{"id":"6db2cd35-b8d1-40f2-b397-865c5ae511da","arxiv_id":"1908.01332","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A Jónsson cardinal κ below ℵ_κ yields 0¶ (the sharp for a strong cardinal), and mutual stationarity with mixed recurring cofinalities yields many high-Mitchell-order cardinals in the core model.","lead":"This set theory paper lowers the consistency-strength bar for mutually stationary sequences with divergent cofinalities, allowing countable cofinalities, and proves that a Jónsson cardinal below ℵ_κ implies the sharp for a strong cardinal, 0¶. It matters because it calibrates the large-cardinal strength of a natural stationarity property with mixed recurring shapes.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proof invokes weak covering reflected to K_X for countable-cofinal successors in §3; below 0¶ weak covering is only for uncountable cofinality, so cof((γ)^*)=μ for μ=ω is unsupported and Theorems 1.2–1.3 are not yet established.","rationale":"The reader's weakest assumption is indeed weak covering reflected to K_X, but I sharpen it: the real problem is its use at countable cofinality. The standard theorem is not available there. The paper cites [ACW, Obs. 25], which was formulated for uncountable cofinalities; the current proof extends to ω without a new argument. Lemmas 3.5–3.6 depend on exact cofinality γ_{n_i+1}=μ_{n_i}; if μ=ω and γ^* is an uncountable-cofinal successor, the contradiction no longer follows. This does not disprove the theorems, but it leaves the advertised countable-cofinality improvement unproved as written. For Theorem 1.4, the same reflected-covering invocation is less obviously problematic because finitely many small n can be discarded and uncountability is eventually available; still, the section would benefit from a spelled-out verification of the weak-covering hypotheses. Hence CONDITIONAL: the paper should be accepted only after the countable-cofinality weak-covering step is justified or replaced, or the affected theorems are restricted to uncountable μ.","tokens_in":11010,"tokens_out":26866,"duration_ms":270742,"concrete_test":"Check the cited [ACW, Obs. 25] and re-derive the equality in §3 for the case μ_{n_i}=ω. A decisive computation: in L (no 0¶), let X be a hull with β=σ^{-1}(ℵ_n)=ℵ_ω and μ=ω; then γ^*=β^{+L}=ℵ_{ω+1}, whose cofinality is ℵ_{ω+1}, not ω, so the displayed equality fails unless some additional alternating-block hypothesis rules out this configuration. If no extra hypothesis is identified, the inference in Lemma 3.5 for countable blocks is unsupported. Alternatively, replace the weak-covering step with a direct proof from Mitchell order alone and retrace Lemma 3.6.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The weakest point is the use of \"weak covering of K reflected down to K_X\" at singular cardinals of countable cofinality. The standard weak-covering theorem below 0¶ gives cf((λ^+)^K)=cf(λ) only when λ is singular of uncountable cofinality; for cf(λ)=ω it can fail dramatically (e.g., in L, cf(ℵ_ω^{+L})=ℵ_{ω+1}≠ω). Section 3 defines γ_{n_i+j}^* := ((β_{n_i})^{+j})^{K_X} for j<l and asserts, for every block including those with μ_{n_i}=ω, that cof(γ^*)=μ_{n_i}, citing [ACW, Obs. 25]. This equality is used in Lemmas 3.5 and 3.6 to force a contradiction when μ_{n_i}≠ρ; without it, the convergence argument for the γ-cofinalities collapses whenever infinitely many blocks have countable μ and ρ is uncountable. Since Theorems 1.2 and 1.3 are expressly advertised as handling countable cofinalities, this is a load-bearing gap. The same reflected-covering invocation in Lemma 4.2(a) also does not spell out why the relevant α is singular of uncountable cofinality in K_X, though the Jónsson case may be repairable by omitting finitely many n. The central proof of Theorem 1.4 is not directly affected, but the paper's headline improvement is.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the consistency strength of mutually stationary sequences with divergent cofinalities. It claims three main results: (1) Theorem 1.2, a lower bound in the core model K below 0¶ from the existence of a stationary set with alternating blocks of cofinalities of fixed length l, where countable cofinalities may appear infinitely often; (2) Theorem 1.3, a variant asserting that if all sequences alternating between two prescribed cofinalities are mutually stationary, then 0¶ exists; and (3) Theorem 1.4, that a Jónsson cardinal κ with κ < ℵ_κ implies 0¶ exists. The proofs analyze co-iterations of K with transitive collapses K_X of Skolem hulls, using fine structure, weak covering, and a pseudo-drop/quasi-iteration technique attributed to Mitchell. The paper explicitly states that it improves earlier work of Adolf, Cox, and Welch by reducing reliance on covering properties and by allowing countable cofinalities.","tokens_in":11325,"tokens_out":6360,"duration_ms":63267,"significance":"If