{"id":"70a07690-090e-4230-a480-cb425a560d5b","arxiv_id":"1908.07160","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"NGC1052-DF2 and NGC1052-DF4 velocity dispersions are consistent with MOND, MOG, Weyl gravity and GR without dark matter, but Verlinde's emergent gravity is disfavored at D=20 Mpc.","lead":"This paper checks whether the low velocity dispersion of the dark-matter-free candidate galaxy NGC1052-DF2 can be explained by four modified gravity theories. It finds that GR without dark matter, MOND, MOG and Weyl gravity are broadly consistent with the data, while Verlinde's emergent gravity fits poorly at the larger assumed distance.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The reported reduced chi-square values, the main quantitative evidence for the consistency claim, are not tied to a valid observable: with only 10 individual GC velocities, a proper likelihood is never defined, and directly comparing |v_i| to sigma_los(R) would bias the test.","rationale":"The reader's weakest assumption (baryonic mass model and truncation) is real, but it shifts profiles by an amount that is at least partially explored through two distances, Re variation, and a discussion of rcut. The statistical comparison is more load-bearing: if the chi^2 statistic is undefined or biased, Tables I and II cannot support any ranking, regardless of the mass model. I am not claiming the authors definitely used the biased comparison; the absence of a definition and of the data table makes it impossible to tell. The requested check would settle it. Read in good faith, this is a plausible consistency analysis, and the broad conclusion may well survive scrutiny, but the central evidence needs to be specified and possibly recomputed. This concern overlaps with the reader's call to tabulate the GC data and report uncertainties, so the conditional verdict remains appropriate.","tokens_in":12409,"tokens_out":15288,"duration_ms":173130,"concrete_test":"Recompute the goodness of fit for all six models at D=20 and D=13.2 Mpc using the ten individual GC velocities from Wasserman et al. (2018) with the Gaussian-likelihood statistic chi^2 = sum [v_i^2/(sigma_model^2(R_i)+epsilon_i^2) + ln(sigma_model^2(R_i)+epsilon_i^2)] (up to constants), and compare to Tables I and II. Also compute the alternative statistic sum (|v_i|-sigma_model(R_i))^2/epsilon_i^2; if the reported values match the latter rather than the former, the central consistency rankings are not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The core demonstration rests on Tables I and II, yet Section V never defines how chi^2/dof is computed from the ten GC velocities. A single globular cluster velocity is one realization of the line-of-sight velocity distribution at its projected radius; it is not itself a measurement of sigma_los(R). The model predicts the scale of that distribution, so the correct statistic is a likelihood such as prod_i N(v_i | 0, sigma_model^2(R_i)+epsilon_i^2). Without this definition, the quoted chi^2 values cannot be checked. The figure caption's 'individual GC velocity dispersion measurements' and a y-axis starting at zero suggest absolute radial velocities may have been compared directly to the predicted dispersion curve. Under Gaussianity, E|v| ~ 0.80 sigma_los, so that comparison biases residuals and alters the rankings. Because the abstract's 'fully consistent' claim is carried by exactly those numbers, the central result is unverified until the statistic and the ten input velocities are specified.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper computes line-of-sight velocity dispersion profiles for the ultra-diffuse galaxy NGC1052-DF2 under general relativity without dark matter, MOND (with and without an external field effect), Weyl conformal gravity, MOG, and Verlinde's emergent gravity, using a Sérsic baryonic mass model at two assumed distances (20 Mpc and 13.2 Mpc). It reports reduced chi-square values comparing predicted dispersion profiles with ten globular-cluster velocity data points and concludes that the data are fully consistent with GR and with all modified gravity theories except emergent gravity at 20 Mpc. It also gives rms dispersion predictions for NGC1052-DF4. The quantitative evidence for the central claim is in Section V: Tables I and II contain reduced chi-square values, and Figure 4 compares rms dispersion bounds.","tokens_in":12682,"tokens_out":6757,"duration_ms":67211,"significance":"If the statistical comparison were valid, the paper would be a useful unified reference for how four modified gravity theories plus GR fare against the DF2 data, with the strength of treating both disputed distance estimates and including MOND's external field effect. The