Pith. sign in

REVIEW 3 major objections 4 minor 37 references

Plurisubharmonic Noncommutative Rational Functions

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read A noncommutative rational function is plurisubharmonic exactly when it factors as a convex rational function composed with an analytic rational function, and a two-line projection test on its minimal realization decides the property.

desk verdict A constructive rational-function extension of the plush decomposition theorem; the main risk is a load-bearing spanning lemma that is cited rather than proved. read the letter →

arxiv 1908.01895 v1 pith:G4KP6KGN submitted 2019-08-05 math.FA math.CV

classification math.FAmath.CV MSC 47A5646L0732A9946L89
keywords plurisubharmonicnoncommutativerationalfunctiondescriptorrealizationfreeanalysiscomplexHessianconvexminimalmatrixpositivity
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 establishes a structure theorem for plurisubharmonic (plush) noncommutative rational functions: a symmetric nc rational function regular at 0 is plush near 0 if and only if it is the composition of a convex nc rational function with an analytic rational mapping. This lifts to rational functions the known polynomial fact that plush polynomials split as convex composed with analytic. The factorization is constructive, built from a minimal descriptor realization via angular operators and finitely many intermediate variables. The paper also proves a computable test: plushness holds exactly when the two projected signature blocks $PKP$ and $P_* K P_*$ are positive semidefinite, for the projections onto the ranges of the two coefficient tuples. Testing plushness therefore reduces to checking a pair of matrix inequalities.

What carries the argument

The engine of the paper is the minimal symmetric descriptor realization $r(x,x^*)=c^*(K-\Lambda_B(x)-\Lambda_{B^*}(x^*))^{-1}c$, with signature matrix $K$, together with the two range projections $P$ onto $\operatorname{rng} B$ and $P_*$ onto $\operatorname{rng} B^*$. The central identity is the direct-sum decomposition of the complex Hessian: reading $r$ at $X\oplus \widetilde{X}$ in the block off-diagonal direction $H$ produces $\operatorname{diag}(r_\downarrow(X,\widetilde{X})[H], r_\uparrow(X,\widetilde{X})[H])$, so plushness is equivalent to positivity of both $r_\downarrow$ and $r_\uparrow$ on free balls; these in turn reduce to $PKP\succeq 0$ and $P_* K P_*\succeq 0$. For the converse construction, maximal $K$-nonnegative subspaces containing the two ranges, parameterized by angular operators $\rho,\rho_*$, are used to build the convex factor $f$ and the analytic map $q$; a finite set of intermediate variables suffices once $\{B_1,\dots,B_g\}$ is taken linearly independent.

What would settle it

Compute the complex Hessian numerically for a candidate minimal symmetric descriptor realization with a non-trivial signature $K$, testing the predicted equivalence between positivity of $PKP$ and $P_* K P_*$ and positivity of all Hessian blocks $r_\downarrow(X,Y)[H]$ and $r_\uparrow(X,Y)[H]$ on a dense grid of direct sums $X\oplus Y$ with off-diagonal $H$; a single instance with PSD projected blocks but a negative Hessian block, or a plush $r$ with one projected block indefinite, would refute Theorem 1.4 and the constructive direction of Theorem 1.3.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is Theorem 1.3: for a symmetric nc rational function $r$ in $g$ variables, regular at 0, $r$ is plush in a neighborhood of 0 if and only if there are a positive integer $h$, a convex nc rational function $f$ in $h$ variables, and an analytic nc rational mapping $q$ with $r=f\circ q$. The companion result, Theorem 1.4, gives a realization-level criterion: if $r$ is presented minimally as $c^*(K-\Lambda_B(x)-\Lambda_{B^*}(x^*))^{-1}c$, then $r$ is plush near 0 exactly when $PKP$ and $P_* K P_*$ are positive semidefinite, where $P$ and $P_*$ are the orthogonal projections onto $\operatorname{rng} B$ and $\operatorname{rng} B^*$. The sufficiency direction uses the $K$-nonnegativity of the two ranges to build a pairwise positive decomposition of the complex Hessian; the necessity direction uses the Hessian decomposition together with a spanning lemma securing that every direction in the ranges can be realized by an off-diagonal block. A parallel statement, Theorem 1.6, characterizes convexity near 0 by $QKQ$ being positive semidefinite for $Q$ the projection onto $\operatorname{rng} B+\operatorname{rng} B^*$.

