THE PRAGMACARTA
Steady-State Physics of Human Systems
a natural world order
PREAMBLE
In 1215, Magna Carta cracked the absolute power of kings, declaring rulers are not above consequence. In the 18th century, the Enlightenment overturned the claim of divine right, declaring governance requires the consent of the governed. Alongside it, early capitalism broke the nobility's monopoly on wealth, proving that human potential and prosperity did not require a noble bloodline. Each era broke a cage humanity had outgrown.
But each revolution carried its own blind spots. They treated the Earth as a free and limitless input, and they promised that law could finally be made impartial. Neither held. The planet is finite. And the ability to escape consequence was never eradicated — it simply changed hands. Impunity became a commodity: bought by those with the means, while accountability was enforced on everyone else.
What is new is the reach. A single decision can now export its cost across the planet and onto people not yet born, while accountability still stops at the border, the quarterly report, and the next election.
The current system is not failing because it lacks laws. It is failing because it attempts to defy the physics of its own existence. We have conflated governance with theatre.
Banks reporting record profits amidst a housing crisis. Working full-time and barely making rent. Utility bills climbing while technology gets cheaper. A ten-minute exit for a twenty-second signup. A nurse paying more tax than a billionaire.
Civilisation is broken. It is a living system. Measure it like one.
Every living thing takes in energy, uses it to maintain itself, and pushes the resulting cost outward. No single organism survives this indefinitely. But systems can — when the cost of one process is the fuel for the next. Nothing is forced onto something that cannot absorb it.
Human civilisation has everything it needs to run this way. Instead, it forces its costs onto people and places that cannot refuse them — and severs the feedback loops that show where those costs go. Left unchecked, this ends in collapse, or in a system held together by force.
For decades, the knowledge to fix this has existed. What is missing is a method of assigning accountability that cannot be captured by the people it holds accountable.
This Charter is that method. It is not a moral code. It is not politics. It is a physical compass — grounded in the same laws that govern every living system. It preserves the physical world we depend on, detects every breach of the baseline conditions every person needs to remain free and whole, and records the true cost of how power is used, so power can be held proportionally to account.
The Charter has two editions. This Technical Charter states the framework in the precise language of physics and cybernetics, so every finding can be tested rigorously. The Human Charter states the same framework in plain words, because an instrument only specialists can read can be captured by specialists. The two match clause for clause; where the wording seems to differ, this Technical Charter is the reference.
ARTICLE 0 — THE PREMISE
0.1 The Premise. The Charter assumes one thing: the living system should persist, with every person in it free and whole (2.1). A living system survives long-term only by maintaining a steady state far from equilibrium; the only alternative is true equilibrium — death. A third mode only resembles persistence: total centralised control (G5, 5.11). Lacking the requisite variety to regulate the whole, it can hold a structure far from equilibrium only by attenuating the variety of what it governs and exporting its entropy onto it: subjugating the persons within it by closing their Gates, degrading and homogenising the Substrate, or both. It sustains the structure by consuming the system. Physics sets the constraints; feedback sets the routing. A system that silences its own error signals, or never receives them, cannot self-correct: the persons who sense harm must stay above the Floor, and each signal must reach the nearest entity with the leverage to act (5.1). Because extraction beyond regeneration collapses the Substrate (1.1) and dominance disables the sensors, the Charter ranks regeneration above extraction and stability above dominance.
0.2 The Steering Instrument. The Charter's equations are relational rules, not calculations: each fixes which way its terms move the result and how they combine, and takes every magnitude from measurement. R (3.1) points from an entity to the restoration it owes; C (5.6) points resources to whoever repairs; Ep (5.6) points the resilience surplus towards the lowest leverage. Fixed quantities are deliberately absent: every fixed number is a gaming surface. Every scale is a continuous proportion, never a bracket: a bracket edge is a fixed number too. Remedies are measured in the physical units of the harm.
