{"id":"e70701b6-4563-4e67-9d22-17f0a5f80e3a","arxiv_id":"2506.14896","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Quasar host galaxies at z~6 are systematically more compact than typical star-forming galaxies, and the most compact hosts contain the most overmassive black holes and the least cold gas.","lead":"Using ALMA observations of 22 quasars at redshift about 6, the authors measure the sizes of their host galaxies and find they are systematically more compact than ordinary star-forming galaxies at the same epoch. The most compact hosts also tend to have the least cold gas and the most overmassive black holes, suggesting a link between black hole growth and compact galaxy structure.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Validate [CII] sizes against JWST stellar sizes; otherwise the compactness claim may be an AGN tracer artifact.","rationale":"The reader's weakest_assumption correctly identifies the [CII]-traces-stars assumption as the most load-bearing step. The paper's observational result is entirely based on [CII] sizes, yet no direct validation of the tracer against stellar light is presented. The reader's CONDITIONAL verdict already accounts for this, and our analysis does not find a reason to shift the verdict: the paper does include several robustness tests (Methods 7.1-7.4) that support the qualitative conclusions under alternative fitting profiles, S/N thresholds, and Mdyn assumptions, and the comparison to ALPINE SFGs uses the same tracer and similar fitting methods, partly controlling for systematic effects. However, the lack of an external size check means the central compactness claim remains provisional. The proposed JWST comparison is directly actionable and would settle whether the compactness is physical or an AGN artifact. We therefore agree with the reader's weakest_assumption and recommend keeping the verdict CONDITIONAL (UNCHANGED). We do not see an internal inconsistency or a clear reason for rejection; the concern is a missing validation rather than a demonstrated error.","tokens_in":20861,"tokens_out":5329,"duration_ms":58663,"concrete_test":"For the subset of z~6 quasars in this sample that also have JWST/NIRCam host detections (e.g., SHELLQs-Ding et al. 2025, arXiv:2505.03876), measure the stellar half-light radius R_e,star in the rest-frame optical and compare with R_e,[CII] from this paper. If the median ratio R_e,[CII]/R_e,star is <0.8 (systematic concentration of [CII] relative to stars), the compactness claim is an artifact of the tracer and the overmassive ratios are overestimated; if the ratio is consistent with 1, the concern is resolved. A fallback test is to inject mock galaxies with R_e~2.5 kpc into the actual ALMA visibilities and run the Spergel fitting pipeline; if the recovered R_e is biased low by >0.1 dex at S/N~20, the size selection is biased.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that z~6 quasar hosts are compact rests on R_e measured from [CII] 158um emission (Methods 6.2). The paper states on p.2 that the FIR-to-mm emission is 'powered mainly by galaxy-wide star formation except for some rare, and extreme types of active galactic nuclei,' but that concerns the continuum, not the [CII] line. AGN can alter the [CII] spatial distribution via X-ray-dominated regions, outflows, or dust-bounded photodissociation regions, so R_e,[CII] need not equal the stellar half-light radius. This matters because (1) 'compact galaxies' is a statement about stellar hosts, (2) M* is derived from Mdyn = 1.16e5*(0.75 FWHM/sin i)^2 * D (Eq. 1, Methods 6.3), so a biased R_e propagates into M* and hence the overmassive ratios, and (3) the evolutionary link to z~4-5 quiescent galaxies compares against their stellar sizes. Methods 7 robustness tests address fitting profile, Mdyn, and AGN contributions to SFR, but none tests whether [CII] traces the stellar distribution: Methods 7.4 checks only the L_bol distributions between compact and extended subsamples, not the size tracer itself. Without an external cross-check, the observed compactness and the lower f_cool trend in compact hosts could both be AGN excitation artifacts.