REVIEW 2 major objections 4 minor 23 references
Operators which are polynomially isometric to a normal operator
T0 review · 2 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash
Pith's one-line read This paper shows that if two operators have matching norms for every polynomial image, and one is normal, then the other is forced to be normal in the matrix setting and in the compact-operator (c,p)-norm setting; for the…
desk verdict Sections 2 and 3 are strong and publishable, but the compact (c,p)-norm theorem in Section 4 has a genuine proof gap that needs repair. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The key object is the comparison of Riesz spectral projections P_i=E_N(Δ_i) and Q_i=E_A(Δ_i) at the same spectral sets. Polynomials that isolate individual spectral subsets give ‖P_i‖=‖Q_i‖ and ‖I−P_i‖=‖I−Q_i‖. Lemma 2.3 — if a projection and an idempotent have the same norm, and their complements have the same norm, then the idempotent is a projection — converts this data into selfadjointness of Q_i. In the operator-norm case, the Lavrentieff hypothesis makes the unital algebra generated by N equal to its C*-algebra, so the polynomial isometry extends to a *-isomorphism; when the hypothesis fails, a unilateral shift summand supplies the non-normal partner. In the compact (c,p)-norm case, finite-dimensional cutoffs H_k constructed from these Riesz projections reduce the problem to the matrix theorem.
What would settle it
For a concrete test, take N=diag(λ_1,λ_2,0,...) and A=N+E_12 in a (c,p)-norm; if some polynomial p satisfies ‖p(A)‖_{c,p}=‖p(N)‖_{c,p} while A is nonnormal, Theorem 4.3 is false. Short of that, computing the singular values of R_k−P_k and R_k−Q_k for such a pair at one spectral cutoff settles whether the proof's operative rank assumption is valid.
Extended reading notes
Core claim
The central discovery is that polynomial isometry to a normal operator is a rigid constraint. In M_n(C), for any unitarily-invariant norm, if ‖p(A)‖_u=‖p(N)‖_u for all polynomials and N is normal, then A is normal; if the norm separates projections by rank, A and N are unitarily similar (Theorem 2.4). In B(H) with the operator norm, normality transfers exactly when σ(N) is Lavrentieff; otherwise a non-normal operator B polynomially isometric to N always exists (Theorem 3.6). For compact A and N with N normal, the same conclusion holds under any (c,p)-norm (Theorem 4.3). The proofs identify the polynomial class with the full C*-algebra generated by N in the Lavrentieff case, and use Riesz projections plus a finite-dimensional lemma to force the partner's spectral idempotents to be selfadjoint in the compact case.
Load-bearing premise
The compact-operator argument hinges on an unshown rank-counting step: the auxiliary finite-rank projection R_k must differ from the Riesz projections P_k and Q_k in such a way that the first n singular values of each difference are all 1; if that count is wrong, Lemma 2.3 cannot be invoked.
Editorial extensions
If this is right
- In any finite-dimensional setting, a normal matrix is completely pinned down, up to unitary similarity, by the polynomial norm behaviour of its spectral projections whenever the norm used can tell projections of different ranks apart.
- For the operator norm, the line between normality forcing and non-normality allowing is drawn precisely by Lavrentieff spectrum: spectra with interior or with more than one complementary component admit non-normal partners.
- The compact (c,p)-norm theorem means that for Schatten-type and Ky Fan type norms, polynomial isometry to a normal compact operator is a genuine normality certificate, even though the analogous statement fails for general bounded operators under the operator norm.
- When only polynomials vanishing at zero are compared, the conclusion fails in general; however, for invertible matrices on finite-dimensional spaces, vanishing-at-zero polynomial data still force normality and unitary similarity.
Reading between the lines
- The finite-dimensional result suggests that in any unitarily invariant norm that separates projections by rank, polynomial isometry to a normal element should be a complete unitary-invariance invariant; one could test this in von Neumann algebra settings where a version of Lemma 2.3 may hold.
- The Lavrentieff dichotomy for the operator norm invites the question of whether other unitarily invariant ideals, such as Schatten p-classes with p<∞, have their own spectral-geometry threshold; the Section 4 counterexample with the constant term omitted shows the answer will depend delicately on the role of the identity.
- A concrete computational test of Theorem 4.3's 'readily verified' rank assumptions on small compact examples would either close the proof gap or reveal a need for additional hypotheses; that is the natural next step for someone wanting to rely on the compact-operator result.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies the question of whether an operator A that is polynomially isometric to a normal operator N with respect to a unitarily-invariant norm must itself be normal. In the finite-dimensional case the authors prove that this is true for every unitarily-invariant norm on M_n(C) (Theorem 2.4). For the operator norm on B(H) they prove a complete dichotomy: if σ(N) is Lavrentieff, every polynomially isometric A is normal, while if σ(N) is not Lavrentieff, a non-normal polynomially isometric operator exists (Theorem 3.6). For compact operators with respect to the (c,p)-norms of Chan, Li and Tu they claim the same automatic normality (Theorem 4.3). A final section shows that if only polynomials without constant term are used, normality can fail, and gives a positive finite-dimensional result for invertible operators.
