Chains of Stakes

Al nome di Dio, amen.

On 12 December 1399, Giovanni Orlandini and Piero Benizi and company in Bruges instructed Francesco da Prato and company in Barcelona to pay Domenico Sancio six hundred écus, reckoned at ten shillings and five pence of Barcelonese money to the écu, and charge their account. The value had been received from Jacopo Goscio. There was a blessing at the end as well as at the beginning. Between the two, the letter found room for the people, the money, and the instruction. It did not explain the business to a stranger.¹

The acceptance was dated 11 January 1399. Read without the calendar convention, it appears to precede the instruction by eleven months. The January belongs to 1400 in our reckoning. Even this small piece of writing, which seems to have stripped a transaction down to what was indispensable, asked something of its reader that it did not undertake to teach him. Payment was due at usance: a customary interval intelligible to people who knew the exchange between these places. The paper could be brief because others knew a great deal.

We can follow the instruction as far as the acceptance, but not watch the money being paid. For that we would need another record. The few lines have already brought several people into relation without telling us very much about any of them: who could afford to wait, who expected further business, who might have to find money that was not presently at hand. Those questions would have been quite real to someone. They need not all have been answered on this sheet.

The invocation is familiar in another way. God has been admitted into a transaction whose amounts are stated with considerable care. We can enjoy the proximity without deciding that either the devotion or the accounting must have been insincere. A merchant did not have to resolve the relation between salvation and gain before giving an instruction. There was business to be done in the meantime.

Papers left in a house

The bill survives through the Datini archive at Prato, among the accumulated writing of Francesco di Marco Datini and the businesses with which he was connected. A few lines of instruction have come down to us amid an abundance of correspondence: letters between partners, letters home, accounts sent back for examination. The collection offers far more company than the bill itself requires.

Margherita appears in these papers attending to a household from which her husband was often absent. Their marriage produced no children, but Datini had a daughter, Ginevra, born to an enslaved woman and brought into the household. His biographer Michele Luzzati describes a correspondence in which affection, business and the vexations of absence repeatedly meet. The merchant's family lived with the consequences of a success conducted elsewhere; letters were among the means by which they continued a life together.²

Datini could be absent for another reason. In December 1386, having been selected as gonfaloniere at Prato, he tried to avoid the public office by hiding for six days in Florence. Luzzati records that the attempt failed. A man capable of supervising business across several cities had found an unwelcome use for his abilities close to home. The episode is worth keeping beside the grander picture of mercantile organization: there were occasions on which the organizer preferred to be difficult to find.

When a company periodically closed its accounts, its books were sent to Datini for examination, accompanied by bundles of correspondence. After his death, the charitable institution founded through his estate became their custodian. The archive's later history includes ordering, storage in disused spaces, and the recovery of bundles from a closed staircase. By then, much of what had once needed to be answered had ceased to concern anyone. The papers could enter another kind of inquiry. Even the familiar nineteenth-century rediscovery story reaches us through incomplete accounts.³

Because these papers have become valuable to us, it is easy to imagine their keepers steadily preserving our future knowledge. An account awaiting examination had a more immediate reason to remain. A letter could be needed to settle what a correspondent had been told. Once those reasons expired, the paper might remain simply because there was room for it; which explanation belongs to which bundle requires evidence. Across that long interval the writing has acquired readers who care intensely about particulars its first readers could take for granted, and know almost nothing of matters a terse instruction could safely leave unsaid.

The bill now looks rather less self-sufficient. Its brevity belonged to people with accounts to consult and correspondents from whom more could be learned. It spared them the trouble of setting all that knowledge down again.

What the correspondent knew

A bill could help move purchasing power without sending the corresponding quantity of coin along the same route. It could also supply credit during the interval before payment. These uses overlapped, but they were not identical: moving money to another place did not mean that every participant had entered the same kind of loan. The exchange of currencies added another relation to be understood. Six hundred écus named in Bruges and the Barcelonese sum payable to Sancio were connected by an agreed rate, not made identical by the lettering on the page.