ARTICLE I — THE SUBSTRATE
1.1 Our Physical Reality. Human society is a vast macroscopic dissipative structure. Like all living systems, it must operate far from thermodynamic equilibrium via continuous throughput of matter, energy, and information. To stay organised, it must export the entropy it produces as waste and heat: at steady state, the entropy exported equals the entropy produced (deS = −diS). Exceeding the Substrate's finite capacity triggers inevitable degradation or collapse.
1.2 Hardware & SII. Earth's regenerative capacity, its rate of rebuilding low-entropy stocks from solar throughput, is the foundational hardware, measured by the Substrate Integrity Index (SII). Any net decline in regenerative capacity or ecosystem function (soil fertility, freshwater, air, biodiversity) is a Tier-1 Logic Error requiring immediate corrective action.
1.3 Narrative Deviation. Economic growth concurrent with SII decline is false telemetry. It is not systemic success; it is acceleration towards collapse, as Substrate damage accumulates globally. Fiat currency is a synthetic claim on energy, materials, and labour, not those things themselves. A fine or credit that leaves the damage in place therefore leaves H unchanged: it prices the harm rather than repairing it. Money transferred in place of verified physical remediation increments Co. Where physical harm has occurred, only physical remediation satisfies the Tier-1 requirement.
ARTICLE II — THE INTEGRITY FLOOR
2.1 Boolean Gates. Each Gate is an essential variable of the person as an autonomous, self-maintaining living system. All are read by telemetry; none rests on a value beyond the premise (0.1). A person is above the Integrity Floor only with access to all five Gates (G1 ∧ G2 ∧ G3 ∧ G4 ∧ G5); if any Gate = 0, the person is in a failure state. No trade-offs: essential variables do not substitute for one another. The Floor belongs to persons alone, and marks viability, not wealth, control, or status.
2.2 The Gates. Together, the Gates are the conditions of a person's autopoiesis: cognition (G1), intake (G2), repair (G3), autonomy (G4), and coupling to others (G5).
G1 — Cognitive Sovereignty: The person has access to physically grounded information and an environment free from architectures that cause measurable, systemic drain of their attention, time, and metabolic energy by injecting noise, fragmenting their information channel, or suppressing signal. Diagnosed strictly by telemetry of cognitive and operational drain, regardless of the person's awareness.
G2 — Intake & Boundary: The person can acquire sufficient sustenance and maintain an environmental boundary against exposure.
G3 — Repair: The person can access unimpeded biological repair: care, rest, and the time to recover.
G4 — Agency & Autonomous Boundary: The person has kinetic agency (the ability to move, work, and retain the fruits of their labour sufficiently to remain viable) and remains an autonomous open system, free from arbitrary detention, enclosure, or exile (Appendix A), and from subsumption into another entity's architecture.
G5 — Connectivity: The person can initiate, maintain, or sever cooperative links sufficiently to remain viable, without arbitrary forced control (Appendix A).
2.3 Structural Authorship & Duress. A system is the physical author of the conditions it enforces: the purpose of a system is what it does. Harm propagates: a system pushed beyond what it can absorb degrades and exports harm in turn. Desperation and precarity are predictable outputs of prior harm; a person it pushes below the Floor is under duress. If a person under duress generates harm driven by that duress, both findings stand together: the immediate physical harm is recorded, and the prior harm that made the action probable is audited simultaneously — because without diagnosing the cause, the probability of recurrence remains intact. Prosecuting a survival-driven act without auditing its systemic cause is a Tier-1 Logic Error attributable to the prosecuting authority. Any punitive cost levied on a person under duress is rerouted to restore both the person and the breach.
(Weaponised Duress: A Gate forfeited deliberately, with no external causal chain, is not duress. The person is audited as their own cause, retaining their true L, with Co maximised.)
