A note on <i>CM</i>-triviality and the geometry of forking (Q4953232)

A well-known problem in stability theory is the classification of the possible forking geometries. The easiest ones are locally modular (or one-based); a slightly more complicated case has been introduced by Hrushovski and called CM-trivial. In this short note the author introduces a whole hierarchy of geometrical conditions. He calls a theory \(n\)-ample if after naming some parameters there are \(a_0,\dots,a_n\) such that \(\bullet\) \(\text{acl}(a_0,\dots,a_i)\cap\text{acl} (a_0,\dots,a_{i-1},a_{i+1})=\text{acl}(a_0,\dots,a_{i-1})\) for all \(i<n\), \(\bullet\) \(a_n\) is independent of \((a_0,\dots,a_i)\) over \(a_i\) for all \(i<n\), and \(\bullet\) \(a_n\) forks with \(a_0\) over \(\emptyset\). The author remarks that failure of \(1\)-ampleness is equal to one-basedness, and shows that failure of \(2\)-ampleness corresponds to CM-triviality. Moreover, a theory of finite Lascar rank is \(n\)-ample if there is a type \(p\) of Lascar rank \(1\) which is \(n\)-ample (i.e.\ all the \(a_i\) in the definition can be taken from \(p^{\text{eq}}\)). In particular, a theory of finite Lascar rank coordinatized by CM-trivial types is itself CM-trivial. Lastly, he shows that a stable field is \(n\)-ample for all \(n\geq 0\). He conjectures that there are \(n\)-ample but not \((n+1)\)-ample strongly minimal sets for all \(n>0\) (this would generalize Hrushovski's counterexample to Zilber's conjecture), and that there is an infinite simple nonabelian group of finite Morley rank which is not \(3\)-ample (which would contradict Cherlin's conjecture).