ABSTRACT:

Adaptation requires that a entity, whether carbon or silicon
based, be able to analyze itself and its environment.  This
analysis, in turn, allows this entity to react or make decisions 
and carry out actions that improve its ability to survive 
and function in its environment.  

Computational reflection defines the basic requirements needed
to support adaptation.  Namely, the mechanisms that allow a
system to examine and modify its own behavior.  However, should
these two aspects of reflection be tightly tied together?

In this talk, we will examine some of the roots of reflection
and propose how and why reflection can be functionally 
divided.  Additionally, we will further examine how these 
ideas might be used in designing and implementing adaptive systems.