The size of \(\widetilde T\) (Q1590663)

For \(\kappa\in \{\omega_1, \omega_2\}\), let \(NS_\kappa\) denote the nonstationary ideal on \(\kappa\) and set \(NS^+_\kappa= P(\kappa)- NS_\kappa\) and \(NS^*_\kappa= \{T\subseteq \kappa:\overline T\in NS_\kappa\}\), where \(\overline T= \kappa- T\). The following definition is due to Woodin. Given \(T\subseteq \omega_1\), \(\widetilde T\) denotes the set of \(\beta\) such that \(\omega_1< \beta< \omega_2\) and \(\{\gamma\in \omega_1: \text{o.t.}(R\mid\gamma)\in T\}\in NS^*_{\omega_1}\) for every well-ordering \(R\) of \(\omega_1\) of length \(\beta\). The paper under review is a detailed study of the size of \(\widetilde T\). Here are some of its results: (1) If \(NS_{\omega_1}\) is saturated, then \(\widetilde C\in NS^*_{\omega_2}\) for every \(C\in NS^*_{\omega_1}\). (2) \(\diamondsuit\) implies that there is \(T\in NS^+_{\omega_1}\) such that \(\overline T\in NS^+_{\omega_1}\) and \(\widetilde T= \overset\simeq T= \phi\), and \(C\in NS^*_{\omega_1}\) such that \(\widetilde C= \phi\). (By a result of Woodin, it is consistent relative to the existence of a Woodin cardinal that \(NS_{\omega_1}\) is presaturated and \(\diamondsuit\) holds.) (3) It is consistent relative to the existence of a Woodin cardinal that \(NS_{\omega_1}\) is saturated and there exists \(T\in NS^+_{\omega_1}\) such that \(\overline T\in NS^+_{\omega_1}\) and \(\widetilde T=\overset\simeq T=\phi\). (4) It is consistent relative to the existence of a supercompact cardinal that PFA holds and \(\widetilde T=\phi\) for some \(T\in NS^+_{\omega_1}\). (5) \(\text{SRP}(\omega_2)\) implies that for any \(T\in NS^+_{\omega_1}\), \(\{\alpha\in\widetilde T: \text{cf}(\alpha)= \omega\}\) is unbounded in \(\omega_2\) and \(\{\alpha\in\widetilde T: \text{cf}(\alpha)= \omega_1\}\in NS^+_{\omega_2}\). A byproduct of the proof is that \(\psi_{AC}\) follows from \(\text{SRP}(\omega_2)\), a result due independently to Woodin. (Woodin showed that MM implies that \(\{\alpha\in\widetilde T: \text{cf}(\alpha)= \omega\}\in NS^+_{\omega_2}\) for every \(T\in NS^+_{\omega_1}\). Todorcevic showed that \(\text{MM}(c)\Rightarrow \text{SRP}(\omega_2)\Rightarrow NS_{\omega_1}\) is saturated.) (6) Let \(\chi\) be the statement that \(\{\alpha\in\widetilde T: \text{cf}(\alpha)= \omega\}\in NS_{\omega_2}\) for every \(T\subseteq \omega_1\) with \(\overline T\in NS^+_{\omega_1}\). Then it is consistent relative to MM that \(\chi\) and ``MM holds for all forcings which do not change the cofinality of \(\omega_2\) to \(\omega\)'' both hold. (7) It is consistent relative to the existence of a pointclass \(\Gamma\subseteq P(\mathbb{R})\) which is closed under continuous preimages and such that \(L(\Gamma, \mathbb{R})\models \text{AD}_{\mathbb{R}}\) \(+\) ``\(\Theta\) is regular'' that \(\text{MM}^{++}(c)\), \((*)\) (the \(\mathbb{P}_{\max}\) axiom) and \(\chi\) all hold simultaneously.