{"id":"e68f66bd-6a62-4ccf-b620-6f04676e4e74","arxiv_id":"1908.08648","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":2.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"In Illustris, the assumed relation between galaxy intrinsic shape and kinematic misalignment fails, so shape-recovery methods relying on it are questionable and 'prolate rotation' is a misnomer.","lead":"Using the Illustris simulation, the authors show that galaxies' intrinsic three-dimensional shapes do not line up with their rotation axes in the simple way that many galaxy surveys assume. This means that popular methods for recovering galaxy shape distributions from telescope data may be biased, and that 'prolate rotation' is a misleading label.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The paper's central implication—that Psi_int has little constraining power and biases shape recovery—is not demonstrated; no quantitative inference test is run, only a scatter plot is shown.","rationale":"The reader's weakest_assumption focuses on Illustris realism. That is a valid external-validity concern, but it is secondary: even if Illustris were a perfect proxy, the paper still would not have shown that the failure of the Weijmans relation degrades shape-recovery accuracy. The missing quantitative link is internal to the argument and more directly load-bearing for the paper's stated conclusion. The concrete test above settles the internal link without needing external data. If the mock recovery shows bias, then the realism question becomes the next bottleneck; if it shows no bias, the paper's main claim is falsified. I therefore partially agree with the reader: the verdict CONDITIONAL remains appropriate, but the condition should be not merely 'simulation realism' but 'demonstrated loss of constraining power in a forward model.'","tokens_in":3977,"tokens_out":9999,"duration_ms":99881,"concrete_test":"Use the 978 Illustris galaxies to run a mock shape-recovery test: project each galaxy at many random viewing angles, measure the projected axis ratio and kinematic misalignment Psi (the observables used by IFS surveys), then feed these measurements through the same Bayesian shape-inference pipeline that observational papers use (e.g., Foster et al. 2017), which assumes the Weijmans et al. (2014) relation. Compare the recovered distribution of intrinsic shapes (p, q, or T) with the true 3D distribution measured in the simulation. If the recovered distribution is significantly biased (e.g., a KS test p < 0.01), the paper's central implication holds; if it is unbiased, then Psi_int retains sufficient constraining power and the claim is over-stated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 3 concludes from the right-hand panel of Figure 1 that 'Psi_int holds little constraining power when inferring the distribution of galaxy intrinsic shapes.' This is an inference, not a measurement. The figure shows conditional distributions: oblate and triaxial galaxies cluster at Psi_int ≈ 0, while prolate and spherical galaxies scatter widely. But a variable can have a non-trivial (even noisy) correlation with shape and still carry information; e.g., Psi_int ≈ 0 may strongly favour oblate/triaxial over prolate/spherical. The paper provides no mutual information, no classification accuracy, no posterior comparison, and no mock shape-recovery experiment. The preceding sentence in the same section—that the failure of the Weijmans et al. (2014) relation 'affects our ability to recover accurate intrinsic shape distributions'—is similarly unsupported by the figure alone. Standard forward-modeling methods (e.g., Foster et al. 2017) marginalize over the full joint distribution; a broken one-to-one relation does not automatically imply a biased recovered shape distribution. The paper's title and abstract make a strong practical claim, but the evidence stops at the scatter plot.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses 978 galaxies from the Illustris cosmological simulation at z=0 to measure intrinsic ellipsoidal axis ratios (p, q) and the intrinsic kinematic misalignment angle Psi_int between the stellar rotation vector and the morphological major axis. The authors compare the joint distribution of triaxiality and Psi_int with the analytical relation suggested by Weijmans et al. (2014). They report that most oblate and triaxial Illustris galaxies have Psi_int near 0 degrees, while prolate and spherical galaxies scatter broadly in Psi_int, with no preference for Psi_int near 90 degrees. From this they conclude that Psi_int holds little constraining power for recovering galaxy intrinsic shape distributions, and that the term 'prolate rotation' is a misnomer.","tokens_in":4186,"tokens_out":3981,"duration_ms":40040,"significance":"The question addressed is timely and important: several large IFS surveys (ATLAS3D, SAMI, MANGA, MASSIVE) have used the assumed relationship between kinematics and intrinsic shape to statistically recover shape distributions, so a demonstration that this relationship fails would have direct practical impact. The paper's strengths are that it uses a large, well-defined simulated sample with a clear stellar-particle selection threshold, an iterative reduced-inertia-tensor shape measurement, and a direct visual comparison to an external analytical relation. However, as written the paper provides only a scatter plot as evidence; there is no quantitative test of constraining power, no uncertainty estimate, and no validation of the simulation against observed misalignment