The Science of Wholeness

Volume II — Foundations: The Paradigm and the Science

Nine movements, four borrowings.

THE PLATE

A watercolour of rolling farmland in ochre and green, the hills repeating into the distance like a wave. — Plate II.04The Standing Wave.The pattern is not in the water and it is not in the flame. It is in the difference between them, and it will hold exactly as long as the difference does.

THE LETTER

There is a sentence that appears in almost every book on regenerative economics, including books this house admires: the economy is a living system. It is usually offered as though it settled something.

This chapter is about what that sentence is allowed to mean.

Four bodies of work sit behind it, and they are not literature — they are science, with theorems, thresholds, definitions and published limits. Ludwig von Bertalanffy showed what an open system is and what follows from openness. Ilya Prigogine showed how order is paid for, and won a Nobel Prize for the mechanism. Humberto Maturana and Francisco Varela wrote a definition of the living so exact that it can be failed, and specified precisely what fails it. Donella Meadows ranked twelve places to intervene in a system and put the ranking in an order she defended.

Each of those four licenses an economist to say some things and not others, and the boundary is sharp in every case. It is also almost never drawn. So an enormous amount of good instinct in this field arrives carrying two kinds of cargo at once — claims that would survive a physicist reading them, and claims that would not — and because they travel in the same sentence, a sceptic who finds one loose discards both.

This chapter is the one place in the edition where you can find out which is which. You will be handed the actual definitions, the actual thresholds and the actual numbers, and you will be handed a test you can run on any sentence of systems language in ninety seconds. Some of what you already believe will come through licensed and stronger for the scrutiny. One thing that is said constantly will not survive, and we will name it, show the criterion it fails, and show that the men who wrote the criterion said so themselves.

The reward for this is not rigour for its own sake. It is that a claim which has been checked can be taken into a room where the other people have physics.

— The Editors

DISCOVERY

Where the borrowing has already been done properly

Begin with the good news, because it is substantial and it is specific: in every one of these four traditions, an economist has already done the transfer rigorously, and the rigorous version is published, citable and better than the loose one.

Kenneth Boulding, 1956. The year after he helped Bertalanffy found the Society for General Systems Research, Boulding published "General Systems Theory — The Skeleton of Science" in Management Science. He set out nine levels of system complexity, from static frameworks through clockworks, thermostats, cells, plants, animals, humans and social organisations up to transcendental systems. Then he did the thing that makes it science: he placed his own discipline on the ladder and said it was in the wrong place. Economics, he wrote, has theoretical models at about the level of the clockwork and the thermostat, while its subject matter is at the level of the social organisation. He used general systems theory to bound a claim rather than to inflate one, and that single move is the whole methodology of this chapter.

Herbert Simon, 1962. "The Architecture of Complexity" gives the parable of the two watchmakers, Hora and Tempus, and from it a genuine result: near-decomposable hierarchic systems assemble far faster than flat ones under interruption, because progress is saved at each stable sub-assembly. This is not an analogy. It is a statement about the expected time to completion of an assembly process, it is derivable, and it directly licenses a claim an economist wants to make about modular organisations and federated structures.

Brian Arthur, 1989. "Competing Technologies, Increasing Returns, and Lock-In by Historical Events", in the Economic Journal, formalises path dependence as a generalised Pólya urn and proves convergence to one of several possible absorbing states. Far-from-equilibrium economics stopped being a mood the day this was published, because the model has theorems and the theorems have conditions. Prigogine's bifurcation became an economist's limit theorem, and it became so by being written down.

Gunther Teubner, 1993. Law as an Autopoietic System takes Maturana and Varela's definition into the social sciences the only way it can honestly go: by naming the components. In Teubner's account, following Niklas Luhmann's Soziale Systeme (1984), the components of the legal system are legal communications, not lawyers — and a legal communication is produced only by other legal communications, which is precisely what the definition requires. Whether one accepts the extension is a live question. What is not in question is that Teubner and Luhmann paid the price of admission: they said what the components were, and they accepted the consequence that people are then the system's environment rather than its parts.

Elinor Ostrom, from 1990. The Institutional Analysis and Development framework is a systems analysis conducted entirely at the level of rules-in-use — who may do what, under what conditions, with what payoff, verified how. It produces eight design principles for enduring common-pool resource institutions and, crucially, the conditions under which they fail. Read against Meadows, the whole of Ostrom's empirical programme sits at leverage points five and six, the rules and the information flows, and it is the most successful systems intervention in the history of economics.