authors are transparent about MOG's parameter flexibility and about idealizations such as spherical symmetry, isotropy, and a sharp truncation radius. However, the central claim rests entirely on chi-square values whose definition is never given; since individual globular-cluster velocities are not dispersion measurements, the quoted numbers cannot be verified as stated. The paper also does not propagate uncertainties from the mass model, distance, or truncation radius into the chi-square tables. The comparison is therefore not yet a sound basis for the abstract's 'fully consistent' conclusion.","major_comments":[{"comment":"The reduced chi-square values in Tables I and II are the sole quantitative evidence for the abstract's central claim that the dispersion data of NGC1052-DF2 are 'fully consistent' with the modified gravity paradigm, yet the manuscript never defines how χ²/dof is computed from the ten globular-cluster velocities. A single line-of-sight velocity v_i is one draw from the velocity distribution at projected radius R_i, not a measurement of σ_los(R_i); if Figure 2's 'individual GC velocity dispersion measurements' are the absolute values |v_i| plotted against the predicted σ_los curve, the test is biased because under a Gaussian the expectation of |v| is about 0.80σ. The authors must state the exact statistic, report the ten input velocities, and either use a Gaussian likelihood such as ∏_i N(v_i | 0, σ_los^2(R_i) + ε_i^2) or bin the data into dispersion estimates with proper errors. Without this, the numbers in Tables I and II cannot be checked and the consistency claim is unverified.","section":"Section V, Tables I and II"},{"comment":"The MOG parameters α = 1.30 and μ = 0.443 kpc⁻¹ are taken from Moffat and Toth [44], who fit the velocity dispersion of NGC1052-DF2 itself; the MOG row in Tables I and II is therefore a same-galaxy re-fit rather than an independent prediction. The paper explicitly acknowledges in Section VI that MOG 'enjoys one additional degree of freedom' because its parameters are mass-dependent. To make the comparison meaningful, the authors should either fix MOG parameters from external galaxy samples or apply a model-comparison penalty (e.g., AIC or BIC) for the fitted parameters. As presented, the lower MOG χ² relative to MOND, Weyl, and Emergent gravity does not by itself establish that MOG is more consistent with the data.","section":"Section IIIc and Section VI"},{"comment":"The error bars in Figure 4 are obtained by varying the effective radius by 50%, but Tables I and II are computed with fixed values of Re, Σ0, rcut, and anisotropy, with no propagation of these uncertainties or of the distance uncertainty into the predicted σ_los profiles. Since the distance is disputed (20 Mpc vs 13.2 Mpc) and rcut is a hand-chosen truncation, the claimed consistency should be demonstrated by showing how χ² changes under these variations, not only by the two distance rows in Tables I and II. The statement in Section VI that the simplifications 'would unlikely to alter the final conclusion much' is an assertion that needs quantitative support.","section":"Section V, Eq. (5.2) and Figure 4; Section VI"}],"minor_comments":[{"comment":"The galaxy parameter is spelled 'Sersic' in Section II but should be 'Sérsic' throughout; the phrase 'Modified Newtonian Dynamcies' in the title and Section IIIa should be 'Modified Newtonian Dynamics'.","section":"Throughout, Section IIIa"},{"comment":"The abstract contains a duplicated article: 'coupled to the the baryonic mass' should read 'coupled to the baryonic mass'.","section":"Abstract"},{"comment":"The exterior moment E_{-1}(r) appears in Eq. (3.3) but is not explicitly defined; the given definition of E_n(r) covers n = -1 only implicitly, and this should be stated.","section":"Section IIIb, Eq. (3.3)"},{"comment":"The captions refer to 'individual GC velocity dispersion measurements' for the blue points; if these points are the original individual radial velocities, they should be relabeled as such and the error bars explained, because individual velocities are draws from a distribution, not dispersion measurements.","section":"Figures 2 and 5"},{"comment":"The degrees of freedom for the quoted χ²/dof are never stated; the authors should specify dof for each theory, for example as the number of globular clusters minus the number of free parameters in the model.","section":"Section V, Tables I and II"},{"comment":"The ten globular-cluster velocities and their uncertainties used in Figures 2 and 5 are not tabulated in the manuscript; adding a data table would make the analysis reproducible and allow readers to verify the chi-square construction.","section":"Section V, data