Conversion
兑换

Writing Declaration: This paper was co-drafted with Claude (Anthropic). All intellectual decisions, framework design, and final editorial judgments were made by the author.

Introduction

In the autumn of Year 24, Sundial Ratings downgraded Apple City's municipal bonds by one notch.

Apple City is a city in the Kingdom State. Two years earlier it had issued municipal bonds to raise funds for the expansion of the City Hospital. Sundial is the largest credit rating agency in the Kingdom State, with a market share of about sixty percent; the rating fee is paid by the party rated. The announcement of the downgrade listed several reasons, one of which was: "The proposed new rules for online marketplaces may reduce the tax base." That year the City Council of Apple City had just declared its support for the draft Online Marketplaces Act then under deliberation in the State Legislature. The draft required online marketplaces to publish their ranking rules and to set up a seller appeals body independent of the marketplace; a further clause allowed the state government to request the Office of Trade to initiate breakup proceedings on the basis of the Competition Review Division's report. After the downgrade, the cost of Apple City's further borrowing rose.

A council member said publicly that the downgrade had to do with Crossways. Crossways Market, run by Crossways Group, is precisely the kind of marketplace the draft is aimed at. Among Sundial's shareholders, the Qiao Family Office holds thirty percent of its shares; this family office holds all of Crossways Group's Class B shares, which on the share register of the spring of Year 21 carried about fifty-six percent of the voting power in Crossways. Crossways Group and its settlement subsidiary, Crossways Pay, are moreover Sundial's two largest clients, and the rating fees they pay Sundial come to about ten percent of Sundial's revenue. Sundial's reply was: "Ratings are independent; the views of the rating committee are not influenced by shareholders." Sundial does not publish the list of members of its rating committee.

There is another side to this matter as well. From Year 18 on, Sundial has published, free of charge, its rating methodology for the bonds of the Kingdom State's small towns, and several small towns issued bonds for the first time as a result. The Kingdom State's Credit Rating Agency Registration Rules require rating agencies to register, and do not review rating methodologies. The laws and bodies here are all examples set up by this series; the provisions this paper refers to serve only as facts, and this series does not claim that they are the best design.

A rating can be raised, and it can be lowered; the reasons for a downgrade can be right, and they can be wrong. The council member's words are a judgment, not a piece of evidence; Sundial's words are also a judgment, not a proof. This paper does not decide for them whose will this downgrade came from. What is to be asked is something else: a body rates other parties' credit, asks a question on behalf of many people, and many people make decisions on the strength of its answer; can accumulated shareholdings and client relations be converted into control over a channel of questioning of this kind?

The earlier papers wrote of how accumulation forms and where it shows a power-face (Paper 1); of how to read it when, at the level of relations, it has become a chokepoint others cannot get around (Paper 2); and, when it has grown so large that many cannot get around it, of how strong checks must be and where they fall (Paper 3). This paper writes of accumulation turned to other effects. Fame bringing opportunities and capital bringing more capital are ordinary matters of the Emergence Layer, and this paper does not aim at reducing them. But there are two kinds of conversion at which this paper sets a gate: accumulation is not converted into domination over people's Base Layer; accumulation is not converted into control over questioning.

The paper goes in five steps. First the first gate, through the employment contracts of Bluekiln (§1); then the second gate, through Sundial and Echo Wall (§2); then the two directions of political–economic separation, through a vote in the State Legislature and a division head who left office (§3); next, that exchange outside the two gates is left open as usual, through several kinds of fame, with a comparison there with Walzer and Sandel (§4); last, public support obtained through structural dependence, through an emergency loan to Crossways Pay (§5).

The reading follows the Prequel: the construct takes the ought, the position takes accountability, the person takes his own law. The division into the normative layer, the rule layer and the judgment layer also follows the Prequel; this paper states which sentence falls at which layer only where this is used.

§1 Accumulation Is Not Converted into Domination over People's Base Layer

Paper 2 has already read the employment contracts of Bluekiln once. That reading asked whether the constraints of dormitory, kiln scrip, wage advance and non-compete, looked at one by one and taken together, still leave a person his way of leaving when they fall on one and the same person (Paper 2, E12); and how the one concerning where he goes after leaving is to be read (Paper 2, E12a). This section asks a different question of the same set of facts: what was given in exchange to obtain each of these constraints?

Bluekiln Porcelain Works employs three thousand four hundred people, about seventy percent of the town's working-age population; its wages are ten percent above the average for the trade in the Kingdom State, its apprentice program charges no tuition, and the company also paid for the road from the town to the state road. Forty percent of the housing in the town belongs to the company; employees live in dormitories, the rent is deducted from their wages, and they must move out within thirty days of leaving. Twenty percent of wages is paid in "kiln scrip," which can be used only at the company's Bluekiln Cooperative Store and canteen; the cooperative store is the largest shop in town, with prices about fifteen percent higher than in Apple City; there are three other small grocery shops in town, and they do not accept kiln scrip. The company advances wages to employees, up to six months' worth; Article 14 of the employment contract reads: "Where an advance has not been repaid, leaving requires the company's consent." Article 19 is the non-compete: for two years after leaving, the employee may not engage in porcelain-making, or work for a porcelain enterprise, anywhere in the Kingdom State.

Among these clauses there are three things the company has accumulated: forty percent of a town's houses, its largest shop, and a sum of money that can be lent out. They are not in themselves the problem. Renting houses to people to live in, selling things in a shop, lending money to those who urgently need it: all of these are exchanges in the Emergence Layer, and the people of the town really do use them. What is to be asked about is conversion: whether these things have been converted into another kind of effect.

Article 14 is the clearest. An advance is money the company has lent out; money lent must be repaid, and that is a legitimate claim; the company demanding it under the loan agreement is exchange. But what Article 14 demands is not repayment; it is something else: where the advance has not been repaid, leaving "requires the company's consent." A debt is thereby converted into a right of consent, and what it governs is whether a person can go. When the veteran kiln worker gave notice of leaving in the spring of Year 22, four months of advance remained unrepaid, and the company did not consent; a month later he borrowed money from relatives and repaid the advance, and only then did this barrier open. By the wording of this article, a sum of money in the company's hands was converted into the decision on whether or not he left.

Dormitories and kiln scrip are written another way. Housing is tied to the job, and one must move out within thirty days of leaving; in a town where forty percent of the houses belong to the company, leaving the job first requires finding somewhere to live. Twenty percent of wages can be spent only at the company's store and canteen; whether the kiln scrip in hand can still be used after leaving the company is not recorded. What they have been converted into depends on whether the person still has other roads: how much other housing there is in town, and how expensive it is, is not recorded; the three small grocery shops do not accept kiln scrip. How they are read one by one and taken together Paper 2 has already written, and it is not repeated here. This section only places them under one and the same question: whether the accumulated houses, shop and money are converted into control over people's leaving, refusing and revoking.

Why must a gate be set at this kind of conversion? Because the place it converts into is the Base Layer. Leaving, refusing and revoking are the few channels by which a person holds his own in a relation. Institution Theory writes this into the construct's first ought toward its internal subjects: a constraint that "does not compress the life-substrate to the point of threatening the bearer, and does not seal off the channels of exit, refusal and revocation is a limit; one that crosses any of these is a violation" (Institution Theory Paper 2); this ought "does not vary with relational thickness and is not discounted under any condition" (Institution Theory Z9). The original speaks of the construct's own acts toward its internal subjects; here it is extended to the conversions an economic position makes with accumulation (own undertaking). The Base Layer does not rise or fall with assets, and is not exchanged for achievement: how much one person owns does not change whether another's way of leaving is that other's own; however much good a company has done, it cannot get in exchange for it that others need its consent to leave. Food, water, electricity, housing: whatever people need to sustain their lives can, once it is hoarded in one place, be converted into control of this kind; that someone is wealthy does not mean that he may use this wealth to dominate others. This gate therefore has two branches. One is converting what people need to sustain their lives into control that compresses their life-substrate, which Paper 2 already wrote of when it read the sole position of supply at the life-substrate (Paper 2, E11, E11a). The other is converting money, housing or the manner of paying wages into control over leaving, refusing and revoking, and the cases of this section fall in this branch. What counts as necessary to sustain life, and what the Base Layer specifically includes, the list lies at the rule layer (Paper 2, E10).

One overreaching reading of this gate must be guarded against. That a relation has terms, conditions and procedures, and that leaving has a cost, does not mean that the way of leaving has been sealed off. Rights Theory P5 drew this threshold: sealing the exit channel "is not the presence in a relation of terms, procedures, conditions, or costs of exit; it is the subject no longer holding a structural channel for refusing, revoking a standing bond, or passing into a reviewable procedure" (Rights Theory P5 §III). Three months' notice of leaving is a term; having to find other housing on moving out of a dormitory is a cost; these are not in themselves what this gate is to stop. What the gate stops is the step of conversion: accumulation converted into the decision over leaving itself, so that the person no longer holds that structural channel. Article 14, by handing leaving over to the company's consent, touches precisely this step.

A signature does not change this reading either. On kiln scrip there is a written reply from the State Labor Agency: paying part of wages in scrip "requires the employee's written consent"; every employee signed a consent form on being hired, and every article of the employment contract was signed as well. Once it is signed, has the matter become voluntary? Rights Theory P5 has answered this question at the Base Layer. A subject declares that it voluntarily hands over its first-person encapsulation, letting another position decide for it: "does this change the judgement? It does not. The criterion lies in structure, not in professed voluntariness." Two things are separated there: revocable authorization is a person disposing of his own exercise under his own law; the surrender of position that seals revocation and exit is another matter, and "professed voluntariness changes no structural reading" (Rights Theory P5 §III). If the consent signed at hiring is an arrangement that can be taken back at any time, it lies within the limit; if it makes leaving require the other party's consent from then on, the signature does not change this structure. Whether basic rights can be waived by the person himself, as a general question, this series, like Institution Theory, does not answer; at the Base Layer, the sentence above has already given the answer.

The sentences of this section fall differently across the three layers. That accumulation is not converted into control over people's leaving, refusing and revoking is at the normative layer; it does not vary with the size of assets, nor with whether something was signed. The Kingdom State's Wage Payment Act says only that wages are "paid in general currency," and sets no upper limit on the share paid in scrip; the Labor Agency's reply requires written consent: these are ways of writing at the rule layer, and they are examples; whether scrip may exist at all, how large a share it may take, and how many days are given for moving out of a dormitory are for the rule layer to decide, and this paper gives only the criteria. Whether, in the veteran kiln worker's situation that year, the several clauses taken together had already crossed the line is a matter for the judgment layer, and Paper 2 has read it with respect to one and the same person under one and the same actual configuration.

It may be said that all these clauses are things the company gives: without the company's dormitories, the people of the town would have nowhere to live; without advances, those who urgently need money could not borrow it; what kiln scrip is exchanged for is also something real. What it says is true: Bluekiln Porcelain Works pays ten percent more than others in the trade, its apprentices pay no tuition, and the road from the town to the state road was built by it. The first gate does not stop these. Houses can be rented, money can be lent, a shop can be opened; what the gate stops is only the step of converting these into a requirement that others have its consent when they leave. A company that wants back the money it lent demands it under the loan agreement, and if it cannot get it, takes the road of debt collection; it cannot convert a debt into a person's going or staying. Giving something, and using what has been given to hold a person fast, are two different things; this section stops only the latter. If the latter still appears in the name of "giving," it touches what this series calls colonization. Rights Theory P6 writes: "Colonization is a boundary-crossing act disguised as cultivation: it lays an external grid over the subject's own law, redraws an encapsulated dimension, seals off exit or interruption, or reduces the subject to material for another's ends." (Rights Theory P6 F26) Under this narrow sense, where the kind that does not pose as "giving" falls, Institution Theory Paper 1 states plainly: "if effluent presses on the bodies and livelihoods of the household downstream, if a configuration seals off a person's way of refusing, these crossings do not necessarily pose as cultivation, and they all fall within the first two oughts" (Institution Theory Paper 1, §8). Using necessities to dominate people need not first pose as cultivation in order to cross the line; the line it crosses is that of the first two oughts, and only if it still appears in the name of "giving" is it colonization. This sentence speaks of the structure of a kind of arrangement, and does not characterize any particular company; whether these clauses at Bluekiln, taken together, have reached that point is still examined by Paper 2's reading.

The veteran kiln worker's two colleagues, for whom no one put up the money, stayed with the company; the full reasons for their staying are not recorded. The first gate can stop conversions of the kind in Article 14; it does not reach a person's own decision to stay.

E21 Gate One: accumulation is not converted into domination over people's Base Layer: what people need to sustain their lives is not converted into control that compresses their life-substrate (Paper 2, E11, E11a), nor may money, housing or the manner of paying wages be used to place people's ability to leave, refuse or revoke a standing bond under another party's control; neither of these is within ordinary exchange in the Emergence Layer. What counts as necessary to sustain life, and what the Base Layer specifically includes, the list lies at the rule layer (Paper 2, E10). The Base Layer does not rise or fall with assets, and is not exchanged for achievement. That a relation has terms, conditions, procedures or exit costs does not mean that exit has been sealed off; sealing off means that the person no longer holds a structural channel for refusing, revoking a standing bond, or passing into a reviewable procedure (Rights Theory P5 §III). Consent signed at hiring does not by itself change the reading: "The criterion lies in structure, not in professed voluntariness" (ibid.); the general question whether basic rights can be waived by the person himself is not answered. The original says that the construct's own acts cross no line against any internal subject (Institution Theory Z9); here it is extended to the conversions an economic position makes with accumulation (own undertaking).

§2 Accumulation Is Not Converted into Control over Questioning

Back to Sundial.

What a rating agency does is ask a question on others' behalf: can this borrower pay it back? When city governments, companies and small towns issue bonds, those who buy them mostly do not investigate for themselves; they look at the rating. A rating here is a channel of questioning. The making of rules, the adjudication of disputes, the rating of credit and the keeping of accounts of transactions all belong to this kind: many people need not ask about everything in person because there is a place asking for them, and what the asking brings out can be taken and used to make decisions.

