Courts Without Thrones

The Soul selects her own Society —
Then — shuts the Door —
To her divine Majority —
Present no more —

— Emily Dickinson, “The Soul selects her own Society”

In 1293, William of Abingdon appeared at the fair court of St. Ives to recover eleven pounds owed to a London cloth merchant, John of Abingdon. The defendant, William Martin, challenged his authority to speak for John, and the hearing was postponed. When they returned two days later, Martin had a document which he had not produced at the first appearance. It purported to be John's acknowledgment that the debt had been paid. William called it false and asked for a jury to examine it. Martin objected that William could not deny another man's deed. The court accepted the objection, discharged Martin, and fined William for his false claim.¹

The document may have been forged. Its arrival is suspicious, and William's protest gives us a reason to doubt it. The temptation is to finish the case on his behalf, a small retrospective victory which the surviving account will not sustain. The court reached a decision about the debt without permitting an inquiry into the disputed acquittance. A rule concerning who could speak had determined what could be examined. William had found the debtor and brought him before an institution capable of making him pay. He left owing a fine himself.

Much of the institution was working. There was a place to bring a claim, a procedure for answering it, a record of the proceeding, and an authority able to give the result a consequence. A creditor could pursue a debt originating elsewhere while the parties were gathered at an English fair. What had been made possible for him depended on several kinds of work, none of which the promise to pay could have accomplished by itself.

The fair belonged to Ramsey Abbey. The abbot's officers administered it under royal authority, collected its revenues, and could seize goods or take defendants into custody. Merchants contributed knowledge and participated in judgment, though the arrangements of the court belonged to a wider world of local jurisdiction and lordship. Royal privilege, mercantile practice, the authority of the abbot, and the presence of property that could be reached all entered the proceeding. St. Ives gives us no republic of merchants exempt from political power. It gives us a court in which political power had acquired a local and commercially useful form.¹

Calling such an arrangement public or private can become an obstacle if the word settles the inquiry too early. The merchant who wanted his money needed to know where the debtor could be found, what counted as evidence there, who could speak, and what property might answer for a judgment. Whether these functions belonged to one institution mattered less to him than whether they could be made to work together. It matters a great deal to anyone trying to understand how the arrangement governed. The office that knew how to hear the claim might possess only a small part of the power that made its hearing effective.

There is an understandable wish to recover from this history an alternative to the state as we have come to know it. Commercial life has often found usable forms of order where a distant authority supplied little practical help. People devised instruments, established forums, and gave promises consequences beyond the immediate encounter. That ingenuity deserves its place in the account. So do the privileges they obtained, the force on which they relied, and the people for whom their institutions offered no useful remedy. A history attentive to those circumstances can teach us how authority was divided without requiring us to believe that power had gone away.

The fair is leaving

A fair court sat among people preparing to leave. Goods were being sold, debts settled, and journeys resumed. A witness available today might soon be on a road where no officer of this court could readily find him. Property within reach could leave with its owner. Even an eventual victory might be worth little if obtaining it consumed the season in which the claimant needed the money. The dusty feet preserved in the name piepowder belong to this practical history of adjudication. The court's calendar was partly imposed by the people passing through it.

Haste was therefore capable of serving either party. It could prevent a debtor from outlasting the creditor; it could also close an inquiry before an absent witness or disputed document received attention. There was no general virtue in making the court slower. One has to ask what the delay would permit, who would pay for it, and whether the means of enforcing a judgment would still be there when the delay ended. The same interval can be an opportunity for inquiry and an opportunity to escape.

The records sometimes preserve the modest expedients by which a case survived its venue. In 1316, Ralph of Houghton was pursuing a debt against John Christian at St. Ives. Goods had been seized, including a barrel, two hogsheads, and a tankard, but the defendant still had not appeared. Ralph asked that the case, together with the goods held against it, pass to the court of the weekly market. The request came near the end of the fair. The roll does not give his reason, although the timing makes the approaching departure of the court a plausible one.²

A barrel and a tankard are an unpromising foundation for a theory of jurisdiction. They make the requested transfer rather more intelligible, though. Ralph wanted the goods already held against his claim to remain available as the case moved elsewhere. The receiving court would acquire unfinished business and property under restraint. Someone would have to account for what had been done, preserve what had been taken, and continue from a proceeding begun by others. The work of making a judgment possible extended across the interval in which there was, as yet, no judgment.

