1.  Aggregation:  The first sense of aggregation is to simplify complex systems we aggregate similar things into categories – trees, cars, banks – and then treat them as equivalent.  
>This one of the chief techniques for constructing models.  We decide which details are irrelevant to the question of interest, and then ignore them.  We can thus collect things into a category which differ only in their abandoned details.  The modeler must decide which features to make salient and which to eliminate.
>A second sense of aggregation is more of a matter of what cas do, rather than how we model them.  It concerns the emergence of complex large-scale behaviors from aggregate interactions of less complex agents, such as the emergence of the complex ant hill from the behavior of the less complex ants.
>Aggregates so formed can in turn act as agents at a still higher level – meta-agents.  This process can proceed to result in the hierarchical organization of cas.
>Aggregation in the second sense (the emergence of complex large-scale behaviors from aggregate interactions of less complex agents) is a basic characteristic of cas, and the emergent phenomena which can result are the most enigmatic aspect of cas.
>The study of cas turns on our ability to discern the mechanisms that enable simple agents to form highly adaptive aggregates.  What kinds of “boundries” demarcate these adaptive aggregates?  How are the agents interactions within these boundries directed and coordinated?  How do contained interactions generate behaviors that transcend the behaviors of the component agents?