Some linked subsets of posets (Q1099170): Difference between revisions

Many times in forcing one has a name f for a function and one would like to have a condition deciding ``a lot'' about this function. In general this will of course not be possible. In a recent paper [Axiomatic Set Theory, Contemp. Math. 31, 79-97 (1984; Zbl 0565.03024)], \textit{W. G. Fleissner}, bringing work of others under one common denominator, introduced lynxes for posets. Roughly speaking a lynx takes a bite out of every maximal antichain of the poset in such a way that the bites satisfy certain compatibility (linking) conditions. In its most general form a lynx has five parameters. Turning back to the function f, the bites out of the antichains making up its name very often will allow you to construct a function much like f in the ground model. This opens the way to showing that certain posets cannot create a certain type of function unless such a function was already present in the ground model. The author investigates what types of lynxes a poset of the form Fn(I,J,\(\kappa)\) can have. For \(\kappa =\omega\) there are many ZFC results, for \(\kappa >\omega\) one generally needs inaccessible cardinals. These problems arose, among others, in work on the consistency of the Normal Moore Space conjecture and some of its generalizations, to which the author also made important contributions.