The Champagne fairs organized a larger circuit. Six fairs rotated through four towns, Lagny, Bar-sur-Aube, Provins, and Troyes, bringing northern and Mediterranean commerce into repeated contact. Merchants could transact in one part of the cycle with obligations falling due in another. The counts' protection, the fair wardens, accommodation, measures, and places for storing and selling goods helped sustain these gatherings. Trade acquired a recurring geography in which people, property, information, and authority could meet.³

That concentration reduced some of the difficulties of dealing with a stranger. He might have no connection to one's home city, yet for a time his goods lay in the same town, his conduct could be observed by other traders, and an officer could act against property he valued. Beyond the town there were other fairs and other counterparties. The prospect of returning gave a present dispute a longer consequence. It also gave the people who administered the venue a reason to care about what happened after this particular dispute had ended.

Paul Milgrom, Douglass North, and Barry Weingast isolated one part of this arrangement in their model of the law merchant. Merchants who would not meet often enough to discipline one another through repeated bilateral dealings could consult a judge's record. A defection reported and left unremedied would affect future opportunities with other traders. Cooperation could then be supported by the value of transactions still to come.⁴ The model explains how information about one encounter may change the incentives in another. It leaves a historical question about which institutions actually supplied that information, enforced the sanction, and restrained those who administered it.

The counts and their officers occupy more of the answer than the image of a private judge suggests. Edwards and Ogilvie's reconstruction of the fairs emphasizes political protection and contract enforcement, including the circumstances in which public authority supported foreign merchants against powerful borrowers.³ Royal or princely involvement was not a small imperfection in an otherwise private mechanism. It affected whose property could be reached and whose refusal could be made costly. A reputation sanction might work very well among merchants and be a poor answer to a debtor who could prevent those merchants from entering his territory.

The interests of the protector could also change. A ruler who benefited from attracting merchants might provide dependable access; the same office could distribute favors, tax selectively, or make a community's presence conditional on demands it had little power to resist. Commerce did not escape politics by becoming useful to it. Usefulness supplied a reason for protection whose strength depended on the ruler's other interests and means of revenue. When the routes, political arrangements, and financial practices supporting the fairs changed, their old position could not be preserved by commercial habit alone.

Nor did a wider reputation network remove the problem encountered by William of Abingdon. Suppose another venue received only the news that he had brought a false claim and been fined. The result would be easier to circulate than the procedural history that produced it. A later recipient might reasonably assume that someone had examined the substance before recording the conclusion. The authority of the first court could travel through that assumption even if its reasons did not. Better communication would enlarge the consequence of whatever the first proceeding had managed to establish, including its mistakes.

What a letter knows

The Cairo Geniza has left us a commercial world through a different kind of preservation. At the Ben Ezra Synagogue in Old Cairo, worn books and documents accumulated over centuries in a chamber used for writings withdrawn from use. Reverence for Hebrew script helped bring ordinary correspondence into the same keeping as religious texts; much of the writing was Arabic expressed in Hebrew characters. Business letters survived among materials whose protection had purposes quite apart from the settlement of accounts.⁵ A letter can cease to matter in the affair for which it was written and, centuries later, become the most substantial evidence that the affair ever occurred. Its survival need not reflect anyone's judgment that this transaction deserved remembrance. The Geniza permits us to recover commercial practices partly because writings were spared destruction for reasons a commercial explanation would otherwise leave out.

The letters were addressed to people who already knew a good deal. A correspondent could leave a circumstance unexplained because the recipient had lived through it, knew the family, or would recognize what a delay implied. The modern reader arrives with other letters and other questions, trying to recover some of that competence. Even a large collection leaves silences whose meaning is uncertain. An institution rarely mentioned may have been irrelevant, expensive, or so familiar that no one thought to explain it. An accusation may tell us more about an endangered relationship than about the conduct it alleges.

