We study a notion of ‘o-amorphous’ which bears the same relationship to ‘o-minimal’ as ‘amorphous’ 191–233) does to ‘strongly minimal’. A linearly ordered set is said to be o-amorphous if its only subsets are finite unions of intervals. This turns out to be a relatively straightforward case, and we can provide a complete ‘classification’, subject to the same provisos as in Truss . The reason is that since o-amorphous is an essentially second-order notion, it corresponds more accurately to 0-categorical o-minimal, (...) and our classification is thus very similar to the one given in 565–592) for that case. More interesting structures arise if we replace ‘interval’ in the definition by ‘convex set’, giving us the class of weakly o-amorphous sets. Here, in fact, there are so many examples that a complete classification seems out of the question. We illustrate some of the structures which these may exhibit, and classify them in certain instances not too far removed from the o-amorphous case. (shrink)

A set is said to be amorphous if it is infinite, but cannot be written as the disjoint union of two infinite sets. The possible structures which an amorphous set can carry were discussed in [5]. Here we study an analogous notion at the next level up, that is to say replacing finite/infinite by countable/uncountable, saying that a set is quasi-amorphous if it is uncountable, but is not the disjoint union of two uncountable sets, and every infinite subset has a (...) countably infinite subset. We use the Fraenkel–Mostowski method to give many examples showing the diverse structures which can arise as quasi-amorphous sets, for instance carrying a projective geometry, or a linear ordering, or both; reconstruction results in the style of [1] are harder to come by in this case. (shrink)