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper compiles archival ALMA [CII] 158 um observations of z~5.5-8 quasars, uniformly reduces them in CASA, and measures host sizes by fitting Spergel profiles to visibilities in GILDAS. From 22 sources with reliable size measurements, it reports that quasar hosts are systematically more compact (median R_e ~1.6 kpc) than z~4-6 [CII]-selected star-forming galaxies, that the most compact hosts tend to host overmassive BHs and have low cold-gas content, and that these systems are plausible progenitors of the first massive quiescent galaxies at z~4-5. Stellar masses are derived as M_dyn - M_gas assuming a thin rotating disk with negligible dark matter, with M_gas obtained from a [CII]-luminosity calibration.","tokens_in":21315,"tokens_out":12885,"duration_ms":130944,"significance":"If the compactness result survives a direct validation of the [CII]-vs-stellar size mapping, the paper would be an important contribution: it connects the overmassive BH population at z~6 to a specific host-galaxy property, offers an evolutionary path to z~4-5 quiescent galaxies and local compact overmassive BH hosts, and provides concrete targets for formation models. The strengths are the uniform ALMA reduction, uv-plane Spergel/visibility fitting, careful rejection of mergers and substructures, and several robustness tests (relaxed S/N threshold, exponential profile, empirical M_dyn). The main caveat is that the paper does not validate [CII] sizes against stellar sizes, and some mass-surface-density comparisons mix dynamical and stellar definitions; these issues are load-bearing for the title-level claims.","major_comments":[{"comment":"The central claim that quasar 'host galaxies' are compact rests on R_e measured from [CII] emission (Methods 6.2), but the paper does not establish that [CII] traces the stellar distribution. The statement in the Introduction that the FIR-mm continuum is powered mainly by star formation concerns the continuum, not the [CII] line; AGN can alter the [CII] spatial distribution through X-ray-dominated regions, outflows, or dust-bounded photodissociation regions. Since M* is derived from Eq. (1) using R_e,[CII], a biased [CII] size propagates into the compactness offset and the overmassive ratios, and the low f_cool trend in compact hosts could likewise be a [CII]-excitation artifact. Methods 7.4 tests only the L_bol distributions between subsamples. Please add a cross-check against JWST stellar sizes for overlapping sources (e.g., Ding et al. 2023, 2025) or a quantitative estimate of the expected [CII]-to-stellar size offset, and adjust the title/abstract wording if such a test is not possible.","section":"Abstract / Methods 6.2, 6.3, 7.4"},{"comment":"The comparison of compactness between z~6 quasars and local galaxies / z~4-5 quiescent galaxies uses different mass definitions: for quasars, Sigma_eff = M_dyn/(pi R_e^2), whereas the local sample uses stellar masses from M*/L_K and the quiescent galaxies use stellar masses and stellar sizes. Because M_dyn includes gas and possibly dark matter, the quasar points are not directly comparable to the stellar-only points. Methods 8 acknowledges this ('despite not tracing stars currently'), but the figures and abstract present the comparison as direct. Please compute stellar mass surface densities for the quasar sample using M* = M_dyn - M_gas, or rescale the comparison and quantify the systematic offset.","section":"Methods 2.3 / Figures 2 (right), 3 (right)"},{"comment":"The main claim that quasar hosts are systematically more compact than SFGs is supported by median R_e values (1.58 vs 2.26 kpc) and by the statement that nearly all quasars fall below the 1-sigma region of the M*-R_e relation, but the paper provides no statistical test for the difference (e.g., a bootstrap or KS test on residuals from the M*-R_e relation, or a test that accounts for the upper-limit M* values). With only 22 vs 21 sources, a quantitative significance statement is needed to support the central claim.","section":"Results / Figure 2 (left)"},{"comment":"The stellar masses depend on the [CII]-to-M_gas calibration of Salvestrini et al. (2025) and on the assumption of negligible dark matter in Eq. (1). Methods 7.1-7.3 vary the sample selection, the fitting profile, and the V_circ normalization, but they do not vary the gas-mass calibration or the dark-matter fraction. The authors should quantify how much the M_BH-M* and M*-R_e conclusions shift for plausible changes in these two inputs, especially for the sources with M* set to upper limits in Methods 6.5.","section":"Methods 6.4, 6.5, 7.1-7.3"}],"minor_comments":[{"comment":"Some entries in Extended Data Table 1 are unphysical as printed, e.g., J0055+0146 has log M_dyn = 19.46 +/- 86.25 and J239-07 has 16.01 +/- 13.34; please correct or explain these values.","section":"Extended Data Table 1"},{"comment":"The text says the compactness conclusion 'holds independently of the methods of size fitting and M* estimation,' but the tests change the profile and M_dyn