According to the sources, complexity is a multi-dimensional concept that is variously described as a property of systems, a mental construct of the observer, or a specific type of ill-structured problem known as a “mess”. Consequently, the advice provided for dealing with it ranges from structural engineering and cybernetic regulation to psychological shifts and the use of narrative.
What is Complexity?
The authors in the sources define complexity through several distinct frameworks:
- Systemic Interaction and Emergence: Alicia Juarrero describes complexity as coherent dynamics arising from interactions and feedback loops where the behaviour of each component is constrained by its relationship to the whole 1. This leads to emergence, where novel properties appear that cannot be predicted by studying parts in isolation 2. Herb Simon adds that in such systems, the whole is more than the sum of its parts, making it non-trivial to infer system properties from individual components 3.
- Variety and States: Cyberneticians like Stafford Beer and Ross Ashby define complexity as variety—the number of possible states a system can exhibit 4, 5. Max Boisot views it as the manifestation of variety at work, distinguishing between crude complexity (noise) and effective complexity (regularities) 6.
- Observer-Dependent Phenomenon: James Wilk argues that complexity is not a feature of the world but a perceptual “bug” or a function of our lack of understanding 7. Similarly, John Warfield defines it as a sensation of frustration experienced when human cognitive limits are exceeded by a multi-variable situation 8. Russ Ackoff notes that perceived complexity tends to decrease as familiarity with a situation increases 9.
- Messes and Wicked Problems: Ian Mitroff and Russ Ackoff characterise complexity as a “mess”—a system of highly interactive, strongly coupled problems that cannot be understood independently 9, 10. Rittel and Krippendorff associate it with “wicked problems” that lack definitive formulations or clear stopping rules 11.
- Limits of Logic and Computation: Relational biologists suggest complexity is an intrinsic property where a system requires multiple, non-equivalent, and irreducible descriptions 12. Tim Allen notes that Rosennean complexity describes systems that cannot be fully captured by a formal model or simulated by a computer 13.
Advice on How to Deal with Complexity
The sources offer diverse strategies for navigating complex environments, categorised by the specific perspectives of the authors:
1. Shift from Reductionism to Holism and Synthesis
- Embrace the Whole: Alicia Juarrero advises moving away from “nothing but-ism” to recognise that top-down constraints exert active power over components 14. William Dettmer recommends synthetic thinking, which focuses on functional interactions and the “containing whole” rather than breaking systems into isolated parts 15.
- Identify Constraints: Dettmer further suggests focusing on the system’s “weakest link” or constraint—the “Archimedes Point” where small efforts produce disproportionate benefits 16. James Wilk advises asking negative questions to identify what stops a desired state from occurring 17.
2. Cybernetic Regulation (Variety Engineering)
- Ashby’s Law of Requisite Variety: To control a system, the variety of the regulator must match the variety of the system 18, 19.
- Amplify and Attenuate: Stafford Beer and Patrick Hoverstadt suggest using attenuators to filter out irrelevant environmental noise and amplifiers (like increasing local autonomy) to boost management’s response capacity 19, 20.
- The Black Box Approach: Beer advises that for “unthinkable” systems, it is not necessary to understand the internal detail; instead, manage the system by manipulating inputs and observing output behaviour 21.
3. Structural and Hierarchical Methods
- Hierarchic Decomposition: Herb Simon suggests managing complexity through hierarchies (“boxes-within-boxes”), allowing for the study of short-term subsystem behaviour independently from long-term aggregate dynamics 22, 23.
- Fractal Organisation: Patrick Hoverstadt recommends the Viable System Model (VSM), which uses a fractal structure to handle complexity at the most appropriate local level 24.
- Trimming: In engineering, Triz proponents suggest “trimming”—eliminating expensive components and redistributing their functions to the remaining system to achieve an Ideal Final Result 25.
4. Cognitive and Visual Tools
- Visual Mapping: Eden and Ackermann advocate for Cognitive Mapping and SODA to manage cognitive load and structure the thinking of diverse actors 26. Peter Checkland and the OU Course Material recommend Rich Pictures to capture the “climate” and “muddle” of a situation 27, 28.
- Linguistic Adjustments: John Warfield advises replacing the term “problem” with “problematique”—a graphic structure showing how interrelated problems aggravate one another 29.
- Shifting Metaphors: Lakoff and Lindstone suggest that because no single model is perfect, one must consciously shift between metaphors and move from reflexive to reflective thought to uncover hidden frames 30.
5. Adaptive Action and Learning
- Dissolve Rather Than Solve: Russ Ackoff argues that instead of finding the “best” outcome within a current model, one should redesign the system or environment to remove the problem entirely 31.
- Safe-Fail Design: Ian Mitroff and Alicia Juarrero recommend building “safe-fail” systems that can absorb shocks and recover, rather than attempting to design “fail-safe” systems that are often brittle 14, 32.
- Narrative Explanations: Tim Allen and Alicia Juarrero suggest using narratives to handle complexity, as stories can stably juxtapose contradictory components that formal models cannot 14, 33.
- Iterative Learning: Peter Checkland and MC Jackson advocate for a never-ending learning cycle (such as SSM or Critical Systems Practice) where models are used to facilitate debate and find accommodations for action 27, 34.
6. Data and Attention Management
- The Slip System: Robert Pirsig recommends a random-access system of index cards to manage information in small “chunks,” allowing the mind to remain “empty” and receptive 35.
- Filter Complexity: James Wilk suggests using iterative “rule-out” questioning to eliminate vast subsets of irrelevant data 36.
- Manage Attention: Herb Simon notes that in an information-rich world, the scarce resource is human attention; therefore, intelligent filtering and the use of emotions to prioritise tasks are essential 37.