Before coinage: debt, tally and the ledger societies

Most people carry a story about the origin of money that they never chose. It goes like this: first there was barter, which was awkward because the baker who wanted shoes had to find a cobbler who wanted bread; then someone hit on the idea of a commodity everyone would accept; then metal, then coins, then paper, then the app on your phone. It is a tidy story. It appears in Adam Smith's Wealth of Nations in 1776 and it has been repeated in textbooks ever since.

The problem is that the archaeological and anthropological record does not support it. What the record shows, in the earliest literate societies we can read, is not barter but accounting: obligations recorded, tracked, netted and settled. Money in those societies looks less like a thing you hand over and more like a position in a ledger. Understanding that changes how you read everything that follows — including the balance in your bank account today, which is also a position in a ledger and not a thing at all.

The barter story and why the evidence is thin

The barter narrative was a thought experiment, not a report. Smith was reasoning about what must have happened before money, and the reasoning is elegant: barter suffers from the double coincidence of wants — for a trade to happen, each side must want what the other has, at the same moment, in divisible quantities. Money solves it. That much is sound economics.

What is not sound is the claim that any society actually lived through the barter stage. The anthropologist Caroline Humphrey put the point sharply in 1985: no example of a barter economy, pure and simple, has ever been described, let alone the emergence of money from one. Where anthropologists do find barter, it tends to be between strangers, at the edges of trade, or in societies where a monetary system has broken down — post-Soviet Russia in the 1990s, prisoner-of-war camps, hyperinflations. Barter is what happens when money fails, not what happens before it exists.

Inside a community, the record shows credit. Neighbours did not haggle over each transaction; they kept score. The baker gave the cobbler bread and both remembered. What is being tracked is a debt, denominated in something, settled later or perhaps never settled at all so long as the relationship holds.

Mesopotamia: the unit of account arrives first

The clearest evidence comes from Mesopotamia, because clay survives. Denise Schmandt-Besserat's work on small clay tokens, sealed inside hollow clay envelopes called bullae, argues that these were accounting devices — one token per unit of grain or livestock — and that cuneiform writing itself grew out of the practice of impressing the tokens on the outside of the envelope so the contents could be read without breaking it. If that reading is right, then writing was invented to keep accounts. Literature came later.

By the third and second millennia BC the temple and palace complexes of Sumer and Babylon were running large internal economies on written obligation. Rations, land rents, tithes and loans were recorded in two units: the shekel of silver, a weight, and volumes of barley. Note carefully what the shekel was. It was a unit of account — a measuring stick for value — long before it was a coin. Debts were denominated in silver by weight; they were frequently paid in barley, wool, oil or labour at conventional rates. The thing money measured and the thing that changed hands were routinely different.

Interest was charged, and it was regulated. The laws of Hammurabi, from the eighteenth century BC, cap interest on silver loans at roughly one fifth and on barley loans at roughly one third. Regulation implies a lending market large enough to be worth regulating. So does the other well-attested Mesopotamian institution: the periodic royal edict cancelling agricultural debts and freeing debt-bondsmen. Rulers cancelled debts because compounding claims on a peasant economy eventually swallow the tax base and the army's recruits. This is the oldest recorded instance of a lesson the rest of this course keeps re-teaching: a debt structure that cannot be serviced will be resolved somehow, and the question is only whether the resolution is orderly.

The tally: a ledger entry you could put in your pocket

Jump forward to medieval England for a mechanism that shows the ledger logic with unusual clarity. The Exchequer used tally sticks — lengths of hazel notched to record a sum, the notch widths coding pounds, shillings and pence. The stick was then split lengthwise through the notches into two unequal parts. The longer piece, the stock, went to the creditor; the shorter foil stayed with the Exchequer. Because the grain of the wood and the notch pattern were unique, the two halves matched only each other. The instrument was self-authenticating.

Now the important part. A tally issued to a royal supplier was a claim on the Crown, and Crown claims were accepted in payment of taxes. That made tallies transferable: a supplier who needed cash sooner could sell his stock to someone with a tax bill, at a discount. A discount on a future claim is an interest rate. So here, centuries before a central bank, you have government debt, a bearer instrument, a secondary market and a discount rate — implemented in wood, in a largely illiterate society, with no metal moving at all.

The system ran until 1826. In 1834 the accumulated sticks were burned in the furnaces under the House of Lords; the fire got out of hand and destroyed the Palace of Westminster. It is a good image to keep: the physical residue of England's ledger money burned down its parliament.

