{"id":"7f371223-4fed-41d2-901c-481b3c7d233f","arxiv_id":"1908.03108","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The Selberg zeta-function of a compact hyperbolic surface is pseudo-prime and right-prime, meaning its only nontrivial decompositions are severely restricted.","lead":"This paper proves that the Selberg zeta-function, an analytic object built from the geometry of a compact curved surface, is pseudo-prime and right-prime: it cannot be factored non-trivially. The result is a step toward showing that this zeta-function is prime, continuing a program that already covered the Riemann zeta-function and the Gamma function.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Prop. 5's preimage-ray computation is wrong: for h(s)=a_d s^d+..., the rays L_j should be (π/2−2arg a_d)/d+2jπ/d, and 'arg L_j≠π' fails, e.g. for d=2, arg a_d=π/4; the resulting negative-real component spoils the Z→1 vs Z→∞ contradiction.","rationale":"Although the reader's three flagged defects are genuine, they are localized: the Lemma 7 constant needs a factor 1/π but the intended bound 0.41 is correct; 'Lemma 11' is a reference typo for Lemma 4; and the Corollary 2 angle is a typo. The more load-bearing problem is in Proposition 5, the only proof of right-primeness. The paper chooses a half-line ℓ with direction π/2−arg a_d and claims its preimages tend to rays L_j with directions π/(2d)−(arg a_d)/d+2jπ/d and that none of these is the negative real axis. The direction formula is inconsistent with the stated ℓ: since h(s)∼a_d s^d, the condition arg h(s)=π/2−arg a_d gives d arg s = π/2−2 arg a_d (mod 2π). More importantly, the intended conclusion is false in a generic case: for h(s)=e^{π i/4}s^2+... the preimages of ℓ are asymptotically the positive and negative real axes, so one preimage curve is not inside A and carries no Z→∞ limit; along the negative real axis Z has zeros at every negative integer and grows only at the midpoints. Hence the contradiction (7) is not established for degree 2 leading coefficients, and the right-primeness step is incomplete. The theorem may well be true and the gap is likely repairable by rotating ℓ or by using the midpoint growth from Lemma 7, but the proof needs that repair. Because the main claim still appears correct and the gap is localized, the appropriate verdict remains conditional pending a corrected Proposition 5; this does not change the reader's verdict.","tokens_in":6912,"tokens_out":32014,"duration_ms":329289,"concrete_test":"Recompute the preimage directions in Prop. 5 with the standard formula α_j=(θ−arg a_d+2jπ)/d for the choice θ=π/2−arg a_d. For h(s)=e^{π i/4}s^2 (so d=2, arg a_2=π/4), verify that the two preimage rays of ℓ={arg s=π/4} are the positive and negative real axes; the negative-real curve is not contained in A and Z does not tend to ∞ along it (Z(−n)=0 while Z(−n−1/2)→∞). Then check whether replacing ℓ by a half-line with θ=0 restores a preimage in σ>2 and one in A. If the repaired choice works, Proposition 5 needs an extra rotation argument before the published proof is complete.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Proposition 5 the proof of right-primeness chooses the half-line ℓ={arg s=π/2−arg a_d} and asserts that its preimage components under h(s)=a_d s^d+... are asymptotic to L_j={arg s=π/(2d)−(arg a_d)/d+2jπ/d}, with arg L_j≠π. This angle formula is incorrect: writing h(s)∼a_d s^d and s=re^{iα}, the condition arg h(s)=θ=π/2−arg a_d gives dα+arg a_d≡θ (mod 2π), hence α_j=(π/2−2 arg a_d+2jπ)/d, not (π/2−arg a_d+2jπ)/d. More importantly, the conclusion 'arg L_j≠π' is false. For d=2 and arg a_d=π/4, the correct preimage rays of ℓ are at angles 0 and π, so one component is the negative real axis. That component is not contained in the set A (which excludes a neighbourhood of the negative real axis), and Z(s) does not tend to ∞ along the negative real axis: Z(−n)=0 for every positive integer n, while the midpoints Z(−n−1/2) grow by Lemma 7. Hence the two limits in (7) cannot be obtained for this preimage, and the contradiction proving d=1 is not established. Since d=2 is a central case for Theorem 1 (quadratic inner factor), this is a load-bearing gap, not just a typographical slip. The gap appears repairable by rotating the chosen half-line ℓ so that no preimage ray lands on the negative real axis, but the published argument needs that additional step.