Haute Lumière

Commerce · II · MMXXVI · daylight

La Bourse  /  Volume II  /  Nº II.E3

The Translation: One Economy, Two Vocabularies

Volume II — Foundations: The Paradigm and the Science · Extension

Nine movements, two vocabularies.

---

THE PLATE

---

THE LETTER

There are two vocabularies in the room and everybody knows it.

One of them says regeneration, flow, wholeness, stock, resilience. The other says depreciation, working capital, consolidation, expected shortfall, covenant. The people who use the first are told, politely, that they are describing values rather than facts. The people who use the second are told, less politely, that they are describing money rather than the world. Both groups then repeat themselves more loudly, and the meeting ends.

The argument is unnecessary for most of what it covers and irreducible for the rest, and the whole value of this chapter is knowing which rows are which.

So here is the object: a term table in which each row is one thing under two names, with the arithmetic or the citation that shows it is one thing. Fourteen rows. A recovery fraction is inventory turns. A leverage point is an elasticity. Resilience is expected shortfall and the buffer that funds it. Care work is a satellite account. These are not analogies. In every row the two sides compute the same number or reconcile to the same published figure, and where they do, the argument was about words.

And then the part that is worth more than the table: the terms that do not translate at all. Five in one direction, six in the other, and six false friends — words that pass between the vocabularies unremarked and change meaning on the way. Twenty-five terms in the joint vocabulary, of which a translation table can carry 56.0 percent. The remaining 44.0 percent is not a gap in the dictionary. It is a bound, and it has a shape: everything finance holds and this vocabulary lacks is about time and the order of claims; everything this vocabulary holds and finance lacks is about quality and direction. The two are missing orthogonal things, which is why neither can absorb the other and why both rooms keep being right.

The false friends are the expensive ones. An untranslatable term fails loudly — somebody asks what it means. A false friend passes, and by the arithmetic below, 4.8 of the six are expected to pass unnoticed into a document that matters.

— The Editors

---

DISCOVERY

What is already working

Translation between these two vocabularies is not an aspiration. It has been built, at national scale, by statisticians, and the builds are public.

The System of Environmental-Economic Accounting. Adopted by the United Nations Statistical Commission in 2012 as an international statistical standard — the same status as the System of National Accounts itself — with the ecosystem accounting volume following in 2021, and compiled now by the great majority of the statistical systems that have tried. The feature that makes it a translation rather than an alternative is this: SEEA does not replace GDP with a better number. It builds physical and monetary accounts for water, energy, timber, fish, land and ecosystem services that reconcile line by line with the national accounts, so a finance ministry reads both without converting. Botswana's water accounts changed how water was allocated. Costa Rica's forest accounts did the same work in a different climate. In both, the accounting came first and the policy followed.

The household satellite account. The ONS values unpaid household service work and publishes it beside GDP rather than inside it — £1,239.0 billion in 2016, which is £1.24 trillion — in hours and in money, with the method open. Its share of GDP is 63.8 percent on the £1,943.0 billion denominator Chapter II.10 uses and 63.1 percent as Chapter II.03 quotes it, a spread of 0.7 points that is entirely the GDP vintage and not the household work; divide and the second figure implies a GDP of £1,963.5 billion. The BEA does the same for the United States: 25.7 percent of GDP in 2010. The move is the one this whole chapter recommends: translate, publish both columns, and let the reader do the reconciliation that the two documents now permit.

IPMVP. The International Performance Measurement and Verification Protocol exists because two parties needed to agree what a saving was before they could finance one. It is a dictionary with a signature block: baseline, method, period, verifier. Every shared-savings facility in this edition rests on it, and it was built by engineers and lenders sitting in the same room and refusing to leave until the words meant one thing each.

The GHG Protocol's scopes. A translation with a known and published failure mode, which is what makes it exemplary rather than merely useful: the standard tells you that chain inventories overcount, which is why corporate inventories cannot be summed into a national total. A dictionary that documents its own mistranslations is doing the thing this chapter is about.

