Peter_de_Blanc comments on Expected utility without the independence axiom