Similes of Symmetry

The cloth went another way. In a document of 1320 discussed by Jeremy Edwards and Sheilagh Ogilvie, Italian merchants bought cloth outside the Champagne fairs and sent it to Italy without bringing it through them. The financing was still arranged at the fairs.¹

There is something arresting about this incomplete departure. The familiar account of an institution's decline encourages us to look for whatever replaced it, as though the successor had taken possession of an office and all its contents. Here the goods have ceased to make the journey, while an arrangement concerning their purchase continues to use the old place. A merchant who no longer needs to bring his cloth to a fair may still need someone he can find there. The expense of carrying merchandise and the difficulty of arranging finance need not disappear together.

Nor were resident agents and fairs incompatible alternatives. Edwards and Ogilvie observe that Italian merchants were using agents at the fairs during their flourishing years. They dispute the account in which sedentary commerce simply superseded the older gatherings, assigning a central role in the fairs' decline to destructive public policy.¹ What interests me here is the separability within an arrangement that could nevertheless hold together. People can cease to need an institution for one purpose, find it useful for another, and remain dependent upon it for a third. The word replacement makes these changes sound more complete than they are.

We have followed records that allowed people to act without assembling everyone who knew the business. An absence became tolerable because something could be inspected in the absent person's place. This achievement also permits a further absence, less welcome to an established intermediary: the person who once had to come back may be able to go elsewhere. A useful result becomes available apart from the continuing relationship through which it was obtained. What must accompany it is not the whole institution, reproduced in miniature. But neither can the relationship's unfinished business be disposed of by sending a copy.

Still payable at the fair

The development of the bill of exchange depended on people who remained in distant places and could act for others there. Meir Kohn's account puts the organization before the letter: permanent connections abroad made it possible to order a payment from someone who would be present to make it. In the uses he describes, usance could fix the interval before payment without waiting for the next fair. Later financial fairs again used periodic settlement.² The calendar was not marching toward liberation. Merchants were finding different ways to make their undertakings meet.

A letter relieved its sender of a journey because someone else had work to do. There is no diminution of the achievement in saying so. To make a distant undertaking possible without requiring its initiator to appear in person is already a considerable freedom. It gives that person time for other business, or lets business begin that would otherwise have been impracticable. But the freedom belongs to a particular participant. For the correspondent, there are instructions to interpret, funds to find, a payment to make or refuse. The arrangement has redistributed attendance. It has not abolished it.

The difference matters when the document fails to obtain what its bearer expected. In the historical procedures described by Kohn and by James Bolton and Francesco Guidi-Bruscoli, a protest recorded a refusal and could support contingent recourse against the relevant party. It did not turn everyone whose name appeared on a bill into an unconditional debtor.³ Carrying the refusal elsewhere was useful because other people could do something with it. The notary's record did not pay the money. It made a particular failure available to a further proceeding.

A disappointment could travel too, without remaining solely in the keeping of the person who disappointed you. Departure need not mean abandoning the means of pursuing an obligation. Yet an instrument that enables pursuit can also keep a person within reach long after the cooperation that made performance possible has broken down.

The historians recount a particularly ugly conjunction in 1297: French authorities confiscated Flemish merchants' goods; merchants unable to make promised payments after the seizures were then subjected to a fair-ban until they paid their debts.¹ The machinery for recognizing default continued to work upon people whose means of performance the authorities had helped destroy.

An institution's ability to maintain a claim across disruption is therefore no assurance that the burden should continue unchanged. We have to ask what happened to the conditions of performance, who altered them, and what authority can now determine the consequence. Otherwise the virtue of continuity becomes an excuse not to examine an intervening act of power. The goods have been seized, but the account remains admirably legible.

These are different departures from the fair: a route bypassing it, financing arranged through it, a merchant excluded from it. To call them all a shift in verification cost would obscure the very thing the comparison ought to make visible. The departure of the goods leaves a financial relationship in place; exclusion can leave a merchant liable within an arrangement from which he has been barred. The institution's continued usefulness and its capacity to impose a burden have to be followed separately.

After the account is closed

Consider, now, a hypothetical business leaving a service through which it has taken orders, received payments and accumulated a trading record. A downloadable account is a substantial improvement over being allowed to view the past only while paying to remain a customer. Another provider might use the exported entries to prepare the books; a prospective customer might inspect evidence of work already done. The business need not recover every conversation that produced each entry. Where the records and their support are adequate for the new purpose, useful reliance can begin without reenacting the whole relationship.

