**Problem:** When considering the issue of homeless people and political boundaries, all kinds of strategies are considered. Most strategies yield unresolvable economic or ethical consequences. So generally, the issues are ignored and get worse.
Boundaries and borders create dilemmas.
One approach is to consider the boundaries for each individual subsystem rather than designating (dreaming up) averaged boundaries for groups of subsystems (individuals).
Such boundaries will be functional boundaries rather than political boundaries.
This is related to Ostrom's polycentric governance and Alexander's A City is Not a Tree, or a semi-lattice.
What if each homeless person or family was considered as a system/subsystem within an ecosystem(s)? Solutions to the situations for that individual subsystem in a particular set of ecosystems is going to be more "solvable".
Patterns that have "worked" should be shared, not imposed. Supporting policies could emerge from shared patterns that work for particular people in particular sets of situations. This is in some ways the opposite of creating policies and force fitting people and situations to compliance with the policies.
> Of course all of this analysis is just trying to learn how to use and talk about Ashby's "requisite variety."
In a way this approach can avoid using pseudo-metasystem approaches when the problem is not one of subsystem to metasystem but simply inappropriate aggregation of parts at the same level.
Only complex adaptive systems can deal with the level of complexity inherent in issues like homelessness. Complicated approaches and Simple approaches are bound to fail and likely allow the situation to worsen.
> Simple, Complicated, Complex, Chaotic refer to Dave Snowden's Cynefin framework for problem solving. The term Cynefin is an attempt to draw attention to the particularity of problems and the need for complex adaptive systems' approaches.
So, the question becomes how can simple and complicated mechanisms help rather than harm complex adaptive mechanisms?