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the Selberg zeta-function Z(s) attached to a compact Riemann surface of genus g≥2 in the sense of decomposition (factorization) of meromorphic functions. Theorem 1 claims that Z is pseudo-prime and right-prime, and that if Z(s)=f(h(s)) with f rational and h meromorphic, then f is a polynomial whose degree divides 2g−2 and h is entire. Corollary 2 claims a rigidity property for entire maps sending the nontrivial zero set of one Selberg zeta-function to that of another. The proofs combine a criterion of Liao and Yang on accumulation lines of value sets, growth estimates for Z and the functional-equation factor X(s), and zero-counting on circles centered at negative integers.","tokens_in":7283,"tokens_out":14145,"duration_ms":158029,"significance":"If the main theorem is correct, it gives a strong rigidity statement: every nontrivial decomposition of the Selberg zeta-function has a very restricted outer or inner component, and in the rational outer case the degree is constrained by the genus through 2g−2. The approach is structurally attractive because it avoids special-function identities and instead uses sectorial growth contrasts plus Rouché-type counting. The paper also has a clear falsifiable consequence, the divisibility constraint on the degree of a polynomial outer component. The main ideas are sound and the argument is largely self-contained modulo the authors' earlier estimates, but the proof of right-primeness contains a concrete geometric error that must be repaired before the central claim is fully established.","major_comments":[{"comment":"The preimage-ray computation in the proof of right-primeness is incorrect. For h(s)=a_d s^d+..., the half-line ℓ={arg s=π/2−arg a_d} has preimage rays asymptotic to L_j with angles α_j=(π/2−2 arg a_d+2πj)/d, not (π/2−arg a_d+2πj)/d as stated. Consequently the assertion 'arg L_j ≠ π' is false in general. For d=2 and arg a_d=π/4, the correct preimage rays of ℓ are the positive and negative real axes; the negative real axis is not contained in the sector A, and along that ray Z does not tend to ∞ (it vanishes at every positive integer, while the growth statements in (5) and (6) do not apply). Thus the two limits in (7) cannot both be obtained for this preimage, and the contradiction proving d=1 is not established. Since d=2 is the central case for a quadratic inner factor, this is a load-bearing gap. The gap is repairable by choosing the half-line ℓ with a generic angle so that none of its preimage rays lands on the negative real axis and at least one lies eventually in σ>2 and one in A, but that additional step is absent.","section":"§2, Proposition 5"},{"comment":"The numerical estimate in case 3 is false as printed: the inequality √3/4 + log(1−e^{−2π√3/4}) ≥ 0.41 is incorrect, since the left-hand side is approximately 0.365. The lemma itself survives because the weaker positive lower bound 0.36 still exceeds the stated δ=0.007. Nevertheless, this is a computational premise in a lemma that is used both in Lemma 8 and in the right-primeness growth argument, so the incorrect estimate should be corrected or replaced by a valid bound.","section":"§2, Lemma 7, case 3"},{"comment":"The proof of Corollary 2 is incomplete as written. The set E is said to have accumulation lines arg s=π/2 and 3π/4, but the zeros of the Selberg zeta-function lie essentially on the line σ=1/2, whose accumulation angles are π/2 and 3π/2 (mod 2π), not 3π/4. More importantly, the final step 'Therefore g(s)=bs and b=1' is not justified: after Lemma 9 gives deg g≤2, one must prove that a polynomial of degree at most two that maps all but finitely many points of the line σ=1/2 into the same line must be the identity. This is plausible, but the argument is missing and the notation is confusing. Since the corollary is a stated result, this gap should be addressed.","section":"§2, proof of Corollary 2"}],"minor_comments":[{"comment":"The proof refers to 'Lemma 11' when citing the asymptotic for the factor X(s); this should be Lemma 4.","section":"§2, Proposition 5"},{"comment":"The sentence 'Then p is a linear function since F is a right-prime function' is too terse. As written, right-primeness does not apply directly to the displayed decomposition F=f(h) with rational f and transcendental h. The intended argument must re-compose as F(z)=G(p(z)) with G(z)=f(e^z−w) and then apply right-primeness to that decomposition; this step should be stated explicitly.","section":"§2, Lemma 6"},{"comment":"The displayed equality in case 4 appears to be a typographical repetition of the same expression on both sides; the intended lower