This is probably the most creative paper I ever wrote. I hope to be able to return to it one day to do some serious research. The approach is to conceptualize people as adaptive systems. So, I propose a "structure of self" analogous to that described for social and computer systems in the Model of Change, my undergraduate thesis.
Reason and emotion, I propose, are actually two points on a spectrum of a single phenomenum, related by a tradeoff of detail for scope: reason is relatively precise, well-defined, well-controlled emotion. Emotional involvement in external problems is a manifestation of top-down problem solving. Our experience of personal growth can then be understood as a product of cycles of accumulation of complexity and reform.
This paper is partially available as RTF and as postscript. Except for the chapter "Learning and the Structure of the Self", the rest has been lost in electronic form.