{"id":"a39a0dd4-7ff4-4f9b-99d3-ff6d2360673b","arxiv_id":"2412.05304","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The whisker topology on fundamental groups preserves products, yields a Hausdorff non-discrete non-abelian topological group for an infinite product of wedges of two circles, and is non-separable on the Hawaiian earring group.","lead":"A topology called the whisker topology is placed on the fundamental group of a space. The paper shows this topology behaves well under products, answers an open question by giving a space whose whisker fundamental group is a non-discrete, non-abelian Hausdorff topological group, and proves the whisker topology on the Hawaiian earring group is not separable.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The headline example rests on Proposition 3.8, whose proof has an invalid converse step for infinite products; the product homeomorphism must be independently verified before the open question is answered.","rationale":"The paper's headline result is Proposition 3.19, and the only route from the hypotheses to the conclusion is Proposition 3.8 together with discreteness of the whisker topology on each factor. The discreteness step is standard and the factor S^1 vee S^1 is semilocally simply connected at the wedge point, so I see no serious issue there. The fragile point is Proposition 3.8 for infinite products. I checked the converse half of the proof and it is genuinely wrong as written: after choosing decompositions only for i in F, it sets the remaining coordinate loops constant, so the constructed element has trivial outside-F coordinates rather than the arbitrary prescribed coordinates. A correct proof is not hard: choose loops kappa_i for every i with [eta_i]=[zeta_i * kappa_i]; the coordinatewise loop is continuous by the universal property of products and has image in U because the restrictions are imposed only on the finitely many coordinates in F. But the paper does not give this argument, and Proposition 3.19 explicitly uses a countably infinite product, so the gap directly affects the central claim. The reader's CONDITIONAL verdict seems exactly right: the mathematics is very likely correct, but the posted proof needs repair before the open question can be regarded as settled. Other issues I noticed, such as the reversed hypothesis in Theorem 4.5 and the wrong internal references to Propositions 3.7 and 3.11 instead of 3.8 and 3.12, are typographical in nature but should also be corrected. I would not move the verdict.","tokens_in":10840,"tokens_out":14763,"duration_ms":143310,"concrete_test":"Verify Proposition 3.8 for the infinite product used in Proposition 3.19. Take X=prod_{i in N}(S^1 vee S^1), zeta the constant loop at the wedge point, F={1}, and U=U_1 x prod_{i>=2}(S^1 vee S^1) with U_1 a small open arc at the wedge point. Check explicitly whether B(1,U) equals B(1,U_1) x prod_{i>=2} pi^wh_1(S^1 vee S^1): in particular, take [eta] whose first coordinate is trivial and whose second coordinate is one of the two generators of F2, and show there is a loop kappa:I->U with [eta]=[kappa]; conversely show any loop in U has first-coordinate class in B(1,U_1). If this equality fails, Proposition 3.8 is false in the regime needed; if it holds, redo the general proof by choosing kappa_i for all i with [eta_i]=[zeta_i * kappa_i] and verifying the coordinatewise map is continuous.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Proposition 3.19 (the affirmative answer to Brazas's Question 3.18) identifies pi^wh_1(prod_{i in N}(S^1 vee S^1)) with (F2)^omega as a topological group. That identification depends entirely on Proposition 3.8, and the proof of Proposition 3.8 is not correct as written. In the converse half, after writing [eta_i]=[zeta_i * kappa_i] for i in F, the text says 'let kappa_i be constant at x_i when i in F' and then concludes ([eta_i]) lies in Phi(B([zeta], U)). This cannot be right: for coordinates outside F the chosen kappa_i are constant, so the constructed loop has components [zeta_i] outside F rather than the arbitrary prescribed [eta_i]. A correct argument would choose, for every coordinate i, a loop kappa_i with [eta_i]=[zeta_i * kappa_i], no restriction on the image of kappa_i for i outside F, and then use the coordinatewise loop kappa=(kappa_i); one must also check continuity and the equality of basic neighborhoods. The paper does not supply that check, and the infinite product case is exactly where the homeomorphism matters. If the homeomorphism fails for infinite products, or if some basic whisker neighborhood in the product does not split as a product of basic neighborhoods in finitely many factors, then W is not (F2)^omega topologically and the example collapses.