Significance. If the results hold, the finite-dimensional theorem for all unitarily-invariant norms is a clean extension of earlier Frobenius- and operator-norm results, and the operator-norm dichotomy in terms of Lavrentieff spectra correctly identifies the role of polynomial approximation. The zero-constant-term counterexample in Section 5 is a useful negative result that delineates why the constant term is essential. The C*-algebra argument in Section 3 is short and conceptually appealing. However, the compact (c,p)-norm theorem has a serious proof gap in the reduction to finite-rank compressions, and until this is repaired the paper's main advertised extension to compact operators is not established.
major comments (2)
- [Section 4.3, proof of Theorem 4.3] The displayed chain showing that A_k and N_k are polynomially isometric proves only the ambient equality ||q(0)I_H + P_k q_1(N)||_{c,p} = ||q(0)I_H + Q_k q_1(A)||_{c,p} in B(H). The following 'Notice also' replaces I_H by R_k and asserts that the norms of the compressions on H_k agree with these ambient norms, but this replacement is not valid. For an operator S supported on H_k, the ambient operator q(0)I_H + S has infinitely many singular values equal to |q(0)| coming from the infinite-dimensional complement H_k^⊥, whereas the norm on B(H_k) is computed from q(0)I_{H_k} + S alone. When |q(0)| is large enough, the ambient norm is identically |q(0)|(Σ c_j)^{1/p} and carries no information about the finite-rank part. Thus equality of ambient norms does not imply that N_k and A_k are polynomially isometric as operators on H_k, and Theorem 2.4 cannot be applied. The finite-spectrum case, described as 'similar', inherits the same problem. A concrete illustration is H_k = C^2 with c_1 = c_2 = 1 and p = 1: for S = diag(-1/2,-1/2) and T = 0, the ambient operators I_H+S and I_H have the same first two singular values, while the norms restricted to H_k are 1 and 2 respectively.
- [Section 4.3, definition of H_k] The assertion that ||R_k - P_k||_{c,p} = ||I - P_k||_{c,p} and ||R_k - Q_k||_{c,p} = ||I - Q_k||_{c,p} is 'readily verified' but actually requires that the complements H_k ⊖ Ran P_k and H_k ⊖ Ran Q_k have dimension at least n, so that the first n singular values are all equal to 1. The definition of H_k via α_k = rank P_k + dim(ker Q_k)^⊥ + rank Q_k + 3k guarantees this only for k with 3k ≥ n, and the proof applies the equalities for every k without this rank verification. This is a repairable technical point, but it is load-bearing for the application of Lemma 2.3 and should be stated and proved explicitly.
minor comments (4)
- [Section 4.3, Case One] In the paragraph defining the limit of the P_k, the text writes 'P = sup{Q_k : k ∈ N} = SOT − lim P_k'; this should be 'sup{P_k : k ∈ N}'.
- [Section 4.3, final norm estimate] In the chain of inequalities near the end of Case One, the expressions 't_j^{(k)}(A)AA' and 't_j^{(k)}(N)NN' appear to contain redundant letters and should read 't_j^{(k)}(A)A' and 't_j^{(k)}(N)N'.
- [Section 4.2, Definition of (c,p)-norm] The definition of the (c,p)-norm does not explicitly state how the singular values s_j(T) are interpreted when dim H < n; the authors should state that the missing singular values are taken to be 0, as is implicitly done in the later use on finite-dimensional spaces.
- [Section 3.6, proof of (b) implies (a)] The step 'Let L = hat{σ(N)}. Then L^o ≠ ∅' is stated without explanation; a sentence justifying why a non-Lavrentieff spectrum forces the polynomially convex hull to have nonempty interior would improve readability.
Circularity Check
No significant circularity: all load-bearing reductions follow from polynomial isometry and independent theorems; the Section 4 gap is a correctness issue, not a circular reduction.
