A new model for all C C -sequences are trivial

Assaf Rinot, Zhixing You, Jiachen Yuan · 2025

Read A new model for all C C -sequences are trivial by Assaf Rinot, Zhixing You, Jiachen Yuan (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-2504-06794.txt.

Introduction

Assaf Rinot, Zhixing You, and Jiachen Yuan build a set-theoretic world with a surprising pair of traits. Their 2025 paper concerns C-sequences, ladders of sets assigned to ordinals, and when they can be trivialized. In the model they construct, every C-sequence is trivial, yet a certain tree still exists with full vanishing levels.

The combination answers a question posed by Lambie-Hanson and Rinot and is optimal, in that compactness and its failure coexist as strongly as possible. The construction extends Kunen's classic forcing method with a device the authors call a mutually exclusive ascent path. Such models chart the exact boundaries of what set theory can force to be true.

Catalog record

CREATORAssaf Rinot, Zhixing You, Jiachen Yuan
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE58 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-06794

Browse the full library · About the Tower of Babel

Full text

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

A new model for all C C -sequences are trivial Assaf Rinot, Zhixing You, Jiachen Yuan Abstract . We construct a model in which all C C -sequences are trivial, yet there exists a κ \kappa -Souslin tree with full vanishing levels. This answers a question from [ RYY23 ] , and provides an optimal combination of compactness and incompactness. It is obtained by incorporating a so-called mutually exclusive ascent path to Kunen’s original forcing construction. 2010 Mathematics Subject Classification Primary 03E35. Secondary 03E05, 03E55 1. introduction Motivated by a characterization of weak compactness in terms of C -sequences due to Todorčević [ Tod87 , Theorem 1.8] , Lambie-Hanson and Rinot [ LHR21 , Definition 1.6] introduced a new cardinal characteristic, the C -sequence number , to measure the compactness of a regular uncountable cardinal \kappa , where the best case \chi(\kappa)=0 amounts to saying that \kappa is weakly compact. This notion is quite useful, for example, if the C -sequence number of a strongly inaccessible \kappa is bigger than 1 (i.e., \chi(\kappa)>1 ), then the \kappa -chain condition is not infinitely productive (see [ LHR18 , Lemma 3.3] and [ LHR21 , Lemma 5.8] ), and there is a \kappa -Aronszajn tree with no ascent path of width less than \chi(\kappa) (see [ IR25 , Lemma 7.9] ). 1 1 1 Undefined terminology can be found in Section 2 . A simple way to argue that \chi(\kappa)=1 holds in a given model is to prove that the weak compactness of \kappa can be resurrected via some \kappa -cc notion of forcing. As a sole example, Kunen’s model [ Kun78 , §3] satisfies this requirement. In [ RS23 ] , Rinot and Shalev put forward the importance of the vanishing levels V(T) of a \kappa -Souslin tree T , deriving from it instances of the guessing principle \clubsuit_{\ad} and using it to solve problems in set-theoretic topology. They proved [ RS23 , Theorem 2.30] that for \kappa weakly compact, \clubsuit_{\ad} fails over every club in \kappa . In addition, if \clubsuit_{\ad} holds over a club in \kappa , then \kappa is immediately seen to be non-subtle. This raises the question of what large cardinal notions are compatible with \clubsuit_{\ad} holding over a club. The best known result in this vein is [ RYY23 , Theorem E] asserting that assuming the consistency of a weakly compact cardinal, it is consistent that for some strongly inaccessible cardinal \kappa satisfying \chi(\kappa)=\omega , \clubsuit_{\ad} holds over a club in \kappa , furthermore, there is a \kappa -Souslin tree T such that V(T)=\acc(\kappa) . Whether this result may be improved to \chi(\kappa)<\omega remained an evasive open problem. Here, this remaining case is resolved. Main Theorem . Suppose that \kappa is a weakly compact cardinal. Then there exists a forcing extension in which \kappa is strongly inaccessible, \chi(\kappa)=1 and there exists a \kappa -Souslin tree T such that V(T)=\acc(\kappa) . A natural attempt to prove the preceding is to follow Kunen’s approach, this time adding a generic \kappa -Souslin tree T such that V(T) covers a club in \kappa , hopefully arguing that the weak compactness of \kappa can be resurrected by a further \kappa -cc forcing. However, in