{"id":"ded4bd45-1385-42ea-a778-6d064f844aee","arxiv_id":"2412.16756","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The spectrum of a self-adjoint extension of a discrete symplectic system in the limit point case is completely determined by analytic behavior of the limiting Weyl-Titchmarsh M-function.","lead":"This paper proves that for a class of discrete symplectic systems on a half-line, the full spectrum of any self-adjoint boundary value problem is encoded in a single limiting Weyl-Titchmarsh matrix function. It also studies how the spectrum changes when the boundary condition is changed, giving interlacing and countability results.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 4.4's pole-counting argument targets the wrong Möbius denominator: it counts zeros of det(αα̂*−αJα̂*M_+(λ,α̂)), which are potential poles of M_+(λ,α), not of M_+(λ,α̂); the at-most-m-poles conclusion is unproven.","rationale":"The reader's verdict is CONDITIONAL and cites the Theorem 4.4 proof gap; my stress-test confirms that this is a genuine, localized flaw. I do not agree with the reader's weakest_assumption that the strong Atkinson condition is the most load-bearing fragility: the paper explicitly assumes it, and Theorem 1.1's proofs appear consistent under that assumption. The real soft spot is the Möbius counting error in Section 4, which undermines the advertised Sturmian boundary-dependence results (Theorem 4.4 and Corollaries 4.5–4.9) but does not invalidate the central resolvent/spectrum characterization in Theorem 1.1. I also considered the square-summability step in Theorem 3.1 for real λ0, where the text invokes an estimate with 1/(Im λ); however, the surrounding reference to [7] and the equality X_+^*JX_+=0 may provide a valid substitute, and I am not confident enough to raise it as a separate objection. Thus the reader's conditional verdict stands: the paper should be accepted only after the proof of Theorem 4.4 is corrected or replaced by an argument that counts poles of M_+(λ,α̂) via the inverse Möbius transformation.","tokens_in":24961,"tokens_out":26770,"duration_ms":214925,"concrete_test":"Invert the Möbius transformation (4.1) to express M_+(λ,α̂) = (αα̂* + M_+(λ,α) αJα̂*)^{-1} (M_+(λ,α) αα̂* − αJα̂*) and count zeros of det(αα̂* + M_+(λ,α) αJα̂*) on (a,b) using Lemma 2.5 applied to the holomorphic function M_+(λ,α). For the scalar case, take α0=0, α̂0=π/2, so αα̂*=0 and αJα̂*=−1, and choose an interval (a,b) where M_+(λ,α)=λ crosses the value 0; the proof's determinant −M_+(λ,α̂) has no zeros on (a,b) (indeed M_+(λ,α̂) = −1/λ has a pole at λ=0), while the inverse denominator M_+(λ,α) has one zero. This directly verifies that the counting argument in Theorem 4.4 targets the wrong determinant.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The main spectral characterization in Theorem 1.1 is supported by Theorems 3.1–3.4 and appears internally sound. The load-bearing defect lies in the advertised Sturmian application, Theorem 4.4, which claims that if (a,b) ⊆ ρ(T_LP(α)), then M_+(λ,α̂) has at most m = rank(αJα̂*) simple poles in (a,b). The proof counts zeros of det(αα̂* − αJα̂* M_+(λ,α̂)). However, in the forward Möbius relation (4.1), this determinant is the denominator of M_+(λ,α) expressed in terms of M_+(λ,α̂); its zeros are therefore singularities of M_+(λ,α), not of M_+(λ,α̂). Since M_+(λ,α) is holomorphic on (a,b) by assumption, these zeros either do not occur or cancel with the numerator, so they say nothing about poles of M_+(λ,α̂). The poles of M_+(λ,α̂) are zeros of the inverse-transformation denominator, obtained by solving (4.1) for M_+(λ,α̂), namely det(αα̂* + M_+(λ,α) αJα̂*) (up to matrix ordering). The scalar case n=1 makes the error transparent: with α=(sinα0,cosα0), α̂=(sinα̂0,cosα̂0), the counted determinant is cos(α0−α̂0) − sin(α0−α̂0)M_+(λ,α̂), which cannot vanish where M_+(λ,α) is holomorphic, whereas the inverse denominator cos(α0−α̂0) + sin(α0−α̂0)M_+(λ,α) has one zero between consecutive poles, exactly as Corollary 4.9 asserts. Thus Lemma 2.5 is applied to K11 built from the wrong function, and the proof of Theorem 4.4 does not establish its conclusion.