{"id":"fc91db11-a769-40e2-a674-8c790248b70f","arxiv_id":"1908.07571","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Every sufficiently large iterate of a Thurston map without Levy cycles and not covered by a torus endomorphism is isotopic to the subdivision map of a finite subdivision rule; the torus-covered case is classified by the eigenvalues of the affine lift.","lead":"This paper proves that almost every Thurston map, a broad family of sphere maps from complex dynamics, becomes a finite subdivision rule after iterating enough times. It also pins down exactly which torus-covered exceptions do or do not have this property, and gives conditions that forbid it.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 2.6 depends on an unproved local connectivity assertion for immediate basins of Böttcher expanding Thurston maps; if it fails, the finite graph G need not exist.","rationale":"The reader's weakest_assumption identifies the same load-bearing concern: the continuous extension of the Böttcher coordinates and local connectivity of immediate basin boundaries. I agree that this is the most critical point in the proof of Theorem 2.6, because the finite graph G and the sets B and C are constructed from rays and arcs in these basins, and the final isotopy argument uses the boundary behavior near ∂F_p. The paper presents the local connectivity assertion as a carry-over of the Douady–Hubbard argument and points to [2, Lemma 4.7] for support, but it supplies no proof and does not explicitly verify that the cited result applies to the full generality of Böttcher expanding Thurston maps. This is not a disagreement with the mathematical consensus; it is a correctness risk in a key step. I do not see a more serious flaw elsewhere: the finiteness lemmas, the CW-complex construction, and the nonrealizability arguments are coherent conditional on this basin-boundary structure. The minor issues noted by the reader (the strength of Theorem 3.1(1) and the unproved arc condition in Example 4.2) are real but secondary. Therefore the conditional verdict stands unchanged.","tokens_in":19147,"tokens_out":20911,"duration_ms":644596,"concrete_test":"Read [2, Lemma 4.7] and the discussion preceding it in [2, Section 4.2], and check that the lemma proves the continuous extension of ψ_p to D and surjectivity of the boundary map onto ∂F_p for every Böttcher expanding Thurston map in the sense of [2, Definition 4.1], with no extra hypotheses such as rationality. If the lemma requires extra hypotheses, identify a Böttcher expanding Thurston map failing them and test whether its immediate basin boundary is locally connected; this would settle whether the graph construction in Section 2 is valid.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 2.6 is proved by constructing a finite graph G whose 1-skeleton supports a finite subdivision rule. The construction of G uses, for each periodic critical point p, a Böttcher homeomorphism ψ_p : D → F_p that must extend continuously to the closed disk, with every boundary point of F_p the landing point of at least one ray. The paper asserts that the Douady–Hubbard proof of local connectivity for subhyperbolic polynomials 'carries over in this setting', citing [2, Lemma 4.7]. This is the load-bearing step: Lemmas 2.3, 2.4 and 2.5, the finiteness of the arc sets B and C, and the final isotopy near ∂F_p all depend on it. Without local connectivity, the finite graph G invariant up to isotopy is not shown to exist, and Theorem 2.6 has no proof. The gap is internal because the paper does not reproduce the argument, and the cited lemma is not verified to cover the full class of Böttcher expanding Thurston maps, including non-rational maps and maps with periodic critical points.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies which Thurston maps have large iterates isotopic to subdivision maps of finite subdivision rules. The main theorem (Theorem 2.6) asserts that every sufficiently large iterate of a Thurston map that is not doubly covered by a torus endomorphism and has no Levy cycle is isotopic to the subdivision map of a finite subdivision rule. The proof reduces to the Böttcher expanding case via the Bartholdi–Dudko characterization, then constructs a finite graph G whose 1-skeleton supports a finite subdivision rule and proves that the nth pullback of a suitable cell structure is isotopic to G. Section 3 treats Thurston maps with Euclidean orbifold of signature (2,2,2,2), giving a trichotomy in terms of the eigenvalues of the affine lift. Section 4 gives nonrealizability conditions: no iterate is Thurston equivalent to a subdivision map if the pullback relation has wandering arcs or wandering univalent curves, with examples.","tokens_in":19341,"tokens_out":11134,"duration_ms":104955,"significance":"If the main theorem is correct, it is a substantial advance: after passing to an iterate and allowing isotopy, finite subdivision rules capture essentially