Based on the sources, here is a structured list of practical advice on “how to act” derived from the specific worldview of each author/source.

Alicia Juarrero

Context-Dependence: Do not treat a system as a mere aggregate of parts. You must account for context-dependent constraints, as these are what enable the parts to self-organize and produce emergent properties[1],[2].

C. West Churchman

Sweep In the Environment: Do not limit your view to the system’s immediate boundaries. You must “sweep in” the environment—including ethical, political, and moral considerations (even your “enemies”)—to truly understand the implications of your design[3],[4].

Claude Shannon (and associated complexity sources)

Probe, Sense, Respond: In unpredictable systems, do not impose rigid control or detailed requirements. Instead, use a “probe-sense-respond” approach, launching safe-to-fail experiments to allow solutions to emerge[2].

Colin Eden

Manage Complexity, Don’t Reduce It: Do not ignore the richness of a problem just to simplify it. Use methods like cognitive mapping to manage the complexity without destroying the interconnected “mess” that makes the problem real[5],[6].

Dave Snowden

Manage Starting Conditions: Stop trying to design an ideal future state, as the complex domain is retrospective. Instead, manage the “starting conditions” and the “evolutionary potential of the present” to steer the system[7],[8].

David L. Abel

Distinguish Choice from Chance: Recognize that true organization requires “Choice Contingency” (purposeful selection). Do not confuse “self-ordering” physical events (like a tornado) with “organized” systems (like a circuit) that require formal controls[9],[10].

Dee Hock

Address the Interior: To lead an organization, you cannot just manage the “exterior” (observable behavior). You must equally address the “interior” domain of collective consciousness, culture, and shared meaning[4],[11].

George Lakoff

Abandon Direct Causation: Stop looking for direct “A causes B” links. Adopt “systemic causation,” recognizing that in social and ecological spheres, causes are often indirect, diffuse, and interactive[12],[13].

Gregory Bateson

Participate, Don’t Control: Abandon the fallacy of “unilateral control.” Move from a stance of manipulating an external object to one of “participation” and mutual learning within the system[14].

Herbert Simon

Use Near-Decomposability: To understand a complex architecture, look for its stable subsystems. Analyze the system as “nearly decomposable,” focusing on the intermediate forms that allow the hierarchy to evolve and function[15],[16].

Horst Rittel

Accept No Stopping Rule: When dealing with social “wicked problems,” do not expect a definitive solution. Accept that there is no “stopping rule” (no point where the problem is strictly “solved”) and that every attempt to fix it leaves irreversible traces[7],[17].

Humberto Maturana

Trigger, Don’t Instruct: Understand that you cannot “instruct” a system to change from the outside. Your interventions are merely “triggers”; the system determines its own structural change based on its internal organization[15],[18].

James Wilk

Filter the Flux: Do not attempt the impossible task of modeling the whole complex reality. Instead, use “minimalist intervention” to set constraints that filter the “flux” of events into a manageable outcome[19],[20].

John Warfield

Structure Your Thinking: Overcome the “frustration” of complexity by using formal logic and “Structural Thinking.” Map the relational patterns among the components to defend against behavioral pathologies[21],[22].

Michael C. Jackson

Use Multi-Methodology: Do not rely on a single tool. Adopt a pluralistic approach, selecting different methodologies (mechanical, organismic, or emancipatory) depending on the specific type of complexity you face[3],[23].

Max Boisot

Operate at the Edge of Chaos: Do not seek total stability. Position the organization at the “edge of chaos,” a zone of instability where interactions between agents are maximized to allow for adaptation and emergence[24],[25].

Mike McMaster

Use Attractors: Since you cannot predict the details of a living system, do not try to control it. Use “attractors” (principles or values) to influence the general shape and direction of the system’s emergence[21],[26].

Niklas Luhmann

Practice Selectivity: Recognize that the system must ignore most of the environment to function. You must practice “selectivity” because it is impossible to connect every element to every other element[27],[28].

Patrick Hoverstadt

Look for Drivers: Do not focus on the surface effects or the stability of the system. Look for the “underlying drivers” and the dynamic processes that are constantly changing the system[29],[30].

Paul Cilliers

Be Modest: Abandon the “arrogance” of seeking a totalizing model. Adopt a “modest” position, acknowledging that all descriptions of complex systems are limited, provisional, and involve ethical choices[2],[31].

Peter Checkland

Focus on Learning: Do not treat the system as a description of the world. Use “the system” as a mental device to structure debate, facilitate learning, and accommodate conflicting worldviews among stakeholders[3],[32].

James Ladyman & Ross Ashby

Simulate Emergence: When a system is too organized for statistics but too complex for analysis, use computational simulation (like agent-based modeling) to observe how simple rules generate complex aggregate behaviors[4],[33].

Relational Biologists (Rosen/Pattee)

Accept Non-Simulability: Acknowledge that complex living systems have no “largest model.” You must use complementary descriptions (e.g., dynamic laws vs. rules) that are formally incompatible but jointly necessary to capture the whole[34],[35].

Robert Flood

Manage Within: Do not try to “manage over” the totality of a system. You must learn to manage and organize “within” the unmanageable and unorganizable[7],[36].

Robert Pirsig

Value Stuckness: If you get “stuck,” do not panic. Value this state as the failure of your static patterns, which creates the necessary opening for “Dynamic Quality” (new insight) to emerge[21],[37].

Russ Ackoff

Dissolve, Don’t Solve: Do not try to “solve” a problem by isolating the parts. Attempt to “dissolve” the problem by redesigning the larger system or its environment so the problem no longer arises[6],[38].

Stafford Beer

Match Variety: To control a system, you must ensure your control mechanism has as much “variety” (number of possible states) as the system itself. “Only variety can absorb variety”[39],[40].

The Other Group (TOG/MOM)

Cultivate Resilience: Stop seeking certainty or “truth” in complex systems. Focus on “cultivating resilience” in your decisions to survive the inevitable uncertainty[39],[41].

Theory of Constraints (Dettmer)

Manage the Constraint: Do not try to manage every variable. Identify the system’s “constraint” (or Archimedes point), as this single leverage point governs the performance of the entire system[29],[42].

Tim Allen

Use Narrative: When formal models fail due to contradiction and uncertainty, use “narrative” to bridge the gap between the different levels and scales of the system[24],[43].

Triz (Bukhman)

Pursue Ideality: Guide the system toward “Ideality” by identifying and resolving the contradictions that prevent the system from performing its function without cost or harm[34],[44].