availability"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses a topical and contested question, and the theoretical dispersion profiles are straightforward to compute and mostly transparent in their parameter choices. The core problem is the undefined and likely biased chi-square statistic; once the authors provide the exact statistic, the ten input velocities, and a likelihood-based comparison, the revised version could be suitable for publication. I would ask the editor to insist on those elements in the revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe one thing you should know: this paper's headline claim—DF2's dispersion is 'fully consistent' with modified gravity—is carried by reduced chi-squares (Tables I and II) that are never defined. With ten GC velocities, no per-GC dispersion exists. The plots look like they compare absolute values of individual velocities directly to the predicted sigma_LOS(R) curve. Under Gaussianity, E|v| ≈ 0.8 sigma, so that comparison biases the residuals. Until the authors specify the statistic, the central numbers are uncheckable. I read the stress-test note and it holds up.\n\nWhat is actually new: the paper extends earlier MOND/MOG consistency checks (Haghi et al., Famaey et al., Kroupa et al., Moffat & Toth) by computing full radial dispersion profiles, adding Weyl conformal gravity and Verlinde's emergent gravity, and giving a rough DF4 estimate. It also checks both distance assumptions (20 and 13.2 Mpc) and includes MOND's external field effect. The Jeans modeling and theory formulas are standard and the caveats about the mass profile, rcut, and anisotropy are acknowledged. That is valuable as a reference.\n\nSoft spots, in order of size. First, the missing likelihood is a load-bearing flaw. The right object is something like L = Π_i N(v_i | 0, σ_model^2(R_i) + ε_i^2), not a chi-square comparing |v_i| to σ_model(R_i). The abstract's 'fully consistent' is also overstated—emergent gravity gets χ²/dof ≈ 10 at D=20 Mpc. Second, the MOG parameters come from Moffat & Toth's scaling relation for this same galaxy; the authors are transparent about that, and they do show an alternative parameter set (Fig. 6), so the circularity concern is real but modest. Third, the mass model and sharp rcut are hand-chosen and not propagated into uncertainties; this affects all predictions, though probably not the broad ranking.\n\nThis is a modest but useful consistency analysis, not a new discovery. The right audience is people following the 'galaxy without dark matter' debate; it deserves a serious referee, but only after the statistics are fixed. I would accept it for peer review with major revision.","headline":"A useful but flawed consistency check: the chi-squares that carry the main claim are never defined, and the 'fully consistent' abstract oversells emergent gravity's poor fit.","tokens_in":13181,"tokens_out":5983,"would_cite":true,"duration_ms":56726,"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":"NGC1052-DF2, the galaxy claimed to have no dark matter, still fits MOND, MOG, Weyl conformal gravity, and general relativity without dark matter, with only emergent gravity failing at the 20 Mpc distance.","keywords":["NGC1052-DF2","ultra-diffuse galaxy","velocity dispersion","modified gravity","MOND","MOG/STVG","Weyl conformal gravity","emergent gravity"],"falsifier":"A decisive check would be an independent distance measurement for NGC1052-DF2 (for example from the tip of the red-giant branch) plus a stellar mass profile from resolved star counts, along with radial velocities for at least thirty globular clusters. Recomputing the six predicted dispersion profiles from that measured mass profile, the consistency claim is falsified if the observed line-of-sight dispersion is excluded at 2σ by all of the baryons-only predictions in any radial bin.","tokens_in":12204,"feed_emoji":"🌌","tokens_out":14141,"duration_ms":136134,"temperature":0.7,"pith_summary":"The paper asks whether NGC1052-DF2, an ultra-diffuse galaxy whose low globular-cluster velocity dispersion has been interpreted as evidence that it contains essentially no dark matter, actually challenges modified gravity theories. Starting from a Sersic stellar mass profile and solving the spherically symmetric Jeans equation, the authors compute the expected line-of-sight velocity dispersion for general relativity without dark matter, MOND with and without the external-field effect, MOG/STVG, Weyl conformal gravity, and emergent gravity. At the usual assumed distance of 20 Mpc, every model except emergent gravity falls within the observed confidence intervals, with reduced chi-square values from 1.47 to 5.00; emergent gravity gives 9.99. At the disputed nearer distance