Load-bearing premise

The result assumes the spanning lemma: for a minimal realization one can always find a small block $X$ and vector $v$ for which the $d$ components of $\Delta(X)(c\otimes v)$ are linearly independent; the paper quotes this from prior Hermitian-variable work rather than proving it here.

Editorial extensions

If this is right

  • Plushness of an nc rational function is decidable: minimal realizations are computable, so Theorem 1.4 reduces the question to checking two positive-semidefiniteness conditions.
  • Free pseudoconvex sets defined by plush rational functions admit proper analytic nc rational maps onto convex free domains, connecting plurisubharmonicity to the linear matrix inequality world.
  • Convexity near 0 for symmetric nc rational functions has the same flavor of test: $QKQ\succeq 0$ for the projection $Q$ onto $\operatorname{rng} B+\operatorname{rng} B^*$ (Theorem 1.6), and convexity implies plushness.
  • The class of plush rationals is closed under right composition with analytic rational maps, so the factorization $r=f\circ q$ is not a one-off: every analytic reparameterization of a convex rational factor remains plush.

Reading between the lines

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

  • Beyond the paper: the finite-variable reduction suggests an explicit bound on the intermediate dimension $h$ in terms of the realization size $d$ (roughly $d^2$ from the algebra generated by $KB_j$), which the paper does not state.
  • Beyond the paper: the necessity direction hangs on a spanning lemma imported from hermitian-variable realization theory; a self-contained proof, or a counterexample in the $x,x^*$ setting, would be the cleanest stress test of the characterization.
  • Beyond the paper: the formal power-series identity in Theorem 5.2 may hold for a wider class of noncommutative formal rational series, e.g., those with non-symmetric coefficients, suggesting a formal-language version of the factorization.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper studies symmetric noncommutative (nc) rational functions in variables x_1,...,x_g and their adjoints, regular at 0, and defines a function to be plurisubharmonic (plush) if its nc complex Hessian is matrix positive semidefinite on an nc neighborhood of 0. The two main results are Theorem 1.3 and Theorem 1.4. Theorem 1.3 asserts that a symmetric nc rational function is plush in a neighborhood of 0 if and only if it can be written as f composed with q, where f is a convex nc rational function and q is an analytic nc rational mapping. Theorem 1.4 gives a computable criterion: for a minimal symmetric descriptor realization r(x,x*) = c*(K - Lambda_B(x) - Lambda_{B*}(x*))^{-1}c, the function r is plush if and only if PKP and P_* K P_* are positive semidefinite, where P and P_* are the projections onto the ranges of B and B*, respectively. The proof strategy is constructive: Section 2 reduces plushness to two separated estimates for the Hessian, Section 3 proves the realization criterion, Section 4 treats convex nc rational functions and proves the easy direction of Theorem 1.3, and Section 5 constructs f and q via formal power series and a finite-dimensional reduction.

Significance. If fully substantiated, these results are significant. Theorem 1.3 gives a clean structural characterization of plush nc rational functions, relating them to convex nc rational functions through analytic nc maps, and the proof is constructive. Theorem 1.4 provides a finite-dimensional, computable test for plushness in terms of a minimal realization, which is a practically useful criterion. The paper is careful in stating its dependence on realization theory and on prior results. However, the proof of the crucial spanning lemma, Lemma 3.2, is not self-contained and is delegated to a cited hermitian-variable version of results from [HMV06]; this lemma is used in the necessity direction of Theorem 3.1 and hence in the converse construction of Section 5. A second, smaller gap is the omitted proof of one identity in Lemma 5.3. These gaps do not appear to be circular or fraudulent, but they are load-bearing and should be repaired before the paper is accepted.