What the second gate stops is accumulation converted into control over channels of this kind. Control means a standing: one in which the party questioned can on its own close off questioning of itself, correction, or necessary delivery. This sentence must be written precisely, so that it is not read too broadly. A rating, a set of accounts, an adjudication may have positions particularly responsible for them: someone must always rate, someone record, someone judge. Being responsible for making a record or a judgment is not acquiring the capacity to monopolize its final review, still less being able to cancel questioning of it from outside. A record must have a clear scope and a reviewable provenance; how it is configured is a matter for the rule layer. Under this interpretation a short sentence can be written: the account does not belong to the party whose accounts are kept, nor to any single keeper of accounts. This sentence does not say that a body may not keep a partial account; it says that no keeper of accounts can turn the account it has kept into something others cannot question and cannot correct.

The ground is in Institution Theory. Its fifth ought reads: "An institution keeps open the channel through which it is questioned and corrected, and the channel meets three conditions: it is independent of the party questioned, internal subjects can actually initiate it, and initiating it brings no retaliation. Independence is functional and does not require being outside the institution; what it guards against is the party under review sealing off review of itself on its own" (Institution Theory Z34). The original speaks of an institution's channel for questioning itself; here it is extended to the channels of questioning that economic positions reach (own undertaking). The source of the three conditions is Proposition IV of How Is Institution Possible (Foundational Paper 6; cited as F6): "An institution premised on non dubito must have a chiseling channel satisfying three structural conditions: independent of what is chiseled; actually initiable by members; initiation not met with retaliation." That proposition has its own reach: its subject is an institution premised on non dubito, and "This proposition establishes conditions only; it establishes no office and no ought." (F6 §3.2) Between Sundial and the cities and companies it rates there is no institution premised on non dubito; so the "ought" here is not taken from Proposition IV but from Institution Theory, which itself states that the step "From conditions to an ought" is its own undertaking.

How is one to judge whether a channel has already been controlled? The Theory of Government gave a measure for this within government: "Capture is judged by structure, not by the name of a body or where it sits: a complaints desk set up in the department complained against is not in itself capture; suppose that it is the only door, can close cases itself, and that referral out needs its consent: then the party questioned can seal off review of itself on its own"; the same proposition also says: "Relations of nomination and of party membership do not automatically prove capture either." (Theory of Government G38) The original speaks of channels of questioning within government; here it is extended to economic positions (own undertaking). By this measure, shareholding and fee relations do not in themselves settle the matter: had Sundial no shareholder in common with a party it rates, that would not thereby prove it independent; that it now has such a shareholder does not thereby prove it captured. What must be looked at is the structure. Theory of Government Paper 6 writes it as three things: "who can activate it; what actual efficacy it has on reasons, actions, rules, or the holding of office; how the party questioned cannot cancel it on its own." (Theory of Government Paper 6, §2)

Moral Law P3 met a structure of the same form in the matter of reputation. It lists "Centralized authority articulation of reputation states" as structurally prohibited: "No single body or institution can pronounce, with binding authority, what each Self's reputation is" (Moral Law P3 §1.3); the reason is that "a centralized authority that pronounced on reputation could have its pronouncement gamed by strategic actors targeting that authority" (§8.3), and that distribution is itself the structural defense against optimization-driven capture (§8.4). The scope of that paper is "the pure 15DD community — an idealized case in which all members are 15DD legislative subjects" (§0.6); the protection it speaks of "is structural rather than institutional — it does not depend on any specific institution's enforcement" (§8.3), and rests on the many people in the community each exercising their own discernment. Only one structure is taken from it here: a single centre can be gamed by a party targeting it. Ratings and scores are records, not reputation (Paper 1, E4); in mixed reality, a concentrated record can likewise be gamed by a party targeting it. The step from this to requiring that records have a reviewable provenance and can be questioned from elsewhere is this paper's own: the original speaks of distributed protection within a community that does not rely on institutions; here it is extended to institutional arrangements for records in mixed reality (own undertaking).

Read Sundial by these. The downgrade weighs on Apple City: its financing costs rose, to be paid in the end by Apple City's taxpayers, and those reached by the hospital expansion are among them. What is asked is how the rating decision and its channel of correction are affected by shareholders, fees and procedure. The facts in hand are these: the rating fee is paid by the party rated; the Qiao Family Office holds thirty percent of Sundial's shares, and through its Class B shares sets the direction of Crossways; Crossways Group and Crossways Pay are Sundial's two largest clients, and their rating fees come to about ten percent of its revenue; what Sundial publishes is its rating methodology for small-town bonds; the list of members of the rating committee is not published; the state's registration rules require only registration, and do not review methodology.

By these, the facts point to the following. One of the reasons listed for Apple City's downgrade mentions precisely the bill Apple City had declared its support for; Sundial says it is independent; the council member says the downgrade has to do with Crossways. Both statements are judgments, and both are on the record. Whether a shareholder ever issued an order to downgrade is not recorded; whether the listed reasons are true or false is not recorded either; this paper does not fill this in for either side. But something else can be said: if Apple City wants to question this downgrade, what does it have in hand? Who the members of the rating committee are is not known; whether the rating methodology for city bonds has been published is not recorded; the state's registration does not review methodology; whether there is any place that can review a rating is not recorded. Where the channel of questioning lies has not, in fact, been given. This is a point of entry for questioning, not a capture already proven.

The side of Crossways Pay is read the same way. A month before the crisis, Sundial still gave Crossways Pay the highest rating (§5). The party rated pays the fee, and the controller of the party rated is also a major shareholder of the party that rates: this is precisely the kind of risk in which "the account belongs to the party whose accounts are kept." That a rating was wrong does not in itself prove it controlled; what must be asked is, if such a rating is wrong, who can question it beforehand, whether questioning it can get it corrected, and whether the party rated can, on its own, make this questioning stop.

Echo Wall is the second matter. It is a platform for reviews and sharing: users give scores to, and write reviews of, shops, workshops, and gig workers such as home repairers and nannies; in Year 21, about seventy percent of adults in the Kingdom State used it. From Year 22, several microlenders in Apple City wrote its scores into their lending conditions, lending nothing to shops scored below four stars; a domestic-help agency sends out only nannies with scores of 4.5 or above. A score here is already no longer only a record; it has become a condition of others' decisions, a door (Paper 1, E5; Paper 2, E7). Echo Wall has also done the work of guarding records: in Year 22 it took down a batch of counterfeit Bluekiln porcelain; buyers from elsewhere mostly find Bluekiln's workshops through it.

In the spring of Year 23, Crossways Group acquired Echo Wall. Crossways Group's own brand, Crossways Home, is one of the sellers on Crossways Market; the reviews on its product pages are also recorded on Echo Wall. A statistician in Apple City took a sample in Year 24: the proportion of negative reviews on these product pages deleted on the ground of "violating community norms" rose from three percent before the acquisition to eleven percent.

In the autumn of Year 23, the veteran kiln worker's workshop received more than a hundred one-star reviews within a week, all from newly registered accounts, and its score fell from 4.7 to 3.1. That month the workshop applied to a small-loan company for a loan to expand its kiln, and was refused. The workshop appealed to Echo Wall; three weeks later, Echo Wall deleted sixty of the reviews, saying "the rest do not constitute violations." Appeals are handled by Echo Wall's content review team, which gives no reasons.

Those who bear this are the veteran kiln worker's workshop, and also everyone whose scores are used to decide on lending and assigning work. The record contains the change in the deletion rate before and after the acquisition, and the particular course of this one episode at the workshop; nor is it thereby proven that Echo Wall before the acquisition had every channel independent and every review true.

Put the three questions to the workshop's appeal, and the facts are these. Can it be activated: yes; the workshop appealed. Does it have actual efficacy: in part; sixty reviews were deleted. Can the party questioned close it on its own: the appeal is handled by Echo Wall's own review team, which gives no reasons; whether the rest stay is decided by it alone, and there is no record of another door to go through. Suppose that this is the only door, and that it can close cases itself: then it is the kind the Theory of Government speaks of, in which the party questioned can seal off review of itself on its own. Who wrote those hundred-odd negative reviews is not recorded, and cannot be filled in as instigation by Crossways; nor can the rise in the deletion rate be filled in as every deletion extinguishing questioning. But several things come together: the party that keeps the record was bought by one of the sellers in the field; that seller's negative reviews came to be deleted more often; a workshop could not get a loan after its score was dragged down by a batch of new accounts; and for appeals against scores, the record shows only this one place, which gives no reasons. This is precisely the kind of structure the second gate is to look at.

The sentences of this section also divide into three layers. That accumulation is not converted into control over channels of questioning is at the normative layer. How far rating methodologies are published, how members of a rating committee recuse themselves when they have ties to shareholders, whether appeals against scores must give reasons, and whether there is review outside the platform: these are matters for the rule layer, and can be written in different ways; those mentioned here are all examples, and are not claimed to be the best design. Whether this downgrade was right, and whether that deletion of reviews was right, are left to the positions that make these judgments, and the account falls on those positions.

It may be said that ratings and scores are both opinions in the market: those who disagree can ignore them, or switch to another provider, and the market will itself punish the one that rates inaccurately. The weight of this objection is that rating is by nature something many can do; and Sundial's free publication of its small-town rating methodology really did let several small towns issue bonds for the first time. But "one can switch to another" does not answer the matter being asked about. Sundial holds sixty percent; seventy percent of adults use Echo Wall; scores have already been written into the conditions of lending and assigning work; the downgraded city and the workshop refused a loan cannot, by "switching to another," take back the record that has already fallen on them. The second gate does not require that ratings and scores not exist, nor that there be only one of each, nor that they all be right; it requires only that no position that makes records can, by accumulated shareholdings and client relations, make the questioning of that record pass through itself alone.

Here one thing is left: the list of members of Sundial's rating committee is not published; how far its rating methodology for city bonds is published is not recorded; where Apple City can ask is for the Kingdom State's rules to write.

E22 Gate Two: accumulation is not converted into control over questioning. Making rules, adjudicating, rating and keeping accounts are channels of questioning. Accumulation may not be converted into a standing in which the party questioned can on its own close off questioning of itself, correction, or necessary delivery. Rating, record-keeping and adjudication may have positions particularly responsible for them; making a record or a judgment does not acquire the capacity to monopolize its final review or to cancel outside questioning; a record must have a clear scope and a reviewable provenance, and how this is configured lies at the rule layer. The account does not belong to the party whose accounts are kept, nor to any single keeper of accounts; this sentence does not forbid a single body from keeping a partial account. Independence is examined as "independent of the party questioned," and is functional (Institution Theory Z34); capture is judged by structure, not by the name of a body or where it sits, and relations such as nomination and party membership do not automatically prove capture (Theory of Government G38): the absence of cross-shareholding does not automatically prove independence, nor does the presence of cross-shareholding automatically prove capture. The source of the three conditions is F6 Proposition IV, which establishes only conditions, not an ought; the ought rests on Institution Theory Z34. Of the prohibition of centralized articulation and the distributed defense in Moral Law P3 (§1.3, §8.3, §8.4), only the structure "a single centre can be gamed" is taken, as the same form; that paper stands within the pure 15DD community as an idealized case (§0.6), and the protection it speaks of is structural, not dependent on institutions (§8.3); extending it to the reviewable provenance of records in mixed reality is an own undertaking. Points of entry for risk and for questioning are not treated as proven capture. The original speaks of an institution's own channels of questioning and of capture within government; here it is extended to the channels of questioning that economic positions reach (own undertaking).

§3 Political–Economic Separation, in Both Directions

The draft Online Marketplaces Act lay before the Economic Committee of the State Legislature of the Kingdom State, which has fifteen members. From Year 23 to Year 24, Crossways Group and its executives made campaign donations to nine of them, all within the limits of the Kingdom State's Campaign Finance Act; the Qiao Foundation donated and built libraries for schools in the districts of four committee members. In the same year Crossways voluntarily disclosed all its political donations. In the autumn of Year 24, the committee, by eight votes to seven, struck out the clause on requesting the Office of Trade to initiate a breakup, and kept the two clauses on publishing ranking rules and on setting up an independent appeals body.

In the winter of the same year, the head of the Competition Review Division left office. He had led the competition review of Crossways Market: the review began in Year 23, and in the spring of Year 24 it issued a report recommending that Crossways Market and Crossways Home be placed in two separate companies, each operating on its own (Paper 3). Four months after leaving office, he became a vice president of Crossways Group, in charge of "Compliance and Government Relations." The Regulations on Departing Public Officials of the United City-States of Ends (UCSE) require that, for one year after leaving office, a person have no contact with his former body on matters he was in charge of; they do not restrict employment. These laws and regulations are likewise examples set up by this series.

Money went into the legislature; a person went out of a public body. Are these two things one and the same?

This paper's answer is that they are two directions of one and the same boundary. Money does not buy that share of political power which reviews and questions it; decisions made in a position are not cashed in, either, as wealth later obtained from the party the position questioned. The first direction is familiar: a company may not use money to buy away the vote that reviews it. The second is easily missed: a person who reviewed a company in a position may not turn the decisions he made in that position into a reward later obtained from that company. Both are conversions, and what both exchange is the road of questioning: the first uses money to obtain the road by which it is questioned; the second uses the road of questioning to obtain money. So this is the second gate in its strongest form, its extreme case, at the level of public power: what reviews and questions it are public positions such as legislation and review. Every level of public power is the same, from a whole state down to the owners' self-government of a residential complex. It is also the reasoning of the separation of powers and their mutual checks carried into economic positions: only where economic power and political power stand apart can they check each other.

How the second direction is taken into the second gate must be seen from the side of the party that pays. The subject of the second gate is accumulation. If donations, promised rewards or employment after the fact are actually used to buy the positions that review it, adjudicate it or correct it, then the paying party is using accumulation to obtain control over questioning; reward after the fact belongs here too because it can be anticipated before the decision, and the decision in the position thereby acquires a later price. Seen from the side of the party that receives the money, it is the use of granted power being exchanged for a private reward; that side belongs to the Theory of Government, and capture is judged by structure (Theory of Government G38); the two places point to each other, and each writes its own part. Neither donations, nor employment, nor their sequence in time proves this conversion by itself alone. This proposition, too, has its reach. "Money does not buy political power" here means not buying that share of political power which reviews and questions it; using money to buy a law that opens a door for oneself, keeps competitors out, or obtains an exclusive franchise is read under chokepoints and checks (Paper 2, E7; Paper 3), not under this proposition.

