Complex Adaptive Modelling System — Working Paper
The Complex Adaptive Modelling System (CAMS) was not designed to be a grand unified theory of history. It began from a deliberately modest ambition: to find the simplest possible model — a putative model — capable of reading the pattern of power in human societies without pretending to predict individual events. That intellectual pressure-free starting point, grounded in classical Socratic reasoning and the conventions of philosophical and scientific realism, turned out to be essential. The goal was not argument but measurement; not persuasion but pattern recognition.
At its core, CAMS is a cataloguing instrument. What it records, most clearly, is the changing pattern of consensus — the shifting distribution of coherence, capacity, stress, and abstraction across a society's institutional nodes over time. Even in its most static reading, this yields something genuinely useful: a thermodynamic snapshot of how power is organised and where it is leaking.
The theoretical foundation of CAMS rests on a claim that, while not yet universally accepted, has not been falsified: societies are complex adaptive systems. More specifically, they belong to the class of dissipative systems — structures that maintain their internal complexity by continuously drawing energy and resources from their environment and exporting entropy. All such systems, from convection cells to ecosystems, share certain universal characteristics. CAMS is built on the expectation that societies are no exception.
This matters because complex adaptive systems behave like weather. They are always different, and yet they rhyme — returning to recognisable patterns in slightly different configurations. They do not follow linear trajectories. They phase-transition. They exhibit sensitivity to initial conditions. They spend long periods within attractors before sudden regime shifts. The five-day weather forecast is unreliable not because the atmosphere is random, but because small perturbations compound exponentially. Yet climate — the long-run behaviour within the attractor — remains modelable, because we are describing probabilities within constraints, not predicting individual outcomes.
Societies operate under analogous constraints. They go to total war on roughly generational timescales. They age through recognisable phases of youth, maturity, and decline. They are stamped by their founding myths and institutional origins in ways that persist across centuries. They undergo revolutions when stress accumulates beyond the capacity of the system to absorb it. These patterns are not accidental. They are signatures of the underlying thermodynamic logic that governs all complex adaptive systems — and CAMS is designed to make that logic legible.
Perhaps the most important intuition behind CAMS is metabolic rather than purely thermodynamic. A biological organism must maintain sufficient metabolic throughput to sustain its internal complexity. Deprive an animal of nutrition and its structural coherence degrades — organs fail, coordination collapses, the capacity to respond to environmental challenges diminishes. The analogy to a society's institutional architecture is direct: a meta-organism with nodes that correspond, functionally, to the organs of a living system.
The selection of eight nodes — Archive, Craft, Flow, Hands, Helm, Lore, Shield, and Stewards — emerged through a combination of theoretical constraint and empirical robustness. Attempts to include a ninth node, whether framed as a truth-seeking function, an environmental monitor, or a measure of international coupling, consistently failed to add explanatory value while significantly increasing computational complexity. The eight-node structure proved not merely convenient but load-bearing: it reflects convergent evidence from military organisational theory, primate social structure, and the architecture of other complex adaptive systems that eight functional roles represent a natural minimum for a self-sustaining, self-replicating social organism.
The four metrics — Coherence, Capacity, Stress, and Abstraction — were arrived at similarly. Coherence measures the degree of internal alignment and institutional trust. Capacity measures functional throughput and resource command. Stress indexes systemic strain as a negative integer: higher stress registers as a larger negative value, directly degrading node health in proportion to systemic load. Abstraction captures the gap between a system's self-narrative and its material reality — the degree to which its governing ideas have become decoupled from operational consequence. These four variables, applied across eight nodes, yield a 32-dimensional state space within which the trajectory of any society can be traced.
The resulting Grand System Metric — calculated as (Average Node Value × Average Bond Strength) / 10 — provides a scalar summary of overall systemic health, directly analogous to the thermodynamic concept of free energy available for useful work.
The following profiles illustrate CAMS applied blindly to five anonymised datasets spanning corporate, national, and indigenous timescales. The framework was asked to derive each entity's "memory" — the historical signature encoded in its thermodynamic trajectory — without prior knowledge of identity.
A 12-year entity that confused complexity for intelligence. The system constructed elaborate narrative structures (high Abstraction) atop collapsing coherence, producing the classic "Smartest Guys in the Room" failure mode. The 2000–2001 phase transition shows textbook node decoupling: a sudden dispersion spike followed by entropic heat death, with Stress at maximum negative and Capacity at minimum. The Archive was sophisticated; the Hands and Shield were hollow. Likely identity: WorldCom.
Where Entity A collapsed suddenly, Entity B bled slowly. Abstraction fell from near-maximum to zero over fourteen years while narrative confidence was maintained — a positive-feedback lie loop. A 2010 recovery was a false attractor: narrative resilience without thermodynamic recovery. The system ended in terminal coordination collapse. Likely identity: Theranos.
An entity operating at the thermodynamic limits of human organisation. A 2015 fragmentation event was successfully navigated, followed by sustained mythic ascent — high Abstraction maintained alongside exceptional Coordination, with low internal variance indicating elite execution. Strong Archive and Lore nodes suggest knowledge codification as a structural competitive advantage. Likely identity: SpaceX.
A 425-year arc of high stability, colonial impact trauma, and slow-burning recovery. Lore and Archive dominance signals deep mnemonic orientation: this is a system whose primary adaptive mechanism is the preservation and transmission of knowledge across generations. Unlike corporate entities with 12-year cycles, recovery here operates on century timescales. The system remembers what it lost and has not yet returned to pre-trauma coordination levels. Likely identity: Indigenous Australian societies.
Pre-contact stability shattered by invasion and land wars, followed by violent oscillation — a different trauma response to Entity D's sustained suppression. The post-1950 phase shows attempted integration of dual knowledge systems: Indigenous Lore and Colonial Archive held in unresolved tension. Likely identity: Māori / Aotearoa.
Several things are notable about these blind analyses. The framework successfully discriminated between rapid-cycle corporate fraud (12-year arcs) and long-cycle societal systems (400-year arcs), using the same node value equation throughout. It distinguished technical-engineering cultures, characterised by high Capacity and managed Stress, from narrative-deceptive cultures, characterised by high Abstraction and inverted Stress/Capacity ratios. And it identified two distinct colonial trauma response signatures — sustained suppression versus violent oscillation — that correspond to historically documented differences in how different peoples experienced and responded to colonial disruption.
The predictive horizon in the corporate cases is also notable. Entity A's entropy peak in 2000 preceded its 2001 collapse. Entity B's entropy peak in 2013 preceded its terminal state by five years. These are not post-hoc rationalisations; they are consequences of the model's thermodynamic logic applied consistently.
Most significantly, the cross-scale robustness of the node value equation — holding across 12-year corporate collapses and 400-year indigenous timelines — suggests that CAMS is capturing something real about the deep structure of complex adaptive systems, rather than merely fitting historical data after the fact.
This paper is part of the ongoing development of the Complex Adaptive Modelling System. Comments and correspondence are welcome.