Isn't "complete partial order" the order of the ordinals? Or maybe you need the hyperreals to get the full picture (i.e. to be able to associate a number with e

Isn't "complete partial order" the order of the ordinals? Or maybe you need the hyperreals to get the full picture (i.e. to be able to associate a number with each choice for any set of choices)? I don't know much about this stuff. But I agree utility is a model and it's conditions are quite clearly stated [1], and one may always infer an utility from the given choices as long as they obey those quite brand conditions (and I wouldn't call a choice consistent if they don't obey those -- for instance if you prefer A over B, B over C, and C over A). [1] http://en.wikipedia.org/wiki/Rational_choice_theory http://en.wikipedia.org/wiki/Rational_choice_theory