{"id":"c2e65b0c-1d7a-4214-a8c5-5a9fc3eb89d2","arxiv_id":"1908.01850","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A Schur or Schur-Agler function on the polydisc factors into two factors of the same class exactly when it has a realization whose colligation block matrix satisfies D21 = 0 and aD12 = C1B2, with the factors built explicitly from the matrix blocks.","lead":"This mathematics paper gives explicit block-matrix conditions on the operator 'colligation' used to represent a Schur or Schur-Agler function, conditions that hold exactly when the function splits into a product of two functions of the same class. A generalist reads it for a clean operator-theoretic criterion that turns a function-theory question, factorization, into a checkable matrix-structure question.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'complete answer' claim rests on Theorem 4.2 and Theorem 5.3, both stated without proof; the algebraic step in Theorem 2.3 reconstructs correctly and does not undermine the core theorem.","rationale":"The reader's weakest_assumption was the multi-line algebra in the converse of Theorem 2.3. I checked that algebra: the garbled displayed line near (2.5) is a typographical corruption of a correct identity, and the conclusion that V1 and V2 are isometries follows once the normalization |beta|^2 = 1 - C2*C2 is used consistently. Thus I do not regard that as a load-bearing defect in the central nonzero-at-origin claim. However, the paper explicitly asserts a 'complete answer' and then states Theorem 4.2 and Theorem 5.3 without proofs. These are not marginal remarks: Theorem 4.2 is the vanishing-at-origin analogue in SA(D^n), directly extending the main classification, and Theorem 5.3 is the advertised Drury-Arveson multiplier analogue. The text itself flags these omissions ('leave out the details', 'We omit the proof'), and Theorem 4.2 additionally contains a missing condition and a typo in the class name. A referee cannot certify the paper's completeness claims without these proofs. Therefore the verdict should remain CONDITIONAL: the core proven theorems are sound as far as the reconstruction shows, but the manuscript needs full proofs (or explicit downgrading to conjectures/statements) for the unproved classifications before the 'complete answer' claim is acceptable. This is a partial agreement with the reader because the reader's formal weakest_assumption concerned the algebra, but the reader also noted the unproved statements in the rationale; my read shifts the emphasis to those omitted proofs as the more salient unresolved issue.","tokens_in":24008,"tokens_out":19944,"duration_ms":178785,"concrete_test":"Provide the omitted proof of Theorem 4.2(1) by adapting the one-variable proof of Theorem 4.1(1) to the n-variable block form. Explicitly, from V with a = 0, Bi(2) = 0, Dij(21) = 0, and C(1)C(1)*D(12) = C(1)*C(1)D(12) with C(1)*C(1) > 0, choose x with |x|^2 = C(1)*C(1) and define V1 = [[0, B(1)], [(1/x)C(1), D(1)]] and V2 = [[x, (1/bar-x)C(1)*D(12)], [C(2), D(2)]]. Then verify directly that V1 and V2 are isometric colligations and that tau_{V1} tau_{V2} = tau_V. If this construction succeeds, the theorem is true but unproved; if any isometry or product identity fails, the stated classification is false and the recommendation changes accordingly.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central nonzero-at-origin characterization (Theorems 2.4 and 3.4) is supported by an algebraic converse whose displayed line near (2.5) is garbled, but the intended identity is recoverable: with |beta|^2 = 1 - C2*C2 and alpha = a/beta, the needed step is (bar-beta/alpha)(|alpha|^2 + (1/|beta|^2)C1*C1)B2 = (bar-a + (1/a)C1*C1)B2, not 'bar-a squared over alpha'. Using this correction, V1 and V2 are indeed isometries and V1V2 = V. So I do not find a correctness failure in the main sufficiency argument. The load-bearing gap is the paper's own claim of a 'complete answer'. Section 4 states Theorem 4.2 for SA(D^n) vanishing at the origin with 'leave out the details to the reader', and Section 5.4 states Theorem 5.3 for Drury-Arveson multipliers with 'We omit the proof'. These are stated classification results, not conjectures, and the introduction advertises the paper as giving a complete answer. Moreover, Theorem 4.2 contains a missing condition -- the line 'one has Bi(2) =, Dij(21) = 0' is visibly incomplete -- and 'AS(D^n)' should read 'SA(D^n)', so the statement is not even fully checkable as printed. Since the vanishing-at-origin case and the multiplier case are part of the paper's advertised scope, the absence of proofs is a real completeness concern, although it does not invalidate the proven nonzero-at-origin theorems.