Hirsch-Plotkin radical of stability groups. (Q1950626): Difference between revisions

Let \(G\) be a subgroup of \(\Aut_FV\), where \(V\) is a vector space over the field \(F\). The group \(G\) is unipotent if for each \(g\in G\), the element \(g-1\) is nilpotent; equivalently if \(g\) stabilizes a series of subspaces in \(V\) of finite length. If \(V\) is finite-dimensional or if \(G\) acts finitarily on \(V\), then \(G\) is unipotent if and only if \(G\) stabilizes a series of subspaces in \(V\) and in these cases \(G\) is at least locally nilpotent. In general this equivalence between unipotence and stabilizing a series does not hold. The authors here investigate whether there is a relationship between these two concepts at least under some sort of local nilpotence hypothesis. The authors' main theorem is the following. Suppose \(V\) has countable dimension over \(F\). Let \(L\) be a series of subspaces in \(V\), \(S\) the (full) stability subgroup of \(L\) in \(\Aut_FV\) and \(H\) the Hirsch-Plotkin radical of \(S\). Then \(H\) is a Fitting group and is the set of all elements \(g\) of \(S\) that stabilize some subseries of \(L\) of finite length. In the opposite direction they prove the following. Let \(G\) be some Fitting group that is either torsionfree or a \(p\)-group for some \(p\). Suppose \(G\) is countable, or more generally suppose \(G\) is \(G\)-generated by a countable subset. Then there is a field \(F\) (of characteristic \(p\) if \(G\) is a \(p\)-group and zero otherwise), a vector space \(V\) over \(F\) and a series \(L\) of subspaces in \(V\) such that \(G\) embeds into the Hirsch-Plotkin radical of the stability group of \(L\). A slight word of warning: the theorem of this reviewer quoted near the foot of page 134 is seriously misquoted by the authors.