{"id":"ea4a1a49-f3ab-4204-a42d-7b886ef9ef03","arxiv_id":"2608.08404","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A multivariate lexicographic functional calculus is defined, and its divided-difference derivative formula unifies the known Fréchet-differentiability results for functions of operators in commutative, finite-dimensional, and Besov-regular settings.","lead":"This paper introduces a new lexicographic functional calculus that evaluates multivariable functions on tuples of noncommuting self-adjoint operators, acting left to right with ideal elements inserted between them. The payoff is a unified derivative formula for maps f(a) that recovers the commutative, finite-dimensional, and Besov-regularity cases in one theorem.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 4.4's proof jumps from f^[k]∈VC(R^{k+1}) to f∈VC^k(R) without justification; the Besov recovery in Theorem 1.12 depends on that jump.","rationale":"The reader correctly identifies the Peller theorem as the main external input, but the most fragile link in the Besov corollary is what happens after Peller's theorem is granted. Theorem 4.13 itself is well supported: the construction of LFC, the norm-control estimates, and the polynomial approximation argument are coherent, and the missing C_I^{i-1} constants in the derivative bounds are harmless. The real issue is in §4.1: the characterization C^k_β(R)=C^k_β(R) is used in Cor 4.10 to convert f^[k]∈VC into f∈VC^k. The final paragraph of the proof of Theorem 4.4 asserts without proof that higher-order Varopoulos regularity of the top divided difference implies lower orders. That implication may be true — likely via Peller-type converses — but it is not demonstrated, and the cited containment C^{k+1}⊆C^k_β does not obviously entail it. Since Theorem 1.1(iii) is one of the three headline recoveries, the central claim is conditionally supported rather than unconditionally established. I therefore recommend CONDITIONAL rather than ACCEPT.","tokens_in":54862,"tokens_out":36898,"duration_ms":386212,"concrete_test":"Re-derive the final paragraph of Theorem 4.4 in the case k=2: from f∈C^2(R) and f^[2]∈VC(R^3), prove that f^[1]∈VC(R^2), e.g., by applying Peller's Theorem 4.9 at level 1 to a compactly supported cutoff fχ or by invoking the converse direction of Peller's theorem. If the implication cannot be established, or a C^2 function with f^[2]∈VC but f^[1]∉VC is exhibited, then Corollary 4.10's use of Theorem 4.4 is a genuine gap and the proof of Theorem 1.12 needs a revised argument.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"In §4.1, the proof of Theorem 4.4 ends with the assertion that, since C^{k+1}(R)⊆C^k_β(R) by Lemma 4.6, it is 'automatic' that if f∈C^k(R), then f^[i]∈VC(R^{i+1}) for 0≤i<k. This is not automatic: top-order divided-difference membership in the Varopoulos algebra does not by itself force lower-order divided differences into VC, and the containment C^{k+1}⊆C^k_β does not supply such an implication. The step is load-bearing because Corollary 4.10 proves B^{k,∞}_1(R)⊆VC^k(R) by combining Peller's Theorem 4.9—which gives only f^[k]∈VC(R^{k+1})—with Theorem 4.4. If the implication fails, the claimed recovery of the Peller/Besov regularity result in Theorem 1.1(iii) is unsupported as written. The main derivative formula (Theorem 4.13) and the abstract LFC framework are unaffected, since they only require f∈C^k_α(R) (the polynomial closure); the weakness is specifically in the advertised unification of the three headline results.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a 'lexicographic functional calculus' (LFC) for tuples of noncommuting self-adjoint elements of a unital C*-algebra, with elements of a symmetrically normed ideal I inserted between the variables. For a family α of possibly infinite norms that controls polynomial LFC and is finite on polynomials, the calculus extends from polynomials to the closure C_α(R^m), with norm estimates in I. Three main examples are developed: the uniform family in commutative algebras, a scaled uniform family in finite-dimensional algebras, and the (A-)Varopoulos family in arbitrary C*-algebras, plus an MOI-based example. The main application, Theorem 4.13, asserts that if f belongs to the α-C^k closure C^k_α(R) and α controls polynomial LFC in I, then the map b ↦ f(a+b)−f(a) sends I_sa into I and is Fréchet C^k, with the k-th derivative expressed as a symmetrized LFC applied to the k-th divided difference f^[k]. Corollaries recover the