Smith met this point long ago. At the end of Book I of The Wealth of Nations, speaking of new laws or regulations of commerce proposed by the order of dealers, he says: "The proposal of any new law or regulation of commerce which comes from this order, ought always to be listened to with great precaution, and ought never to be adopted till after having been long and carefully examined, not only with the most scrupulous, but with the most suspicious attention" (Wealth of Nations, Book I, ch. XI, Conclusion of the Chapter, per Project Gutenberg #3300; tier A). The point of contact is this: legislative proposals from dealers call for suspicious scrutiny. In the present case, Crossways is the party targeted by the proposed law. The original concerns proposals put forward by dealers; this paper extends the point to the influence a party targeted by a law may exert on those reviewing it (own undertaking). The text just before the original sentence speaks of narrowing competition, and that side this series reads under chokepoints and checks (Paper 2, E7; Paper 3). Smith's measure is the interest of the public, his reason being that the interest of that order differs from the interest of the public; this paper's measure lies in structure: what is to be examined is whether the road of questioning has been bought, and it does not judge where the interest of any order lies.

What is to be examined is whether the road by which it is questioned has been bought. What most needs guarding against here is letting one thing stand in for the verdict on another. A lawful donation does not, by being lawful, prove that nothing was bought; that there was a donation does not, by that fact, prove that something has been bought. Employment after leaving office is the same: that the employment complies with the rules does not prove that the position was not traded; that there was employment does not prove that it was. The Theory of Government says: "Relations of nomination and of party membership do not automatically prove capture either." (Theory of Government G38) The original speaks of nomination and party membership within government; here it is extended to donations and employment (own undertaking). The judgment lies in structure and sequence: from whom the money came, into whose hands it went, before which decision; from which position the person left, what he was in charge of in that position, and whether after leaving he has had contact with the matters he was formerly in charge of.

Read the matter of the State Legislature by this measure. Those fifteen committee members were elected by the voters of the Kingdom State; with the breakup clause struck out, the sellers concerned in the matter under review have one road fewer before them. What the record contains is this: nine members received donations, four districts got libraries, the vote was eight to seven, the donations were all within the limits, and they have been disclosed. These must be recorded separately: whether the nine who received donations and the eight who voted in favor are the same people is not recorded; whether there were prior promises is not recorded; how the four members whose districts got libraries voted is not recorded either. So the facts point only this far: the party at which the law is aimed spent money on those who review the law, and disclosed it; those who review the law struck out its clause requesting the initiation of a breakup. This is a point of entry for questioning: voters can question their members, the question can be put at the next election, and the limits of the Campaign Finance Act can also be questioned. It is not a purchase already proven. What the vote would have been without these donations is a supposition this paper does not make.

The division head's employment must be read separately. The review he led produced a report recommending that the marketplace and the self-operated business be separated, a recommendation unfavorable to Crossways; he left office after the clause was struck out, and was employed four months after leaving. So his employment cannot be filled in as a factor before the striking out of the clause. What is to be asked about is the second direction: in his position he had charge of matters concerning Crossways, and now he is at Crossways in charge of "Government Relations." The departure regulations govern a period of one year, during which he may not have contact with his former body on matters he was in charge of; whether his new post will touch those matters within the year, and whether there has been any contact, is not recorded, and cannot be filled in as contact already made during the prohibited period. That the regulations do not restrict employment is one way of writing at the rule layer; whether to restrict it, and for how long, is for the rule layer to decide. If the rule layer is to restrict him, for a period, from being employed by the party he reviewed, that provision protects the channels through which the public can question that party; it follows the same reasoning as the constraints Paper 2 speaks of, set to protect others and not among non-compete agreements (Paper 2, E12a). Whether it holds requires stating which of these channels it protects, how this employment would bring that channel under constraint, why recusal or restricting contact is not enough, and how wide its scope is; it does not hold merely because it is called a cooling-off period. The normative layer says only one thing: decisions in a position may not be the price of a later reward; and the decisions a position has made remain open to questioning as before, and the questioning does not stop because the person who made them has left.

That Crossways voluntarily disclosed all its political donations must also be recorded. Disclosure is itself an act that makes questioning possible: that nine people received donations, the voters know because it disclosed them. This side holds at the same time as the side above, and neither cancels the other.

It may be said that donating is political participation and taking employment is the freedom to choose one's occupation; restricting them amounts to restricting people's participation in politics and where they may work. The weight of this objection is that what it guards is real: citizens and companies alike may support candidates they agree with; and a person who leaves public office must have somewhere to go. The second gate does not prohibit donation, nor does it prohibit employment; donations within the limits and lawful employment are left open as usual in the Emergence Layer (§4). What it stops are only two things: buying away the road by which one is questioned, and exchanging the road of questioning for a reward. How these two are recognized depends on structure and sequence, and lies at the judgment layer; the limits on donations, the cooling-off period after leaving office and the scope of recusal lie at the rule layer, and can be written in different ways.

After the committee struck out the breakup clause, the road by which the state government could request the initiation of a breakup was gone; whether the Office of Trade will still proceed under its own procedures is not recorded. Breakup as a means of keeping positions from being final is in Paper 3 (Paper 3, E18).

E23 Political–economic separation, in both directions, is the extreme case of the second gate at the level of public power. Money does not buy that share of political power which reviews and questions it; decisions made in a position are not cashed in, either, as wealth later obtained from the party the position questioned. What is exchanged in both directions is the road of questioning; donations before the fact and employment after it are the two sequences this case shows; whether a conversion has occurred must still be examined, and the sequence alone does not prove it. The use of granted power on the side receiving money is read under Theory of Government G38; the two places point to each other, and each writes its own part. Every level of public power is the same, from a whole state down to the owners' self-government of a residential complex; this is the reasoning of the separation of powers and their mutual checks carried into economic positions: only where economic power and political power stand apart can they check each other. What is to be examined is whether the road by which it is questioned has been bought, or whether the road of questioning has been exchanged for a reward. Buying the road by which one is questioned, and ordinary employment or donation, do not stand in for each other's verdict: a lawful donation does not thereby prove that nothing was bought, nor does a donation thereby prove that something has been; employment after leaving office likewise. The judgment lies in structure and sequence, not in the relation itself: relations such as nomination and party membership do not automatically prove capture (Theory of Government G38; the original speaks of nomination and party membership within government; here it is extended to donations and employment; own undertaking). Buying a law to open a door for oneself is read under chokepoints and checks (Paper 2, E7; Paper 3), not under this proposition.

§4 Outside the Gates, Exchange in the Emergence Layer Is Left Open as Usual

Both gates have now been written; one must turn back and guard against a misunderstanding: once gates have been set, is every conversion suspect?

In Year 22, a gallery in Apple City held the "New Bluekiln Porcelain" exhibition, and the tickets were printed with the name of the third-generation heir of the Qiao family, as "co-curator." The third-generation heir was twenty-four that year, a graduate of Oakgrove College, with nine hundred thousand followers on Echo Wall, often posting about travel and wine tasting. In the first week of the exhibition, attendance was four times that of a similar exhibition the year before. The veteran kiln worker had two pieces in the second hall. He had thrown pots and fired kilns at Bluekiln Porcelain Works for thirty-one years; the old masters in town and several of the old porcelain shops of Apple City all said "his glazes are steady," but outside the town few people knew of him. Elsewhere, Crossways Market gives some sellers the "Gold Seller" badge, assessed by sales volume and by the amount spent on paid promotion.

Here there are the shapes of several conversions. The tickets carry the name of a person with nine hundred thousand followers on Echo Wall, as co-curator; in the first week, attendance was four times that of a similar exhibition the year before. The name, the position and the numbers appear together; how they are connected the record does not say; yet this is already the common shape of fame bringing opportunities. Two pieces by the veteran kiln worker were also in this exhibition, in the second hall. A seller's money obtains paid promotion, and the amount of paid promotion is in turn one of the grounds on which the "Gold" badge is assessed. On this series' reading, these are all exchanges in the Emergence Layer: fame bringing opportunities, capital bringing more capital, qualifications bringing posts, none of them in itself the problem (Paper 1, E1). This series does not aim at reducing them, nor does it seal off exchange in the Emergence Layer on the ground that "they may be converted."

One point must be written precisely. The "Gold" badge is obtained with money, but what money obtains is a badge, a record, not reputation. In the sense of Moral Law P3, reputation cannot be bought: "Reputation, in SAE, is not a kind of capital"; "It is not exchangeable with any 14DD primary medium" (Moral Law P3 §2.2, §1.1; the conclusions there stand within the pure 15DD community as an idealized case, and which of them this series takes in mixed reality is set out in Paper 1, E4). The veteran kiln worker's standing for "steady glazes" no one can buy; Crossways' badge a seller can obtain by spending money. With the two kept apart, money buying a badge is not money buying reputation. If a badge is taken by others as a condition of lending or assigning work, it becomes a door, and that is a matter read in Papers 1 and 2 (Paper 1, E5; Paper 2, E7); here there is no such record.

So the two gates speak only of the restrictions on conversion this paper handles. Accumulation is not converted into domination over people's Base Layer, nor into control over questioning; that a transaction has not run into these two gates shows only that it is not stopped at these two places, and does not certify it as warranted in every respect. It must still keep the requirements of this series elsewhere: not crossing the line, not sealing off roads (Paper 2); being open to questioning, with checks increasing with the power-face (Paper 3); for what does the work of the Base Layer, securing beyond checks (Paper 3, E14a); giving conditions for people's emergence without deciding their direction for them (Paper 6). The donations and employment of the previous section belong here too: donations within the limits and lawful employment are in themselves ordinary matters in the Emergence Layer; only if they exchange away the road by which one is questioned do they enter the second gate.

This section can be compared with two authors.

Walzer, in Spheres of Justice (1983), asked the same thing: what money ought not to buy. As summarized by G. den Hartogh, Walzer sets out in the book a list of what he takes to be regarded in Western societies as falling outside the sphere of money, namely "blocked exchanges": political power and office; personal liberty; the freedoms of speech, press, religion and assembly; the rights to emigrate and to leave; exemption from communally imposed duties; love, in particular love of a particular person; and so on; among them, the prohibition of desperate exchanges is explained by the principle of need (paraphrased from den Hartogh 1999, which draws on pp. 100–103 of the original; original not consulted; tier C). According to the summary, his ground lies in the distributive principles of each sphere. The question is the same question: where what has been accumulated may not be exchanged. The original has not been consulted, and this paper only lists the points of comparison. There are three: political power and office on the list correspond to the direction of §3 of this paper; desperate exchanges correspond to what the first gate stops, people being driven by their situation to hand over their way of leaving; and the grounds: the grounds of this paper's two gates lie in structure, one in people's Base Layer and one in channels of questioning, while his, according to the summary, lie in spheres and distributive principles. Whether at these points the two meet or part, and at which step in each case, will be written only once the original has been consulted. Nor does this paper list everything money cannot buy; the two gates are only the two restrictions on conversion this paper handles, and the other requirements are elsewhere in this series.

Sandel, in What Money Can't Buy (2012), asks whether there are some things money ought not to buy. As summarized in D. M. Kennedy's review, he gives two objections. One concerns fairness: under conditions of inequality, being willing to buy and being able to afford are not the same thing, and "free choice" in the market may not be free. The other concerns corruption: putting a thing on the market can dissolve the value it had (paraphrased from Kennedy 2012; original not consulted; tier C). The original has not been consulted, and what follows only lists the points of comparison. The first objection meets the same kind of situation as "a signature does not change this reading either" in §1: when a situation presses on a person, what does consent in a transaction show? The second objection asks about the meaning of a thing; this paper's two gates do not take the meaning of things as their ground. Under what circumstances the market changes the meaning of a thing, and with which of this series' boundaries this question connects, can be written only once the original has been consulted.

From the other direction there is a worry too: leaving everything outside the gates open amounts to letting pass the inequality that accumulation produces elsewhere; a person with nine hundred thousand followers is joined to a curator's position, while a man who has fired kilns for thirty-one years is unknown outside his town. The difference this worry sees is real. But this series does not take distributive outcomes as its endpoint, nor does it aim at reducing accumulation (the Prequel); what it asks is whether this difference has been converted into domination over others' Base Layer, whether it has been converted into control over questioning, and whether it crosses the line or seals off roads elsewhere. Two pieces by the veteran kiln worker were also in that exhibition; on the record of this occasion, nowhere does a conversion of fame appear to have closed his road. If one day a person's followers can decide whose work enters the exhibition hall and who gets a loan, then the question will be one not for this section but for Papers 1 and 2.

Here one thing is left: the "Gold" badge is assessed by sales volume and paid promotion; whether buyers know that part of it is bought is not recorded; how a badge marks its provenance is a matter for the rule layer.

E24 Outside the gates, exchange in the Emergence Layer is left open as usual. Fame bringing opportunities, capital bringing more capital, qualifications bringing posts, donations within the limits and lawful employment are not, in themselves, the problem; exchange in the Emergence Layer is not sealed off on the ground that it "may be converted." Money can buy badges and records; it cannot buy reputation (Moral Law P3 §1.1, §2.2; see Paper 1, E4). The two gates speak only of the restrictions on conversion this paper handles: not running into the two gates does not certify a transaction as warranted in every respect, nor does it cancel this series' requirements elsewhere of guarding the line, being open to questioning, securing, and giving conditions.

§5 Public Support Is Not an Endorsement

Crossways Pay is the settlement subsidiary of Crossways Group. It settles payments for buyers and sellers on Crossways Market; from Year 25 it has also collected payments for small merchants in the towns of the Kingdom State. By Year 26, the payments of about forty percent of the state's small merchants passed through it.

In the summer of Year 26, a batch of Crossways Pay's short-term investments made losses, and settlements were delayed by three days. Two workshops in Bluekiln could not pay wages that month. The State Financial Agency and the State Office of Finance gave Crossways Pay an emergency loan at an interest rate below the market rate, on condition that it pay no dividends to its parent company for three years. A spokesperson for the state government said: "If it falls, what falls is forty percent of the small merchants." A month before the crisis, Sundial's rating of Crossways Pay was still the highest rating.

The small merchants who collect payments through Crossways Pay, the workers of those two workshops who could not get their wages, and the taxpayers of the Kingdom State are all in this matter: the loan uses taxpayers' money. Two things are asked below: what the ground of this loan is; and what it gave Crossways Pay, and what it did not give. Who in the end bears the losses on those investments is not recorded; whether the forty percent of merchants would necessarily have fallen is not recorded either.