There is an immediate temptation to make the export stand for independence. But an unpaid balance is not settled by exporting its amount. A copy may make a disputed charge easier to challenge without supplying a way to obtain its correction. Nor does an obligation to correct an account evaporate when the person affected ceases to buy the service. A transition can transfer the work, assign responsibility to a successor, or provide for the winding up of outstanding claims. The important condition is that leaving should not itself deprive someone of what the arrangement still owes them. Which claims survive, and in what form, depends on their terms and the authority governing their alteration; continued subscription is a different question.⁴

Much of the attraction of an institution lies in not having to arrange these matters separately. People buy a relationship because the separate purchases would exhaust them. A customer may have no desire to select one service for keeping records, another for establishing their origin, a third for recovering unpaid money, and a fourth for deciding whether the first three have done their work. Remaining with a capable provider can be the sensible use of one's freedom. There is no obligation to make a hobby of independence.

An alternative acquires political significance when it allows that continued relationship to be chosen on terms that can be refused. It need not reproduce every feature of the incumbent. A less elaborate service may be quite sufficient for someone who no longer needs the old bundle, while an arrangement among several providers may make the same work available in a different way. The possibility of leaving can affect bargaining before anyone leaves. It can also prove worthless if customers cannot reach the alternative, finance the transition or obtain recognition of what they bring. These are conditions to establish about a route, not reasons to assume that only a complete institutional replica could count as one.

The former provider has interests here too. Preserving an archive, answering questions about old transactions and administering unresolved claims can consume real resources after new business has gone elsewhere. Someone must pay for the duties that remain. But the fact that a duty costs money does not give its bearer an unlimited right to attach unrelated conditions to its performance. If correcting a charge is already owed, the institution must not make the correction depend on purchasing another year of services. A price for new work and a demand made against an existing obligation are different claims upon the customer.

This distinction is especially important for people who were never customers. A credit report describes someone for the benefit of someone else. When that report moves, the subject may acquire no new choice at all; other institutions have simply acquired another reason to act upon them. The person who wants evidence of completed work to become usable elsewhere and the person who wants an old accusation to stop circulating are not making contradictory demands. One seeks recognition of an achievement. The other disputes an assertion, or the authority for its continued use. The difficulty begins when the ease of moving a record decides both questions in advance.

There are records we need to retain precisely because an institution would prefer departure to end the matter. There are also accurate records whose availability supplies no authority for a new adverse use. A discharged obligation should not continue to operate as a debt merely because a copy survives. These distinctions do not require a fresh jurisdiction for every ordinary use: an existing lawful arrangement can authorize later reliance. They require that the record remain distinguishable from the powers under which something may be done with it. Losing the paper is not forgiveness; keeping it is not a sentence.⁵

Accepted elsewhere

The hardest part of leaving may be neither obtaining the file nor paying for the journey. It may be persuading another institution that what one brings is something it can use. A record of competence, for example, need not command the same conclusion from every recipient. The work required in a different practice may be different; an examination may have tested something narrower than the document's title suggests. Equally, insisting that everyone begin again can protect an established profession from people already capable of doing the work. Recognition is an occasion for judgment, and a position from which competition can be refused.

We should not settle this conflict by choosing, in advance, between total portability and unanswerable discretion. An institution can be required to give due weight to evidence without being required to treat unlike achievements as identical. It can explain what a previous assessment establishes for its purposes, identify what is missing and offer a practicable way of establishing the rest. It may conduct its own investigation. What needs a defense is the demand that a person repeat the old route when the relevant work has already been adequately done and can be examined. The institution should be able to name what the repetition adds beyond obedience to its preferred itinerary.

Titling exposes a further difficulty. Hernando de Soto's insistence on recognition was never simply an instruction to print better certificates. In his 2001 interview with the Federal Reserve Bank of Minneapolis, he describes property through social agreements and the legal organization of claims.⁶ The documentary burden can be reduced while the disagreement remains: another occupant, a rival account of possession, a boundary that one office accepts and a neighbor contests. It would be a strange liberation that made the applicant's claim cheap to establish by making everyone else's expensive to hear.

Nor does an accepted title compel a lender to advance money. Recognition of a right and a profitable use of it belong to different inquiries. A better record can allow an undertaking that ignorance previously prevented, but it cannot supply demand for what will be produced or funds on terms the borrower can meet. We shall encounter this separation again: an obstacle can be removed without creating the opportunity that seemed to lie just behind it. That does not make its removal worthless. It prevents us from charging the obstacle with every disappointment that follows.

Consider a hypothetical system that can recover the documents needed to support an application, compare their terms and identify the discrepancy an official needs to resolve. If the work is done adequately, someone who once had to employ an intermediary for the search and comparison may no longer need to do so. The resulting inquiry can still call for a professional judgment; a recipient may still obtain new grounds through examination or testing. But the surviving need does not restore a necessity for every operation formerly sold alongside it.