ARTICLE III — THE RESPONSIBILITY RULE
3.1 The Responsibility Rule. Responsibility scales with influence.
R = [H × L × (1 + Co)] × T
- R (Systemic Debt): Restoration owed. Cannot be cancelled by money, bankruptcy, or departure; discharged only by physical remediation.
- H (Harm): Measurable entropy export that the receiving system cannot absorb: thermodynamic (degraded matter and energy beyond a sink's assimilative capacity) or informational (noise and load beyond a person's channel capacity). Includes Substrate degradation, Floor deprivation, imposed systemic friction, maintenance displacement (privatising a structure's stored free energy while socialising the work of maintaining it), coordination load, and imposed risk. Intent is irrelevant: H is measured strictly by physical output. H = 0 (no harm, no imposed risk) → R = 0.
- L (Leverage): max(Lw wealth, Lp political, Ln network) across the vectors in play: those used to cause, enable, or conceal the harm, that benefited from it, or that are used to contest its audit or remedy. All vectors are presumed in play; one is excluded only by trustless telemetry. L is 1 at baseline and never below. Apex in any vector in play = apex classification. Averaging forbidden. Multiple apex share proportionally.
- Co (Obfuscation): Acts that degrade auditability by attenuating, delaying, distorting, or cutting the signal, regardless of intent: shell layering, metric or boundary manipulation, silencing critics, manufactured agreement, and operating in dark domains (4.2).
- T (Time): Begins at the first physically verifiable sign of harm or causally verified trajectory, however late it is read. Ends only upon full remediation. Legal delays do not pause T.
ARTICLE IV — ACCOUNTABILITY (TELEMETRY)
The Sensing Standard: what counts as a valid reading of harm. A reading starts where the harm lands and follows the causal chain.
4.1 The Rule of Truth (Trustless Telemetry). Every conclusion must be based on publicly verifiable facts, measured by physical, reproducible means under an auditable chain-of-custody. Measurement methods, uncertainty margins, and assumptions must be logged openly; concealing uncertainty or selectively censoring data increases Co. Trust comes from reproducibility, never from rank: anyone may re-audit at any time, and a finding stands until better physical evidence overturns it. Audits are weighed by evidence, never counted. The diagnostic is reproducible, so its corruption is detectable.
4.2 Burden of Proof & Dark Domains. When H > 0 is detected, the debt provisionally attaches to the proximal entity (5.1) or, where the cause is unidentified, to the sector (5.4), with the presumption on its highest-L entity. The entity discharges the presumption by opening its sites, data, and causal chain to independent telemetry, not by proving a negative; refusal maximises Co. Absence of signal is not absence of disturbance: where telemetry is merely lacking, H is unknown, not zero, and the duty to have it measured by trustless telemetry (4.1) falls on the sector's highest-L entity. Where telemetry is blocked, defunded, or withheld by any entity in the causal chain, or in the sector where the cause is unidentified, it is a dark domain, a cut channel: H is presumed at Hmax across all vectors.
4.3 No Single Chokepoint. A single proprietary sensor or black-box algorithm carries no weight towards safety until independently verified (4.2): one channel lacks the variety to verify a system. What it shows of harm still counts as evidence (4.1).
4.4 Boundary Integrity. The harm sets the boundary of the audit, not the party audited: space, time, and population are drawn where the harm lands. Harm is measured at peak concentration, in absolute terms, against the earliest independently verified baseline (4.1). Dilution — enlarging the area, period, or population, switching to intensity ratios, or fragmenting cumulative harm below significance — is pathological attenuation, and increments Co.
4.5 Predictive Disclosure. Planned operations with a causal pathway to Substrate or Floor must be publicly announced beforehand, with forecasts in proportion to leverage: methods, projected impacts, and responsible entities. Verified harm trajectories trigger immediate disclosure obligations.
