Model-theoretic characterizations of large cardinals (Re) 2 visited
Will Boney, Jonathan Osinski · 2025
Read Model-theoretic characterizations of large cardinals (Re) 2 visited by Will Boney, Jonathan Osinski (2025) 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-2505-15574.txt.
Introduction
Will Boney and Jonathan Osinski characterize giant infinite sets through logic in a 2025 paper. Large cardinals are infinities so big that standard axioms cannot prove they exist. The authors ask which of these notions match neat properties of stronger logics, continuing an earlier joint work.
Several strong cardinal types turn out to equal compactness style properties for logics enriched with special model notions. Huge cardinals match a compactness property for a logic that can demand well founded models. The team also pins down the strength of second order logic and its sorted cousins to the first extendible cardinals of matching kinds. The results draw a tidy map between infinity sizes and logical strength.
Catalog record
| CREATOR | Will Boney, Jonathan Osinski |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 65 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-15574
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Model-theoretic characterizations of large cardinals (Re) 2 visited Will Boney, Jonathan Osinski Abstract We characterize several large cardinal notions by model-theoretic properties of extensions of first-order logic. We show that Π n \Pi_{n} -strong cardinals, and, as a corollary, “Ord is Woodin” and weak Vopěnka’s Principle, are characterized by compactness properties involving Henkin models for sort logic. This provides a model-theoretic analogy between Vopěnka’s Principle and weak Vopěnka’s Principle. We also characterize huge cardinals by compactness for type omission properties of the well-foundedness logic 𝕃 ( Q WF ) \mathbb{L}(Q^{\text{WF}}) , and show that the compactness number of the Härtig quantifier logic 𝕃 ( I ) \mathbb{L}(I) can consistently be larger than the first supercompact cardinal. Finally, we show that the upward Löwenheim-Skolem-Tarski number of second-order logic 𝕃 2 \mathbb{L}^{2} and of the sort logic 𝕃 s , n \mathbb{L}^{s,n} are given by the first extendible and C ( n ) C^{(n)} -extendible cardinal, respectively. All results of this paper are also part of the second author’s PhD thesis [ Osi24 ] . This material is based upon work while the first author was supported by the National Science Foundation under Grant No. DMS-2137465 and DMS-2339018. 1 Introduction Large cardinals are closely tied to model-theoretic properties of strong logics , i.e., of extensions of first-order logic. Early on in the study of large cardinals, Tarski proposed weakly and strongly compact cardinals still prominent today as cardinals witnessing certain compactness principles for infinitary logics (cf. [ Tar62 ] ). Subsequently, several other large cardinals such as supercompact and extendible cardinals received model-theoretic characterizations (cf. [ Mag71 , Ben78 ] ). Further, Vopěnka’s Principle was shown to be equivalent to axiom schemas postulating that all logics haver certain downward Löwenheim-Skolem and compactness properties (cf. [ MV11 , Mak85 ] ). In the last years, there has been a resurgence of this type of research and the large cardinal strengths of a series of model-theoretic properties of strong logics were considered (cf. [ MV11 , BV16 , Bon20 , GKV25 , BDGM24 , HLM24 , GO25 , OP25 , Osi24 ] ). While the above cited results on connections between Vopěnka’s Principle and model-theoretic properties of logics provided some global equivalence, crucially, many of these newer results (e.g. from [ Bon20 , GO25 , OP25 , Osi24 ] ) allow us to understand this relationship in a much more fine grained way. More concretely, the first author [ Bon20 ] showed that, for a natural number n , the existence of C^{(n)} -extendible cardinals is equivalent to the existence of a compactness number for the sort logic \mathbb{L}^{s,n} (cf. Sections 2.1 and 2.2 for definitions). The C^{(n)} -extendible cardinals were introduced by Bagaria (cf. [ Bag12 ] ) as a strengthening of the more familiar extendible cardinals and he showed that axioms stating their existence form, with rising n , a hierarchy of stronger and stronger axioms bounded by Vopěnka’s Principle. Moreover, he showed that Vopěnka’s Principle can be restricted in uniform ways such that these restrictions are precisely