First, how Crossways Pay was able to obtain this support. Its position was accumulated: a subsidiary of a group, collecting payments for others one transaction at a time, until it reached forty percent of the state's small merchants. The payments of about forty percent of the state's small merchants pass through it; whether they could switch at short notice, and what switching to another provider would cost, is not recorded. But when many people's payments are concentrated in one place, the consequences of its falling do not land on it alone. Rights Theory P4, speaking of how exit shows a power-face, says: "Only when structural dependence is present does Exit Rights evolve into power through F15." (Rights Theory P4 §V) The subject term of the original sentence is an exit, and what is said is that it shows a power-face; here two things are changed: the subject term is changed to a position's possible failure, and what is said is that it thereby obtains public support (own undertaking). Crossways Pay did not use "I may fall" to apply pressure; the state gave the loan, by the spokesperson's words, for the sake of the forty percent of small merchants it reaches. Structural dependence here is not a threat but a situation; the public support was given within this situation.

Next, what this support does not give. The Theory of Government, on a government's guarantee of the rights it embeds, wrote: "A guarantee is not an endorsement. Neither constitution nor operation succeeding is the same as normative legitimacy"; the same proposition also says that "long existence is not in itself a reason for continuation (own undertaking)" (Theory of Government G22). The original speaks of a government's guarantee of the recognition-based rights it embeds; here it is extended to public support of an accumulated position (own undertaking). An emergency loan was given; how it connects with the recovery after the delay is not recorded. It does not thereby prove that those short-term investments were sound, does not prove that Crossways Pay's earlier configuration was without problems, and does not prove that it ought from now on to go on existing in this way. Obtaining public support does not issue legitimacy to its existing acts and its continued existence.

The word "guarantee" must be used precisely here. A loan with conditions attached is not the state's setting up of a new general guarantee. The loan has an interest rate, conditions and a term, and it was made by two particular bodies; it is not a promise that "from now on, whatever happens, it will be covered." Whether, after this, anyone came to think that the state would step in again next time, and so more confidently placed their payment collection with Crossways Pay, or whether Crossways Pay itself more confidently made short-term investments, is not recorded. This question is real; it is left here, and is not filled in as something that has already happened.

What, then, must the party that gives the support do? Institution Theory's fifth ought contains a sentence: "An institution bears the burden of explanation for what it has itself produced, and must answer the very matter it is questioned about." (Institution Theory Z34) This loan, made by the State Financial Agency and the State Office of Finance, is something they have produced. They must explain its ground, authority, object and cost: on what basis it was lent, who had the authority to lend it, whom lending to it was meant to protect, and on whom the cost falls. The spokesperson's line "what falls is forty percent of the small merchants" speaks of the object: what this money is to protect is not Crossways Pay but those who collect payments through it. By this account, the arrangements of support must answer to this object. The three-year ban on dividends stops money flowing from the subsidiary back to the parent company; whether this loan reached the wages the two workshops could not pay is not recorded; whether the bodies that made the loan explained this point is not recorded either. This is one point the arrangements of support must answer to. A rescue as such is a grant, and it is the government that makes it; how it is made, and what questioning it is subject to, belongs to the Theory of Government; the half extended to accumulated positions is this series': the position that receives support is questioned by its particular conditions, the accountability it bears and its power-face, and is not questioned one bit less for having received support (own undertaking).

This must be kept apart from a proposition of Paper 3. Paper 3 says that where what a construct does falls in the Base Layer, the institution must, beyond checks, also secure this supply (Paper 3, E14a); there the ground lies in that supply itself: people depend on it to sustain their lives. The ground here is different: whether settlement is a supply in the Base Layer this paper does not decide; Crossways Pay obtained support, by the spokesperson's account, for the sake of the forty percent of small merchants it reaches: the ground lies in scale and dependence, not in whether settlement as such is in the Base Layer. Public support obtained by scale itself needs a ground of its own, and cannot borrow the ground of securing the Base Layer. The two can fall on the same matter: the workers of the two workshops who could not get their wages are touched in their livelihood; but then what is to be secured is wages and livelihood, not that settlement company.

Sundial's side connects here with §2. A month before the crisis, Crossways Pay was given the highest rating. The highest rating is not a guarantee of no risk; this rating and the delay that followed do not thereby prove the rating wrong, still less that it was controlled. But there is a structure here to be asked about: if a channel of questioning cannot itself stand up to questioning, a risk that was not questioned beforehand may afterwards be taken up by public support; on whom the final cost of this loan falls is not recorded. The second gate and this section are, here, the two ends of one and the same matter.

The sentences of this section also divide into three layers. That support is not an endorsement, that the party giving support must explain, and that the party receiving it is to be questioned, are at the normative layer. The interest rate, conditions and term of a loan, and whether the condition is a ban on dividends or something else, are at the rule layer, and those mentioned here are all examples. Whether to lend, how much and when are left to the positions that make this judgment, and the account falls there.

There is one more concern: when matters are urgent, a full explanation cannot be demanded first; in the three days of delay, the workers of the two workshops had already gone without their wages. To support, and to explain, and not to endorse: might the state hesitate on this account and spoil matters instead? That the cost of delay falls on particular people and cannot wait is true. But not endorsing is not the same as not supporting, nor need the explanation be completed before the support is given. Support can be given first; the explanation comes after it has been given, answering the very matter it is questioned about. What is to be guarded against is only one thing: an emergency loan being read afterwards as approval of that company, or as a promise to cover everything from then on.

Two things are left undecided in this section: who in the end bears the losses on those investments; and whether this loan reached the two workshops' wages.

E25 Public support is not an endorsement. Structural dependence can bring an economic position public support (Rights Theory P4 §V reads structural dependence; the original says that where structural dependence is present exit shows a power-face; here the subject term is changed to a position's failure, and what is said is that it obtains public support; own undertaking); obtaining support does not issue legitimacy to its existing acts and its continued existence (Theory of Government G22, "A guarantee is not an endorsement"; the original speaks of a government's guarantee of embedded rights; here it is extended to public support of accumulated positions; own undertaking). The public position that gives the support must state its grounds, authority, object and cost, and must answer the very matter it is questioned about (Institution Theory Z34); the position that receives the support is questioned by its particular conditions, the accountability it bears and its power-face. This is kept apart from Paper 3, E14a: securing a supply in the Base Layer has its ground in that supply; public support obtained by scale itself needs a ground of its own. A loan with conditions attached is not the setting up of a new general guarantee. A rescue as such is a grant, and belongs to the Theory of Government; the half extended to accumulated positions is this series' (own undertaking).

Where This Paper Stops

This paper has proved these things. Accumulation is not converted into domination over people's Base Layer: what people need to sustain their lives is not converted into control that compresses their life-substrate, nor may money, housing or the manner of paying wages be used to place people's ability to leave, refuse or revoke a standing bond under another party's control, and a signature does not change this reading (E21). Accumulation is not converted into control over questioning: channels of questioning may have responsible positions, and are not closed by the party questioned on its own; the account does not belong to the party whose accounts are kept, nor to any single keeper of accounts; the judgment lies in structure, not in shareholding and fee relations as such (E22). Political–economic separation runs in both directions, and is the extreme case of the second gate at the level of public power: money does not buy that share of political power which reviews and questions it, and decisions in a position are not cashed in, either, as wealth later obtained from the party the position questioned; donation, employment and purchase do not stand in for each other's verdict (E23). Outside the two gates, exchange in the Emergence Layer is left open as usual; not running into the two gates does not certify the rest (E24). Public support obtained through structural dependence is not an endorsement; the party that gives the support must explain, and the party that receives it is questioned as usual (E25).

What it has not proved: the list of what counts as necessary to sustain life; the detailed rules of campaign finance and of departure from office; the design of the supervision of rating agencies; the conditions of rescue; tax. These are matters for the rule layer and for specific theories of institutions; this paper gives only criteria, and does not decide in their place. Sundial's downgrade, Echo Wall's deletion of reviews, the State Legislature's vote, the division head's employment: in each, this paper writes only as far as where the facts point, and does not make the final judgment in place of the positions that make it.

Handed to the later papers: accumulation grows by itself, and how growth as such is read is in Paper 5; the rate of return set as the sole goal is a signal to be examined (Paper 5, E27), and control arrangements must meet the three conditions of self-chiseling (Paper 5, E30).

The remainders left: the reasons the veteran kiln worker's two colleagues stayed; the list of members of Sundial's rating committee, and how far the methodology for city bonds is published; whether there is any other door besides Echo Wall's appeals; the nine who received donations and the eight votes; whether the division head's new post will touch the old matters; how the "Gold" badge marks its provenance; who bears Crossways Pay's losses, and the two workshops' wages. All of these are recorded, and each has its place to go.

The Five Propositions of This Paper

E21 Gate One: accumulation is not converted into domination over people's Base Layer: what people need to sustain their lives is not converted into control that compresses their life-substrate (Paper 2, E11, E11a), nor may money, housing or the manner of paying wages be used to place people's ability to leave, refuse or revoke a standing bond under another party's control; neither of these is within ordinary exchange in the Emergence Layer. What counts as necessary to sustain life, and what the Base Layer specifically includes, the list lies at the rule layer (Paper 2, E10). The Base Layer does not rise or fall with assets, and is not exchanged for achievement. That a relation has terms, conditions, procedures or exit costs does not mean that exit has been sealed off; sealing off means that the person no longer holds a structural channel for refusing, revoking a standing bond, or passing into a reviewable procedure (Rights Theory P5 §III). Consent signed at hiring does not by itself change the reading: "The criterion lies in structure, not in professed voluntariness" (ibid.); the general question whether basic rights can be waived by the person himself is not answered. The original says that the construct's own acts cross no line against any internal subject (Institution Theory Z9); here it is extended to the conversions an economic position makes with accumulation (own undertaking).

E22 Gate Two: accumulation is not converted into control over questioning. Making rules, adjudicating, rating and keeping accounts are channels of questioning. Accumulation may not be converted into a standing in which the party questioned can on its own close off questioning of itself, correction, or necessary delivery. Rating, record-keeping and adjudication may have positions particularly responsible for them; making a record or a judgment does not acquire the capacity to monopolize its final review or to cancel outside questioning; a record must have a clear scope and a reviewable provenance, and how this is configured lies at the rule layer. The account does not belong to the party whose accounts are kept, nor to any single keeper of accounts; this sentence does not forbid a single body from keeping a partial account. Independence is examined as "independent of the party questioned," and is functional (Institution Theory Z34); capture is judged by structure, not by the name of a body or where it sits, and relations such as nomination and party membership do not automatically prove capture (Theory of Government G38): the absence of cross-shareholding does not automatically prove independence, nor does the presence of cross-shareholding automatically prove capture. The source of the three conditions is F6 Proposition IV, which establishes only conditions, not an ought; the ought rests on Institution Theory Z34. Of the prohibition of centralized articulation and the distributed defense in Moral Law P3 (§1.3, §8.3, §8.4), only the structure "a single centre can be gamed" is taken, as the same form; that paper stands within the pure 15DD community as an idealized case (§0.6), and the protection it speaks of is structural, not dependent on institutions (§8.3); extending it to the reviewable provenance of records in mixed reality is an own undertaking. Points of entry for risk and for questioning are not treated as proven capture. The original speaks of an institution's own channels of questioning and of capture within government; here it is extended to the channels of questioning that economic positions reach (own undertaking).

E23 Political–economic separation, in both directions, is the extreme case of the second gate at the level of public power. Money does not buy that share of political power which reviews and questions it; decisions made in a position are not cashed in, either, as wealth later obtained from the party the position questioned. What is exchanged in both directions is the road of questioning; donations before the fact and employment after it are the two sequences this case shows; whether a conversion has occurred must still be examined, and the sequence alone does not prove it. The use of granted power on the side receiving money is read under Theory of Government G38; the two places point to each other, and each writes its own part. Every level of public power is the same, from a whole state down to the owners' self-government of a residential complex; this is the reasoning of the separation of powers and their mutual checks carried into economic positions: only where economic power and political power stand apart can they check each other. What is to be examined is whether the road by which it is questioned has been bought, or whether the road of questioning has been exchanged for a reward. Buying the road by which one is questioned, and ordinary employment or donation, do not stand in for each other's verdict: a lawful donation does not thereby prove that nothing was bought, nor does a donation thereby prove that something has been; employment after leaving office likewise. The judgment lies in structure and sequence, not in the relation itself: relations such as nomination and party membership do not automatically prove capture (Theory of Government G38; the original speaks of nomination and party membership within government; here it is extended to donations and employment; own undertaking). Buying a law to open a door for oneself is read under chokepoints and checks (Paper 2, E7; Paper 3), not under this proposition.

E24 Outside the gates, exchange in the Emergence Layer is left open as usual. Fame bringing opportunities, capital bringing more capital, qualifications bringing posts, donations within the limits and lawful employment are not, in themselves, the problem; exchange in the Emergence Layer is not sealed off on the ground that it "may be converted." Money can buy badges and records; it cannot buy reputation (Moral Law P3 §1.1, §2.2; see Paper 1, E4). The two gates speak only of the restrictions on conversion this paper handles: not running into the two gates does not certify a transaction as warranted in every respect, nor does it cancel this series' requirements elsewhere of guarding the line, being open to questioning, securing, and giving conditions.

E25 Public support is not an endorsement. Structural dependence can bring an economic position public support (Rights Theory P4 §V reads structural dependence; the original says that where structural dependence is present exit shows a power-face; here the subject term is changed to a position's failure, and what is said is that it obtains public support; own undertaking); obtaining support does not issue legitimacy to its existing acts and its continued existence (Theory of Government G22, "A guarantee is not an endorsement"; the original speaks of a government's guarantee of embedded rights; here it is extended to public support of accumulated positions; own undertaking). The public position that gives the support must state its grounds, authority, object and cost, and must answer the very matter it is questioned about (Institution Theory Z34); the position that receives the support is questioned by its particular conditions, the accountability it bears and its power-face. This is kept apart from Paper 3, E14a: securing a supply in the Base Layer has its ground in that supply; public support obtained by scale itself needs a ground of its own. A loan with conditions attached is not the setting up of a new general guarantee. A rescue as such is a grant, and belongs to the Theory of Government; the half extended to accumulated positions is this series' (own undertaking).

References

All SAE works below are cited by concept DOI, which always resolves to the latest version. The version records used for sentence-by-sentence verification are given in the accompanying audit table.