Te Awa Tupua. In 2017 the New Zealand Parliament recognised the Whanganui River as a legal person with rights, powers, duties and liabilities, and created an office to speak in its name. Read it here as a translation event: a concept with no counterpart in commercial law was given one by legislation, because it could not be given one by modelling. That is the honest route for an untranslatable term, and it is the exception that shows why the list in the Arithmetic matters.

The exergy-indexed tariff. Chapter II.06's heat offtake prices by grade rather than by quantity, converting a subjective quarrel about quality into a thermometer reading. It exists precisely because the vocabulary does not: there is no accounting standard for work capacity, so the parties had to write one into a contract.

Six builds, one pattern: in each, somebody wrote the dictionary down, in a document with a signature on it, and the argument stopped being about words and started being about numbers.

---

THE ARITHMETIC

The table, the failures, and the bound

Every figure below is computed in lib/verify/II_E3.py, which prints each input with its unit and its source before it prints a result.

First, the table. Fourteen rows, each one object under two names.

The arithmetic of four rows, so the table can be checked rather than believed.

  circulation multiplier, M = 1 / (1 - p)
     p = 30 %   M =  1.43 x        p = 50 %   M =  2.00 x
     p = 80 %   M =  5.00 x        p = 90 %   M = 10.00 x

  leverage, S* = aG/(a+k), with a = 0.30 and k = 0.10
     elasticity to the goal                    1.000
     elasticity to the parameter               0.250
     ratio                                      4.00 x

  trust versus contract, crossover membership n = 1 + 1.667 T
     T =  12 transactions/member/yr     n =  21 members
     T =  30                            n =  51
     T =  60                            n = 101

  the kernel, at an attribute vector of k = 12
     gradient one price recovers                8.3 %
     directions it cannot see                     11

Read the second block as a translation and it does real work. Work on the rules, not the parameters is a sentence from one vocabulary; the elasticity of the outcome to the goal is four times its elasticity to the parameter is a sentence from the other; and they are the same sentence, with the second one telling you when the first is false — because at other loop dominances the ratio collapses toward one and the parameter is the better lever.

Second, and this is the chapter: where the translation fails.

  rows that translate                                14
  living-systems terms with no counterpart            5
  finance terms with no counterpart                   6
  ------------------------------------------------------
  joint vocabulary                                   25 terms
  share a translation table can carry              56.0 %
  share it cannot                                  44.0 %

The five with no counterpart in finance. Grade of energy — no accounting standard recognises work capacity, and a megawatt-hour is a megawatt-hour in every contract ever signed. The kernel directions of a price — by the theorem, a price cannot name what it has no component in, and naming them requires a second channel with a bill attached. Surplus created and not captured — innovators capture on the order of 2.2 percent of the surplus they generate, and there is no accounting object for the rest; it is not goodwill, not an intangible, not deferred revenue, not anything. Group-level fitness — Muir selected cages rather than hens, and no consolidated reporting entity corresponds to the cage. Regeneration of a stock nobody owns — the aquifer is an asset and it is on no balance sheet, because title is precisely the thing it lacks.

The six with no counterpart in living-systems economics, and this list is the one this tradition does not like to write down. The discount rate — a forest has no weighted average cost of capital, and nothing in this vocabulary converts a future flow into a present number. Liquidity — the ability to exit a position at short notice, for which no ecological quantity exists, because an ecosystem cannot exit. Seniority and the capital stack — an ordering over claims in insolvency; nothing in an ecosystem ranks claimants when the system fails. Legal personality and fiduciary duty — a rule rather than a flow, which is why Te Awa Tupua had to be legislated. The cost of price discovery — 0.67 percent of aggregate market value a year, about $10,421 per bit, and no term in this vocabulary names the cost of finding out what a thing is worth. An arm's-length price where no market exists — five permitted methods and a range, because a number is required and there is nothing to read it off.

Now the cut, and it is a fact about the two lists rather than an argument.

Look at what each list contains. Every finance-only term is about time and the order of claims: when the money arrives, who gets paid first, how fast you can leave, who owes a duty to whom. Every living-only term is about quality and direction: what grade the energy is, which way the instrument is pointing, what was created and not captured, at what level the thing is being selected.

