Uniformity of Consistency in Arithmetic and Gödel’s Second Incompleteness Theorem: Ein Märchen

Harald Grobner · 2026

Read Uniformity of Consistency in Arithmetic and Gödel’s Second Incompleteness Theorem: Ein Märchen by Harald Grobner (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-00266.txt.

Introduction

Harald Grobner examines a recent claim about the consistency of arithmetic. Artemov had argued a subtle point. Gödel's second theorem bars a strong theory from proving its own consistency as one sentence, yet a uniform selector method can still verify each proof attempt in turn. Grobner asks how far that phenomenon reaches.

In this note he argues that it extends to all sufficiently strong, uniformly reflexive theories of a certain form, including set theory. A fixed mechanical selector takes any derivation, picks out the finite set of axioms it uses, and builds a proof that it does not end in a contradiction. The note also analyzes why this step-by-step verification does not collapse into the single forbidden sentence, and places selector proofs within the wider theory of provability and reflection.

Catalog record

CREATORHarald Grobner
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE29 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.00266

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Uniformity of Consistency in Arithmetic and Gödel’s Second Incompleteness Theorem: Ein Märchen Harald Grobner Abstract . In much discussed work Artemov has recently shown that, for PA \mathrm{PA} , the consistency schema admits a form of uniform verification via selector proofs, despite the unprovability of the corresponding uniform consistency sentence Con ⁡ ( PA ) \mathrm{Con}(\mathrm{PA}) . In this note, we recast that this phenomenon extends to all sufficiently strong arithmetizable theories: For such theories T T , there exists a primitive recursive selector producing proofs of all instances of the associated consistency schema. This results – a soft version of a classical result of Pudlák – yields a form of computational uniformity, despite the fact that it cannot be internalized as the uniform consistency sentence of Gödel’s Second Incompleteness Theorem. Our main goal is to analyze this gap and to locate selector proofs within the broader framework of provability and reflection. 2010 Mathematics Subject Classification 03F40; 03F30, 03F07, 03F45; 03E35 Introduction Gödel’s Second Incompleteness Theorem (2.UVS 1 1 1 The unconventional abbreviation stands for Zweiter (=2.) Unvollständigkeitssatz . ), cf. [ Göd31 ] , is commonly taken to establish a definitive limitation on formal theories: No sufficiently strong, recursively axiomatized, consistent theory T , based on a system of axioms \mathcal{L} , can prove its own consistency, provided that consistency is expressed in the standard arithmetical form \mathrm{Con}(T)\equiv\forall x\,\neg\mathrm{Proof}_{T}(x,\ulcorner 0=1\urcorner). This formulation has played a central role in the interpretation of Hilbert’s program and in subsequent debates about the epistemological significance of consistency proofs, cf. [ Res74 , Aue78 , Aue89 , Ste91 ] . The standard conclusion, drawn from Gödel’s 2.UVS, is that a theory capable of formalizing a sufficiently rich fragment of arithmetic cannot, by its own resources, certify the absence of a derivation of contradiction from its own axioms. There is, however, a complementary perspective on this limitation, arising from the analysis of serial properties and their formalization, leading to a form of uniform provability which is not captured by a single universal sentence: For each standard natural number n , let \mathrm{Con}_{n}(T)\equiv\neg\mathrm{Proof}_{T}(\overline{n},\ulcorner 0=1\urcorner), and consider the associated consistency schema \mathrm{Con}^{S}(T):=\{\mathrm{Con}_{n}(T):n\in\mathbb{N}\}. It was shown in [ Art19 , Art25a ] that, for \mathcal{L}=\mathrm{PA} , this schema is provable in the sense of “selector proofs”: More precisely, there exists a primitive recursive procedure producing, for each n\in\mathbb{N} , a proof of \mathrm{Con}_{n}(\mathrm{PA}) . Thus, while the universal sentence \mathrm{Con}(\mathrm{PA}) is unprovable, the corresponding series admits a form of uniform computational witnessing. Otherwise put, \mathrm{PA} proves, in the selector sense, its own associated consistency series. In this short note 2 2 2 A note, which admittedly bears a title that can rightly be considered stolen from R. P. Langlands, or at least contains a clear allusion to a work written by him and highly valued by the author of these lines, see [ Lan79 ] . we first recast that this phenomenon is not specific to \mathrm{PA} , but extends in a natural way to all sufficiently strong arithmetizable theories. More