I · This system

Qin, H. (2026). How Is Institution Possible: From Inter-Ontological Remainder to Co-Constructive Framework (SAE Foundational Paper 6) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.19328662

Qin, H. (2026). The Interface and Transformation between Recognition-Based Rights and Power (SAE Rights Theory, Paper 4) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.21093939

Qin, H. (2026). On the Limitation and Violation of Basic Rights (SAE Rights Theory, Paper 5) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.21212595

Qin, H. (2026). On the Cultivation and Diachronic Unfolding of Rights (SAE Rights Theory, Paper 6) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.21252863

Qin, H. (2026). The Reputation Economy · Dao (SAE Moral Law, Paper 3) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.20046168

Qin, H. (2026). To Whom the Ought Is Addressed (SAE Institution Theory, Paper 1) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.23269635

Qin, H. (2026). Not Crossing the Line (SAE Institution Theory, Paper 2) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.23270384

Qin, H. (2026). Open to Questioning (SAE Institution Theory, Paper 6) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.23271324

Qin, H. (2026). Guarantee (SAE Theory of Government, Paper 3) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.23413112

Qin, H. (2026). Questioning and Bearing (SAE Theory of Government, Paper 6) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.23413511

II · External works (originals, tier A)

Smith, A. (1776). An Inquiry into the Nature and Causes of the Wealth of Nations. Project Gutenberg, EBook #3300. https://www.gutenberg.org/ebooks/3300 (Book I, ch. XI, Conclusion of the Chapter)

III · External works (originals not consulted; per secondary summaries, tier C)

Walzer, M. (1983). Spheres of Justice: A Defense of Pluralism and Equality. New York: Basic Books. (original not consulted; paraphrased from den Hartogh 1999)

den Hartogh, G. (1999). The Architectonic of Michael Walzer's Theory of Justice. Political Theory, 27, 491–522.

Sandel, M. J. (2012). What Money Can't Buy: The Moral Limits of Markets. New York: Farrar, Straus and Giroux. (original not consulted; paraphrased from Kennedy 2012)

Kennedy, D. M. (2012). The Root of All…: Michael Sandel's salvos against the conventional wisdom of what markets should—and should not—do. Democracy: A Journal of Ideas, No. 26 (Fall 2012). https://democracyjournal.org/magazine/26/the-root-of-all/

Interfaces

Main upstream: Institution Theory Z9 (extended to the conversions an economic position makes with accumulation; own undertaking), Z34 (extended to the channels of questioning that economic positions reach; the burden of explanation); Theory of Government G38 (extended to economic positions, and to donations and employment; own undertaking; the receiving side in the second direction of political–economic separation is read under it), G22 (extended to public support of accumulated positions; own undertaking); Rights Theory P5 §III.

Also used: F6 §3.2, Proposition IV (the source of the three conditions, its reach as the original sentence states it); Theory of Government Paper 6, §2 (the three questions); Rights Theory P4 §V (structural dependence, extended to what a failure reaches; own undertaking); Rights Theory P6 F26 (the narrow sense of colonization); Institution Theory Paper 1, §8 (crossings that do not pose as cultivation also fall within the first two oughts); Moral Law P3, the signature claim, §0.6, §1.1, §1.3, §2.2, §8.3, §8.4 (the same form; an idealized case; extended to the reviewable provenance of records in mixed reality; own undertaking).

Interfacing: Smith (The Wealth of Nations, Book I, ch. XI, Conclusion of the Chapter, legislative proposals from dealers; tier A, per Project Gutenberg #3300; §3); Walzer (Spheres of Justice, ch. 4, "blocked exchanges"; tier C, per den Hartogh 1999); Sandel (What Money Can't Buy; tier C, per Kennedy's 2012 review).

This series: the Prequel (not aiming at reduction; not taking distributive outcomes as the endpoint; the three layers); Paper 1, E1, E4, E5; Paper 2, E7, E10, E11, E11a, E12, E12a; Paper 3, E14a, E18; Paper 5, E27, E30; Paper 6.

Handed to Paper 5: accumulation grows by itself; how growth as such is read.

写作声明：本文与 Claude（Anthropic）共同起草，所有思想决策、框架设计和最终编辑判断由作者做出。

引言

第 24 年秋，日晷评级（Sundial Ratings）把苹果城（Apple City）的市政债下调了一级。

苹果城是王国州（the Kingdom State）的一座城。两年前，它发行市政债，为市立医院扩建筹款。日晷是王国州最大的信用评级机构，市场份额约六成；评级费由被评的一方付。下调的公告列了几条理由，其中一条是："拟议中的线上市集新规可能削减税基。"那一年，苹果城市议会刚刚表态支持州议会正在审议的《线上市集法》草案。草案要求线上市集公开排序规则，设独立于市集的卖家申诉机构；另有一款，许州政府请求商贸署依竞争审查处的报告启动拆分程序。下调以后，苹果城再借钱，成本上升了。

一位市议员公开说，这次下调与通衢有关。通衢集团（Crossways Group）经营的通衢市集，正是那部草案所针对的那一类市集。日晷的股东里，乔氏家族办公室（the Qiao Family Office）持有它三成股份；这个家族办公室持有通衢集团的全部 B 类股，第 21 年春的股东名册上，占通衢约五成六的投票权。通衢集团与它的结算子公司通衢支付（Crossways Pay），又是日晷最大的两家客户，付给日晷的评级费约占日晷营收的一成。日晷的答复是："评级独立；评审委员会的意见不受股东影响。"日晷不公开评审委员的名单。

这件事里也有另一面。第 18 年起，日晷免费公开它对王国州小市镇债券的评级方法，几个小镇因此第一次发了债。王国州的《信用评级机构登记办法》要求评级机构登记，不审评级方法。这里的法与机构，都是这个系列设的例子；本篇引到的条文只作事实，本系列不主张它们是最好的设计。

评级可以上调，也可以下调；下调的理由可以对，也可以错。那位市议员的话是一个判断，不是一项证据；日晷的话也是一个判断，不是一项证明。本篇不替他们判这一次下调出自谁的意思。要问的是另一件事：一家机构评别人的信用，替许多人问了一个问题，许多人凭它的回答作决定；积累起来的股份与客户关系，能不能换成对这样一条问的通道的控制？

前几篇写的，是积累怎样形成、在哪里显出权力面（第一篇），在关系一级成了别人绕不过去的关口时怎样读（第二篇），大到许多人都绕不过去时制衡要多强、落在哪里（第三篇）。这一篇写积累换成别的作用。名声换来机会，资本换来更多的资本，这是涌现层的常事，本篇不以减少它为目的。可是有两种兑换，本篇要设闸：积累不兑换为对人的基础层的支配；积累不兑换为对追问的控制。

这一篇分五步。先说第一道闸，用青窑镇的劳动合同（第一节）；再说第二道闸，用日晷与回音壁（第二节）；然后说政经分离的两个方向，用州议会的一票与一位离任的处长（第三节）；接着说两道闸以外的交换照常放开，用几样名声，并在那里与沃尔泽、桑德尔对照（第四节）；最后说结构依赖换来的公共支持，用通衢支付的一笔紧急贷款（第五节）。

读法照前置篇：构用应当，位置用职责，人用自己的法。规范层、规则层、判断层的分法也照前置篇；本篇只在用到的地方说明哪一句落在哪一层。

第一节　积累不兑换为对人的基础层的支配

青窑镇（Bluekiln）的劳动合同，第二篇已经读过一遍。那一遍问的是：宿舍、窑券、预支与竞业几条约束，一条一条看与合起来看，落在同一个人身上时，还留不留得下他离开的路（第二篇 E12）；离开以后往哪里去的那一条，又怎样读（第二篇 E12a）。这一节换一个问题，再看同一组事：这些约束，各是拿什么换来的？

青窑瓷业（Bluekiln Porcelain Works）雇三千四百人，约占全镇适龄劳动人口的七成；工资比王国州同行平均高一成，学徒班免学费，公司还出资修了镇上通往州道的路。镇上四成的住房属于公司；职工住宿舍，房租从工资里扣，离职后三十日内搬出。工资的两成以"窑券"发放，窑券只能在公司的青窑合作商店与食堂用；合作商店是镇上最大的商店，物价比苹果城高约一成五，镇上另有三家小杂货铺，不收窑券。公司给职工预支工资，最多六个月；劳动合同第 14 条写："预支未清的，离职须经公司同意。"第 19 条是竞业：离职后两年内，不得在王国州境内从事制瓷或为制瓷企业工作。

这几条里，有三样是公司积累起来的东西：一镇四成的房子，一家最大的商店，一笔可以借出去的钱。它们本身不是问题。房子租给人住，商店卖东西，钱借给急用的人，都是涌现层里的交换，镇上的人也确实用着它们。要问的是兑换：这些东西，有没有被换成另一种作用。

第 14 条最清楚。预支是公司借出去的钱；借出去的钱要还，这是一项正当的请求，公司照借约去要，是交换。可第 14 条要的不是还钱，是另一件事：预支没有还清，离职"须经公司同意"。一笔债由此换成了一项同意权，管的是人能不能走。老窑工第 22 年春提出离职时，有四个月的预支没有还清，公司没有同意；一个月以后，他向亲戚借钱还清了预支，这一道才开。照这一条的写法，公司手里的一笔钱，换成了对他离开与否的决定。

宿舍与窑券是另一种写法。住处与工作拴在一起，离职三十日内搬出；在一个四成房子属于公司的镇上，离开工作，先要找到住处。工资的两成只能在公司的商店与食堂花；离开公司以后，手里的窑券还能不能用，没有记载。它们换成了什么，要看这个人还有没有别的路：镇上别的住房有多少、多贵，没有记载；三家小杂货铺不收窑券。逐项看与合起来看怎样读，第二篇已经写了，这里不重复。这一节只把它们放在同一个问题下：积累起来的房子、商店与钱，换不换成对人离开、拒绝与撤回的控制。

为什么这一种兑换要设闸？因为它换到的地方是基础层。离开、拒绝与撤回，是人在一段关系里保住自己的那几条通道。制度论把它写进构对内部主体的第一条应当：一项约束「不压缩生命基底至威胁承载者，不封死退出、拒绝与撤回的通道，是限；越出任何一条，是侵」（制度论第二篇）；这一条「不随关系厚度调节，也不随任何条件打折」（制度论 Z9）。原说构自己对内部主体的施为，这里推到经济位置拿积累所作的兑换（自担）。基础层不随资产增减，也不以成就交换：一个人拥有多少，不改变别人离开的路是不是别人自己的；一家公司做了多少好事，也换不来别人离开时须经它同意。粮食、饮水、电力、住处，凡人维持生活所必需的东西，一旦囤在一处，都可以被换成这样的控制；有人富有，不等于他可以拿这份富有去支配他人。这一道闸因此有两支：一支是把人维持生活所必需的东西换成压缩其生命基底的控制，第二篇读生命基底处唯一的供给位置时已经写到（第二篇 E11、E11a）；一支是把钱、住处或工资的发放方式换成对离开、拒绝与撤回的控制，这一节的事例落在这一支。什么算维持生活所必需，基础层具体包括什么，清单在规则层（第二篇 E10）。

这一条要防一个过头的读法。关系里有期限、有条件、有程序，离开要付代价，不等于离开的路已被封死。权利论第五篇划过这一道阈值：封死退出通道，「不等于关系中存在期限、程序、条件或退出成本，它指主体不再保有拒绝、撤回持续绑定或转入可审查程序的结构通道」（权利论第五篇 §III）。提前三个月的离职通知是一个期限；搬出宿舍要另找住处，是一项代价；它们本身不是这道闸要挡的东西。闸挡的是兑换的那一步：积累换成了对离开本身的决定，人不再保有那一条结构通道。第 14 条把离职交给公司同意，碰着的正是这一步。

签了字，也不改这一读。窑券有一份州劳动厅的复函：以代用券支付部分工资，"须经职工书面同意"；职工入职时都签了同意书，劳动合同的每一条也都签了字。签了字，事情就成了自愿的吗？权利论第五篇在基础层答过这一问。一个主体宣称自愿交出它的第一人称封装，让另一个位置替它决定，「这改变判定吗。不改变。判据在结构，不在宣称的自愿」。那里分开两样：可撤回的授权，是人以自己的法处置自己的行使；封死撤回与退出的位置让渡，是另一回事，「宣称的自愿不改变结构读数」（权利论第五篇 §III）。入职时签下的同意，若是一项随时可以收回的安排，它在限之内；若它让离开从此须经对方同意，签字不改这一结构。基础权利能不能由本人放弃，这一般的问题，本系列与制度论一样不答；在基础层，上面那一句已经作了答。

这一节的几句话，落在三层上各不相同。积累不兑换为对人离开、拒绝与撤回的控制，在规范层；它不随资产大小，也不随签字与否。王国州的《工资支付法》只写工资"以通用货币支付"，没有写代用券的比例上限，劳动厅的复函要书面同意：这些是规则层的写法，是例子；代用券可不可以有、可以占多少，宿舍搬出给多少日子，由规则层去定，本篇只给判据。老窑工那一年的处境里，几条合起来是否已经越线，是判断层的事，第二篇照同一个人、同一实际配置读过。

有人会说，这几条都是公司给的：没有公司的宿舍，镇上的人没处住；没有预支，急用钱的人借不到钱；窑券换来的，也是实实在在的东西。它说的是真的：青窑瓷业的工资比同行高一成，学徒不交学费，镇上通往州道的路是它修的。第一道闸不挡这些。房子可以租，钱可以借，商店可以开；闸挡的，只是把这些换成别人离开时须经它同意的那一步。公司要回借出去的钱，照借约去要，要不到，走追债的路；它不能把一笔债换成一个人的去留。给出去，与拿给出去的东西卡住人，是两件事；这一节只拦后一件。后一件若还以"给"的名义出现，就碰着这个系列所说的殖民。权利论第六篇写：「殖民是伪装成涵育的越界施为，它以一张外部网格代写主体自己的法，改画封装维度，封死退出或打断通道，或把主体降为实现他人目的的材料」（权利论第六篇 F26）。窄义之下，不装成"给"的那一种落在哪里，制度论第一篇说得清楚：「排放若压到下游人家的身体与生计，一套配置若封死一个人拒绝的路，这些越线不一定装成涵育，也都在前两条里」（制度论第一篇第八节）。拿必需之物去支配人，不必先装成涵育才越线；它越的是前两条的线，若还以"给"的名义出现，才是殖民。这一句说的是一种安排的结构，不是给哪一家公司定性；青窑这几条合起来是否已到那一步，仍照第二篇的读法核。