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies factorization of Schur functions and Schur-Agler class functions on the polydisc through isometric colligations, building on Agler's transfer-function realization. For functions with nonzero value at the origin, it characterizes when theta in SA(D^n) is a product of two SA(D^n) factors (Theorem 3.4) or a product of functions in disjoint sets of variables (Theorem 2.4) by explicit block-structure conditions on any isometric colligation realizing theta, called property F(n) and F_m(n), respectively; explicit formulas for the factors are given in (3.12). A one-variable vanishing-at-origin classification is proved in Theorem 4.1, and extensions to vanishing functions in SA(D^n) and to Drury-Arveson multipliers are stated as Theorems 4.2 and 5.3. The paper also provides a product construction for several one-variable factors (Theorem 5.1), examples involving Blaschke factors, a connecting result between F_m(n) and F(n) (Theorem 5.2), and a discussion of the reversibility of the canonical constructions.","tokens_in":24178,"tokens_out":3575,"duration_ms":35553,"significance":"If the main results hold as stated, the paper provides a clean, checkable colligation-level answer to a natural factorization question for Schur-Agler class functions, with explicit factor realizations and a two-way correspondence. The proofs of Theorems 2.2-2.4, 3.2-3.4, 4.1, 5.1, and 5.2 are direct algebraic verifications; the constructions of V from V1 and V2 and the reverse construction are explicit and, as shown in Section 5.5, reversible up to a unimodular phase. These are genuine strengths. However, the advertised 'complete answer' is weakened by the fact that two stated classification results, Theorem 4.2 and Theorem 5.3, are given without proof and that the statement of Theorem 4.2 is not fully checkable as printed.","major_comments":[{"comment":"Theorem 4.2 is presented as a classification of factorizations of functions in SA(D^n) vanishing at the origin, yet the paper explicitly says 'we leave out the details to the reader' and the statement contains an incomplete condition ('one has Bi(2) =, Dij(21) = 0') and a typo ('AS(D^n)' instead of 'SA(D^n)'). Because the introduction advertises a complete answer and Section 4 is framed as a complete description, this unproved and partially unreadable statement is load-bearing for the paper's central claim. The authors must either supply a full proof and correct the statement, or visibly downgrade this part to a conjecture and adjust the advertised scope.","section":"Section 4, Theorem 4.2"},{"comment":"Theorem 5.3, concerning factorization of multipliers of the Drury-Arveson space, is stated with the sentence 'We omit the proof'. Since the introduction and the section title present multipliers as part of the paper's scope, the absence of a proof leaves the 'complete answer' claim unsupported for this case. A complete proof should be added, or the result should be explicitly labeled as a conjecture with the completeness claim restricted accordingly.","section":"Section 5.4, Theorem 5.3"},{"comment":"The displayed line after equation (2.2) contains an incorrect algebraic factor: it reads '\\bar a^2 / alpha' where the argument requires '(\\bar beta / alpha)'. This identity is the key decoupling step showing that V2 is an isometry, and it is reused in Theorem 3.3. As printed, the proof is not correct, although the intended identity is recoverable from V*V = I together with (2.5). The displayed computation must be corrected before publication.","section":"Section 2, proof of Theorem 2.3, around equation (2.5)"}],"minor_comments":[{"comment":"There are several typographical errors, for example 'out results' in the introduction, 'Schu r' in the abstract, and 'opera tor matrix' in the abstract; a careful proofreading pass is needed.","section":"Throughout"},{"comment":"In addition to the incomplete condition already noted, 'AS(D^n)' should read 'SA(D^n)' in the statement of Theorem 4.2.","section":"Section 4, Theorem 4.2"},{"comment":"In the statement of Theorem 5.3 the domain is the unit ball B^n, but the text writes 'z in D^n'; this should be corrected.","section":"Section 5.4, Theorem 5.3"}],"recommendation":"major_revision","confidential_remarks":"The core nonzero-at-origin theorems appear sound and are supported by explicit, reconstructible algebra, so the paper is publishable in principle. The main obstacle is not the mathematics itself but the gap between the advertised complete answer and the actual content: two stated classification theorems are unproved, and one of them is not even readable as printed. I would encourage the editor to require that the authors either supply the missing proofs or clearly mark those results as conjectures, and to correct the erroneous displayed identity in Section 2."