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the spectral theory of discrete symplectic systems (S_λ) on the half-line in the limit point case, under the strong Atkinson condition. The main result (Theorem 1.1, proved as Theorems 3.1–3.4) asserts that for any self-adjoint extension T_LP(α) of the minimal linear relation, the spectrum is completely encoded in the limiting Weyl–Titchmarsh function M_+(λ): λ0 is in the resolvent set iff M_+ is holomorphic at λ0; isolated eigenvalues correspond exactly to simple poles of M_+, with the residue related to the jump of the limiting spectral function; and the essential spectrum is split into point-continuous and continuous parts according to whether lim_{ν→0} ν M_+(λ0+iν) is nonzero or zero. The paper also investigates how the spectrum depends on the boundary condition α through the Möbius transformation (4.1), and derives consequences for Sturmian theory (Theorems 4.3, 4.4, 4.7 and Corollaries 4.5–4.9).","tokens_in":25406,"tokens_out":18496,"duration_ms":130055,"significance":"The complete characterization of the spectrum in terms of a single limiting M-function is a substantial and valuable result for discrete symplectic systems, extending to the linear-relation setting results known for Hamiltonian differential systems (Hinton–Shaw) and correcting earlier work on linear Hamiltonian difference systems. The proofs of Theorems 3.1–3.4 are detailed and use a coherent toolkit: the extended Lagrange formula, integral representations of Nevanlinna functions, and careful square-summability arguments. The paper gives explicit credit to prior work and identifies exactly which steps require the strong Atkinson condition. If the Section 4 applications were supported by correct proofs, the Sturmian interlacing results would be a useful complement to the existing oscillation theory in [8]. At present, the advertised boundary-dependence results rest on flawed proof arguments and require revision.","major_comments":[{"comment":"The proof analyzes the determinant det(αα̂* − αJα̂* M_+(λ,α̂)), but by the Möbius relation (4.1) this is the denominator in the expression of M_+(λ,α) in terms of M_+(λ,α̂). Its zeros are therefore potential singularities of M_+(λ,α), which is holomorphic on (a,b) by the hypothesis (a,b) ⊆ ρ(T_LP(α)); they do not locate poles of M_+(λ,α̂). Poles of M_+(λ,α̂) correspond to zeros of the inverse-transformation denominator, obtained by solving (4.1) for M_+(λ,α̂), namely det(αα̂* + M_+(λ,α)αJα̂*) (up to matrix ordering). Consequently, the subsequent application of Lemma 2.5 to the block K11 of K = V* M_+(λ,α̂)V is not justified: M_+(λ,α̂) is not known to be holomorphic on (a,b), and that holomorphy is exactly the conclusion to be proved. The scalar case n=1 makes the error transparent: with α=(sinα0,cosα0) and α̂=(sinα̂0,cosα̂0), the counted determinant is cos(α0−α̂0) − sin(α0−α̂0)M_+(λ,α̂), which cannot vanish when M_+(λ,α) is holomorphic, whereas the inverse denominator cos(α0−α̂0) + sin(α0−α̂0)M_+(λ,α) has one zero between consecutive poles, as asserted in Corollary 4.9. Thus the proof of Theorem 4.4 does not establish the at-most-m-poles conclusion; the same defect affects Theorem 4.7 and Corollaries 4.8 and 4.9.","section":"§4, Theorem 4.4, proof after Eq. (4.1)"},{"comment":"The proof of the invariance of ρ(T_LP(α)) ∪ σ_d(T_LP(α)) states that if M_+(λ,α) is holomorphic on O(λ0), then 'the same is true' for αJα̂* + αα̂*M_+(λ,α̂) and αα̂* − αJα̂*M_+(λ,α̂), i.e., M_+(λ,α̂) is holomorphic on O(λ0). This does not follow from the Möbius relation (4.1): M_+(λ,α̂) may have a pole at λ0 