bound should be written as an inequality.","section":"§2, Lemma 7, case 4"},{"comment":"The angle '3π/4' should presumably be '3π/2', and the reuse of the symbol b for both the Hadamard constant and the coefficient of g is confusing.","section":"§2, proof of Corollary 2"}],"recommendation":"major_revision","confidential_remarks":"The main theorem is likely correct and the errors found are repairable within the manuscript's scope: the right-primeness proof needs the additional step of choosing the half-line ℓ generically, and Lemma 7 needs a corrected numerical estimate. The Corollary 2 proof needs substantial elaboration. The paper is a short note that relies heavily on the authors' earlier work, which is acceptable here because those results are independently established. I would not recommend rejection, but the current text does not yet fully prove the stated claims."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my read. The genuinely new thing is Theorem 1: the Selberg zeta-function of a compact surface is pseudo-prime and right-prime, and in the rational-component case the polynomial degree divides 2g−2. That's a real addition to the short list of zeta-functions with primeness properties, and the degree-divisibility statement is new. The proof strategy is sensible: pseudo-primeness comes from Liao–Yang's lemma, right-primeness from the contrast between Z→1 in the right half-plane and Z→∞ in the excluded sectors, and Lemma 8 uses Rouché on the trivial-zero circles to force the divisibility. The paper is concise and mostly readable.\n\nBut the right-primeness step as written does not work. In Proposition 5 the half-line ℓ is defined as {arg s = π/2 − arg a_d} but the preimage rays L_j are computed as if ℓ had angle π/2. The correct preimage angles of the stated ℓ would be (π/2 − 2 arg a_d + 2jπ)/d. More importantly, the assertion 'arg L_j ≠ π' is false. Take d = 2 and h(s) = i s^2: the preimage of the positive imaginary axis contains the negative real axis, where Z has trivial zeros and certainly does not tend to ∞. The contradiction in (7) then isn't available for that preimage. This is load-bearing for the proof of right-primeness, though I think it's repairable: a generic rotation of the half-line avoids the excluded sector around the negative real axis. The intended dichotomy is right; the execution is missing a step.\n\nSmaller issues: Lemma 7, case 3, writes '√3/4 + log(...)' where the 1/π in front of the log is dropped; with the factor restored the estimate is true, so this is a typo, not a fatal flaw. The reference to 'Lemma 11' in Proposition 5 should be Lemma 4. Corollary 2's 'two accumulation lines, namely arg s = π/2 and 3π/4' looks like a typo for 3π/2, and the application of Lemma 3 with a_2 = ∞ to a function with no poles is underexplained. None of these undermine the central idea.\n\nThe self-citations to [4] and [5] are fine; those are independent growth results, not circular.\n\nMy verdict: the theorem is probably true and worth a serious referee, but the printed proof needs repair. I'd send it to review and tell the authors to fix the preimage computation in Proposition 5 before acceptance. It's the kind of paper a value-distribution person will want to know about.","headline":"First primeness result for the Selberg zeta-function, likely true but the proof as printed has a repairable gap in the right-primeness step plus several typos; send to a referee but require a fix before acceptance.","tokens_in":7839,"tokens_out":12193,"would_cite":false,"duration_ms":128746,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11M36"],"pacs":[],"model":"deepseek-v4-flash","headline":"The Selberg zeta-function of a compact Riemann surface is pseudo-prime and right-prime, forcing any rational outer factor in a decomposition to be a polynomial whose degree divides 2g−2 and the inner factor to be entire.","keywords":["Selberg zeta-function","compact Riemann surface","prime function","pseudo-prime","right-prime","function decomposition","value distribution","trivial zeros"],"falsifier":"A direct counterexample would be an explicit decomposition Z(s)=f(h(s)) with rational f of degree not dividing 2g−2 for some compact Riemann surface (e.g., a cubic outer factor for genus 2); short of that, evaluating the Lemma 7 bound at |s|=1/2, arg s=π/3 (where the paper's value 0.41 is wrong, the true value being about 0.365) and finding a non-positive value there would break Lemma 7 and with it the zero-counting proof of the divisibility