REVIEW 3 major objections 4 minor 5 references
$T$-convexly valued o-minimal fields are definably spherically complete
T0 review · 3 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash
Pith's one-line read Every model of $T_{\mathrm{convex}}$ is definably spherically complete: definable nested families of valuation balls always have non-empty intersection.
desk verdict A short, promising note: the theorem is likely correct, but Lemma 1's proof needs to be written out before this is referee-ready. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The workhorse is the cofinality classification in Lemma 1: for $(E,O)\models T_{\mathrm{convex}}$, every $(E,O)$-definable subset $X\subseteq v(E,O)$ has cofinality in $\{\mathrm{cof}(r(E,O)), \mathrm{cof}(v(E,O)^{<0}), \mathrm{cof}(E), 1\}$. The proof of the theorem chooses a cardinal $\kappa$ larger than all three infinite cofinalities and passes to a maximal $\kappa$-bounded wim-constructible extension $(E^*,O^*)$, where 'wim' abbreviates weakly immediate: the extension is generated by pseudolimits of weakly immediate sequences that have no pseudolimit in the ground field. Fact 2 ensures this extension preserves the residue field, leaves $E$ cofinal in $E^*$, and leaves $v(E,O)^{<0}$ cofinal in $v(E^*,O^*)^{<0}$; with those cofinalities fixed, Lemma 1 forces every definable nested ball family to intersect, making the extension (and by elementarity the original structure) definably spherically complete.
What would settle it
Produce a model $(E,O)$ of $T_{\mathrm{convex}}$ and an $(E,O)$-definable nested family of valuation balls whose intersection is empty; equivalently, find an $(E,O)$-definable subset of $v(E,O)$ whose cofinality is none of $\mathrm{cof}(r(E,O))$, $\mathrm{cof}(v(E,O)^{<0})$, $\mathrm{cof}(E)$, or $1$. Either would directly contradict Theorem 3 or Lemma 1 and settle the matter.
Extended reading notes
Core claim
The central claim is Theorem 3: every model of $T_{\mathrm{convex}}$ is definably spherically complete. In detail, if $(E,O)$ is an o-minimal field equipped with a non-trivial $T$-convex valuation ring, then every $(E,O)$-definable nested family of valuation balls has non-empty intersection. The proof builds a maximal $\kappa$-bounded wim-constructible elementary extension $(E^*, O^*)$ of $(E,O)$ for a cardinal $\kappa$ exceeding the relevant cofinalities; such an extension is $\kappa$-spherically complete, and Fact 2 shows it preserves the residue field, the cofinality of the field, and the cofinality of the negative part of the value group. Lemma 1 then classifies the cofinality of every $(E,O)$-definable subset of the value group as one of four values, forcing every definable ball nest in $(E^*, O^*)$—and hence in $(E,O)$—to meet. Since known results show that non-trivially $T$-convexly valued o-minimal fields defining exponentiation are not spherically complete, the paper concludes that Question 1.1 of [1] has a negative answer for such expansions.
Load-bearing premise
The load-bearing premise is the lemma that every definable subset of the value group has one of four possible 'sizes at infinity'; if that classification fails, the proof collapses, and the printed Fact 2 also appears to contain a sign typo ('>0' where '<0' seems intended) in the cofinality-preservation step used next.
Editorial extensions
If this is right
- Every model of $T_{\mathrm{convex}}$ is definably spherically complete, so no definable nested family of valuation balls can have empty intersection.
- In expansions defining exponentiation, definable spherical completeness coexists with the absence of spherical completeness, giving a negative answer to the motivating question about whether definably spherically complete expansions always have spherically complete elementary extensions.
- The cofinality classification of Lemma 1 is a structural fact about definable subsets of the value group in all $T_{\mathrm{convex}}$ models and may hold independently of the completion argument.
- Maximal $\kappa$-bounded wim-constructible extensions provide $\kappa$-spherically complete elementary extensions that preserve the residue field and the cofinalities of the field and of the negative value group.
Reading between the lines
- One could test whether the same conclusion holds for other classes of valued fields with restricted cofinality spectra for definable subsets of the value group, such as Hensel-minimal or dp-minimal fields.
