What the heck is the “Interation” in “Human-Computer Interaction”?

This is my current best stab at it:

1. It’s making new bodies out of computer parts.  We engage the world by seeking out useful sensori-motor loops.  Our bodies are critical parts of these sensori-motor loops.  Some parts of tools are best described as part of our bodies.
2. It’s about modeling computers in our heads.  We construct internal models of external systems, and use these to predict useful interventions in the world.  Over time these model/intervention/result/percept loops become sensori-motor loops as in (1) or they are externalized into extra-personal structures as in (3).
3. It’s being a part of distributed cognitive systems that have computers in them too.  We create and engage structures external to our bodies (other people, objects, culture, social structure) which allow us to take on larger cognitive problems than 1 & 2 alone would permit, or to do the same problems with less resources.

I’d like to draw a picture of this.  Obviously, modeling activities (2) get offloaded into external structures (3).  And sometimes we use modeling (2) to help us find and build external structures (3).  Over time, the structures in (3) can become part of our bodies (1), and we use both our bodies (1) and external structures (3) to scaffold the modeling in (2).  Basically, it’s a big rats nest.  But I still want to draw it.