full rationale
The derivation chain is not circular. In Theorem 2.4, the spectral projections P_i are obtained from the normal matrix N via polynomial functional calculus, and the equalities ||P_i||_u=||Q_i||_u and ||I-P_i||_u=||I-Q_i||_u are immediate specializations of the polynomial-isometry hypothesis; Lemma 2.3 then converts these equalities into the conclusion that the Q_i are projections. No displayed equation restates the theorem's input as its output. The infinite-dimensional operator-norm argument (Theorem 3.3) relies on the Lavrentieff characterization Alg(N)=C*(N), cited to [8] (Farenick-Forrest-Marcoux). This is a self-citation, but the cited theorem is an independent, parameter-free result about normal operators whose assumptions do not include polynomial isometry or the target normality conclusion; by Rule 4 it is real evidence and does not raise the circularity score. Section 4.3's compact-(c,p) proof attempts to reduce to Theorem 2.4 via finite-rank compressions. The displayed chain equates q(0)I+P_k q_1(N) with q(0)I+Q_k q_1(A), while the finite-dimensional application needs equality for q(0)R_k+q_1(N_k) versus q(0)R_k+q_1(A_k); and the 'readily verified' rank conditions for ||R_k-P_k||_{c,p}=||I-P_k||_{c,p} are not displayed. That is a proof gap, i.e. a correctness risk, but it is not a circular reduction: no fitted value is renamed as a prediction, and no theorem is defined in terms of the result it is used to prove. Therefore there are no circular steps.
Assumptions & free parameters
assumptions (7)
- standard math Von Neumann correspondence: every unitarily-invariant norm on M_n(C) is given by a symmetric gauge function of singular values.
- standard math A unital contractive algebra homomorphism between unital C*-algebras is automatically a *-homomorphism.
- standard math Lavrentieff's theorem: polynomials are uniformly dense in C(K) iff C \ K is connected and K has empty interior.
- standard math For a subnormal operator S, ||p(S)|| = spr(p(S)) for every polynomial p.
- standard math Mergelyan's theorem: holomorphic functions on a compact set with connected complement are uniformly approximable by polynomials.
- standard math Riesz functional calculus and spectral mapping for compact operators, including finite-rank Riesz idempotents for isolated spectral sets.
- domain assumption Domain assumptions: H is a complex separable Hilbert space; the norms are unitarily invariant, and for the (c,p)-norm theorem, A and N are compact with c1 >= c2 >= ... >= cn > 0 and 1 <= p < infinity.
Cite this review
Pith. "Pith review of Operators which are polynomially isometric to a normal operator." pith.science (2026). https://pith.science/paper/6X3PKL7F
@misc{pith2026190807029,
author = {Pith},
title = {Pith review of: Operators which are polynomially isometric to a normal operator},
year = {2026},
howpublished = {\url{https://pith.science/paper/6X3PKL7F}},
note = {Machine review of arXiv:1908.07029}
}
abstract
Let $\mathcal{H}$ be a complex, separable Hilbert space and $\mathcal{B}(\mathcal{H})$ denote the algebra of all bounded linear operators acting on $\mathcal{H}$. Given a unitarily-invariant norm $\| \cdot \|_u$ on $\mathcal{B}(\mathcal{H})$ and two linear operators $A$ and $B$ in $\mathcal{B}(\mathcal{H})$, we shall say that $A$ and $B$ are \emph{polynomially isometric relative to} $\| \cdot \|_u$ if $\| p(A) \|_u = \| p(B) \|_u$ for all polynomials $p$. In this paper, we examine to what extent an operator $A$ being polynomially isometric to a normal operator $N$ implies that $A$ is itself normal. More explicitly, we first show that if $\| \cdot \|_u$ is any unitarily-invariant norm on $\mathbb{M}_n(\mathbb{C})$, if $A, N \in \mathbb{M}_n(\mathbb{C})$ are polynomially isometric and $N$ is normal, then $A$ is normal. We then extend this result to the infinite-dimensional setting by showing that if $A, N \in \mathcal{B}(\mathcal{H})$ are polynomially isometric relative to the operator norm and $N$ is a normal operator whose spectrum neither disconnects the plane nor has interior, then $A$ is normal, while if the spectrum of $N$ is not of this form, then there always exists a non-normal operator $B$ such that $B$ and $N$ are polynomially isometric. Finally, we show that if $A$ and $N$ are compact operators with $N$ normal, and if $A$ and $N$ are polynomially isometric with respect to the $(c,p)$-norm studied by Chan, Li and Tu, then $A$ is again normal.
