New consequences of PFA( T
Carlos Martínez-Ranero, Lucas Polymeris · 2025
Read New consequences of PFA( T by Carlos Martínez-Ranero, Lucas Polymeris (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-2510-22807.txt.
Introduction
Set theory axioms can force strong facts about trees and graphs. Carlos Martinez-Ranero and Lucas Polymeris study one such axiom in 2025. It is a variant of the Proper Forcing Axiom. The variant protects one almost Suslin tree. Krueger first introduced this axiom.
The pair extend its known consequence list. They prove it implies the Mapping Reflection Principle. It also implies the Open Graph Axiom. All special Aronszajn trees become club isomorphic under it. Yet not all almost Suslin trees do. Aronszajn trees are tall trees with no long branch. Club isomorphism is a strong form of sameness. The results sharpen what this axiom can and cannot do.
Catalog record
| CREATOR | Carlos Martínez-Ranero, Lucas Polymeris |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 47 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-22807
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
New consequences of PFA( T ∗ T^{*} ) Thanks: MSC 2020 : 03E35, 03E65, 03E05 Thanks: Keywords : Aronszajn tree, MRP, OGA, Forcing Axiom Thanks: The first named author was partially supported by Proyecto VRID-Investigación No. 220.015.024-INV Carlos Martínez-Ranero, Lucas Polymeris Abstract . Let T ∗ T^{*} be an almost Suslin tree, that is, an Aronszajn tree with no stationary antichains. Krueger [ Kru20b ] introduced a forcing axiom, PFA ( T ∗ ) \mathrm{PFA}(T^{*}) , for the class of proper forcings that preserve that T ∗ T^{*} is almost Suslin. He showed that PFA ( T ∗ ) \mathrm{PFA}(T^{*}) implies several well-known consequences of the Proper Forcing Axiom ( PFA \mathrm{PFA} ), including Suslin’s Hypothesis and the P-ideal dichotomy. We extend this list by proving that PFA ( T ∗ ) \mathrm{PFA}(T^{*}) also implies the Mapping Reflection Principle ( MRP \mathrm{MRP} ) and the Open Graph Axiom ( OGA \mathrm{OGA} ). Additionally, we show that PFA ( T ∗ ) \mathrm{PFA}(T^{*}) implies that all special Aronszajn trees are club-isomorphic, but it does not imply that all almost Suslin trees are club-isomorphic. An Aronszajn tree is called almost Suslin if it has no stationary antichains. For such a tree T^{*} , Krueger introduced in [ Kru20b ] the forcing axiom \mathrm{PFA}(T^{*}) , a relativized version of \mathrm{PFA} which applies to all proper forcings preserving that T^{*} is almost Suslin. Krueger showed that \mathrm{PFA}(T^{*}) already implies many of the well-known consequences of the proper forcing axiom, including the Suslin’s Hypothesis, the P -ideal dichotomy, and the failure of a very weak form of club guessing. He raised the question of whether further familiar consequences of \mathrm{PFA} also follow from \mathrm{PFA}(T^{*}) , specifically: the mapping reflection principle ( \mathrm{MRP} ), the open graph axiom ( \mathrm{OGA} ) and the non-existence of weak Kurepa trees. A second question posed in [ Kru20b ] , attributed to Moore, asks whether \mathrm{PFA}(T^{*}) implies that all normal almost Suslin trees are club isomorphic. This is a natural analogue of the fact that \mathrm{PFA} implies all normal Aronszajn trees are club-isomorphic; a positive answer would suggest that models of \mathrm{PFA}(T^{*}) are somewhat canonical. In this work, we address these questions. We show that \mathrm{PFA}(T^{*}) imply both \mathrm{MRP} and \mathrm{OGA} , by constructing suitable modifications of the standard proper forcings for these principles. In contrast, we answer Moore’s question negatively: starting from a model of \diamondsuit , we construct two almost Suslin trees T and S that are not club-isomorphic and show they remain so after forcing to obtain a model of \mathrm{PFA}(T+S) . As a positive counterpart, we prove that \mathrm{PFA}(T^{*}) does imply that any two normal special Aronszajn trees are club-isomorphic, thus establishing a partial analogue of the situation under \mathrm{PFA} . We note that subsequent work by Lambie-Hanson and Stejskalová [ LHS25 ] has established