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the whisker topology on fundamental groups of path-connected, locally path-connected based spaces. Its main results are: (1) Proposition 3.8, asserting that the whisker topology on the fundamental group preserves products, i.e. the canonical map π1^wh(∏ Xi) → ∏ π1^wh(Xi) is a homeomorphism; (2) Proposition 3.19, an affirmative answer to Brazas's Question 3.18, exhibiting X = ∏_{i∈N}(S^1∨S^1) for which π1^wh(X,x0) is a non-discrete, non-abelian, Hausdorff topological group; (3) Theorem 5.9, showing that π1^wh(E1,b0) is not separable; and (4) a pseudometric on π1^wh(X,x0) for metrizable X (Section 5.1) and a dichotomy for one-dimensional Peano continua (Theorem 5.10). The paper also contains results on when the whisker topology makes the fundamental group a topological group (Theorem 3.9) and on separation axioms (Theorem 4.5).","tokens_in":90,"tokens_out":13113,"duration_ms":276577,"significance":"If the results are correct, the paper makes a meaningful contribution to wild topology: it resolves an open question of Brazas on the existence of a space whose whisker fundamental group is non-discrete, non-abelian, Hausdorff, and a topological group, and it establishes non-separability of the whisker topology on the earring group. The broad strategy is sound and the paper connects several known tools (Cannon–Conner reduction, shape injectivity, product arguments). However, as written, the proofs contain gaps in load-bearing places: the product homeomorphism proof is incomplete, Theorem 4.5 is garbled, and Theorem 3.9 has an underived step. These are local and repairable, but they must be fixed before the claims can be accepted.","major_comments":[{"comment":"The converse half of the proof that Φ is open is incorrect as written. After writing [η_i] = [ζ_i·κ_i] for i∈F, the text says \"let κ_i be constant at x_i when i∈F\"; this should almost certainly read \"when i∉F\", but even with that correction the constructed loop has components [ζ_i] outside F, not the prescribed [η_i]. A correct argument must choose, for every coordinate i, a loop κ_i with [η_i] = [ζ_i·κ_i] (for i∉F one may take any representative of [ζ_i^{-1}·η_i]), then form the coordinatewise loop κ = (κ_i) by the universal property of the product. One must then verify that κ is continuous and that Im(κ) ⊆ U = ∏_{i∈F} U_i × ∏_{i∉F} X_i, which is true because each κ_i is a loop at x_i and for i∈F its image lies in U_i. The paper does not supply this check, and the infinite product case is exactly where the homeomorphism assertion is used in Proposition 3.19.","section":"§3, Proposition 3.8"},{"comment":"The theorem's statement and proof are garbled. Item 4 is a biconditional statement (\"π1^wh is Hausdorff if and only if X is homotopically Hausdorff at x0\"), not a condition in the same form as items 1–3. In the proof of 2⇒4, the text assumes ⟨1⟩ is not closed and concludes that X is not homotopically Hausdorff; this is the contrapositive of one direction of the biconditional, not a proof from the Hausdorff assumption. The subsequent \"conversely\" paragraph says \"suppose X is homotopically Hausdorff at x0\" but then asserts the existence of a non-null loop ζ that is path homotopic to loops in every neighborhood of x0, which is the negation of homotopically Hausdorff. The theorem should be restated (e.g., all four items are equivalent, with item 4 understood as the biconditional) and the two directions proved separately and correctly. Since Theorem 5.6 and Theorem 5.10 rely on this result, the repair is load-bearing.","section":"§4, Theorem 4.5"},{"comment":"In the proof of 1⇒2, the step \"Im(κ^{-})⊆V and so [κ][η]=[η][ι] for Im(ι)⊆U\" is underived. It can be obtained from the assumed continuity of inversion at [η^{-}]: for κ a loop in V, one has [η^{-}·κ]∈B([η^{-}],V), so its inverse [κ^{-}·η] lies in B([η],U), meaning [κ^{-}·η]=[η·ι] for some loop ι in U; multiplying on the left by [κ] and rearranging gives [κ][η]=[η][ι^{-}], with ι^{-} also having image in U. This derivation should be written out. The proof also contains typographical errors, including a missing parenthesis in \"inv(B([η^{-}, V))\" and the appearance of \"k^{-}\" instead of \"κ^{-}\".","section":"§3, Theorem 3.9"},{"comment":"The theorem asserts in both cases that eXwh is metrizable, but the paper never proves that eXwh is Hausdorff. Pseudometrizability is established in Lemma 5.4 and Theorem 5.5 for π1^wh, not for eXwh, and the metric ρ is defined only on π1^wh(X,x0). In the proof, the sentence \"since all one-dimensional Hausdorff spaces are homotopically Hausdorff, eXwh and π1wh are both Hausdorff and thus both metrizable\" presupposes a criterion for eXwh to be Hausdorff that is not stated or proved in the manuscript. A supporting argument or an explicit reference for the Hausdorff property of eXwh is needed before the metrizability and separability conclusions for eXwh can be accepted.","section":"§5, Theorem 5.10"}],"minor_comments":[{"comment":"The manuscript contains many typos and incorrect cross-references. For example, \"simly\" and \"nullhomotpic\" in Proposition 3.12, \"πwh_1(E1,b0) is a topological