The two vocabularies are not rivals describing one world in different words. They are two projections of it, and the things each one is missing are orthogonal to the things the other is missing. A price is a scalar functional and loses k − 1 directions; this tradition has no operator at all that carries a value across time. Neither contains the other, and 56.0 percent is not a measure of how hard people have tried. It is the bound.

Which is why the correct object is not a better translation. It is a dictionary with a marked untranslatable section — the thing this chapter is.

Third, the false friends, which are worse than the blanks.

Six words pass between the vocabularies without anybody stopping them, and change meaning on the way across.

And the arithmetic of why they cost more than the blanks:

  P(an untranslatable term is noticed)        100 %   somebody asks what it means
  P(a false friend is noticed)                 20 %   it reads as ordinary English

  false friends expected to pass unnoticed      4.8
  untranslatable terms expected to pass         0.0

The dangerous half of a dictionary is not the blank entries. A term with no counterpart announces itself at the first reading; a word that translates silently reaches the term sheet, the covenant and the board minute, and is discovered in the quarter when the two parties turn out to have signed different agreements.

Fourth, the honest negative, and it is about the table rather than the failures.

A row that rests on an identity round-trips exactly: translate recovery fraction into inventory turns and back, and the number returns unchanged, because 1/(1−p) is the same function read in either direction. A row that rests on a chosen parameter does not, and the loss is silent.

  rows resting on an identity                  10        71.4 %
  rows resting on a chosen parameter            4        28.6 %

Four of the fourteen — the regeneration rate, the externality, wellbeing and abundance — carry a parameter that somebody selected: a rate, an attribute vector, an inequality aversion, a choice about which sense of scarce is meant. Translate one of those and the parameter travels invisibly inside the number.

The worked case is the sharpest, and it belongs to the other vocabulary:

  a sum a century out                          GBP 1,000,000
  horizon                                            100 years
  at 1.4 % — the low end of the published debate
     present value                             GBP   249,003
  at 4.3 % — the higher end
     present value                             GBP    14,845
  ------------------------------------------------------------
  ratio                                              16.77 x

Same sum, same century, two defensible published rates, and a factor of 16.77 between them — and there is no living-systems operator that produces either number. A practitioner who translates a long-horizon regeneration claim into present value has imported the entire intergenerational argument, uncosted, in one cell of a spreadsheet. That is not an argument for refusing to discount. It is an argument for writing the rate on the same page as the answer, which is what the instrument in the ninth movement does.

---

DREAM

What becomes ordinary

In the practice that has absorbed this, every serious document carries a definitions schedule with two parts, and the second part is the one people read.

Part A is the translated terms: recovery fraction and inventory turns, regeneration rate and useful economic life, systemic position and concentration limit, each with the arithmetic that identifies them, so that a disagreement about the word is settled by a division. Nobody argues about whether the terms mean the same thing. They check.

Part B is the untranslatable list, and it names the party who carries the risk of each gap. The discount rate is stated on the face of the paper with the answer, not buried in a model. The grade of the energy is a column in the tariff. The surplus the firm created and did not capture is described in words, in the narrative, and explicitly not totalled — because there is no account it belongs in, and saying so is more honest than inventing one.

The false friends have been retired from documents that matter. Nobody writes efficiency without a qualifier; nobody writes diversification without saying whose portfolio; nobody writes capital across the two vocabularies without naming which kind. The words are still used, freely and warmly, in the letter and the conversation, where ambiguity is a feature. They are not used in a covenant.

Meetings between the two rooms have become shorter and more productive, and the reason is unromantic: the participants now know which disagreements are about words and can be settled in ninety seconds, and which are genuinely about the 44.0 percent that does not translate — where the honest answer is that each side is holding something the other's vocabulary cannot express, and the conversation is about who carries that risk rather than about who is right.

And a generation of analysts can do both. They compute a Carnot factor and a weighted average cost of capital in the same afternoon and do not find it unusual, because they were taught the dictionary in the second week along with the list of words that are not in it.