Donella Meadows and Jay Forrester, from 1969. Urban Dynamics, then Limits to Growth in 1972, then the leverage-points essay in 1999. The through-line is that stocks, flows, delays and feedback loops are measurable objects, and a model built from them makes predictions that can be checked against a later world. Some of those predictions held and some did not, and the tradition has published both — which is the behaviour of a science.

Six borrowings, six disciplines, one pattern: in every case the transfer became rigorous at the moment somebody specified what the terms referred to in the new domain. Not when they found a better analogy. When they named the components, stated the threshold, or wrote the equation down. That is the whole technique, and the next movement is what it costs.

THE ARITHMETIC

What each source licenses, with the numbers

Bertalanffy: openness, and the steady state that is not an equilibrium.

Bertalanffy's central object is the Fließgleichgewicht — the flow equilibrium, or steady state. A closed system runs down to thermodynamic equilibrium, where nothing further happens. An open system settles into a state where something is happening constantly and the composition stays put. These are not the same condition and conflating them is the commonest error in the whole field.

His own growth law makes the point in one line:

   dW/dt  =  η·W^(2/3)  −  κ·W          anabolism minus catabolism
   W∞     =  (η/κ)³

With η = 0.5 and κ = 0.1 per period, the asymptote is 125.0 units, and it is reached from anywhere: starting at 1.0 gives 124.986 after 300 periods, starting at 40.0 gives 124.995, starting at 300.0 gives 125.006. The relaxation time constant is τ = 3/κ = 30 periods, so 300 is ten time constants and the convergence is real rather than asserted.

The ceiling is a ratio of the system's own two coefficients. Nothing outside imposed it. That is Bertalanffy's actual claim, and it licenses an economist to say something quite strong: the size a firm, a city or an industry settles at is often set by the ratio of its build rate to its maintenance rate, not by a market limit — which means it can be moved by changing either coefficient, and by nothing else.

The property that follows is equifinality: an open system reaches the same final state from different initial conditions and by different routes. Take a two-variable open system:

   dx/dt = −0.5x + 0.2y + 10
   dy/dt =  0.3x − 0.6y + 4      →   x* = (28.3333, 20.8333)

Started at (0, 0), at (100, 5) and at (3, 90), it arrives at (28.3333, 20.8333) every time. Now run the control — the same relaxation with the source terms removed, so the system is closed and conservative:

   from (0, 40)    →  (20.0000, 20.0000)     total  40.0 conserved
   from (100, 5)   →  (52.5000, 52.5000)     total 105.0 conserved
   from (3, 90)    →  (46.5000, 46.5000)     total  93.0 conserved

Same relaxation law, three different destinations, because a closed system remembers its initial total. Equifinality is a property of openness. It is not a property of being complicated, and it is not a property of being alive.

What this licenses: the end state of an open economic system need not carry the memory of its starting conditions. What it does not license: any claim that the end state is chosen, or that all end states are reachable. The steady state is still determined, rigidly, by the coefficients. Equifinality says your history does not bind you. It does not say your parameters do not.

Prigogine: order is paid for, and the bill is continuous.

The balance is exact and it is the whole of the matter:

   dS  =  d_eS  +  d_iS          with  d_iS ≥ 0 always

Internal entropy production is never negative — that is the second law. But the exchange term with the surroundings can be negative, and a structure holds its order by making it negative enough to cancel the production. A body dissipating 100 W into a 293 K sink exports 0.3413 W/K, which is 29.49 kJ/K per day, and at steady state its internal production is exactly the same figure. Halve the export and 0.1706 W/K of entropy begins accumulating inside, every second.

Now the number that ought to change the conversation.

   intercepted solar   1361 W/m² × π(6.371×10⁶ m)²    = 173,549 TW
   absorbed            × (1 − 0.294)                  = 122,526 TW
   entropy in          (4/3)·absorbed / 5772 K        = 2.8304×10¹³ W/K
   entropy out         (4/3)·absorbed / 255 K         = 6.4066×10¹⁴ W/K
   planetary export    out − in                       = 6.1235×10¹⁴ W/K

   economy             19 TW degraded to heat at 288 K = 6.5972×10¹⁰ W/K

   planetary export / economic production             = 9,282×

(The factor of 4/3 is the entropy of blackbody radiation, from Planck. Using F/T instead understates the export by a third, and it is the commonest slip in this literature.)

