{"id":"287674b8-bc99-47b0-afe5-645849d72c6b","arxiv_id":"2509.02567","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Physical determinism, read with ensemble and canonicalization clauses, can change between V=L and large-cardinal/projective-determinacy backgrounds.","lead":"This paper argues that whether a physical theory counts as deterministic can depend on which set-theoretic axioms we adopt, once 'for almost all' and 'canonical' claims are taken seriously. If true, debates about black holes, quantum bases, and inverse problems inherit hidden commitments to V=L or large cardinals.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central claim is conditional on defining determinism to include the Π^1_2 universal-tameness clause; the advertised unconditional theorems that would remove this dependence are absent from the body.","rationale":"The reader's weakest_assumption identifies the same load-bearing concern: the conclusion depends on including a Π¹₂ robustness ideal in the meaning of 'determinism.' I agree with that assessment. My stress-test adds that the two unconditional theorems promised in the abstract—the Ising Σ¹₂-completeness theorem and the Kerr E₀/canonicalization theorem—would have provided non-definitional evidence for the dependence, but they are not actually proved in the body. Section 6 gives only the observation that a Π¹₂ uniformization problem arises, not a proof that the particular fixed Ising system's profile is Σ¹₂-complete. Section 7.2 similarly describes a projective canonicalization problem but gives no E₀-embedding and no proof that V=L supplies a projectively definable rule while PD forbids one. This matters because complexity by itself does not establish a toggle: a Π¹₂ statement can be absolutely true in both V=L and LC, or its regularity can fail only for an arbitrary Δ¹₂ set that is not shown to arise in the physical construction. The paper's own Section 3 concession—that none of the later physical examples produces non-measurable coefficients for any fixed instance—reinforces that the body's examples are not the advertised unconditional toggles. Since the reader's CONDITIONAL verdict already captures this state of affairs, I do not move the verdict. The descriptive set-theoretic facts invoked are standard and I see no internal inconsistency in the mathematics; the weakness is at the level of what the examples establish and what definition of determinism is doing the work.","tokens_in":38440,"tokens_out":9500,"duration_ms":118879,"concrete_test":"Re-formulate each example in §§4–7 with the paper's own §1 notion of determinism (for every admissible datum: existence, uniqueness, continuous dependence), deleting the universal-tameness clause UT(x,y). Show, via Shoenfield absoluteness and the ZFC-absolute nature of the analytic estimates, that the resulting verdict is the same under V=L and LC for every fixed datum. If it is, the central dependence is an artifact of adding UT to the definition of determinism.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's set-theoretic background facts are standard, but the step from 'this selection problem has Π^1_2 complexity' to 'the determinism verdict differs' is not secured. Section 1 defines determinism in the standard Hadamard sense; that analytic layer is ZFC-absolute. The toggles appear only after adding the 'universal tameness' clause UT(x,y): ∀ρ∈𝒫 ∃q∈ℚ⁺ Good(x,y,ρ,q) in Section 7, and Section 7.4 explicitly states: 'if determinism is defined through such a robustness ideal, then even the determinism profile of a single physical system can differ.' This is a philosophical premise, not a theorem. Without UT, the examples in Sections 4–7 leave existence, uniqueness, and continuous dependence unchanged for every fixed admissible datum. The abstract's two unconditional theorems—the Σ¹₂-complete Ising tail-readout profile and the Kerr E₀ canonicalization dichotomy—would provide independent, non-definitional content, but neither is proved in the body. Section 6 contains no Σ¹₂-completeness construction; Section 7.2 contains no E₀ embedding or V=L/PD projectivity proof. Section 3 concedes that 'none of the later, more physical examples...produces non-measurable coefficients for any fixed instance.' Thus the strongest advertised claim is unsupported, and the remaining conclusion is conditional on accepting UT as part of 'determinism.'","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper argues that the choice between V=L and large-cardinal/PD set-theoretic frameworks can change whether physical theories are deterministic, at a 'regularity layer' beyond