that \mathrm{PFA}(T^{*}) implies the non-existence of weak Kurepa trees. Furthermore, [ LHS25 , Fact 6.19(4)] –a fact originally remarked in [ Kru20b ] – can be used to show that \mathrm{PFA}(T^{*}) implies the non-existence of \omega_{2} -Aronszajn trees, via the standard argument from \mathrm{PFA} . The structure of our paper is as follows. After establishing the necessary preliminaries in Section 1 , we proceed to demonstrate the main combinatorial consequences of \mathrm{PFA}(T^{*}) . Specifically, Section 2 is dedicated to proving that the axiom implies the Mapping Reflection Principle ( \mathrm{MRP} ), and Section 3 proves that it implies the Open Graph Axiom ( \mathrm{OGA} ). We then turn to structural consequences for trees in Section 4 , where we resolve the question of whether \mathrm{PFA}(T^{*}) implies that all almost Suslin trees are club-isomorphic. 1. Preliminaries A set-theoretic tree is a partial order set (T,<_{T}) such that for any node t\in T , the set of predecessors \{s\in T:s<_{T}t\} is well-ordered by <_{T} . The height of t , written \hht_{T}(t) , is the unique ordinal representing the order-type of the set of its predecessors. For an ordinal \alpha , the \alpha -th level of T , written T_{\alpha} , consists of the elements of T of height \alpha . The height of T is the least ordinal \alpha for which T_{\alpha} is empty. A branch of T is any downwards closed chain of T , and a branch is called cofinal if its height is equal to the height of T . For each t\in T and \xi\leq\hht(t) , we let t\restriction\xi to be the unique s\in T of height \xi such that s\leq_{T}t . If the tree is clear from the context, then t\perp s denotes the fact that t and s are incomparable in the poset (T,\leq_{T}) . The tree T is said to be well-pruned if for all t\in T , the subtree \{s\in T:s\not\perp t\} has the same height as T . For incomparable t,s\in T , we let \Delta(t,s) to be the least ordinal \alpha such that t\restriction\alpha\neq s\restriction\alpha . We leave \Delta undefined if t and s are comparable. A tree is called Hausdorff if \Delta(t,s) is always a successor ordinal. An \omega_{1} -tree is a tree of height \omega_{1} with all its levels countable. An \omega_{1} -tree is called an Aronszajn tree if it has no uncountable chains, equivalently, no cofinal branches. An \omega_{1} -tree is called a Suslin tree if it has no uncountable chains and no uncountable antichains. A typical argument shows that if T is a well-pruned \omega_{1} -tree, then it is Suslin iff it has no uncountable antichains. A tree is called normal if it is Hausdorff, well-pruned, and every node has \aleph_{0} immediate successors. Definition 1.1 . Let T be a tree with height \omega_{1} . A subset A\subseteq T is called a stationary antichain if A is an antichain and the set \{\hht_{T}(t):t\in A\} is a stationary subset of \omega_{1} . Definition 1.2 . An almost Suslin tree is any Aronszajn tree without stationary antichains. Fix an Aronszajn tree T . The remaining definitions and lemmas of this section are all from [ Kru20b ] . We define a relation <_{T}^{-} by letting x<_{T}^{-}y if \hht_{T}(x)<\hht_{T}(y) and x\not<_{T}y . Definition 1.3 . Let N be a countable set such that \delta:=N\cap\omega_{1} is an ordinal. A node x\in T_{\delta} is said to be (N,T) -generic if for any set A\subseteq T which is a member of N and any relation R\in\{<_{T},<_{T}^{-}\} , if x\in A , then there exists yRx such that y\in A . Lemma 1.4 . ( [ Kru20b , Lemma 1.4] ) Let \theta be a regular cardinal with T\in H(\theta) , and N a countable elementary submodel of H(\theta) containing T as an element. Let \delta:=N\cap\omega_{1} . Assume that x\in T_{\delta} is (N,T) -generic. Let B\subseteq T be in N and R\in\{<_{T},<_{T}^{-}\} . If for all yRx , y\in B , then x\in B . As mentioned in the introduction, we will be mainly interested in the case that T is almost Suslin, one of the reasons being the following. Lemma 1.5 . ( [ Kru20b , Lemma 1.5] ) Let \theta be a regular cardinal such that T\in H(\theta) . Let N be a countable elementary submodel of H(\theta) containing T as an element. If T is almost Suslin,