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core of this paper is solid. Theorems 2.4 and 3.4 give explicit, reversible block-matrix conditions for when a Schur-Agler function splits as a product, and I went through the isometry normalizations in the converse proof. The algebra works; the garbled line near (2.5) is a typesetting error, not a mathematical one, and the intended identity can be reconstructed from the surrounding equations. The explicit formulas for the factors in (3.11)-(3.12) are a genuine convenience, and the forward direction is properly credited to the classical product-of-colligations construction. If you work with transfer-function realizations, this is a directly usable criterion.\n\nThe novelty is real, though not dramatic. The forward direction is classical; the contribution is the converse and the clean statement of the block conditions, especially for the Schur-Agler class in more than two variables. That is worth having in the literature.\n\nThe soft spots are in the packaging. The introduction says the paper gives \"a complete answer\" to the factorization question, but Theorem 4.2 (the n-variable vanishing-at-origin case) and Theorem 5.3 (Drury-Arveson multipliers) are both stated without proof. Theorem 4.2's statement is also visibly corrupted: one condition reads \"Bi(2) =\" with nothing after it, and the class is misspelled as AS(D^n). That is not a minor typo; it makes the theorem uncheckable as printed. The one-variable vanishing-at-origin case (Theorem 4.1) is proven, so the gap is specifically in the advertised n-variable and multiplier claims. The authors should either prove these statements or explicitly mark them as conjectures and trim the \"complete answer\" language. The many smaller typos throughout the paper, while annoying, do not compromise the main argument.\n\nThis paper deserves a serious referee, not a desk reject. The main theorems are correct and the explicit criteria are useful. A referee should be asked to check the unproved statements carefully and require revision before the paper is accepted. If you use realization theory, cite this for the proven nonzero-origin characterization; just don't cite it for the vanishing-at-origin or multiplier cases until they appear with proofs.","headline":"The main factorization criteria are correct and useful, but the paper's 'complete answer' claim outruns what is actually proven: the n-variable vanishing-at-origin and Drury-Arveson results are merely stated.","tokens_in":24939,"tokens_out":1639,"would_cite":true,"duration_ms":20304,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["32A10","32A38","32A70","47A48","47A13","46E15","93B15","15A40"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that a Schur-Agler class function with nonzero value at the origin factors into two Schur-Agler class functions exactly when its realizing isometric colligation has an explicit block structure, and gives explicit…","keywords":["Schur functions","Schur-Agler class","factorization","isometric colligations","transfer function realizations","polydisc","block operator matrices"],"falsifier":"Take any explicit isometric colligation $V$ of the form with $a \\neq 0$, $D_{21}=0$, and $aD_{12}=C_1B_2$, compute $\\tau_V$ and the two factors from (3.12), and check whether both factors are Schur functions whose product equals $\\tau_V$; a single such $V$ for which the factors leave the Schur class or fail to multiply back to $\\tau_V$ would disprove the characterization. The 3-by-3 Blaschke-product example in Section 5.2 is a concrete case where the formulas should recover the two Blaschke factors exactly.","tokens_in":23590,"feed_emoji":"🧩","tokens_out":10591,"duration_ms":84536,"temperature":0.7,"pith_summary":"The paper establishes a transfer-function criterion for when a Schur-Agler class function on the unit polydisc $\\mathbb{D}^n$ factors as a product of two Schur-Agler class functions. The criterion lives entirely on the isometric colligation that realizes the function: the colligation must have a block shape, called property $F(n)$, in which certain off-diagonal blocks vanish and others factor as products of the $C$ and $B$ rows. When the function does not vanish at the origin, the criterion is an if-and-only-if, and the paper gives explicit formulas for the two factors as transfer functions of reduced colligations. The same pattern classifies split-variable factorizations, where one factor depends on the first $m$ variables and the other on the rest, and it