the standard analytic Hadamard notion. It introduces three hinges—coherence, uniqueness-as-genericity, and identity/canonical selection—and a universal tameness clause UT(x,y) of Pi-1-2 complexity claimed to formalize robustness across admissible encodings. It surveys inverse problems, thin-barrier Markov uniqueness, zero-temperature Ising dynamics, decoherence/preferred basis, and Kerr interiors. The abstract advertises two unconditional theorems (a Sigma-1-2-complete Ising tail-readout profile and a Kerr E_0 canonicalization dichotomy) and concludes with a Carnap/Quine dilemma about relativizing physics to set theory. The body, however, contains no statement or proof of the two advertised theorems, and Section 7.4 explicitly frames the main single-system conclusion as conditional on defining determinism through the UT robustness ideal.","tokens_in":38850,"tokens_out":5834,"duration_ms":66580,"significance":"The topic is timely and potentially important: if the advertised results were established, they would create a genuine bridge from descriptive set theory to philosophy of physics and give content to the proposed 'reverse physics' program. The paper deploys standard facts correctly (Shoenfield absoluteness, V=L pathologies, PD regularity) and is commendably transparent about its own limitations, notably in Section 3 and Section 7.4. But as it stands, the central claim does not rise above a conditional philosophical premise, and the abstract's two unconditional theorems are absent from the body. The paper is a promising research proposal rather than a completed proof of metatheoretic dependence of determinism.","major_comments":[{"comment":"The abstract states: 'Two unconditional theorems are proved. First, a fixed computable nearest neighbor Ising Hamiltonian ... yields a Sigma-1-2-complete tail-readout profile ... Second, at the Kerr Cauchy horizon, canonicalizing continuous extension germs ... contains E_0.' No such theorem statements or proofs appear in the body. Section 6 discusses tie-breaker invariance and uniformization but never defines a tail-readout profile or proves Sigma-1-2-completeness. Section 7.2 defines Gamma, Gamma*, UT and asserts complexity facts, but gives no E_0 embedding and no V=L/PD projectivity dichotomy. This is load-bearing because the advertised theorems are what would make the dependence non-definitional.","section":"Abstract; Sections 6 and 7.2"},{"comment":"The paper first defines determinism in the standard Hadamard sense (existence, uniqueness, continuous dependence), which is ZFC-absolute. The V=L/PD divergence is introduced only through the universal-tameness clause UT(x,y) : <=> forall rho in P exists q in Q+ Good(x,y,rho,q), described as formalizing a 'robustness ideal.' Section 7.4 concedes: 'if determinism is defined through such a robustness ideal, then even the determinism profile of a single physical system can differ.' The claim that this idealization is 'not an optional flourish' is asserted, not argued. A physicist who declines to include universal robustness in the definition of determinism loses nothing the paper shows to be part of standard determinism talk. The main conclusion is therefore a philosophical premise, not a theorem.","section":"Section 1 and Section 7.4"},{"comment":"The examples establish at most that certain Pi-1-2-defined families 'may fail' to be measurable or to admit measurable uniformization under V=L; they do not show that any specific, physics-relevant set is actually non-measurable under V=L. Section 3 explicitly says that none of the later examples produces non-measurable coefficients for any fixed instance. For the ensemble-level claims, the text says measurability 'can fail' or 'need not be guaranteed,' but it supplies no diagonal construction or proof that the particular sets arising in inverse problems, Markov uniqueness, Ising dynamics, decoherence, or Kerr are V=L-nonmeasurable. 'Can fail' or 'may fail' does not yield the abstract's stronger claim that determinism verdicts 'can differ.'","section":"Sections 4-7"},{"comment":"The Kerr discussion is presented as a live case, but its analytic claims are asserted rather than proved: Gamma(d) is non-empty and sequentially compact with analytic graph; the canonicalization problem has Pi-1-2 complexity; under PD measurable selectors exist while under V=L definable but non-measurable tie-breakers are available. No E_0 embedding into the canonicalization germs is constructed, and no proof is given that V=L supplies a projectively definable rule or that PD forbids one. As it stands, the section is a framework for a possible