view of the implicit requirement to kill subtle-ness, this demands adding at least \kappa^{+} -many branches through T , which has a negative effect on the chain condition of the further forcing. Instead, our approach is to produce a model admitting a family of generic elementary embeddings living in further \kappa -cc forcing extensions, as follows. We will start with \kappa a weakly compact cardinal, carry out some preparatory forcing below \kappa , and then force to add a uniformly homogeneous \kappa -Souslin subtree T of {}^{<\kappa}\kappa such that V(T) is a club in \kappa and such that T admits an \mathcal{F} -ascent path for an educated choice of a filter \mathcal{F} over \omega . Our plan is to argue that \chi(\kappa)=1 holds in the final model by showing that for every C -sequence \vec{C} over \kappa , a nontrivial elementary embedding j:M\rightarrow N between two \kappa -models with \crit(j)=\kappa and \vec{C}\in M exists in some further \kappa -cc forcing extension. In the inevitable case that our tree T belongs to M , it would admit a branch b:\kappa\rightarrow\kappa , as witnessed by any element of the \kappa^{\text{th}} -level of j(T) . Meanwhile, since V(T) covers a club, \kappa\in j(V(T))=V(j(T)) , and hence the tree j(T) would have a vanishing \kappa -branch b^{\prime}:\kappa\rightarrow\kappa . As b is non-vanishing and b^{\prime} is vanishing, the fact that T is uniformly homogeneous implies that b and b^{\prime} must disagree cofinally often. By iterating such an argument, we infer that the tree must have at least \omega -many cofinal branches that disagree with each other cofinally often. Thus, we shall need a notion of forcing that introduces such cofinal branches in a \kappa -cc fashion. This is exactly where the definition of a mutually exclusive \mathcal{F} -ascent path arises. To connect on a previous remark, instead of adding \kappa^{+} -many branches, in our approach only countably many (mutually exclusive ones) are added. 1.1. Organization of this paper In Section 2 , we provide some necessary background on trees, ascent paths and C -sequences. We motivate the new notion of a mutually exclusive \mathcal{F} -ascent path by showing that variations of Kunen’s forcing that do not make use of this kind of ascent path are not well behaved. In Section 3 , we present our main notions of forcing. Large cardinals play no role in this section, and the results are applicable as low as at \aleph_{2} . In Section 4 , we prove the main theorem by iterating the posets of Section 3 below and at a weakly compact cardinal. 2. Preliminaries 2.1. Trees and vanishing levels For simplicity, instead of working with abstract trees, we opt to work with the following more concrete implementation. A streamlined tree is a subset T\subseteq{}^{<\kappa}H_{\kappa} for some cardinal \kappa that is downward-closed, i.e., for every t\in T , \{t\mathbin{\upharpoonright}\alpha\mathrel{|}\allowbreak\alpha<\kappa\}\subseteq T . The height of a node x\in T is \dom(x) . The height of T , denoted \h(T) , is the least ordinal \beta such that T_{\beta}:=\{x\in T\mathrel{|}\allowbreak\dom(x)=\beta\} is empty. Note that for every ordinal \beta , T\mathbin{\upharpoonright}\beta:=\{x\in T\mathrel{|}\allowbreak\dom(x)<\beta\} is as well a streamlined tree. A streamlined tree T is said to be normal iff for every x\in T and every \alpha<\h(T) , there exists some y\in T_{\alpha} with either x\subseteq y or y\subseteq x . A streamlined \kappa -tree [ BR21 , Definition 2.3] is a downward-closed subset T of {}^{<\kappa}H_{\kappa} satisfying that 0<|T_{\alpha}|<\kappa for every \alpha<\kappa . 2 2 2 In this case, the poset (T,{\subsetneq}) is a set-theoretic \kappa -tree in the usual abstract sense. A streamlined \kappa -Souslin tree is a streamlined \kappa -tree having no \kappa -branches and no \kappa -sized antichains with respect to the ordering \subsetneq . Convention 2.1 . Hereafter, by a tree, we mean a streamlined tree. Definition 2.2 . For all s,t\in{}^{<\kappa}H_{\kappa} , let \Delta(s,t):=\min(\{\dom(s),\dom(t)\}\cup\{\delta\in\dom(s)\cap\dom(t)\mathrel{|}\allowbreak s(\delta)\neq t(\delta)\}). In addition, we define (s*t):\dom(t)\rightarrow H_{\kappa} via: (s*t)(\varepsilon):=\begin{cases}s(\varepsilon),&\text{if }\varepsilon\in\dom(s);\\ t(\varepsilon),&\text{otherwise.}\end{cases} Definition 2.3 . A \kappa -tree T is uniformly homogeneous iff for all s,t\in T , s*t

Read the complete text