while the transformation (4.1) yields a holomorphic M_+(λ,α) through cancellation of poles in numerator and denominator (explicitly in the scalar case, M_+(λ,α̂) ~ c/(λ−λ0) gives M_+(λ,α) → −cot(α0−α̂0)). The proof should treat (4.1) as an identity of meromorphic functions and use the identity theorem in that category, or otherwise justify the holomorphy assertion. The conclusion of Theorem 4.3 may be correct, but the argument as written relies on an unproven premise.","section":"§4, Theorem 4.3, proof"}],"minor_comments":[{"comment":"The definition of the matrix K is ambiguous: the notation 'Kpαq – V ˚ M`pα, ˆαq V' does not clearly indicate whether the argument is λ, α, or both; the text should specify K(λ) = V* M_+(λ, · ) V with the appropriate boundary condition.","section":"§4, Theorem 4.4 proof"},{"comment":"The proof of part (i) is terse; the strong Atkinson condition is used to conclude 0 < ||X_+(λ)||²_Ψ, which indeed follows from the uniqueness property, but a sentence making this explicit would aid the reader.","section":"Lemma 2.5"},{"comment":"There is a typo: 'if any only if' should be 'if and only if'.","section":"Theorem 3.2 proof"}],"recommendation":"major_revision","confidential_remarks":"The manuscript's header indicates it was published in Linear Algebra Appl. 634 (2022). If this is a resubmission of a published paper, the editor should consider whether it meets the journal's criteria for submitted work. My report is based solely on the scientific content."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core of this paper is strong. Theorem 1.1 gives a complete description of the spectrum of an arbitrary self-adjoint extension via the limiting Weyl–Titchmarsh function M_+: resolvent set, isolated eigenvalues (simple poles), point-continuous and continuous spectrum are all characterized cleanly. The proofs in Sections 2 and 3 look coherent, and the treatment of the non-densely defined operator through linear relations is necessary and well executed. The extension from Hinton–Shaw's continuous Hamiltonian systems and from Shi's discrete linear Hamiltonian systems to time-reversed discrete symplectic systems with general linear dependence on the spectral parameter is real and nontrivial. Lemmas 2.8 and 2.9, connecting jumps of the spectral function to the limiting value of nu M_+(lambda0 + i nu) and showing that isolated singularities are simple poles, are carefully proven. The author also deserves credit for correcting Shi's omission regarding non-density of the domain.\n\nThe soft spot is Section 4, specifically Theorem 4.4. The proof counts zeros of det(alpha alphahat* − alpha J alphahat* M_+(lambda, alphahat)), but by the Möbius relation (4.1), that determinant is the denominator of M_+(lambda, alpha) expressed in terms of M_+(lambda, alphahat). Its zeros are singularities of M_+(lambda, alpha), not poles of M_+(lambda, alphahat). Since M_+(lambda, alpha) is holomorphic on (a,b) by assumption, those zeros either do not occur or cancel; they say nothing about the poles of M_+(lambda, alphahat). The actual poles of M_+(lambda, alphahat) are zeros of the inverse transformation denominator, det(alpha alphahat* + M_+(lambda, alpha) alpha J alphahat*) up to ordering. The scalar case n=1 makes the mistake transparent: the counted denominator is cos(alpha0−alphahat0) − sin(alpha0−alphahat0)M_+(lambda, alphahat), which cannot vanish where M_+(lambda, alpha) is holomorphic, whereas the inverse denominator has exactly one zero between consecutive poles, as Corollary 4.9 asserts. Additionally, the proof applies Lemma 2.5 to M_+(lambda, alphahat) before establishing its holomorphicity on (a,b), which is circular. The advertised \"at most m points\" result and the interlacing corollaries are not supported as written. Theorem 4.3 may be repairable, but the argument needs a corrected proof or a scaled-back claim.