group whenever x, b0\" in Remark 3.15, and references to Proposition 3.7, Proposition 3.11, Proposition 3.18, and Theorem 4.6 where Propositions 3.8, 3.12, 3.19, and Theorem 4.5 (or 4.3) are meant.","section":"Throughout"},{"comment":"The proof says \"W is not semilocally simply connected at any of its points\"; this should refer to the space X = ∏(S1∨S1), not the group W. Also, the proof cites \"Proposition 3.7\" and \"Proposition 3.11\" but should cite Proposition 3.8 and Proposition 3.12.","section":"§3, Proposition 3.19"},{"comment":"The proof of Lemma 5.2 contains garbled inequalities. In the case x,y∈[0,1/2], the expression \"diam(ζ(x/2),ζ(y/2))\" is not meaningful; it should be d(ζ(2x),ζ(2y))≤diam(ζ). In the second case, \"d(η(2x−1),ζ(2y−1))\" should read \"d(η(2x−1),η(2y−1))\".","section":"§5, Lemma 5.2"},{"comment":"The proof invokes \"shape injectivity\" without defining the term or giving a precise reference. A brief definition or citation (e.g., to Cannon–Conner) would help the reader verify the claim that [ℓ_n·ℓ_1]∉B([ℓ_1],U_n).","section":"§3, Proposition 3.14"},{"comment":"The one-line proof should justify why a retract of a space with non-topological-group whisker fundamental group must itself fail to be a topological group; the argument requires that the retraction-induced homomorphism gives a homeomorphic embedding of π1^wh(E1) into π1^wh(X).","section":"§3, Corollary 3.16"},{"comment":"In the semilocally simply connected case, the assertion that π1^wh(X,x0) is finitely generated is not justified; a reference for finite generation of the fundamental group of a semilocally simply connected Peano continuum should be supplied.","section":"§5, Theorem 5.10"}],"recommendation":"major_revision","confidential_remarks":"The manuscript has a promising core and the main examples seem genuinely valuable, but the current version has too many proof gaps and cross-referencing errors to be published as is. The most serious risk is Proposition 3.8 for infinite products, since the affirmative answer to Brazas's question depends on it; the proof can likely be repaired by the coordinatewise construction described in the report. Theorem 4.5 also needs a full rewrite. I recommend major revision rather than reject because the central claims are defensible and the issues are local in nature."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does three things that are actually new: it proves the whisker topology preserves arbitrary products (Prop 3.8), uses that to answer Brazas's open question with an infinite product of S1∨S1 (Prop 3.19), and shows the whisker topology on the earring group is non-separable (Thm 5.9) with a one-dimensional dichotomy (Thm 5.10). The pseudometric in Section 5 is also a useful addition. The exposition of basic properties—functoriality, left-topological group structure, equivalence with being a topological group—is careful and mostly correct. The paper is built on standard techniques, and I see no sign of circularity or parameter fitting.\n\nThe stress-test note is right about Prop 3.8. The converse half says \"let κ_i be constant at x_i when i∈F,\" which is backwards: for coordinates in the finite set F, the κ_i must realize the prescribed homotopy classes; for coordinates outside F, constant loops are fine. As written, that step does not establish the claimed homeomorphism, and this is the load-bearing result for the headline example. The intended repair is obvious—choose κ_i for every coordinate with κ_i constant for i∉F, then verify continuity of the coordinatewise loop and that the basic neighborhood splits. But the continuity check is not in the paper, and in the infinite product case that is exactly where the topology matters. I consider this a typo with a clear fix, not a deep flaw, but it is still a real gap in the posted version.\n\nOther soft spots are smaller but worth listing. Theorem 4.5 is misstated: item 4 is an iff statement, not one of the equivalent conditions, and the proof conflates \"X is homotopically Hausdorff\" with \"π1^wh is Hausdorff.\" Proposition 3.9's 1⇒2 step uses an underived identity [κ][η]=[η][ι]. There are also several wrong cross-references, e.g., Prop 3.19 cites Prop 3.7 instead of 3.8 and Prop 3.11 instead of 3.12. These are the kind of errors a referee would catch quickly.