Ramsey theory for highly connected monochromatic subgraphs (Q2036568)

The authors consider the following variant of the Ramsey theorem. A graph \(G=(V,E)\) is:\par \(\kappa\)-\textit{connected} if it remains connected after the deletion of any fewer than \(\kappa\) vertices;\par \textit{highly connected} if it is \(\vert V\vert \)-connected.\par For cardinals \(\kappa,\lambda, \mu,\nu\) the symbol \(\nu\to_{hc}(\mu)^2_\lambda\) means that every coloring of the edges of the complete graph on \(\nu\) into \(\lambda\) many colors contains some size-\(\mu\) monochromatic subgraph which is highly connected. The authors study the problem for which cardinals \(\kappa,\lambda, \mu,\nu\) the relation \(\nu\to_{hc}(\mu)^2_\lambda\) holds. The following facts are proven in the first part of the article.\par (1) If \(\mu\) is an infinite cardinal and \(k<\omega\) then \(\mu\to_{hc}(\mu)_k^2\) holds.\par Now assume that \(\lambda\) is an infinite cardinal. Then \par (2) \(\lambda^+\not\to_{hc}(\mu)^2_\lambda\) for any \(\mu\ge 3\).\par (3) If \(\mu\le 2^\lambda\) then \(\mu\not\to_{hc} (\mu)^2_\lambda\).\par (4) If \(\mu=\mu^\lambda\) then \(\mu^+\to_{hc}(\mu)_\lambda^2\).\par The above facts can be proved in ZFC. The main part of the paper contains proofs of two theorems on consistency with ZFC. \par (5) It is consistent with ZFC+GCH that \(\mu\not\to_{hc}(\mu)_\lambda^2\) holds for all infinite cardinals \(\lambda<\mu\). \par (6) Assuming the existence of a weakly compact cardinal, the relation \(2^{\aleph_1}\to_{hc} (2^{\aleph_1})^2_{\aleph_0}\) is consistent with ZFC.\par To prove (5), the authors introduce the following property \(\square (\mu)\) of a regular uncountable cardinal \(\mu\): there exists a sequence \(\mathcal{C}=\langle C_\alpha\colon \alpha<\mu\rangle\) such that \par (i) \(C_\alpha\) is a closed unbounded subset of \(\alpha\), for each \(\alpha\);\par (ii) \(C_\beta\cap\alpha=C_\alpha\), for every limit point \(\alpha\) of \(C_\beta\);\par (iii) no club \(C\subset\mu\) satisfies \(C\cap\alpha=C_\alpha\) at every limit point \(\alpha\) of \(C\).\par \(\mathcal{C}\) is called a \textit{\(\lambda\)-stationary \(\square(\mu)\)-sequence} if the set \(\{ \alpha<\mu\colon \text{otp}(C_\alpha)=\lambda\}\) is stationary in \(\mu\). The authors observe, for each infinite regular cardinal \(\lambda<\mu\), if \(\mu\) is not Mahlo in the constructible universe then there exists a \(\lambda\)-stationary \(\square(\lambda)\)-sequence, and prove that this yields the relation \(\mu\not\to_{hc} (\mu)^2_\lambda\).\par The statement (6) is a consequence of the following theorem. Let \(\nu\) be a weakly compact cardinal, and let \(\lambda<\mu=\mu^{<\mu}<\nu\) be given. Then there exists a cardinal-preserving forcing \(\mathbb{P}\) which forces ``\(2^\mu=\nu\) and \(\nu\to_{hc}(\nu)_\lambda^2\)''.\par In the last part of the article, the authors present a list of open problems related to this topic.