major comments (3)
  1. [3, Lemma 3.2] Lemma 3.2 is the central spanning lemma: it asserts that for a minimal realization and every epsilon > 0 there exist n, X in the free ball of radius epsilon, and v such that z = Delta(X)(c tensor v) has d linearly independent components. This lemma is used in Proposition 3.4 to prove the necessity direction of Theorem 3.1, and the converse construction in Section 5 relies on Theorem 3.1. The proof given in the manuscript, however, is limited to the sentence: substitute x_j = y_j + i y'_j and apply a hermitian version of [HMV06, Lemmas 7.2 and 7.4], justified by the local-global principle of [BK13]. This is not a proof in the manuscript; it is a citation to an adaptation that the paper does not state or verify. Since Remark 1.7 explicitly promises that results for symmetric variables transfer to hermitian variables but does not prove that transfer at the level of spanning lemmas, the necessity direction of the main characterization is not fully supported. Please provide a complete proof of Lemma 3.2, or a precise statement and proof of the hermitian-variable versions of the cited lemmas.
  2. [5.2, Lemma 5.3] Lemma 5.3 states two identities used to simplify alternating words in the moment-matching proof of Theorem 5.2. The first identity, B_l^* K [psi^* J psi] K B_j^* = B_l^* K B_j^*, is proved. The second identity, B_l K [psi^* J psi] K B_j^* = B_l K B_j^*, is dismissed with 'We omit the details.' This second identity is used in the simplification of the alternating products in equations (5.4)-(5.7) and is therefore essential to the conclusion that f(q(x)) equals the original realization r(x). The omission is not merely cosmetic. Please include a complete derivation, or at least a fully detailed sketch analogous to the proof of the first identity.
  3. [Remark 1.7 and global transfer of results] The paper repeatedly applies theorems from the literature that are stated for symmetric variables or for hermitian variables, while the paper works with the (x,x*) formalism. Remark 1.7 acknowledges this and gives a heuristic justification via the change of variables x_j = y_j + i y'_j. This transfer is used at load-bearing points: Lemma 3.2 invokes a hermitian version of [HMV06, Lemmas 7.2 and 7.4], and Proposition 4.1 invokes [HMV06, Proposition 5.1] together with Remark 1.7. The manuscript does not supply the transfer proofs that Remark 1.7 promises. Because the characterization theorems depend on these transferred statements, the paper should either prove the relevant hermitian-variable versions or restructure the argument to avoid relying on them.
minor comments (4)
  1. [Abstract] The abstract lists the variables as x_1,...,x_g,x_1^*,...,x_g; the last symbol should presumably be x_g^* rather than x_g.
  2. [Section 2.3, first paragraph] The formal domain of the symmetric descriptor realization is described using 'J - Lambda_B(X) - Lambda_B(X)^*', but the realization in equation (1.5) uses the signature matrix K, not J. This appears to be a typo.
  3. [Section 3, proof of Proposition 3.5] In the proof, the notation R and R_* is used both for the subspaces rng B and rng B^* and, near the end, for the inclusions of these subspaces. This dual use of the same symbol is confusing; using distinct symbols for subspaces and inclusions would improve readability.
  4. [Section 5.3.1, Remark 5.4] The term 'convexotonic map' is used without definition or reference to a specific definition. A one-sentence definition or a pointer to the precise definition in [HKMV] would help the reader verify the claim made in the remark.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: plush characterization is proved constructively from realization theory; cited prior work is independent support.

full rationale

The derivation chain is not circular. Theorem 1.4 is proved in Section 3 from the Hessian formulas (2.3)-(2.4), the direct-sum decomposition Proposition 2.1, and the spanning Lemma 3.2; Lemma 3.2 is quoted from published realization-theoretic results [HMV06, BK13] rather than from the theorem being proved. Theorem 1.3's forward direction is the explicit Section 5 construction: assuming plushness, Theorem 3.1 supplies K-nonnegativity of rngB and rngB*, and then f and q are built from angular operators and word substitutions; the equality r=f∘q is verified by formal-power-series moment matching (Theorem 5.2), not assumed. The reverse direction uses Corollary 4.3, which follows from Proposition 4.1 and Theorem 3.1, both proved independently. No fitted parameter is renamed as a prediction, and no definition smuggles the conclusion into the hypotheses. The only flagged weaknesses are proof gaps: Lemma 3.2 is justified by a 'hermitian version' of [HMV06, Lemmas 7.2 and 7.4] via [BK13] rather than a full proof, and Lemma 5.3 omits details of one identity. These are completeness and correctness concerns, not circularity: they do not make any conclusion equal to an input by construction.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