Avner Greif read the correspondence of the Maghribi traders as evidence of a multilateral reputation mechanism. A merchant who entrusted goods to an agent abroad could not continuously observe their handling. Other merchants might nevertheless learn how the agent had behaved and withhold future employment after misconduct. In Greif's account, the commercial future exposed to this collective response helped secure performance in a present relation.⁶ An agent could have something substantial at risk without placing it in the principal's hands before the goods departed.

The attraction of the mechanism is easy to understand. Where information circulates among people who expect to remain in business, a private disappointment need not remain private. Other people have reason to listen, since the same conduct could affect them later. They need not share the injured merchant's affection or animosity to care whether his account is true. The network can make each person attentive to transactions in which he had no direct part.

Whether the surviving documents establish the particular coalition Greif describes has been disputed. Jeremy Edwards and Sheilagh Ogilvie argue that the evidence does not support an exclusive group with clearly defined membership and that the traders used legal means to enforce agency relations. Greif's response contests their reading of the legal evidence and defends the importance of multilateral reputation.⁷ Their disagreement bears on how much explanatory weight can be placed on exclusion, on formal enforcement, and on a boundary around the group.

The distinction between personal knowledge and formal evidence remains useful without requiring two pure civilizations to embody it. A trader can consult a correspondent and retain a legal instrument. A court can matter even when a relationship makes litigation unnecessary. Conversely, the existence of courts does not establish that they offered a practical remedy for every overseas agency dispute. The available combinations have to be reconstructed from the relations and jurisdictions in which people were actually acting. It would be a strange way to honor the richness of the Geniza to turn its merchants into representatives of an institutional choice they never had to make in that form.

Formal instruments change what a stranger needs to know. A notarial act can provide names, terms, and an authenticated event that a receiving institution knows how to examine. Its holder need not reproduce the whole history of a friendship before making a claim. That economy can widen access, though it depends on the office, the cost of obtaining its work, and the willingness of the destination to recognize it. Someone pays for the form and for the institutions that let it mean something elsewhere.

There are freedoms in this impersonality which a defense of community can forget. A person may wish to trade without making his family, manner of life, or loyalties available to the judgment of everyone with whom he does business. He may have left a place where those things were too well known, or known in a way he could never correct. A document that establishes a limited entitlement can permit an encounter to remain limited. One can owe a stranger payment without owing him an account of oneself.

Yet limited forms can become painfully inadequate when the relation breaks down. A correspondent who knows the person may understand a failure as illness, misfortune, or an uncharacteristic mistake. A distant office receives a date and an unmet obligation. Sometimes the office's narrowness protects the claimant against the indulgence shown to a well-connected debtor. Sometimes it prevents the debtor from making a relevant circumstance understood. Neither intimacy nor impersonality can be trusted to distribute this advantage justly on every occasion.

The same person may need both. He wants a commitment to be intelligible beyond the circle that knows him, and he wants room to explain what the commitment's portable form has omitted. An institution able to receive the form and give it ordinary effect still needs people who can hear a reason why its ordinary treatment would be wrong in this case.

Among the dealers

Lisa Bernstein's 1992 study of the New York diamond trade describes an industry in which arbitration, reputation, and access to a commercial community did much of the work that general contract law might otherwise be expected to do. The Diamond Dealers Club supplied a specialized forum, and its connection to other bourses allowed judgments to have consequences beyond the room in which they were given. This was an arrangement observed in a particular trade, with public law still surrounding it.⁸

Among the details in Bernstein's account are letters posted on club bulletin boards by dealers who believed themselves injured by baseless gossip. Some rebuttals were also circulated in the trading hall or on Forty-seventh Street. A reputation system required these awkward acts of self-defense alongside its more orderly procedures. A man whose commercial standing was under attack could need to address people who had heard the accusation from someone else, perhaps without ever discovering how many had heard it.⁸

The notice on the wall belongs to the institution as much as the arbitration rule does. It is an attempt to catch a claim while it is still moving. The person answering may sound angry or self-interested; it would be surprising if he were altogether detached. Reputation gives a community an inexpensive means of remembering conduct, but the apparent economy partly depends on how much work a disputed memory obliges its subject to perform.