commutative, finite-dimensional, and Besov/Hölder regularity results of Theorem 1.1 and generalize several results on differentiable operator functions in ideals.","tokens_in":55118,"tokens_out":18360,"duration_ms":187126,"significance":"If the proof gaps are repaired, this is a valuable unifying framework: Theorem 4.13 yields a single derivative formula whose commutative, finite-dimensional, and Varopoulos specializations reproduce three previously separate regularity statements, with explicit norm-control hypotheses. The paper gives detailed proofs of the main theorem by polynomial approximation, supplies concrete evaluation formulas in the commutative, finite-dimensional, and Varopoulos settings, and offers a new characterization of Jekel's noncommutative C^k space. It also recovers Peller's Besov regularity theorem and the author's earlier results as corollaries, and it is transparent about which supporting results are sketches or left to the reader. These are concrete, falsifiable claims, and the central derivative theorem is supported by explicit norm estimates rather than by heuristic arguments.","major_comments":[{"comment":"The final step of the proof of Theorem 4.4 is invalid as written. The text asserts: 'since C^{k+1}(R)⊆C^k_β(R) by Lemma 4.6, it is automatic that if f∈C^k(R), then f^[i]∈VC(R^{i+1}) ...'. Lemma 4.6 only establishes that C^{k+1}(R) is contained in the closure of W^k(R) in C^k_α(R); it does not establish that every f∈C^{k+1} has all divided differences in the Varopoulos algebra. In fact, for k=1 the asserted implication would make every C^2 function operator Lipschitz (equivalently, f^[1]∈VC(R^2)), a much stronger statement than anything proved or cited in the paper. This step is load-bearing: Corollary 4.10 combines Peller's Theorem 4.9 (which gives only f^[k]∈VC(R^{k+1})) with Theorem 4.4 to conclude B^{k,∞}_1⊆VC^k, and Theorem 1.12 uses Corollary 4.10 for the Besov recovery advertised in Theorem 1.1(iii). The abstract LFC framework and Theorem 4.13 are not affected, since they use the closure space C^k_α(R) directly. Note that Remark 4.11 sketches an alternative proof of ˙B^{k,∞}_1⊆VC^k that avoids this disputed implication; if that route is promoted to a formal proof, the Besov recovery can be salvaged. Please either supply a correct proof of the implication in Theorem 4.4 or restructure §4.1 so that Theorem 1.12 depends on the Remark 4.11 argument.","section":"§4.1, proof of Theorem 4.4; used in Corollary 4.10 and Theorem 1.12"}],"minor_comments":[{"comment":"The notation C^k_α(R) and C^k_α(R) is nearly indistinguishable in plain text; in Lemma 4.6 and Theorem 4.4 this ambiguity appears to have contributed to the invalid inference discussed above. Please use an unambiguous notation, e.g., \\(\\overline{C^k_\\alpha}(\\mathbb{R})\\), throughout.","section":"Throughout §4.1"},{"comment":"Lemma 3.10 is stated without proof, although Theorem 3.11's extension argument relies on it. Since the proof is a straightforward polynomial estimate, it should be included or sketched in the paper.","section":"Lemma 3.10"},{"comment":"Lemma A.4 is only sketched. For a self-contained appendix, please expand the proof that the divided-difference map sends Hol(U) into Hol_0(U^{k+1}).","section":"Lemma A.4"},{"comment":"Theorem 3.30 is presented as a sketch and depends on external MOI facts. If it is to be used as a main example, please supply the details of the closure argument and explicitly state the precise role of [32, Thm. 4.2.4(iii)] and [30, Prop. 4.1.7].","section":"Theorem 3.30"},{"comment":"The case j=0 in formula (4.2) uses Δ_0 and ρ_0, but Notation 4.5 defines Δ_k and ρ_k only for k∈N. Please handle k=0 separately or extend the notation.","section":"Lemma 4.6"},{"comment":"In the case ε>1, the inclusion 'C^{k,ε}_loc(R)⊆P' is unexplained. Please spell out that a locally Hölder-continuous k-th derivative with exponent greater than 1 is locally constant, hence the function is locally polynomial.","section":"Proof of Theorem 1.12"}],"recommendation":"major_revision","confidential_remarks":"The main issue is the proof of Theorem 4.4, which is load-bearing for the Besov recovery in Theorem 1.12. However, the manuscript itself contains an alternative route in Remark 4.11, so the problem is repairable without affecting the central derivative theorem. I recommend revision primarily to move the Besov recovery onto a valid proof and to remove the C^k_α/C^k_α notational ambiguity. The self-citations appear appropriate because old results are reproved as corollaries. The paper fits the scope of math.FA and the length is substantial but largely justified by the breadth of the application section."