extends to multipliers of the Drury-Arveson space and to functions vanishing at the origin, where extra rigidity identities appear. This matters because factorization of bounded analytic functions in several variables is generally delicate, and the paper turns the question into checkable linear algebra.","feed_headline":"One block-matrix condition decides when Schur functions factor","feed_subtitle":"A Schur-Agler function factors exactly when its colligation has an explicit block form; the paper builds the factors.","key_machinery":"The load-bearing object is the isometric colligation $V = \\begin{bmatrix} a & B \\\\ C & D \\end{bmatrix}$ and its transfer function $\\tau_V(z) = a + B(I - E(z)D)^{-1} E(z) C$. A colligation satisfying property $F(n)$ has its $D$-block shaped so that the resolvent $(I - E(z)D)^{-1}$ becomes upper triangular after a flip, which makes the transfer function split into a product of two transfer functions. The factorization is produced by splitting each coordinate space into $M_i \\oplus N_i$, rearranging the direct sum with a flip operator, and normalizing the off-diagonal data with scalars $\\alpha$ and $\\beta$; the identities $|\\beta|^2 = |a|^2 + C(1)^*C(1)$ and $\\alpha = a/\\beta$ are exactly what force the two reduced blocks to be isometric colligations. The same machine, with different block conditions, handles split-variable factors and the vanishing-at-origin case.","core_discovery":"The central claim is Theorem 3.4: for $\\theta \\in SA(\\mathbb{D}^n)$ with $\\theta(0) \\neq 0$, $\\theta = \\phi\\psi$ for some $\\phi,\\psi \\in SA(\\mathbb{D}^n)$ if and only if $\\theta = \\tau_V$ for some isometric colligation $V$ satisfying property $F(n)$. Property $F(n)$ means that, after decomposing each coordinate space as $M_i \\oplus N_i$, the colligation has the block form with $D_{ij}(21) = 0$ and $a D_{ij}(12) = C_i(1) B_j(2)$ for all $i,j$, where $a = \\theta(0)$. The factors are not merely shown to exist; they are constructed explicitly as $\\tau_{V_1}$ and $\\tau_{V_2}$ for reduced colligations $V_1, V_2$ whose entries are normalized by scalars $\\alpha,\\beta$ satisfying $|\\beta|^2 = |a|^2 + C(1)^*C(1)$ and $\\alpha = a/\\beta$. The same equivalence, with property $F_m(n)$, classifies split-variable factorizations (Theorem 2.4), and Theorem 4.1 gives the vanishing-at-origin analogue on the disc, where the block conditions become rigidity identities such as $C_1 C_1^* D_2 = C_1^* C_1 D_2$ and $D_2 = XY$ with $X$ an isometry.","pith_inferences":["Because property $F(n)$ is written as explicit equations among the blocks, a symbolic or numerical solver could turn the characterization into a computational factorization test for rational Schur-Agler functions with finite-dimensional realizations.","The proof mechanism uses only the isometry equations of the colligation, so the same scheme is likely to extend to operator-valued Schur-Agler functions and to other multiplier algebras that admit transfer-function realization theorems; Theorem 5.3 already points in that direction.","A natural extension question, not addressed in the paper, is whether all minimal isometric realizations of a factorable function share the $F(n)$ block shape, or only the canonical one constructed here; if they do, the block shape would be an invariant of factorability rather than of a particular realization."],"forward_implications":["For any Schur-Agler function on the polydisc with nonzero value at the origin, deciding whether it factors into two Schur-Agler factors reduces to checking the explicit block conditions of property $F(n)$ on a realizing isometric colligation.","When the conditions hold, the two factors can be written down directly from the colligation via formulas (3.12), so the factorization is algorithmic rather than existential.","A Schur-Agler function on $\\mathbb{D}^n$ with nonzero value at the origin factors as $\\phi(z_1,\\ldots,z_m)\\psi(z_{m+1},\\ldots,z_n)$ exactly when its colligation satisfies property $F_m(n)$; in the two-variable case this detects products of two one-variable Schur functions.","In the vanishing-at-origin case, factorization forces rigid identities on the colligation: on the disc, one-factor-nonzero factorizations correspond to $C_1 C_1^* D_2 = C_1^* C_1 D_2$ with $C_1^*C_1 > 0$, and both-factors-vanishing factorizations correspond to $D_2 = XY$ with $X$ an isometry and $X^*D_1 = 0$.","The same transfer-function criterion classifies split-variable factorizations of multipliers in the unit ball of the Drury-Arveson space, as stated in Theorem 5.3."],"supporting_citations":[{"why":"supplies the transfer-function realization theorem for the Schur-Agler class on the polydisc