distributions. If the central claim were supported by quantitative inference tests, the result would be significant for the interpretation of IFS shape-recovery studies.","major_comments":[{"comment":"The conclusion that 'Psi_int holds little constraining power when inferring the distribution of galaxy intrinsic shapes' is not established by the scatter plot alone. A variable with a noisy or non-monotonic joint distribution can still be informative; for example, the strong clustering of oblate and triaxial galaxies at Psi_int near 0 degrees is itself a constraint that separates them from prolate and spherical galaxies. The paper should provide a quantitative measure of constraining power, such as mutual information, classification accuracy, or a likelihood-based comparison of P(shape | Psi_int), or a mock shape-recovery experiment that measures how well the true shape distribution is recovered with and without the assumed relation.","section":"§3, Fig. 1 (right panel)"},{"comment":"The statement that the failure of the Weijmans et al. (2014) relation 'affects our ability to recover accurate intrinsic shape distributions' is an inference that goes beyond the figure. Standard forward-modeling methods, such as those used by Foster et al. (2017), marginalize over the full joint distribution of intrinsic shape and misalignment; a broken one-to-one relation does not automatically imply a biased recovered shape distribution. The authors should demonstrate the claimed bias explicitly, for instance by applying a shape-recovery method to simulated galaxies with known intrinsic shapes and comparing the inferred distribution with and without the assumed relation.","section":"§3"},{"comment":"The text states that Psi_int is 'the angle between the short axis of the equivalent ellipsoid and the stellar angular rotation vector', but Eq. (2.4) defines it using the angle between the rotation vector and the major axis e1. This is a direct inconsistency in the definition of the central quantity of the paper. Please reconcile the text and equation, and also explain how the major axis is determined robustly for spherical galaxies, whose major-axis direction is stochastic, since those galaxies contribute to the scatter shown in Figure 1.","section":"§2.2, Eq. (2.4)"},{"comment":"The paper draws observational implications from Illustris without validating that the simulation reproduces the observed distribution of kinematic misalignments. The conclusion that real shape-recovery methods are biased depends on Illustris being a faithful proxy for galaxy angular-momentum content and shapes, especially at the low-mass, prolate end. Please include a comparison of the projected kinematic misalignment distribution of the simulated sample with observed IFS samples (e.g., ATLAS3D, SAMI, or MANGA), or otherwise justify that the simulated dynamics are representative, before generalizing to observational surveys.","section":"§2–§3"}],"minor_comments":[{"comment":"The phrase 'Many recent integral integral field spectroscopy' contains a duplicated word 'integral'; please correct.","section":"Abstract"},{"comment":"The notation 'Ln ⃗ vn' appears to be a typesetting error; it should likely be the luminosity L_n times the velocity vector \\vec{v}_n. Please clarify the vector notation in this equation.","section":"Eq. (2.3)"},{"comment":"The right-hand panel would benefit from error bars or at least a statement of the typical measurement uncertainty on Psi_int and triaxiality, since the visual scatter is compared qualitatively with the Weijmans et al. relation.","section":"Fig. 1"},{"comment":"The reference to Mendez-Abreu (2016) is malformed ('ASSL, 15, ASSL..418'); please provide the full bibliographic entry.","section":"References"},{"comment":"The functional form of the Weijmans et al. (2014) relation that is drawn as a red dashed line is not specified in the text; please state the equation or describe the relation quantitatively so that the comparison in Figure 1 is reproducible.","section":"§3"}],"recommendation":"major_revision","confidential_remarks":"This manuscript is a two-page IAU symposium summary of results published in Bassett & Foster (2019). The claim that Psi_int has little constraining power is presented with only a scatter plot and no quantitative inference test; this is a load-bearing issue for the stated conclusions. If the format of the proceedings does not permit additional analysis, the authors should soften the conclusions to match the evidence presented rather than asserting bias in shape-recovery methods."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know: this is a short IAU proceedings paper, and it says so itself—the results were published in Bassett & Foster 2019. There is no new analysis here. What it does do is restate one key claim: Illustris galaxies do not follow the Weijmans et al. (2014) relation between triaxiality and intrinsic kinematic misalignment, and prolate galaxies do not preferentially spin around their major axis.\n\nThe paper does some things well. It is transparent about provenance. It defines Psi_int relative to the major axis rather than the minor axis, with a good reason (minor axis is ill-defined in near-prolate systems). It also flags that spherical galaxies have a stochastic major axis, so their wide scatter in Psi_int is partly expected. Those are honest, useful choices.