Respect for expertise does not require preserving avoidable dependence upon its custodian. An expert can contribute something indispensable while other parts of the transaction become available more widely. The recipient's new means of checking a result may improve the relationship with the expert, make another expert a practical option, or permit some work to proceed without either. Whether the result was produced statistically, through a declared rule or by a person does not decide which of these changes has occurred. Evidence supporting a result and responsibility for its use must each be established in the arrangement; a deployment can provide both.⁷

There is a modesty in useful equivalence that its rhetoric often conceals. To accept two things for a purpose is to leave much about them undecided. The receiving office does not need to agree with the sender about everything, and the person carrying the record should not have to carry an entire institutional world along with it. What matters is the correspondence on which the next act depends. An explanation that survives inspection can make that correspondence usable without requiring submission to every authority that helped produce it.

Similes of symmetry names these limited correspondences. Their promise is not that unlike institutions can be made identical. It is that some of what they have learned or established can become useful beyond them, under conditions another can understand. The same accomplishment that makes cooperation possible can make a particular cooperative relationship less compulsory. Its limits are not a refutation of that freedom. They tell us what has actually become available and what remains to be obtained.

Who pays for the alternative

Yoram Barzel begins his account of measurement costs with an orange. Its weight is easier to learn than the quantity and quality of the juice inside. Much of the organization of exchange follows from such discrepancies between what people can readily inspect and what they want to obtain. In his account, repeated dealings, reputation and warranties reduce the need to measure everything afresh.⁸ The dependable intermediary has an economic reason to exist: its customers can get on with other things because they have reason to accept what it supplies.

The difficulty is what follows when some of that knowledge becomes usable apart from the intermediary. One cost may fall while the benefit goes to someone who controls a different scarce condition. An inexpensive assessment is of limited competitive value if only one organization can give access to the customers who require it. Conversely, several institutions able to use the same evidence may make a new undertaking possible without abolishing the expense of maintaining the evidence. Useful work can become less exclusive without becoming free.

This is why a gross fee will not disclose the rent within it. The coherence fee, as the term is used here, concerns work needed to keep reliance warranted across settings. The trust tax is a premium enabled by control of an accepted route to that use. To distinguish them in an actual arrangement requires an account of the service, its risks and the alternatives available. Finding a cheaper check does not complete that account. But neither should the inability to price every component suspend criticism of a provider that withholds a usable record or makes an existing remedy depend on continued custom. A supported objection to its power need not wait for an estimate of its monopoly profit.⁹

An alternative also needs a means of staying in business. The people who establish a rival route must obtain equipment, organize work and meet expenses before the hoped-for traffic pays for them. Existing users may want the route to be available without wanting to finance its beginnings. An incumbent may be expensive and still be the only provider able to carry the obligations already accumulated. These are reasons to take the construction of an alternative seriously. They are not reasons to grant the incumbent a perpetual franchise over work others could undertake.

The achievement pursued through this volume is a form of continuity that does not require captivity. People should be able to make further use of what has been established without purchasing again every part of the relationship that established it. They must also remain able to pursue what is still owed, and to resist consequences that the record does not authorize. An institution that has done useful work deserves to have that work understood. It does not thereby acquire title to all the future cooperation made possible by it.

Yet the choice to leave is thin if no one can afford to build another route. As some of the work becomes reproducible, bargaining can move toward what remains difficult to obtain: finance, productive capacity, access to a market, permission to operate. Those positions may be inherited, purchased, contested or newly created. Their owners can collect even when they did none of the work that made the new possibility apparent. A claim has found a way out of its first keeping; the person relying on it must still find someone willing and able to act. That person is entitled to ask more than whether the next gatekeeper has better records. On what terms can the gate itself be built—and can anyone else afford to build one?

Witness claims

These pointers identify related formal objects, not proofs of the institutional argument. Applications require an explicit correspondence and the model's assumptions. A query or equivalence object does not provide a forum, a remedy or a commercially viable alternative.

WC-08-01 property="composition"

Where a platform's power depends on keeping identity, reputation, payment or dispute records local, genuine portability can create a route around the incumbent. Its effect depends on access, recognition, transition costs and the work still required. The route-shift explanation fails for a particular claim of reduced dependence if no practicable alternative has been created. It predicts neither an inevitable replacement nor a universal sequence of concentration and competition.

Formal models: → A21 (coherence cost framework) or A24 (predicate package)

WC-08-02 property="recourse"

Preserving effective recourse after a relationship ends requires more than retaining a copy. A relevant forum must be able to consider the evidence and act upon the consequence; it may be the existing forum, a successor or another authorized institution. Continuity of remedy does not require transferring every function to a single new provider. If effective remedies operate without the functions asserted as necessary in a particular application, that application over-specifies the problem.