that the whole characterization builds on.","marker":"[1]"},{"why":"provides the one-variable lurking-isometry realization and the transfer-function background that motivates Theorems 3.4 and 4.1.","marker":"[3]"},{"why":"gives the prior product-of-colligations construction, which the paper contrasts and extends with explicit reversible formulas.","marker":"[4]"},{"why":"establishes the von Neumann inequality for commuting pairs, so the two-variable specialization of the Schur-Agler theorem is a statement about ordinary Schur functions on the bidisc.","marker":"[6]"},{"why":"supplies the realization theorem for multipliers of the Drury-Arveson space used in the multiplier factorization section.","marker":"[8]"},{"why":"connects factorization of transfer functions to invariant subspaces and colligation products, serving as the earlier colligation-factorization reference.","marker":"[9]"},{"why":"shows the von Neumann inequality fails for three or more variables, so the Schur-Agler class is genuinely smaller than the Schur class there and the realization approach is needed.","marker":"[33]"}],"fun_headline_variants":["A single block form dictates Schur-Agler factorizations","When does a Schur function split? Check its colligation","Explicit factors from one colligation condition","Block-matrix test reveals Schur function factors","Factorize Schur functions via colligation block form"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The converse of the main equivalence assumes that the two reduced block matrices built from a colligation satisfying property $F(n)$ are themselves isometric colligations whose product recovers the original colligation; the entire sufficiency direction rests on that algebraic identity.","fun_headline_variants_meta":{"raw":{"variants":["A single block form dictates Schur-Agler factorizations","When does a Schur function split? Check its colligation","Explicit factors from one colligation condition","Block-matrix test reveals Schur function factors","Factorize Schur functions via colligation block form"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000257,"raw_usage":{"total_tokens":1728,"prompt_tokens":1246,"completion_tokens":482,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":862,"completion_tokens_details":{"reasoning_tokens":405}},"tokens_in":862,"tokens_out":482,"duration_ms":4441,"temperature":1.0,"reasoning_tokens":405,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:03:23.241035+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take any explicit isometric colligation $V$ of the form with $a \\neq 0$, $D_{21}=0$, and $aD_{12}=C_1B_2$, compute $\\tau_V$ and the two factors from (3.12), and check whether both factors are Schur functions whose product equals $\\tau_V$; a single such $V$ for which the factors leave the Schur class or fail to multiply back to $\\tau_V$ would disprove the characterization. The 3-by-3 Blaschke-product example in Section 5.2 is a concrete case where the formulas should recover the two Blaschke factors exactly.","supporting_citations":[{"cited_title":"Agler, On the representation of certain holomorphic functions defi ned on a polydisc , Topics in operator theory: Ernst D","cited_arxiv_id":null,"evidence_quote":"supplies the transfer-function realization theorem for the Schur-Agler class on the polydisc that the whole characterization builds on."},{"cited_title":"Agler and J","cited_arxiv_id":null,"evidence_quote":"provides the one-variable lurking-isometry realization and the transfer-function background that motivates Theorems 3.4 and 4.1."},{"cited_title":"Alpay, A","cited_arxiv_id":null,"evidence_quote":"gives the prior product-of-colligations construction, which the paper contrasts and extends with explicit reversible formulas."},{"cited_title":"Andˆ o,On a pair of commutative contractions , Acta Sci","cited_arxiv_id":null,"evidence_quote":"establishes the von Neumann inequality for commuting pairs, so the two-variable specialization of the Schur-Agler theorem is a statement about ordinary Schur functions on the bidisc."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the realization theorem for multipliers of the Drury-Arveson space used in the multiplier factorization section."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"connects factorization of transfer functions to invariant subspaces and colligation products, serving as the earlier colligation-factorization reference."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"shows the von Neumann inequality fails for three or more variables, so the Schur-Agler class is genuinely smaller than the Schur class there and the realization approach is needed."}],"review_version":1}