normalization, not the M_gas calibration or the [CII]-to-stellar mapping; please revise the wording.","section":"Main text / Methods 7"},{"comment":"The aperture radius for spectrum extraction is described as 'manually chosen'; please specify the selection criteria and test the sensitivity of the measured line FWHM and flux to this choice.","section":"Methods 3.3"},{"comment":"The reference list contains duplicate numbers in the sentence 'Ref 2,4,38-48,48,49,49-51,51,52'; please clean up the citation list.","section":"References"},{"comment":"The figure captions should state how many sources have M* upper limits and how these are treated in the median values and in the plotted relations.","section":"Figures 1-3 captions"},{"comment":"The paper would benefit from a brief discussion of the known offset between [CII] sizes and rest-frame UV/optical stellar sizes in high-z SFGs, since that offset is directly relevant to interpreting the median R_e values as host-galaxy sizes.","section":"Discussion / Introduction"}],"recommendation":"major_revision","confidential_remarks":"My recommendation is driven by the need for an external size-tracer validation, not by any apparent inconsistency in the data reduction; the authors are transparent about most assumptions. If a JWST cross-check is not feasible, a reframing of the claims from 'compact host galaxies' to 'compact [CII] reservoirs' would be acceptable, though it would reduce the impact of the paper. The mixed use of dynamical versus stellar mass surface densities in the comparison figures should also be addressed before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe headline: this is the first systematic [CII]-based size sample for z~6 quasars—22 objects, uniformly reduced, Spergel-fitted in the uv-plane—and it makes a credible statistical case that these hosts are more compact than similarly measured SFGs at the same redshift. If that holds, it links overmassive black holes to compact, gas-poor hosts that resemble the first quiescent galaxies. It deserves a serious referee.\n\nThe paper does several things well. The sample construction is careful: companions and substructures are excluded, and a strict S/N threshold is tested by relaxing it. The robustness section is honest—fixing an exponential profile and adopting a lower empirical Mdyn both preserve the compactness result. The authors also flag the main assumptions in Methods 7 rather than burying them.\n\nMy main reservation is the tracer itself. The quasar-vs-SFG size comparison is apples to apples because both come from [CII]. But the evolutionary claim—that these are progenitors of z~4–5 quiescent galaxies—compares against stellar sizes of quiescent galaxies. If AGN radiation or outflows concentrate the [CII] emission, the compactness is exaggerated. Methods 7.4 checks only that the AGN contribution to L_bol is similar between compact and extended subgroups; it does not test the spatial distribution of the line. I would like the authors to confront this with JWST stellar sizes for at least a few objects, or provide a quantitative argument for why R_e,[CII] equals the stellar half-light radius.\n\nThe stellar mass scale is the second soft spot. M* = Mdyn - Mgas assumes a thin disk, no dark matter, and a gas calibration from z>7 quasars. The paper tests a lower Mdyn and the result holds, which is reassuring. But the absolute M* values and the overmassive ratios are model-dependent, and several sources are upper limits. The comparison SFG stellar masses come from SED fitting, so a systematic method offset could contribute to the M*-Re shift.\n\nThe sample is size-selected (22 of 97 starting quasars), but the relaxed S/N test helps. That is a minor worry given the state of the field.\n\nOverall, the statistical compactness claim is likely real, with the [CII]-tracer question as the main risk. The paper is a solid contribution and I would want it published with that question addressed more directly. Send it to peer review.