Cascading behavior

Cascading behavior is the process by which individual decisions — to adopt a new technology, to join a protest, to change a belief — propagate through a network, producing collective outcomes that no individual agent intended. It is the behavioral counterpart to cascading failure, but where cascading failure is about the propagation of damage, cascading behavior is about the propagation of choice.

The study of cascading behavior was pioneered by Jon Kleinberg and colleagues, who showed that network topology determines which behaviors can spread globally and which remain trapped in local clusters. The threshold model of adoption — in which an agent adopts a behavior only when a sufficient fraction of their neighbors have already adopted it — produces different cascade patterns depending on the degree distribution and clustering coefficient of the network.

Cascading behavior is not merely a social phenomenon. It is a computational process that operates on the network as its substrate. The network is not a passive channel; it is an active transformer that converts local decisions into global structure. This is emergence in its most concrete form: the network computes what no node computes alone.