\n\nThe soft spot is exactly where the reader's and stress-test notes point. Figure 1 is a scatter plot with no error bars, no significance test, and no quantitative comparison to the Weijmans relation. From that, the paper concludes that 'Psi_int holds little constraining power' and that the broken relation 'affects our ability to recover accurate intrinsic shape distributions.' Those are inferential leaps. A scatter plot can look noisy and still contain information—if oblate/triaxial galaxies cluster at Psi_int near 0, that is exactly the kind of conditional signal a recovery method can exploit, especially since forward-modeling approaches marginalize over the joint distribution. Without a mutual information estimate, a classification test, or a mock recovery experiment, the practical claim is not demonstrated.\n\nThe 'prolate rotation is a misnomer' point is on firmer ground: if prolate galaxies in Illustris have Psi_int all over the map rather than peaking at 90 degrees, then the usual label is misleading, at least in that simulation. But that is a simulation statement. The paper jumps from Illustris to observational practice without validating the simulated kinematic misalignment distribution against observed IFS galaxies, so the external punchline carries an unstated assumption.\n\nFor a conference proceedings, this is fine: it is a readable summary that points to the real paper. The citation practice is not a problem here because the original is openly cited. The problem is only that the strong abstract/title claims are supported by a figure rather than a measurement. If it crossed my desk as a journal submission, I would send it back for the quantitative test, but it would be reasonable to read this alongside Bassett & Foster 2019. I would not cite this proceedings; I would cite the MNRAS paper.","headline":"A clear conference summary of an already-published Illustris result; the plotted trend is plausible, but the paper's strong conclusion about 'little constraining power' is not backed by quantitative analysis.","tokens_in":4688,"tokens_out":2467,"would_cite":false,"duration_ms":25636,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that the assumed galaxy shape–rotation relation fails in Illustris, making 'prolate rotation' a misnomer.","keywords":["galaxies: kinematics and dynamics","galaxies: structure","galaxies: statistics","galaxies: fundamental parameters","intrinsic shapes","kinematic misalignment","prolate rotation","Illustris"],"falsifier":"One could falsify the central claim by showing that real IFS galaxies satisfy the $\\Psi_{\\rm int}$–shape relation: for example, if a large observed sample of galaxies classified as prolate by an independent method shows projected major-axis rotation near $\\Psi_{\\rm int}\\simeq 90^\\circ$ as predicted by the analytic relation, the Illustris-based conclusion would fail. A second check would be measuring $\\Psi_{\\rm int}$ distributions in galaxies from another independent cosmological simulation: if that simulation reproduces the analytic relation, the result is simulation-specific rather than general.","tokens_in":3770,"feed_emoji":"🌌","tokens_out":8302,"duration_ms":74149,"temperature":0.7,"pith_summary":"The paper tests a widely used assumption in galaxy-shape recovery: that a galaxy's intrinsic shape is tightly related to the angle between its rotation axis and its morphological axes, the intrinsic kinematic misalignment $\\Psi_{\\rm int}$. Using 978 galaxies from the Illustris cosmological simulation, the authors measure ellipsoid-equivalent shapes and angular momentum vectors directly in three dimensions. They find that the assumed relationship does not hold: most oblate and triaxial galaxies are aligned ($\\Psi_{\\rm int}\\simeq 0$), whereas prolate and spherical galaxies scatter widely in $\\Psi_{\\rm int}$, with no preference for the $\\Psi_{\\rm int}\\simeq 90^\\circ$ configuration commonly called 'prolate rotation'. If real galaxies behave like Illustris, then published intrinsic-shape distributions inferred from integral-field spectroscopy using this relation are unreliable, and the phrase 'prolate rotation' should be abandoned.","feed_headline":"Simulated galaxies break the assumed shape–rotation link","feed_subtitle":"Oblate and triaxial galaxies align, but prolate ones rotate any which way, so the angle cannot infer shapes.","key_machinery":"The central object is the intrinsic kinematic misalignment angle $\\Psi_{\\rm int}$, defined as the angle between the galaxy's angular momentum vector and the major axis of its ellipsoid-equivalent shape, with $\\Psi_{\\rm int}=90^\\circ$ corresponding to rotation around the major axis. The shape itself is measured with an iterative reduced inertia tensor calculation on stellar particles inside the half-mass radius, giving axis ratios $p=b/a$ and $q=c/a$; galaxies are sorted into spherical, oblate, prolate, and triaxial classes from these ratios, and the triaxiality $T=(1-p^2)/(1-q^2)$ is computed. The argument is a direct comparison: the measured $T$–$\\Psi_{\\rm int}$ scatter in Illustris is checked against the analytic curve that earlier work derived from Stäckel-potential models.","core_discovery":"On the paper's own terms, the discovery is that the analytic relationship between triaxiality $T$ and intrinsic kinematic misalignment $\\Psi_{\\rm int}$ suggested in earlier theoretical work is not reproduced in Illustris. For oblate and triaxial galaxies the angular momentum vector almost always lies close to the morphological minor axis, giving $\\Psi_{\\rm int}\\simeq 0^\\circ$; for prolate and spherical galaxies $\\Psi_{\\rm int}$ spans the full range from $0^\\circ$ to $90^\\circ$. In particular, prolate