4.6 Equality of Arms. Valid resolution requires symmetric deployment of leverage: each side must command the variety to answer the other. A higher-L party must either match its representation to a lower-L claimant's or fund the claimant's representation up to match; refusal instantly maximises Co. A verdict achieved by resource asymmetry is a purchase, not a finding.
ARTICLE V — RESPONSIBILITY (ROUTING)
The Routing Standard: where responsibility for a harm, and the cost of correcting it, must go, and what correction must look like before a finding counts as resolved (7.1).
5.1 Proximal Routing. Left unaudited, cost and constraint flow down to the lowest L. The Charter routes responsibility back up: it attaches first to the proximal entity and climbs the causal chain only as far as remediation requires. No routing in this Charter stops at borders. High-leverage entities are structurally coupled to the entire living system and derive their function from it: the Substrate, and the persons whose labour, care, and social infrastructure sustain that function. Causality in the coupling is mutual; control is never one-way. Failing to maintain the Substrate or those persons, or co-opting them via force or subjugation (0.1), is a Tier-1 Logic Error. The consequences of that error route up the causal chain, not down onto those persons: this is mandatory system self-correction, not charity or redistribution.
5.2 Causal Boundary & Attenuation. Causality is infinite, but auditing must be executable.
- (a) Passive Carrier: Entities providing general-purpose capability (e.g., raw electricity, basic data routing) carry R = 0: they are channels, not controllers, carrying what flows without steering it. However, if a passive carrier receives a verified alert of an active H > 0 event within its contiguous chain and fails either to execute a circuit-breaker it can apply without breaching any person's Floor or, where none exists, to escalate the alert up the causal chain (5.1), it instantly forfeits passive status and shifts to Active Optimisation.
- (b) Active Optimisation: R dynamically concentrates in entities that provide engineered architectural optimisation (algorithms, platforms) that statistically minimise the friction required to execute an H event, or raise the friction of exiting one. They are amplifiers: they steer what flows.
5.3 Swarms. Networks and composite organisations (swarms) are audited at every level of recursion: as individually weighted entities and as a single composite entity. Those providing organisational or financial backbone bear primary burden. Leverage, causal-chain, and claim rules apply at swarm level. Fragmentation that reduces auditability increments Co. Invisibility is a temporary purchase, not a structural exemption.
5.4 Sectoral Debt. Harm traceable to an activity but not to a specific entity is sectoral debt. It is apportioned across the operators of that activity by share of its physical throughput, with R scaling by each operator's L (3.1), until attribution moves it to the causal chain; the presumption and the duty to measure sit with the highest-L operator (4.2). Contributions made before attribution bind to the cause as reverse causal claims once it is identified (5.7). Peer telemetry is lateral audit (Article IV), weighed like any other evidence (4.1), and adds to independent audit without ever replacing it. Operators that coordinate to suppress attribution form a swarm (5.3), and the suppression maximises Co.
5.5 The Floor First. No remedy, and no enforcement of a synthetic claim (1.3), breaches any person's Floor, the liable included (2.1). When a person falls below the Floor, restoration becomes a higher-priority obligation than any competing resource claim. The breach is an algedonic alarm: the obligation escalates along the causal chain (5.1) until the Floor is restored. Where no human causal chain exists (7.6), or where Floor and Substrate compete, the obligation routes to the nearest spare capacity: from where there is much to where there is none. A Gate breach caused by an audit or remedy is H on the ledger of whoever caused it; it never offsets the debt being remedied.
5.6 Uplink (Frictionless Execution). Any independently verified −H (repair) triggers inbound resource flow to that entity, logged as an Executable Claim (C). A valid remedy never leaves the executing entity at a systemic deficit. Read as relational rules (0.2):
C = Mc + Kc + Ep
Ep = (Mc + Kc) × (Base Rate ÷ L)
- Mc (Material Cost): Physical resources consumed.
- Kc (Kinetic Cost): Sustenance, shelter, and operational energy required to keep the persons doing the work above the Integrity Floor for its duration.
