Pith. sign in

REVIEW 2 major objections 4 minor 21 references

A Riemann-Hilbert Approach to Asymptotic Analysis of Toeplitz+Hankel Determinants

T0 review · 2 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read A 4×4 Riemann–Hilbert problem computes large-$n$ asymptotics of Toeplitz+Hankel determinants whose Toeplitz and Hankel symbols are not assumed related.

desk verdict A genuinely new 4x4 Riemann–Hilbert framework for Toeplitz+Hankel determinants with independent symbols, with a main theorem that is honestly conditional on a non-degeneracy hypothesis; deserves a serious referee. read the letter →

arxiv 1909.00963 v6 pith:6BKSLT5S submitted 2019-09-03 math-ph math.MP

classification math-phmath.MP MSC 15B0530E1535Q15
keywords Toeplitz+HankeldeterminantsRiemann-HilbertproblemnonlinearsteepestdescentorthogonalpolynomialsasymptoticanalysisSzegő-typesymbolsunitcircleHankelmatrices
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper develops a 4×4 Riemann–Hilbert method for the large-$n$ asymptotics of determinants of sums of Toeplitz and Hankel matrices, with no a priori relation between the Toeplitz symbol $\varphi$ and the Hankel symbol $w$. The main theorem gives the ratio of consecutive determinants, $h_{n-1}=D_n/D_{n-1}$, as $-\alpha(0)E(n)/E(n-1)$ with an exponentially small relative error, where $E(n)$ is an explicit combination of contour integrals and $\alpha$ is the Szegő function of $\varphi$. The same model problem emerges when the Hankel weight is supported on an interval inside $(0,1)$, so the analysis covers both geometries. This matters because such determinants control norms of associated orthogonal polynomials and arise, for instance, in the characteristic polynomials of Hankel matrices.

What carries the argument

The load-bearing object is a 4×4 Riemann–Hilbert problem (the $X$-problem) whose jump matrix is an ordinary multiplicative jump; it is obtained by doubling a 2×2 problem with the Carleman shift $z\mapsto z^{-1}$. Nonlinear steepest descent reduces it to a model Riemann–Hilbert problem on the unit circle. For the solvable class, the condition $d\tilde d=1$ makes one entry of the model jump vanish, and the jump is factorized using the Szegő functions $\alpha,\beta$ of $\varphi$ and $d$, yielding an explicit model solution $\Lambda$. The error is controlled by a small-norm problem for $R=S\Lambda^{-1}$, whose jump is exponentially close to the identity; the leading terms $R_{1,23}$ and $R_{1,43}$ combine into the functional $E(n)$ appearing in the determinant-ratio formula.

What would settle it

Take the explicit family (4.38)–(4.39) with a parameter choice where the constant $\kappa$ in (4.41) vanishes, so that $E(n)$ violates condition (1.9), and compute $h_{n-1}=D_n/D_{n-1}$ numerically to high precision for large $n$; if the asymptotic $-\alpha(0)E(n)/E(n-1)$ still holds, the non-degeneracy condition is not necessary, while a breakdown would confirm it is load-bearing.

Watch

Extended reading notes

Core claim

The paper's central claim, Theorem 1.1, is that for smooth nonvanishing Szegő-type symbols $\varphi$ and $w=d\varphi$ on the unit circle, with $d\tilde d=1$ on the circle, the determinant ratio obeys $$h_{n-1}=-\$\alpha$(0)\frac{E(n)}{E(n-1)}\left(1+O($e^{{-c_1 n}}$)\right),\qquad n\to\infty,$$ where $E(n)=2\alpha(0)R_{1,43}(0;n)-C_\rho(0)R_{1,23}(0;n)$ is built from two exponentially small contour integrals and $c_1>0$. The proof reconstructs the original 2×2 orthogonal-polynomial problem from the 4×4 problem, solves the model problem explicitly for this symbol class, and controls the correction by a small-norm Riemann–Hilbert analysis. As a consequence the paper also obtains large-$n$ asymptotics of the orthogonal polynomials themselves, and for an explicit family shows $D_n\sim C(-b_1)^n n^{\alpha_1-1}$, an oscillatory behavior the authors report as confirmed numerically.

Load-bearing premise

The proof requires the exponentially small functional $E(n)$ to stay bounded below by $C r^n$ for some $r<1$ as $n\to\infty$; if $E(n)$ decays faster than every such bound or has zeros for large $n$, the reconstruction of the 2×2 problem from the 4×4 problem is not justified.