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core of this paper is good and worth engaging. The lexicographic functional calculus is a new object, the integral description of the Varopoulos algebra in Theorem 2.7 is a useful addition, and Theorem 4.13 is a real contribution: the derivative formula is proven with explicit norm-control estimates and a polynomial-approximation argument, and it genuinely unifies the commutative, finite-dimensional, and Besov-regularity results in a way that was not there before. The paper also gives a new characterization of Jekel's noncommutative C^k space, and it is honest about the Schatten-ideal case it does not recover. Those are substantial achievements. The soft spot flagged in the stress-test is real. In the proof of Theorem 4.4, the final step says that because C^{k+1}(R) sits in the closure of W^k(R) in C^k_α(R), it is automatic that every f∈C^k(R) has f^[i]∈VC(R^{i+1}) for i<k. That does not follow from the stated Lemma 4.6. Lemma 4.6 gives W^k⊆C^k_α and C^{k+1} contained in the closure of W^k; it does not give membership of lower divided differences in the Varopoulos algebra for arbitrary C^k functions. Corollary 4.10 needs exactly that: Peller's theorem supplies only f^[k]∈VC(R^{k+1}), and passing to f∈VC^k(R) requires lower-order VC control. As written, the Besov recovery in Theorem 1.12 is unsupported. This is not a fatal blow to the main framework — Theorem 4.13 and the abstract LFC machinery only need f∈C^k_α(R) — but it does mean one of the three advertised headline recoveries needs a fix. The repair may be straightforward: use the Besov embedding B^{k,∞}_1⊆B^{j,∞}_1 and apply Peller's theorem at each order, or prove a separate inheritance lemma for lower divided differences. But it should not be left as is. The other listed sketches — Lemma 3.10, Proposition 2.4, Lemma A.4, Theorem 3.30 — are minor by comparison; they are either standard or peripheral. The external reliance on Peller's theorem is acceptable. Who is this for? Operator theorists and anyone working on noncommutative functional calculus or higher derivatives of operator functions. It deserves a serious referee, not a desk reject, but the referee should demand a repaired Theorem 4.4 before acceptance.","headline":"Lexicographic functional calculus is a new and mostly solid construction with a real unifying derivative formula, but the advertised Besov recovery rests on an under-justified step in Theorem 4.4 that needs repair.","tokens_in":816,"tokens_out":1819,"would_cite":true,"duration_ms":114502,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["47A60","47A13","26E15","47L20","46E10","46E35"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper introduces lexicographic functional calculus and proves one derivative formula that makes the three known regularity results for operator functions corollaries of a single theorem.","keywords":["lexicographic functional calculus","functional calculus calculus","divided differences","symmetrically normed ideals","C*-algebra","Fréchet derivatives","Varopoulos algebra","Besov spaces"],"falsifier":"Take any explicitly given function in the homogeneous Besov space $\\dot{B}^{1,\\infty}_1(\\mathbb{R})$ that is $C^1$ but not $C^2$, and compute whether its first divided difference restricted to $[-1,1]^2$ has finite projective tensor norm; if it does not, the embedding $\\dot{B}^{1,\\infty}_1(\\mathbb{R})\\subseteq V C^1(\\mathbb{R})$ is false and the Besov recovery through Varopoulos LFC fails. The same derivative formula can be tested directly in $M_2(\\mathbb{C})$ with $f(t)=t^4$ and a noncommuting self-adjoint pair $a,b$ to check Theorem 4.13 in a finite-dimensional setting.","tokens_in":54639,"feed_emoji":"🧮","tokens_out":14097,"duration_ms":136634,"temperature":0.7,"pith_summary":"The paper introduces lexicographic functional calculus (LFC), a multivariate functional calculus for tuples of noncommuting self-adjoint elements in which the variables act left to right and chosen elements of an ideal are inserted between consecutive actions. Its purpose is to prove one derivative formula for the map $b \\mapsto f(a+b)-f(a)$: when a norm family $\\alpha$ controls polynomial LFC in a symmetrically normed ideal $I$, every $f \\in C^k_\\alpha(\\mathbb{R})$ makes this map a $C^k$ map from $I_{\\mathrm{sa}}$ into $I$, and its $k$th