claim.","tokens_in":6696,"feed_emoji":"🧩","tokens_out":10814,"duration_ms":103498,"temperature":0.7,"pith_summary":"This paper proves that the Selberg zeta-function associated with a compact Riemann surface of genus g is pseudo-prime and right-prime in the sense of value-distribution theory: it cannot be written as a nontrivial composition f(h(s)) with meromorphic f and h. In every such decomposition with f rational, f must be a polynomial of degree k dividing 2g−2, and h must be entire. This is the Selberg-zeta analogue of a known primeness result for the Riemann zeta-function, and it implies that the zero structure of Z is irreducible in a strong compositional sense. The proof combines a classical pseudo-primeness criterion with contrasting growth estimates of Z in different sectors, a zero-counting argument on circles around the trivial zeros, and a bound on the number of zeros of exponential-type inner factors.","feed_headline":"No hidden factor splits the Selberg zeta-function","feed_subtitle":"Rational outer factors must be polynomials of degree dividing the genus count, with entire inner factors.","key_machinery":"The proof is carried by three mechanisms: (i) a classical lemma from complex analysis stating that a finite-order meromorphic function whose two distinct a-value sets have finitely many accumulation lines is pseudo-prime, applied to the zeros and poles of Z; (ii) the contrasting growth behaviour of Z in the right half-plane (Z(s)→1 as σ→∞) and in a sector containing the negative real axis (Z(s)→∞), which forces any polynomial inner factor h to be linear by comparing preimages of a half-line; and (iii) a zero-counting and Rouché argument on the circles |s+n|=1/2 around the trivial zeros, using the fact that Z(s)→∞ on these circles, to show the degree k of a polynomial outer factor must divide 2g−2. Lemma 7, which supplies the growth on those circles, is reduced via the functional equation to a positive lower bound on t + (1/π) log|1−$e^{{2πis}}$| for |s|=1/2, proven by a four-case split.","core_discovery":"The central discovery is that the Selberg zeta-function Z(s) of a compact Riemann surface X of genus g is prime in the decomposition sense: any representation Z(s)=f(h(s)) with f and h meromorphic forces one component to be trivial. Specifically, Z is pseudo-prime (every decomposition has f rational or h a polynomial) and right-prime (h is linear whenever f is transcendental); and if f is rational then f is a polynomial of degree k dividing 2g−2 and h is entire. The argument shows that the only possible nontrivial inner factors would be exponentials of linear functions, which are ruled out by the quadratic growth of the zero-counting function of Z, and that polynomial outer factors of degree k would force the zero multiplicities (2g−2)(2n+1) at the trivial zeros s=−n to be divisible by k for all large n, which is possible only if k divides 2g−2.","pith_inferences":["The same growth-dichotomy plus zero-counting strategy may apply to other zeta-functions with an Euler product and a functional equation, such as Selberg zeta-functions for finite-volume non-compact surfaces, provided a suitable analogue of Lemma 7 holds.","In the hypothetical quadratic decomposition, the equality N(d,Z)=N(e,g)∪N(f,g) together with the clustering of a-points suggests that any inner function g would itself need to satisfy a functional equation of Selberg type, which would likely rule out such decompositions entirely.","The numerical error found in the bound of Lemma 7, while repairable, indicates that the proof is delicate at that point; a rigorous re-proof of Lemma 7 is needed before the theorem can be considered fully established.","Corollary 2 shows that the zero set of Z is a complete invariant under entire maps; a natural extension would be to ask whether meromorphic maps (rather than entire) also force h to be the identity."],"forward_implications":["The Selberg zeta-function of a compact Riemann surface is pseudo-prime: in every decomposition, either the outer map is rational or the inner map is a polynomial.","Any transcendental outer map forces the inner map to be linear, so the only possible nontrivial decompositions have a polynomial outer map of degree dividing 2g−2.","If two compact Riemann surfaces have Selberg zeta-functions whose nontrivial zeros are connected by an entire map h, then h is the identity and the zero sets coincide.","The value distribution of Z is prime in the sense that its zero sets and preimages cannot be written as unions of level