The named recipient of the instruction had work of his own. He needed to recognize the house that had written, know which account it meant, and decide how this request bore on his dealings with it. Familiar handwriting and specimens circulated among correspondents could support recognition; letters of advice could carry information alongside the bill. An accustomed reader could use both while still needing to consult his accounts or send an inquiry of his own. These were arrangements for making payment possible. Kohn's history keeps that ordinary purpose in view where an account confined to lending can lose it.⁴

Repeated dealings gave a correspondent reasons to attend to a request that would have looked different coming from an unknown house. He could know more than a new lender about its business; there might be further business he wanted, or existing obligations which a refusal would disturb. A troublesome transaction need not exhaust the relation in which it occurred. Continuing could be prudent, expensive, welcome, or merely preferable to the available alternatives. An account which records continuation will not always let us choose among those explanations.

A named payer, an amount and a due date make an arrangement look finished. Yet the parties might still have things to decide together, especially if the money could not be paid at the expected time. One could insist on the original terms, accept a different date, or seek payment by another route. The ability to recognize what had been agreed supplied a place from which to make a change.

Another archive makes the difficulty unusually clear. The Borromei ledgers take us to Bruges and London in the 1430s, several decades after the Datini bill. The original bills and much of the accompanying correspondence are missing; Jim Bolton and Francesco Guidi-Bruscoli reconstruct their uses from the accounts. Their study finds considerable flexibility in dates and methods of payment. The ledger entries expose accommodations that the standard description of the instrument makes easy to overlook.⁵

On 28 June 1438 the Bruges bank gave John Young, factor and attorney of the merchant Thomas Cannings, £121 Flemish. Young drew a bill on London for £110 sterling. It was not sent. The bank held it as a guarantee for repayment at the August fair at Antwerp. Here a document addressed to another city was useful while remaining where it had been drawn: it supplied a course of action which the arrangement allowed the bank to withhold.

The bill was canceled on 23 September. Young had not brought the cash back: he drew another bill on London, at an exchange rate more favorable to the bank. Four days later, in a separate transaction, the Borromei advanced him £130 Flemish and again held a London bill rather than dispatching it. Young was to repay at the following fair at Bergen-op-Zoom. This time the money was not repaid, and on 19 November the bank sent the bill. The account takes us to that decision; the later outcome in London is not established by this entry.

It would be possible to treat the interval before dispatch as a delay in the instrument's proper operation. I think that would miss much of what the bank had sold. Young obtained money with an opportunity to repay through another route; the bank kept a contingent means of seeking payment. Both had reason to know which bill was still held, which had been canceled, what had replaced it, and what would cause the remaining one to be sent. The usefulness of the written form lay partly in allowing these distinctions to be maintained while the business changed around them. Merely preserving the first instruction would have given a worse account of the obligation.

The exchange rate favored the Borromei. Bolton and Guidi-Bruscoli consider whether the difference reflected a penalty, a longer maturity, or some combination; the entry does not isolate a price for each. The accommodation remained useful without necessarily being generous. Young had obtained money and choices about repayment, while the bank retained claims and terms of its own. A customer whom it had no reason to favor might have found a less accommodating institution behind the same form of bill.

This was commerce among people who had reasons to keep dealing with one another, conducted through an instrument which helped them do so. Its success did not depend on turning them into strangers. Their knowledge of each other's affairs was still doing work.

Keeping the books

The accommodation leaves a small administrative difficulty in its wake. Someone has to distinguish cancellation from payment, and keep track of the obligation that took the canceled bill's place. A clerk consulting the account later should not have to find the people who arranged the change before knowing what remains to be done.

Double-entry bookkeeping gave the keeper a disciplined way to relate postings. Each debit required a corresponding credit; a disagreement between their totals revealed that something had gone wrong. The comparison did not require someone to revisit every transaction before discovering the discrepancy. It was a powerful economy of attention. But the equality could survive an omitted transaction, or a false amount entered on both sides, and it could coexist with a debt that would never be collected. A business could be scrupulously recorded on its way to failure.