all Thurston maps except those with Levy cycles and contracting torus covers. The proof strategy is coherent, and the paper is explicit about its dependence on the deep Bartholdi–Dudko theorem and on earlier work in [12] for the fat path distance lemma. The nonrealizability criteria in Section 4 are concrete and falsifiable. However, several load-bearing topological claims, especially the local connectivity of immediate basins and the existence of landing rays, are asserted rather than fully justified, and these need to be substantiated before the central theorem can be considered established.","major_comments":[{"comment":"The assertion that each ψ_p: D → F_p extends continuously to D and that every boundary point of F_p is the landing point of at least one ray is stated by saying that the Douady–Hubbard proof 'carries over', citing [2, Lemma 4.7] and Milnor's paper. This assertion is load-bearing: the finiteness of the set B (Lemma 2.3), the definition of B using rays landing at every p∈P^1_f, and the final isotopy near ∂F_p in the proof of Theorem 2.6 all rely on ray landing. The manuscript does not reproduce the argument, and the exact statement and hypotheses of [2, Lemma 4.7] are not given. Please either prove the local connectivity and ray-landing statement or quote a theorem that covers the full class of Böttcher expanding Thurston maps considered here, including non-rational maps and maps with periodic critical points.","section":"Section 2, 'Fatou and Julia sets' (before Lemma 2.3)"},{"comment":"The proof that only finitely many rays land at a preperiodic boundary point is compressed at the point where an infinite sequence R_1,R_2,... of rays is said to converge to a ray in R fixed by f, and then Milnor's Lemma 18.12 is invoked to conclude finiteness. The convergence to a ray in R and the applicability of Milnor's lemma in the present non-rational setting need to be justified. If the sequence of rays can converge to a non-landing ray, or if Milnor's lemma has hypotheses not satisfied here, then the finiteness of B is not established, and the construction of the finite graph G in Theorem 2.6 would fail.","section":"Section 2, Lemma 2.3"},{"comment":"The contradiction argument relies on the fat path distance nonincreasing property of lifts of inverse maps of subdivision maps, quoted from [12, Lemma 6.1]. It should be explicitly verified that the lift G^{-1} of g^{-1} satisfies the hypotheses of that lemma, in particular that the lifted initial cell structure on the covering space is such that G^{-1} maps initial tiles to initial tiles. As written, the proof says only that 'the discussion at the beginning of this section applies'; since this is the mechanism that produces the contradiction, the verification should be spelled out.","section":"Section 3, Theorem 3.1(3)"}],"minor_comments":[{"comment":"The open unit disk is denoted D, but in Lemma 2.3 a closed topological disk bounded by two rays is also denoted D. Rename one of these to avoid ambiguity.","section":"Throughout"},{"comment":"In the definition of P^0_f, the expression 'p < union' appears to be a typo; it should read 'p ∉ union'.","section":"Section 2, paragraph after Theorem 2.1"},{"comment":"The symbols used for the union of inessential and peripheral classes ('⊙' and the symbol following it) are not rendered cleanly in the arXiv text; please use readable notation and define all symbols explicitly.","section":"Section 4, opening"},{"comment":"The sentence 'Because α0,α1,α2,... are mutually not homotopic rel Pf, the order of Pf must be at least 4' is not justified and appears too strong; the subsequent statement that the universal cover is the open disk only needs at least three points. Please correct or justify this.","section":"Section 4, proof of Theorem 4.1"},{"comment":"The proof of Theorem 2.6 is very long and refers repeatedly to Figures 1–5; the submitted manuscript should ensure all figures are included and clear, since several topological arguments (e.g., the construction of the disk Δ_p and the disk D in Lemma 2.5) are described with reference to these figures.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The main mathematical risk is the local connectivity and ray-landing assertion in Section 2. If [2, Lemma 4.7] indeed contains the needed theorem for all Böttcher expanding Thurston maps, then the requested changes are mostly clarificatory and the paper should be acceptable after revision. If not, the proof of Theorem 2.6 is incomplete. I recommend asking the authors to make the cited result precise and to expand the arguments around Lemmas 2.3 and 3.1(3) before acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Let me give you the short version. The paper proves that sufficiently large iterates of Thurston maps without Levy cycles (and not doubly covered by an expanding torus endomorphism) are isotopic to subdivision maps of finite subdivision rules. That is a genuinely broad realizability result, and the proof is coherent: reduce to Bottcher expanding via Bartholdi-Dudko, build a finite graph, then isotope the subdivision complex into its nth pullback. The finiteness lemmas (2.3–2.5) are the heart, and they mostly hold together. The classification of (2,2,2,2) maps in Theorem 3.1 is a clean addition, and the wandering-arc/curve criterion in Theorem 4.1 is a genuinely useful nonrealizability tool.