Keisler's order is not simple (and simple theories may not be either)

M. Malliaris, S. Shelah

Solving a decades-old problem we show that Keisler's 1967 order on theories has the maximum number of classes. The theories we build are simple unstable with no nontrivial forking,…