Boxes in Boxes: Recursion as a Feature of Reality

Introduction to Recursive Systems

Recursiveness (or recursion) is a way to describe the fact that systems are made up of subsystems that replicate or resemble the system itself. A classic example is a fractal, where the same pattern repeats at every scale. In cybernetics and systems theory, recursive systems describe organizations or processes that contain subsystems, each of which functions similarly to the larger system.

Cyberneticist Stafford Beer and systems theorist Jay Forrester were pioneers in this field. They explored the mechanics of control, feedback, and adaptation within recursive systems. Beer's Viable System Model examines how the behavior of organizational systems reflects the behavior of their component structures and relationships. Forrester’s work on system dynamics highlights how feedback loops create recursiveness in dynamic systems.

Donella Meadows, a systems thinker, who studied under Forester, extended these ideas by identifying leverage points—places where small changes can lead to large impacts. Her work integrates well with the concept of recursion, as intervention points exist at multiple levels within recursive structures. In this piece, we will explore how these three thinkers contribute to our understanding of recursive systems and systemic change.

A Box of Beer

For me, Stafford Beer’s Viable System Model (VSM) is central to understanding recursive systems in the context of organizations. The VSM posits that for any system to be viable, it must contain subsystems that mirror the organizational structure of the whole. Each subsystem, whether a department in a company or a unit in a larger operation, behaves autonomously but in alignment with the overarching system. This self-similar, recursive organization allows for flexibility and resilience.

Beer’s recursive model is not purely hierarchical but integrates feedback loops at every level. For instance, a company’s sales department might function as an independent system, making decisions autonomously, but it receives constant feedback from higher-level strategic goals and the broader market environment. In turn, it sends feedback upwards, creating a dynamic recursive structure.

This recursive framework allows systems to adapt and respond to environmental changes. Beer's work stresses that recursion enables both top-down and bottom-up governance, with autonomy preserved at each level. The interplay between different subsystems through feedback is key to maintaining the viability of the entire system.

By framing systems in this recursive way, Beer provided a powerful tool for understanding organizational resilience and control, paving the way for more effective intervention strategies at both micro and macro levels.

The Whole World in a Box

Jay Forrester’s work on system dynamics, particularly his concept of feedback loops, demonstrates how recursion operates in dynamic systems. System dynamics explores how systems evolve over time, often due to the interaction of various feedback loops. These feedback loops, which can either amplify (positive feedback) or dampen (negative feedback) changes, create complex and recurring behavior within systems.

One of Forrester’s most famous models is World Dynamics, which simulates global population growth, resource use, and environmental impact. Within this model, recursive social behaviors (such as short-term profit maximization, in-group/out-group identities, and narrow geographic thinking) become evident through the interaction of feedback loops governing population size, economic activity, and environmental degradation. Each of these variables influences and is influenced by others, creating a web of interdependent and recursive feedback.

Forrester’s approach highlights that systems are not linear; instead, they exhibit patterns where small changes in one part of the system cascade through recursion to create large, often unforeseen consequences elsewhere. Feedback loops create cycles of cause and effect that repeat across different levels of the system. It is the complex interactions between a large number relatively simple and recursive beliefs and responses of human actors that makes predicting long-term outcomes difficult. Forrester teaches us that if we can take and inventory the behaviors at work, we can insight into where interventions can disrupt harmful patterns.

Finding the Right Box for the Job

Read Donella Meadows’ Places to Intervene in a System. (https://www.donellameadows.org/wp-content/userfiles/Leverage_Points.pdf). It’s awesome.

Meadows’ leverage points approach gives us a structure for taking meaningful action somewhat detatched from the systems we participate in. In her hierarchy of leverage points, Meadows identifies twelve places where small interventions can cause significant system-wide changes. These range from changing numerical parameters, like subsidies or taxes, to shifting paradigms, which govern the mindset of the entire system.

In recursive systems, leverage points exist at multiple levels. For instance, an intervention aimed at altering feedback loops (one of Meadows’ leverage points) can ripple through a system recursively. At the level of feedback loops, small changes, such as adjusting a production quota, can propagate through the system, causing shifts in behavior at both local and higher levels.

The most powerful leverage points are those that influence the underlying goals, rules, and paradigms of the system. By recognizing the recursive structure, Meadows argues that changes to the deep, foundational levels of a system can have profound cascading effects, affecting every recursive layer.

This perspective reinforces the idea that recursive systems are not impenetrable. By understanding where the most effective leverage points lie, systemic change can be achieved, even within the most complex recursive frameworks.

Stacking and Unstacking Boxes

Systems often resist change. But the concept of recursive systems gives us a tool. If we recognize that all systems (including complex ones) are built of components that behave in recursive ways, then we can look to the small behaviors and the small structures that drive those behaviors. If we understand the logic, the values, and behaviors of the system, we can change the outputs of the system and the outcomes for the wider world.

To be effective, interventions must be strategic and tailored to fit the system. Sometimes brute opposing force will work, but usually, something more akin to the Jujitsu concept of Kuzushi – redirecting the force generated by the system itself – is the better approach.