{"id":"dc03eb1b-56bd-4788-8261-4d31dccd3a0a","arxiv_id":"1908.03310","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"COMPASS reports preliminary qT-weighted Drell-Yan transverse spin asymmetries from 2015 plus half of 2018 data, and extracts the first transverse moment of the pion Boer-Mulders TMD PDF.","lead":"This paper presents new measurements of qT-weighted transverse-spin-dependent asymmetries in Drell-Yan scattering from the COMPASS experiment, combining 2015 data with about half of the 2018 data. The results are used to test the predicted sign change of the Sivers function and to extract the first transverse moment of the pion Boer-Mulders function.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The pion Boer–Mulders extraction in §3.2 sets all sea-quark Boer–Mulders and transversity contributions to zero with no quantitative bound, so the 'first extraction' shown in Fig. 4b is only as solid as that untested assumption.","rationale":"The reader's weakest-assumption analysis identifies the same load-bearing point: the novel pion Boer–Mulders extraction in §3.2 depends on setting sea-quark Boer–Mulders and transversity terms to zero, with no quantitative justification. I agree that this is the most serious threat to the central claim. The extracted quantity is obtained by a single algebraic inversion of Eq. (3.2); any unmodeled sea TMD contamination enters directly into the numerator and changes the extracted valence pion Boer–Mulders function, not just its uncertainty. The paper helpfully states the assumption and several other neglected effects, but it does not estimate their magnitude. Because this is a short conference proceedings, a full global fit is not expected, but a sensitivity check using an existing parametrization or a model sea TMD would settle whether the assumption is numerically safe. The rest of the paper—the qT-weighted asymmetry measurement itself and the Sivers sign-change comparison—is fairly presented as preliminary and statistically inconclusive, so it does not support a stronger verdict. A conditional acceptance is therefore appropriate; the concern should be resolved before the extraction is cited as a definitive first determination of the pion Boer–Mulders function. No issue with the experimental method or the internal consistency of the presented asymmetries rises to the level of rejection.","tokens_in":6916,"tokens_out":14290,"duration_ms":153744,"concrete_test":"Recompute Fig. 4b using the Boer (1999) parametrization cited as Ref. [21], extended to include non-zero sea-quark Boer–Mulders functions for the pion and non-zero sea transversity for the proton, or using a global TMD set with comparable sea TMD content. Generate pseudo-data for the COMPASS qT-weighted sin(2φ−φS) asymmetry from the full flavor-sum Eq. (1.4), then apply the paper's Eq. (3.2) extraction procedure with the sea TMD terms forced to zero. Compare the recovered xπ h_1^{⊥(1)\\bar u}_{1,π} with the input truth. If the deviation exceeds the plotted error bars at any xπ bin, the extraction must be presented as explicitly conditional on the zero-sea-TMD assumption and should carry a corresponding systematic uncertainty.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 3.2, immediately after Eq. (3.2), states that 'the Boer–Mulders and Transversity PDFs of sea quarks are assumed to be zero and only u, d and s quark contributions are considered in the denominator.' This assumption is not a harmless convention: the measured qT-weighted asymmetry is a flavor sum, and Eq. (3.2) solves for h_1^{⊥(1)\\bar u}_{1,π} by inverting the ratio [A × (unpolarized denominator)] / (−2 e_u^2 h_1^{u,p}). Any non-zero sea-quark Boer–Mulders function of the pion or sea transversity of the proton contributes additively to that ratio. The denominator keeps full unpolarized u, d, s sea contributions, so the dropped terms are not suppressed by the same valence cut in the denominator; they enter the numerator with their own charge and x-dependent weights. The paper gives no estimate, upper bound, or model-based check of the size of these terms. It also states that the hard-scale dependence of the input transversity TMD PDF is neglected, so the input h_1^{u,p}(x_N) is taken at a different scale than the DY data. If either the sea TMDs or Q2 evolution are sizeable in the COMPASS xπ, xN range, the extracted xπ h_1^{⊥(1)\\bar u}_{1,π} in Fig. 4b is biased, not merely imprecise. Since the central claim of the paper is precisely this first extraction, the missing quantification of this assumption is a load-bearing weakness.