Formal: → A23 (identity maintenance) or A10 (witnessed sameness)

WC-08-03 property="binding"

A model output may support an inquiry without identifying who answers for its consequential institutional use. Evidence for a relation and answerability for an act are distinct. A deployed system can supply both. A criticism that a particular deployment lacks adequate grounds or responsibility fails when that deployment demonstrably supplies them; their joint provision does not erase the distinction.

Formal: → A1 (commitment set), A2 (provenance judgement), A10 (witnessed sameness)

WC-08-04 property="verification-cost"

Automation can reduce one component of verification cost while leaving or increasing access, interpretation, correction and governance costs. Lower technical cost does not by itself reveal a measurable rent or disperse power. An empirical claim about the trust tax must identify the work, the party controlling access and the alternative against which a premium can be assessed. An objection to exclusion or an obstructed remedy does not depend on quantifying that premium.

Formal model: → A21 (coherence cost framework)

WC-08-05 property="temporal-jurisdiction"

An accurate record can remain available after a particular authority for adverse use has ended. Existing authority may also validly cover continued use. Availability alone settles neither question. The chapter does not infer mercy from physical decay, require a fresh jurisdiction for every reuse or specify the later volume's rules for temporal release.

Source notes

Footnotes

1. Jeremy Edwards and Sheilagh Ogilvie, “What Lessons for Economic Development Can We Draw from the Champagne Fairs?”, Explorations in Economic History 49 (2012), pp. 136, 139. Their historical analysis is inspected here, rather than the medieval originals. The distinction among dependencies is this chapter's interpretation. ↩ ↩² ↩³

2. Meir Kohn, “Bills of Exchange and the Money Market to 1600,” Dartmouth Economics Working Paper 99-04 (1999), pp. 1–4: pp. 1–2 and p. 4 n. 10 on distant organization; p. 3 on usance and n. 7 on a return to periodic settlement. Kohn's synthesis distinguishes different instruments and markets. The discussion does not make correspondent networks the cause of the Champagne fairs' decline or claim that all trade adopted the same calendar. ↩

3. Kohn, pp. 3–4; James L. Bolton and Francesco Guidi-Bruscoli, “Your Flexible Friend: The Bill of Exchange in Theory and Practice in the Fifteenth Century,” Economic History Review 74 (2021), pp. 876–877. Acceptance, refusal and the parties' positions affected recourse. These passages support the limited historical comparison; they establish no universal remedy attached to every transferable record. ↩

4. The departing business is a hypothetical inquiry into obligations and alternatives, not a report about any named platform or a statement of a particular jurisdiction's law. The normative argument concerns duties already owed under an arrangement, their authorized alteration and their performance after departure; it creates no general entitlement to all records, all customer data or uncompensated new services. Other people's privacy and rights continue to govern disclosure. No quantitative rent or switching-cost estimate is asserted. ↩

5. Satisfaction, discharge and authority for later use are distinct from preserving the evidence of what occurred. This carries forward the selected Chapter 7 and the mature Volume III treatment of records. The chapter does not supply a new expiry rule, reset an existing threshold or require erasure of evidence against an institution. “What is still owed” can be altered or discharged through an authorized proceeding; it does not mean that every original claim must persist unchanged. ↩

6. “Interview with Hernando de Soto,” The Region, Federal Reserve Bank of Minneapolis (2001), especially the exchanges defining property as a social contract and discussing law, description and dispute settlement. This is evidence of de Soto's account, not independent verification of the interview's valuations or the causal effects of titling. The earlier numerical inventories are not used. Recognition, creditworthiness and the availability of a willing lender remain separate. ↩

7. The application is conditional and hypothetical. A check establishes the relation it actually tests under its assumptions; neither its presence nor a statistical origin decides the adequacy of the whole deployment. Selected Chapters 4–7 develop institutional adoption, transformations, evidentiary support and reliance. The present chapter follows the resulting alternatives rather than repeating their examples or reinstating a hierarchy of witness functions. ↩

8. Yoram Barzel, “Measurement Cost and the Organization of Markets,” Journal of Law and Economics 25 (1982), pp. 27–29 and 46–48. His models connect costly measurement with arrangements including warranties, reputation and integration. He describes the evidence as illustrative rather than a test of the model. The present argument grants that contribution; it claims no new economic theory of transaction costs or interoperability. ↩

9. The terms retain the selected Chapter 3's distinction between necessary work and an asserted rent. They are not measurable components furnished by a posted commission, and the necessity of particular work does not establish one provider's exclusive right to perform it. A21 remains a formal framework whose application requires stated correspondences and empirical inputs, not a demonstrated economic law. ↩