result, not a proof of the 'unconditional theorem' promised in the abstract.","section":"Section 7.2"},{"comment":"The summary claims that 'the regularity layer—the assumptions that make ensemble or canonical claims well-defined—does' depend on set-theoretic background. But the only demonstrated mechanism is the added UT clause; without it, the paper's own Section 3 concedes fixed instances remain analytic and ZFC-absolute. The step from 'physicists sometimes use genericity idioms' to 'these idioms are part of determinism itself' is the entire weight of the argument, and it is not defended against the obvious response that robust across all admissible policies is a methodological ideal added by the author, not a component of determinism.","section":"Section 7.5"}],"minor_comments":[{"comment":"Typo: 'caees' should be 'cases'.","section":"Section 7.4"},{"comment":"Typo: 'dsicsussed' should be 'discussed'.","section":"Objection 1"},{"comment":"Typo: 'parition' should be 'partition'.","section":"Objection 4"},{"comment":"Typo: 'aluded' should be 'alluded'.","section":"Objection 5"},{"comment":"The notation 'Σ¹ ₂' has inconsistent spacing; should be uniform throughout.","section":"Section 4, footnote 4"},{"comment":"The proposed experimental protocols are described as testing whether nature 'enforces measurable, coding-invariant behavior,' but all experimental measurements are finite and Borel; the stated contrast between the LC and V=L profiles cannot be operationalized as written. This appendix should be toned down or clarified.","section":"Appendix F"}],"recommendation":"major_revision","confidential_remarks":"There is a serious mismatch between what the abstract promises (two unconditional theorems) and what the body delivers (a conditional philosophical argument plus asserted examples). The paper should not be accepted in its current form. I would require either that the advertised theorems be stated and proved in the body or appendix, or that the abstract and conclusions be substantially revised to reflect the actual conditional content. The descriptive set theory is used correctly, and the author's explicit caveats are helpful, but the central claim needs to be either proved or honestly downgraded."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nRead Clarke-Doane's paper on set-theoretic metatheory and physical determinism. The core idea is worth your time. He distinguishes the analytic layer of a physical theory (well-posedness in Hadamard's sense, ZFC-absolute) from a 'regularity layer' where ensemble talk ('almost all', 'generic') and canonical selection presuppose measurability or Baire regularity. He argues this layer sits at the Pi^1_2 level, where V=L and large-cardinal-plus-PD diverge. That framing--plus the 'reverse physics' program, modeled on reverse mathematics--is genuinely new as far as I can tell, and it connects known descriptive set theory to physical practice in a way the literature hasn't. The paper is also honest in places: Section 3 explicitly concedes that none of the later, more physical examples produces non-measurable coefficients for any fixed instance.\n\nThe problem is the gap between the abstract and the body. The abstract promises two unconditional theorems: a Sigma^1_2-complete Ising tail-readout profile and an E_0-embedding at the Kerr Cauchy horizon. Neither is proved in the text. Section 6 gives no completeness construction; Section 7.2 gives no E_0 embedding. What the paper actually shows is conditional: if determinism is defined to include the 'universal tameness' clause UT (forall rho exists q Good(x,y,rho,q)), then the robustness profile lands at Pi^1_2 and toggles between V=L and PD. Section 7.4 states exactly that. So the strongest advertised results are missing, and the central conclusion rests on a philosophical premise about what determinism includes. A physicist who insists determinism is just the analytic layer will see no dependence.\n\nThe examples themselves mostly show possibility ('may fail') rather than actual divergent verdicts. They demonstrate that ensemble idioms can lose determinate sense under V=L, not that the same theory has a different unique future. That's still interesting, but it is a weaker claim than the headline.\n\nI'd like a revision that either proves the two theorems or drops them, and that explicitly separates 'ill-posed ensemble language' from 'different determinism verdicts.' As it stands, this is a stimulating conceptual contribution, not a mathematical demonstration. I'd send it to a good referee, but they should be asked to check the gap between claims and proofs. For my own work, I'd cite the analytic/regularity distinction as a framing device, but not the purported theorems.