Currency backed by nothing (fiat) is at least 4000 years old.
Before paper money there was Chinese coins from base metal, not silver or gold, but before that there were credits, fiat money.
To make his point as clear as possible, Graeber (p. 29) quotes from Caroline Humphrey's Cambridge University dissertation as the definitive anthropological work on barter. Her statement is as clear as it is emphatic. "No example of a barter economy, pure and simple, has ever been described, let alone the emergence from it of money; all available ethnography suggests there has never been such a thing." Innes knew this 100 years ago, yet the myth persists.
So if there has never been a land of barter, where did we get money and credit? Innes (p. 397) argues that systems of credit pre-date coins by over a thousand years. "The earliest known coins of the western world are those of ancient Greece, the oldest of which, belonging to the settlements on the coast of Asia Minor, date from the sixth or seventh centuries B.C." In contrast, the law of debt goes back to at least the Code of Hammurabi in Babylonia 2000 years B.C. Innes saw that the foundation of society and thereby of credit was that promises or obligations were and are viewed as sacred. In all societies (p. 391) the breaking of the pledged word, or the refusal to carry out an obligation is held equally disgraceful." He goes on to explain how wooden tally sticks and clay shubati tablets were used to track credits/purchases and debits/sales long before the existence of coins. And that one could repay a debt by returning a credit of the same amount to the lender. In fact, village fairs were convened so that those holding the debts of others could match credits and debits together and thereby clear their accounts. Over time others showed up to buy and sell other goods and services or to cater to those in this most basic business of banking.
There are a variety of reasons why this matters for monetary theory and macroeconomic policy. But let me leave you with just one. From the Smithian story, it was gold and silver that backed the issuance of a paper currency. However, if Innes is right, the banking system never worked in that way.
In Innes's world, money is and always has been a token representing a socially constructed debit-credit relationship. A stamped coin, $20 bill, or tax refund check is an asseta credit to those who hold it and a liabilitya debitfor the government who issues it. When the federal government spends, perhaps by directly depositing a Social Security recipient's check into her account, a special kind of credit is created. This credita new "debt" of the federal governmentsatisfies all four functions that are used to define money. It serves as a medium of exchange, store of value, means of payment, and a unit of account. But what gives this money value? The money is valuable because it is the only token acceptable for the payment of taxes. And when those taxes are paid, the money that had been spent into existence is extinguished. Thus, it is through federal government spending that money enters the economy and through taxation that it is destroyed. This is where Innes's 100- year-old insights lead. If these ideas are hold up under academic scrutiny, are further disseminated, and become the basis of how we understand money and credit, an entirely new paradigm will need to emerge in the study of monetary economics.
http://neweconomicperspectives.org/2013/09/money-created-overcome-barter.html