Expertise helps a private forum because the dispute need not begin with an education in the trade. Adjudicators may already understand the conventions of delivery, the significance of a defect, or the way an apparently informal agreement is ordinarily concluded. They can attend to a difference which a generalist would scarcely notice. The same familiarity can make an assumption difficult to challenge. A practice may appear reasonable to everyone whose position has been formed within it, while imposing a cost on the newcomer whom the practice has kept at the edge.

Competition among forums can improve the situation when both sides have a practical influence over where disputes will be heard. The prospect of losing participants gives a tribunal reason to care about competence and fairness. It is less reassuring when the forum is selected by the party that brings it repeated business and accepted by the other as a condition of obtaining something else. The occasional participant may encounter the institution only after the choice has become costly to revisit. An honest adjudicator can work inside an arrangement whose commercial dependencies run mostly toward one side.

We should distinguish that pressure from a finding that every decision favors the repeat participant. A forum can decide many cases fairly and still fail to confront a rule which supplies its business. It may correct departures from the contract while treating the terms through which it acquired authority as settled. The relevant dependence can remain outside the questions its adjudicators are ordinarily asked to answer. No private instruction to decide dishonestly is required.

An exclusion sanction has a related limit. It works against someone who needs continued access and can be excluded at a tolerable cost to the people imposing it. As one participant becomes indispensable, the cost of his exclusion rises. Others may still disapprove of his conduct and even record it accurately, while finding that the sanction they would impose on a smaller participant would now damage too much of their own business. A rule can survive this change word for word and bind very different people in practice.

The problem is familiar wherever a service accumulates the conditions of its users' livelihoods. A merchant may enter for useful access to customers and spend years building relationships there. The operator's authority then reaches something which did not exist in full when the merchant first accepted the terms. Departure means more than declining the next transaction. The history of previous transactions, the means of reaching customers, and the evidence through which a new counterparty might judge him can all remain bound to the service.

Some of that cost belongs to a worthwhile association. A new marketplace cannot supply the same customers merely because it offers an export function, and customers have their own freedom to remain. Obligations incurred inside the old relation may survive departure. The difficulty is to distinguish these consequences from restrictions by which an operator keeps unrelated interests at risk. A seller's evidence of completed work should not become inaccessible merely because he wants another venue to judge it. Nor should leaving a venue make a debt disappear. A useful portability rule has to preserve both sides of that distinction.

The operator's own dependence deserves equal attention. Members can appear to have many choices while every practical route to their customers passes through the same intermediary. Conversely, an institution with considerable local power may be answerable because those it serves can act together, establish a rival forum, or obtain a remedy from elsewhere. Counting venues will tell us less than following the relations that keep them in business. Somewhere in those relations there must remain a way to make the successful institution answer for the terms it imposes.

The sword in the basement

Hobbes's covenants without the sword have lost none of their difficulty. A person contemplating performance wants some assurance that the other party cannot take its benefit and leave the obligation behind. A legal order that announces consequences which nobody can bring about offers little security against someone prepared to test it.⁹ The histories of commercial order do not dispose of this difficulty. They show how several institutions could contribute to a consequence, sometimes without any one of them governing the whole relation.

A court examines claims and gives judgment. The force that makes the judgment effective may be situated elsewhere, under an authority whose reach is broader and whose knowledge of the dispute is narrower. An abbot's officers could seize goods in a fair whose commercial practices they did not invent. A community could deny access to its members while depending on a ruler to keep the roads open. The distribution was often uncomfortable, with each participant trying to enlarge its authority or avoid the costs incurred by another. Nevertheless, the work could be divided. The institution that understood the transaction did not have to possess every means by which a judgment concerning it might be enforced.

A forum can appear self-sufficient while compliance is ordinary. Its dependencies become visible when someone refuses, moves the assets, or invokes an authority which the forum cannot command.

