Extending ideals (Q1093631)

An ideal \({\mathcal I}\) on the set X is said to be \(\kappa\)-extendible if whenever \({\mathcal A}\) is a \(\kappa\)-size family of subsets of X, then there is a \(\sigma\)-ideal on X extending \({\mathcal I}\) which measures each A in \({\mathcal A}\) (i.e. either A or X-A is in \({\mathcal I}\). The ideal \({\mathcal I}\) is said to be (\(\kappa\),\(\lambda)\)-extendible if whenever \({\mathcal A}\) is a \(\kappa\)-size family of subsets of X there is a \(\lambda\)-size subfamily \(\bar {\mathcal A}\) of \({\mathcal A}\) and a \(\sigma\)-ideal on X extending \({\mathcal I}\) which measures each set in \(\bar {\mathcal A}\). The ideal \({\mathcal I}\) is said to be \(\lambda\)-completable if there is a \(\lambda\)-complete ideal on X which contains \({\mathcal I}.\) The first part of the paper concerns \(\kappa\)-extendibility. The main result is that \({\mathcal I}\) is \(\kappa\)-extendible if and only if \({\mathcal I}\) is \((\kappa^{\omega})^+-\) completable, unless \(\kappa\) is greater than or equal to either a weakly compact or another type of large cardinal. There are other results valid provided there are no measurable cardinals in an inner model. The second part of the paper is an extensive discussion of (\(\kappa\),\(\lambda)\)-extendibility, concentrating on the extendibility of the ideals \([\omega_ 1]^{<\omega_ 1}\) and \([2^{\omega_ 1}]^{<2^{\omega_ 1}}\).