Fréchet derivative is a permutation sum of the $k$th divided difference $f^{[k]}$ applied through LFC to the directions. The formula is established on polynomials by direct computation and extended by approximation, so no multiple-operator-integral machinery is needed. A single theorem (4.13) then contains the three known smoothness results as corollaries: commutative $C^*$-algebras with $f \\in C^k$, finite-dimensional $C^*$-algebras with $f \\in C^k$, and arbitrary $C^*$-algebras with $f$ slightly better than $C^k$. This is what the author means by supplying a single framework for 'functional calculus calculus.'","feed_headline":"One functional calculus unifies three operator smoothness results","feed_subtitle":"The kth derivative of f(a+b)-f(a) is a lexicographic divided-difference sum, covering commutative, finite-dimensional, and Besov cases.","key_machinery":"The central object is lexicographic functional calculus: for a function $\\varphi$ of $m$ variables and a tuple $a=(a_1,\\ldots,a_m)$ of self-adjoint elements, the expression $\\varphi_{A,\\alpha}(a)\\sharp b$ is built to mean that $a_1$ acts first, then the perturbation $b_1$ is inserted, then $a_2$ acts, and so on, ending with $a_m$. The construction starts from the polynomial rule $P_\\otimes(a)\\sharp b=\\sum c_\\delta a_1^{\\delta_1}b_1\\cdots a_{m-1}^{\\delta_{m-1}}b_{m-1}a_m^{\\delta_m}$ and extends it by continuity to $\\alpha$-continuous functions once the norm family $\\alpha$ controls the size of these polynomial evaluations. The same expression, with the $k$th divided difference $f^{[k]}$ in place of $P$ and all $k+1$ entries set to $a+b$, gives the $k$th derivative after symmetrizing over the $b_i$'s. The paper's examples are the uniform family, the finite-dimensional scaled uniform family, and the Varopoulos family, whose norm is the projective tensor-product norm on continuous functions over spectra and which is defined in every $C^*$-algebra.","core_discovery":"The load-bearing claim is Theorem 4.13. For a unital $C^*$-algebra $A$, a symmetrically normed ideal $I$ of $A$, and a family $\\alpha$ of possibly infinite norms that controls polynomial lexicographic functional calculus in $I$, the paper proves that $f(a+b)-f(a)\\in I$ for all $b\\in I_{\\mathrm{sa}}$, that $f_{a,I}:I_{\\mathrm{sa}}\\to I$ is Fréchet $C^k$ with respect to the ideal norm, and that $\\partial_{b_k}\\cdots\\partial_{b_1}f_{a,I}(b)=\\sum_{\\pi\\in S_k} f^{[k]}_{I,\\alpha}(a+b,\\ldots,a+b)\\sharp[b_{\\pi(1)},\\ldots,b_{\\pi(k)}]$. The proof first verifies the formula for polynomials, using the holomorphic perturbation formula, and then extends it by continuity to $C^k_\\alpha(\\mathbb{R})$. The paper shows that the uniform family controls LFC in commutative and finite-dimensional algebras, and that the Varopoulos family controls LFC in every $C^*$-algebra, so each of the three classical regularity statements follows from the same theorem by checking the relevant divided-difference class.","pith_inferences":["The norm-control condition is the actual engine: any new family $\\alpha$ satisfying it automatically yields a regularity theorem, so the three cases treated in the paper are the first instances of a general recipe rather than isolated results.","Identifying $V C^k(\\mathbb{R})$ with functions whose $k$th divided difference is Varopoulos-continuous gives an explicit, checkable description of an abstract noncommutative $C^k$ completion; this characterization should transfer to settings where divided differences are already the natural language, such as free probability.","A direct next step the paper itself leaves open is whether a scaled uniform family controls polynomial LFC in the Schatten ideals $S_q$; if it does, Theorem 4.13 would immediately reproduce higher-order $S_q$-differentiability without Besov assumptions."],"forward_implications":["In every commutative unital $C^*$-algebra, $f\\in C^k(\\mathbb{R})$ implies that $a\\mapsto f(a)$ is Fréchet $C^k$, with derivative $f^{(k)}(a+b)b_1\\cdots b_k$.","In every finite-dimensional unital $C^*$-algebra, the same $C^k$ regularity holds and the derivative is an explicit sum over the spectral projections of $a+b$ with the $b_i$'s inserted between them.","For every unital $C^*$-algebra, membership in the homogeneous Besov space $\\dot{B}^{k,\\infty}_1(\\mathbb{R})$ or in $C^{k,\\varepsilon}_{\\mathrm{loc}}(\\mathbb{R})$ suffices for $f_A$ to be Fréchet $C^k$.","For any symmetrically normed ideal $I$, $f(a+b)-f(a)$ lies in $I$ for every self-adjoint