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports preliminary COMPASS measurements of qT-weighted transverse-spin-dependent azimuthal asymmetries in π−p↑ Drell-Yan, combining the full 2015 data set with about 50% of the 2018 data in the high-mass range M_μμ ∈ [4.3,8.5] GeV/c². Three weighted asymmetries are extracted as functions of x_N, x_π, x_F, and M using the modified double-ratio method. The sin ϕS asymmetry is compared with a projection from COMPASS SIDIS under the sign-change hypothesis for the Sivers function, and the sin(2ϕ−ϕS) asymmetry is used, together with external unpolarized PDFs and a previous COMPASS transversity extraction, to obtain the first transverse moment of the valence pion Boer-Mulders distribution h_1^{⊥(1)\\bar u}_{1,π}(xπ). The central claim is this first extraction, shown in Fig. 4b.","tokens_in":7252,"tokens_out":4366,"duration_ms":50917,"significance":"If the extraction is robust, the pion Boer-Mulders transverse moment is genuinely new experimental information, because the qT-weighted method avoids assumptions on the kT shape of TMDs. The paper also provides a useful, if not yet conclusive, test of the Sivers sign-change hypothesis; the authors appropriately state that the present significance is insufficient to draw conclusions. Strengths include the use of the modified double-ratio method, the inclusion of false-asymmetry systematic studies, and the combination of two data-taking periods. The reliability of the central claim, however, hinges on the approximations in Eq. (3.2) and on the completeness of the quoted uncertainties, both of which need quantitative support before the 'first extraction' statement can be taken at face value.","major_comments":[{"comment":"The assumption that Boer-Mulders and transversity PDFs of sea quarks are zero is load-bearing and has no quantitative justification. The measured asymmetry is a charge-weighted flavour sum: any nonzero sea-quark Boer-Mulders function of the pion or sea transversity of the proton contributes additively to the numerator of Eq. (3.2) before the ratio is inverted, while the denominator keeps the full unpolarized u, d, s sea. The valence-region argument used for the Sivers projection in Section 3.1 therefore does not apply to the dropped terms in the same way. Please provide a model-based estimate, an upper bound, or a sensitivity test quantifying the bias on each extracted point of xπ h_1^{⊥(1)\\bar u}_{1,π} in Fig. 4b, or explicitly present the result as conditional on the sea-quark-zero approximation.","section":"Section 3.2, Eq. (3.2)"},{"comment":"The text states that the hard-scale dependence of the transversity TMD PDF is neglected, while the input transversity h_1^{u,p} comes from a point-by-point extraction from COMPASS SIDIS data and is applied to DY data at M_μμ ∈ [4.3,8.5] GeV/c². No estimate of the scale mismatch is given, even though transversity evolution is not expected to be negligible in this range. The same paragraph also states that uncertainties on the unpolarized proton and pion PDFs are neglected. Please quantify the Q2-evolution effect and the PDF-uncertainty effect, or show that they are small compared with the quoted statistical and asymmetry systematic errors.","section":"Section 3.2, Fig. 4b"},{"comment":"The figures show statistical and, in Fig. 4b, combined statististical-plus-systematic error bars, but the caption of Fig. 3 states that normalization uncertainties from target polarization (5%) and dilution factor (8%) are not shown. Since these are not automatically expected to cancel in the ratio defining the Boer-Mulders moment, the paper should either demonstrate the cancellation explicitly or include the normalization uncertainty in the error budget of the central extraction. Without this, comparisons with the parametrization of Ref. [21] in Fig. 4b are incomplete.","section":"Figures 3 and 4"}],"minor_comments":[{"comment":"The phrase 'was never extracted before' is a strong novelty claim. Since the comparison in Fig. 4b is with a parametrization rather than with a previous data-based extraction, please define what qualifies as an extraction and cite all prior phenomenological determinations of the pion Boer-Mulders function (e.g., from unpolarized cos 2ϕ Drell-Yan analyses and model fits) so that the novelty statement is verifiable.","section":"Section 3.2, abstract"},{"comment":"The notation for the azimuthal angle is inconsistent: Eq. (1.4) uses sin(2ϕ−ϕS), Eq. (3.2) and Fig. 3 use sin(2φ−ϕS) or sin(2φ−φS), and the text mixes φ and ϕ. Please unify the symbol for the lepton azimuthal