Elimination of hyperimaginaries and stable independence in simple CM-trivial theories (Q372631)

In this paper the authors prove several results related to the important questions on elimination of hyperimaginaries and on stable forking in simple theories. The authors prove that in a CM-trivial theory (a generalization of a 1-based theory) every hyperimaginary is interbounded with a sequence of finitary hyperimaginaries and conclude that any such theory eliminates hyperimaginaries whenever it eliminates finitary hyperimaginaries, and that small CM-trivial theories always eliminate hyperimaginaries. In addition they prove that in a supersimple CM-trivial theory the relation \(R(x;y,z)\) defined by: ``\(\mathrm{tp}(x/yz)\) doesn't fork over \(y\)'' is stable; the proof of this applies the machinary of regular types that is available in the supersimple case. A nice corollary of this (obtained in the current paper) is that any omega-categorical supersimple CM-trivial theory has stable forking. They also prove a related result in the general simple case: the relation \(R(x;y,z)\) defined by ``\(\mathrm{tp}(x/yz)\) doesn't fork over \(\mathrm{Cl}(y)\)'' is stable, where \(\mathrm{Cl}(y)\) is the closure of \(y\) with respect to the family of all non-1-based types.