International arbitration gives the division an explicit legal form. Parties can select a tribunal and procedure, identify a seat, and arrange for a dispute to be heard by people with relevant expertise. The resulting award can receive effect in another jurisdiction under the New York Convention of 1958. The Convention provides for recognition of arbitration agreements and enforcement of foreign and non-domestic awards within its scope, subject to its conditions and grounds for refusal. A domestic court supplies coercive reach where voluntary compliance has ended.¹⁰

The arrangement answers a practical difficulty which confidence in one's own courts cannot necessarily solve. Each contracting party may have a home jurisdiction and little enthusiasm for submitting to the other's. Even a trusted home court may be unable to reach assets abroad without help. By agreeing on a forum in advance, the parties can separate the question of who should decide the merits from the question of where property may later be found. A clause which appears remote from the business at hand is arranging the conditions under which the business can survive a breakdown in confidence.

People who expect to do profitable business together arrange for a third party to hear them if the relationship fails. The name of that person may still be unknown. They settle how the appointment will be made and how an argument they have not yet had will be conducted, while they are still able to agree about something. It is an accommodation to their own possible unreliability, or to the less offensive possibility that both will act honorably and discover that they understood the bargain differently.

The receiving court does not simply accept everything the tribunal has done. Article V includes grounds for refusal concerning an invalid arbitration agreement, inability to present a case, an award exceeding the submission, and other specified defects; arbitrability and public policy also enter. The Convention supplies no general appeal on the merits.¹⁰ Its usefulness depends in part on that restraint. An award that merely opened the same factual controversy afresh wherever assets were found would carry a much less valuable kind of finality.

Finality leaves room for disappointment which no institution can wholly remove. A party may have been heard fairly and still believe the tribunal misunderstood the business. Another may believe, with good reason, that renewed litigation would consume what the award has finally secured. Broadening review changes the position of both. A small claimant can need protection from a bad decision and from a well-funded opponent's ability to prevent any decision from becoming final. The cost of error has to be considered together with the cost of continuing to inquire.

This is also why the growth of public commercial law cannot be treated simply as the loss of a more intimate justice. Public institutions can give a stranger standing which a closed trading community would deny, reach property beyond a particular venue, and provide a means of challenging those who administer it. Their generality can be valuable precisely when the local relation has become oppressive. It may also make a proceeding slower, more expensive, or less able to distinguish the features of a trade on which the dispute turns. These costs are borne unevenly, and the label public tells us little about whether the person who needs the remedy can obtain it.

Bruno Leoni's criticism of legislation attends to the knowledge lost when general rules are made at a distance from the claims they will govern. Law developed through disputes can draw on circumstances an advance formulation failed to anticipate.¹¹ That is a reason to preserve the work of particular judgment. It gives no assurance that the judge knows enough, that the claimant can afford to appear, or that the accumulation of prior decisions has remained open to people whom earlier proceedings excluded. A rule developed slowly can carry an old advantage as faithfully as a newly enacted one.

Nor should we assume that a formal limit imposed from elsewhere is necessarily alien to the relation it constrains. Someone inside the relation may have been asking for it for years. A prohibition can give effect to a claim which the institution's ordinary procedures had learned to disregard. Its justification still depends on the claim, the reach of the remedy, and the consequences of imposing it. The defender of private order has reason to scrutinize public power and equal reason to avoid lending the language of freedom to an institution whose members can no longer make it answer.

Elinor Ostrom's account of durable self-governance is useful here because it directs attention to the arrangements by which participants monitor conduct, alter rules, impose proportionate sanctions, and resolve disputes at a practicable cost. These arrangements vary with the resource and its users; in larger systems, they can belong to institutions organized at several scales.¹² Her work gives us grounds for examining who does the governing and how the governors are constrained. It would lose much of that value if converted into a warrant for whichever institution happens to call itself a community.

The terms of the bargain

Henry Maine's movement from status to contract described a change in the formal source of obligation. A relation once attached to one's place in a family or inherited order could increasingly be entered through agreement.¹³ The phrase has an amplitude which the history cannot consistently sustain. People continued to bargain from positions shaped by birth, wealth, sex, citizenship, and associations they had not chosen. Yet the possibility of undertaking a particular obligation, with terms capable of being stated and examined, could alter what another person was entitled to demand.