The planet exports entropy roughly nine thousand times faster than the entire human economy produces it. The sink is not the binding constraint, and it is not close. A city of one million people, at the global mean of 2,375.0 W per person, runs a metabolism of 2.375 GW — 2.79 percent of the 85.0 GW of sunlight falling on its own 500 km² of ground.

This is the chapter's cut, so take it slowly. The thermodynamic argument is the one this field leans on hardest and it is the one the arithmetic will not carry. Not because thermodynamics is wrong — it is the most secure knowledge we have — but because it is not the scarce thing. What binds is specific: a particular material cycle, a particular gradient, a particular rate. A claim of the form thermodynamics forbids this is almost always doing no work, and a sceptic with a calculator will find that out in four minutes and then discount everything else in the paragraph.

Two further limits belong here, both named and both load-bearing. First, the Glansdorff–Prigogine theorem of minimum entropy production holds only in the linear regime near equilibrium. The attempt to extend it into the far-from-equilibrium domain as a general evolution criterion does not yield a variational principle, and Prigogine's own later work says so. Second, the maximum entropy production principle is a conjecture, not a theorem — the derivations are contested and the review literature is explicit about it. So the sentence systems self-organise to maximise entropy production is not available. It sounds like physics. It is a hypothesis wearing the clothes.

And one further correction, offered as the courtesy it is: Nicholas Georgescu-Roegen's proposed fourth law, that matter degrades irrevocably so that complete recycling is impossible in principle, was shown by Bianciardi, Tiezzi and Ulgiati in 1993 to be false as stated. Complete recycling is thermodynamically permitted at finite, large energy cost. The practical constraint is entirely real; the law is not. The right claim is the expensive one, and it is stronger for being true.

The threshold that everyone skips. Throughput is necessary for self-organisation and it is not sufficient. The Rayleigh–Bénard system gives the exact condition — cells appear only above a critical Rayleigh number of 1,708:

   layer   1.0 mm   →   ΔT_crit = 120.7587 K
   layer   5.0 mm   →   ΔT_crit =   0.9661 K
   layer  10.0 mm   →   ΔT_crit =   0.1208 K
   layer  50.0 mm   →   ΔT_crit =   0.0010 K

The threshold falls as the cube of depth — a tenfold deeper layer needs 1,000× less gradient. Below the threshold the same energy crosses the layer and organises precisely nothing. This is the condition the living-systems approach needs and usually does not check. If the gradient across your system is subcritical, enabling self-organisation will produce nothing at all, and the honest move is to build the structure deliberately instead. The design lever the physics actually hands you is the depth term, because it enters cubed: make the system deeper — longer horizons, more layers of intermediation, more slack — and the gradient you need collapses.

Maturana and Varela: a definition sharp enough to fail.

Varela, Maturana and Uribe gave a six-point key in 1974. In their order: a boundary determinable by an observer; enumerable constitutive components; a mechanistic system, where relations follow from component properties; boundary components produced by the network itself; those boundary components produced by interactions of components themselves produced by the network; and every other component either produced by the network or permanently imported into that production.

The key is a conjunction. Varela and colleagues are explicit: fail any one point and the system is not autopoietic. Score four candidates honestly:

   a bacterial cell                               6/6  = 100.0 %   autopoietic
   a legal system (components: legal comms)       6/6  = 100.0 %   autopoietic
   a limited company                              3/6  =  50.0 %   NOT autopoietic
   a market                                       2/6  =  33.3 %   NOT autopoietic

Here is the honest negative, and it is the most-repeated sentence in this field: a firm is not an autopoietic system. It fails at criterion four. A company's boundary is produced by a registrar acting under company law — by an external legal process, not by the company's own components. It fails at six as well: people are not produced by the firm, they are recruited intact and leave intact. And by Maturana and Varela's own taxonomy a machine that produces something other than itself is allopoietic, which is exactly what a firm producing carpet or software is. They supplied the word. It is in the same book.

Maturana himself declined the extension to human social systems, treating them instead as third-order structural couplings between autopoietic individuals. One may disagree with him. One may not do so silently while citing him.

The distance between 50.0 percent and nought is the entire misuse. Partial credit on a conjunctive definition is not partial truth; it is a category error, and it is what makes the borrowing decoration rather than argument.