Editorial extensions

If this is right

  • For every symbol pair satisfying Theorem 1.1, $D_n\neq0$ for all sufficiently large $n$, and the norm $h_{n-1}$ is known up to exponentially small relative error.
  • Because the same model problem governs the interval-supported Hankel case, the method transfers to $D_n(\varphi,w;1,s)$ whenever the induced function $d=-\varphi^{-1}\tilde u$ satisfies $d\tilde d=1$ on the unit circle.
  • For the explicit family (4.38)–(4.39), the determinant has the oscillatory asymptotic $D_n\sim C(-b_1)^n n^{\alpha_1-1}$, with the constant $C$ still to be determined.
  • The asymptotics in Remark 4.1 give explicit expressions for the orthogonal polynomials $P_n$ inside, on, and outside the unit circle, uniformly in $z$.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the non-degeneracy condition (1.9) is generic, as the explicit example suggests, the functional $E(n)$ may be the universal object controlling smooth Toeplitz+Hankel asymptotics; testing other symbol families would reveal how far that universality reaches.
  • A direct numerical check of the explicit family should reproduce $-b_1(n/(n-1))^{\alpha_1-1}$ up to $O(n^{-2})$ at moderate $n$; this would independently verify the error terms in (4.37).
  • Applying the same 4×4 problem to $D_n(-\lambda,w;0,0)$ would give Hankel eigenvalue asymptotics, a case the paper identifies as simpler in symbol but harder because the model jump no longer simplifies.
  • Because the method drops the condition $d(\pm1)=1$ required by the operator-theoretic approach, overlapping cases provide a built-in consistency test: the two routes must agree wherever both apply.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 4 minor

Summary. The paper develops a 4x4 Riemann-Hilbert formalism for Toeplitz+Hankel determinants in which the Toeplitz and Hankel symbols are not assumed to be related. For the case where the Hankel symbol is supported on the unit circle, the authors carry out a steepest descent analysis and reduce the problem to a model Riemann-Hilbert problem. For the special class w = dφ with d satisfying d(z)d(z^{-1})=1, they construct the model solution explicitly and derive a conditional asymptotic formula for the ratio h_{n-1}=D_n/D_{n-1}, namely (1.10), under the non-degeneracy hypothesis (1.9). They also show that the same model problem emerges when the Hankel symbol is supported on an interval [a,b], and they provide a concrete one-parameter family (4.38)-(4.39) for which the non-degeneracy condition is verified. Section 5 lists open questions, including the characterization of the classes C and C0.

Significance. If the proof is completed, this is a substantial contribution to the asymptotic theory of Toeplitz+Hankel determinants. The 4x4 framework is new in allowing unrelated Toeplitz and Hankel symbols, it connects the unit-circle and interval-supported Hankel cases through the same model problem, and it yields large-n asymptotics for the associated orthogonal polynomials as well as for the norm parameter h_n. The paper is honest about the conditional nature of the main result: Theorem 1.1 explicitly depends on (1.9), and Section 5.5 acknowledges that the class of symbols satisfying this condition is not characterized. The worked example (4.38)-(4.41) providing a nonempty subclass C0 is a genuine strength, as is the comparison with the operator-theoretic results of Basor and Ehrhardt. No parameter fitting or circular reasoning is involved: the asymptotics are expressed through the explicit functional E(n) in (1.6). However, the proof of the final ratio formula contains an index discrepancy that must be repaired before the theorem is established as stated.

