Does the method of cases disconcert a theory?