perturbation $b\\in I_{\\mathrm{sa}}$, and the map $b\\mapsto f(a+b)-f(a)$ is $C^k$ with respect to the ideal norm.","The derivative formula is explicit enough to serve as a computation rule: symmetrize the $k$th divided difference evaluated at $a+b$ through LFC over the $k$ directions."],"supporting_citations":[{"why":"Supplies the integral representation of $k$th divided differences of Besov functions used to embed $B^{k,\\infty}_1(\\mathbb{R})$ into $V C^k(\\mathbb{R})$.","marker":"[38]"},{"why":"Provides the stronger form of the same divided-difference representation cited to support Theorem 1.12.","marker":"[39]"},{"why":"Gives the detailed proof of the Besov divided-difference representation that the main text invokes without proof.","marker":"[30, App. A]"},{"why":"Supplies the divided-difference identities and representation formulas used throughout the construction of $C^k_\\alpha(\\mathbb{R})$.","marker":"[32]"},{"why":"Introduced the abstract noncommutative $C^k$ space that Theorem 4.4 concretely identifies with $V C^k(\\mathbb{R})$.","marker":"[25, Ch. 18]"},{"why":"Provides the multilinear Schur multiplier estimates that sharpen the finite-dimensional LFC control constants.","marker":"[46]"},{"why":"Proved the finite-dimensional Fréchet $C^k$ regularity result that the paper recovers as a corollary of Theorem 4.13.","marker":"[11]"}],"fun_headline_variants":["One calculus unifies three smoothness results","Lexicographic functional calculus unifies operator regularity","Unified proof for commutative, finite-dim, and Besov smoothness","LFC: the k-th derivative is a lexicographic divided-difference sum","Three regularity results, one lexicographic calculus"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof that Besov-class functions are Varopoulos-$C^k$ rests on an external theorem, quoted without proof, that every function in $B^{k,\\infty}_1(\\mathbb{R})$ has its $k$th divided difference represented as an integral of $k+1$ bounded measurable functions with finite projective-type norm; if that representation failed, the claimed Besov recovery through Varopoulos LFC would collapse, although the abstract LFC framework and derivative formula would remain intact.","fun_headline_variants_meta":{"raw":{"variants":["One calculus unifies three smoothness results","Lexicographic functional calculus unifies operator regularity","Unified proof for commutative, finite-dim, and Besov smoothness","LFC: the k-th derivative is a lexicographic divided-difference sum","Three regularity results, one lexicographic calculus"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001667,"raw_usage":{"total_tokens":6813,"prompt_tokens":1346,"completion_tokens":5467,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":962,"completion_tokens_details":{"reasoning_tokens":5385}},"tokens_in":962,"tokens_out":5467,"duration_ms":40069,"temperature":1.0,"reasoning_tokens":5385,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:36:30.050594+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take any explicitly given function in the homogeneous Besov space $\\dot{B}^{1,\\infty}_1(\\mathbb{R})$ that is $C^1$ but not $C^2$, and compute whether its first divided difference restricted to $[-1,1]^2$ has finite projective tensor norm; if it does not, the embedding $\\dot{B}^{1,\\infty}_1(\\mathbb{R})\\subseteq V C^1(\\mathbb{R})$ is false and the Besov recovery through Varopoulos LFC fails. The same derivative formula can be tested directly in $M_2(\\mathbb{C})$ with $f(t)=t^4$ and a noncommuting self-adjoint pair $a,b$ to check Theorem 4.13 in a finite-dimensional setting.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the integral representation of $k$th divided differences of Besov functions used to embed $B^{k,\\infty}_1(\\mathbb{R})$ into $V C^k(\\mathbb{R})$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the stronger form of the same divided-difference representation cited to support Theorem 1.12."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the divided-difference identities and representation formulas used throughout the construction of $C^k_\\alpha(\\mathbb{R})$."},{"cited_title":"Skripka and A","cited_arxiv_id":null,"evidence_quote":"Provides the multilinear Schur multiplier estimates that sharpen the finite-dimensional LFC control constants."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Proved the finite-dimensional Fréchet $C^k$ regularity result that the paper recovers as a corollary of Theorem 4.13."}],"review_version":1}