of 13.2 Mpc, all six models fit, with chi-square values between 1.79 and 4.55. The paper concludes that the galaxy's low dispersion does not falsify modified gravity and can also be explained by baryons alone under general relativity.","feed_headline":"DF2's missing dark matter is no blow to modified gravity","feed_subtitle":"All modified gravity models except emergent gravity at 20 Mpc fit DF2's data; at 13.2 Mpc even emergent gravity fits.","key_machinery":"The central object is the projected line-of-sight velocity dispersion obtained from the isotropic, spherically symmetric Jeans equation: $\\sigma^2(r) = \\frac{1}{\\rho(r)}\\int_r^\\infty \\rho(r') a(r')\\,dr'$ and $\\sigma_{\\mathrm{LOS}}^2(R) = \\frac{\\int_R^\\infty r\\,\\sigma^2(r)\\rho(r)/\\sqrt{r^2-R^2}\\,dr}{\\int_R^\\infty r\\,\\rho(r)/\\sqrt{r^2-R^2}\\,dr}$. Into this equation the authors feed the truncated Sersic stellar density of Eq. (2.1) and each theory's acceleration law: MOND's interpolation function with $a_0=1.21\\times10^{-10}$ m/s$^2$, Weyl conformal gravity's fourth-order acceleration with constants $R_0=24$ kpc, $M_0=5.6\\times10^{10}\\,M_\\odot$, $\\kappa=9.54\\times10^{-54}$ cm$^{-2}$, MOG's vector-field acceleration with $\\alpha=1.30$, $\\mu=0.443$ kpc$^{-1}$, and emergent gravity's extra term $a_v = a_0 M(r)/[d(M(r)r)/dr]$. The Jeans pipeline converts each theory's acceleration into a dispersion curve that can be compared with the ten globular-cluster velocities.","core_discovery":"On the paper's own terms, the finding is that the ten globular-cluster velocity dispersions of NGC1052-DF2 are consistent with the baryonic mass alone, both in general relativity without dark matter and in most modified gravity theories. The quantitative ranking is distance-dependent: for D=20 Mpc the reduced chi-square values are 1.82 for GR, 1.47 for MOG, 3.60 for MOND, 2.01 for MOND with the external-field effect, 5.00 for Weyl conformal gravity, and 9.99 for emergent gravity; for D=13.2 Mpc they improve to 2.04, 1.81, 1.89, 1.79, 3.35, and 4.55, respectively. The authors trace emergent gravity's poor 20-Mpc performance to its additional acceleration term, which is two to three orders of magnitude above the Newtonian value in the inner galaxy. They extend the conclusion to NGC1052-DF4, a second claimed dark-matter-free galaxy, whose predicted rms dispersions are compatible with most of the same theories at the 2σ level.","pith_inferences":["With only ten tracers and asymmetric error bars, the reduced chi-square differences among the better-fitting models (1.47–3.60 at 20 Mpc) may not be statistically distinguishable; a full likelihood treatment could make several models essentially equivalent.","The sharp truncation radius is a simplified proxy for tidal stripping; a theory-specific tidal-radius calculation would alter the outer parts of every predicted profile and could reorder the chi-square ranking.","The same Jeans-profile pipeline could be applied to other ultra-diffuse galaxies discovered in wide surveys; a galaxy whose measured dispersion lies well below the baryon-only prediction would be the real falsification test that DF2 does not provide."],"forward_implications":["A dark-matter-free interpretation of NGC1052-DF2 does not discriminate between general relativity plus baryons and modified gravity: the same stellar mass profile reproduces the data in both paradigms.","If the galaxy is at 13.2 Mpc, the data become compatible with all six models considered, so a decisive distance measurement would settle which model rankings matter.","MOG's good fit is aided by its mass-dependent parameters $\\alpha$ and $\\mu$, which give it a flexibility the other modified theories lack.","At 20 Mpc, emergent gravity is the only theory examined that fails to match the data, because its extra acceleration term dominates the baryonic Newtonian acceleration in the inner galaxy.","The external-field effect improves MOND's fit at 20 Mpc from 3.60 to 2.01, and the true MOND prediction with a realistic external acceleration lies between those values."],"supporting_citations":[{"why":"Supplies the original globular-cluster kinematics, the D=20 Mpc distance, and the initial baryonic mass/luminosity model for NGC1052-DF2.","marker":"[1]"},{"why":"Provides the re-derived 1σ/2σ/3σ confidence intervals for the intrinsic velocity dispersion used to judge the model profiles.","marker":"[8]"},{"why":"Gives the competing D=13.2 Mpc distance and the corresponding lower mass and effective radius used in the second analysis.","marker":"[9]"},{"why":"Is the source of the individual globular-cluster line-of-sight velocity dispersion measurements used for the chi-square fits.","marker":"[10]"},{"why":"Defines MOND and