What survives, and it is a great deal, is structural coupling: the medium perturbs, the structure determines. That is not soft, and it has a number. Put an identical ten percent demand step into two firms with the same natural frequency and different damping:

   firm A, ζ = 0.9   →   peak overshoot   0.152 %
   firm B, ζ = 0.2   →   peak overshoot  52.662 %
                          ratio  345.6 ×

Same shock, responses differing by a factor of 346, and every bit of that difference is internal. No instructive interaction — the phrase that reads as philosophy — is a claim about where explanatory weight sits, and it is right.

Meadows: the ordering, and what sets its size.

Her twelve, in her order, weakest to strongest: constants and parameters; buffer sizes relative to flows; the structure of stocks and flows; the length of delays; the strength of balancing feedback; the gain around reinforcing feedback; the structure of information flows; the rules; the power to self-organise; the goals; the paradigm; the power to transcend paradigms. The deep points, one through four, are 4 of 12 — 33.3 percent of the list, and they are the ones nobody costs.

The ordering is defensible from a two-line model. A stock filled by goal-seeking inflow a·(G − S) and drained at k·S settles at S* = aG/(a+k). With a = 0.3, k = 0.1, G = 100, that is 75.00.

   +20 % on the parameter a   →   S* = 78.2609    (+4.35 %)
   +20 % on the goal G        →   S* = 90.0000    (+20.00 %)

   elasticity to G  =  1.000 exactly
   elasticity to a  =  k/(a+k)  =  0.250
   ratio            =  4.00×

Meadows' ordering falls out of two lines of algebra. The goal enters with elasticity exactly one; the parameter enters with elasticity k/(a+k), which is less than one whenever the filling loop is strong.

And here is where it fails. Set a = 0.1 and k = 0.9, so the draining loop dominates:

   elasticity to G = 1.000 ;  elasticity to a = 0.900 ;  ratio = 1.111×

The advantage of the goal over the parameter collapses from 4.00× to 1.11×. Meadows' list is an ordering of kinds, and its magnitude is set by loop dominance in the particular system. She knew this and wrote that the leverage points are counterintuitive and that the higher ones are the ones the system resists most. A practitioner who treats the twelve as a calibrated scale is claiming something the source never offered. A practitioner who computes the elasticities for their own system is doing exactly what she asked.

DREAM

What becomes ordinary

In the economics that has absorbed this, every systems claim in a paper carries its licence in the sentence next to it.

A modelling paper that calls a firm an open system says which flows cross the boundary and at what rate, because that is what the word means and the number is short. A paper that invokes self-organisation states the gradient and the threshold, because the threshold exists and it is computable. Nobody writes thermodynamics requires unless they have divided something by a temperature, and nobody writes autopoietic unless they can name the components and the process that makes the boundary. The words have gone back to meaning what they mean, and the effect is not a narrowing. It is that they now carry weight.

Referees ask the licence question as a matter of routine, the way they ask for a standard error. It takes ninety seconds and it is never experienced as hostile, because the answer is usually yes, and here is the equation — the borrowing was sound and the author simply had not been asked to show it.

The physicists have stopped smiling at this literature. That is worth more than it sounds, because several of the hardest problems in regenerative economics — material cycles, rate limits, the thermodynamic cost of recovering dispersed elements — are problems where a physicist and an economist have to write one paper together, and until recently the vocabulary made that awkward in a way the mathematics never did.

Students learn the four sources in the original. They read Bertalanffy's steady state, Prigogine's balance equation, the six-point key, and the twelve, and they learn each one together with the thing it forbids. They arrive in their first job able to tell the difference between an argument and an image, which turns out to be the most commercially useful thing in the curriculum.

And the field's best claims have got louder, because they are no longer standing next to claims that will not hold. Structural coupling, the rules level, information asymmetry as a leverage point, near-decomposability, path dependence as a limit theorem — all of them were always strong. What changed is that they now travel alone.

DESIGN

The borrowing licence

Here is the instrument, and it is four questions. Run them on any sentence of systems language, your own included, before it goes into a paper, a board memo or a prospectus.

One — what do the terms refer to in this domain? Name the components. Name the boundary. Name the flows and their units. If system here means the firm, say whether people are components or environment, because the answer changes everything downstream. The Teubner test: he could answer this, which is why his extension is arguable rather than empty.

Two — is the source claim a theorem, a measurement, or an ordered intuition? All three are useful and they carry different weight. Equifinality is a theorem about open systems. The critical Rayleigh number is a measurement. Meadows' twelve are an ordered intuition offered as one — she says so in the essay's own opening. Say which you are holding.

