Everyone knows what money is. Whether denominated in paper, coins, or something else entirely, transactions get paid for by a physical commodity that enables humans not to waste their time bartering with each other. Felix Martin's book, published this week in the United States, convincingly argues this view is not only wrong, but dangerous.

Martin, a polymath classics graduate and ex-World Bank economist who now works as a UK fund manager, says money develops from a collection of obligations: i.e. credit. Bilateral credit becomes money when a third party can rely on its value. Credit can turn into a currency if the promise to repay is generally considered trustworthy, or if the issuer has enough authority to force its use.

Why does this alternative definition of money matter? First, because it is historically accurate; there has never been a sophisticated barter economy. Second, because since at least the fourth century B.C., rulers and subjects alike have known that controlling the value of money is an important manifestation of political power. Money is inherently political.

Control of the money supply allows governments to tax by stealth - either by clipping coins or paying their bills with newly created paper. The response to debasing governments is always to try to create a private alternative, what Martin calls "the monetary maquis." Bitcoin is the latest example.

Although he doesn't discuss that crypto-currency, Martin is very good on why schemes like it aren't sustainable. Money is credit, not a commodity, so it is liable to sudden losses of confidence, leading to sharp losses of value. For a stable currency, someone has to be able to offer trustworthy support in a panic, and that someone is always a sovereign state which can collect taxes.

Martin sees the current monetary arrangement as basically private credit with sovereign guarantees. He says the creation of the private but government supported Bank of England in 1694 first enabled both king and country to have a say on the monetary standard.

The replacement of royal power with taxpayer-funded democracies might have led to an enlightened age of money. Private institutions would spread credit efficiently, while wise government authorities, central bankers, would protect them from their own excesses by shifting the value of money in response to economic developments.

That clearly didn't happen - and Martin blames John Locke. The 17th century philosopher designed British parliamentary democracy to limit despotic power. He also wanted to limit the sovereign's role in deciding what money was worth. He demanded money be fixed as a certain weight of silver.

It may seem over the top to view this as the biggest economic clanger in the last three hundred years, but Martin makes a good case. Fixing the monetary standard made economic value seem a natural phenomenon, to be measured like weight or height, rather than a political question to be debated. In future, devaluing the currency to inflate away debts would be frowned on not because it merely made the rich slightly poorer, but because it interfered with the natural order of things.

Locke's fundamental monetary error led modern economists to see commodity money as logically anterior to credit, rather than the exact other way round. With this backwards view, the professionals focused on price stability and ignored the buildup of debt before the financial crisis 2007.

The Lockean economists also forget that money is trust-dependent credit, so they endorsed light-touch financial regulation. The result was banks which expanded out of control, and a new monetary maquis, a $35 trillion shadow banking system in the United States and EU.

The repudiation of Lockean monetary theory may make Martin sound radical. But critics will be missing out on an important insight into how finance and economics blindsided each other in the runup to the financial crisis. Their loss.