Pacioli's exposition, printed in Venice in 1494, is more companionable about the work than the later phrase “self-checking ledger” would suggest. When the merchant wishes to balance the books and carry the accounts into a new ledger, he should enlist a helper. One person takes the journal, the other the ledger. The helper calls the relevant pages; the keeper finds the entries and checks the amounts. Neither is to mark an entry as checked independently of the other. Once the mark has been made, someone coming afterward will understand that the comparison was done.⁶

That last precaution has an almost excessive modesty beside the claims subsequently made for double-entry. A mark which saves the next reader work is worth making correctly. The book can retain the mark after the checking has passed beyond observation. Pacioli's instruction therefore reaches the moment when a useful abbreviation might begin to misdescribe what people have done. He does not cure the problem with a more impressive notation. He tells the two people how to work together.

Later in his summary he insists on consent: the merchant must not enter someone as a debtor without that person's permission, or add conditions to a creditor's entry without agreement. This was advice in a manual, not evidence that every merchant obeyed it. It nevertheless makes it difficult to attribute to the book a power its teacher was so careful to restrict. A keeper could write an amount under a person's name. The writing alone did not entitle him to make that person owe it. Arithmetic would have no trouble with the entry.

Books can acquire an authority extending beyond such internal work. Poovey's account of mercantile writing follows the association between formal precision and the merchant's rectitude; the arrangement of particulars could help make a writer credible. The inference remains dangerous when the particulars are numbers. Order is visible before honesty has been established, and the person most interested in the reader's confidence may also have decided what the account will contain.⁷

Orderly books make an enormous amount of business intelligible, including business their keeper would prefer others not to understand. Their convenience also allows a recipient to decide how much further to inquire. He may accept the balance because the comparison it performs is enough for his present purpose, because he knows the keeper, because he has other evidence, or because he cannot afford the examination that would distinguish these reasons. The page will look the same while the grounds for relying on it differ.

Even merchants handling similar bills did not organize this work in the same way. Bolton and Guidi-Bruscoli note that English merchants used exchange instruments without necessarily keeping Italian double-entry books. A useful means of payment could be adopted while the surrounding accounts were kept differently. The connection between the instrument and the books was close; it still left room for more than one way of doing business.

Where the paper could be heard

Behind the books lay a question which no style of bookkeeping could answer by itself: where could someone go when the other party would not do what the account required? A distant correspondent might negotiate, refuse, or lack the means to pay. Continued business could support an accommodation; it could also be the very thing one party no longer wanted. The network needed ways to deal with relations which had ceased to be agreeable.

Long before the Borromei transactions, the Champagne fairs had concentrated trade through a recurring series of six fairs in four towns. Merchants met under protection whose value was inseparable from the power providing it. The counts supplied safe-conduct and institutions for settling disputes; towns, ecclesiastical authorities and merchants' organizations also participated in the business of the fairs. Measures and commercial conduct were subject to public regulation. A common measure was useful because people could expect it to be recognized and departures from it to have consequences. The material standard, however durable, could not confer that authority on itself.⁸

Those who supplied protection had interests in keeping the trade. That fact need not discredit the service. A traveler might welcome a ruler's practical reason for wanting merchants to return. The difficulty was that commercial usefulness gave the ruler opportunities beyond supplying it. Edwards and Ogilvie recount a loan in 1221 to the Countess of Flanders and Hainaut, secured by an agreement that her merchants would be excluded from the fairs if she failed to repay. The condition made other people's access answer for the debtor's obligation. We know the condition; we need not imagine the ban being carried out to see whose business it placed at risk.

The attractive picture of a chain in which every participant stakes something of his own has become less comfortable. Here part of the force came from people who had not borrowed the money. Their interest in continued admission could make the debtor's promise more credible to a lender. It could also give them a reason to resent an arrangement made in their name. The same commercial concentration which saved merchants the trouble of finding one another made exclusion expensive. A fair was useful as a meeting place and formidable as a place from which one might be barred.

The terms of protection therefore belonged to the trade as much as the price of carriage or the supply of credit. Merchants moved through overlapping jurisdictions, finding assistance in institutions whose authority did not extend everywhere they might have to go. Connections could help a claimant find a forum and make himself understood there. A person without the language, standing, money or introductions required could possess an intelligible claim and still have difficulty bringing it before anyone able to act.