\n\nRecommendation: engage, but don't take the advertised theorems at face value.","headline":"A stimulating but over-sold paper: the analytic/regularity distinction is valuable, but the advertised theorems are absent and the main conclusion is conditional on a definitional premise.","tokens_in":39215,"tokens_out":2732,"would_cite":true,"duration_ms":30131,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["03E15","03E35","03E45"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper argues that whether a physical theory is deterministic can depend on which extension of ZFC one adopts, because ensemble and canonical claims require a regularity layer that V=L and large-cardinal frameworks settle differently.","keywords":["determinism","set-theoretic foundations","V=L","projective determinacy","regularity layer","Cauchy horizon","Ising model","reverse physics"],"falsifier":"Construct, in ZFC, a universally measurable selector for the Kerr Cauchy-horizon canonicalization multifunction; that would contradict the paper's claim that the relation contains $E_0$ and hence admits no such rule. Alternatively, exhibit a Borel (or even $\\Sigma^1_1$) witness to the Ising tail-readout, contradicting its claimed $\\Sigma^1_2$-completeness.","tokens_in":38368,"feed_emoji":"🕳️","tokens_out":11634,"duration_ms":113230,"temperature":0.7,"pith_summary":"This paper argues that the question of whether a physical theory is deterministic—whether it fixes a unique future from the present—is not answered by the equations alone. Once determinism is read as physicists actually use it, with “almost all” and “canonicl” claims required to be robust across every admissible way of sampling or refining the system, the verdict lands in a regularity layer whose properties differ between Gödel's constructible universe $V=L$ and large-cardinal frameworks implying projective determinacy. The paper proves two unconditional results in this vein: a zero-temperature Ising model on $\\mathbb{Z}^3$ has a $\\Sigma^1_2$-complete tail-readout profile, and canonicalizing extensions at the Kerr Cauchy horizon is so irregular that no universally measurable rule exists in ZFC, while $V=L$ and PD diverge on which projective rules exist. The upshot is a dilemma: physical theories must either be relativized to their set-theoretic background, or the search for new axioms is empirical, as Quine held.","feed_headline":"Physics determinism can flip with set-theoretic axioms","feed_subtitle":"The choice between V=L and large-cardinal axioms changes whether black-hole interiors have a unique future.","key_machinery":"The central mechanism is the $\\Pi^1_2$ “universal tameness” clause $UT(x,y)$: for every admissible sampling or refinement policy $\\rho$ there is a tolerance $q$ such that $\\text{Good}(x,y,\\rho,q)$. It encodes the physicist's robustness ideal and sits at the first projective level where $V=L$ and PD diverge on regularity, so the existence of measurable, representation-independent selections — the identity hinge — becomes metatheory-dependent.","core_discovery":"Determinism, the paper claims, has a “regularity layer” beyond analytic well-posedness. Its universal-tameness clause $UT(x,y)$ — $\\forall \\rho\\,\\exists q\\,\\text{Good}(x,y,\\rho,q)$ over admissible sampling policies — is $\\Pi^1_2$, the level where $V=L$ and large-cardinal projective determinacy split: $V=L$ permits non-measurable $\\Delta^1_2$ sets and merely definable uniformizations; PD gives every projective set measurable and uniformly selectable. The paper proves two unconditional theorems: a fixed computable nearest-neighbor Ising Hamiltonian on $\\mathbb{Z}^3$ with fixed zero-temperature Glauber schedule yields a $\\Sigma^1_2$-complete tail-readout profile; and Kerr Cauchy-horizon canonic","pith_inferences":["If the regularity layer is accepted, then other modal idioms — prediction, explanation, causation, physical necessity — presumably inherit the same metatheory-relative semantics, a consequence the paper only gestures at.","The $\\Sigma^1_2$-completeness result for the Ising tail-readout, if it withstood scrutiny, would give physics its own complete problems inside the projective hierarchy, opening the door to transfer of undecidability between statistical mechanics and descriptive set theory.","The $E_0$-embedding at the Kerr horizon suggests that cosmic censorship interacts with the theory of countable Borel equivalence relations; asking whether the canonicalization is hyperfinite or