\n\nThe strong Atkinson condition is a genuine restriction, though Remark 2.1 notes that weaker versions suffice for parts of the theory. The self-citations are legitimate; they form a coherent research program and do not hide dependence on unverified results.\n\nBottom line: the main spectral theorem is likely correct and valuable, and the paper deserves a serious referee. I would send it to peer review and ask the author to fix or qualify the Sturmian claims.","headline":"Main M-function spectral characterization for discrete symplectic systems is solid, but the Sturmian section rests on a pole-counting argument that targets the wrong Möbius denominator.","tokens_in":25933,"tokens_out":4752,"would_cite":true,"duration_ms":37784,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["47B39","47A10","47A06","39A06","39A12"],"pacs":[],"model":"deepseek-v4-flash","headline":"The spectrum of every self-adjoint extension of a discrete symplectic system in the limit point case is completely characterized by the limiting Weyl–Titchmarsh matrix function $M_+(\\lambda)$.","keywords":["Discrete symplectic system","spectrum","eigenvalue","limit point case","M(λ)-function","Weyl–Titchmarsh theory","linear relation","Green's function"],"falsifier":"Take a concrete limit-point discrete symplectic system for which the truncated Weyl functions $M_N$ can be evaluated, choose a real $\\lambda_0$ where the limiting spectral function $\\tau$ has a jump, and compute $L=\\lim_{\\nu\\to 0}\\nu\\,M_+(\\lambda_0+i\\nu)$ both analytically and numerically. Then independently solve the boundary-value problem for a square-summable solution of $(S_{\\lambda_0})$ with $\\alpha\\,z_0=0$. If $L\\neq 0$ but no such nonzero solution exists, or if $L=0$ while one does exist, then Theorem 1.1(iii)–(iv) is wrong; agreement on several such systems would corroborate the $M_+$-dictionary.","tokens_in":24733,"feed_emoji":"🧮","tokens_out":8770,"duration_ms":73084,"temperature":0.7,"pith_summary":"Discrete symplectic systems cover the standard difference equations of mechanics and spectral theory, including even-order Sturm–Liouville difference equations and linear Hamiltonian difference systems. This paper proves that, when such a system is in the limit point case and satisfies the strong Atkinson condition, the spectrum of every self-adjoint extension is completely determined by the limiting Weyl–Titchmarsh matrix function $M_+(\\lambda)$: resolvent points are exactly the real points where $M_+$ is holomorphic, isolated eigenvalues are exactly its simple poles, and the two parts of the essential spectrum are separated by the size of the limit $\\nu\\,M_+(\\lambda_0+i\\nu)$ as $\\nu\\to 0$. This gives an explicit Green-function formula for the resolvent and, through the pole structure, a quantitative description of how the discrete spectrum moves when the boundary condition changes. If correct, it reduces spectral analysis of a large class of infinite difference systems to the study of one matrix-valued Herglotz function.","feed_headline":"One function M+ encodes the whole spectrum","feed_subtitle":"Resolvent, isolated eigenvalues, and continuous spectrum all reduce to the boundary behavior of the limiting Weyl–Titchmarsh function.","key_machinery":"The load-bearing object is the limiting Weyl–Titchmarsh function $M_+(\\lambda)$, an $n\\times n$ matrix Herglotz function obtained as the limit of the regular Weyl functions $M_N(\\lambda,\\alpha,\\beta)$ as the right endpoint $N$ tends to infinity. In the limit point case the nested Weyl disks collapse to a single point, so $M_+$ is independent of the auxiliary $\\beta$, and the columns of the Weyl solution $X_+(\\lambda)=\\hat{Z}(\\lambda)+\\tilde{Z}(\\lambda)M_+(\\lambda)$ span all square-summable solutions. The strong Atkinson condition makes the derivative $M'_+(\\lambda)$ strictly positive on intervals where $M_+$ is holomorphic, and the extended Lagrange identity connects boundary values of solutions to $\\Psi$-inner products; together with the matrix Green function $G_{kj}$, these tools convert questions about the resolvent of the linear relation $T_{LP}$ into questions about the singularities of $M_+$ and the jumps of its representing spectral function $\\tau$.","core_discovery":"The central claim can be stated as a dictionary. Write $T_{LP}(\\alpha)$ for the self-adjoint extension selected by the boundary condition $\\alpha\\,z_0=0$ with $\\alpha\\in\\Gamma$. Theorem 1.1 says that $\\lambda_0$ lies in the resolvent set of $T_{LP}(\\alpha)$ if and only if $M_+(\\lambda)$ is holomorphic at $\\lambda_0$, and then the resolvent is represented by the Green matrix built from the Weyl solution $X_+$; $\\lambda_0$ is an isolated eigenvalue if and only if $M_+$ has a simple pole there, with residue $K_{-1}$ equal to minus the jump of the limiting spectral function $\\tau$ at $\\lambda_0$ and with the columns of $\\tilde{Z}(\\lambda_0)K_{-1}$ giving the eigenfunctions; $\\lambda_0$ belongs to the point-continuous spectrum if and only if $M_+$ is nonholomorphic at $\\lambda_0$, the limit $L=\\lim_{\\nu\\to 0}\\nu\\,M_+(\\lambda_0+i\\nu)$ is nonzero, and $M_+-iL(\\lambda-\\lambda_0)^{-1}$ is still nonholomorphic; $\\lambda_0$ belongs to the continuous spectrum if and only if $M_+$ is nonholomorphic and that same limit is zero. In other words, the whole real spectrum, including its fine structure, is read off from the singularity type of $M_+$ on the real line.","pith_inferences":["The classification suggests a numerical route to spectra of discrete symplectic systems: compute truncated $M_N$ over large $N$, locate poles, and evaluate $\\nu\\,M_N(\\lambda_0+i\\nu)$; the theorem predicts that these finite-data objects stabilize exactly to the spectral decomposition, but this route is not tested in the paper.","Because $M_+$ encodes eigenfunctions through $\\tilde{Z}(\\lambda_0)$ times the residue or limit matrix, the same formalism could export spectral measures to eigenfunction-expansion formulas for singular systems, extending the expansion theory developed for regular systems; that step is implicit rather than proved here.","The interlacing result for $n=1$ is the discrete analogue of classical Sturm eigenvalue interlacing under boundary-condition changes, and it may connect to discrete oscillation counts and spectral flow along paths of self-adjoint extensions, but the paper does not develop that link.","Relaxing the strong Atkinson condition would likely enlarge the class of systems covered; the paper notes that a weaker version suffices for parts of the auxiliary analysis, but whether the full four-way classification survives is left open."],"forward_implications":["If the theorem is right, pure discrete spectrum of $T_{LP}(\\alpha)$ is equivalent to $M_+$ being meromorphic on $\\mathbb{C}$, because then every real singularity is an isolated pole.","The essential spectrum is independent of the boundary condition $\\alpha$, while the resolvent set and isolated eigenvalues can be exchanged; changing $\\alpha$ introduces at most $\\operatorname{rank}(\\alpha J\\hat{\\alpha}^*)$ isolated eigenvalues on any resolvent interval.","For the scalar case $n=1$, between two consecutive eigenvalues of one boundary condition with no essential spectrum in between there is exactly one eigenvalue of any other boundary condition, giving an interlacing theorem.","Boundary conditions with $\\alpha J\\hat{\\alpha}^*=0$ give the same resolvent set, so the spectral dependence on $\\alpha$ is governed by the overlap matrix $\\alpha J\\hat{\\alpha}^*$.","The resolvent formula gives a constructive Green-function representation, so the theorem supplies a way to write the resolvent kernel explicitly whenever $M_+$ is known."],"supporting_citations":[{"why":"Supplies