{"id":"dadde9b7-bced-4797-9865-1ff9cd3d1d29","arxiv_id":"2506.08046","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For meromorphic KdV potentials, the Schrödinger equation in the Lax pair is solvable by quadrature if and only if the potential is reflectionless, but the paper proves this only under extra hypotheses on the scattering coefficient and on analyticity at infinity.","lead":"The author proves that the Schrödinger equation in the Lax pair for KdV is solvable by quadrature exactly when the meromorphic potential is reflectionless, under particular decay and analyticity conditions. The result extends the author's earlier work from exponentially decaying potentials to meromorphic ones, and it includes a negative result for rational potentials with a simple pole at infinity.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proposition 2.1(iv) does not establish that a(k) has finitely many zeros in C+ when b=0; zeros may accumulate at k=0, and Theorem 1.3's finite-sum quadrature proof depends on that finiteness.","rationale":"The reader's verdict is CONDITIONAL, and I agree that the abstract overstates the proved theorems: Remark 1.5(i) restricts Theorem 1.3 to the case where a(k) has a zero in C+, and Theorem 1.4 additionally assumes analyticity at infinity. My stress-test focus is different. The technically load-bearing step is the finiteness of zeros in Proposition 2.1(iv), used in Theorem 1.3 to write the solution as a finite rational function of x and e^{ik_j x}. The proof of that proposition gives no reason for a(k) to be nonzero near k=0; accumulation of zeros at 0 is compatible with analyticity in C+, continuity on R*, and b=0. L^1 reflectionless potentials with infinitely many bound states exist, and condition (A1) was not shown to exclude them. If such a potential is meromorphic in a neighborhood of R, the proof of Theorem 1.3 fails for it even if the statement is true. This is a genuine gap rather than a stylistic mismatch, so the paper needs a revision of Proposition 2.1(iv) or an added hypothesis such as finiteness of the discrete spectrum or a first-moment condition. Because I do not know of a counterexample to the theorem's statement, my recommended verdict remains conditional; the reader's specific weakest assumption about analytic continuation along Γ is related but not the same, hence agreement is 'disagree'.","tokens_in":16246,"tokens_out":23635,"duration_ms":273847,"concrete_test":"Construct the reflectionless potential from infinite-soliton data κ_n=1/n^2 with positive norming constants c_n chosen so that u(x)=-2 d^2/dx^2 log det(I+C(x)) is meromorphic in a neighborhood of R and satisfies (1.8); compute a(k)=∏_{n≥1}(k-iκ_n)/(k+iκ_n). If the product has zeros accumulating at 0 and the potential obeys (A1), then Proposition 2.1(iv) is false and the finite-sum proof of Theorem 1.3 does not cover this case; if no such L^1 meromorphic potential exists, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Proposition 2.1(iv) argues that if b(k)=0 on R*, then 'by its analyticity a(k) has no zero near k=0 in C+.' This step is not justified: a(k) is analytic in C+ and nonzero on R*, but k=0 is a boundary point of C+, so the identity theorem does not forbid zeros accumulating there. Condition (A1), which only requires L^1 decay on the real tails and meromorphy in a neighborhood of R, does not exclude infinitely many bound states with eigenvalues κ_n → 0. Standard inverse scattering admits reflectionless potentials with infinitely many discrete eigenvalues accumulating at 0, and such potentials can be L^1. If a(k) has infinitely many zeros in C+ accumulating at 0, the proof of Theorem 1.3, which represents the Jost solution as a finite sum over the zeros k_j via (3.5)-(3.6), does not apply. Thus the central construction currently rests on an unproved, likely false, finiteness assertion.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies solvability by quadrature of the Schrödinger equation (1.5) in the Lax pair for the KdV equation when the potential is meromorphic and satisfies the L¹-tail condition (A1). It proves three results: Theorem 1.3, if the potential is reflectionless and a(k) has a zero in the upper half-plane, then (1.5) is solvable by quadrature; Theorem 1.4, if u is analytic at infinity and (1.5) is solvable by quadrature for all k∈C*, then u is reflectionless; and Theorem 1.6, rational potentials whose denominator degree exceeds the numerator degree by one are not solvable by quadrature for some k∈R*. The abstract states a general if-and-only-if characterization