The notary's protest addressed one particular difficulty within that world. When payment or acceptance was refused, the refusal had occurred somewhere. A party seeking recourse elsewhere needed a recognized account of the relevant acts: what had been presented, when, and what had followed. A protest could put the event into a documentary form another forum could consider. It did not turn everyone named in a bill into an unconditional guarantor. The instrument, timely procedure and the receiving law still governed who could be pursued and on what grounds.⁹

There was considerable value in being able to establish a refusal without requiring the later tribunal to reconstruct the entire encounter unaided. But documenting the refusal and recovering the money remained different work. A perfectly authentic protest could accompany a claim against someone with no assets available, or reach a forum in which the proposed defendant was not liable. The notary could preserve what had happened at the place of presentment. He could not make every other institution owe the claimant the result he wanted.

The paper had reduced part of the difficulty of acting at a distance. What remained had addresses, offices and people in it. Some charged for the work; some exercised powers acquired for other purposes; some bore costs because somebody else had made their interests useful as security. The bill could travel through this arrangement without describing how unevenly its burdens were shared.

Another holder

The changing uses of paper eventually altered who could make a demand. We must move forward again, beyond the fifteenth-century Borromei accounts. The wider use of assignment and endorsement in the sixteenth century and after helped instruments circulate beyond the parties to their issue. The legal histories of bills, letters obligatory and bearer or order instruments overlap; they do not form one neat succession. Assigning a claim, transferring a document and acquiring a right to pursue an earlier holder are separate achievements. Their conjunction required more than a sentence authorizing payment to someone else.¹⁰

An endorsement could name a new holder and, under the applicable rules, leave a previous party exposed if payment failed. That exposure gave the new holder something more to consider than the apparent wealth of the ultimate payer. But the names on the back were not a universal insurance policy. Presentment, notice, protest where required, and defenses recognized by the governing law could affect recourse. Nor should these later arrangements be read backward into the Borromei entries, where a bill could be held, replaced or redirected without establishing a later chain of endorsers.

The new holder might know less about the original business and more about the person from whom he took the paper. That changed what he needed to ask and from whom he could seek an answer. An instrument could circulate more readily because someone who had passed it on remained liable: the freedom of the paper depended in part on an obligation which its previous holder had not escaped by parting with it. How far that obligation reached was a matter of considerable interest to both of them.

People remained involved at different times, with different opportunities to inspect what the others had done. One wrote; another recognized the hand; a correspondent adjusted an account; a keeper distinguished the bill that had been canceled from the one still held; an official recorded a refusal. No participant needed to know everything the arrangement relied upon. Each could nevertheless become indispensable to someone who needed it to work.

That is also why usefulness could become control. A house which repeatedly made distant payment possible acquired knowledge and relationships a new entrant could not instantly reproduce. An office whose acts were widely recognized could charge for access to a form others needed. The charge might pay for necessary work. It might also exploit the absence of another recognized route, and a person standing outside the arrangement could have difficulty discovering how much of the price belonged to each. Knowing that the institution had made commerce easier was not yet knowing what one owed it forever.

We can return to the short bill without asking it to contain this whole history. The invocation, names, rate and instruction were enough to begin a course of action among people prepared to read them. When we ask why they were enough, the answer takes us outward into accounts, correspondence, accommodations and jurisdictions. Sometimes those arrangements kept an original undertaking in force. Sometimes they made a change possible. Sometimes they bound someone who had been given little say in the undertaking at all.

The people who supplied this work could be trusted, distrusted, replaced, or impossible to replace. Their authority grew partly from having done something worth doing. A history which sees only the service will miss the power it creates; a history which sees only the power will be unable to explain why people kept coming back. The next question belongs to both histories. Who controlled the verification?