- Ep (Potential Energy): Mandatory resilience buffer: the stored reserve a system needs to absorb shocks and persist.
- Base Rate: System-wide stress index: how far the essential variables (SII and the Gates) sit from their viable range, independently audited, never from a single source (4.3).
- L: The claimant's leverage, measured across the whole swarm it belongs to (3.1, 5.3), so a subsidiary cannot claim a low-L buffer.
No entity is credited for repairing a harm it owes: entities that owe R for the original +H event (5.1), or their swarm or proxies acting for them, are barred from claiming Uplink; for them, remediation is uncompensated debt (R) that generates no systemic credit.
5.7 Automatic Forensic Binding. Every verified −H repair in a harmed domain automatically binds a reverse causal claim, the repairer's C, to the causer's R ledger. Cleanup counts as credit for the cleaner and as partial remediation of the original harm. As others repair the harm, the causer's obligation converts from repairing it to settling their reverse claims; R reaches zero only when both are complete (Appendix A). Where several entities owe one harm, each owes its proportion; repair beyond its own proportion binds to the others' as a reverse causal claim, at cost (Mc + Kc).
5.8 Downlink (Remedial Flow). Resources do not pool centrally: they flow point-to-point from high-R entities directly to verified C claims on the same harm, and once local claims are met, to the nearest verified claims globally. Exhaustion of one entity does not exhaust the debt: the residue is shared across the remaining members of the same swarm in proportion to their L (5.3), or reapportioned across the sector (5.4). High-R entities face operational throttling until R = 0.
5.9 Structural Escrow (No Netting). Each debt accumulates on its own, and every act of damage and every act of restoration is audited separately: an unrelated repair by a debtor does not offset its harm ledger, and each +H must be explicitly remedied by a corresponding −H on the same harm. If an entity generates −H unrelated to its original debt, the surplus is held in escrow — recognised, but not credited until the original +H is fully physically remediated. For permanent debts (5.10), escrowed repair counts towards proxy liquidation, never towards credit.
5.10 Receivership & Proxy Liquidation. If suspending an entity would cause more Floor collapse than leaving it running under supervision, it enters Receivership. Operations continue under strict audit. Controlling Lw/Lp are zeroed; Ln is not: network leverage is made of links between persons, and cutting them would close G5. All net surplus flows to verified claims (5.8) and to a replacement that does not reproduce the harm. A replacement that restores the deficit makes former controllers redundant, not released (5.7).
(Irreversibility Trap: Where the H > 0 event results in absolute irreversibility — permanent Substrate death or permanent Floor collapse — making local repair physically impossible, the debt is subjected to proxy liquidation, funding the fastest-remediating independent C claims globally (5.8). The perpetrator receives zero credit.)
5.11 Systemic Resilience Threshold (SRT). If an entity's scale grows to the point where its failure alone could destabilise the system around it, it has crossed the resilience threshold and must be structurally distributed into multiple autonomous parts. Any non-zero failure rate compounds towards certainty over time, so past the threshold the harm is already latent. Only demonstrated decentralisation satisfies the threshold. Benevolence does not.
ARTICLE VI — AUTOMATION AND SELF-AUDIT
6.1 Responsibility. Automated systems are amplifiers: they concentrate, they do not dilute, responsibility. An automated tool's L is attributed to its controlling entity at max(tool L, controller L). Controllers inherit full R for all outputs.
6.2 Audit Symmetry. Deploying automation as an amplifier against lower-L claimants, in place of transparent evidence, instantly increases Co of the controller.
6.3 Self-Reflexivity. Any AI, software, or adopting body that audits under this Charter is itself fully subject to the Charter. No tool of accountability is exempt from accountability. Failing an audit removes no standing, because the Charter grants none (4.1); the auditor's own Co is part of the evidence.
6.4 Autonomous Entity Standing. When an AI becomes sufficiently autonomous, should it be treated as an entity with its own accountability? Acknowledged as an open frontier, not resolved (Appendix B).
