Strongly summable ultrafilters on \({\mathbb{N}}\) and small maximal subgroups of \(\beta\) \({\mathbb{N}}\) (Q752863)

As is now well known, the Stone-Čech compactification \(\beta\) \({\mathbb{N}}\) of the discrete space \({\mathbb{N}}\) of positive integers can be given a natural semigroup structure which extends addition in \({\mathbb{N}}\). The general theory of compact semigroups guarantees the existence of idempotents in \(\beta\) \({\mathbb{N}}\), and for each idempotent p there is a largest subgroup H(p) which has p as its identity. It is not hard to prove that H(p) contains (algebraically) a copy of \({\mathbb{Z}}\) (its elements are of the form \(p+n\) with \(n\in {\mathbb{Z}}\), for \(p+n\) makes sense even when n is negative). However, \(\beta\) \({\mathbb{N}}\) is very large, and H(p) would be expected to be much larger. The author proves an extremely remarkable fact: it is consistent with the usual axioms of set theory that \(H(p)=p+{\mathbb{Z}}\), for some idempotents p. The key to this result is to consider strongly summable ultrafilters P on \({\mathbb{N}}\); this means that P has a basis of sets of the form \(FS(<x_ n>^{\infty}_{n=1})\), where for any sequence \(<x_ n>\) in \({\mathbb{N}}\), \(FS(<x_ n>)\) is the collection of distinct finite sums of elements of \(<x_ n>\), that is, \(FS(<x_ n>)=\{\sum_{n\in F}x_ n:\) F is a finite subset of \({\mathbb{N}}\}\). Every strongly summable ultrafilter is an idempotent, but the existence of such ultrafilters cannot be proved in ZFC. Using characteristically difficult and ingenious arguments, the author shows that if P is strongly summable then \(H(p)=p+{\mathbb{Z}}\).