what he really means, despite the use of "sine qua non", is that the vast vast majority of people really don't like dealing with more than one currency, and if their government uses a given currency, most people will use it because they have to pay taxes with it.
> Sure, taxes increase demand for money, but that is a far stretch from saying money can't exist without taxes
it really depends what you mean by "money". prior to government meddling in markets, most historical evidence points to economies relying on ad-hoc, non-denominated credit systems (aka, i can drink at the tavern all year because the barkeep knows i'll give him some of my grain come harvest time). these tabs were, notably, not denominated in an abstract currency. it wasn't "you've spent 16oz of gold at the bar this year", it was just "you've had a lot to drink, you better give me a good portion of your harvest".
you only really start seeing professional merchants and money-based trade after currency has been standardized by axial age governments for the purposes of funding their armies. to fund the armies, governments would mint money out of precious metal reserves, pay the army, and require the coins back from the population at large. all of a sudden, the whole population is now legally required to, in some manner, economically service the army (as they are the vehicle through which all coinage enters the market). the armies would then take over neighboring countries, enslave their population, and the slaves would mine more metal for coining. a nice little feedback loop.
anyway, sorry i got a bit off track there, but my real point was: currency was, by all historical evidence, definitely an invention of/majorly promoted by the state.