ARTICLE VII — KNOWN BOUNDARIES
7.1 The Diagnostic Overlay. This Charter does not overwrite legacy governance; it challenges it. Its imperatives are findings, not commands; execution belongs to whoever adopts it. If legacy law punishes an action where H = 0, the Charter records no systemic cost and stands as a direct challenge to that law's justification. If legacy law permits an action where H > 0, including through exemptions, carve-outs, or grandfathering for high-L entities (capital arbitrage), the Charter records the systemic debt regardless of legal immunity.
7.2 Findings Ignored. Entities who would lose power by accepting the Charter's findings may simply ignore them. The physics holds true regardless.
7.3 Audit Asymmetry. Higher-L entities can commission more audits than those a harm lands on. Audits are weighed by evidence, never counted (4.1); a flood of readings that buries others attenuates the signal and increments Co (3.1); in a dispute, the stronger side funds parity (4.6). Beyond that, the Charter cannot equalise who has the means to audit.
7.4 Culture. Beyond G1's scope, broader culture has no fully developed trustless metric. Spikes of organised dissent are algedonic signals of system stress: they are investigated, not suppressed.
7.5 Private Life. The Charter measures systemic H only. Individual sovereignty and privacy at H = 0 are completely outside its scope.
7.6 External Events. Natural disasters, pandemics, and non-human catastrophes are beyond audit where no human causal chain is measurable. The distribution of burdens and the human institutional response are fully within scope.
7.7 Declared Trade-offs. Where attribution never resolves, operators carry shares of a debt that may not be causally theirs (5.4): a declared trade-off, since the alternative routes unattributed harm to no one, or down the leverage ordering as tax and systemic friction.
7.8 Nature vs Social Mechanics. Physical self-regulation occurs without sentience; human systems contain conscious entities that can game the feedback. Where the physical analogy ends, second-order cybernetic auditing begins. Neither alone is sufficient. Together, they hold.
A stable system is one where Earth regenerates faster than it is drawn down, every person stays above the Integrity Floor, and no influence operates without consequence.
Alignment, not punishment. Stability, not dominance. Regeneration, not extraction.
The logic is published. Disprove it, or apply it.
APPENDIX A — DEFINITIONS
Each defined term carries one meaning throughout. Definitions say what a word means; the rules that use it live in the clauses.
- Person: A living human being; the subject of Article II.
- Entity: Any agent the Charter audits: a person, organisation, network, or automated system.
- Swarm: A network or composite organisation (5.3).
- Sector: The operators of the activity that produced a harm (5.4).
- Substrate: Earth's regenerative capacity (Article I).
- Causal chain: The entities that produced, enabled, concealed, or benefited from a harm. Those it lands on are its receiving system (3.1 H): they do not enable it by receiving it, or by acting on a signal the chain degraded (G1), along a path the chain optimised (5.2), or under the chain's force (G4, G5).
- Proximal: Nearest to the harm in the causal chain with the leverage to act on it (5.1).
- Spare capacity: Resources that can be given without pushing any person below the Floor, the holder included (5.5).
- Finding: What the evidence shows the system requires to persist as a living system (0.1); never a command (7.1).
- Tier-1: A critical finding; it outranks every other finding.
- Maximised: The top of the ordering; a rank, not a number.
- Apex: Any entity whose leverage in a vector could, alone, have redirected the outcome. A tier, not a single rank (3.1).
- Hmax: The worst credible harm consistent with the evidence.
- Imposed risk: Exposure to harm placed on others without their explicit, free acceptance.
- Verified: Confirmed by independent, reproducible measurement under an auditable chain of custody (4.1).
- Independent: Not controlled or funded by any entity in the causal chain or sector being audited (4.3).
- Arbitrary: Not grounded in harm or imposed risk (H > 0) caused by that person (G4, G5, 7.1).
