Part Ix Afterword

Someone has to go last and in this case the honor falls to Mike Byrne. I did not commission Mike to write “a last chapter” but that is what he did. In a way his chapter 30 and my chapter 1 can be viewed as bookends for the 28 chapters that lie between. I began the book with a short essay on three types of control of cognition that need to be considered when building integrated models of cognitive systems. Not wanting to overfit the data, I did not try to sharply define what these distinctions were. Instead my goal was to sketch some broad distinctions that might make conversations between research camps easier and, indeed, might give those who build integrated models a new vocabulary to discuss their models and to communicate their needs to those who build single-focus models of cognitive functions or, as Mike puts it, local theories. Before writing chapter 1, my discussion of Type 1, 2, and 3 control was limited to several paragraphs on the now defunct workshop Web site. I am pleased that, in addition to Mike’s chapter, three other chapters have found a good use for this vocabulary (Gluck, Ball, and Krusmark in chapter 2; Gunzelman, Gluck, Price, Van Dongen, and Dinges in chapter 17; and Ritter, Reifers, Klein, and Schoelles in chapter 18). I am especially pleased that Mike seems to have used the Type 1 versus Type 2 distinction as a call-to-arms for the cognitive community to come together and start building the types of models needed to create integrated models of cognitive systems. His perspective is that of a Type 1 modeler in search of Type 2 components that can be plugged and played in his models. His appeal is very direct, very scholarly, and very personal.