\n\nThe non-separability theorem depends on Cannon–Conner's reduced-path uniqueness, which is standard and appropriately cited. Overall, the central arguments hold up despite the presentation problems. This paper is for people who work on topological fundamental groups and wild topology, and it deserves a serious referee. My recommendation: send it to peer review; ask for a corrected proof of Prop 3.8, a cleaned-up Theorem 4.5, and a few cross-reference fixes. After that, it should be publishable.","headline":"Three genuinely new results and a plausible affirmative answer to Brazas's question, but the posted proof has a load-bearing typo in Proposition 3.8 that needs repair before the example is fully credible.","tokens_in":11596,"tokens_out":3033,"would_cite":true,"duration_ms":47944,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["54H11","55Q05","54E35"],"pacs":[],"model":"deepseek-v4-flash","headline":"The whisker topology on fundamental groups respects products, giving a non-discrete, non-abelian, Hausdorff topological group from a countable product of figure eights.","keywords":["whisker topology","fundamental group","topological group","Hawaiian earring","one-dimensional Peano continuum","product preservation","homotopically Hausdorff","separability"],"falsifier":"Take an infinite product of spaces that are not semilocally simply connected and compare a basic whisker neighborhood of a loop in the product with the product of basic whisker neighborhoods in the factors. If some basic neighborhood cannot be expressed as a product of factor neighborhoods, or if the canonical map $\\Phi$ is not open on such a neighborhood, then the product-preservation theorem fails and the main example loses its proof.","tokens_in":10636,"feed_emoji":"⭕","tokens_out":9134,"duration_ms":79571,"temperature":0.7,"pith_summary":"This paper studies the whisker topology, a topology on the fundamental group designed to detect local complications in spaces where classical covering-space theory fails. Its main aims are to prove that the whisker topology is preserved by products, to use that fact to settle an open question by building a space whose whisker fundamental group is simultaneously non-discrete, non-abelian, and Hausdorff, and to show that the whisker fundamental group of the Hawaiian earring is not separable. If these results stand, they give a way to compute whisker fundamental groups of infinite products and sharpen the contrast between the Hawaiian earring and tamer examples such as the infinite torus.","feed_headline":"Infinite product of figure-eights yields a wild topological group","feed_subtitle":"Countably many wedge circles make the whisker fundamental group Hausdorff, non-abelian, and non-discrete, settling an open question.","key_machinery":"The whisker topology on $\\pi_1(X,x_0)$ is generated by basic sets $B([\\zeta],U)$ consisting of classes $[\\zeta\\cdot\\eta]$ for loops $\\eta$ based at $x_0$ with image in an open neighborhood $U$ of $x_0$. Two structural facts carry the argument: these basic sets are clopen, so a Hausdorff whisker group is totally separated, and the product-preservation theorem identifies the whisker fundamental group of a product with the product of whisker fundamental groups via the map sending $[\\zeta]$ to $([p_i\\circ\\zeta])_i$. This converts an infinite product of figure eights into an infinite product of discrete free groups, and because products of topological groups are topological groups, the resulting group inherits the desired combination of properties.","core_discovery":"The central claim is Proposition 3.19. Let $X=\\prod_{i\\in\\mathbb{N}}(S^1\\vee S^1)$, based at the point whose every coordinate is the wedge point of a figure eight. Then the whisker fundamental group $W=\\pi_1^{wh}(X,x_0)$ is a non-discrete, non-abelian, Hausdorff topological group. The proof rests on the product-preservation theorem: the canonical map $\\Phi$ from $\\pi_1^{wh}(\\prod_i X_i)$ to $\\prod_i\\pi_1^{wh}(X_i)$ is a homeomorphism. Because each factor $S^1\\vee S^1$ is semilocally simply connected, its whisker fundamental group is the discrete free group $F_2$, and $W$ becomes the countable power $(F_2)^{\\omega}$ with the product topology. The paper also claims that $\\pi_1^{wh}(E^1,b_0)$, the whisker fundamental group of the Hawaiian earring, is not separable, and derives a metrizability and separability dichotomy for one-dimensional Peano continua.","pith_inferences":["The same product argument should generalize to products of arbitrary semilocally simply connected spaces with non-abelian fundamental groups, producing a large class of non-discrete topological whisker groups rather than a single example.","The non-separability of $\\pi_1^{wh}(E^1,b_0)$ likely forces the generalized universal cover $\\widetilde{E}^1$ to be non-separable too, so the whisker topology is a finer invariant than mere metrizability for one-dimensional spaces.","One could test whether the product-preservation theorem extends to inverse limits or uncountable products; the written proof relies on finite-support neighborhoods, so uncountable products may behave differently.","The dichotomy in Theorem 5.10 suggests that non-separability of the whisker fundamental group is a robust marker of non-semilocally-simple-connected one-dimensional spaces, which could be used to classify generalized covering spaces by separability alone."],"forward_implications":["Whisker fundamental groups of arbitrary products can be computed factor by factor, making infinite products of locally nice spaces a reliable source of whisker groups.","The