equivalent to the existence of C^{(n)} -extendible cardinals. The first author’s results about compactness numbers of sort logic thus show that this hierarchy below Vopěnka’s Principle is precisely mirrored by compactness properties of logics. Combining results of Poveda [ Pov20 , Theorem 5.2.3] and the first author [ Bon20 , Theorem 4.13] one can show that a similar connection holds between Löwenheim-Skolem properties of sort logics and C^{(n)} -extendible cardinals. And strengthening the notions of Henkin models for abstract logics (cf. Section 3 ) to strong Henkin models, Poveda and the second author [ OP25 ] showed that certain compactness principles for strong Henkin models of sort logic provide yet another column to this hierarchy. This paper can be seen as a follow up paper to [ Bon20 ] and [ BDGM24 ] and adds new model-theoretic characterizations of several large cardinals. This provides new insights into the relation of Vopěnka’s Principle to model theory of strong logics. We will show that by considering upward Löwenheim-Skolem properties of strong logics provides a fourth characterization of the hierarchy of large cardinal axioms below Vopěnka’s Principle by model-theoretic means (cf. Section 6 ). More concretely, we show that the ULST number introduced in [ GKV25 ] of second-order logic \mathbb{L}^{2} is the first extendible cardinal. This confirms a conjecture from [ GKV25 ] . The result generalizes to sort logic, in the sense that the existence of a ULST number of the sort logic \mathbb{L}^{s,n} is equivalent to the existence of a C^{(n)} -extendible cardinal. As a corollary, we get a characterization of Vopěnka’s Principle. Moreover, we will consider what is known as weak Vopěnka’s Principle , which is an axiom schema motivated by a category-theoretic characterization of Vopěnka’s Principle (cf. [ ART88 ] ). Recently, Wilson [ Wil20 , Wil22b ] showed that weak Vopěnka’s Principle is indeed strictly weaker than Vopěnka’s Principle, solving a longstanding open problem. Bagaria and Wilson [ BW23 ] showed that below weak Vopěnka’s Principle, there is also hierarchy of stronger and stronger large cardinal axioms, stating the existence of what they call \Pi_{n} -strong cardinals. There are thus analogous patterns (cf. [ BL24 ] ) in the large cardinal hierachy below Vopěnka’s Principle and weak Vopěnka’s Principle. Using compactness properties for Henkin models, in Section 3 we will give a characterization of \Pi_{n} -strong cardinals. This shows that the pattern below weak Vopěnka’s Principle can also be obtained by model-theoretic properties of strong logics, as was the case for the pattern below Vopěnka’s Principle, establishing a novel analogy between the two. As a corollary, we obtain a model-theoretic characterization of weak Vopěnka’s Principle. We will further show that huge cardinals – and thus, in particular, cardinals with higher consistency strength than Vopěnka’s Principle – can be characterized by a model-theoretic property of the logic \mathbb{L}(Q^{\text{WF}}) , answering a question of Wilson (cf. Section 4 ). Finally, we will show how to separate the compactness number of the logic \mathbb{L}(I) from supercompact cardinals (cf. Section 5 ). 2 Preliminaries 2.1 Abstract model theory We will use the following standard notions from abstract model theory. A language is a set of (typically) finitary relation, function, and constant symbols. An abstract or strong logic \mathcal{L} consists of a definable class function that assigns to every language \tau a set of sentences \mathcal{L}[\tau] , and of a definable relation \models_{\mathcal{L}} that may hold between \tau -structures \mathcal{A} and members of \mathcal{L}[\tau] . If \mathcal{A}\models_{\mathcal{L}}\varphi , we say that \mathcal{A} is a model of \varphi . There are several standard niceness properties one assumes for the relation \models_{\mathcal{L}} , e.g., that isomorphic structures are models of the same sentences. A precise definition of the properties needed can be found in [ BDGM24 , Definition 2.1] (which is itself a sublist of classical definitions). Note that we assume that \mathcal{L}[\tau] is always a set, as opposed to a proper class. If \sigma and \tau are languages, we say that a renaming is a bijection f:\sigma\rightarrow\tau which restricts to bijections between the sets of relation, function, and constant symbols in \sigma and \tau , all while preserving their respective arities. We may turn a