angle throughout.","section":"Eqs. (1.4), (3.2), and surrounding text"},{"comment":"The curve from Ref. [21] is shown as a line without uncertainty or a statement of whether it is a model prediction or a fit to other data; please clarify its status in the caption or text.","section":"Fig. 4b"},{"comment":"The displayed title contains 'COMP ASS' due to a line-break artifact; the correct spelling should be restored in the final version.","section":"Title page"}],"recommendation":"major_revision","confidential_remarks":"This is a proceedings-style manuscript whose headline result, the first extraction of the pion Boer-Mulders transverse moment, is presented with strong caveats embedded in Section 3.2. If the paper is intended for archival publication, the central claim needs either a quantitative validation of the sea-quark-zero and scale-mismatch assumptions or a clearly conditional wording. The comparison with the Boer parametrization is not a substitute for that validation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: this is a preliminary COMPASS proceedings note that adds about half of the 2018 Drell-Yan data to the qT-weighted asymmetries already reported for 2015. The genuinely new item is the extraction of the first transverse moment of the pion Boer-Mulders function in Fig. 4b. The Sivers comparison is an update, not a new result.\n\nWhat the paper does well: the analysis follows the modified double-ratio method used for the 2015 data, the presentation is clear, and the authors state their assumptions openly. They explicitly say the Sivers sign-change test is not yet conclusive, and the systematic band is shown. That is honest. The new BM extraction is presented as an independent estimate, and the comparison with Boer's parametrization is a reasonable first look.\n\nThe soft spots are real. The BM extraction assumes zero sea-quark Boer-Mulders and transversity. That is not a harmless convention: the measured asymmetry is a flavor sum, and those dropped terms enter with their own weights. The paper provides no estimate or upper bound, so the result in Fig. 4b is model-dependent rather than a direct measurement. The same section neglects Q2 evolution of the input transversity TMD and neglects PDF uncertainties. Those are acknowledged in the text, but not quantified. The normalization uncertainties from target polarization and dilution factor (5% and 8%) are omitted from the plots; for a weighted asymmetry they may partly cancel, but the reader cannot tell. The Sivers comparison in Fig. 4a is too low-statistics to constrain the sign-change hypothesis; the paper says so.\n\nNone of this makes the paper wrong. It makes the central new claim exploratory. The assumptions could bias the extraction if sea-quark TMDs are non-negligible; the stress-test concern is on target. A revised version that gives a model-based estimate of the dropped terms, or at least a scan over plausible sea-quark sizes, would materially improve it. The 2015+2018 combined asymmetries are useful to the COMPASS TMD community as a status report.\n\nWho this is for: people tracking the TMD program at COMPASS, and those interested in the pion Boer-Mulders function. It is not a definitive result. I would send it to a referee if it were submitted as a regular paper, because the data and extraction deserve scrutiny; the referee should push for a quantitative treatment of the sea-quark assumption and scale dependence. As a proceedings note, it is fine.","headline":"A clean, honest preliminary COMPASS status report; the one new physics claim, the first pion Boer-Mulders extraction, is real but rides on untested zero-sea assumptions.","tokens_in":7779,"tokens_out":2598,"would_cite":false,"duration_ms":27118,"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":"First data extraction of the pion Boer-Mulders function from Drell-Yan spin asymmetries.","keywords":["Drell-Yan","transverse-spin asymmetries","qT-weighted asymmetries","Boer-Mulders function","Sivers function","TMD PDFs","COMPASS","semi-inclusive deep inelastic scattering"],"falsifier":"Include non-zero sea-quark transversity and sea-quark Boer-Mulders terms in the analysis of the same COMPASS data and see whether the extracted pion Boer-Mulders moment moves by more than the quoted uncertainty; alternatively, measure the same $q_T$-weighted asymmetry in $\\pi^+ p$ Drell-Yan, where the sea-quark terms enter with different charge and flavor weights, and check whether the resulting $h^{\\perp(1)}_{1,\\pi}$ is consistent with the $\\pi^-$ result.","tokens_in":6655,"feed_emoji":"📐","tokens_out":12562,"duration_ms":115957,"temperature":0.7,"pith_summary":"This paper reports the first extraction of the first transverse moment of the pion Boer-Mulders function from data, using the $q_T$-weighted transverse-spin asymmetry $A^{\\sin(2\\phi-\\phi_S) q_T/M_\\pi}_T$ measured in COMPASS Drell-Yan scattering of a 190 GeV/$c$ $\\pi^-$ beam on a transversely polarized proton target. In the TMD approach, weighted asymmetries turn convolutions of TMD parton distributions into products of their transverse moments, which is what allows the Boer-Mulders moment to be solved for directly. The paper also compares the $q_T$-weighted Sivers asymmetry in Drell-Yan with a projection from SIDIS data that assumes the predicted sign change of the Sivers function. This matters because the pion Boer-Mulders function is a time-reversal-odd transverse-momentum-dependent distribution that had previously been constrained only by models; a direct data point gives the TMD program a new anchor.","feed_headline":"First direct extraction of the pion Boer-Mulders distribution","feed_subtitle":"qT-weighted spin asymmetries from COMPASS yield the first data-based view of the pion Boer-Mulders function.","key_machinery":"The machinery is the $q_T$-weighted transverse-spin asymmetry itself: a ratio of a $q_T$-weighted integral of a spin-dependent structure function to the unweighted integrated unpolarized structure function. The weight is chosen so the convolution over the two intrinsic quark transverse momenta collapses into a product of transverse moments of TMD PDFs; for the Boer-Mulders channel, Eq. (1.4) expresses the $q_T/M_\\pi$-weighted asymmetry as the product of the pion Boer-Mulders first moment and the proton transversity, divided by the unpolarized flavor-sum denominator. Inserting the proton transversity from a point-by-point extraction, unpolarized PDFs, and a valence pion PDF leaves the Boer-Mulders moment as the only unknown, read off bin by bin in $x_\\pi$. The asymmetries are extracted with the modified double-ratio method, which cancels most acceptance effects.","core_discovery":"Combining the full 2015 Drell-Yan data set with about half of the 2018 data, the collaboration extracts the three $q_T$-weighted transverse-spin asymmetries in the high-mass range $M_{\\mu\\mu}\\in[4.3,8.5]$ GeV/$c^2$. The central new result is the first data extraction of $x_\\pi h^{\\perp(1)\\bar{u}}_{1,\\pi}(x_\\pi)$, the first transverse moment of the valence anti-up quark Boer-Mulders distribution of the pion, obtained by rewriting the weighted asymmetry $A^{\\sin(2\\phi-\\phi_S) q_T/M_\\pi}_T$ as a product of this moment with the proton transversity distribution and dividing out the latter. The extraction sets sea-quark Boer-Mulders and sea-quark transversity terms to zero and keeps only u, d, and s contributions in the denominator. The quoted values in Fig. 4b are compared with an existing parametrization. The $q_T$-weighted Sivers Drell-Yan asymmetry measured on the same samples is consistent, within current statistical precision, with the SIDIS-based projection that assumes the Sivers sign change, though the precision does not yet allow a conclusion on that hypothesis.","pith_inferences":["Because the extraction drops sea-quark transversity and Boer-Mulders terms, a natural follow-up is a global fit that includes those sea-quark TMDs and checks whether the valence Boer-Mulders point moves; that is an extension beyond the paper's assumption, not something the paper performs.","The same $q_T$-weighting procedure could be applied to $\\pi^+$ or kaon beams to separate valence flavors; the pattern of the $x_\\pi$ dependence of the Boer-Mulders moment across flavors would discriminate between model predictions more sharply than a single beam.","Feeding the extracted pion Boer-Mulders moment into a prediction for the $\\cos 2\\phi$ asymmetry in unpolarized $\\pi^- p$ Drell-Yan, following Ref. [7] of the paper, would test the TMD against a target-polarization-free observable."],"forward_implications":["The pion Boer-Mulders moment now has a data point that model calculations and global fits must reproduce, turning a previously model-only quantity into an observable constraint.","The same weighting technique can be applied to the other modulations in Eq. (1.1), giving access to other transverse moments of pion and nucleon TMDs without assuming a functional form for the $k_T$ dependence.","When the remaining half of the 2018 data is added, the statistical uncertainty on both the Sivers comparison and the Boer-Mulders point will shrink; if the Sivers