major comments (2)
  1. [Section 4.2, Eqs. (4.31) and (4.37)] There is a load-bearing index error in the passage from (4.36) to (4.37). With the definitions as printed, (4.12) gives R_{1,jk}(0;n) = (1/2πi)∫_{Γ'_i} μ^n g_{jk}(μ) dμ for jk=12,14,23,43, while (4.31) defines R^{(1)}_{jk}(n) = (1/2πi)∫_{Γ'_i} μ^{n-2} g_{jk}(μ) dμ for the same index set. Hence the correct relation is R_{1,jk}(0;n) = R^{(1)}_{jk}(n+2), not R_{1,jk}(0;n) = R^{(1)}_{jk}(n+1) as stated in the text. Consequently the denominator in (4.36) equals E(n-2) under the displayed definitions, not E(n-1), and the identity (4.37), which is used to prove formula (1.10) of Theorem 1.1, does not follow. If the intended definition in (4.31) is μ^{n-1} on Γ'_i, the exponent should be corrected consistently; otherwise the derivation leading to (4.37) needs to be repaired. This issue also propagates to the explicit asymptotics stated after (4.41), which rely on the ratio E(n)/E(n-1).
  2. [Section 1, Theorem 1.1; Section 5.5] The non-degeneracy hypothesis (1.9) is essential to the proof: it is used in (4.19)-(4.21) to guarantee condition (4.18) of Lemma 2.7, which in turn is needed to reconstruct the Y-RHP from the X-RHP. The hypothesis is verified only for the explicit family (4.38)-(4.39), and Section 5.5 explicitly leaves open the characterization of the classes C and C0. This is not an internal inconsistency, since Theorem 1.1 is honestly conditional, but the advertised scope of the asymptotic result is accordingly narrower than a first reading of the abstract suggests. I would ask the authors to state prominently that Theorem 1.1 applies subject to (1.9), and that beyond the explicit family the condition is verified only in particular cases; this is a scope clarification rather than a request for new theorems.
minor comments (4)
  1. [Section 4.2, after Eq. (4.41)] The sentence "This fact is confirmed numerically" is not accompanied by any numerical data, figure, or description of the computation; either provide the numerical verification or remove the sentence.
  2. [Abstract and Introduction] The abstract's phrase "This in turn will allow us to find the asymptotics" could be read as unconditional; it would be helpful to mention explicitly that the asymptotic theorem requires the non-degeneracy condition (1.9), whose validity for the general class is presently open.
  3. [Remark 4.2, Eq. (4.43)] The sufficiency condition (4.43) for the interval-supported case is stated without derivation; a brief indication of where this condition comes from, or a reference to a future work, would improve readability.
  4. [Throughout] There are occasional spelling and formatting slips (for example, "offset" for "offset" and some inconsistent spacing in displayed equations) that should be corrected in copy-editing.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the asymptotic ratio in Theorem 1.1 is expressed through an explicitly defined symbol functional E(n), derived self-contained via the 4x4 RHP, with no fitted input called a prediction.

full rationale

The derivation chain runs: Y-RHP -> doubled 4x4 X-RHP -> lens openings and normalizations -> global parametrix -> explicit model solution (4.6) -> small-norm R-RHP -> reconstruction (2.49) -> P(n) asymptotics (4.17) -> condition (4.18) -> formula (4.37). The object E(n) is defined in (1.6) as a combination of contour integrals (R1,23, R1,43) with integrands g23, g43 in (1.5), all built directly from the Szego data phi and d; it is not defined in terms of h_n or D_n, and no parameter is fitted to the determinant to make (1.10) hold. The non-degeneracy hypothesis (1.9) is an explicit condition on that symbol functional; the authors verify it only for the family (4.38)-(4.39) and state in Section 5.5 that characterizing the class C0 is open. That is an acknowledged scope limitation, not a circular reduction. Self-citations to [13] appear as historical or technical references (Remark 1.3, Section 5.4), but Theorem 1.1's proof does not rely on an unproved assertion from [13] for its central step; the model problem is solved in the paper and the final estimate (4.37) follows from the RHP transformations. Hence no circular step is exhibited.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central claim depends on explicit structural assumptions on the symbols and on a non-degeneracy bound on the error functional E(n). No free parameters are fitted to data; the result is a conditional theorem with a verified example.

assumptions (5)
  • domain assumption The symbols phi and d are analytic in a neighborhood of the unit circle and have zero winding number (Szego-type).
    Needed for the Szego function construction and for the primary lens opening in Sections 2 and 4.
  • domain assumption The ratio d satisfies d(z)d(z^{-1}) = 1 on the unit circle.
    This makes the 23-element of the model jump matrix vanish (Section 4, after equation (4.2)), enabling the explicit factorization and solution of the model problem.
  • domain assumption The non-degeneracy condition |E(n)| >= C r^n for some r in [r0,1) holds (inequality (1.9)).
    Required in Theorem 1.1 to guarantee that condition (2.36) holds for n and n-1, which lets the Y-RHP be reconstructed from the X-RHP; not proven for the whole class, only for the example family in Section 4.2.
  • domain assumption For the interval case, w is supported on [a,b] with 0<a<b<1 and has no Fisher-Hartwig singularities.
    Assumed in Section 3 to define the orthogonality relations and to avoid local parametrices.
  • standard math Standard results on small-norm Riemann-Hilbert problems, Plemelj-Sokhotskii formulae, Cauchy integrals, and Watson's lemma are used as black boxes.
    Invoked in Sections 2 through 4; these are established theorems in the literature.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Riemann-Hilbert Approach to Asymptotic Analysis of Toeplitz+Hankel Determinants." pith.science (2026). https://pith.science/paper/6BKSLT5S

@misc{pith2026190900963,
  author       = {Pith},
  title        = {Pith review of: A Riemann-Hilbert Approach to Asymptotic Analysis of Toeplitz+Hankel Determinants},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6BKSLT5S}},
  note         = {Machine review of arXiv:1909.00963}
}
abstract