老窑工的两位同事，没有人替他们出钱，留在了公司；他们留下的全部理由，没有记载。第一道闸拦得住第 14 条那一类兑换，管不到一个人自己决定留下。

E21　闸一：积累不兑换为对人的基础层的支配：不把人维持生活所必需的东西换成压缩其生命基底的控制（第二篇 E11、E11a），也不拿钱、住处或工资的发放方式去锁住人的离开、拒绝与撤回；这两样都不在涌现层的正常交换里。什么算维持生活所必需、基础层具体包括什么，清单在规则层（第二篇 E10）。基础层不随资产增减，不以成就交换。关系里有期限、条件、程序或退出成本，不等于退出已被封死；封死，指人不再保有拒绝、撤回持续绑定或转入可审查程序的结构通道（权利论第五篇 §III）。入职时签下的同意不独自改变读数：「判据在结构，不在宣称的自愿」（同上）；基础权利能不能由本人放弃，这一般的问题不答。原说构自己的施为对每一个内部主体都不越线（制度论 Z9），这里推到经济位置以积累所作的兑换（自担）。

第二节　积累不兑换为对追问的控制

回到日晷。

一家评级机构做的事，是替别人问一个问题：这一个借钱的一方，还不还得起？市政府、公司、小镇发债，买债的人多半不会自己去查，他们看评级。评级在这里是一条问的通道。规则的制定，争议的裁判，信用的评级，交易的记账，同属这一类：许多人不必每件事亲自去问，是因为有一处在替他们问，而问出来的东西，可以被拿去作决定。

第二道闸挡的，是积累换成对这类通道的控制。控制，指的是一种地位：被问的一方能够独自关闭对它自身的追问、纠错或必要的兑现。这一句要写准，免得读宽。一份评级、一本账、一项裁判，可以有具体负责的位置：总要有人评，有人记，有人判。负责作出一份记录或判断，不等于取得独占其终审的权能，更不等于能取消外面对它的追问。记录要有明确的范围，有可以复核的来路；它怎样配置，是规则层的事。在这一释义之下，可以写一句短的：账不归被记账的一方，也不归任何单一的记账者。这一句不是说一家机构不能记一份局部的账；它说的是，没有一处记账者，能把它记下的账变成别人问不了、改不了的东西。

根据在制度论。制度论的第五条应当写：「制度保持被追问与修正的通道，且通道满足三项条件：独立于被追问的一方，内部主体实际可以发动，发动不被报复。独立是功能义，不要求在制度之外，要防的是被审的一方独自封住对自己的审查」（制度论 Z34）。原说制度对它自身的追问通道，这里推到经济位置所及的那些问的通道（自担）。三项条件的出处在《制度如何可能》（下称 F6）的命题 IV：「一个以不疑为前提的制度，其修凿通道须满足三项结构条件：独立于被凿者；成员可实际发动；发动不被报复。」那一命题有它自己的射程：它的主语是一个以不疑为前提的制度，而且「本命题只立条件，不立职务与应当」（F6 §3.2）。日晷与它评的那些城、那些公司之间，不是一个以不疑为前提的制度；所以这里的"应当"不从命题 IV 取，取自制度论，制度论自己写明了"从条件到应当"这一步是它自担的。

怎样判一条通道是不是已被控制？政府论在一级政府里给过分寸：「判捕获看结构，不看机构的名称与设在哪里：投诉窗口设在被投诉的部门里，本身不等于捕获；设若它是唯一的门、可以自行结案、转出须经它同意，被问的一方就能独自封住对自己的审查」；同一条又说，「提名关系、同党关系，也不自动证明捕获」（政府论 G38）。原说政府里的追问通道，这里推到经济位置（自担）。照这一分寸，股权与收费关系本身不定案：日晷若没有一个与被评一方相交的股东，不因此就证明它独立；现在有这样一个股东，也不因此就证明它已被捕获。要看的是结构。政府论第六篇把它写成三件：「谁能启动；对理由、行动、规则或任职位置有什么实际效力；被问的一方怎样不能独自取消它。」（政府论第六篇第二节）

道德律第三篇在声誉上碰到过同形的结构。它把「中心化权威表述声誉状态」列为结构上不许的：「没有任何单一机构可以以约束力下判决每个Self的声誉是什么」（道德律第三篇 §1.3）；理由是「一个对声誉下判决的中心化权威，可被针对该权威的策略性行动者钻空子」（§8.3），而「分布本身就是抵御优化驱动之捕获的结构性防御」（§8.4）。那一篇的范围，是「纯15DD共同体——所有成员都是15DD立法主体的理想型」（§0.6）；它说的保护「是结构性而不是制度性的——它不依赖任何具体institution的enforcement」（§8.3），靠的是共同体里许多各自作辨识的人。这里只取其中一样结构：一个单一的中心，可被针对它的一方钻空子。评级与评分是记录，不是声誉（第一篇 E4）；在混合现实里，一处集中的记录，同样可以被针对它的一方钻空子。由此要求记录有可以复核的来路、可以从别处问，这一步是本篇自己走的：原句说的是共同体里不靠制度的分布式保护，这里推到混合现实里对记录的制度安排（自担）。

照这些读日晷。下调压在苹果城身上：融资成本上升了，最后由苹果城的纳税人来付，医院扩建所及的那些人也在其中。问的是评级的决定与它的纠错通道，怎样受股东、收费与程序的影响。手里的事实是这几样：评级费由被评的一方付；乔氏家族办公室持有日晷三成股份，又以 B 类股定着通衢的方向；通衢集团与通衢支付是日晷最大的两家客户，评级费约占它营收的一成；日晷公开的是小市镇债券的评级方法；评审委员的名单不公开；州的登记办法只要求登记，不审方法。

照这几样，事实指到这几处。苹果城的下调所列的一条理由，提到的正是苹果城表态支持的那部法；日晷说独立，市议员说与通衢有关。两句都是判断，都在记录里。股东有没有下达过降级的命令，没有记载；所列的理由是真是假，也没有记载；本篇不替哪一方补上。可是另一件可以说：苹果城若要问这一次下调，它手里有什么？评审委员是谁，不知道；城市债的评级方法有没有公开，没有记载；州的登记不审方法；有没有一处能复核一次评级的地方，没有记载。问的通道在哪里，事实上没有给出来。这是一个追问的入口，不是一项已经证明的捕获。

通衢支付那一面，也照此读。危机前一个月，日晷给通衢支付的仍是最高一级（第五节）。被评的一方付费，被评一方的控制人又是评的一方的大股东：这正是"账归被记账的一方"的那一种风险。一次评级错了，本身不证明它被控制；要问的是，这样一份评级若有错，谁能在事前问到它，问了能不能改，被评的一方能不能独自让这一问停下。

回音壁（Echo Wall）是第二件。它是一个点评与分享的平台：用户给店铺、作坊与上门维修、保姆这一类零工打分、写评；第 21 年，王国州约七成的成年人用它。第 22 年起，苹果城几家小额贷款公司把它的评分写进放贷条件，店铺评分四星以下不贷；一家家政中介只派评分四点五以上的保姆。评分在这里已经不只是一份记录，它成了别人作决定的条件，成了一道门（第一篇 E5；第二篇 E7）。回音壁也做过守护记录的事：第 22 年，它下架了一批冒牌的青窑瓷；外地的买家多半经它找到青窑镇的作坊。

第 23 年春，通衢集团收购了回音壁。通衢集团的自营品牌"通衢家用"（Crossways Home），是在通衢市集上卖东西的一个卖家；它的商品页上的评论，也记在回音壁上。苹果城一位统计学者第 24 年抽样：这些商品页上的差评，以"违反社区规范"为由被删的比例，由收购前的百分之三，升到百分之十一。

第 23 年秋，老窑工的窑坊一周内收到一百多条一星差评，都来自新注册的账号，评分由四点七跌到三点一。那个月，窑坊向一家小贷公司申请扩窑贷款，被拒。窑坊向回音壁申诉；三周以后，回音壁删了其中六十条，说"其余不构成违规"。申诉由回音壁的内容审核组处理，不说理由。

受这件事的，是老窑工的窑坊，也是所有评分被拿去决定放贷、派工的人。记载里有收购前后删评比例的变化，有窑坊这一次的具体经过；收购以前的回音壁，也不因此证明它的每一条通道都独立、每一条评论都真。

照那三件去问窑坊的申诉，事实是这样。能不能启动：能，窑坊申诉了。有没有实际效力：有一部分，六十条删了。被问的一方能不能独自把它关上：申诉由回音壁自己的审核组处理，不说理由；其余的留不留，由它一家定，没有别的门可走的记载。设若这是唯一的门，又可以自行结案，那就是政府论说的那一种：被问的一方能独自封住对自己的审查。那一百多条差评是谁写的，没有记载，不能补成通衢指使；删评比例上升，也不能补成每一次删除都在消灭追问。可是几样事合在一起：记录的一方被场里的一个卖家买下，那个卖家的差评被删得多了，一家作坊的评分被一批新账号拉下以后贷不到款，而对评分的申诉，记载里只有这一处，不说理由。这正是第二道闸要看的那种结构。

这一节的几句话，也分三层。积累不兑换为对问的通道的控制，在规范层。评级方法公开到什么范围，评审委员与股东有关联时怎样回避，评分的申诉要不要说理由、有没有平台以外的复核：这些是规则层的事，可以有不同的写法，这里提到的都是例子，不主张是最好的设计。这一次下调对不对，那一次删评对不对，交给作这些判断的位置，账落在那些位置上。

有人会说，评级与评分都是市场里的意见：不同意的人可以不看，可以换一家，市场自会惩罚评得不准的那一家。这一说的分量在于，评级本来是许多家都可以做的事；日晷免费公开小市镇的评级方法，也确实让几个小镇第一次发了债。可是"可以换一家"答不到被问的那件事。日晷占六成，回音壁七成的成年人在用，评分已经写进放贷与派工的条件；被下调的城、被拒贷的作坊，不能靠"换一家"收回已经落在自己身上的那一份记录。第二道闸不要求评级与评分不存在，不要求它们只有一家，也不要求它们都对；它只要求，没有一处作记录的位置，能凭积累起来的股份与客户关系，让对这份记录的追问只经过它自己。

这里留下一处：日晷评审委员的名单不公开；它对城市债的评级方法公开到什么范围，没有记载；苹果城可以从哪里问，要由王国州的规则去写。

E22　闸二：积累不兑换为对追问的控制。规则的制定、裁判、评级与记账，是问的通道。积累不得兑换成这样一种地位：被问的一方能够独自关闭对它自身的追问、纠错或必要的兑现。评级、记账与裁判可以有具体负责的位置；作出一份记录或判断，不取得独占其终审、取消外部追问的权能；记录要有明确的范围与可复核的来路，怎样配置在规则层。账不归被记账的一方，也不归任何单一的记账者；这一句不禁止单一机构记一份局部的账。独立照「独立于被追问的一方」核，是功能义（制度论 Z34）；判捕获看结构，不看机构的名称与设在哪里，提名、同党一类关系也不自动证明捕获（政府论 G38）：没有股权交叉，不自动证明独立；现有交叉，也不自动证明捕获。三项条件的出处在 F6 命题 IV，那里只立条件，不立应当；应当据制度论 Z34。道德律第三篇 §1.3、§8.3、§8.4 对中心化表述的禁止与分布式的防御，只取"单一的中心可被钻空子"这一结构作同形；那一篇立在纯 15DD 共同体的理想型里（§0.6），所说的保护是结构性的，不靠制度（§8.3）；推到混合现实里记录的可复核来路，自担。风险与追问的入口，不当作已证的捕获。原说制度自身的追问通道与政府里的捕获，这里推到经济位置所及的问的通道（自担）。

第三节　政经分离，双向

《线上市集法》草案在王国州议会的经济委员会，委员十五人。第 23 年至第 24 年，通衢集团与它的高管给其中九人捐了竞选款，都在王国州《竞选财务法》的上限之内；乔氏基金会（the Qiao Foundation）给四位委员所在选区的学校捐建了图书馆。通衢同年自愿公开了它的全部政治捐款。第 24 年秋，委员会以八票对七票，删去了请求商贸署启动拆分的一款，保留了公开排序规则与设独立申诉机构两款。

同年冬，竞争审查处的处长离任。通衢市集的那次竞争审查是他主持的：审查从第 23 年开始，第 24 年春出了报告，建议通衢市集与"通衢家用"分属两家公司，各自经营（第三篇）。离任四个月以后，他出任通衢集团的副总裁，主管"合规与政府关系"。目的联合城邦（UCSE）的《公职人员离任规定》要求，离任一年内不得就本人主管过的事务与原机关接触；它没有限制受聘。这些法与规定，同样是这个系列设的例子。

钱进了议会，人出了机关。这两样，是不是同一件事？

本篇的回答是：它们是同一条界线的两个方向。钱不买成审它、问它的那份政治权力；位置上作出的决定，也不变现为日后从被它问过的一方拿到的财富。前一个方向人们熟悉：一家公司不能用钱买走审它的那一票。后一个方向容易漏掉：一个在位置上审过一家公司的人，不能把他在位置上作出的决定，变成日后从那家公司拿到的报酬。两样都是兑换，换的都是问的那条路：前一样拿钱去换问它的路，后一样拿问的路去换钱。所以这一条是第二道闸最强的样子，是它的极端情形，在公共权力一级：审它、问它的，是立法与审查这类公共位置。公共权力的每一级都一样，大到一国，小到一个小区的业主自治。它也是分权制衡的道理在经济位置上的延伸：经济的权力与政治的权力分开站，才能彼此制衡。

后一个方向怎样收进第二道闸，要从付钱的一方看。第二道闸的主语是积累。捐款、许诺的报酬或事后的受聘，若实际被用来买下审它、裁断它或纠正它的位置，付钱的一方就是在用积累取得对追问的控制；事后的报酬之所以也在这里，是因为它可以在决定以前就被预期，位置上的决定于是有了日后的价钱。从收钱的一方看，是授予权力的运用被换成了私人的报酬；那一面归政府论，捕获照结构判（政府论 G38），两处互相指引，各写各的。捐款、受聘与先后的时序本身，都不单独证明这项兑换。这一条也有它的射程。"钱不买成政治权力"，在这里说的是不买成审它、问它的那份政治权力；拿钱买一部法给自己开门，替自己挡住竞争者、换一份专营，读在关口与制衡（第二篇 E7；第三篇），不在这一条里。