\n\nThere are two soft spots, both minor. First, Theorem 3.1(1) says 'is the subdivision map' but the proof cites Theorem 2.6, which only gives 'isotopic to'. Unless there is an unstated upgrade for the Euclidean case, the statement overreaches. Easy fix. Second, Example 4.2 asserts the arc condition of Theorem 4.1 without showing the computation. A referee should ask for at least a sketch.\n\nThe stress-test note about local connectivity of immediate basins is, in my reading, a citation rather than a gap. The authors explicitly say the Douady–Hubbard argument carries over and point to [2, Lemma 4.7] and Milnor's discussion. Since the entire Bottcher-expanding framework is imported from [2], the local connectivity is part of that framework. I can't verify the reference without pulling [2], but nothing in the text suggests an unproved claim internal to this paper. If I were refereeing, I'd ask them to quote the lemma, but I would not call this a load-bearing flaw.\n\nThe citation pattern is fine: they use their own earlier work for the fat path distance and for Lemma 2.2, but those are published, non-circular results. The reliance on Bartholdi-Dudko is appropriate; it's a deep external theorem doing real work.\n\nBottom line: this is a solid, serious paper for anyone working on Thurston maps, finite subdivision rules, or expanding maps. The main theorem is a significant step toward the iterate question, and the proof is detailed enough that experts can check it. I'd send it to a strong referee. The minor issues should be fixed in revision, but none of them undermines the central result.","headline":"Genuinely broad realizability theorem with a clean proof; the local-connectivity worry is a citation, not a gap, and the only real blemish is an overstatement in Theorem 3.1(1).","tokens_in":19871,"tokens_out":4118,"would_cite":true,"duration_ms":103648,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["37F10","52C20","57M12"],"pacs":[],"model":"deepseek-v4-flash","headline":"Every sufficiently large iterate of a Thurston map with no Levy cycle and no contracting torus lift is isotopic to the subdivision map of a finite subdivision rule, making finite subdivision rules the generic combinatorial description of…","keywords":["finite subdivision rule","Thurston map","Böttcher expanding","Levy cycle","torus endomorphism","postcritically finite","cellular Markov partition","isotopy"],"falsifier":"Look for a Thurston map satisfying the hypotheses of Theorem 2.6 (no Levy cycle, and if it lifts to a torus endomorphism the affine eigenvalues lie outside the unit circle) whose pullback relation on arcs has a wandering orbit, as defined in Section 4. Theorem 4.1 would then imply that no iterate is Thurston equivalent to a subdivision map, contradicting Theorem 2.6; if such a map exists, the proof must fail at the local-connectivity or finiteness step.","tokens_in":18923,"feed_emoji":"🧩","tokens_out":13190,"duration_ms":112639,"temperature":0.7,"pith_summary":"Finite subdivision rules are combinatorial recipes that tile the sphere and then replace each tile by a finite pattern, refining the tiling. This paper proves that for any Thurston map (a postcritically finite branched self-cover of the sphere) with no Levy cycle and not doubly covered by a contracting torus endomorphism, every sufficiently large iterate is isotopic --- continuously deformable keeping the postcritical points fixed --- to the subdivision map of a finite subdivision rule. For maps that are double covers of torus endomorphisms, the torus-covered case is settled completely: expanding eigenvalue matrices give subdivision maps on all large iterates, an eigenvalue $\\pm 1$ gives a one-tile-type rule, and a contracting eigenvalue rules out every iterate. The paper also gives a general non-realizability condition: if the pullback relation on arcs has a wandering orbit, or on curves has a wandering univalent orbit, then no iterate is Thurston equivalent to a subdivision map. The upshot is that finite subdivision rules are not a narrow special class but the generic combinatorial description of expanding Thurston dynamics.","feed_headline":"No Levy cycle, expanding torus lift: some iterate becomes a subdivision map","feed_subtitle":"Essentially every expanding Thurston map agrees with a finite subdivision rule after passing to an iterate.","key_machinery":"The load-bearing mechanism is the Böttcher-expanding property: a complete length metric on the sphere minus the attracting periodic critical cycles that is uniformly compressed by $f$, with each attracting cycle locally modeled on $z\\mapsto z^k$. The proof also relies on local connectivity of immediate basins, which makes each basin a closed topological disk whose boundary is completely visible by rays, so boundary points can be manipulated by ray-tails and equipotentials. On that structure the paper assembles three finite sets of curves $A$, $B$, $C$: equipotentials plus postcritical points, ray tails, and connecting arcs satisfying ten explicit combinatorial conditions; their union is a graph $G$ invariant up to isotopy. Pulling back the cell complex induced by $G$ under $f^n$ and adding subdivisions inside the Fatou disks produces the finite subdivision rule, with finiteness lemmas --- finitely many rays landing at a preperiodic boundary point, at most two accessible boundary points from outside, and finitely many separating boundary points --- closing the construction.","core_discovery":"The central claim is Theorem 2.6: if $f$ is a Thurston map such that (i) it has no Levy cycles when it does not lift to a torus endomorphism, and (ii) when it does lift, the associated affine map $x\\mapsto Ax+b$ has eigenvalues outside the unit circle, then every sufficiently large iterate $f^n$ is isotopic rel the postcritical set to the subdivision map of a finite subdivision rule. The proof constructs a finite graph $G\\subseteq S^2$ containing the postcritical set, built from three finite families of curves --- equipotentials, ray tails in immediate basins, and connecting arcs --- and shows $G$ is invariant up to isotopy. The Böttcher-expanding property then forces the pullbacks of the induced cell structure to have arbitrarily small tiles, and subdividing the Fatou disks yields the required finite subdivision rule. Theorem 3.1 completes the torus-covered case: eigenvalues outside the unit circle put every sufficiently large iterate in the subdivision-map class, an eigenvalue $\\pm 1$ makes $f$ itself a subdivision map with one tile type, and a contracting eigenvalue forbids every iterate. Theorem 4.1 supplies an obstruction from pullback dynamics: wandering orbits of arcs, or of curves with univalent pullback, imply that no iterate is Thurston equivalent to a subdivision map.","pith_inferences":["The construction is sufficiently explicit that, given a Böttcher expanding map, a finite subdivision rule for a large iterate could in principle be computed; making the finiteness constants effective would turn the theorem into an algorithm for producing cellular Markov partitions.","The theorem suggests the obstruction to being a subdivision map, as opposed to an iterate of one, is concentrated in the early dynamics: a map that fails must fail before its Böttcher-expanding behavior dominates, so the class of maps with subdivision-map iterates is the closure of the expanding class under finite iterates.","The fat-path distance argument behind Theorem 3.1 is a general necessary condition --- any map with a subdivision-map iterate has bounded pullback complexity --- and could be applied to maps outside the theorem's scope, such as those with Levy cycles, to test for non-realizability.","The wandering-orbit criterion, combined with canonical-decomposition constructions, suggests a systematic family of non-realizable maps: take an elliptic piece with at least four marked points and twist it by a pseudo-Anosov map so that curves inside the piece wander under pullback; such maps would lie just beyond the boundary of Theorem 2.6."],"forward_implications":["For every map satisfying the two hypotheses of Theorem 2.6, all sufficiently large iterates are isotopic rel postcritical set to subdivision maps of finite subdivision rules; finite subdivision rules are thus universal for the expanding side of Thurston dynamics up to isotopy and passing to an iterate.","A map doubly covered by an affine torus endomorphism with eigenvalues of absolute value greater than 1 has every sufficiently large iterate equal (not just isotopic) to a subdivision map; eigenvalue $\\pm 1$ yields a subdivision map with one tile type.","A map doubly covered by an affine torus endomorphism with an eigenvalue of absolute value less than 1 has no iterate that is Thurston equivalent to a subdivision map, completing the classification of this exceptional family.","Wandering orbits under the pullback relation on arcs, or wandering univalent orbits on curves, form a necessary obstruction: no iterate of such a map is Thurston equivalent to a subdivision map.","The proof produces, for each large $n$, a finite cell structure $\\mathcal{S}$ whose pullback by $f^n$ refines it, so each such map carries explicit cellular Markov partitions."],"supporting_citations":[{"why":"Supplies the cited theorem that every Thurston map with no Levy cycles and not doubly covered by a torus endomorphism is Böttcher expanding; this is the entry point for Theorem 2.6.","marker":"[1]"},{"why":"Defines Böttcher expanding maps and supplies the