space $\\prod_{i\\in\\mathbb{N}}(S^1\\vee S^1)$ provides a canonical example where the whisker topology is Hausdorff, non-discrete, and non-abelian simultaneously, and it is a topological group, so the open question has an affirmative answer.","Because $\\pi_1^{wh}(E^1,b_0)$ is metrizable but not separable, any one-dimensional Peano continuum containing the Hawaiian earring as a retract has a non-separable whisker fundamental group.","For one-dimensional Peano continua, the whisker fundamental group is either a discrete separable metric group or a non-separable metric group, with no intermediate behavior.","The pseudometric constructed in Section 5 makes $\\pi_1^{wh}(X,x_0)$ pseudometrizable whenever $X$ is metrizable, and metrizable when $X$ is additionally homotopically Hausdorff."],"supporting_citations":[{"why":"Supplies Lemma 5.8, the reduced-path existence and uniqueness result used in the non-separability proof for the Hawaiian earring group.","marker":"[7]"},{"why":"Supplies the shape-injectivity and reduced-word rigidity of the Hawaiian earring group used in the earring-space arguments.","marker":"[8]"},{"why":"Supplies the standard fact that products of topological groups are topological groups, used in Proposition 3.19.","marker":"[2]"},{"why":"Provides the definition of homotopically Hausdorff spaces that Theorem 4.5 equates with Hausdorffness of the whisker fundamental group.","marker":"[9]"},{"why":"Provides the classical covering-space framework that identifies the whisker path space with a universal covering and supports the quotient-topology equivalence in Section 5.","marker":"[18]"},{"why":"Supplies the generalized universal covering space theory that places the whisker path space as a covering-like object outside the classical setting.","marker":"[13]"}],"fun_headline_variants":["Infinite product of eights gives a Hausdorff, non-discrete, non-abelian group","Countable product of eights yields non-discrete non-abelian Hausdorff group","Open question settled: product of eights yields non-discrete, non-abelian group","Hawaiian earring whisker group not separable"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The main example rests on the claim that the whisker topology on the fundamental group of a product of spaces is exactly the product of the whisker topologies of the factors; if that homeomorphism fails for infinite products, the proposed example collapses.","fun_headline_variants_meta":{"raw":{"variants":["Infinite product of eights gives a Hausdorff, non-discrete, non-abelian group","Countable product of eights yields non-discrete non-abelian Hausdorff group","Open question settled: product of eights yields non-discrete, non-abelian group","Hawaiian earring whisker group not separable"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.002733,"raw_usage":{"total_tokens":10383,"prompt_tokens":866,"completion_tokens":9517,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":482,"completion_tokens_details":{"reasoning_tokens":9427}},"tokens_in":482,"tokens_out":9517,"duration_ms":59765,"temperature":1.0,"reasoning_tokens":9427,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T12:13:02.965565+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take an infinite product of spaces that are not semilocally simply connected and compare a basic whisker neighborhood of a loop in the product with the product of basic whisker neighborhoods in the factors. If some basic neighborhood cannot be expressed as a product of factor neighborhoods, or if the canonical map $\\Phi$ is not open on such a neighborhood, then the product-preservation theorem fails and the main example loses its proof.","supporting_citations":[{"cited_title":"Cannon, G.R","cited_arxiv_id":null,"evidence_quote":"Supplies Lemma 5.8, the reduced-path existence and uniqueness result used in the non-separability proof for the Hawaiian earring group."},{"cited_title":"Cannon, G.R","cited_arxiv_id":null,"evidence_quote":"Supplies the shape-injectivity and reduced-word rigidity of the Hawaiian earring group used in the earring-space arguments."},{"cited_title":"Arhangegel’skii and M","cited_arxiv_id":null,"evidence_quote":"Supplies the standard fact that products of topological groups are topological groups, used in Proposition 3.19."},{"cited_title":"Conner, M","cited_arxiv_id":null,"evidence_quote":"Provides the definition of homotopically Hausdorff spaces that Theorem 4.5 equates with Hausdorffness of the whisker fundamental group."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the classical covering-space framework that identifies the whisker path space with a universal covering and supports the quotient-topology equivalence in Section 5."},{"cited_title":"Fischer, A","cited_arxiv_id":null,"evidence_quote":"Supplies the generalized universal covering space theory that places the whisker path space as a covering-like object outside the classical setting."}],"review_version":1}