under (A1). The paper extends the author's previous analytic-potential work [29] and uses inverse scattering, local differential Galois theory, and Kovacic's algorithm.","tokens_in":16498,"tokens_out":17050,"duration_ms":174159,"significance":"If the stated results were fully established, they would provide a Galois-theoretic marker for closed-form KdV evolution under meromorphic initial data, going beyond the analytic potentials of [29] and connecting inverse scattering with differential Galois theory. The paper contains useful concrete material: Example 3.2 gives an explicit two-soliton-type potential with a confirmed scattering picture, and Remark 3.1 corrects a sign error in formula (4.12) of [29]. The use of Ramis's local Galois group theorem and Kovacic's algorithm is appropriate. The significance is, however, reduced by the fact that the abstract's unrestricted claim is not what the theorems prove, and by unresolved technical gaps in the analytic continuation of the scattering data and in the finiteness of bound states.","major_comments":[{"comment":"The abstract claims that, under condition (A1), Eq. (1.5) is solvable by quadrature if and only if the meromorphic potential is reflectionless. This is stronger than anything proved in the paper. Theorem 1.3 assumes a(k) has a zero in C+, and Theorem 1.4 assumes u is analytic at infinity. Remark 1.5(i) explicitly concedes that reflectionless potentials satisfying (A1) with a(k)=1, such as the Adler-Moser potentials, are not covered by Theorem 1.3. The abstract's characterization therefore does not follow from the theorems and should be weakened or proved separately.","section":"Abstract; Theorems 1.3, 1.4; Remark 1.5(i)"},{"comment":"The proof of Proposition 2.1(iv) asserts that if b(k)=0 on R*, then 'by its analyticity a(k) has no zero near k=0 in C+.' This inference is invalid: k=0 is a boundary point of the domain C+, and the identity theorem does not prevent zeros of a bounded analytic function from accumulating at a boundary point. Condition (A1), which only gives L¹ decay on the real tails, does not by itself rule out infinitely many bound states with eigenvalues accumulating at zero; standard inverse scattering admits reflectionless potentials with infinite discrete eigenvalues accumulating at 0. If a(k) has infinitely many zeros in C+ accumulating at 0, the finite sum in (3.5) and the subsequent linear-algebraic representation of N^r_j(x) do not apply. Thus Proposition 2.1(iv) must either be proved under (A1) plus reflectionlessness or replaced by an explicit finite-bound-state hypothesis in Theorem 1.3.","section":"Proposition 2.1(iv); proof of Theorem 1.3"},{"comment":"The paper does not justify the complex-analytic continuation of b(k) that the main proofs require. Proposition 2.1 only states analyticity of b on R*, yet the proof of Theorem 1.3 evaluates b(k) and its derivatives at the complex zeros k_j of a(k) in equations (3.4) and (3.5), and the proof of Theorem 1.4 uses the identity theorem for b on a neighborhood of R*. For general L¹ potentials, b is defined on the real axis and need not be holomorphic in a complex neighborhood of R*. The curve Γ introduced in the proof of Proposition 2.1 does not show that the Wronskians computed on Γ agree with the scattering data defined by the real-axis asymptotics (1.4), because no connection from Γ to R avoiding the poles of u is established. These continuation issues are load-bearing for the proofs of both Theorems 1.3 and 1.4 and must be addressed.","section":"Proposition 2.1(i)-(iii); Sections 3 and 4"}],"minor_comments":[{"comment":"The text says 'This means that m2−m1 = −4', but from u(x)=Cq(x)^{-4} one gets m2−m1 = 4deg(q), not −4. The contradiction with m2−m1=1 still works because deg(q)≥1, so the sign should be corrected.","section":"Section 5, after equations (5.3)-(5.4)"},{"comment":"The displayed formulas for the formal fundamental matrix are typeset incorrectly, for example 'V(y) = ( * * Φ(y;k) (0 c− ) )' and the corresponding line for Ψ, which makes the argument difficult to follow; these displays should be rewritten with explicit matrix entries.","section":"Lemma 4.1 proof"},{"comment":"There is a duplicated word in 'we have have α− or α+ = 0'; it should read 'we have α− or α+ = 0'.","section":"Proof of Theorem 1.4"},{"comment":"The word 'triangulariable' in Proposition B.1 should be 'triangularizable', matching the usage in Proposition B.2.","section":"Appendix B.1"},{"comment":"There is a typo 'meromporphic' in the abstract, and in the proof of Proposition 2.1 the inference