David Glasner had an article called Wicksteed on the Value of Paper Money.
David discusses the idea that it is the imposition of taxes by government, payable in fiat money tokens, that gives fiat money its value. This is chartalism, the very same idea that MMTers trumpet as their unique addition to economic discussion. This is one theory for why fiat money has value. Another is Mike Sproul's backing theory in which, just as a mutual fund's assets give its units value, a central bank's assets support the value of its liabilities. When it comes to explaining modern money, I'm partial to the latter. I'm not averse to the MMT explanation, but only as science fiction. I don't think any real monetary system has actually worked in this way.
David mentions that Adam Smith advocated the taxes-drive fiat money theory. Incidentally, Randall Wray says the same in his book Understanding Modern Money. I disagree. To make a long story short, in Smith's world, fiat money was any issue of paper money which was temporarily unredeemable and therefore circulated at a discount. In forcing people to pay taxes with this money, the sovereign created a built-in liquidity premium. But the tax obligation did not give the paper money it's original value - the probability of future redemption did. Here's the a better explanation that I left on David's blog:
I’ve read Smith pretty carefully on this. The offhanded comment comes after he talks about money that is no longer instantly convertible but redeemable at some future point in time. He uses as his example Scottish notes for which the option clause has been invoked and US colonial paper which is only redeemable after a few years.
Given deferred redemption, “such a paper money would, no doubt fall more or less below the value of gold and silver, according as the difficulty or uncertainty of obtaining payment was supposed to be greater or less, or the greater or less distance of time at which payment was exigible.”
The point being that Smith thought such money was valued according to its discounted probability of being redeemed at par.
Smith then brings up the point about taxes.
“The paper of each colony being received in the payment of the provincial taxes, for the full value for which it had been issued, it necessarily derived from this use some additional value, over and above what it would have had, from the real or supposed distance of the term of its final discharge and redemption.”
Thus acceptability in payment of taxes added a liquidity premium to colonial paper. But it didn’t give that paper its original value. I read Smith as providing a chartal theory of the liquidity premium, and not an explanation for a positive value of fiat money.