OCA and automorphisms of \({\mathfrak P}(\omega)/\text{fin}\) (Q1803755): Difference between revisions

An automorphism \(\varphi\) of \({\mathcal P}(\kappa)/\text{fin}\) is said to be trivial iff there is some permutation \(e\) of \(\kappa\) so that \(\varphi[a]=^* e''(a)\) for all \(a\subset\kappa\). For the purpose of this review, let us define \(\text{AAT}(\kappa)\) to be the statement ``all automorphisms of \({\mathcal P}(\kappa)/\text{fin}\) are trivial''. Shelah showed via oracle-cc forcing that \(\text{AAT}(\omega)\) is consistent; Shelah and Steprāns showed that \(\text{AAT}(\omega)\) follows from PFA. This paper shows: 1. \(\text{MA}_{\omega_ 1}+ \text{OCA}\) (= open coloring axiom) implies both \(\text{AAT}(\omega)\) and \(\text{AAT}(\omega_ 1)\); the amount of OCA used does not require large cardinals for consistency. 2. PFA implies \(\text{AAT}(\kappa)\) for all infinite \(\kappa\). The author also gives, in ZFC, a sufficient criterion for triviality of automorphisms of \({\mathcal P}(\omega)/ \text{fin}\) which is weaker than previously known criteria. Aronszajn trees, specializing functions, and collapsing maps play key roles in moving from \(\text{AAT}(\omega)\) to \(\text{AAT}(\kappa)\), \(\kappa>\omega\). The proof of \(\text{AAT}(\omega)\) uses almost disjoint families -- under \(\text{MA}_{\omega_ 1}+ \text{OCA}\), if \(\varphi\) is nontrivial on \({\mathcal P}(\omega)/ \text{fin}\) then there is a particularly nice uncountable almost disjoint family of \({\mathcal A}\) so that \(\varphi\) is nontrivial on \({\mathcal P}(a)/ \text{fin}\) for each \(a\in{\mathcal A}\), which contradicts an OCA result proved earlier in the paper. This clear and elegant paper has a typographical error which may cause the reader some difficulty: in the claim appearing in the proof of 4.2, ``\(I\)'' should be ``\(\kappa\)''.