{"id":"f6589fe1-e9d8-434d-b5e7-3cf6520d26e5","arxiv_id":"2502.07385","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":6,"one_line_summary":"Non-uniqueness of C^{1,alpha} weak solutions for the nonlinear dynamical Lame system is claimed via convex integration, but the constructed solutions fail the paper's H2 weak-solution regularity condition.","lead":"A convex integration proof claims infinitely many C^{1,alpha} solutions of a nonlinear Lame wave system from the same small initial data, but the constructed solutions do not meet the paper's own H2 regularity requirement for weak solutions. The new building blocks that exploit double wave speeds are the technical heart, and the result may hold under a weaker weak-solution definition.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proof's parameter condition \\bar b(\\beta)>1+6\\beta fails for \\beta near 1/60, so Theorem 1.3's claimed exponent range is not established.","rationale":"The reader's verdict (REJECT) was based on the mismatch between Definition 1.1 requiring H^2 regularity and the constructed C^{1,\\alpha} limit. That concern is real but ultimately patchable by relaxing the weak-solution definition to the distributional sense, since the integral formulation (1.2) only needs first derivatives in L^1. My stress-test identifies a different, more sharp obstruction: the proof of Theorem 1.3 explicitly chooses b=\\bar b(\\beta) and requires \\bar b(\\beta)>1+6\\beta, yet this inequality fails for \\beta near 1/60. Because Proposition 3.3 requires b>1+6\\beta, the mollification step breaks for exactly the \\beta values needed to reach \\alpha close to 1/60. Thus the central quantitative claim is not proven as stated. The paper's building-block construction (Lemma 3.4) and the convex-integration framework appear internally coherent; the issue is a concrete parameter-regime error, not a vague concern about the method. A single arithmetic check settles it, so the concern is testable. I therefore keep the REJECT verdict but disagree with the reader's identification of the weakest assumption.","tokens_in":45838,"tokens_out":28631,"duration_ms":217953,"concrete_test":"Evaluate S(\\beta)=1-36\\gamma(\\beta)-36\\beta-288\\beta^2 at \\beta=1/60; since \\bar b(\\beta)>1+6\\beta is equivalent to S(\\beta)>0, showing S(1/60)<0 disproves the stated range. Also compute S at \\beta=1/63 and at the numerical threshold ~0.01588 to confirm where the condition changes sign.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 1.3 claims C^{1,\\alpha} non-uniqueness for every 0<\\alpha<1/60. In the proof (Section 6.3), \\alpha<\\beta<1/60, and Proposition 2.3 fixes b=\\bar b(\\beta), where Proposition 5.1 requires \\bar b(\\beta)=(1+12\\beta-36\\gamma)/(48\\beta)>1+6\\beta, with \\gamma=(1/3)((1+12\\beta)/24-\\sqrt{\\beta-6\\beta^2}/6). For \\beta=1/60, direct arithmetic gives \\gamma\\approx0.00986, hence \\bar b(1/60)\\approx1.056, whereas 1+6/60=1.1. Thus the required inequality fails for all \\beta in an interval approaching 1/60 (numerically \\beta\\gtrsim0.01588\\approx1/63). Since Proposition 3.3 needs b>1+6\\beta for the mollification estimates, and b is taken equal to \\bar b, the induction cannot close for these \\beta. Consequently the claimed range \\alpha<1/60 is not supported by the proof; at best the construction appears to yield \\alpha up to about 1/63. This is an internal parameter inconsistency, independent of the separate issue that the constructed C^{1,\\alpha} limit may not meet the H^2-based weak-solution Definition 1.1.