In order to properly examine the relation between emotions and the self-concept, Frijda believes it is necessary to distinguish between the self as object of knowledge and the transcendental self. He makes a compelling case for the importance of the self as object of knowledge, but I have trouble understanding exactly what he means by the transcendental self. At one point he defines it as “the sum total of the information processes at a functional level” and at another as a person's “set of concerns.” These are both vague constructs that require clarification, and, unless I am missing something, they are not even the same.

Others, as well as I, have proposed more clearly articulated distinctions between two somewhat similar constructs to the ones proposed by Frijda. For example, William James (1910) initially, and Gordon Allport (1955) later, expressed the difference between the two selves in terms of the self as an object of knowledge and the self as an executive self that is the source of behavior. Both, however, eventually disowned the executive self as scientifically indefensible, likening it to a homunculus residing inside a person's head and directing the person's behavior. The problem with the executive self and its homunculus analogy is that understanding their behavior is no simpler than understanding the behavior of the whole person and therefore does nothing to further our understanding. I will discuss a more meaningful representation of the executive self later, when I present my own views.