sets of a simpler function, restricting attempts to model the nontrivial zeros by a composition.","The paper's concluding remark suggests the stronger and still open possibility that no nontrivial decomposition exists at all."],"supporting_citations":[{"why":"provides the proof of Lemma 3, the pseudo-primeness criterion applied to the zeros and poles of Z.","marker":"[1]"},{"why":"established the analogous primeness of the Riemann zeta-function and the strategy of applying Lemma 3 to zeta-functions.","marker":"[10]"},{"why":"supplies the definition of the Selberg zeta-function, its functional equation, the growth formula (5), and the location and multiplicities of its zeros.","marker":"[8]"},{"why":"gives the asymptotic expansion of the factor X(s) used as Lemma 4 to derive Z(s)→∞ in a sector.","marker":"[5]"},{"why":"introduces the antiderivative of z tan(πz) used in Lemma 7 to estimate the factor X(s).","marker":"[13]"},{"why":"provides Lemma 9, used in Corollary 2 to identify the inner function mapping zeros to zeros.","marker":"[3]"},{"why":"defines the notion of factorization and prime functions that the paper's main theorem builds on.","marker":"[7]"}],"fun_headline_variants":["Why the Selberg zeta-function is indecomposable","Selberg zeta is pseudo-prime and right-prime","Selberg zeta refuses nontrivial factorization","Selberg zeta has no nontrivial factor","Selberg zeta-function: prime under decomposition"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof hangs on Lemma 7's claim that Z(s) becomes unbounded on the circles |s+n|=1/2 as n→−∞, and that claim rests on a numerical lower-bound check for t + (1/π) log|1−$e^{{2πis}}$| on |s|=1/2 that is computed incorrectly in one of the four angular cases, leaving a gap that must be repaired for the proof to be complete.","fun_headline_variants_meta":{"raw":{"variants":["Why the Selberg zeta-function is indecomposable","Selberg zeta is pseudo-prime and right-prime","Selberg zeta refuses nontrivial factorization","Selberg zeta has no nontrivial factor","Selberg zeta-function: prime under decomposition"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00116,"raw_usage":{"total_tokens":4698,"prompt_tokens":734,"completion_tokens":3964,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":350,"completion_tokens_details":{"reasoning_tokens":3887}},"tokens_in":350,"tokens_out":3964,"duration_ms":32196,"temperature":1.0,"reasoning_tokens":3887,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:28:18.182944+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct counterexample would be an explicit decomposition Z(s)=f(h(s)) with rational f of degree not dividing 2g−2 for some compact Riemann surface (e.g., a cubic outer factor for genus 2); short of that, evaluating the Lemma 7 bound at |s|=1/2, arg s=π/3 (where the paper's value 0.41 is wrong, the true value being about 0.365) and finding a non-positive value there would break Lemma 7 and with it the zero-counting proof of the divisibility claim.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"provides the proof of Lemma 3, the pseudo-primeness criterion applied to the zeros and poles of Z."},{"cited_title":"Liao and C.-C","cited_arxiv_id":null,"evidence_quote":"established the analogous primeness of the Riemann zeta-function and the strategy of applying Lemma 3 to zeta-functions."},{"cited_title":"Hejhal, The Selberg trace formula for P SL(2,R)","cited_arxiv_id":null,"evidence_quote":"supplies the definition of the Selberg zeta-function, its functional equation, the growth formula (5), and the location and multiplicities of its zeros."},{"cited_title":"Garunkˇstis and R","cited_arxiv_id":null,"evidence_quote":"gives the asymptotic expansion of the factor X(s) used as Lemma 4 to derive Z(s)→∞ in a sector."},{"cited_title":"Randol , The Riemann hypothesis for Selberg’s zeta-function and the asymptotic behav- ior of eigenvalues of the Laplace operator , Trans","cited_arxiv_id":null,"evidence_quote":"introduces the antiderivative of z tan(πz) used in Lemma 7 to estimate the factor X(s)."},{"cited_title":"Edrei , Meromorphic functions with three radially distributed val ues, Amer","cited_arxiv_id":null,"evidence_quote":"provides Lemma 9, used in Corollary 2 to identify the inner function mapping zeros to zeros."},{"cited_title":"Gross , On factorization of meromorphic functions , Trans","cited_arxiv_id":null,"evidence_quote":"defines the notion of factorization and prime functions that the paper's main theorem builds on."}],"review_version":1}