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a convex-integration construction of Hölder-regular weak solutions to the Cauchy problem for the nonlinear dynamical Lamé system (1.1) on T^3 (and T^2), with the same small initial data. The main claim (Theorem 1.3) is that for μ>0 and λ+μ>0, for every 0<α<1/60 there are infinitely many distinct C^{1,α} weak solutions with the same small initial data. The proof follows the De Lellis–Székelyhidi scheme: an induction on approximate solutions (u_q,c_q,R_q), mollification, a six-step perturbation using planar building blocks whose transverse amplitudes solve an algebraic system (Lemma 3.4), estimates on the Reynolds error (Proposition 5.1), and a bifurcation argument (Proposition 2.4). The paper also announces a two-dimensional analogue with exponent 1/30 (Theorem 1.4).","tokens_in":46057,"tokens_out":10616,"duration_ms":101347,"significance":"The novelty is genuine: applying convex integration to a quasi-linear wave system with double wave speeds, exploiting the linear-degeneracy and null-condition structure of the nonlinearity, is new compared with existing Euler/MHD constructions. The paper includes a detailed algebraic building-block lemma, explicit linear-degeneracy computations, and self-contained appendices; the main technical work is not outsourced. If the construction can be repaired as discussed below, it would give the first non-uniqueness result for the nonlinear dynamical Lamé system in a Hölder class. The proof follows the standard convex-integration pattern and contains a large amount of useful technical detail, which is a strength.","major_comments":[{"comment":"The claimed range 0<α<1/60 is not supported. Proposition 5.1 requires b̄(β)=(1+12β−36γ)/(48β)>1+6β with γ=(1/3)((1+12β)/24−√(β−6β²)/6). Direct arithmetic at β=1/60 gives γ≈0.00986 and b̄(1/60)≈1.056<1.1=1+6/60. The inequality fails numerically for all β≳0.01588≈1/63. Since Theorem 1.3 fixes β∈(α,1/60) and Proposition 2.3 sets b=b̄(β), the induction cannot close for α in (0.01588,1/60); for such α no admissible β exists. At best the proof establishes the construction up to about 1/63, not up to 1/60. This is an internal parameter inconsistency, not merely an endpoint-limiting issue.","section":"Proposition 5.1, §6.3"},{"comment":"The limit obtained in Section 6.3 is only shown to lie in C^{1,α′}([0,T]×T³). Definition 1.1, however, defines a weak solution as a function in C([0,T];H²(T³))∩C¹([0,T];H¹(T³)) with ∂_t²u∈L²(0,T;L²(T³)) and initial data in H²×H¹. Hölder C^{1,α} regularity does not imply H² spatial regularity or L² second time derivatives on T³, and no estimate in the paper supplies these norms. The proof verifies only that the equation holds after passing to the limit in the approximate equations; it never checks the regularity clauses of Definition 1.1. Thus Theorem 1.3 as stated is not established. Either Definition 1.1 must be weakened to a distributional/continuous weak-solution notion consistent with the construction, or additional H² estimates must be provided.","section":"Definition 1.1; §6.3"}],"minor_comments":[{"comment":"Theorem 1.4 is stated for d=2 with exponent 1/30, but no proof is given; the introduction says the three-dimensional result is a corollary of the two-dimensional one, which is also confusing because the 3D theorem is the one proved. Please either prove Theorem 1.4 explicitly or state it as a remark with the necessary changes.","section":"Theorem 1.4"},{"comment":"The notation C^{1,1/60−} and C^{1,1/30−} should be replaced by the precise quantifier 'for every α<1/60' (respectively α<1/30) used in Theorems 1.3 and 1.4, to avoid ambiguity about endpoint regularity.","section":"Introduction"},{"comment":"The condition b̄(β)>1+6β appears only in Proposition 5.1, but the proof of Proposition 2.3 sets b=b̄(β) and relies on it. Since the admissible β interval is exactly what is at stake, this hypothesis should be stated explicitly in Proposition 2.3.","section":"Proposition 2.3"}],"recommendation":"major_revision","confidential_remarks":"The two major issues are independent. The parameter-range failure directly affects the advertised exponent 1/60, while the weak-solution regularity gap affects the interpretation of the theorem. Both appear repairable within the scope of a revision: the exponent can be lowered to the range where b̄(β)>1+6β actually holds, and the weak-solution definition can be adjusted to the regularity that the convex-integration construction actually produces. Given the substantial and self-contained technical work, I would not reject the paper outright, but the current statement cannot be accepted as it stands."