\n\nBest.","headline":"First systematic [CII] size sample of z~6 quasars shows compact hosts, but the claim hinges on [CII] tracing the stellar distribution—worth peer review with that caveat on the table.","tokens_in":21754,"tokens_out":4284,"would_cite":true,"duration_ms":45201,"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":"Host galaxies of $z\\sim6$ quasars with the most overmassive black holes are compact and gas-poor.","keywords":["high-redshift quasars","[CII] 158 micron emission","host galaxy sizes","overmassive black holes","compact galaxies","quiescent galaxies","ALMA","galaxy evolution"],"falsifier":"Resolve the rest-frame optical stellar light of several $z\\sim6$ quasar hosts at the same angular scale as the ALMA [CII] maps: if the stellar half-light radii are found to be comparable to those of normal star-forming galaxies while the [CII] sizes remain compact, the claimed compactness of the stellar hosts is falsified.","tokens_in":20678,"feed_emoji":"🕳️","tokens_out":12612,"duration_ms":109741,"temperature":0.7,"pith_summary":"The paper assembles 22 quasars at $z\\sim6$ with ALMA measurements of the [CII] 158-micron line and derives reliable host-galaxy sizes. It finds that these hosts have a median half-light radius of about 1.6 kpc, making them systematically more compact than typical star-forming galaxies of similar mass and redshift. The most compact hosts also contain the most overmassive black holes and the least cold gas, and their sizes already match the first massive quiescent galaxies seen at $z\\sim4$–5. The authors conclude that these compact systems are late-stage galaxies whose black holes grew early and whose host growth has nearly finished, and that they are the likely progenitors of the first quiescent galaxy population.","feed_headline":"Early universe black holes live in compact galaxies","feed_subtitle":"A 22-quasar sample shows these compact hosts likely become the first quiescent galaxies.","key_machinery":"The load-bearing object is the [CII] 158-micron half-light radius $R_{\\rm e,[CII]}$ of quasar hosts, obtained by fitting a flexible single-component light profile to ALMA visibilities in the uv-plane and converting it to an effective (half-light) radius. That size is combined with the [CII] line width under a rotating thin-disk model to give a dynamical mass, from which a [CII]-luminosity-based gas mass is subtracted to estimate stellar mass. These measurements are placed in two diagnostic planes: the stellar-mass–size plane, where quasar hosts sit below normal star-forming galaxies, and the black-hole-to-stellar-mass versus effective surface-density plane, where they overlap local galaxies with the most overmassive black holes.","core_discovery":"The central discovery is an observed connection between overmassive black holes and compact host galaxies at $z\\sim6$. Using uniformly reduced ALMA [CII] 158$\\mu$m observations and fitting a flexible single-component surface-brightness profile directly in the visibility plane, the authors measure half-light radii for 22 quasars and find $R_{\\rm e}\\sim1.6$ kpc, below the mass–size relation of typical star-forming galaxies at the same redshifts. Stellar masses are obtained by subtracting [CII]-based gas masses from dynamical masses, and the resulting black-hole-to-stellar-mass ratios are high, with the most compact, gas-poor hosts carrying the most massive black holes. The paper interprets these systems as being at the end of their rapid growth, with compactness comparable to $z\\sim4$–5 quiescent galaxies, making them plausible direct progenitors of those galaxies.","pith_inferences":["If the compactness reflects a deep gravitational potential funneling gas inward, the JWST-discovered Little Red Dots—compact AGN at $z\\sim4$–8—may be lower-mass versions of the same phenomenon; a direct test is measuring their host sizes and gas fractions at comparable resolution.","A direct observational test would be to resolve rest-frame optical stellar light in several $z\\sim6$ quasar hosts: if the stellar distribution is as compact as the [CII] gas, the structural claim is confirmed, but if it is extended, the compactness is not a property of the stars.","The paper's evolutionary scheme predicts that the most compact, gas-poor quasar hosts should have older stellar populations and higher stellar surface densities than the extended, gas-rich ones; spectroscopy of a larger sample could check this ordering.","Whether these systems become today's compact overmassive relics or normal ellipticals depends on the merger history; counting ultra-compact massive galaxies at intermediate redshifts would constrain which path dominates."],"forward_implications":["The host galaxies of $z\\sim6$ quasars are systematically more compact than normal star-forming galaxies of the same mass and redshift, with a median half-light radius of about 1.6 kpc.","Among quasar hosts, the most compact galaxies have the most overmassive black holes and the lowest cold-gas fractions.","If current black-hole and star-formation growth rates continue for one gas-depletion timescale, these overmassive systems will move toward the local black-hole-to-stellar-mass relation, not further away from