the results are correct, they constitute a genuine advance: the previous treatment in [ACW] required all cofinalities to be uncountable, and the new theorems remove that restriction. The Jónsson-cardinal result Theorem 1.4 is also a strong lower consistency bound, improving on earlier work. The paper contains serious technical work: it develops a pseudo-drop lemma (Lemma 4.4) and adapts comparison arguments to cases where the K_X side of the iteration is nontrivial. However, the paper relies on a to-appear citation [ACW] for a key lemma (Lemma 2.1) and for the observation about reflected weak covering, and several central proof steps are compressed. The most important issue is that the treatment of countable cofinalities in Section 3 appears to invoke weak covering in a regime where the standard theorem does not apply; this directly affects the paper's main advertised improvement.","major_comments":[{"comment":"The claim that cof((γ^X_{n_i+j})*) = μ_{n_i+j} for all j < l, including the case μ_{n_i} = ω, is asserted to follow from 'weak covering reflected down to K_X' and cited to [ACW, Obs. 25]. Standard weak covering below 0¶ gives cf((λ^+)^K) = cf(λ) only when λ is singular of uncountable cofinality; for cf(λ) = ω it can fail, as in L where cf(ℵ_ω^{+L}) = ℵ_{ω+1} ≠ ω. Since the hypotheses of Theorem 1.2 and 1.3 explicitly allow μ_n = ω infinitely often, this equality is not established. The equality is used in Lemmas 3.5 and 3.6 to force a contradiction when μ_{n_i} ≠ ρ, so without it the convergence argument collapses for blocks with countable μ. Thus Theorems 1.2 and 1.3 are not yet supported in the advertised countable-cofinality case.","section":"§3, paragraph after Lemma 3.3"},{"comment":"The step 'It follows from weak covering that cof((α^+)^{K_X}) = (μ^X_0)^{+(n-1)}' needs verification that the relevant cardinal in K_X is singular of uncountable cofinality. The text does not spell this out. Since (μ^X_0)^{+(n-1)} is uncountable if μ^X_0 ≥ ℵ_2, this is likely repairable, but as written the reflected covering step is not justified and the lemma's proof is incomplete.","section":"Lemma 4.2(a)"},{"comment":"The proof of Lemma 4.4 is central to the quasi-iteration argument for Theorem 1.4, but it is too compressed. The 'easy induction' showing M^X_α = Hull^{M^X_α}_ω(κ^X_α ∪ {π_{2,α}(f̄_n) : n < ω}) is not shown, and the reflection claim '∀δ∀ξ_0...∀ξ_{m-1}∃β' is stated without a full derivation. Since this lemma substitutes for Lemma 2.1 in a case where the usual soundness argument fails, the proof should be expanded or the lemma should be stated with a pointer to a complete proof in the literature.","section":"Lemma 4.4"}],"minor_comments":[{"comment":"The text says 'In Section 5 we will have to consider special iterations', but the paper has no Section 5; the reference should be to Section 4.","section":"Preliminaries, second paragraph"},{"comment":"In the proof, 'we have some X_A ∈ S with X ≺ A' uses the variable X without introducing it; this should be 'with X_A ≺ A' to avoid confusion.","section":"Proposition 3.2 proof"},{"comment":"The expression Ult(K;σ_X↾(K||α^X_0)) uses the notation K||α^X_0 without definition; a brief reminder of the standard Zeman notation would improve readability.","section":"Lemma 3.3"},{"comment":"There are numerous typographical errors and inconsistent cross-reference labels (for example, '1.2Introductionthm.1.2'). A careful proofreading pass is needed.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The paper relies heavily on the to-appear reference [ACW] for Lemma 2.1 and for the reflected weak-covering observation. Since the current manuscript's main new claim (countable cofinalities) depends on a form of weak covering that is not part of the standard theorem, the editor may wish to ask the authors to clarify the status of [ACW, Obs. 25] and to verify whether it genuinely covers the countable-cofinality case. If it does not, the main theorems of Section 3 will need to be weakened or proved by a different route."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: if correct, this paper significantly improves the known lower bounds for mutual stationarity with divergent cofinalities, and the Jónsson result (Theorem 1.4) is new. But I have a concrete concern about the use of weak covering at countable cofinalities that may affect Theorems 1.2 and 1.3, the very results the paper advertises as its main improvement over [ACW].\n\nWhat the paper does well: it reduces the reliance on covering assumptions, extends the earlier arguments to sequences where countable cofinalities occur infinitely often, and proves that a Jónsson cardinal with κ<ℵ_κ implies 0^¶ exists. The proof strategy is coherent: mutual stationarity is used to find hulls with good cofinality patterns, then the co-iteration of K and K_X is analyzed, with Mitchell's pseudo-drop idea applied in the Jónsson case. The writing is clear about where results come from, and the historical note at the end is a plus. The main external citation, [ACW], is the author's own prior work, which is standard in this research program, but it does make verification harder because Lemma 2.1 is cited rather than reproduced.