Hierarchy and networks

Susan notes that there are modern political systems (such as various terrorist groups) that aren't amenable to traditional attacks on hierarchical systems, because they operate more like a network than a hierarchy. The Internet was designed to be hard to attack from the start, as envisioned by Paul Baran in a 1964 paper about distributed communications, which showed that a distributed network structure was harder to attack than a centralized or "decentralized" (hierarchical) structure; link to cool maps courtesy of Sterling's blog at Infinite Matrix.

Which left me wondering: people often say that human hierarchies are built into our biology, part of our primate heritage. Is human nature shifting as our political and technological systems become more oriented toward the distributed-network approach and less toward top-down hierarchy?

Susan replied that Donna Haraway (who has an agenda, so may not be entirely accurate, but that could be said of most people) suggests that dominance hierarchies in primates are far from the simple linear or tree structures that one might suppose; are in fact networks.

Which leads me to wonder if the line between hierarchy and network in social systems is more blurry than it initially appears. Are things that appear to be clear hierarchies (like feudal systems) actually less clearly structured than they appear? (If there's a king, where do the queen and the king's advisers fit into the power structure?) In a modern computer company, org charts are often filled with "dotted line" indirect-reporting, and executives of small computer companies sometimes ask employees to contact them directly about various concerns, leapfrogging the chain of command.

Then again, Baran's abovementioned 1964 paper distinguishes between "centralized" (one central point that everyone reports to) and "decentralized" (a hierarchical tree); the tree hierarchy has some of the attributes of a distributed network. So maybe I'm talking through my hat, and human systems tend toward tree hierarchies. Certainly terrorist organizations often have a central leader, even if the individual cells behave more like nodes of a distributed network in some ways.

Dunno. Really addressing this stuff would take a lot more research and thought than I have time for, and I imagine others have already addressed it in more detail than I could manage anyway, so I'll leave it at this disconnected rambling.

Join the Conversation