On \({\mathfrak c}^+\)-chromatic graphs with small bounded subgraphs (Q916653)

The graph \(G\) is said to be \(\kappa\)-good (for cardinal \(\kappa\)) if the number of vertices of \(G\) and the chromatic number of \(G\) both equal \(\kappa\), but every subgraph of \(G\) with cardinality less than \(\kappa\) has chromatic number at most \(\omega\). The author shows that relative to ZF, it is consistent to have ZFC \(+\) \((2^{\omega}\) is as large as you wish) \(+\) (there is a \((2^{\omega})^+\)-good graph). \textit{J. E. Baumgartner} [J. Symb. Logic 49, 234-249 (1984; Zbl 0573.03021)] used a forcing construction to show the relative consistency of ZFC \(+\) GCH \(+\) (there is an \(\omega_2\)-good graph); whether the existence of an \(\omega_3\)-good graph is consistent with GCH is not known. The straight forward generalization of Baumgartner's forcing poset to \(2^{\omega}\) rather than \(\omega_1\) does not work; however a refinement of that ordering restricted to a subset of the original does the job.