A contract limits as well as binds. The parties can agree to a sale without acquiring authority over each other's lives. The obligation may be demanding and its enforcement unpleasant; it is still possible to ask what was undertaken, by whom, and within what bounds. This is one reason the inadequacies of actual consent deserve attention. A relation that obtains its legitimacy from agreement owes an account of the agreement through which that legitimacy is claimed.

The agreement, in turn, needs institutions. Someone may have to authenticate it, determine whether performance occurred, or recover what is due. An enforceable promise can free a person from dependence on the continuing goodwill of a counterparty while making him dependent on the offices through which enforcement is obtained. Long-distance commerce exposed these dependencies because a document could move more easily than the circumstances required to give it effect. A claim intelligible at home might become an expensive explanation abroad.

Computational verification can reduce the cost of some of that work. A signature can be checked against an authorized key, a delegated power compared with its bounds, and a recorded event tested against a declared condition. Where money has already been placed under an agreed mechanism's control, a verified condition can trigger a payment. The arrangement brings some evidence and consequence into the transaction itself, before a claimant has to find a forum and reconstruct the sequence after failure.

Consider an agreement to release payment when a consignment reaches a checkpoint and a designated inspector attests that it passed a specified test. A system can check the inspector's signature and apply the agreed payment rule. It can establish that the authorized attestation was supplied. The inspector may still be mistaken or dishonest, and the test may have been a poor account of the quality the buyer needed. An exact execution of the payment rule leaves those questions where the parties put them, or where they failed to make provision for them at all.

There may be considerable value in that narrower achievement. The parties can establish which message triggered payment and which authorization the mechanism accepted. If a bond or escrow has been properly constituted, some assets can remain available while a specified objection is examined. A later tribunal may receive a more intelligible disagreement because parts of the sequence were established when they occurred. The reduction in work is conditional on reliable implementation, usable evidence, and an arrangement which actually controls the promised assets. A declaration of collateral supplies no security if the funds can be withdrawn before the claim matures.

The difficulty also begins before execution. Someone selected the test. It may be entirely adequate for one use of the goods and inadequate for another. A supplier can learn to satisfy every measured property while letting an unmeasured one deteriorate. The buyer who accepted the specification may have misunderstood its limits; the seller may reasonably have relied on its words. A dispute of this kind concerns the relation between the formal condition and the undertaking it was meant to express. Checking the condition again will not complete the inquiry.

The penumbra includes these ordinary disagreements. Interpretation is required where the declared rule does not settle its own application or where applying it raises a claim the parties have reason to hear. Ordinary language, imperfect observation, unforeseen circumstances, and disagreement about authority can bring judgment into an otherwise routine transaction. A formal procedure can make the disagreement easier to locate without deciding how it should be resolved.

Much commercial expense arises from recovering facts that could have been established more cheaply at the time. A bounded system which reliably records authorization or keeps pledged assets available may spare both parties substantial work. The saving does not have to encompass the whole dispute to matter. It can change which transactions are economical, which counterparties can deal with one another, and how much activity needs to remain inside a firm with a common administration.

The arrangements that produce the saving acquire power of their own. A standard can determine which claims are easy to express. A credential issuer can control whose authority is recognized. A settlement service can become difficult to leave because everyone else already relies upon it. Even when the verification itself is inexpensive, entry into the network that recognizes its result may be costly. The merchant looking for a usable forum would recognize something of this difficulty. The means by which a claim travels can become the place where its bearer must seek permission.

It follows that moving verification closer to the transaction requires attention to what remains outside it. An affected person may need access to the relevant evidence after a provider's interface has disappeared. A tribunal may need to examine why a formally valid act exceeded the authority its principal intended to grant. People beyond the agreement may have suffered an injury which the agreement gave neither party an incentive to consider. A system that answers its participants well can still leave an outsider addressing a relation in which nobody has offered him standing.

These are reasons to specify the reach of a computational arrangement and the institutions that can receive what it leaves unresolved. They also qualify the hope that adjudication will become a smaller remainder around an expanding formal core. Better records may reduce reconstruction; cheaper transactions may increase the number of disputes worth bringing; new forms of automated action may create kinds of disagreement the old procedures never encountered. The future workload of courts cannot be inferred from the speed at which a signature can be checked.