the interpolating-function acceleration law that produces the MOND dispersion profiles.","marker":"[16]"},{"why":"Introduces MOG/STVG and its vector-field acceleration, from which the MOG profile is computed.","marker":"[18]"},{"why":"Formulates Weyl conformal gravity and its field equations, providing the conformal acceleration used in the Weyl profile.","marker":"[20]"},{"why":"Presents emergent gravity and its additional acceleration term, which the paper identifies as the source of emergent gravity's poor fit at 20 Mpc.","marker":"[21]"},{"why":"Supplies the host-galaxy external acceleration estimate used to model the MOND external-field effect.","marker":"[42]"},{"why":"Supplies the MOG parameters α=1.30 and μ=0.443 kpc^{-1} adopted for NGC1052-DF2.","marker":"[44]"}],"fun_headline_variants":["Modified gravity passes DF2 no-dark-matter test","DF2's missing dark matter compatible with modified gravity","Modified gravity survives DF2 dark matter challenge","Emergent gravity stumbles on DF2, modified gravity holds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the visible stellar mass is well described by a truncated Sersic profile with the adopted central surface density, effective radius, and cutoff radius (and either the 20 or 13.2 Mpc distance); if the true stellar profile, tidal truncation, or distance differs, every predicted dispersion curve shifts and the reported chi-square rankings change.","fun_headline_variants_meta":{"raw":{"variants":["Modified gravity passes DF2 no-dark-matter test","DF2's missing dark matter compatible with modified gravity","Modified gravity survives DF2 dark matter challenge","Emergent gravity stumbles on DF2, modified gravity holds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000794,"raw_usage":{"total_tokens":3561,"prompt_tokens":1076,"completion_tokens":2485,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":692,"completion_tokens_details":{"reasoning_tokens":2422}},"tokens_in":692,"tokens_out":2485,"duration_ms":19805,"temperature":1.0,"reasoning_tokens":2422,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:24:16.062830+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive check would be an independent distance measurement for NGC1052-DF2 (for example from the tip of the red-giant branch) plus a stellar mass profile from resolved star counts, along with radial velocities for at least thirty globular clusters. Recomputing the six predicted dispersion profiles from that measured mass profile, the consistency claim is falsified if the observed line-of-sight dispersion is excluded at 2σ by all of the baryons-only predictions in any radial bin.","supporting_citations":[{"cited_title":"Moffat and Toth","cited_arxiv_id":null,"evidence_quote":"Provides the re-derived 1σ/2σ/3σ confidence intervals for the intrinsic velocity dispersion used to judge the model profiles."},{"cited_title":"A distance of 13 Mpc resolves the claimed anomalies of the galaxy lacking dark matter","cited_arxiv_id":"1806.10141","evidence_quote":"Gives the competing D=13.2 Mpc distance and the corresponding lower mass and effective radius used in the second analysis."},{"cited_title":"A Deficit of Dark Matter from Jeans Modeling of the Ultra-diffuse Galaxy NGC 1052-DF2","cited_arxiv_id":"1807.07069","evidence_quote":"Is the source of the individual globular-cluster line-of-sight velocity dispersion measurements used for the chi-square fits."},{"cited_title":"Milgrom, The Astrophysical Journal 270, 365 (1983)","cited_arxiv_id":null,"evidence_quote":"Defines MOND and the interpolating-function acceleration law that produces the MOND dispersion profiles."},{"cited_title":"Tidal stripping as a possible origin of the ultra diffuse galaxy lacking dark matter","cited_arxiv_id":"1804.06421","evidence_quote":"Formulates Weyl conformal gravity and its field equations, providing the conformal acceleration used in the Weyl profile."},{"cited_title":"However, this difference would not change our result","cited_arxiv_id":null,"evidence_quote":"Presents emergent gravity and its additional acceleration term, which the paper identifies as the source of emergent gravity's poor fit at 20 Mpc."},{"cited_title":"Globular clusters as a probe for Weyl Conformal Gravity","cited_arxiv_id":"1811.00065","evidence_quote":"Supplies the host-galaxy external acceleration estimate used to model the MOND external-field effect."},{"cited_title":"However, these studies lack a de- tailed comparison between the observed dispersion velocities for individual GCs and the predicted values, which we will investigate thoroughly","cited_arxiv_id":null,"evidence_quote":"Supplies the MOG parameters α=1.30 and μ=0.443 kpc^{-1} adopted for NGC1052-DF2."}],"review_version":1}