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a genuine convex integration paper for a quasilinear wave system with two wave speeds, and the new longitudinal-transverse building blocks are a real idea worth engaging. I don't think the paper is correct as stated, but the core mechanism is promising and the problems look repairable.\n\nWhat's new: first non-uniqueness construction for the nonlinear dynamical Lamé system; building blocks exploit λ+μ>0 to get a planar wave with a longitudinal main wave and small transverse corrections. The appendices carefully verify linear degeneracy and solve the algebraic system with Newton iteration. The induction follows the standard De Lellis–Székelyhidi/Buckmaster pattern, with honest parameter chasing.\n\nSoft spots, in order of importance. First, the claim α<1/60 is not established. Proposition 5.1 requires \\bar b(β)=(1+12β−36γ)/(48β)>1+6β. At β=1/60, γ≈0.00986, so \\bar b≈1.056, while 1+6β=1.1. The inequality fails for β≳1/63, so the construction only gives α up to about 1/63. This is an internal parameter inconsistency in the proof, independent of the weak-solution definition issue. It can be fixed by shrinking the exponent range or reworking the parameter choice, but Theorem 1.3 as written is false as proven.\n\nSecond, Definition 1.1 defines weak solutions with H^2 spatial regularity, but the constructed limit is only C^{1,α} with α<1/60, which does not imply H^2 on T^3. The convergence argument passes to the limit in the distributional formulation, so the natural interpretation is distributional weak solutions; the definition and theorem statement need to be aligned. This is a mismatch, not a hidden contradiction in the construction.\n\nMinor: the paper is long and the estimates are dense; some notation (δ_{q,i}, I+σ) is not always defined clearly, but a referee can sort that out.\n\nFor whom: researchers in convex integration and low-regularity well-posedness of quasilinear waves. It deserves a serious referee: the technical core is novel and mostly standard, and the flaws are correctable rather than fatal. My recommendation is to send to peer review with a request for significant revision, specifically to fix the exponent range and reconcile the weak-solution definition.","headline":"Solid convex integration construction for the dynamical Lamé system, but the stated Hölder range and the weak-solution definition both need fixing before the theorem as written is supported.","tokens_in":46635,"tokens_out":4143,"would_cite":false,"duration_ms":38624,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35A02","35D30","35L05","35L15","35L72"],"pacs":[],"model":"deepseek-v4-flash","headline":"The nonlinear dynamical Lamé system with two distinct wave speeds admits infinitely many weak solutions in $C^{1,\\alpha}$ for any $\\alpha<1/60$ in three dimensions (and $\\alpha<1/30$ in two), all sharing the same small initial data.","keywords":["Lamé system","non-uniqueness","convex integration","weak solutions","Hölder regularity","double wave speeds","linearly degenerate hyperbolic system","Cauchy problem"],"falsifier":"Compute the algebraic system (3.18) in the equal-speed case $\\lambda+\\mu=0$ for a generic amplitude tensor $A$: if the $2\\times 2$ coefficient matrix becomes singular or the unique solution violates the bound $|a_{A,2}|+|a_{A,3}|\\lesssim \\varepsilon$, then Lemma 3.4 fails and the non-uniqueness result cannot be obtained by this method. Alternatively, run the same construction with $\\lambda+\\mu=0$ and check whether the perturbation estimates in Proposition 3.7 degrade as $\\lambda+\\mu\\to 0^+$.","tokens_in":45570,"feed_emoji":"🌊","tokens_out":5624,"duration_ms":48800,"temperature":0.7,"pith_summary":"The paper proves that the nonlinear dynamical Lamé system, a model of elastic waves with two distinct wave speeds, is non-unique at low Hölder regularity: the same arbitrarily small initial data can produce infinitely many distinct weak solutions in $C^{1,\\alpha}$ for $\\alpha<1/60$ in three dimensions and $\\alpha<1/30$ in two. This is the first non-uniqueness result of this kind for the Lamé system, built with a convex integration scheme. The central