No numerical parameters are fitted and no new entities are postulated. The proof relies on standard realization theory, finite-dimensional algebra, and prior results from the same research group. The dependency with the most risk is Lemma 3.2, where a Hermitian-variable spanning property is imported rather than proved in full.

assumptions (6)
  • domain assumption Every symmetric nc rational function regular at 0 admits a symmetric descriptor realization, and minimal realizations exist and are unique up to unitary similarity.
    Invoked in Sections 1.2 and 2.3 to set up Theorems 1.3, 1.4, and the construction in Section 5; imported from realization theory (BGM05, HMV06, Vol18).
  • domain assumption Minimality implies the local spanning property in Lemma 3.2: for every epsilon there exist X in the free ball and v such that z = Delta(X)(c tensor v) has d linearly independent components.
    This is the load-bearing import. The proof defers to [HMV06, Lemmas 7.2, 7.4] and the local-global principle of [BK13], adapted to Hermitian variables via Remark 1.7. It is used to prove necessity in Theorem 3.1 and to start the converse in Theorem 5.1.
  • domain assumption Results proved for symmetric variables extend to Hermitian (x, x*) variables via the change of variables in Remark 1.7.
    Remark 1.7 asserts this extension across the paper, for example to apply [HMV06] convexity and controllability results in noncommuting variables with adjoints. The extension is not proved in full detail.
  • standard math For a signature matrix K and a K-nonnegative subspace N, there is a maximal K-nonnegative subspace containing N, with an angular operator, and the K-neutral subspace decomposition used in Propositions 3.5 and 4.1.
    Cited to Ando [And79] and used in the sufficiency part of Theorem 3.1 and in the construction in Section 5.
  • standard math Formal power series identity in a neighborhood of 0 implies equality of the corresponding nc rational functions.
    Used to pass from the formal composition identity in Theorem 5.2 to the rational decomposition in Theorem 5.1.
  • standard math The algebra generated by {K B_1, ..., K B_g} in M_{a+b}(C) is finite-dimensional, so it has a finite basis C_1, ..., C_h with structure constants Xi satisfying equation (5.9).
    Used in Section 5.3.1 to reduce infinitely many intermediate variables to finitely many.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Plurisubharmonic Noncommutative Rational Functions." pith.science (2026). https://pith.science/paper/G4KP6KGN

@misc{pith2026190801895,
  author       = {Pith},
  title        = {Pith review of: Plurisubharmonic Noncommutative Rational Functions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/G4KP6KGN}},
  note         = {Machine review of arXiv:1908.01895}
}
abstract

A noncommutative (nc) function in $x_1,\dots,x_g,x_1^*,\dots,x_g$ is called plurisubharmonic (plush) if its nc complex Hessian takes only positive semidefinite values on an nc neighborhood of 0. The main result of this paper shows that an nc rational function is plush if and only if it is a composite of a convex rational function with an analytic (no $x_j^*$) rational function. The proof is entirely constructive. Further, a simple computable necessary and sufficient condition for an nc rational function to be plush is given in terms of its minimal realization.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 35 canonical work pages

  1. [1]

    Ando, Linear operators on Kreĭn spaces

    T. Ando, Linear operators on Kreĭn spaces. Hokkaido University, Research Institute of Applied Electricity, Division of Applied Mathematics, Sapporo, 1979. ii+59 pp

  2. [2]

    Agler, J

    J. Agler, J. McCarthy: Global holomorphic functions in several non-commuting variables, Canad. J. Math. 67 (2015) 241--285