Three — what is the threshold or the condition? Almost every real systems result has one: a critical gradient, a linear regime, a conjunction that must hold in full. Find it in the source and state whether your case satisfies it. This question alone catches most of the field's loose claims, and it catches them early enough that the paragraph can simply be rewritten.

Four — what would falsify the transferred claim? If nothing would, you have an image. Images are legitimate and this house uses them; they belong in the letter and the plate, not in the arithmetic.

Scored on four sentences one hears constantly:

   the economy is an open system with a steady state       4/4   licensed
   rules and information flows outrank parameters          4/4   licensed
   firms are autopoietic                                   1/4   decoration
   systems self-organise to maximise entropy production    1/4   decoration
                                            licensed 2 of 4 = 50.0 %

The sequence of intervention follows Meadows' order and the elasticity result, with one addition of our own. Before choosing a leverage point, compute the elasticity of the outcome you care about with respect to each candidate, in your own model, with your own coefficients. It is an afternoon. It will sometimes tell you that in your system the parameter is nearly as powerful as the goal — the 1.11× case — and when it does, take the parameter, because it is the one you can actually reach this quarter.

Governance. One person in the room holds the licence questions, and the role rotates. It is not a veto and it must never become one; it is a standing question, asked out loud, of every systems claim including the chair's. The failure mode of rigour is that it becomes a status weapon, and the defence against that is rotation plus the rule that the asker has to have their own last claim scored first.

DESTINY

How it holds when nobody is enforcing it

Three things keep this alive once the enthusiasm for it has worn off, and only three.

It is cheaper than the alternative. A licensed claim survives contact with a hostile reader, and an unlicensed one costs you the whole paper. Once a group has lost one argument to a sceptic who checked a division, the four questions stop needing to be sold.

The computation is in the repository, not the prose. Every figure in this chapter is computed in lib/verify/II_04.py and prints its inputs with their units and sources before it prints a result. That file is the thing that persists. Prose decays into paraphrase within two citations; a module that runs does not.

Somebody who did not write the claim reads it. The author is the one person who cannot see the gap between what their source says and what their sentence says, because they read the sentence through the source. This is structural rather than a failing of care.

Now the honest part, because this discipline has its own failure modes and they are specific.

It fails when the licence test becomes a purity test, and the room learns that the safest thing is to make no systems claim at all. That is a worse outcome than the loose version: the instinct behind the economy is a living system is correct and the field needs it.

It fails when a genuinely good ordered intuition is discarded because it is not a theorem. Meadows' twelve are not calibrated and they are still the most useful single page in the practice.

It fails when the arithmetic is done once, at publication, and never again as the system changes — an elasticity computed in 2019 on a system whose loop dominance has since shifted is a number that is now lying quietly.

And it fails most often in the way every rigour programme fails: somebody computes the easy thresholds, finds them satisfied, and stops before the one that binds. The four questions are a floor. They are not a ceiling, and a paper that passes all four can still be wrong.

DELIGHT

What it feels like

There is a specific pleasure in the moment a metaphor turns out to be a theorem. You have been saying a thing for years because it felt true; you go and read the source properly, expecting to have to soften it; and instead there is an equation that says it, exactly, with a constant in it.

It happens more often than the sceptical mood of this chapter suggests. Equifinality is real. Structural coupling is real and it is a factor of 346. The elasticity of a system's equilibrium with respect to its goal is exactly one, and the first time you derive that on the back of something it is genuinely delightful, because Meadows put the goal near the top of her list on judgement in 1999 and the algebra has been agreeing with her ever since without being asked.

And there is the other pleasure, the one that comes from subtraction. Letting go of a claim you cannot support is lighter than defending it. You stop bracing. The next time a physicist is in the room you find you are looking forward to it, which is not a feeling this field has had very much of, and it is entirely available.

OPERATIONALIZE THIS

At the level of finance

Prigogine's result has a direct balance-sheet consequence and almost nobody has written the instrument. Here it is.

The structure: a covenanted maintenance-of-order reserve.

If order is maintained by continuous throughput, then the throughput cost of an asset base is not discretionary operating expenditure. It is the price of the asset continuing to be the asset. The instrument funds it ahead of distributions and covenants it, on the model of the replacement reserve in commercial real estate lending and the asset retirement obligation under IAS 37 and ASC 410-20.

