A Jacobin magazine article by Mike Beggs reviews Graeber's Debt: The first 500 years. Among the many, many well-founded statements is the following, discussing the establishment of gold as the standard for valuation of currency:
In the modern period, state after state committed to metallic anchors as strategic decisions to enhance trust in their national currencies. Gold eventually beat out the other metals on a world scale thanks to various accidents and a snowballing network effect. The point was never to drive out state paper money, but to promote its acceptance as a stable standard of value. Neither was it intended to wipe out credit-money, but to tend and grow it by taming the wild fluctuations of bank credit.
Wouldn't you love to know what the "various accidents" and the "snowballing network effect" were, at least in Beggs' opinion?