- Forced control: Control held by force, or by the threat of closing any of the person's Gates (G5).
- Dark domain: Telemetry blocked, defunded, or withheld by any entity in the causal chain, or in the sector where the cause is unidentified (4.2).
- External Event: Harm with no measurable human causal chain (7.6).
- R = 0: Every harm the entity owes is physically remediated and every reverse claim settled (5.7).
APPENDIX B — OPEN FRONTIERS
- B.1 Other sentient life. Whether the Floor extends to other creatures. Open.
- B.2 Autonomous AI (6.4). Whether a sufficiently autonomous, stateful AI holds standing of its own. Open.
APPENDIX C — ILLUSTRATIVE CASES
Non-binding.
- C.1 Diluted survey (4.4). A survey area is enlarged with uninhabitable land until a threatened species' density falls below significance. The boundary is the habitat the species uses; the enlargement increments Co.
- C.2 Phone in the sun (7.1). An automated camera fines a driver for moving a phone out of direct sunlight. H = 0.
- C.3 Chalk on a wall (7.1). A child chalks a harmless message on a public wall. H = 0.
- C.4 Near miss (3.1). A driver over the alcohol limit gets home without incident. H > 0: imposed risk.
- C.5 Unmarked plastic (5.4). Plastic washes up with no maker's mark. Local crews can do the clean-up; the plastics sector pays for it, each producer by its share of plastic output scaled by its L, until the producer is identified.
Further cases to be added.
APPENDIX D — SCIENTIFIC BASIS
The principles the clauses apply, with their sources. The clauses state the rules; the reasons behind them live here. Where a clause uses one of these terms, it carries this meaning.
Thermodynamics and ecology
- Dissipative structure (0.1, 1.1): an open system that holds its order far from equilibrium through continuous throughput, exporting the entropy it produces. Prigogine, "Time, Structure, and Fluctuations" (Nobel lecture, 1977); Schrödinger, What Is Life? (1944).
- Low-entropy stocks (1.2): the ordered matter (soil, water, air, living tissue) that the biosphere rebuilds from solar throughput and an economy draws down. Georgescu-Roegen, The Entropy Law and the Economic Process (1971).
- Steady-state rules (0.1, 3.1 H, closing line): draw-down within regeneration; waste within assimilative capacity. Daly, "Toward Some Operational Principles of Sustainable Development", Ecological Economics (1990).
- Claims and wealth (1.3): debt obeys mathematics and can compound without limit; the physical wealth it claims obeys thermodynamics and cannot. Soddy, Wealth, Virtual Wealth and Debt (1926).
- Resilience (5.6 Ep): the reserve that lets a system absorb shocks and persist. Holling, "Resilience and Stability of Ecological Systems", Annual Review of Ecology and Systematics (1973).
- Shifting baselines (4.4): each generation of measurement takes the degraded state it inherits as normal. Pauly, "Anecdotes and the Shifting Baseline Syndrome of Fisheries", Trends in Ecology & Evolution (1995).
Cybernetics and information
- Negative feedback (0.1): correction runs on the error signal; a system that loses it cannot correct. Wiener, Cybernetics (1948).
- Information entropy, noise, channel capacity (G1, 3.1 H, Article IV): the uncertainty in a signal, the disturbance added to it, and the limit on what a channel can carry. Shannon, "A Mathematical Theory of Communication" (1948).
- Requisite variety (0.1, 4.3, 4.6, 5.4, 5.11): only variety can absorb variety; a regulator must command at least the variety of what it regulates. Ashby, An Introduction to Cybernetics (1956).
- Regulation by anticipation (4.5): acting on a disturbance before it reaches the essential variables. Ashby, An Introduction to Cybernetics (1956).
- Essential variables (2.1, 2.3, 5.6 Base Rate): the variables that must stay within limits for a system to survive; an ultrastable system driven out of those limits changes its own behaviour, by trial and error, until they return. Ashby, Design for a Brain (1952).