to part (iii) from the identity theorem is too terse: analyticity in C+ alone does not give discreteness of zeros in C+∪R unless the function is known to be analytic across the real axis.","section":"Abstract and Proposition 2.1 proof"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is largely a self-extension of the author's own paper [29], and the technical gaps are concentrated exactly in the new meromorphic setting: the analytic continuation of b to complex k and the finiteness of the zero set of a. These are fixable in principle by adding explicit hypotheses (e.g., finitely many bound states, and a growth condition ensuring the needed continuation) and by rewriting the abstract to match the theorems. If the author can supply those fixes, the paper would be publishable after a substantial revision. I do not see grounds for rejection, but the current abstract and Theorem 1.3 as stated overreach the proofs."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this paper extends the earlier Galois-theoretic characterization of reflectionless Schrödinger potentials from exponentially decaying analytic potentials to meromorphic ones satisfying a mild L1 condition, and it does so with real mathematical work. But the abstract promises an iff that the theorems do not deliver, and one supporting claim about finiteness of bound states looks genuinely shaky.\n\nWhat is new: Theorem 1.3 and Theorem 1.4 extend the author's prior result [29] from condition (A0) to condition (A1). Theorem 1.3 gives quadrature solvability for reflectionless potentials with a zero in C+, and Theorem 1.4 gives the converse for potentials analytic at infinity. Theorem 1.6 provides a new non-integrability criterion for rational potentials with m2-m1=1. The proofs are detailed, the worked example in Section 3 is helpful, and Remark 1.5(i) is honest about the scope of Theorem 1.3.\n\nThe first soft spot is the abstract. It says \"if and only if the meromorphic potential is reflectionless\" under just (A1), but the theorems require extra hypotheses: a zero of a(k) in C+ for the forward direction, and analyticity at infinity for the converse. Remark 1.5(i) even concedes that Adler-Moser potentials have a(k)=1 and b=0, so Theorem 1.3 does not cover them. The abstract should be aligned with the body.\n\nThe second, more serious soft spot is Proposition 2.1(iv). The proof claims that b(k)=0 on R* implies a(k) has no zero near k=0 in C+, citing analyticity. But a(k) is only shown analytic in C+ and near R*, not at 0. The identity theorem does not rule out zeros accumulating at the boundary point 0. Reflectionless potentials with infinitely many bound states accumulating at zero energy are standard objects in inverse scattering, and they can satisfy (A1) when the potential is a suitably chosen infinite sum of solitons, meromorphic on a strip and L1. If a(k) has infinitely many zeros, the finite-sum representation used in the proof of Theorem 1.3 collapses. The paper does not address this, and the theorem as stated may be false if such potentials are not solvable by quadrature in finite terms.\n\nThe rest is in better shape. The Kovacic-based proof of Theorem 1.6 looks careful, and the local differential Galois analysis in Section 4 is sound. The citation pattern is normal, with [29] as a template; the new proofs are carried out here.\n\nWho is this for? Specialists in integrable systems and differential Galois theory working on KdV and reflectionless potentials. They will find the question interesting and the techniques useful, but the finiteness issue needs to be resolved. I would send it to a referee with an explicit request to examine Proposition 2.1(iv), and I would expect a revision.","headline":"Extends a known Galois-theoretic criterion for reflectionless KdV potentials to meromorphic data, but the abstract overclaims an iff and a key finiteness claim in Proposition 2.1(iv) is unproved.","tokens_in":16986,"tokens_out":9090,"would_cite":false,"duration_ms":99881,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q53","37K15","34M03","34M15","34M35","34M40","35P25"],"pacs":[],"model":"deepseek-v4-flash","headline":"Reflectionless meromorphic potentials are exactly the ones for which the KdV Schrödinger equation is solvable by quadrature.","keywords":["Korteweg-de Vries equation","meromorphic potentials","inverse scattering transform","solvability by quadrature","differential Galois theory","reflectionless potential","Jost solutions","Kovacic algorithm"],"falsifier":"Take the potential (3.9) from Example 3.2 and numerically solve (1.5) at a real k away from the pole; if the solution does not match the closed form (3.8) up to numerical precision, Theorem 1.3 fails for that case. A sharper test would be to find any meromorphic potential satisfying (A1) with b(k)=0 and a zero of a(k) in C+ whose Jost solution is not a rational function of x and finitely many exponentials, or to find a rational potential with m2-m1=1 for which Kovacic's algorithm produces a Liouvillian solution on an open interval of k.","tokens_in":16067,"feed_emoji":"🌊","tokens_out":6594,"duration_ms":63319,"temperature":0.7,"pith_summary":"This paper asks when the Korteweg-de Vries equation with meromorphic initial data can be solved in closed form, in the precise sense that the Schrödinger equation in the Lax pair is solvable by quadrature in differential Galois theory. Under a mild decay condition on the meromorphic potential, the paper proves that reflectionless potentials are quadrature-solvable whenever the scattering coefficient a(k) has at least one zero in the upper half-plane, and that, for potentials analytic at infinity, quadrature solvability forces the reflection coefficient to vanish identically. A separate theorem shows that rational potentials whose numerator and denominator degrees differ by one are not quadrature-solvable for all real k. The abstract states the equivalence without these side conditions; the theorem statements carry them.","feed_headline":"Reflectionless meromorphic KdV potentials have closed-form solutions","feed_subtitle":"New theorems tie quadrature solvability of the KdV Schrödinger equation to a vanishing reflection coefficient.","key_machinery":"The carrier is the pair of scattering coefficients a(k) and b(k) attached to the Schrödinger equation through the Jost solutions, the two solutions tending to $e^{{-ikx}}$ and $e^{{ikx}}$ at minus and plus infinity; the potential is reflectionless when b(k)=0 on the real axis. In the forward direction, b(k)=0 plus a zero of a(k) in C+ turns the inverse-scattering integral equation into residue sums at finitely many zeros k_j, so the Jost solution and the potential are built from rational functions of x and $e^{{ik_j x}}$. In the converse, the paper rewrites the equation near infinity, reads off the formal monodromy, exponential torus and Stokes matrices, and uses the classification of algebraic subgroups of SL(2,C) to force the Stokes matrices to be triangular, which by Lemma 4.1 kills b(k). The rational-potential theorem uses Kovacic's algorithm, whose pole data at infinity must be independent of k.","core_discovery":"On the paper's own terms, the discovery is a differential-Galois characterization of when the inverse scattering transform for KdV runs on closed-form functions. Theorem 1.3 says that if a meromorphic potential satisfying the decay condition (A1) has scattering coefficient a(k) with a zero in C+ and the reflection coefficient b(k) vanishes on the real axis, then the Schrödinger equation (1.5) is solvable by quadrature for every nonzero k. Theorem 1.4 supplies the converse direction under the extra assumption that the potential is analytic at infinity: if (1.5) is solvable by quadrature for all nonzero k, then b(k) vanishes identically. Theorem 1.6 completes the picture for rational potentials that decay like 1/x, showing they are not quadrature-solvable on an open set of real k. The paper also exhibits a negaton-type potential whose Jost solution is an explicit rational-exponential function.","pith_inferences":["Inference: The zero-of-a(k) condition in Theorem 1.3 may be removable; the paper's Remark 1.5(i) shows a whole family of reflectionless potentials with a(k)=1, so a unified proof would need a different mechanism, possibly Darboux transformations, for the no-zero case.","Inference: The same Stokes-matrix argument should extend to other integrable PDEs in the Zakharov-Shabat class, and the companion paper cited as [30] is the natural place to look; if the meromorphic analogue holds there, quadrature solvability would again be equivalent to reflectionlessness under analyticity at infinity.","Inference: A concrete testable extension is to compute a(k) for positon-type potentials with more than one pole; Theorem 1.3 predicts an explicit formula for the Jost solution whenever a zero exists in C+, and the formula should match direct numerical integration of (1.5)."],"forward_implications":["For any reflectionless meromorphic potential satisfying (A1) with a zero of a(k) in C+, the KdV solution is an explicit combination of exponentials and rational functions of