---

DESIGN

The bilingual schedule, built

One artifact, two parts, a fortnight.

Part A — the translated terms. Four columns: the term in each vocabulary, the identification (an equation or a published reconciliation), and whether it rests on an identity or on a chosen parameter. The last column is the one people leave out and it is the one that matters, because it tells a later reader which rows can be relied on to round-trip.

Part B — the untranslatable schedule. Three columns: the term, the vocabulary it lives in, and the party who bears the risk of the gap. That third column will be resisted, exactly as the fifth column of Chapter II.03's boundary register is resisted, and for the same reason: it names somebody. Often the answer is benign — the lender carries the discount-rate risk because the facility is fixed-rate — and writing it down is what makes the benign cases visible as benign.

The build.

Days one to five: list the terms that appear in the document and mean something financial. Most agreements have between fifteen and forty. Take them from the document, never from a glossary.

Days six to ten: sort into A, B and false friend. The test for Part A is mechanical and it is the only test that works: can you write the identity? If two people cannot compute the same number from both sides, the row is not in Part A. The test for a false friend is likewise mechanical: does the word appear in both vocabularies with different referents? If yes, it is banned from the operative clauses and defined explicitly wherever it survives.

Days eleven to fourteen: fill the risk column in Part B, and take it to the person named in it before it goes anywhere else.

Governance. The schedule is owned by whoever owns the document — counsel on a facility, the controller on a set of accounts, the programme director on a business case. It is reviewed when the document is amended and at no other time, because a schedule with its own review cycle becomes a separate artifact that drifts from the agreement it defines.

Sequence, and why this order. Part B first, always. The untranslatable list takes an afternoon, needs no arithmetic, and is the half that prevents disputes rather than settling them — and a team that starts with Part A will spend three weeks perfecting identities and never reach the part where the money is.

What to resist. The schedule is not a campaign to translate everything. A term that has been forced across when it does not go is worse than a term left in Part B, because it now carries a number somebody will rely on. The purpose is to make the translations checkable and the gaps visible — and some of the gaps should stay gaps.

---

DESTINY

How it holds when the drafters have gone

Three things keep a dictionary alive, and they are all about attachment.

It is attached to a document that is signed. A glossary published on its own is read once. A Schedule 1 to an agreement is read every time somebody argues about a clause, which is the only reliable form of readership.

It is used to settle something in its first year. An instrument that has never been used in an argument is quietly dropped at the next amendment. Use it the first time two parties disagree about efficiency — and they will, because that word does three jobs.

Somebody who is not the drafter reads it. The drafter is the one person who cannot see the gap between what the word means to them and what it means on the page, because they read the page through the meaning.

Now the failure modes, named plainly.

It fails when Part B is dropped at the second revision, because it is the part with no numbers and it looks like commentary. The guard is the risk column: a term with a named party attached is harder to delete than a term with a definition attached.

It fails when a translated row is quietly regraded — an identity row whose parameter has been changed so that it no longer round-trips, and nobody versioned it. Version the schedule; keep the previous identity visible.

It fails, most expensively, when the dictionary is used in one direction only. Translating a living-systems claim into finance vocabulary makes it financeable, which makes it worth misstating — and a schedule that only ever converts regeneration into cash flows has become a marketing instrument. The countermeasure is Part B: a schedule that names six things finance can say and this vocabulary cannot is visibly not a piece of advocacy.

And it fails when somebody mistakes the dictionary for an agreement. Two parties who share definitions can disagree completely, and should be able to. The schedule buys a shorter argument, not a shared position, and a room that expects the second from it will abandon the first.

---

DELIGHT

What it feels like

There is a specific pleasure in the moment a year-long argument dissolves because somebody wrote down what a word meant. Two teams discover they have been using efficiency for different quantities, both correctly, and the disagreement was never about the plant. The hour they get back is not the good part. The good part is the quality of the argument that follows, which is about something real.