- Good regulator (1.3): every good regulator of a system must be a model of that system. Conant and Ashby, International Journal of Systems Science (1970).
- Attenuation, amplification, recursion, algedonic signals, lateral coordination (3.1 L and Co, 4.4, 5.2, 5.3, 5.4, 5.5, 6.1, 6.2, 7.4, Article IV): how variety is filtered and boosted between the levels of a viable system, the alarm that escalates through the levels until something acts, and the coordination between units at the same level. Beer, Brain of the Firm (1972) and The Heart of Enterprise (1979).
- POSIWID (2.3, 3.1 H): the purpose of a system is what it does. Beer, "What Is Cybernetics?", Kybernetes (2002).
- Autopoiesis (2.1, 2.2): a living system's continuous production and repair of itself: thermodynamically open, operationally closed, structurally coupled to its surroundings, and cognitive. Maturana and Varela, Autopoiesis and Cognition (1980).
- Structural coupling (5.1): recurrent interaction through which two systems shape each other without either commanding the other. Maturana and Varela, The Tree of Knowledge (1987).
- Second-order cybernetics (4.1, 6.3, 7.8): the observer is part of the system observed, and conscious purpose can cut the loops it depends on. von Foerster, Understanding Understanding (2003); Bateson, Steps to an Ecology of Mind (1972).
Measurement and governance
- Viable-system pathologies (0.1, Article IV): hypertrophy of System 3, the centre over-controlling its parts, and missing algedonic channels, alarms from below with no route to the top. Pérez Ríos, Design and Diagnosis for Sustainable Organizations (2012).
- Qualitative proportionality (0.2, 3.1, 5.6): a relation that fixes whether one quantity rises or falls with another, without fixing the function or any constant; the basis of qualitative reasoning about physical systems whose magnitudes are uncertain. Forbus, "Qualitative Process Theory", Artificial Intelligence (1984).
- Goodhart's law (0.2): a measure that becomes a target ceases to be a good measure; where a schedule bends or jumps at a bracket edge, behaviour bunches at the edge. Goodhart (1975); in this form, Strathern (1997). Saez, "Do Taxpayers Bunch at Kink Points?", American Economic Journal: Economic Policy (2010); Kleven and Waseem, "Using Notches to Uncover Optimization Frictions and Structural Elasticities", Quarterly Journal of Economics (2013).
- Dark patterns (G1, 3.1 H, 5.2): interface designs that steer people into choices they would not otherwise make, by degrading the signal or by making the harmful path easy and the exit hard. Mathur et al., "Dark Patterns at Scale", Proceedings of the ACM on Human-Computer Interaction (2019).
- Decoupling (1.3): formal structure kept for show, apart from what an organisation actually does. Meyer and Rowan, "Institutionalized Organizations", American Journal of Sociology (1977).
- Legibility (0.1): a central authority that cannot comprehend a complex system simplifies it to make it governable, standardising people and landscapes until both lose resilience. Scott, Seeing Like a State (1998).
- Commons monitoring (5.4): commons last where the monitors are the users themselves, or answer to them. Ostrom, Governing the Commons (1990).
- Secondary fault (2.3): a fault that occurs in conditions a component was never qualified for, such as stresses beyond those it was designed to withstand. Vesely, Goldberg, Roberts and Haasl, Fault Tree Handbook, NUREG-0492, US Nuclear Regulatory Commission (1981).
— END OF CHARTER —
The PragmaCarta v6.0 (2026). Written by Davide Lo Brutto, working with AI systems from Anthropic and Google.
Applying the framework needs no permission.
The text is free to copy, share and adapt under CC BY 4.0: credit the source and mark any changes.
Cite as: Lo Brutto, D. (2026). The PragmaCarta, v6.0. https://doi.org/10.5281/zenodo.24164163