & Paiement , A","cited_arxiv_id":null,"evidence_quote":"showed co-planar loop orbits couple most strongly to the perturber, motivating the disc-geometry choice."},{"cited_title":"& van den Bosch , F","cited_arxiv_id":null,"evidence_quote":"demonstrated torque cancellation and core stalling in cored potentials, the baseline the warm-disc results contrast with."},{"cited_title":"& Sridhar , S","cited_arxiv_id":null,"evidence_quote":"analyzed energy exchange between a massive perturber and stellar orbits in a cored galaxy, informing the anisotropic-orbit response."}],"review_version":1}