1. Overview: A Major Shift in Network Theory and the Landscape of Innovation

In the history of Social Network Analysis (SNA) and organizational theory, explanations for the origins of innovation and performance have long rested on a balance between two perspectives. One is the theory of structural holes, associated with Ronald S. Burt, and the other is the theory of network closure, associated with James S. Coleman. Together, these views have shaped a familiar message in both academia and practice: to succeed, either become a broker or build strong cohesion.

Burt’s structural holes theory argues that brokers who connect otherwise disconnected groups gain competitive advantage by controlling information flows and accessing non-redundant information. This is a strategy that turns the absence of ties into opportunity, creating value through information arbitrage. In contrast, Coleman’s closure theory emphasizes that dense internal connections foster trust and shared norms, which in turn promote cooperation and prevent free riding. Here, cohesion itself is the key asset, maximizing collective execution. For years, research has focused on which of these values is superior, or how they might be optimally combined, for example in ideas such as “brokerage plus closure.”

In 2010, however, Balázs Vedres and David Stark challenged this established framework with their paper “Structural Folds: Generative Disruption in Overlapping Groups.” Their argument disrupted the existing debate. They claimed that innovation does not arise simply from importing external ideas through brokerage, nor solely from internal cohesion. Instead, they reframed innovation as a process of recombination and introduced a distinctive network configuration that enables it: structural folds.

Structural folds refer to points where multiple cohesive groups overlap. Actors located there are not mere brokers but function as multiple insiders. They are not outsiders bridging distant worlds, but participants who are fluent in the languages and norms of several groups at once, while still being able to connect them. According to Vedres and Stark, it is the generative friction that emerges at these overlapping points that truly drives innovation.

In this post, we revisit the evolution of the “holes versus folds” debate to explore what it means for data analysis and interpretation, and how these seemingly opposing concepts can ultimately be brought together.

☞ Read more