x, so it can be written down without solving differential equations.","The Adler-Moser rational potentials fall outside Theorem 1.3 because their scattering coefficient is a(k)=1 with no zero in C+, even though they are known to be quadrature-solvable; the theorem is not an exhaustive test.","Rational potentials with denominator degree one greater than numerator degree cannot be solved by quadrature for all real k, so the inverse scattering route fails for them despite their simple form.","If the converse direction holds, quadrature solvability itself becomes a spectral test: a closed-form Schrödinger solution for all k forces the potential to be reflectionless."],"supporting_citations":[{"why":"Supplies the previous reflectionless/quadrature theorem under condition (A0) and the overall inverse-scattering strategy that Theorems 1.3 and 1.4 extend.","marker":"[29]"},{"why":"Kovacic's algorithm is the case analysis whose pole data proves Theorem 1.6.","marker":"[13]"},{"why":"Ramis's theorem on local differential Galois groups gives the Zariski closure result linking Stokes matrices to the global Galois group in Theorem 1.4.","marker":"[24,25]"},{"why":"Supply the differential Galois theory foundations and the classification of algebraic subgroups of SL(2,C) used throughout.","marker":"[10,23]"},{"why":"Provides the existence of Jost solutions for potentials that are absolutely integrable away from a bounded interval, which is condition (A1).","marker":"[9]"},{"why":"Establishes the Adler-Moser rational potentials' Galois groups and the a(k)=1 example that motivates the limitations discussed in Remark 1.5.","marker":"[11]"},{"why":"The companion paper proves the analogous statements for Zakharov-Shabat systems, cited as the meromorphic extension of the ZS result.","marker":"[30]"}],"fun_headline_variants":["KdV solvable by quadrature iff reflectionless meromorphic potential","Quadrature solvability of KdV tied to zero reflection coefficient","Meromorphic KdV: closed-form solutions only when scattering is reflectionless","Inverse scattering for KdV: integrable exactly when reflection vanishes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the scattering coefficients a(k) and b(k), defined initially on the real axis, can be analytically continued along a contour in the upper half-plane that avoids the poles of u, and that the Wronskian identities still identify these continuations with the real-axis scattering data.","fun_headline_variants_meta":{"raw":{"variants":["KdV solvable by quadrature iff reflectionless meromorphic potential","Quadrature solvability of KdV tied to zero reflection coefficient","Meromorphic KdV: closed-form solutions only when scattering is reflectionless","Inverse scattering for KdV: integrable exactly when reflection vanishes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001231,"raw_usage":{"total_tokens":5039,"prompt_tokens":908,"completion_tokens":4131,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":524,"completion_tokens_details":{"reasoning_tokens":4054}},"tokens_in":524,"tokens_out":4131,"duration_ms":24612,"temperature":1.0,"reasoning_tokens":4054,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T05:43:49.884755+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the potential (3.9) from Example 3.2 and numerically solve (1.5) at a real k away from the pole; if the solution does not match the closed form (3.8) up to numerical precision, Theorem 1.3 fails for that case. A sharper test would be to find any meromorphic potential satisfying (A1) with b(k)=0 and a zero of a(k) in C+ whose Jost solution is not a rational function of x and finitely many exponentials, or to find a rational potential with m2-m1=1 for which Kovacic's algorithm produces a Liouvillian solution on an open interval of k.","supporting_citations":[{"cited_title":"Kovacic, An algorithm for solving second order lin ear homogeneous differential equations, J","cited_arxiv_id":null,"evidence_quote":"Kovacic's algorithm is the case analysis whose pole data proves Theorem 1.6."},{"cited_title":"Jim´ enez, J.J","cited_arxiv_id":null,"evidence_quote":"Establishes the Adler-Moser rational potentials' Galois groups and the a(k)=1 example that motivates the limitations discussed in Remark 1.5."},{"cited_title":"Solvability of the Zakharov-Shabat systems with meromorphic potentials by quadrature","cited_arxiv_id":"2506.07246","evidence_quote":"The companion paper proves the analogous statements for Zakharov-Shabat systems, cited as the meromorphic extension of the ZS result."}],"review_version":1}