- Because the proof is non-constructive through a maximal chain, it only asserts existence of intersection points; explicit descriptions in concrete exponential-logarithmic fields remain an open direction.
- The apparent sign typo in Fact 2 should be checked: if the intended statement is that $v(E,O)^{<0}$ is cofinal in $v(E^*,O^*)^{<0}$, the induction works; otherwise the proof of cofinality preservation needs repair.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper claims that every model of the theory Tconvex (non-trivially T-convexly valued o-minimal fields) is definably spherically complete: every definable nested family of valuation balls has non-empty intersection. The proof passes to a maximal κ-bounded wim-constructible elementary extension (E*,O*), uses Fact 2 to preserve the cofinalities of the value group, residue field, and field sort, and applies Lemma 1 to conclude that no definable nested family of balls of small cofinality can exist. The author notes this gives a negative answer to Question 1.1 of [1] for expansions defining exponentiation, since by [3] such fields are not spherically complete.
Significance. If the proof is correct, this is a substantial result: it sharply distinguishes definable spherical completeness from ordinary spherical completeness in a broad class of valued o-minimal fields, and it resolves a question from [1] for a natural family of expansions. The paper is short and relies on the author's earlier work [2] for the construction of κ-bounded wim-constructible extensions; that reliance is not circular because [2] is about weakly immediate types and spherical completions, not about definable spherical completeness. The argument's architecture is coherent: maximal extensions with controlled cofinalities are a standard and promising route. However, the correctness of the proof hinges on a few unproved cofinality identifications in Lemma 1, and on a likely sign typo in the printed statement of Fact 2. The paper is not yet self-contained at the load-bearing step.
major comments (3)
- [Lemma 1, proof] The proof asserts, without proof, that cof(E<b) = cof(r(E,O)) and cof(E>b) = cof(v(E,O)<0). These equalities are load-bearing: they are exactly what allows the cofinality of an arbitrary (E,O)-definable subset of E>0 to be read off from the endpoints of its o-minimal traces. Since the proof of the lemma gives no argument and cites no reference for these identifications, the reader cannot verify the central transfer from subsets of the field to subsets of the residue field and value group. Please supply a proof or a precise citation to a result that establishes these cofinality formulas in Tconvex.
- [Lemma 1, final sentence] The sentence "if X ⊆ E>0, then cof(v(X)) ∈ {1, cof({1/x : x ∈ X})}" is not justified as written. The valuation map v is order-reversing, so cofinality of v(X) is not automatically equal to the cofinality of X or of its reciprocal set; the cofinality of the image under a reversing map depends on initial segments of X, not on the cofinality of X in the usual upward sense. A separate argument is needed for the definable pieces at hand. Without it, the lemma does not establish the claimed bound on cofinalities of definable subsets of v(E,O).
- [Fact 2] The proof of Fact 2 states that [2, Thm. A] gives that "v(E,O)<0 is cofinal in v(E⟨x⟩, O∗∩E⟨x⟩)>0", but the statement of Fact 2 (and its use in Theorem 3) requires cofinality in v(E∗,O∗)<0, i.e. with the inequality '<0' on both sides. This looks like a sign typo in the printed proof. If the '<0' on the target side is indeed meant, please correct the typo; if the intended statement really is '>0', then Fact 2 as stated does not follow and Theorem 3 lacks the cofinality equality cof(v(E,O)<0)=cof(v(E∗,O∗)<0) that it invokes. This is a load-bearing point and must be clarified.
minor comments (4)
- [Abstract and Introduction] There are minor typographical errors in the abstract ('Thi s', 'field') and in the text ('cofinality', 'cofinal') that should be fixed in a revision.
- [Lemma 1, proof] The notation E<b and E>b is used without an explicit definition; please define these as the sets of elements less than b and greater than b, respectively, where b is the element with O < b < E>O.
- [Fact 2] The term "p.c. sequence" is used without recalling its definition; although the reader is referred to [2, Def. 3.15] for the notion of κ-bounded wim-constructible extension, the abbreviation p.c. (probably 'pseudo-Cauchy') should be spelled out or defined.