turbulent would refine which definability obstruction is at work.","The cumulative protocols in the paper's appendix are a concrete, if very hard, route to empirical input on the Quinean question: observation of coding-invariance breakdown would be evidence against the large-cardinal regularity profile."],"forward_implications":["If the paper is right, determinism is not an intrinsic property of the equations; it is a three-layer verdict (coherence, uniqueness-as-genericity, identity) that is only fully determined once a set-theoretic background is fixed.","Strong Cosmic Censorship, phrased as a claim about generic data and canonical continuations, lacks a determinate truth value in ZFC alone; it is relative to $V=L$ or PD.","The preferred-basis problem in decoherence and bulk reconstruction in AdS/CFT inherit projective uniformization problems, so their well-posedness toggles with the metatheory.","Even fully discrete, Borel dynamics (zero-temperature Ising) can exhibit $\\Sigma^1_2$-complete tail behavior, so the phenomenon is not an artifact of continuum PDEs.","A “reverse physics” program becomes possible: classify physical theorems by the axiom schemes needed to make their ensemble and canonical claims meaningful, mirroring reverse mathematics."],"supporting_citations":[{"why":"Supplies the constructible universe $L$ and the $\\Delta^1_2$ well-order used under $V=L$ to build pathological, merely definable selections.","marker":"Gödel 1940; Jech 2003, Ch. 13"},{"why":"Proves projective determinacy from large cardinals, the engine of regularity for all projective sets under LC.","marker":"Martin & Steel 1989"},{"why":"Gives the regularity and uniformization theorems that PD yields at the projective levels.","marker":"Kechris 1995, §28, §38–39"},{"why":"Provides the descriptive-set-theoretic background for projective determinacy, selection, and the Baire/measurable properties.","marker":"Moschovakis 2009, Ch. 6"},{"why":"Fixes the absoluteness boundary: $\\Sigma^1_2/\\Pi^1_2$ truths are absolute, so the $V=L$ vs PD divergence is one of regularity, not membership.","marker":"Shoenfield 1961; Jech 2003, Thm. 25.20"},{"why":"Establishes measurable selection for analytic graphs, the baseline that projective-level uniformization must extend.","marker":"Kuratowski & Ryll-Nardzewski 1965; Castaing & Valadier 1977"},{"why":"Gives the capacity criterion $\\text{cap}_{1,2}(S)=0$ for Markov uniqueness, the analytic core of the thin-barrier example.","marker":"Ma & Röckner 1992; Fukushima, Oshima & Takeda 2011, Ch. 2"},{"why":"Provides the Kerr-interior energy estimates and $C^0$-stability results that frame the Cauchy-horizon canonicalization problem.","marker":"Wald 1984; O’Neill 1995; Dafermos & Luk 2017"}],"fun_headline_variants":["Determinism in physics hinges on set theory axioms","Set-theory axioms can flip physics determinism verdicts","V=L vs large-cardinal axioms flip determinism verdicts","Physics determinism not absolute: depends on set theory","ZFC axioms decide if black-hole futures are unique"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The load-bearing premise is that physical determinism includes the universal-tameness robustness clause UT — the demand that verdicts survive every admissible sampling or refinement policy — rather than only analytic well-posedness; if one rejects that definition of determinism, the set-theoretic dependence does not arise.","fun_headline_variants_meta":{"raw":{"variants":["Determinism in physics hinges on set theory axioms","Set-theory axioms can flip physics determinism verdicts","V=L vs large-cardinal axioms flip determinism verdicts","Physics determinism not absolute: depends on set theory","ZFC axioms decide if black-hole futures are unique"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000809,"raw_usage":{"total_tokens":3455,"prompt_tokens":881,"completion_tokens":2574,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":625,"completion_tokens_details":{"reasoning_tokens":2495}},"tokens_in":625,"tokens_out":2574,"duration_ms":21169,"temperature":1.0,"reasoning_tokens":2495,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T19:56:22.204061+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct, in ZFC, a universally measurable selector for the Kerr Cauchy-horizon canonicalization multifunction; that would contradict the paper's claim that the relation contains $E_0$ and hence admits no such rule. Alternatively, exhibit a Borel (or even $\\Sigma^1_1$) witness to the Ising tail-readout, contradicting its claimed $\\Sigma^1_2$-completeness.","supporting_citations":[],"review_version":1}