galaxies do not preferentially rotate around their major axis, so the term 'prolate rotation' misdescribes their kinematics. Consequently, $\\Psi_{\\rm int}$ carries very little information about intrinsic shape, and methods that use this angle to constrain the distribution of galaxy intrinsic shapes are not on solid ground.","pith_inferences":["If the Illustris result is representative, earlier observational shape-recovery results that used the $\\Psi_{\\rm int}$ relation likely underestimate the fraction of misaligned prolate galaxies; re-analyzing existing IFS data with simulation-calibrated priors would test this.","The same test could be run in other independent cosmological simulations to see whether the breakdown of the $\\Psi_{\\rm int}$–shape relation is a generic prediction of galaxy formation physics or specific to Illustris.","A useful observable extension would be to compare the projected distribution of kinematic misalignment angles predicted by simulating Illustris snapshots with observed IFS samples; a mismatch would help refine the feedback or angular momentum implementations."],"forward_implications":["Published intrinsic shape distributions inferred from IFS surveys such as ATLAS3D, SAMI, MANGA, and MASSIVE that rely on the $\\Psi_{\\rm int}$–shape relation may be systematically biased, because a key prior in the inference is invalid.","Kinematic maps alone cannot break the shape degeneracy: $\\Psi_{\\rm int}$ should be treated as a weak or uninformative constraint rather than a sharp predictor.","Prolate galaxies should not be assumed to rotate around their projected major axis; classification schemes that equate 'prolate rotation' with a particular $\\Psi_{\\rm int}$ value need revision.","Forward-modeling simulated galaxies through observational selection effects would be a safer route to infer intrinsic shape distributions than the analytic relation."],"supporting_citations":[{"why":"Supplies the analytic $T$–$\\Psi_{\\rm int}$ relationship that the paper tests and finds not to hold in Illustris.","marker":"Weijmans et al. (2014)"},{"why":"The original study this proceedings contribution summarizes and the source of the Illustris measurements.","marker":"Bassett & Foster (2019)"},{"why":"Introduced the use of kinematic misalignment to constrain intrinsic shapes and defined the triaxiality parameter.","marker":"Franx, Illingworth & de Zeeuw (1991)"},{"why":"Presents the Illustris simulation whose $z=0$ snapshot provides the 978 simulated galaxies.","marker":"Vogelsberger et al. (2014a)"},{"why":"Describes the galaxy formation model and sample selection used to build the Illustris galaxies.","marker":"Genel et al. (2014)"},{"why":"Supplies the axis-ratio thresholds used to sort galaxies into spherical, oblate, prolate, and triaxial classes.","marker":"Li et al. (2018b)"},{"why":"Provides the reduced inertia tensor method for measuring ellipsoid-equivalent shapes.","marker":"Allgood et al. (2006)"},{"why":"The public data release through which the Illustris snapshots are accessed.","marker":"Nelson et al. (2015)"}],"fun_headline_variants":["Prolate rotation is a misnomer, Illustris shows","Simulations quash prolate rotation idea","Kinematics angle can't infer galaxy shape","No sign of 'prolate rotation' in Illustris"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument depends on Illustris being a faithful stand-in for real galaxies: if the simulated population's angular momentum content or shape distribution differs from the observed Universe, the conclusion that observational shape-recovery is biased does not necessarily follow.","fun_headline_variants_meta":{"raw":{"variants":["Prolate rotation is a misnomer, Illustris shows","Simulations quash prolate rotation idea","Kinematics angle can't infer galaxy shape","No sign of 'prolate rotation' in Illustris"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000936,"raw_usage":{"total_tokens":3956,"prompt_tokens":848,"completion_tokens":3108,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":464,"completion_tokens_details":{"reasoning_tokens":3045}},"tokens_in":464,"tokens_out":3108,"duration_ms":23451,"temperature":1.0,"reasoning_tokens":3045,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:32:45.615627+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"One could falsify the central claim by showing that real IFS galaxies satisfy the $\\Psi_{\\rm int}$–shape relation: for example, if a large observed sample of galaxies classified as prolate by an independent method shows projected major-axis rotation near $\\Psi_{\\rm int}\\simeq 90^\\circ$ as predicted by the analytic relation, the Illustris-based conclusion would fail. A second check would be measuring $\\Psi_{\\rm int}$ distributions in galaxies from another independent cosmological simulation: if that simulation reproduces the analytic relation, the result is simulation-specific rather than general.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the analytic $T$–$\\Psi_{\\rm int}$ relationship that the paper tests and finds not to hold in Illustris."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduced the use of kinematic misalignment to constrain intrinsic shapes and defined the triaxiality parameter."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Describes the galaxy formation model and sample selection used to build the Illustris galaxies."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The public data release through which the Illustris snapshots are accessed."}],"review_version":1}