Reference graph
Works this paper leans on
-
[8]
D. R. Farenick, B. E. Forrest, and L. W. Marcoux, Amenable operators on Hilbert spaces , J. Reine Angew. Math. 582 (2005), 201–228
work page 2005
-
[1]
D. P. Blecher and C. Le Merdy, Operator algebras and their modules – an operator space appr oach, London Math. Soc. Monographs 30, Clarendon Press, Oxford, 2004
work page 2004
-
[2]
C. D. Brooks and A. A. Condori, A resolvent criterion for normality , Amer. Math. Monthly 125 (2018), no. 2, 149–156
work page 2018
-
[3]
Jor-Ting Chan, Chi-Kwong Li, and Charlies C. N. Tu, A class of unitarily invariant norms on B(H), Proc. Amer. Math. Soc. 129 (2001), no. 4, 1065–1076
work page 2001
-
[4]
J. B. Conway, A course in functional analysis , second ed., Graduate Texts in Mathematics, vol. 96, Springer-Verlag, New York, 1990
work page 1990
-
[5]
K. R. Davidson, Nest algebras , Pitman Research Notes in Mathematics Series, vol. 191, Longman Scientific & Technical, Harlow, 1988, Triangular forms for operator algebras on Hilbert space
work page 1988
-
[6]
L. Elsner and Kh. D. Ikramov, Normal matrices: an update , Linear Algebra Appl. 285 (1998), no. 1-3, 291–303
work page 1998
-
[7]
K. Fan, Maximum properties and inequalities for the eigenvalues of completely continuous operators , Proc. Nat. Acad. Sci., U. S. A. 37 (1951), 760–766
work page 1951
Show all 23 references
-
[9]
Farenick, V
D.R. Farenick, V. Futorny, T. G. Gerasimova, V. V. Sergeichuk, and N. Shvai, A criterion for unitary similarity of upper triangular matrices in general positio n, Linear Algebra Appl. 435 (2011), no. 6, 1356–1369
2011
-
[10]
Farenick, T.G
D.R. Farenick, T.G. Gerasimova, and N. Shvai, A complete unitary similarity invariant for unicellular matrices, Linear Algebra Appl. 435 (2011), no. 2, 409–419
2011
-
[11]
Gerasimova, Unitary similarity to a normal matrix , Linear Algebra Appl
T.G. Gerasimova, Unitary similarity to a normal matrix , Linear Algebra Appl. 436 (2012), no. 9, 3777–3783
2012
-
[12]
Gohberg and M.G
I.C. Gohberg and M.G. Krein, Introduction to the theory of linear nonselfadjoint operat ors, Amer. Math. Soc., Providence, R.I., (1969)
1969
-
[13]
Greenbaum and L.N
A. Greenbaum and L.N. Trefethen, Do the pseudospectra of a matrix determine its behavior? , Technical Report TR 93-1371, Cornell University, 1993. ecommons.cornell.ed u/bitstream/handle/1813/6145/93- 1371.pdf
1993
-
[14]
Grone, C
R. Grone, C. R. Johnson, E. M. Sa, and H. Wolkowicz, Normal matrices , Linear Algebra Appl. 87 (1987), 213–225
1987
-
[15]
C. S. Kubrusly, The elements of operator theory , second ed., Birkh¨ auser/Springer, New York, 2011
2011
-
[16]
Lavrentieff, Sur les fonctions d’une variable complexe repr´ esentables par les series de polynˆ omes, Acta
M.A. Lavrentieff, Sur les fonctions d’une variable complexe repr´ esentables par les series de polynˆ omes, Acta. Sci. Ind. 441 (1936)
1936
-
[17]
Li and N.K Tsing, On the unitarily-invariant norms and some related results , Linear and Multi- linear Algebra 20 (1987), no
C.K. Li and N.K Tsing, On the unitarily-invariant norms and some related results , Linear and Multi- linear Algebra 20 (1987), no. 2, 107–119
1987
-
[18]
Mirsky, Symmetric gauge functions and unitarily-invariant norms , Quart
L. Mirsky, Symmetric gauge functions and unitarily-invariant norms , Quart. J. Math. Oxford Ser. (2) 11 (1960), 50–59
1960
-
[19]
Pietsch, Eigenvalues and S-numbers, Cambridge Studies in Advanced Mathematics 13, Cambridge Univ
A. Pietsch, Eigenvalues and S-numbers, Cambridge Studies in Advanced Mathematics 13, Cambridge Univ. Press, Cambridge, 1987
1987
-
[20]
J. R. Ringrose, Super-diagonal forms for compact linear operators , Proc. London Math. Soc. (3) 12 (1962), 367–384
1962
-
[21]
Rudin, Real and complex analysis , third ed., McGraw-Hill Book Co., New York, 1987
W. Rudin, Real and complex analysis , third ed., McGraw-Hill Book Co., New York, 1987
1987
-
[22]
Viswanath and L.N
D. Viswanath and L.N. Trefethen, Matrix behaviour, unitary reducibility, and Hadamard prod ucts, Tech- nical Report TR96-1596, Cornell University, 1996. ecommons.cor nell.edu/handle/1813/7251
1996
-
[23]
Wolf, On the essential spectrum of partial differential boundary p roblems, Comm
F. Wolf, On the essential spectrum of partial differential boundary p roblems, Comm. Pure Appl. Math. 12 (1959), 211-228. 14 Laurent W. Marcoux, Department of Pure Mathematics, Univer sity of W aterloo, W a- terloo, Ontario N2L 3G1 CANADA E-mail address : Laurent.Marcoux@uwater...
1959
Reviewed August 14, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.