And there is the pleasure of the blank entry, which is rarer and better. Writing no counterpart in a column is a small act of honesty that costs nothing and changes the temperature of a room. The finance people relax, because somebody from the other side has just conceded that a discount rate does real work that nothing in their vocabulary does. The ecologists relax, because somebody has written down that a megawatt-hour at sixty degrees and a megawatt-hour at three hundred are not the same object and no contract in the world currently knows it.

Best of all is the moment somebody from the other vocabulary uses one of your rows in their own argument, in their own words, without noticing that it came across. That is a translation having worked — not when both sides agree, but when one side stops being able to tell which language the sentence started in.

---

OPERATIONALIZE THIS

At the level of finance

The dictionary becomes real when it is a numbered schedule to something with a signature block.

The structure: Schedule 1 to a facility agreement, joint venture or offtake — Part A, defined terms with their identifications; Part B, the untranslatable list with the party who bears each gap.

The mechanics.

• Where it sits. As a schedule, incorporated by reference in the definitions clause, not as a side letter. A side letter is not read by the person who administers the agreement three years later, and that person is the reader this instrument is for.
• Part A rows carry their identity. Recovery fraction is defined as the ratio whose denominator is stated in the same clause — the denominator definition being the single most gameable object in this whole edition, and the one Chapter II.01 tells you to sign before anything else.
• Part B rows carry a risk allocation. For each untranslatable term, one sentence: who bears the consequence if it turns out to matter. The discount rate, the grade specification, the title to a regenerating stock, the ordering of claims — these are the four that appear in almost every agreement of this kind.
• The false-friend clause. Six words, listed, each defined once and used only in that sense throughout the document. This clause costs one paragraph and it is the highest-return paragraph in the schedule.
• Amendment. A change to any Part A identity requires the previous identification to remain visible in the amended schedule. The same rule Chapter II.03 puts on a boundary election, for the same reason: it removes almost all of the value of changing a definition quietly.

The balance-sheet treatment. The schedule itself is a drafting cost and runs through the income statement. What it changes is elsewhere and it is material: where Part A identifies a regeneration rate with a useful economic life, that identification is the evidence supporting a depreciation estimate, and your auditors will ask for it. Take it to them at the planning meeting. The conversation is about useful economic life, which they have every year.

The counterparty. Counsel on both sides, at the same table, with one person from each vocabulary present. The schedule drafted by lawyers alone will contain Part A and no Part B, because Part B is not a legal object; the schedule drafted by practitioners alone will not survive a dispute.

The number that decides it.

  drafting cost, Schedule 1 Parts A and B       EUR    18,000
  enforcement cost, share of the claim              21.5 %
    (World Bank enforcing-contracts indicator,
     high-income OECD)
  ------------------------------------------------------------
  break-even claim, if a dispute is certain     EUR    83,721
  P(an ambiguous term is disputed)                  25.0 %
  break-even value at stake                     EUR   334,884

Above roughly a third of a million euros of value resting on ambiguous terms, the bilingual schedule pays for itself out of one avoided argument. Below it, write Part B anyway and skip Part A: the untranslatable list is an afternoon, it is the half that prevents disputes rather than settling them, and it is worth having whether or not anything is signed.

The first ninety days.

Read alongside the other instruments in this edition, this one is the connective tissue: it is the only instrument in the catalogue that does not raise, save or allocate money, and every other instrument in the catalogue is unenforceable without something like it, because each of them turns on a word that means two things.

---

APPRECIATIVE QUESTIONS

Twelve, for a room

Discovery — what is already working

1. Where have two people here already agreed on a number while using completely different words for it? What made that possible?
2. Which of our documents already defines a term that both vocabularies use — and who drafted it, and what did they see that the standard did not require?
3. Think of a time somebody said we have no word for that and it turned out to be the most useful sentence in the meeting. What was the term?

Dream — what becomes possible

1. If every agreement we signed carried a list of the terms that do not translate, what is the first dispute we would never have again?
2. Imagine our analysts could compute a Carnot factor and a cost of capital in the same afternoon and thought nothing of it. What would we ask them for?
3. If the parameter inside every translated number were printed on the same page as the number, which of our decisions would look different?