- [Theorem 3] The conclusion of Theorem 3 is stated as "every model of Tconvex is definably spherically complete", but the proof only treats the expansion (E,O) after passing to E∗. It would help to explicitly state that definable spherical completeness of (E∗,O∗) implies the same for (E,O), since any (E,O)-definable family is also (E∗,O∗)-definable and the intersection property is elementary.
Circularity Check
No significant circularity: Theorem 3 is a new corollary of the author's prior [2] plus a novel Lemma 1; the self-citation is load-bearing but independent.
full rationale
The derivation chain is linear and does not reduce to its inputs by construction. Lemma 1 bounds the cofinality of any (E,O)-definable subset of the value group using [4, (3.10)] and o-minimality; this is a new statement, not a rephrasing of definable spherical completeness. Fact 2 transfers cofinalities from a base model to a wim-constructible extension, citing [2, Thm. A]; this is a separate parameter-free result whose assumptions do not include definable spherical completeness. Theorem 3 combines Fact 2 with Lemma 1 and the existence of maximal κ-bounded wim-constructible extensions from [2]. Every load-bearing input is an independent theorem with its own proof, and the target conclusion is not assumed in any of them. The self-citation to [2] is heavy, but per the rules it is real evidence rather than circularity. The proof of Lemma 1 contains compressed assertions and the printed Fact 2 has a sign typo ('>0' where '<0' appears intended), but these are correctness risks, not circular reductions. No equation or definition in the paper makes Theorem 3 equivalent to Lemma 1 or to [2, Thm. A] by construction.
Assumptions & free parameters
assumptions (4)
- domain assumption T is an arbitrary o-minimal theory expanding RCF and Tconvex is the common theory of non-trivial T-convex valuation ring expansions.
- standard math van den Dries-Lewenberg (3.10): every (E,O)-definable subset of E>0 is a boolean combination of intervals and preimages of O by monotone E-definable functions.
- domain assumption Theorem A of [2]: a pseudolimit of a p.c. sequence without pseudolimits preserves the residue field, makes E cofinal, and preserves cofinality of the negative value group.
- domain assumption A maximal κ-bounded wim-constructible extension is κ-spherically complete.
Cite this review
Pith. "Pith review of $T$-convexly valued o-minimal fields are definably spherically complete." pith.science (2026). https://pith.science/paper/UY6XDSO3
@misc{pith2026241116706,
author = {Pith},
title = {Pith review of: $T$-convexly valued o-minimal fields are definably spherically complete},
year = {2026},
howpublished = {\url{https://pith.science/paper/UY6XDSO3}},
note = {Machine review of arXiv:2411.16706}
}
read the original abstract
I prove the statement in the title using results from arXiv:2404.07646(2). This shows that Question~1.1 in [1] has negative answer for certain expansions of a valued field.
Reference graph
Works this paper leans on
-
[2]
P. Freni. T-convexity, weakly immediate types and T -λ- spherical completions of o-minimal structures, 2024. URL https://arxiv.org/abs/2404.07646
arXiv 2024
-
[1]
D. B. Bradley-Williams and I. Halupczok. Spherically co mplete models of Hensel minimal valued fields. MLQ Math. Log. Q. , 69(2):138–146,
-
[3]
F.-V. Kuhlmann, S. Kuhlmann, and S. Shelah. Exponentiat ion in power series fields. Proc. Amer. Math. Soc. , 125(11):3177–3183, 1997. ISSN 0002-9939,1088-6826. doi:10.1090/S0002-9939-97-0 3964-6. URL https://doi.org/10.1090/S0002-9939-97-03964-6
-
[4]
L. van den Dries and A. H. Lewenberg. T -convexity and tame exten- sions. J. Symbolic Logic, 60(1):74–102, 1995. ISSN 0022-4812,1943-5886. doi:10.2307/2275510. URL https://doi.org/10.2307/2275510
doi:10.2307/2275510 1995
-
[2023]
ISSN 0942-5616,1521-3870
Reviewed August 12, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.