The mechanics.

Sizing. Measured maintenance throughput as a percentage of replacement cost, from the organisation's own three-year history — not a benchmark. Worked: an asset base of £40,000,000 at replacement cost, measured throughput 4.2 percent per year, gives £1,680,000 annually and a three-year required reserve of £5,040,000.
The account. Restricted cash held by the agent bank, released against invoiced maintenance only. Restricted, not notional — an unrestricted reserve is an intention.
The covenant. No distribution while coverage is below 1.000. Against £3,100,000 currently funded, coverage is 0.615, the shortfall is £1,940,000, and funding it over eight quarters costs £242,500 per quarter. That is the negotiation, and it is a small one.
Term. Coterminous with the senior facility, with the reserve surviving refinancing. It attaches to the asset, not to the loan.
Security. The reserve account itself, plus a negative pledge on the assets it maintains.

The balance-sheet treatment. Where a present obligation exists — a contracted overhaul, a regulatory inspection cycle, a restoration duty — it is a provision under IAS 37, measured at the best estimate and discounted where the effect is material. Where it does not, it is an appropriation within equity, disclosed, with the restricted cash shown separately from cash and cash equivalents. Take it to the auditors early; this is a conversation about useful economic life and the timing of obligations, which they have every year.

The counterparty. The senior lender, and the argument to them is not environmental. A borrower whose asset condition is covenanted has lower loss given default, and the reserve is cash the lender can see. Start with the agent on an existing facility at the next amendment, where the marginal cost of adding a schedule is close to nothing.

Why it is priced against a threshold and not a trend. This is the part that comes from the physics, and it is the part that makes the covenant defensible. Model an asset condition index that decays faster the worse it already is — dC/dt = −6·(1 + 2·(100 − C)/100) + m — starting from a realistic 99 rather than a pristine 100, with structural failure at C = 55:

   maintenance at 105 % of decay rate  →  does not reach C = 55 in 200 years
   maintenance at 100 %                →  C = 55 in 31.75 years
   maintenance at  95 %                →  C = 55 in 21.75 years
   maintenance at  90 %                →  C = 55 in 17.68 years
   maintenance at  75 %                →  C = 55 in 12.09 years
   maintenance at  50 %                →  C = 55 in  8.26 years

A five percent funding shortfall costs 31.5 percent of the asset's life, not five percent. A dissipative structure does not decay in proportion to the throughput it is denied. It holds, and then it does not — and that convexity is the entire commercial case for funding the reserve rather than trusting the budget.

The number that decides it.

                   funded reserve
   ------------------------------------------------  ≥  1.000
    three-year measured maintenance throughput cost

One ratio, on the front page, tested before every distribution.

The first ninety days.

Day	Action	Artifact
1–15	Pull three years of maintenance spend by asset class	Throughput history
16–30	Compute throughput as a percentage of replacement cost	The sizing paper
31–45	Run the convexity model on your own decay rates	The threshold chart
46–60	Draft the reserve schedule; open the restricted account	Reserve schedule
61–75	Agree the covenant at the next amendment	Signed amendment
76–90	First coverage test published in the standing pack	The coverage ratio

APPRECIATIVE QUESTIONS

Twelve, for a room

Discovery — what is already working

Where in our work have we already made a systems claim that turned out to be exactly right, and what did we do to find that out?
Which of our models already names its components, its boundary and its units without being asked — and who built it that way?
Think of a time someone in this room corrected a borrowed term and the argument got stronger. What made that safe to do?

Dream — what becomes possible

If every systems claim we made carried its licence in the next sentence, what would we be able to say to people who currently discount us?
Imagine a physicist joining this team next month and reading everything we have published. Which three pieces would we be most pleased for them to start with?
If we knew the elasticity of our main outcome with respect to every lever we have, what would we do differently on Monday?

Design — what we build

What is the gradient across our system, and is it above or below the threshold where structure forms on its own?
Which of our interventions are at Meadows' rules and information levels, and which are at parameters — and have we ever computed which pays more here?
Who holds the licence questions in our room, and how do we make sure the role rotates before it becomes a veto?

Destiny — how it holds

What would have to be true for the computation behind our claims to still run in five years, on a machine none of us owns yet?
Which of our numbers was computed once and has quietly stopped being true? How would we find out?
When somebody outside this room checks one of our figures, what do we want the second thing they check to be?