What "ledger money" actually buys you

The economic function of a ledger is netting, and netting is why credit systems economise on the scarce settlement asset.

Take four parties. A owes B 40. B owes C 40. C owes A 40. A owes C 10. Gross obligations total 130. Compute each party's net position: A pays 40 and 10 and receives 40, so A is short 10. B pays 40 and receives 40, so B is flat. C pays 40 and receives 40 and 10, so C is long 10. A single payment of 10 from A to C extinguishes all 130 of gross claims. The settlement asset — silver, coin, central bank reserves — is needed only for the residual.

That ratio, gross obligations to net settlement, is the quiet engine of every payment system since. It is why the medieval fairs cleared bills in batches rather than shipping bullion (the subject of the lesson on bills of exchange), and why modern real-time gross settlement systems are engineered so carefully around liquidity.

There is also a well-known illustration that money need not move at all. On the island of Yap, large limestone discs called rai served as high-value money; ownership changed by public acknowledgement while the stone stayed put, and at least one stone that had sunk at sea reportedly continued to be owned and traded. The economist Milton Friedman used the case in a 1991 essay to make exactly this point: what circulates is title, not substance.

Switzerland's own ledger money

Switzerland is not a footnote here. Before the federal state, the Confederation was a patchwork of cantonal coinages, and merchants coped through account-keeping and correspondent relationships rather than through any single national coin; a unified Swiss franc arrives only with federal legislation around 1850, a story taken up later in this course.

More striking is the WIR, founded in Zurich in 1934 in the depths of the Depression, when bank credit to small firms had collapsed. WIR is a mutual credit-clearing circle: members extend each other credit denominated in a unit tied to the franc, and balances are cleared on a shared ledger rather than paid in cash. It has operated continuously ever since, and its complementary-currency turnover has tended to expand when ordinary credit contracts — a modern demonstration that trade can be carried on the ledger when the settlement asset is scarce.

And every franc you spend today is a ledger entry. Retail payments net through the banks; interbank obligations settle in Swiss Interbank Clearing, operated by SIX on behalf of the Swiss National Bank, in central bank reserves. Notes and coin are a small tail on the system. The Mesopotamian scribe would recognise the architecture immediately; he would only be surprised by the speed.

Why this matters before you read another chapter

Three habits of mind follow. First, distinguish the unit of account from the means of payment from the settlement asset — they are three different jobs and history routinely assigns them to three different things. Second, when someone tells you money "is" a commodity, ask what the ledger looks like underneath; there is almost always one. Third, take debt seriously as the primary object. Coinage, which the next lesson takes up, did not replace credit. It sat on top of it, and it gave states a new and lucrative lever.

What to take away

• The barter-to-money sequence is a thought experiment that has never been observed in the field; the earliest readable economies ran on recorded obligation, with writing itself apparently growing out of accounting.
• A unit of account can exist and dominate contracts for centuries before any coin exists — Mesopotamian debts were denominated in weights of silver and routinely discharged in barley or labour.
• Netting is the reason credit systems economise on scarce settlement assets: in the worked example, 130 units of gross obligation cleared with a single payment of 10.
• The English Exchequer tally shows that transferable government debt, a secondary market and a discount rate can all exist without banks, paper, or metal changing hands.
• Switzerland's WIR clearing circle, founded in Zurich in 1934, and today's SIC settlement system are both ledger money — the franc in your account is an entry, not an object.

Sources

• Denise Schmandt-Besserat, Before Writing (University of Texas Press, 1992), on clay tokens, bullae and the accounting origins of cuneiform.
• Caroline Humphrey, "Barter and Economic Disintegration," Man, 1985 — the standard anthropological statement that no pure barter economy has been described.
• David Graeber, Debt: The First 5,000 Years (Melville House, 2011), for the credit-first reading of monetary history; read alongside its critics.
• Sidney Homer and Richard Sylla, A History of Interest Rates (Wiley), for Mesopotamian interest ceilings and the long record of lending rates.
• A. Mitchell Innes, "What is Money?", Banking Law Journal, 1913, and Georg Friedrich Knapp, The State Theory of Money (1905), the classic credit and chartalist statements.
• Milton Friedman, "The Island of Stone Money," Hoover Institution working paper, 1991, drawing on William Henry Furness III's 1910 account of Yap.
• William N. Goetzmann, Money Changes Everything (Princeton University Press, 2016), for the finance-as-technology framing of the ancient record.
• Swiss National Bank and SIX, published documentation on the Swiss Interbank Clearing (SIC) system, for how franc obligations are netted and settled today.