这一处，斯密早已碰到过。他在《国富论》第一卷末尾说，凡出自经营者这一阶层的新的商业法律或规章的提议，「都应当以极大的审慎去听，非经长期、仔细的审查，不应采纳；不仅要以最认真的注意，也要以最怀疑的注意去审查」（The proposal of any new law or regulation of commerce which comes from this order, ought always to be listened to with great precaution, and ought never to be adopted till after having been long and carefully examined, not only with the most scrupulous, but with the most suspicious attention；Wealth of Nations, Book I, ch. XI, Conclusion of the Chapter，据 Project Gutenberg #3300，A 档）。接触面在这里：出自经营者的立法提议，审它的人要以怀疑的注意去审；本例里那一方，正是这部法所针对的一方；原说经营者的立法提议，这里推到被这部法针对的一方对审它的人的影响（自担）。原句的上文说的是缩小竞争，那一面本系列读在关口与制衡（第二篇 E7；第三篇）。斯密的尺度是公众的利益，理由是那一阶层的利益与公众的利益不同；本篇的尺度在结构，要核的是问的那条路有没有被买走，不判哪一个阶层的利益所在。

要核的，是有没有买走问它的那条路。这里最要防的，是互相代判。合法的捐赠，不因为合法就证明没有买；有捐赠，也不因为有捐赠就证明已经买了。离任以后受聘，同此：受聘合规，不证明没有拿位置去换；受聘了，也不证明换过。政府论说，「提名关系、同党关系，也不自动证明捕获」（政府论 G38）；原说政府里的提名与同党，这里推到捐赠与受聘（自担）。判在结构与时序：钱从谁那里来，到了谁手里，在哪一个决定之前；人从哪一个位置离开，在那个位置上管过什么，离开以后与原来管过的事有没有接触。

照这一分寸读州议会这件事。那十五位委员是王国州的选民选的；拆分一款删了，在审的那些卖家面前少了一条路。记载里有的是：九名委员受了捐，四个选区有了图书馆，八票对七票，捐款都在上限之内，而且已经公开。这几样要分开记：受捐的九人与投赞成票的八人是不是同一批人，没有记载；有没有事先的许诺，没有记载；四个选区的图书馆与那四位委员怎样投票，也没有记载。所以事实只指到这里：那部法所针对的一方，在审这部法的人身上花了钱，并且公开了；审这部法的人，删去了其中请求启动拆分的一款。这是一个追问的入口：选民可以问他们的委员，下一次选举可以问，《竞选财务法》的上限也可以被问。它不是一项已经证明的买卖。若没有这些捐款，那一票会是什么结果，本篇不作这一假设。

处长的受聘要另读。他主持的那次审查，报告建议把场与自营分开，这一建议对通衢不利；他离任在删款之后，受聘又在离任四个月以后。所以他的受聘，不能补成删款以前的一个因素。要问的是后一个方向：他在位置上管过通衢，如今到了通衢主管"政府关系"。离任规定管的是一年之内，不得就本人主管过的事与原机关接触；他的新职位会不会在一年里碰到那些事，有没有接触过，没有记载，不能补成已在禁期内接触。规定没有限制受聘，这是规则层的一种写法；要不要限制、限制多久，由规则层去定。若规则层要限他一段时间不受聘于他审过的一方，那一条护的是公众问它的那条路；它与第二篇所说为保护他人而设、不在竞业禁止之列的约束同理（第二篇 E12a）。成不成立，要说明护的是哪一条问路，这一受聘怎样使它受制，为什么回避或限制接触不够，范围多宽；不因名为冷却期就成立。规范层只说一件：位置上的决定，不能是日后报酬的价钱；而一个位置作出的决定照旧可被追问，不因作出它的人走了就停下。

通衢自愿公开全部政治捐款，这一处也要记下。公开本身是一项让问成为可能的施为：九个人收了捐，选民知道，是因为它公开了。这一面与上面那一面同时成立，一件不抵销另一件。

有人会说，捐款是政治参与，受聘是择业的自由；限制它们，等于限制人参与政治、限制人去哪里工作。这一说的分量在于，它守的东西是真的：公民与公司都可以支持他们认同的候选人；一个人离开公职以后，也要有地方去。第二道闸不禁止捐赠，也不禁止受聘；上限之内的捐赠与合规的受聘，在涌现层里照常放开（第四节）。它挡的只是两样：买走问它的那条路；拿问的那条路去换报酬。这两样怎样认，看结构与时序，在判断层；捐款的上限、离任的冷却期、回避的范围，在规则层，可以有不同的写法。

委员会删去拆分一款以后，州政府请求启动拆分的那一条路没有了；商贸署依它自己的程序还走不走，没有记载。拆分作为让位置不终局的手段，在第三篇（第三篇 E18）。

E23　政经分离，双向，是第二道闸在公共权力一级的极端情形。钱不买成审它、问它的那份政治权力；位置上作出的决定，也不变现为日后从被它问过的一方拿到的财富。两个方向换的都是问的那条路；事前的捐款与事后的受聘，是本例展示的两种时序，是否发生兑换仍须核，时序本身不单独证明它。收钱一侧所涉授予权力的运用，照政府论 G38 读，两处互相指引，各写各的。公共权力的每一级都一样，大到一国，小到一个小区的业主自治；这是分权制衡的道理在经济位置上的延伸：经济的权力与政治的权力分开站，才能彼此制衡。要核的是有没有买走问它的那条路，或者拿问的那条路换了报酬。买走问它的路，与普通的受聘、捐赠，不互相代判：合法的捐赠不因此证明没有买，有捐赠也不因此证明已经买了；离任后的受聘同此。判在结构与时序，不在关系本身：提名、同党一类关系不自动证明捕获（政府论 G38；原说政府里的提名与同党，这里推到捐赠与受聘，自担）。拿钱买一部法给自己开门，读在关口与制衡（第二篇 E7；第三篇），不在这一条。

第四节　闸以外，涌现层的交换照常放开

两道闸都写过了，要回头防一个误会：设了闸，是不是凡兑换都可疑？

第 22 年，苹果城一家画廊办"青窑新瓷展"，门票上印着乔家第三代的名字，身份是"联合策展人"。乔家第三代那一年二十四岁，橡园学院毕业，在回音壁上有九十万关注者，常发旅行与品酒的内容。开展一周，参观人数是上一年同类展的四倍。老窑工有两件作品在第二厅。他在青窑瓷业拉坯、烧窑三十一年，镇上的老师傅和苹果城几家老瓷器铺都说"他烧的釉色稳"，镇外很少有人知道他。另一处，通衢市集给部分卖家"金牌卖家"标识，按销量与付费推广的额度评定。

这里有好几次兑换的样子。门票上印着一个在回音壁上有九十万关注者的人的名字，身份是联合策展人；开展一周，参观人数是上一年同类展的四倍。名字、位置与人数同时出现，它们之间怎样连着，记载没有说；可这已是名声换来机会的常见样子。老窑工有两件作品也在这次展里，在第二厅。一个卖家的钱换来付费推广，付费推广的额度又是评"金牌"标识的依据之一。照本系列的读法，这些都是涌现层里的交换：名声换来机会，资本换来更多的资本，资质换来职位，本身都不是问题（第一篇 E1）。本系列不以减少它们为目的，也不以"它们可能被兑换"为由，把涌现层的交换封起来。

有一处要写准。"金牌"标识是钱换来的，可钱换来的是一个标识，一份记录，不是声誉。照道德律第三篇的意思，声誉买不到：「声誉在SAE里不是某种资本」；「它不与任何14DD主要介质交换」（道德律第三篇 §2.2、§1.1；那里的结论立在纯 15DD 共同体的理想型里，本系列在混合现实里取它的哪一条，见第一篇 E4）。老窑工"釉色稳"的口碑，没有谁买得到；通衢的标识，卖家可以花钱换来。两样分开，钱买标识，就不是钱买声誉。标识若被人拿去作放贷、派工的条件，它就成了一道门，那是第一篇与第二篇读的事（第一篇 E5；第二篇 E7）；这里没有这样的记载。

所以两道闸只说本篇处理的兑换限制。积累不兑换为对人的基础层的支配，不兑换为对追问的控制；一次交易没有撞上这两道闸，只说明它不在这两处被拦，不认证它在一切方面都正当。它仍要守本系列别处的要求：不越线，不封路（第二篇）；可被追问，制衡随权力面递增（第三篇）；做基础层之事的，制衡之外还要保障（第三篇 E14a）；为人的涌现给条件，不替人定方向（第六篇）。上一节的捐款与受聘也在这里：上限之内的捐赠，合规的受聘，本身是涌现层里的常事；它们若换走了问它的那条路，才进第二道闸。

这一节可以与两家对照。

沃尔泽在《正义诸领域》（1983）里问过同一件事：钱不应当买到什么。按 G. den Hartogh 的概括，沃尔泽在书中列出一张他认为在西方社会里被看作落在钱的领域之外的清单，即"受阻的交换"（blocked exchanges）：政治权力与公职，人身的自由，言论、出版、信仰与集会的自由，移居与离开的权利，免于共同摊派的差事，爱，尤其是对具体某人的爱，等等；其中禁止出于绝望的交换，由需要的原则来说明（据 den Hartogh 1999 的转述，所据原书第 100—103 页；原书未核，C 档）。据转述，他的根据在各个领域各自的分配原则。题是同一个题：积累起来的东西，在哪里不能换。原书未核，本篇只列可对照之处。可对照的有三处：清单里的政治权力与公职，对着本篇第三节的方向；出于绝望的交换，对着第一道闸所挡的、人被处境逼到交出离开的路；还有根据：本篇两道闸的根据在结构，一道在人的基础层，一道在问的通道，他的根据据转述在领域与分配原则。这几处是相遇还是分开、各在哪一步，原书核到才写。本篇也不列钱不能买的全部东西，两道闸只是本篇处理的两项兑换限制，其余的要求在本系列别处。

桑德尔在《钱买不到的东西》（2012）里问，有没有一些东西是钱不应当买的。按 D. M. Kennedy 书评的概括，他给出两种反对。一种关于公平：在不平等的处境里，愿意买与买得起不是一回事，市场上的"自由选择"可能并不自由。一种关于腐蚀：把一样东西放上市场，会消解它本来的价值（据 Kennedy 2012 的转述；原书未核，C 档）。原书未核，以下只列可对照之处。前一种反对，与第一节"签了字，也不改这一读"碰着同一类处境：处境逼人时，交易里的同意说明什么。后一种反对问的是一样东西的意义；本篇的两道闸不以东西的意义为据。在什么情形下，市场改变了一样东西的意义，这一问与本系列的哪一条边界相接，原书核到才能写。

反过来也有一种担心：闸以外一律放开，等于把积累起来的东西在别处造成的不平等也放过了；九十万关注者的人与一个策展人的位置连在一起，一个烧了三十一年窑的人在镇外无人知道。这一担心看见的差别是真的。可是本系列不以分配结果为终点，也不以减少积累为目的（前置篇）；它要问的，是这份差别有没有换成对别人的基础层的支配，有没有换成对追问的控制，有没有在别处越线、封路。老窑工的两件作品也在那次展里；照这一次的记载，没有一处显出名声的兑换关掉了他的路。若有一天，一个人的关注者能决定谁的作品进不进展厅、谁贷得到款，那时问的就不是这一节，而是第一篇与第二篇。

这里留下一处："金牌"标识按销量与付费推广评定，买家知不知道它的一部分是买来的，没有记载；标识怎样标明它的来路，是规则层的事。

E24　闸以外，涌现层的交换照常放开。名声换来机会，资本换来更多的资本，资质换来职位，上限之内的捐赠与合规的受聘，本身都不是问题；不以"可能被兑换"为由封住涌现层的交换。钱买得到标识与记录，买不到声誉（道德律第三篇 §1.1、§2.2；见第一篇 E4）。两道闸只说本篇处理的兑换限制：没有撞上两道闸，不认证一次交易在一切方面正当，也不取消本系列别处守线、可被追问、保障与给条件的要求。

第五节　公共支持不作背书

通衢支付是通衢集团的结算子公司。它替通衢市集上的买卖双方结算；第 25 年起，也替王国州各镇的小商户收款。到第 26 年，州内约四成小商户的收款经它。

第 26 年夏，通衢支付的一批短期投资亏损，结算延迟了三日。青窑镇有两家作坊，当月发不出工资。州金融厅与州财政署给了通衢支付一笔紧急贷款，利率低于市价，条件是三年内不得给母公司分红。州政府的发言人说："它倒了，倒的是四成的小商户。"危机前一个月，日晷给通衢支付的评级仍是最高一级。

经通衢支付收款的小商户，那两家作坊领不到工资的工人，还有王国州的纳税人，都在这件事里：贷款用的是纳税人的钱。下面问两件：这笔贷款的根据是什么；它给了通衢支付什么，没有给什么。那批投资的亏损最后由谁承担，没有记载；四成的商户是否必定会倒，也没有记载。

先说通衢支付何以能取得这笔支持。它的位置是积累来的：一家集团的子公司，一笔一笔替人收款，收到了全州四成的小商户。全州约四成小商户的收款经它；它们一时换不换得开、换一家要付什么，没有记载。可许多人的收款压在一处时，它倒下的后果就不只落在它自己身上。权利论第四篇讲退出怎样显出权力面时说：「只有当结构依赖在场时，退出权才经 F15 演化为权力。」（权利论第四篇 §V）原句的主词是一项退出，说的是它显出权力面；这里换了两处：主词换成一个位置可能的失败，说的是它由此取得公共支持（自担）。通衢支付并没有拿"我可能倒"去施压；州给出贷款，照发言人的话，理由在它所及的那四成小商户。结构依赖在这里不是一句威胁，是一个处境；公共支持，是在这个处境里作出的。

再说这笔支持不给什么。政府论在政府保证它载入的权利时写过：「保证不是背书。构成成功与运行成功，都不等于规范正当」；同一条又说，「久存本身不是续存的理由（自担）」（政府论 G22）。原说政府对它载入的承认性权利的保证，这里推到公共对一个积累位置的支持（自担）。一笔紧急贷款给了出去；它与延迟以后的恢复怎样连着，没有记载。它不因此证明那批短期投资做得对，不证明通衢支付此前的配置没有问题，也不证明它从此该一直这样存续下去。取得公共支持，不替它既有的施为与存续发放正当性。