WORKS CITED

Bertalanffy, L. von (1950). "The Theory of Open Systems in Physics and Biology." Science, 111(2872), 23–29.

Bertalanffy, L. von (1968). General System Theory: Foundations, Development, Applications. George Braziller.

Boulding, K. E. (1956). "General Systems Theory — The Skeleton of Science." Management Science, 2(3), 197–208.

Simon, H. A. (1962). "The Architecture of Complexity." Proceedings of the American Philosophical Society, 106(6), 467–482.

Prigogine, I. (1967). Introduction to Thermodynamics of Irreversible Processes, 3rd edn. Interscience.

Nicolis, G. and Prigogine, I. (1977). Self-Organization in Nonequilibrium Systems: From Dissipative Structures to Order through Fluctuations. Wiley.

Prigogine, I. and Stengers, I. (1984). Order out of Chaos: Man's New Dialogue with Nature. Bantam.

Chandrasekhar, S. (1961). Hydrodynamic and Hydromagnetic Stability. Oxford University Press.

Landsberg, P. T. and Tonge, G. (1979). "Thermodynamic Energy Conversion Efficiencies." Journal of Applied Physics, 50(4), R1–R20.

Schrödinger, E. (1944). What is Life? The Physical Aspect of the Living Cell. Cambridge University Press.

Morowitz, H. J. (1968). Energy Flow in Biology. Academic Press.

Georgescu-Roegen, N. (1971). The Entropy Law and the Economic Process. Harvard University Press.

Bianciardi, C., Tiezzi, E. and Ulgiati, S. (1993). "Complete Recycling of Matter in the Frameworks of Physics, Biology and Ecological Economics." Ecological Economics, 8(1), 1–5.

Martyushev, L. M. and Seleznev, V. D. (2006). "Maximum Entropy Production Principle in Physics, Chemistry and Biology." Physics Reports, 426(1), 1–45.

Maturana, H. R. and Varela, F. J. (1980). Autopoiesis and Cognition: The Realization of the Living. D. Reidel.

Varela, F. J., Maturana, H. R. and Uribe, R. (1974). "Autopoiesis: The Organization of Living Systems, Its Characterization and a Model." BioSystems, 5(4), 187–196.

Maturana, H. R. and Varela, F. J. (1987). The Tree of Knowledge: The Biological Roots of Human Understanding. Shambhala.

Luhmann, N. (1984). Soziale Systeme: Grundriß einer allgemeinen Theorie. Suhrkamp. (English: Social Systems, Stanford University Press, 1995.)

Teubner, G. (1993). Law as an Autopoietic System. Blackwell.

Meadows, D. H. (1999). Leverage Points: Places to Intervene in a System. The Sustainability Institute.

Meadows, D. H. (2008). Thinking in Systems: A Primer. Chelsea Green.

Forrester, J. W. (1969). Urban Dynamics. MIT Press.

Meadows, D. H., Meadows, D. L., Randers, J. and Behrens, W. W. (1972). The Limits to Growth. Universe Books.

Abson, D. J. et al. (2017). "Leverage Points for Sustainability Transformation." Ambio, 46(1), 30–39.

Fischer, J. and Riechers, M. (2019). "A Leverage Points Perspective on Sustainability." People and Nature, 1(1), 115–120.

Arthur, W. B. (1989). "Competing Technologies, Increasing Returns, and Lock-In by Historical Events." The Economic Journal, 99(394), 116–131.

Ostrom, E. (1990). Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press.

Kopp, G. and Lean, J. L. (2011). "A New, Lower Value of Total Solar Irradiance: Evidence and Climate Significance." Geophysical Research Letters, 38, L01706.

Loeb, N. G. et al. (2018). "Clouds and the Earth's Radiant Energy System (CERES) Energy Balanced and Filled (EBAF) Top-of-Atmosphere Edition-4.0 Data Product." Journal of Climate, 31(2), 895–918.

Hartmann, D. L. (2016). Global Physical Climatology, 2nd edn. Elsevier.

Note on figures. Every figure in this chapter is computed in lib/verify/II_04.py, which prints each input with its unit and its source before it prints a result. The planetary entropy budget, the Rayleigh thresholds, the autopoiesis scoring, the leverage elasticities and the reserve convexity model are all reproducible there and their inputs are editable. Where a source claim is contested — maximum entropy production, the extension of autopoiesis to social systems, the calibration of Meadows' ordering — the contest is stated in the text rather than resolved in a footnote.