The record that remains

At St. Ives, the proceeding ended with a result which can still be read. The objection to the document survived in the record even though the court did not permit the inquiry William requested. That survival allows us to distinguish the court's disposition from a finding that the acquittance was genuine. A later reader has something more than the winner's account, although much less than the evidence needed to decide the original debt.

The distinction can be lost each time a result passes onward. A second institution receives a judgment; a third receives the classification which the second made from it. After several such transfers, a decision that depended on a rule of procedure may appear as an established fact about the person. Nobody in the chain need have intended the enlargement. Each may have accepted an earlier institution's work without retaining the limits under which it was done.

A useful receipt therefore has to preserve more than the conclusion if another institution is expected to rely on it. For the relevant act, it should be possible to recover the authority invoked, the evidence examined, and the bounds of what was decided. An unresolved allegation needs to remain distinguishable from a finding. A refusal to hear a claim needs to remain distinguishable from its refutation. Otherwise the recipient receives an assurance whose apparent scope increases as the means of questioning it recede.

This does not require a permanent transcript of everyone whose life touches the transaction. The purpose is to make a consequential use of authority examinable, with access and retention appropriate to the claim. Much of the surrounding life may have no place in that account. The merchant who wants his completed commitments to be recognizable elsewhere need not surrender every conversation through which they were negotiated. There is a difference between preserving the grounds of a decision and giving future institutions an unrestricted account of the person against whom they might act.

Correction has a cost, and so does the opportunity to request it. A procedure that pauses every payment on an unsupported objection can be used to make ordinary performance impossible. A procedure that permits no pause can execute an injury before the evidence of a mistake is assembled. The relevant limits concern the kind of defect alleged, the evidence available, the consequences of delay, and the means of preserving a remedy while the dispute is heard. They will differ across transactions. A constitutional promise becomes credible through these particular provisions, including the provisions that prevent one party from exhausting another through continued review.

The institution receiving the record has responsibilities of its own. It can accept a limited finding without treating the issuing forum as competent to decide everything now at issue. It may have reason to ask whether the affected party could present a case, whether the relevant authority existed, or whether the finding bears on the proposed consequence. The reception of evidence is itself an act of judgment. Making evidence portable should make that judgment better informed, rather than allow every receiving institution to suppose that someone elsewhere has already assumed its responsibility.

There will still be cases in which a rule was followed and a community decides that the result cannot stand. An exception then needs an answerable author and a record of the authority under which it was made. Those who relied on the prior rule have claims too. A declared emergency cannot by itself settle the legitimacy of the power exercised in its name, any more than an unbroken formal procedure can settle every objection to what it has done. Both ordinary execution and intervention need to remain within reach of an institution capable of examining them.

The history of commercial forums gives us reason to take the division of this work seriously. A specialized court can hear a claim which a distant generalist would struggle to understand. An authority outside the trade can give a remedy to someone whom its insiders would prefer to ignore. A record can spare a later institution the labor of reconstructing an event and can conceal how little the first institution established. The value of the arrangement depends on how these activities meet, on what can pass between them, and on what the receiving party is entitled to question.

Much of this labor can be gathered inside a firm. Employees share records and procedures; authority can be delegated within a continuing organization; a dispute that would have required negotiation between businesses can become a matter for internal decision. Keeping the activity together has costs as well. Coase's question concerns where the balance falls between organizing within the firm and arranging transactions outside it.¹⁴ Cheaper means of finding counterparties, verifying performance, and settling obligations can move that balance. They do not tell us where all the work displaced from the firm will be done.

A computational agent may negotiate and complete an exchange in less time than a conventional organization would need to assign it. Its temporary presence need not prevent a valid agreement. It does make the duration of the surrounding institutions newly consequential. The goods may arrive later; their quality may become apparent only in use; an obligation may mature after the process that accepted it has ended. Deciding whether this activity can leave the firm requires an account of who will carry those obligations and receive those claims. The saving in coordination can be assessed only when the work that continues after execution has found somewhere to reside.

