Limitations of traditional, tree-like models of knowledge
We have traditionally relied on hierarchical, tree-like models to interpret knowledge and causality. These models, rooted in linear and binary thinking, assume that a single underlying cause can be identified by tracing a clear, direct path from effect back to cause. This approach oversimplifies complex systems, and reduces them to a single-threaded structure that can overlook the complex, interconnected nature of real phenomena. As a result, a tree-like model like Root Cause Analysis is limited in its ability to capture the fluid dynamics and multiplicity inherent in complex systems.
By complex systems, I mean e.g.:
- healthcare systems,
- supply chains,
- the interaction of formal and informal organizational structures,
- ecological systems,
- financial markets and
- social media networks.
The rhizomatic approach - A decentralized, interconnected framework
In contrast to hierarchical models, a rhizomatic approach (Deleuze & Guattari, 1987) offers a more precise framework for understanding complexity. Rhizomes represent knowledge as a vast, interconnected network that is decentralized and nonhierarchical. Rather than focusing on isolated root causes, a rhizomatic model recognizes multiple inputs and outputs, as well as paths and influences that connect and collectively define the system. This model includes escape routes: channels that allow elements within the system to escape rigid structures, evolve, and form new relationships. This flexibility makes the rhizomatic approach better suited to capture the true complexity of systems, where boundaries and categories are fluid.
Adapting to changing conditions
Deterritorialization is the dynamic process by which familiar structures in complex systems are reinterpreted and reshaped in response to internal and external influences. Rather than remaining static, complex systems continually reorganize and adapt, resulting in an evolving redefinition of relationships within the system. RCA, with its focus on isolating a single root cause, struggles to capture this adaptive process. The linear approach of RCA ignores the reality that causes and effects in complex systems do not follow a simple path; instead, they evolve in response to each other, creating complex feedback loops and dependencies.
Dynamic exploration over static replication
In models like RCA, tracing causes produces a static copy of a pre-existing structure. This process fails to accommodate the adaptive nature of complex systems. Mapping, on the other hand, is a dynamic process that is experimental, adaptive, and open. It enables exploration across disciplines and perspectives, rather than simply replicating known paths. Unlike tracing, a map is not a fixed representation, but a tool for navigating the unpredictability and variability of the system. By adopting a mapping approach, analysts can better understand complex systems as they really are: shifting networks of interactions.
Fluidity over static causality
A model like Root Cause Analysis enforces a static, state-centered perspective. It restricts analysis to rigid boundaries that prevent the recognition of emerging patterns and interdisciplinary influences. In contrast, a rhizomatic perspective treats knowledge in complex systems as an evolving network, one that continually redefines itself in response to changing circumstances. Complex systems, like history, are not static or state-centered. They are moving and adaptive, and they prioritize fluidity and interconnectedness over rigid, hierarchical causality.
A networked understanding of complexity
My critique: when dealing with complex systems, please move from fixed, hierarchical frameworks to a more networked, flexible way of understanding. The rigidity of tree-like models limits our understanding of these systems. They are better understood as dynamic, inclusive networks. By adopting the rhizomatic approach, we can come to a richer, more adaptive understanding of the complexity of real-world systems.
The Functional Resonance Analysis Method (FRAM) - A rhizomatic alternative
FRAM (Hollnagel, 2012) is a good fit for a rhizomatic approach. It offers a practical method for analyzing complex socio-technical systems without enforcing a tree-like structure. FRAM captures the dynamic interaction of functions within systems. It focuses on interdependencies and adaptability. Six ways in which FRAM embodies the rhizomatic model:
- FRAM examines complex systems without imposing a hierarchical or linear structure. It recognizes that socio-technical systems consist of dynamic, interdependent functions rather than fixed cause-and-effect chains.
- Rather than simplifying systems into binary success/failure outcomes, FRAM’s focus on task variability and adaptability aligns with the rhizomatic perspective. It captures the resilience of the system by analyzing how daily adjustments contribute to emergent behavior, both expected and unexpected.
- FRAM avoids prescriptive tracing of linear paths and instead opts for a mapping approach that encourages open exploration of potential interactions. This process allows for understanding work as done and capturing the dynamic, real interactions and adjustments within the system.
- In rhizomatic structures, elements continually redefine their connections, allowing new patterns to emerge. Functional resonance aligns with this idea: small variabilities can resonate across interconnected functions, and grow into significant outcomes that defy prediction based on individual components alone.
- FRAM avoids hierarchical decomposition by treating each function as part of a flexible network, where contextual and situational factors allow functions to influence each other. This absence of a central root or rigid order reflects the decentralized nature of the rhizomatic model, which allows each point to be linked to every other point within the system.
- FRAM’s analysis takes into account the daily adjustments to circumstances, resources, and timing that keep systems operational. This attention to contextual adjustment, rather than the imposition of fixed rules, is central to rhizomatic thinking.
Conclusion
By moving from a linear root cause analysis to a rhizomatic approach such as FRAM, we can better understand and manage complex systems. FRAM’s emphasis on mapping, adaptability, and functional interactions provides a robust framework that recognizes the interconnected, evolving nature of complex phenomena. A rhizomatic perspective provides a richer, more accurate picture of complex systems, one that is attuned to the fluid and multifaceted interactions that define them.
We can still use the tree-like models, of course, for e.g. structural systems like bridges and buildings, predictable environments like assembly lines, individual power generation units, automated warehousing systems, et cetera. When a product fails due to contamination or defect, RCA can identify the failure point within a structured manufacturing process. Think of e.g. identifying a contaminated batch in food or pharmaceutical production.
By matching the approach to the system’s complexity, we can use models that reflect the way the system operates, so we can adapt analysis and protocols accordingly.
References
- Deleuze, G., & Guattari, F. (1987). A Thousand Plateaus: Capitalism and Schizophrenia. University of Minnesota Press.
- Hollnagel, E. (2012). FRAM: The Functional Resonance Analysis Method - Modelling Complex Socio-technical Systems. Farnham/Burlington: Ashgate.