comparison then shows the sign pattern predicted from SIDIS with high significance, that would support the sign-change hypothesis.","The measured $q_T$-weighted Sivers asymmetry provides a scale-matched counterpart to the SIDIS weighted Sivers measurement, making the comparison less sensitive to modeling assumptions than standard asymmetries."],"supporting_citations":[{"why":"Supplies the TMD factorization theorem for SIDIS and Drell-Yan that lets the measured structure functions be read as convolutions of TMD PDFs.","marker":"[1]"},{"why":"Provides the SIDIS $P_T$-weighted Sivers asymmetry whose first transverse moment is projected to Drell-Yan in Eq. (3.1).","marker":"[8]"},{"why":"Defines the standard COMPASS Drell-Yan TSA measurement, the high-mass range, and the sign-change context for the Sivers comparison.","marker":"[10]"},{"why":"Describes the modified double-ratio fit method and the projection procedure used here.","marker":"[12]"},{"why":"Provides the single-polarized Drell-Yan cross-section and structure function definitions on which the weighted asymmetries are based.","marker":"[13]"},{"why":"Supplies the unpolarized proton PDFs used in the denominator and in the Sivers projection.","marker":"[14]"},{"why":"Supplies the fragmentation functions used in the SIDIS weighted-Sivers extraction that feeds the Drell-Yan projection.","marker":"[15]"},{"why":"Gives the sign-change relation between the Sivers function in SIDIS and Drell-Yan that underlies the projection in Eq. (3.1).","marker":"[16]"},{"why":"Supplies the valence pion PDF used in the denominator of the Boer-Mulders extraction.","marker":"[17]"},{"why":"Supplies the point-by-point transversity distribution used to divide out the proton contribution in the Boer-Mulders extraction.","marker":"[20]"}],"fun_headline_variants":["First extraction of pion Boer-Mulders from COMPASS","COMPASS pins down pion Boer-Mulders function","Pion's hidden spin structure revealed by COMPASS","qT-weighted asymmetry yields pion Boer-Mulders"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that sea quarks contribute nothing to the asymmetry used for the Boer-Mulders extraction; if sea-quark transversity or sea-quark Boer-Mulders effects are non-negligible at COMPASS kinematics, the extracted value is biased.","fun_headline_variants_meta":{"raw":{"variants":["First extraction of pion Boer-Mulders from COMPASS","COMPASS pins down pion Boer-Mulders function","Pion's hidden spin structure revealed by COMPASS","qT-weighted asymmetry yields pion Boer-Mulders"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000414,"raw_usage":{"total_tokens":2231,"prompt_tokens":1128,"completion_tokens":1103,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":744,"completion_tokens_details":{"reasoning_tokens":1036}},"tokens_in":744,"tokens_out":1103,"duration_ms":7987,"temperature":1.0,"reasoning_tokens":1036,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:16:53.306544+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Include non-zero sea-quark transversity and sea-quark Boer-Mulders terms in the analysis of the same COMPASS data and see whether the extracted pion Boer-Mulders moment moves by more than the quoted uncertainty; alternatively, measure the same $q_T$-weighted asymmetry in $\\pi^+ p$ Drell-Yan, where the sea-quark terms enter with different charge and flavor weights, and check whether the resulting $h^{\\perp(1)}_{1,\\pi}$ is consistent with the $\\pi^-$ result.","supporting_citations":[{"cited_title":"Collins, F oundations of perturbative QCD, Camb","cited_arxiv_id":null,"evidence_quote":"Supplies the TMD factorization theorem for SIDIS and Drell-Yan that lets the measured structure functions be read as convolutions of TMD PDFs."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the SIDIS $P_T$-weighted Sivers asymmetry whose first transverse moment is projected to Drell-Yan in Eq. (3.1)."},{"cited_title":"Weighted transverse spin asymmetries in 2015 COMPASS Drell-Yan data","cited_arxiv_id":"1812.08505","evidence_quote":"Describes the modified double-ratio fit method and the projection procedure used here."},{"cited_title":"Glück, E","cited_arxiv_id":null,"evidence_quote":"Supplies the valence pion PDF used in the denominator of the Boer-Mulders extraction."},{"cited_title":"Extracting the transversity distributions from single-hadron and dihadron production","cited_arxiv_id":"1412.5946","evidence_quote":"Supplies the point-by-point transversity distribution used to divide out the proton contribution in the Boer-Mulders extraction."}],"review_version":1}