\n\nThe soft spot: in Section 3, after defining (γ_{n_i+j})^* as the successor of β in K_X, the paper asserts cof((γ)^*) = μ_{n_i} even when μ_{n_i}=ω, citing [ACW, Obs. 25]. Standard weak covering below 0^¶ gives cf((λ^+)^K) = cf(λ) only for singular λ of uncountable cofinality; for countable cofinality it can fail (in L, cf(ℵ_ω^{+L}) = ℵ_{ω+1} ≠ ω). The paper gives no proof of the countable-cofinality version, and the reference is a to-appear paper. If [ACW, Obs. 25] does not actually cover this case, the convergence arguments in Lemmas 3.5 and 3.6 collapse whenever countable-cofinality blocks appear. That is load-bearing. The Jónsson theorem (Section 4) seems less exposed, because the cofinalities there are uncountable, but even there Lemma 4.2(a) invokes weak covering without spelling out why the relevant ordinal has uncountable cofinality.\n\nMy own view: the reader's UNVERDICTED verdict is right. I could not find a fatal flaw, but the fine-structure details are beyond what I can verify quickly, and the weak-covering gap is concrete enough that the claims about countable cofinalities should not be taken at face value. I would want a referee who works in core model theory to check exactly what [ACW, Obs. 25] proves.\n\nRecommendation: send it to a good journal and get a serious referee. If the countable-cofinality step is fixed, or the cited lemma already covers it, this is a solid paper worth publishing. If not, Theorems 1.2 and 1.3 need significant revision, and Theorem 1.4 might survive alone. Either way, it deserves referee time, not a desk reject.","headline":"A serious inner-model-theory paper with a potentially load-bearing gap in its countable-cofinality claims; worth refereeing, but the referee must check the weak-covering step against the cited [ACW].","tokens_in":11867,"tokens_out":8264,"would_cite":false,"duration_ms":80256,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["03E45","03E55","03E05"],"pacs":[],"model":"deepseek-v4-flash","headline":"Mutually stationary sequences with alternating cofinality blocks force the core-model sharp 0¶ to exist.","keywords":["mutual stationarity","divergent cofinalities","core model","0¶","Jónsson cardinal","weak covering","consistency strength","inner model theory"],"falsifier":"A direct refutation would be a model of ZFC in which 0¶ does not exist but there is a stationary S ⊂ P(ℵ_ω) with alternating blocks of size 3 whose cofinality pattern uses ω infinitely often, and in which no κ_n < ℵ_n satisfies (κ_n^+)^K < ℵ_n with the stated Mitchell-order lower bound. Equally decisive would be a model with no 0¶ and a Jónsson cardinal κ < ℵ_κ.","tokens_in":10785,"feed_emoji":"♾️","tokens_out":9000,"duration_ms":86655,"temperature":0.7,"pith_summary":"This paper proves new lower consistency bounds for mutual stationarity, a combinatorial property of stationary sets attached to the cardinals below ℵ_ω. It shows that if a mutually stationary sequence concentrates on points whose cofinalities follow an alternating-block pattern, allowing the same cofinality to be ω infinitely often, then the core model K below 0¶ must already exhibit large covering failures. From that failure it derives the existence of 0¶, the sharp for an inner model with a strong cardinal, both for such sequences and for any Jónsson cardinal κ with κ < ℵ_κ. The advance over earlier work is a reduced reliance on covering hypotheses, which is what lets countable cofinalities enter the pattern infinitely often.","feed_headline":"Alternating cofinality blocks imply 0¶ exists","feed_subtitle":"Mutual stationarity with countable cofinalities infinitely often already has the strength of a strong cardinal.","key_machinery":"The core of the proof is the co-iteration of the core model K below 0¶ with K_X, the transitive collapse of X ∩ K for an elementary Skolem hull X of H_{ℵ_ω} or H_κ. The engine is a fine-structural lemma (Lemma 2.1) stating that a regular cardinal sitting between two projecta of a J-structure has the same cofinality as the lower projectum; repeated application converts the prescribed oscillation of the cofinalities μ_n into forced truncations and drops in the iteration. For the Jónsson theorem, a pseudo-drop construction provides a small iterable model whose fixed cofinalities on a club force infinitely many truncations, producing the contradiction.","core_discovery":"The paper's central claim is a dichotomy below 0¶. If S ⊂ P(ℵ_ω) is stationary and has alternating blocks of size l — meaning its elements have fixed cofinalities μ_n and blocks of length l repeat with more than one cofinality appearing infinitely often — then for infinitely many n there is κ_n < ℵ_n with (κ_n^+)^K < ℵ_n and o^K(κ_n) ≥ max(ℵ_n, ($κ_n^{{+(l+1)}}$)^K). Theorem 1.3 strengthens this: if for some k < l every sequence choosing cofinalities from {k,l} on a tail is mutually stationary, then 0¶ exists. Theorem 1.4 uses the same co-iteration analysis to show that a Jónsson cardinal κ with κ < ℵ_κ already yields 0¶.","pith_inferences":["The value of the cofinalities appears immaterial; only the recurrence pattern of the blocks matters. This suggests mutual-stationarity strength is controlled by oscillation patterns, and one can test this by forcing sequences whose cofinalities come from very different intervals while keeping the same block pattern.","A natural endpoint is that the exact consistency strength of Theorem 1.2 is the existence of a cardinal κ with o(κ) ≥ κ^{+(l+1)}; constructing a model with such a cardinal in which the mutually stationary sequence exists would confirm the bound as optimal.","Theorem 1.4 raises the possibility that Jónsson cardinals below ℵ_κ are consistency-equivalent to a strong cardinal or stronger; the paper explicitly asks whether a Woodin cardinal follows, so this is an open direction rather than a paper claim."],"forward_implications":["Below 0¶, a stationary set with alternating blocks of size l yields, for infinitely many n, a cardinal κ_n < ℵ_n whose K-successor is small and whose Mitchell order is at least ℵ_n and (κ_n^{+(l+1)})^K.","If for some k < l every choice of cofinalities from {k,l} on a tail is mutually stationary, then 0¶ exists, so that hypothesis has at least the consistency strength of a strong cardinal.","A Jónsson cardinal κ with κ < ℵ_κ implies 0¶, giving a new lower bound for such Jónsson cardinals.","The argument no longer requires all cofinalities to be uncountable, so the countable-cofinality cases carry the same covering-strength consequences as the uncountable ones."],"supporting_citations":[{"why":"Sets up the co-iteration and fine-structural framework for mutual stationarity with divergent uncountable cofinalities that this paper extends to countable cofinalities.","marker":"[ACW]"},{"why":"Supplies the covering and iterability lemmas that force the K-side to truncate in co-iteration with hulls.","marker":"[Cox09]"},{"why":"Provides the pseudo-drop construction used to control cofinalities on a club in the Jónsson argument.","marker":"[Mit99]"},{"why":"Introduces mutual stationarity and supplies the baseline result that countable-cofinality stationary sets are mutually stationary.","marker":"[FM01]"},{"why":"Shows the existence of the relevant elementary-substructure cofinality patterns is consistent relative to large cardinals.","marker":"[LS97]"},{"why":"Supplies the generator and pcf analysis used in the consistency side of the alternating-block configurations.","marker":"[Sha00]"},{"why":"Gives the consistency result for the countable-cofinality case used as a boundary for Theorem 1.3.","marker":"[CFM06]"},{"why":"Provides the fine-structural framework for the core model K and its hulls.","marker":"[Zem01]"}],"fun_headline_variants":["Alternating cofinality blocks force 0¶ without covering","Countable cofinality infinitely often gives 0¶","Jónsson cardinal below ℵ_κ implies 0¶","Less covering, same mutual stationarity strength","Divergent cofinalities: infinite blocks yield 0¶"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the core model below 0¶ satisfies weak covering at singular cardinals of uncountable cofinality, and that this covering property remains true when K is collapsed to a Skolem hull K_X. If reflected weak covering failed in one of these hulls, the cofinality computations in Lemmas 3.3 and 4.2(a) — and with them the contradiction — would break.","fun_headline_variants_meta":{"raw":{"variants":["Alternating cofinality blocks force 0¶ without covering","Countable cofinality infinitely often gives 0¶","Jónsson cardinal below ℵ_κ implies 0¶","Less covering, same mutual stationarity strength","Divergent cofinalities: infinite blocks yield 0¶"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000297,"raw_usage":{"total_tokens":1652,"prompt_tokens":804,"completion_tokens":848,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":420,"completion_tokens_details":{"reasoning_tokens":766}},"tokens_in":420,"tokens_out":848,"duration_ms":8313,"temperature":1.0,"reasoning_tokens":766,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:16:32.868003+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct refutation would be a model of ZFC in which 0¶ does not exist but there is a stationary S ⊂ P(ℵ_ω) with alternating blocks of size 3 whose cofinality pattern uses ω infinitely often, and in which no κ_n < ℵ_n satisfies (κ_n^+)^K < ℵ_n with the stated Mitchell-order lower bound. Equally decisive would be a model with no 0¶ and a Jónsson cardinal κ < ℵ_κ.","supporting_citations":[],"review_version":1}