In this paper we will formulate $4\times4$ Riemann-Hilbert problems for Toeplitz+Hankel determinants and the associated system of orthogonal polynomials, when the Hankel symbol is supported on the unit circle and also when it is supported on an interval $[a,b]$, $0<a<b<1$. The distinguishing feature of this work is that in the formulation of the Riemann-Hilbert problem no specific relationship is assumed between the Toeplitz and Hankel symbols. We will develop nonlinear steepest descent methods for analysing these problems in the case where the symbols are smooth (i.e., in the absence of Fisher-Hartwig singularities) and admit an analytic continuation in a neighborhood of the unit circle (if the symbol's support is the unit circle). We will finally introduce a model problem and will present its solution requiring certain conditions on the ratio of Hankel and Toeplitz symbols. This in turn will allow us to find the asymptotics of the norms $h_n$ of the corresponding orthogonal polynomials and, in fact, the large $n$ asymptotics of the polynomials themselves. We will explain how this solvable case is related to the recent operator-theoretic approach in [Basor E., Ehrhardt T., in Large Truncated Toeplitz Matrices, Toeplitz Operators, and Related Topics, Oper. Theory Adv. Appl., Vol. 259, Birkh\"auser/Springer, Cham, 2017, 125-154, arXiv:1603.00506] to Toeplitz+Hankel determinants. At the end we will discuss the prospects of future work and outline several technical, as well as conceptual, issues which we are going to address next within the $4\times 4$ Riemann-Hilbert framework introduced in this paper.

Figures

Figures reproduced from arXiv: 1909.00963 by the authors.

Figure 1
Figure 1. The jump contour Γ for the Z, T and the global parametrix Riemann–Hilbert problems. • RH-Z2 Z+(z; n) = Z−(z; n)JZ(z), where JZ(z) =    JX,T(z), z ∈ T, JX,i(z), z ∈ Γi , JX,o(z), z ∈ Γo. • RH-Z3 As z → ∞ we have Z(z; n) = I + O [PITH_FULL_IMAGE:figures/full_fig_p021_1.png] view at source ↗
Figure 2
Figure 2. The jump contour ΓS of the S-RHP. In the usual way, we will first try to solve this Riemann–Hilbert problem by disregarding the jump matrices which depend on n, this solution is denoted by ◦ S and will be referred to as the global parametrix. Once we construct the global parametrix, we will consider the small￾norm Riemann–Hilbert problem for the ratio R := S( ◦ S) −1 and discuss its solvability in the forthcoming se… view at source ↗
Figure 3
Figure 3. The jump contour Σ. We omit the proof here as it is similar to the proof of Theorem 2.1. Corollary 3.2. Suppose that the Y -RH problem has a unique solution for n and n − 1. Then Dn 6= 0, Dn−1 6= 0, and hn−1 6= 0. Moreover, hn−1 = − limz→∞ z n−1 /Y21(z; n). 3.1 The associated 2 × 4 and 4 × 4 Riemann–Hilbert problems The formulation of the 2 × 4 and 4 × 4 Riemann–Hilbert problems are very similar to those of Section … view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: The jump contour ΣS ≡ Σ ∪ Σo ∪ Σi of the S-RHP. • RH-T1 T is holomorphic in C \ Σ. • RH-T2 For z ∈ Σ 0 , we have T+(z; n) = T−(z; n)JT (z; n), where JT (z; n) = ( Jb(z; n), z ∈ T, JX(z), z ∈ (a, b) ∪ [PITH_FULL_IMAGE:figures/full_fig_p031_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

21 extracted references · 17 canonical work pages

  1. [2]

    Asymptotic formulas for determinants of a special class of Toeplitz + Hankel matrices

    Basor E., Ehrhardt T., Asymptotic formulas for determinants of a special class of Toeplitz + Hankel matrices, in Large Truncated Toeplitz Matrices, Toeplitz Operators, and Related Topics, Oper. Theory Adv. Appl. , Vol. 259, Birkh¨ auser/Springer, Cham, 2017, 125–154, arXiv:1603.00506

  2. [13]

    Deift P., Its A., Krasovsky I., Asymptotics of Toeplitz, Hankel, and Toeplitz+Hankel determinants with Fisher–Hartwig singularities, Ann. of Math. 174 (2011), 1243–1299, arXiv:0905.0443

  3. [14]

    Deift P., Its A., Krasovsky I., Eigenvalues of Toeplitz matrices in the bulk of the spectrum, Bull. Inst. Math. Acad. Sin. (N.S.) 7 (2012), 437–461, arXiv:1110.4089