---

Footnotes

1. Stephen E. Sachs, “From St. Ives to Cyberspace: The Modern Distortion of the Medieval ‘Law Merchant’,” American University International Law Review 21, no. 5 (2006), 685–812. The case and the allocation of authority are also set out in the author's earlier, openly accessible thesis, “The ‘Law Merchant’ and the Fair Court of St. Ives, 1270–1324” (2002), chapters I–II. The claim that the acquittance was forged remains an allegation; the account here does not adopt Sachs's stronger inference about its authenticity. ↩ ↩²

2. Sachs, The ‘Law Merchant’ and the Fair Court of St. Ives, chapter II.B.2, note 73, citing Charles Gross, ed. and trans., Select Cases Concerning the Law Merchant, vol. I (Selden Society, 1908), 101. The roll records a request to transfer the case and the distresses; the explanation based on the fair's approaching end is an inference, not a reason stated in the roll. ↩

3. Jeremy Edwards and Sheilagh Ogilvie, “What Lessons for Economic Development Can We Draw from the Champagne Fairs?,” Explorations in Economic History 49 (2012), 131–148, especially the account of institutional provision and the changing political conditions of the fairs. ↩ ↩²

4. Paul R. Milgrom, Douglass C. North, and Barry R. Weingast, “The Role of Institutions in the Revival of Trade: The Law Merchant, Private Judges, and the Champagne Fairs,” Economics & Politics 2, no. 1 (1990), 1–23. ↩

5. Cambridge University Library, “Discovery,” Discarded History; Geoffrey A. Khan, “Arabic Manuscripts in the Genizah,” part one, Genizah Fragments 10 (1985). Khan discusses reverence for Hebrew script, the prevalence of Arabic written in Hebrew characters, and the presence of documents in Arabic script. ↩

6. Avner Greif, “Reputation and Coalitions in Medieval Trade: Evidence on the Maghribi Traders,” The Journal of Economic History 49, no. 4 (1989), 857–882; Institutions and the Path to the Modern Economy: Lessons from Medieval Trade (Cambridge University Press, 2006), chapter 3. ↩

7. Jeremy Edwards and Sheilagh Ogilvie, “Contract Enforcement, Institutions, and Social Capital: The Maghribi Traders Reappraised,” The Economic History Review 65, no. 2 (2012), 421–444; Avner Greif, “The Maghribi Traders: A Reappraisal?,” The Economic History Review 65, no. 2 (2012), 445–469. ↩

8. Lisa Bernstein, “Opting Out of the Legal System: Extralegal Contractual Relations in the Diamond Industry,” The Journal of Legal Studies 21, no. 1 (1992), 115–157, especially 121, note 10, on rebuttals to gossip, and the discussion of bourses and arbitration. Subsequent paragraphs develop institutional possibilities; they do not attribute every failure described to the Diamond Dealers Club. ↩ ↩²

9. Thomas Hobbes, Leviathan (1651), chapter XVII. The argument about dividing institutional work is this chapter's, not Hobbes's conclusion about sovereignty. ↩

10. United Nations, Convention on the Recognition and Enforcement of Foreign Arbitral Awards (1958), articles I–V, especially V(1)(a)–(e) and V(2). ↩ ↩²

11. Bruno Leoni, Freedom and the Law (1961; expanded Liberty Fund edition, 1991). ↩

12. Elinor Ostrom, Governing the Commons: The Evolution of Institutions for Collective Action (Cambridge University Press, 1990), chapter 3; “Beyond Markets and States: Polycentric Governance of Complex Economic Systems,” Nobel lecture (2009), especially the discussion of design principles. The application to commercial and computational institutions is an analogy, not an assertion that those institutions are common-pool resources. ↩

13. Henry Sumner Maine, Ancient Law (1861), chapter V. The checkpoint and inspector in the subsequent discussion are an illustrative arrangement, not a reported deployment or a claim of present product capability. ↩

14. R. H. Coase, “The Nature of the Firm,” Economica 4, no. 16 (1937), 386–405. ↩