precisely, we show that for any recursively axiomatized theory T whose syntax is representable in a weak arithmetic base, there exists a primitive recursive selector producing proofs of all instances \mathrm{Con}_{n}(T) . In particular, this applies to theories such as \mathrm{ZF} . We refer to Thm. 2.3 and to the discussion in Sect. 2.2 for more information on the latter. One should note that the above mentioned selector proofs are, in a certain way, a reintreptation of results in proof-complexity, notably by Pudlák, [ Pud86 , Thm. 5.5] . We refer to our Rem. 2.4 for a detailed explanation. Hence, given our Thm. 2.3 above, the main goal of this note is to locate selector proofs within the broader framework of provability and reflection and so to draw conclusions for their possible implications for the goals of Hilbert’s program, more specifically, for the provability of consistency, or, even more precisely, for how close or how far this attempt must come to failure. Indeed, in that spirit, and taken all together, the above results suggest a refinement of the usual (rather apodictic) interpretation of Gödel’s 2.UVS: The relevant boundary is not simply between provability and unprovability of consistency statements within a given theory, but between two distinct forms of uniformity : On the one hand, there is a form of computational uniformity, embodied – implicitly by Pudlák’s results on proof-complexity and explicitly by selector proofs – in which a single effective procedure produces proofs of all instances of a given property (proofs, which are in this sense uniform). On the other hand, there is the ordinary uniform consistency assertion, expressed by the single universal arithmetical sentence \mathrm{Con}(T) or by a reflection principle. Selector proofs show that the former can exist even when the latter is provably unattainable. In this sense, they isolate a genuine intermediate level of proof: stronger than mere and trivial instantwise verification, yet strictly weaker than uniform reflection. In particular, the limitation identified by Gödel’s 2.UVS is not a limitation on the existence of uniform computational witnesses, but on the passage from such witnesses to the ordinary universal consistency sentence. From this perspective, the significance of schema-based constructions clearly lies not in challenging Gödel’s 2.UVS, but in revealing a form of uniform computational content that does not collapse into uniform reflection. Acknowledgments: This short note is an outgrowth of the author’s lecture-course "Philosophie und Mathematik", held at the University of Vienna in spring 2026 together with Georg Schiemer. We are grateful to Emil Jeřábek for hinting us to Pudlák’s paper [ Pud86 ] . 1. Standing assumptions and minor recaps 1.1. First-order theories Unless otherwise stated, in this note, T denotes a recursively axiomatized theory, formulated in first-order predicate logic (out of a system \mathcal{L} of axioms, say), extending a weak arithmetic base sufficient for elementary syntactic coding. By the latter we shall (at least) mean elementary arithmetic \mathrm{EA} in the sense of I\Delta_{0}+\mathrm{Exp} , i.e. Robinson arithmetic together with induction restricted to \Delta_{0} -formulas and the totality of exponentiation. However, for stronger metamathematical arguments, in particular those involving derivability conditions and reflection principles, we work over I\Sigma_{1} , i.e., arithmetic with induction restricted to \Sigma_{1} -formulas and we shall indicate this in the statement of our results. 1.2. Basics on provability and consistency Given a system of axioms \mathcal{L} with theory T as in Sect. 1.1 , we choose and fix a standard Gödel numbering and a standard proof predicate \mathrm{Proof}_{T}(p,x), where \mathrm{Proof}_{T}(p,x) expresses that p codes a T -derivation of the formula with Gödel number x . The very details in its formalization as an expression of \mathcal{L} are assumed to be fixed once and for all throughout the paper. Definition 1.1 . The associated provability predicate is \mathrm{Prov}_{T}(x)\equiv\exists p\,\mathrm{Proof}_{T}(p,x). The standard consistency sentence is \mathrm{Con}(T)\equiv\neg\mathrm{Prov}_{T}(\ulcorner 0=1\urcorner), i.e., \mathrm{Con}(T)\equiv\forall p\,\neg\mathrm{Proof}_{T}(p,\ulcorner 0=1\urcorner). Lemma 1.2 . For a standard arithmetization of syntax, \mathrm{Proof}_{T}(p,x) is elementary, indeed primitive recursive, and hence representable in \mathrm{EA} . In common codings it may be taken to be \Delta_{0} after the usual coding of bounded computations. Proof. This is the standard representability theorem for primitive recursive relations. (Verification that a finite sequence is a formal derivation is obtained by checking,

Read the complete text