  3. [3]

    Agler, J.E

    J. Agler, J.E. McCarthy: Pick interpolation for free holomorphic functions, Amer. J. Math. 137 (2015) 1685--1701

  4. [4]

    Augat, J.W

    M. Augat, J.W. Helton, I. Klep, S. McCullough: Bianalytic Maps Between Free Spectrahedra, Math. Ann. 371 (2018) 883--959

  5. [5]

    J.A. Ball, G. Groenewald, T. Malakorn: Structured noncommutative multidimensional linear systems. SIAM J. Control Optim. 44 (2005) 1474--1528

  6. [6]

    J.A. Ball, G. Marx, V. Vinnikov: Interpolation and transfer-function realization for the noncommutative Schur--Agler class , Operator theory in different settings and related applications, 23--116, Oper. Theory Adv. Appl. 262, Birkh \"a user/Springer, Cham, 2018

  7. [7]

    Berstel, C

    J. Berstel, C. Reutenauer: Noncommutative rational series with applications , Encyclopedia of Mathematics and its Applications 137, Cambridge University Press, Cambridge, 2011

  8. [8]

    Blekherman, P.A

    G. Blekherman, P.A. Parrilo, R.R. Thomas (editors): Semidefinite optimization and convex algebraic geometry , MOS-SIAM Series on Optimization 13 , SIAM, 2013

Show all 37 references
  1. [9]

    Bre s ar, I

    M. Bre s ar, I. Klep: A local-global principle for linear dependence of noncommutative polynomials. Israel J. Math. 193 (2013) 71--82

  2. [10]

    Cohn: Skew fields

    P.M. Cohn: Skew fields. Theory of general division rings, Encyclopedia of Mathematics and its Applications, 57. Cambridge University Press, 1995

  3. [11]

    D'Angelo: Several complex variables and the geometry of real hypersurfaces , CRC Press, 1993

    J.P. D'Angelo: Several complex variables and the geometry of real hypersurfaces , CRC Press, 1993

  4. [12]

    de Oliveira, J.W

    M. de Oliveira, J.W. Helton, S. McCullough, M. Putinar: Engineering systems and free semi-algebraic geometry, in: Emerging applications of algebraic geometry (edited by M. Putinar, S. Sullivant), 17--61, Springer-Verlag, 2009

  5. [13]

    Effros, S

    E.G. Effros, S. Winkler: Matrix convexity: operator analogues of the bipolar and Hahn-Banach theorems, J. Funct. Anal. 144 (1997) 117--152

  6. [14]

    Forstnerič: Stein manifolds and holomorphic mappings

    F. Forstnerič: Stein manifolds and holomorphic mappings. The homotopy principle in complex analysis. Second edition. Ergebnisse der Mathematik und ihrer Grenzgebiete, 3rd Series, 56. Springer, Cham, 2017

  7. [15]

    Greene: Noncommutative plurisubharmonic polynomials part II: Local assumptions, J

    J.M. Greene: Noncommutative plurisubharmonic polynomials part II: Local assumptions, J. Math. Anal. Appl. 396 (2012) 481--496

  8. [16]

    Greene, J.W

    J.M. Greene, J.W. Helton, V. Vinnikov: Noncommutative Plurisubharmonic Polynomials Part I: Global Assumptions, J. Funct. Anal. 261 (2011) 3390--3417

  9. [17]

    Helton, I

    J.W. Helton, I. Klep, S. McCullough: The tracial Hahn-Banach theorem, polar duals, matrix convex sets, and projections of free spectrahedra, J. Eur. Math. Soc. 19 (2017) 1845--1897

  10. [18]

    Helton, I

    J.W. Helton, I. Klep, S. McCullough, M. Schweighofer: Dilations, Linear Matrix Inequalities, the Matrix Cube Problem and Beta Distributions, Mem. Amer. Math. Soc. 257 (2019), no. 1232, 104 pp

  11. [19]

    Helton, I

    J.W. Helton, I. Klep, S. McCullough and J. Vol c i c . Bianalytic free maps between spectrahedra and spectraballs, preprint https://arxiv.org/abs/1804.09743