Design — what we build

1. Which three words in our live documents are doing two jobs, and which of them is in an operative clause?
2. Which of our rows rest on an identity and which on a chosen parameter — and has anybody ever checked that one of them round-trips?
3. Who bears the risk of the gap on each term in our Part B, and have we asked them?

Destiny — how it holds

1. What would have to be true for the schedule to still be accurate when the people who drafted it have gone?
2. What is the first sign we would see that the dictionary had started being used in one direction only, and who would notice it first?
3. Ten years from now, which untranslatable term would we most want to have been given a counterpart — and what would it take, in law or in a contract, to give it one?

---

WORKS CITED

Ayres, R. U. and Warr, B. (2009). The Economic Growth Engine: How Energy and Work Drive Material Prosperity. Edward Elgar.

Bridgman, B., Dugan, A., Lal, M., Osborne, M. and Villones, S. (2012). "Accounting for Household Production in the National Accounts, 1965–2010." Survey of Current Business, 92(5), 23–36.

Dyer, J. H. and Chu, W. (2003). "The Role of Trustworthiness in Reducing Transaction Costs and Improving Performance: Empirical Evidence from the United States, Japan, and Korea." Organization Science, 14(1), 57–68.

Efficiency Valuation Organization. International Performance Measurement and Verification Protocol (IPMVP), Core Concepts. Successive editions.

Financial Accounting Standards Board (2003). FIN 46(R): Consolidation of Variable Interest Entities.

Gabaix, X., Gopikrishnan, P., Plerou, V. and Stanley, H. E. (2003). "A Theory of Power-Law Distributions in Financial Market Fluctuations." Nature, 423, 267–270.

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

Mount, K. and Reiter, S. (1974). "The Informational Size of Message Spaces." Journal of Economic Theory, 8(2), 161–192.

New Zealand Parliament (2017). Te Awa Tupua (Whanganui River Claims Settlement) Act 2017.

Nordhaus, W. D. (2004). "Schumpeterian Profits in the American Economy: Theory and Measurement." NBER Working Paper 10433.

Nordhaus, W. D. (2007). "A Review of the Stern Review on the Economics of Climate Change." Journal of Economic Literature, 45(3), 686–702.

Office for National Statistics (2018). Household Satellite Account, UK: 2015 and 2016.

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

Price, G. R. (1970). "Selection and Covariance." Nature, 227, 520–521.

Ravallion, M. (2012). "Troubling Tradeoffs in the Human Development Index." Journal of Development Economics, 99(2), 201–209.

Stern, N. (2007). The Economics of Climate Change: The Stern Review. Cambridge University Press.

Szargut, J., Morris, D. R. and Steward, F. R. (1988). Exergy Analysis of Thermal, Chemical and Metallurgical Processes. Hemisphere.

Ulanowicz, R. E. (2009). A Third Window: Natural Life Beyond Newton and Darwin. Templeton Press.

United Nations, European Commission, FAO, OECD, IMF and World Bank (2014). System of Environmental-Economic Accounting 2012 — Central Framework.

United Nations (2021). System of Environmental-Economic Accounting — Ecosystem Accounting.

World Bank. Doing Business: Enforcing Contracts. Successive editions; programme discontinued 2021.

World Resources Institute and World Business Council for Sustainable Development (2011). Corporate Value Chain (Scope 3) Accounting and Reporting Standard.

Note on figures. The term table and its identifications, the circulation multiplier, the leverage elasticities, the trust–contract crossover, the kernel share, the counts of untranslatable terms in both directions and the recoverable share, the detection arithmetic for false friends, the identity-versus-parameter split, the century-horizon present values, and the bilingual schedule's break-even are all computed in lib/verify/II_E3.py and reproduced by python3 lib/verify.py II.E3, which prints every input with its unit and its source before it prints a result. Figures carried in from the eleven chapters are cited to the chapter that computed them. The two discount rates are the low and high ends of a published and unresolved debate and are printed as inputs rather than adopted.