mechanism is a new class of building blocks: planar waves combining a longitudinal component of size $O(1)$ with a transverse correction of size $O(\\varepsilon)$, which exist only when the two wave speeds genuinely differ, i.e. $\\lambda+\\mu>0$. If correct, the result shows that low Hölder regularity does not select a unique solution for this quasi-linear hyperbolic system, even though the nonlinearity satisfies the null condition and classical small-data well-posedness holds in higher regularity.","feed_headline":"One initial condition, infinitely many elastic solutions","feed_subtitle":"Convex-integration proof shows double wave speeds make low-regularity weak solutions non-unique.","key_machinery":"The central object is the planar-wave building block of Lemma 3.4: a nontrivial solution $w_{A,f}$ of the linearized Lamé-type equation, expressed as $(f+a_{A,2}f^\\perp+a_{A,3}\\frac{f}{|f|}\\times f^\\perp)e^{i\\xi_{A,f}}$ with $\\xi_{A,f}=f\\cdot x-((\\lambda+2\\mu)|f|^2-c_A)^{1/2}t$. The longitudinal component propagates at speed $\\sqrt{\\lambda+2\\mu}$, while the transverse amplitudes $a_{A,2}$, $a_{A,3}$ are fixed by the algebraic system (3.17)–(3.18), whose solution bounds rely on $\\lambda+\\mu>0$. These building blocks are inserted through a geometric decomposition of the identity (Lemma 2.2) to cancel the Reynolds error, and the perturbation is added in only one direction at a time, six times per iteration, to control the low–high frequency interactions that cannot be eliminated; this six-step pattern is what makes the 3D argument work.","core_discovery":"The paper establishes that for $\\mu>0$ and $\\lambda+\\mu>0$, for every $0<\\alpha<1/60$ (in $d=3$) and $0<\\alpha<1/30$ (in $d=2$), the Cauchy problem (1.1) with suitably small initial data has infinitely many distinct weak solutions in $C^{1,\\alpha}([0,T]\\times\\mathbb{T}^d)$. The proof constructs a sequence of approximate solutions whose Reynolds error is driven to zero; at each stage the perturbation is a high-frequency superposition of planar waves chosen so that the quadratic nonlinearity cancels the leading error. The novelty is a building block of Lemma 3.4, a planar wave written as $(f+a_{A,2}f^\\perp+a_{A,3}\\frac{f}{|f|}\\times f^\\perp)e^{i\\xi_{A,f}}$, where the longitudinal wave travels at the $P$-wave speed and the small transverse amplitudes $a_{A,2}, a_{A,3}$ solve the algebraic system (3.18). This system is solvable precisely when $\\lambda+\\mu>0$, so the double wave speed property is load-bearing rather than a benign assumption. A bifurcation step then produces two solutions with identical initial data but different values in $L^2$, and repeating the argument yields infinitely many distinct solutions.","pith_inferences":["One might expect the same double-speed mechanism to be adaptable to other two-speed hyperbolic systems, such as full elastodynamics with distinct wave speeds, provided an analogue of the algebraic system (3.18) remains solvable.","The exponent $1/60$ is tied to the specific parameter choices, mollification, and geometric decomposition in the iteration; the paper does not claim this threshold is optimal, and a natural open question is whether it can be raised substantially.","A separate question, not addressed by the paper, is whether the constructed non-unique solutions survive additional selection criteria such as an energy admissibility condition; the construction only shows weak-solution non-uniqueness in $C^{1,\\alpha}$.","Since the initial data are required to be small and the admissible size $\\varepsilon$ depends on $\\lambda+\\mu$, a testable consequence is that the non-uniqueness regime shrinks as the two wave speeds approach each other, potentially disappearing in the equal-speed limit $\\lambda+\\mu=0$."],"forward_implications":["For any Hölder exponent below $1/60$ in three dimensions and below $1/30$ in two dimensions, the Cauchy problem for the nonlinear dynamical Lamé system is not well-posed: the same small initial data has infinitely many weak solutions in $C^{1,\\alpha}$.","The double wave speed condition $\\lambda+\\mu>0$ drives the construction: if the two wave speeds coincide, the algebraic system (3.18) for the building block cannot be solved with the required bounds, so the perturbation scheme collapses.","The result transplants the convex-integration non-uniqueness phenomenon from fluid equations to a quasi-linear hyperbolic system with a null-condition nonlinearity, showing that the null condition alone does not enforce uniqueness at $C^{1,\\alpha}$ regularity.","The bifurcation estimate gives a quantitative lower bound on the $L^2$ distance between two solutions, of order $\\delta_{q+1}^{1/2}/\\lambda_{q+1}$, and the support of their difference can be localized to a prescribed time interval.","In the two-dimensional case, the same argument yields the stronger exponent $\\alpha<1/30$, and the three-dimensional result is deduced as a corollary of the two-dimensional construction."],"supporting_citations":[{"why":"Supplies the geometric lemma (Lemma 2.2) that decomposes a symmetric tensor into rank-one projections on fixed lattice directions, used to cancel the Reynolds error.","marker":"[28]"},{"why":"Provides the inductive framework and bifurcation argument for constructing distinct Hölder-continuous approximate solutions from the same initial data.","marker":"[25]"},{"why":"Supplies the mollification, cut-off, and inductive estimates that control the Reynolds stress through the iteration.","marker":"[30]"},{"why":"Introduces the strategy of adding perturbations in only one direction at each substep, used here to handle the low–high frequency interactions in the 3D construction.","marker":"[53]"},{"why":"Provides the inverse divergence estimates and oscillatory integral bounds (Lemmas D.1 and D.2) needed to estimate the new Reynolds error.","marker":"[13]"},{"why":"Gives the definition and criterion of linear degeneracy used to verify that the transformed Lamé system is linearly degenerate.","marker":"[55]"}],"fun_headline_variants":["Double wave speeds yield infinite elastic solutions","Non-unique elastic waves: one start, many outcomes","Infinite weak solutions from same initial data","Lamé system: convex integration proves non-uniqueness"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The construction requires the two elastic wave speeds to be genuinely different (that is, $\\lambda+\\mu>0$); if they coincide, the algebraic system fixing the transverse amplitudes of the building blocks cannot be solved with the needed bounds, and the entire iteration collapses.","fun_headline_variants_meta":{"raw":{"variants":["Double wave speeds yield infinite elastic solutions","Non-unique elastic waves: one start, many outcomes","Infinite weak solutions from same initial data","Lamé system: convex integration proves non-uniqueness"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000153,"raw_usage":{"total_tokens":1185,"prompt_tokens":899,"completion_tokens":286,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":515,"completion_tokens_details":{"reasoning_tokens":226}},"tokens_in":515,"tokens_out":286,"duration_ms":3244,"temperature":1.0,"reasoning_tokens":226,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-08T12:57:19.812403+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the algebraic system (3.18) in the equal-speed case $\\lambda+\\mu=0$ for a generic amplitude tensor $A$: if the $2\\times 2$ coefficient matrix becomes singular or the unique solution violates the bound $|a_{A,2}|+|a_{A,3}|\\lesssim \\varepsilon$, then Lemma 3.4 fails and the non-uniqueness result cannot be obtained by this method. Alternatively, run the same construction with $\\lambda+\\mu=0$ and check whether the perturbation estimates in Proposition 3.7 degrade as $\\lambda+\\mu\\to 0^+$.","supporting_citations":[{"cited_title":"Buckmaster, C","cited_arxiv_id":null,"evidence_quote":"Provides the inverse divergence estimates and oscillatory integral bounds (Lemmas D.1 and D.2) needed to estimate the new Reynolds error."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the definition and criterion of linear degeneracy used to verify that the transformed Lamé system is linearly degenerate."}],"review_version":1}