The standard story of money begins with barter. People trade goods directly, run into the problem of the "double coincidence of wants" (the hungry shoemaker must find a baker who happens to need shoes), and so gradually settle on a common medium of exchange, usually a precious metal. Money, on this account, is a technology for making trade easier. The debt argument reverses the order: money did not replace barter, it grew out of credit, and at bottom it is a way of recording who owes what to whom.

The argument's earliest modern statement came from the economist Alfred Mitchell-Innes in 1913, who held that a coin or a note is just a portable IOU, and that the value of money lies in the relationship between creditor and debtor rather than in any commodity. The anthropologist David Graeber later gave it its fullest popular form. His case rests on several observations.

First, anthropologists have never found a society that ran on pure barter. Small communities tend to run on gift-giving, informal obligation, and mutual credit: you give me grain now, and I remember that I owe you. Barter in the textbook sense shows up mostly between strangers, or among people used to a currency who have suddenly lost access to it.

Second, the earliest written records point to accounts, not coins. In Bronze Age Mesopotamia, temples and palaces tracked grain, silver, and labor in ledgers, and ordinary people bought on credit from alehouses and merchants, settling up at harvest. The shekel served as a unit of account, a measure for writing down debts, long before anyone minted it. Coinage appears only around the seventh century BCE, in Lydia and elsewhere, and it spread largely through states paying and provisioning soldiers.

Third, the debt theory points to the role of the state. Rulers who demanded taxes payable only in the currency they issued created a standing debt for every subject, and with it a reason for everyone to accept and hold that currency. Money, in this reading, is less a spontaneous invention of the marketplace than a relationship of obligation that governments formalized and enforced.

Defenders add that even humble tools of record-keeping fit the pattern. Medieval English tally sticks, split so that creditor and debtor each held half, were in effect money: a claim that could be passed along and eventually used to settle accounts with the Crown.

Critics reply that the evidence is thinner than it looks, that markets and commodity exchange are also very old, and that the two stories may not be rivals so much as descriptions of different layers of the same history. But the debt argument has changed the question. It asks us to see money first as a social promise, and only secondarily as a thing.