  4. [1]

    Baik J., Deift P., Johansson K., On the distribution of the length of the longest increasing subsequence of random permutations, J. Amer. Math. Soc. 12 (1999), 1119–1178, arXiv:math.CO/9810105

  5. [3]

    Basor E.L., Ehrhardt T., Asymptotic formulas for determinants of a sum of finite Toeplitz and Hankel matrices, Math. Nachr. 228 (2001), 5–45, arXiv:math.FA/9809088

  6. [4]

    Theory Adv

    Basor E.L., Ehrhardt T., Asymptotic formulas for the determinants of symmetric Toeplitz plus Hankel matrices, in Toeplitz Matrices and Singular Integral Equations (Pobershau, 2001), Oper. Theory Adv. Appl. , Vol. 135, Birkh¨ auser, Basel, 2002, 61–90, arXiv:math.FA/0202078

  7. [5]

    Matrices 3 (2009), 167–186, arXiv:0804.3073

    Basor E.L., Ehrhardt T., Determinant computations for some classes of Toeplitz–Hankel matrices, Oper. Matrices 3 (2009), 167–186, arXiv:0804.3073

  8. [6]

    Basor E.L., Ehrhardt T., Fredholm and invertibility theory for a special class of Toeplitz + Hankel operators, J. Spectr. Theory 3 (2013), 171–214, arXiv:1110.0577

Show all 21 references
  1. [7]

    B¨ ottcher A., Grudsky S.M., Spectral properties of banded Toeplitz matrices, Society for Industrial and Applied Mathematics (SIAM), Philadelphia, PA, 2005

  2. [8]

    B¨ ottcher A., Silbermann B., Introduction to large truncated Toeplitz matrices,Universitext, Springer-Verlag, New York, 1999

  3. [9]

    B¨ ottcher A., Silbermann B., Analysis of Toeplitz operators, 2nd ed., Springer Monographs in Mathematics , Springer-Verlag, Berlin, 2006

  4. [10]

    Charlier C., Asymptotics of Hankel determinants with a one-cut regular potential and Fisher–Hartwig singularities, Int. Math. Res. Not. 2019 (2019), 7515–7576, arXiv:1706.03579

  5. [11]

    Charlier C., Gharakhloo R., Asymptotics of Hankel determinants with a Laguerre-type or Jacobi-type potential and Fisher–Hartwig singularities, arXiv:1902.08162. 6The very recent work [18] shows that indeed one can explicitly describe the leading terms of the asymptotics of Dn(...

  6. [12]

    Chelkak D., Hongler C., Mahfouf R., Magnetization in the zig-zag layered Ising model and orthogonal polynomials, arXiv:1904.09168

  7. [15]

    Pure Appl

    Deift P., Its A., Krasovsky I., Toeplitz matrices and Toeplitz determinants under the impetus of the Ising model: some history and some recent results, Comm. Pure Appl. Math. 66 (2013), 1360–1438, arXiv:1207.4990

  8. [16]

    Pure Appl

    Deift P., Kriecherbauer T., McLaughlin K.T.-R., Venakides S., Zhou X., Strong asymptotics of orthogonal polynomials with respect to exponential weights, Comm. Pure Appl. Math. 52 (1999), 1491–1552

  9. [17]

    Pure Appl

    Deift P., Kriecherbauer T., McLaughlin K.T.-R., Venakides S., Zhou X., Uniform asymptotics for polyno- mials orthogonal with respect to varying exponential weights and applications to universality questions in random matrix theory, Comm. Pure Appl. Math. 52 (1999), 1335–1425

  10. [18]

    Fedele E., Gebert M., On determinants identity minus Hankel matrix, Bull. Lond. Math. Soc. 51 (2019), 751–764, arXiv:1808.08009

  11. [19]

    Its A., Krasovsky I., Hankel determinant and orthogonal polynomials for the Gaussian weight with a jump, in Integrable Systems and Random Matrices, Contemp. Math. , Vol. 458, Amer. Math. Soc., Providence, RI, 2008, 215–247, arXiv:0706.3192

  12. [20]

    Krasovsky I., Correlations of the characteristic polynomials in the Gaussian unitary ensemble or a singular Hankel determinant, Duke Math. J. 139 (2007), 581–619, arXiv:math-ph/0411016

  13. [21]

    Probab., Vol

    Krasovsky I., Aspects of Toeplitz determinants, in Random Walks, Boundaries and Spectra, Progr. Probab., Vol. 64, Birkh¨ auser/Springer Basel AG, Basel, 2011, 305–324, arXiv:1007.1128

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.