the linear-relation framework, the Green function $G_{kj}$, and the nonhomogeneous boundary-value representation used to write the resolvent.","marker":"[7]"},{"why":"Establishes the Weyl disks and their collapse, defines $M_+$ via the limit of $M_N$, and provides the formula linking $M_+$ for different boundary conditions.","marker":"[21]"},{"why":"Provides the Riemann–Stieltjes representation of regular $M_N$, the spectral function $\\tau$, and the Helly-type convergence that yields the limiting spectral function.","marker":"[28]"},{"why":"The continuous Hamiltonian-system prototype whose pole-versus-jump and resolvent arguments are adapted here to discrete systems and linear relations.","marker":"[14]"},{"why":"Characterizes self-adjoint extensions of the minimal relation by boundary conditions $\\alpha\\in\\Gamma$, so the spectrum of $T_{LP}(\\alpha)$ is the object studied.","marker":"[30]"},{"why":"Supplies abstract linear-relation spectral theory and the Nevanlinna-function pole and jump facts used to translate singularities of $M_+$ into spectral data.","marker":"[3]"}],"fun_headline_variants":["Singularities of M+ map out the whole spectrum","M+ singularities decode spectral fine structure","Spectrum read from singularity type of M+","One function's singularities dictate full spectrum"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole classification rests on the strong Atkinson condition, which says that every nontrivial solution has strictly positive $\\Psi$-weight on some fixed initial interval; if that positivity fails, the uniqueness of representatives, the strict positivity $M'_+>0$, and with them the Green-function and pole-residue arguments are no longer justified.","fun_headline_variants_meta":{"raw":{"variants":["Singularities of M+ map out the whole spectrum","M+ singularities decode spectral fine structure","Spectrum read from singularity type of M+","One function's singularities dictate full spectrum"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000172,"raw_usage":{"total_tokens":1264,"prompt_tokens":926,"completion_tokens":338,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":542,"completion_tokens_details":{"reasoning_tokens":279}},"tokens_in":542,"tokens_out":338,"duration_ms":3051,"temperature":1.0,"reasoning_tokens":279,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T10:18:28.387109+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a concrete limit-point discrete symplectic system for which the truncated Weyl functions $M_N$ can be evaluated, choose a real $\\lambda_0$ where the limiting spectral function $\\tau$ has a jump, and compute $L=\\lim_{\\nu\\to 0}\\nu\\,M_+(\\lambda_0+i\\nu)$ both analytically and numerically. Then independently solve the boundary-value problem for a square-summable solution of $(S_{\\lambda_0})$ with $\\alpha\\,z_0=0$. If $L\\neq 0$ but no such nonzero solution exists, or if $L=0$ while one does exist, then Theorem 1.1(iii)–(iv) is wrong; agreement on several such systems would corroborate the $M_+$-dictionary.","supporting_citations":[{"cited_title":"Zemánek, Eigenfunctions expansion for discrete symplectic systems with general linear de- pendence on spectral parameter , J","cited_arxiv_id":null,"evidence_quote":"Provides the Riemann–Stieltjes representation of regular $M_N$, the spectral function $\\tau$, and the Helly-type convergence that yields the limiting spectral function."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The continuous Hamiltonian-system prototype whose pole-versus-jump and resolvent arguments are adapted here to discrete systems and linear relations."},{"cited_title":"Behrndt, S","cited_arxiv_id":null,"evidence_quote":"Supplies abstract linear-relation spectral theory and the Nevanlinna-function pole and jump facts used to translate singularities of $M_+$ into spectral data."}],"review_version":1}