Posets in which every cutset meets every fibre (Q1297476)

In the case of (finite usually) posets there are many subclasses of constructively interesting posets as well as many classes of axiomatically interesting posets. An example of the former is the class of posets of the form \(C^n_2\) (or here \(K^n_{1,1})\), i.e., the Boolean lattices. An example of the latter is the class of chain-meets-antichain posets. This class is known to be the same as the class of \(N\)-free (4-fence-free) posets according to Rival-Zaguia. It is almost always an interesting as well as a tricky problem to match a class of the first type precisely (i.e., by an if and only if result) with a class of the second type, usually via some \(P\)-freeness theorem as in the result quoted above. In this case the axiomatically interesting class is the class of skeletal, i.e., cutset-meets-fiber posets, where a cutset meets all maximal chains and a fiber meets all antichains and is thus seen to be somewhat analogous to the class of chain-meets-antichain posets. In this paper the authors investigate to what extent the class of skeletal posets is invariant (closed) with respect to various standard constructions of interest. Thus, Theorem 2.3 is important in that it states that a lexicographic sum (product) \(\sum_{x\in P}Q_x\) is skeletal if and only if \(P\) and each of the ``components'' \(Q_x\) are skeletal, which permits the conclusion that finite series-parallel posets are skeletal. If \(P\) and \(Q\) are (finite) posets, then it is not true that \(P\times Q\) must be skeletal if \(P\) and \(Q\) are so. However the converse is true in the finite case, while Theorem 3.3 permits the removal of the finiteness condition on \(P\) and \(Q\) provided each satisfies a ``finitelike'' condition, \(P\) being well-founded (no finite descending chain \(x_1>x_2>x_3\dots)\) and \(Q\) being chain complete (maximal chains have endpoints). Using these results and others in the literature the authors are able to address other specific results as well as arriving at certain useful questions and interesting conjectures, e.g., Theorem 4.1. For any finite poset \(P\), \(C^p_2=2^p=(K_{1,1})^p\) is skeletal if \(P\) is a linear (ordinal) sum of antichains.