Kernberg's and Kohut's ideas concerning the idealizing transference are explored. It is postulated that their disparate ideas are explicable in terms of their different beliefs concerning human nature. The ontological status of these beliefs cannot be verified but they lead Kernberg and Kohut to adopt different observational stances. These stances influence their interpretations of the data perceived and the theories they construct. It is argued that Kernberg and Kohut are not unique in this regard. The link between psychological theories and deeply held convictions about human nature is an inextricable one. The controversy between Kernberg and Kohut simply serves to illustrate this.