local-connectivity of immediate basins (Lemma 4.7) and the Fatou/Julia decomposition used throughout Section 2.","marker":"[2]"},{"why":"Provides the combinatorial topology lemma (arc in the 1-skeleton of a disk connecting three vertices) used in the proof of Theorem 2.6, along with the original method of constructing subdivision rules from rational maps.","marker":"[6]"},{"why":"Establishes the one-tile-type subdivision rule for expanding maps covered by torus endomorphisms, underpinning the remark after Theorem 3.1 and the comparison in the torus-covered case.","marker":"[7]"},{"why":"Supplies the local-connectivity proof for filled Julia sets of subhyperbolic polynomials, cited in the paper to show the conjugating maps extend continuously to the closed disk.","marker":"[11]"},{"why":"Provides the fat-path distance nonincreasing lemma for subdivision maps, the tool used in the proofs of Theorems 3.1 and 4.1.","marker":"[12]"},{"why":"Gives the canonical decomposition and pullback relation framework that motivates and supports the wandering-orbit non-realizability condition in Theorem 4.1.","marker":"[19]"}],"fun_headline_variants":["No Levy cycle, expanding lift: some iterate becomes a subdivision map","Expanding torus lifts: iterates become subdivision maps, contracting never","Thurston maps with no Levy cycles: iterates yield finite subdivision rules","Every large iterate of an expanding Thurston map is a subdivision map"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that every immediate basin of attraction is a closed topological disk with a locally connected boundary, so that the conjugating maps extend to the closed disk and every boundary point is the landing point of at least one ray; if some basin boundary failed to be locally connected, the finite sets of arcs used to build the invariant graph $G$ need not exist and the proof of Theorem 2.6 would collapse.","fun_headline_variants_meta":{"raw":{"variants":["No Levy cycle, expanding lift: some iterate becomes a subdivision map","Expanding torus lifts: iterates become subdivision maps, contracting never","Thurston maps with no Levy cycles: iterates yield finite subdivision rules","Every large iterate of an expanding Thurston map is a subdivision map"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00033,"raw_usage":{"total_tokens":1808,"prompt_tokens":885,"completion_tokens":923,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":501,"completion_tokens_details":{"reasoning_tokens":845}},"tokens_in":501,"tokens_out":923,"duration_ms":9697,"temperature":1.0,"reasoning_tokens":845,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:04:42.267584+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Look for a Thurston map satisfying the hypotheses of Theorem 2.6 (no Levy cycle, and if it lifts to a torus endomorphism the affine eigenvalues lie outside the unit circle) whose pullback relation on arcs has a wandering orbit, as defined in Section 4. Theorem 4.1 would then imply that no iterate is Thurston equivalent to a subdivision map, contradicting Theorem 2.6; if such a map exists, the proof must fail at the local-connectivity or finiteness step.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the cited theorem that every Thurston map with no Levy cycles and not doubly covered by a torus endomorphism is Böttcher expanding; this is the entry point for Theorem 2.6."},{"cited_title":"Expanding maps , Trans","cited_arxiv_id":null,"evidence_quote":"Defines Böttcher expanding maps and supplies the local-connectivity of immediate basins (Lemma 4.7) and the Fatou/Julia decomposition used throughout Section 2."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the combinatorial topology lemma (arc in the 1-skeleton of a disk connecting three vertices) used in the proof of Theorem 2.6, along with the original method of constructing subdivision rules from rational maps."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the one-tile-type subdivision rule for expanding maps covered by torus endomorphisms, underpinning the remark after Theorem 3.1 and the comparison in the torus-covered case."},{"cited_title":"Douady and J","cited_arxiv_id":null,"evidence_quote":"Supplies the local-connectivity proof for filled Julia sets of subhyperbolic polynomials, cited in the paper to show the conjugating maps extend continuously to the closed disk."},{"cited_title":"Pilgrim, Expansion properties for finite subdivision rules I , Sci","cited_arxiv_id":null,"evidence_quote":"Provides the fat-path distance nonincreasing lemma for subdivision maps, the tool used in the proofs of Theorems 3.1 and 4.1."},{"cited_title":"Pilgrim, Combinations of complex dynamical systems, Springer Lec- ture Notes in Mathematics 1827, 2003","cited_arxiv_id":null,"evidence_quote":"Gives the canonical decomposition and pullback relation framework that motivates and supports the wandering-orbit non-realizability condition in Theorem 4.1."}],"review_version":1}