it.","The compact, gas-poor quasar hosts match the sizes of the first massive quiescent galaxies at $z\\sim4$–5 and are strong candidates for their progenitors.","Compact quasar hosts at $z\\sim6$ occupy the same region of the overmassive-black-hole compactness plane as local compact galaxies hosting overmassive black holes, suggesting a delayed evolutionary link."],"supporting_citations":[{"why":"Supplies the [CII]-based half-light sizes of normal star-forming galaxies at redshifts 4 to 6 that form the comparison sample.","marker":"14"},{"why":"Provides stellar masses and ancillary measurements for the star-forming-galaxy comparison sample.","marker":"15"},{"why":"Extends the [CII] size comparison to an additional star-forming galaxy survey.","marker":"16"},{"why":"Establishes and validates the uv-plane profile-fitting method used to measure quasar host sizes.","marker":"12"},{"why":"Supplies the [CII]-luminosity-to-gas-mass calibration used to derive gas masses and hence stellar masses.","marker":"13"},{"why":"Provides two post-starburst quasar hosts whose stellar properties support the evolutionary link to quiescent galaxies.","marker":"23"},{"why":"Supports the anti-correlation between black hole mass and cool gas content that underlies the quenching interpretation.","marker":"22"},{"why":"Gives the local galaxy sample with direct black hole masses used for the compactness versus overmassive-black-hole comparison.","marker":"31"}],"fun_headline_variants":["Overmassive black holes found in compact early galaxies","Compact early galaxies host overmassive black holes","Early universe's overmassive black holes live in compact hosts","Quasar hosts are compact, hinting at first quiescent galaxies"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central claim assumes that [CII] 158-micron emission traces the host galaxy's interstellar medium in quasars in the same way it does in normal star-forming galaxies, with negligible AGN heating, so that the compact size measured in [CII] is a property of the stellar host rather than an artifact of the active nucleus.","fun_headline_variants_meta":{"raw":{"variants":["Overmassive black holes found in compact early galaxies","Compact early galaxies host overmassive black holes","Early universe's overmassive black holes live in compact hosts","Quasar hosts are compact, hinting at first quiescent galaxies"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000148,"raw_usage":{"total_tokens":1190,"prompt_tokens":944,"completion_tokens":246,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":560,"completion_tokens_details":{"reasoning_tokens":187}},"tokens_in":560,"tokens_out":246,"duration_ms":2645,"temperature":1.0,"reasoning_tokens":187,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T19:46:58.662140+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Resolve the rest-frame optical stellar light of several $z\\sim6$ quasar hosts at the same angular scale as the ALMA [CII] maps: if the stellar half-light radii are found to be comparable to those of normal star-forming galaxies while the [CII] sizes remain compact, the claimed compactness of the stellar hosts is falsified.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the [CII]-based half-light sizes of normal star-forming galaxies at redshifts 4 to 6 that form the comparison sample."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides stellar masses and ancillary measurements for the star-forming-galaxy comparison sample."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Extends the [CII] size comparison to an additional star-forming galaxy survey."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes and validates the uv-plane profile-fitting method used to measure quasar host sizes."},{"cited_title":"P.et al.Kiloparsec-scale ALMA Imaging of [C II] and Dust Continuum Emission of 27 Quasar Host Galaxies at z∼6.The Astrophysical Journal904, 130 (2020)","cited_arxiv_id":null,"evidence_quote":"Supplies the [CII]-luminosity-to-gas-mass calibration used to derive gas masses and hence stellar masses."},{"cited_title":"L.et al.The ALPINE-ALMA [C II] Survey: Multiwavelength Ancillary Data and Basic Physical Measurements.The Astrophysical Journal Supplement Series247, 61 (2020)","cited_arxiv_id":null,"evidence_quote":"Provides two post-starburst quasar hosts whose stellar properties support the evolutionary link to quiescent galaxies."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supports the anti-correlation between black hole mass and cool gas content that underlies the quenching interpretation."}],"review_version":2}