  12. [20]

    Helton, S.A

    J.W. Helton, S.A. McCullough: Convex polynomials have Degree Two or Less, SIAM J. Matrix Anal. Appl. 25 (2004) 1124--1139

  13. [21]

    Helton, S

    J.W. Helton, S. McCullough: Every convex free basic semi-algebraic set has an LMI representation. Ann. of Math. (2) 176 (2012) 979--1013

  14. [22]

    Helton, S

    J.W. Helton, S. McCullough: Free convex sets defined by rational expressions have LMI representations, J. Convex Anal. 21 (2014) 425--448

  15. [23]

    Helton, S.A

    J.W. Helton, S.A. McCullough, V. Vinnikov: Noncommutative convexity arises from linear matrix inequalities, J. Funct. Anal. 240 (2006) 105--191

  16. [24]

    Helton, V

    J.W. Helton, V. Vinnikov: Linear matrix inequality representation of sets, Comm. Pure Appl. Math. 60 (2007) 654--674

  17. [25]

    Kaliuzhnyi-Verbovetskyi, V

    D.S. Kaliuzhnyi-Verbovetskyi, V. Vinnikov: Singularities of rational functions and minimal factorizations: the noncommutative and the commutative setting, Linear Algebra Appl. 430 (2009) 869--889

  18. [26]

    Marcus, D.A

    A.W. Marcus, D.A. Spielman, N. Srivastava: Interlacing families II: Mixed characteristic polynomials and the Kadison--Singer problem, Ann. of Math. 182 (2015) 327--350

  19. [27]

    McCarthy, R.M

    J.E. McCarthy, R.M. Timoney: Non-commutative automorphisms of bounded non-commutative domains, Proc. Roy. Soc. Edinburgh Sect. A 146 (2016) 1037--1045

  20. [28]

    Muhly, B

    P.S. Muhly, B. Solel: Schur class functions and automorphism of Hardy algebras, Doc. Math. 13 (2008) 365--411

  21. [29]

    Pascoe, R

    J. Pascoe, R. Tully-Doyle: The royal road to automatic noncommutative real analyticity, monotonicity, and convexity, preprint https://arxiv.org/abs/1907.05875

  22. [30]

    Pascoe, R

    J. Pascoe, R. Tully-Doyle: Free Pick functions: representations, asymptotic behavior and matrix monotonicity in several noncommuting variables, J. Funct. Anal. 273 (2017) 283--328

  23. [31]

    Passer, O.M

    B. Passer, O.M. Shalit, B. Solel: Minimal and maximal matrix convex sets, J. Funct. Anal. 274 (2018) 3197--3253

  24. [32]

    Paulsen: Completely bounded maps and operator algebras , Cambridge Univ

    V. Paulsen: Completely bounded maps and operator algebras , Cambridge Univ. Press, 2002

  25. [33]

    Popescu: Free pluriharmonic majorants and commutant lifting, J

    G. Popescu: Free pluriharmonic majorants and commutant lifting, J. Func. Anal. 255 (2008) 891--939

  26. [34]

    Popescu: Free holomorphic automorphisms of the unit ball of B(H)^n , J

    G. Popescu: Free holomorphic automorphisms of the unit ball of B(H)^n , J. reine angew. Math. 638 (2010) 119--168

  27. [35]

    Skelton, T

    R.E. Skelton, T. Iwasaki, K.M. Grigoriadis: A Unified Algebraic Approach to Linear Control Design, Taylor and Francis, 1996

  28. [36]

    Salomon, O.M

    G. Salomon, O.M. Shalit, E. Shamovich: Algebras of bounded noncommutative analytic functions on subvarieties of the noncommutative unit ball, Trans. Amer. Math. Soc. 370 (2018) 8639--8690

  29. [37]

    Vol c i c : Matrix coefficient realization theory of noncommutative rational functions, J

    J. Vol c i c : Matrix coefficient realization theory of noncommutative rational functions, J. Algebra 499 (2018) 397--437

Pith tools

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