Source notes

Footnotes

1. John H. Munro, The Bill of Exchange, Draft, or Acceptance Bill, p. 2, transcription and translation of the Bruges–Barcelona bill dated 12 December 1399, citing Raymond de Roover, Money, Banking and Credit in Mediaeval Bruges (1948), pp. 56, 72. The acceptance is 11 January 1399 old style, 1400 new style. The body paraphrases the transaction; only the Italian invocation is quoted. The source is a transcription, not inspection of the original sheet. Usance is a customary payment interval; its computation varied by route and usage. ↩

2. Michele Luzzati, “Datini, Francesco,” Dizionario Biografico degli Italiani, vol. 33 (1987), discussions of marriage, Ginevra, correspondence, and the December 1386 attempt to avoid public office. No particular domestic letter or motive for preservation is reconstructed here. ↩

3. Istituto Internazionale di Storia Economica “F. Datini,” “L'Archivio Datini,” account of returning branch papers, later custody, ordering and storage; Jérôme Hayez, “L'Archivio Datini: de l'invention de 1870 à l'exploration d'un système d'écrits privés,” Mélanges de l'École française de Rome—Moyen Âge 117(1) (2005), 121–191, at 121, on the incomplete transmitted history of rediscovery. The conditional observations about why individual papers might remain are interpretations, not an asserted chain of custody for every bundle. ↩

4. Meir Kohn, Bills of Exchange and the Money Market to 1600, Dartmouth Working Paper 99-04 (1999), preface and pp.1–4. The account of letters of advice and recognition is general history, not a reconstruction of the particular 1399 bill. The earlier cambium agreement, often notarized, and the later written instruction to a correspondent should not be treated as identical legal acts. ↩

5. Jim Bolton and Francesco Guidi-Bruscoli, “‘Your Flexible Friend’: The Bill of Exchange in Theory and Practice in the Fifteenth Century,” Economic History Review 74(4) (2021), 873–891; open text. Documentary basis and missing bills: 876. John Young transactions, cancellation, replacement and dispatch: 888, n. 63, citing Bruges ledger folios 80.6d–a, 117.2d, 121.1a, 312.3d–a, 348.2d. Flexible payment and the authors' account of ongoing relationships: 884–889. English merchants and accounting practice: 878. Interpreting the held bill as part of the service supplied, and asking who obtained similar treatment, are this chapter's inferences. ↩

6. Luca Pacioli, De Computis et Scripturis (1494), in John B. Geijsbeek, Ancient Double-Entry Bookkeeping (Denver, 1914), ch. 14 on counterpart postings; ch. 32, p. 69 on joint checking and marking; ch. 36, p. 77 on trial balance, consent and conditions. The discussion concerns this published prescription; it does not assume universal compliance. Equal debit and credit totals do not establish completeness, the truth of entries, or collectability. ↩

7. Mary Poovey, A History of the Modern Fact: Problems of Knowledge in the Sciences of Wealth and Society (University of Chicago Press, 1998), 29–33. The relation between formal precision and credibility is her historical interpretation. The wider discussion here distinguishes the appearance of an account from the grounds on which a particular reader relies on it. ↩

8. Jeremy Edwards and Sheilagh Ogilvie, “What Lessons for Economic Development Can We Draw from the Champagne Fairs?” Explorations in Economic History 49(2) (2012), 131–148, especially 131–136; author-hosted text. The 1221 condition is at 135, citing Bourquelot (1865), I, 194. The chapter distinguishes the recorded exclusion condition from evidence of its execution. The distribution of exposure and implications of concentrated access are the chapter's interpretation. ↩

9. Bolton and Guidi-Bruscoli (2021), 876–877; Kohn (1999), 3–4. The legal effect of a protest depends on instrument and jurisdiction. A refusal of acceptance and a refusal of payment are distinct events; neither makes liability unconditional. The discussion identifies the work done by an attestation without attributing a uniform law to medieval Europe. ↩

10. Kohn (1999), 23–26, distinguishes the histories of letters obligatory, bills of exchange and transfer practices. Assignment and endorsement are not used interchangeably here. The narrative makes no claim of a universal first date, general negotiability under a single 1537 ordinance, or endorsement demonstrated by the fifteenth-century Borromei examples. The broader law and forum question is developed separately in Courts Without Thrones. ↩