"担保"一词，在这里要用准。一笔附条件的贷款，不等于州新设了一份普遍的担保。贷款有利率，有条件，有期限，作出它的是两个具体的机关；它不是一句"以后无论怎样都兜底"。这件事以后，会不会有人以为下一次州还会出手，因而更放心地把收款放在通衢支付，或者通衢支付自己更放心地去做短期投资，没有记载。这一问是真的，留在这里，不补成已经发生的事。

那么，作出支持的一方要做什么？制度论的第五条应当里有一句：「制度对它自己造出的东西负说明之责，并须答到被问的那件事上」（制度论 Z34）。州金融厅与州财政署作出的这笔贷款，是它们造出的东西。它们要说明根据、权限、对象与代价：凭什么借，谁有权借，借给它是为了保住谁，代价落在谁身上。发言人那一句"倒的是四成的小商户"，说的是对象：这笔钱要保住的，不是通衢支付，是经它收款的那些人。照这一说法，支持的安排就要答到这个对象上。三年不得分红，挡的是钱从子公司流回母公司；那两家作坊发不出的工资，这笔贷款到了没有，没有记载；作出贷款的机关有没有说明这一处，也没有记载。这是支持的安排要答到的一处。救助本身是授予，作出它的是政府，它怎样作出、受怎样的问，归政府论；推到积累位置的那一半，是本系列的：接受支持的位置，照它的具体条件、所负的职责与权力面受问，不因受了支持就少受一分问（自担）。

这一条要与第三篇的一条分开。第三篇说，构做的事落在基础层时，制度在制衡之外还要保障这一供给（第三篇 E14a）；那里的根据在那一项供给本身，是人维持生活要靠它。这里的根据不同：结算是不是基础层的供给，本篇不定；通衢支付取得支持，照发言人的说法，理由在它所及的那四成小商户：根据在规模与依赖，不在结算本身是不是基础层。规模本身换来的公共支持，要另立根据，不能借基础层保障的根据。两处可以落在同一件事上：两家作坊的工人领不到工资，碰着的是人的生计；可那时要保障的是工资与生计，不是那家结算公司。

日晷那一面，在这里接上第二节。危机前一个月，评给通衢支付的是最高一级。最高一级不是没有风险的保证；这一次评级与此后的延迟，也不由此证明评级错了，更不证明它被控制了。可这里有一个要问的结构：问的通道若本身经不起问，一个事前没被问到的风险，事后就可能由公共的支持来接；这笔贷款最后的代价落在谁身上，没有记载。第二道闸与这一节，在这里是同一件事的两头。

这一节的几句话，也分三层。支持不作背书，作出支持的一方要说明，接受支持的一方要受问，在规范层。贷款的利率、条件、期限，不得分红还是别的条件，在规则层，这里提到的都是例子。借不借、借多少、什么时候借，交给作这一判断的位置，账落在那里。

还有一种顾虑：事情急的时候，不能先要一套说明；三日的延迟里，两家作坊的工人已经领不到工资了。要支持，又要说明，又不背书，州会不会因此迟疑，反倒误了事？延迟的代价落在具体的人身上，而且等不得，这一点是真的。可是不背书不等于不支持，说明也不必在支持之前作完。支持可以先作出；说明在作出以后，答到被问的那件事上。要防的只是一样：一笔急时的贷款，事后被读成对那一家公司的认可，或者读成以后一律兜底的许诺。

这一节未定的两处：那批投资的亏损最后由谁承担；两家作坊的工资，这笔贷款到了没有。

E25　公共支持不作背书。结构依赖可以使一个经济位置取得公共支持（权利论第四篇 §V 读结构依赖；原说结构依赖在场时退出显出权力面，这里主词换成一个位置的失败，说的是它取得公共支持，自担）；取得支持，不为它既有的施为与存续发放正当性（政府论 G22「保证不是背书」；原说政府对已载入权利的保证，这里推到对积累位置的公共支持，自担）。作出支持的公共位置要说明根据、权限、对象与代价，答到被问的那件事上（制度论 Z34）；接受支持的位置，照具体条件、所负职责与权力面受问。与第三篇 E14a 分开：对一项基础层供给的保障，根据在那一项供给；规模本身换来的公共支持，要另立根据。一笔附条件的贷款，不等于新设了一份普遍的担保。救助本身是授予，归政府论；推到积累位置的那一半，是本系列的（自担）。

这一篇停在哪里

这一篇证了这些。积累不兑换为对人的基础层的支配：不把人维持生活所必需的东西换成压缩其生命基底的控制，也不拿钱、住处或工资的发放方式去锁人离开、拒绝与撤回，签了字也不改这一读（E21）。积累不兑换为对追问的控制：问的通道可以有负责的位置，不由被问的一方独自关上；账不归被记账的一方，也不归任何单一的记账者；判在结构，不在股权与收费关系本身（E22）。政经分离是双向的，是第二道闸在公共权力一级的极端情形：钱不买成审它、问它的那份政治权力，位置上的决定也不变现为日后从被它问过的一方拿到的财富；捐赠、受聘与买卖不互相代判（E23）。两道闸以外，涌现层的交换照常放开；没有撞上两道闸，也不认证其余（E24）。结构依赖换来的公共支持，不作背书；作出支持的一方要说明，接受支持的一方照常受问（E25）。

它没有证的：什么算维持生活所必需的清单，竞选财务与离任规定的细则，评级机构的监管设计，救助的条件，税。这些是规则层与具体的制度论的事，本篇只给判据，不代定。日晷那一次下调，回音壁那一次删评，州议会那一票，处长那一次受聘，本篇都只写到事实指向哪里，不替作判断的位置作终判。

交给后面各篇的：积累会自己长大，长大本身怎样读，在第五篇；回报率被置于唯一目标时是一项须核的信号（第五篇 E27），控制安排要满足自修凿的三项条件（第五篇 E30）。

留下的余项：老窑工两位同事留下的理由；日晷评审委员的名单，城市债的方法公开到什么范围；回音壁的申诉之外有没有别的门；受捐的九人与那八票；处长的新职位会不会碰到旧事；"金牌"标识怎样标明来路；通衢支付的亏损由谁承担，两家作坊的工资。这些都照记，各有去处。

本篇命题（合列）

E21　闸一：积累不兑换为对人的基础层的支配：不把人维持生活所必需的东西换成压缩其生命基底的控制（第二篇 E11、E11a），也不拿钱、住处或工资的发放方式去锁住人的离开、拒绝与撤回；这两样都不在涌现层的正常交换里。什么算维持生活所必需、基础层具体包括什么，清单在规则层（第二篇 E10）。基础层不随资产增减，不以成就交换。关系里有期限、条件、程序或退出成本，不等于退出已被封死；封死，指人不再保有拒绝、撤回持续绑定或转入可审查程序的结构通道（权利论第五篇 §III）。入职时签下的同意不独自改变读数：「判据在结构，不在宣称的自愿」（同上）；基础权利能不能由本人放弃，这一般的问题不答。原说构自己的施为对每一个内部主体都不越线（制度论 Z9），这里推到经济位置以积累所作的兑换（自担）。

E22　闸二：积累不兑换为对追问的控制。规则的制定、裁判、评级与记账，是问的通道。积累不得兑换成这样一种地位：被问的一方能够独自关闭对它自身的追问、纠错或必要的兑现。评级、记账与裁判可以有具体负责的位置；作出一份记录或判断，不取得独占其终审、取消外部追问的权能；记录要有明确的范围与可复核的来路，怎样配置在规则层。账不归被记账的一方，也不归任何单一的记账者；这一句不禁止单一机构记一份局部的账。独立照「独立于被追问的一方」核，是功能义（制度论 Z34）；判捕获看结构，不看机构的名称与设在哪里，提名、同党一类关系也不自动证明捕获（政府论 G38）：没有股权交叉，不自动证明独立；现有交叉，也不自动证明捕获。三项条件的出处在 F6 命题 IV，那里只立条件，不立应当；应当据制度论 Z34。道德律第三篇 §1.3、§8.3、§8.4 对中心化表述的禁止与分布式的防御，只取"单一的中心可被钻空子"这一结构作同形；那一篇立在纯 15DD 共同体的理想型里（§0.6），所说的保护是结构性的，不靠制度（§8.3）；推到混合现实里记录的可复核来路，自担。风险与追问的入口，不当作已证的捕获。原说制度自身的追问通道与政府里的捕获，这里推到经济位置所及的问的通道（自担）。

E23　政经分离，双向，是第二道闸在公共权力一级的极端情形。钱不买成审它、问它的那份政治权力；位置上作出的决定，也不变现为日后从被它问过的一方拿到的财富。两个方向换的都是问的那条路；事前的捐款与事后的受聘，是本例展示的两种时序，是否发生兑换仍须核，时序本身不单独证明它。收钱一侧所涉授予权力的运用，照政府论 G38 读，两处互相指引，各写各的。公共权力的每一级都一样，大到一国，小到一个小区的业主自治；这是分权制衡的道理在经济位置上的延伸：经济的权力与政治的权力分开站，才能彼此制衡。要核的是有没有买走问它的那条路，或者拿问的那条路换了报酬。买走问它的路，与普通的受聘、捐赠，不互相代判：合法的捐赠不因此证明没有买，有捐赠也不因此证明已经买了；离任后的受聘同此。判在结构与时序，不在关系本身：提名、同党一类关系不自动证明捕获（政府论 G38；原说政府里的提名与同党，这里推到捐赠与受聘，自担）。拿钱买一部法给自己开门，读在关口与制衡（第二篇 E7；第三篇），不在这一条。

E24　闸以外，涌现层的交换照常放开。名声换来机会，资本换来更多的资本，资质换来职位，上限之内的捐赠与合规的受聘，本身都不是问题；不以"可能被兑换"为由封住涌现层的交换。钱买得到标识与记录，买不到声誉（道德律第三篇 §1.1、§2.2；见第一篇 E4）。两道闸只说本篇处理的兑换限制：没有撞上两道闸，不认证一次交易在一切方面正当，也不取消本系列别处守线、可被追问、保障与给条件的要求。

E25　公共支持不作背书。结构依赖可以使一个经济位置取得公共支持（权利论第四篇 §V 读结构依赖；原说结构依赖在场时退出显出权力面，这里主词换成一个位置的失败，说的是它取得公共支持，自担）；取得支持，不为它既有的施为与存续发放正当性（政府论 G22「保证不是背书」；原说政府对已载入权利的保证，这里推到对积累位置的公共支持，自担）。作出支持的公共位置要说明根据、权限、对象与代价，答到被问的那件事上（制度论 Z34）；接受支持的位置，照具体条件、所负职责与权力面受问。与第三篇 E14a 分开：对一项基础层供给的保障，根据在那一项供给；规模本身换来的公共支持，要另立根据。一笔附条件的贷款，不等于新设了一份普遍的担保。救助本身是授予，归政府论；推到积累位置的那一半，是本系列的（自担）。

参考文献

以下 SAE 文献一律为概念 DOI，始终指向最新版本。逐句核验所据的版本记录，见随稿的转述审计表。

一 · 本系统

秦汉. (2026). 制度如何可能：从主体间余项到共构框架（SAE 基础论文 Paper 6） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.19328662

秦汉. (2026). 论承认性权利与权力的界面与转化（SAE 权利论系列 Paper 4） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.21093939

秦汉. (2026). 论基础权利的限与侵（SAE 权利论系列 Paper 5） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.21212595

秦汉. (2026). 论权利的涵育与历时展开（SAE 权利论系列 Paper 6） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.21252863

秦汉. (2026). 声誉经济 · Dao（SAE 道德律系列 Paper 3） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.20046168

秦汉. (2026). 应当说给谁（SAE 制度论 第一篇） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.23269635

秦汉. (2026). 不越线（SAE 制度论 第二篇） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.23270384

秦汉. (2026). 可被追问（SAE 制度论 第六篇） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.23271324

秦汉. (2026). 保证（SAE 政府论 第三篇） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.23413112

秦汉. (2026). 追问与承担（SAE 政府论 第六篇） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.23413511

二 · 外部文献（原文，A 档）

Smith, A. (1776). An Inquiry into the Nature and Causes of the Wealth of Nations. Project Gutenberg, EBook #3300.（Book I, ch. XI, Conclusion of the Chapter）

三 · 外部文献（原书未核，据二手转述，C 档）

Walzer, M. (1983). Spheres of Justice: A Defense of Pluralism and Equality. New York: Basic Books.（据 den Hartogh 1999 转述）

den Hartogh, G. (1999). The Architectonic of Michael Walzer's Theory of Justice. Political Theory, 27, 491–522.

Sandel, M. J. (2012). What Money Can't Buy: The Moral Limits of Markets. New York: Farrar, Straus and Giroux.（据 Kennedy 2012 转述）

Kennedy, D. M. (2012). The Root of All…: Michael Sandel's salvos against the conventional wisdom of what markets should—and should not—do. Democracy: A Journal of Ideas, No. 26 (Fall 2012). https://democracyjournal.org/magazine/26/the-root-of-all/

接口

主承： 制度论 Z9（推到经济位置以积累所作的兑换，自担）、Z34（推到经济位置所及的问的通道；说明之责）；政府论 G38（推到经济位置，及捐赠与受聘，自担；政经分离后一方向拿钱的一方照此读）、G22（推到对积累位置的公共支持，自担）；权利论第五篇 §III。

另用： F6 §3.2 命题 IV（三项条件的出处，射程照原句）；政府论第六篇第二节（三问）；权利论第四篇 §V（结构依赖，推到失败所及，自担）；权利论第六篇 F26（殖民的窄义）；制度论第一篇第八节（不装成涵育的越线也在前两条里）；道德律第三篇签名句、§0.6、§1.1、§1.3、§2.2、§8.3、§8.4（同形，理想型；推到混合现实里记录的可复核来路，自担）。

对接： 斯密《国富论》第一卷第十一章结论，经营者的立法提议（A 档，据 Project Gutenberg #3300；第三节）；沃尔泽《正义诸领域》第四章"受阻的交换"（C 档，据 den Hartogh 1999）；桑德尔《钱买不到的东西》（C 档，据 Kennedy 2012 书评）。

本系列： 前置篇（不以减少为目的；不以分配结果为终点；三层）；第一篇 E1、E4、E5；第二篇 E7、E10、E11、E11a、E12、E12a；第三篇 E14a、E18；第五篇 E27、E30；第六篇。

交给第五篇： 积累会自己长大；长大本身怎样读。