True cofinality and bounding number for small products of partial orders. (Q1400599)

The present paper extends part of \textit{S. Shelah}'s pcf theory [Cardinal arithmetic. Oxford: Clarendon Press (1994; Zbl 0848.03025)] to the more general context of products of partial orders. The development of the theory splits into an ideal-theoretic and a combinatorial part. Fix a cardinal \(\kappa\). A function \(\tau\) associating a cardinal \(\tau (I)\) with every ideal \(I\) on \(\kappa\) is called continuous if \(\tau (\bigcap \Gamma) = \min \{ \tau (I) : I \in \Gamma\}\) for all nonempty families \(\Gamma\) of ideals. An ideal \(I\) is stable for \(\tau\) if \(\tau (J) = \tau (I)\) for all ideals \(J\) containing \(I\). For continuous \(\tau\), define the ideal \(J_{<\lambda} (\tau) = \bigcap \{ I : I\) ideal on \(\kappa\) and \(\tau (I) \geq\lambda \}\). Given a partial order \(P\), the bounding number \({\mathfrak b} (P)\) is the least size of an unbounded subset of \(P\). If \(P\) has a cofinal subset which is linearly ordered, the true cofinality \({\text{tcf}} (P)\) is the smallest cardinality of such a set. A family \(\overline P = (P_i : i < \kappa )\) of partial orders is said to be progressive if \({\mathfrak b} (P_i) > \kappa^+\) for all \(i < \kappa\). \({\mathfrak{pb}} (\overline P) = \{ {\mathfrak b} (\prod \overline P / I) : I\) ideal on \(\kappa \}\) is the set of possible bounding numbers. \(\prod \overline P / I\) naturally generalizes \(\prod \overline \lambda /I\) from pcf theory where \(\overline \lambda = (\lambda_i : i < \kappa )\) is a family of regular cardinals. \({\mathfrak b} (\prod \overline P /I)\), then, corresponds to \({\text{tcf}} (\prod \overline\lambda / I)\) (if existing), and \({\mathfrak{pb}} (\overline P)\), to the set of possible cofinalities \({\text{pcf}} (\overline\lambda) = \{ {\text{tcf}} (\prod \overline\lambda / I) : I\) ideal on \(\kappa \}\). The author shows that \(\tau_{\overline P} :I \mapsto{\mathfrak b} (\prod \overline P / I)\) is a continuous function for progressive \(\overline P\). Using Shelah's club guessing, he establishes that for such \(\overline P\) and \(\lambda \in {\mathfrak{pb}} (\overline P)\), \(J_{< \lambda^+} (\tau_{\overline P})\) is generated over \(J_{< \lambda} (\tau_{\overline P})\) by a single set, thus generalizing a well-known result from pcf theory. He also proves that if \(\overline\lambda = (\lambda_i : i < \kappa)\) is a progressive sequence of regular cardinals, \({\mathfrak b} (\prod \overline\lambda / I) = \lambda\) and \(I\) is stable for \(\tau_{\overline \lambda}\), then \({\text{tcf}} (\prod \overline \lambda / I) = \lambda\). This provides the connection to pcf theory.