I’m not a fan of Graeber’s debt model for currency (i.e. the IOU). I prefer Aristotle’s nomisma and Szabo’s work on top of it [1].
Based on what I got from this post, humans are evolutionarily wired to like shiny things, and this gives us a competitive advantage. Instead of lazying around to enjoy the sunset after a good harvest, our ancestors would have used the time to make elaborate collectables for its artistic qualities. Despite collectables not being something you can eat or live in, they happen to be a good insurance mechanism during hard times. Other tribes, though untrusting and hostile, also like shiny things and will share their food with you for collectables during emergencies where you need to depend on the kindness of strangers, turning them into a way to store the value of the leisure time used to create them. I wouldn't call this currency because collectables have no standardized value, its value being entirely subjective based on artistic preferences. One hunter might really like the color of your beads and give you two pheasants. Another might only give you half, but will give you three for your ivory comb instead. But this system is enough to sustain basic barter circles between foreign tribes and account for social obligations over time.
This evolutionary tendency to like shiny collectables combined with standardized states is what I think creates our first metallic coins. King Croesus of Lydia is facing a major war crisis as the ancient near east enters an age of chaos and decides to hire foreign mercenaries. He can't pay them with standardized tax tokens since they aren't his citizens to tax. He can't negotiate a tribute with their leader in luxury goods and jewelry because they're ragtags gathered from all over the land with no single leader. What he can do is melt down his golden collectable statues, divide them into standard units reminiscent of existing tax tokens, stamp them with his symbol, give each individual soldier a coin for every battle they fight, and appeal to our inherent love for shiny things. And so the first gold coin based on the existing idea of clay tax tokens is born.
You’re right that the social/cultural value associated with those luxury gold collectibles is an important part of the story. But I don’t think you need any other weird evolutionary argument in addition to Graeber’s view to complete the full picture.
The reason was because now the various city states needed a way to convince people outside their jurisdiction to also provide goods and services for their wars, and those people could not be cowed into using clay. Clay only works within a states jurisdiction because they could punish counterfeits and ensure it's value among subjects as a tax credit to the state. Outside a state's jurisdiction, foreign soldiers have no guarantee that those clay tokens will continue to be a secure store of value. What if the state you were fighting for collapsed tomorrow? Now their tax credits are worthless. Gold, on the other hand, holds it's value because people like shiny things regardless of which king is stamped on it. But because the gold is issued by a state with their tendency for standardization, now your gold trinkets come in nice standardized units instead of a random assortment of jewelry, what gold would have been used for previously.
A paragraph from Chapter 8:
> Why? The single most important factor would appear to be war. Bullion predominates, above all, in periods of generalized violence. There's a very simple reason for that. Gold and silver coins are distin guished from credit arrangements by one spectacular feature: they can be stolen. A debt is, by definition, a record, as well as a relation of trust. Someone accepting gold or silver in exchange for merchandise, on the other hand, need trust nothing more than the accuracy of the scales, the quality of the metal, and the likelihood that someone else will be willing to accept it. In a world where war and the threat of violence are everywhere-and this appears to have been an equally ac curate description of Warring States China, Iron Age Greece, and pre Mauryan India-there are obvious advantages to making one's trans actions simple. This is all the more true when dealing with soldiers. On the one hand, soldiers tend to have access to a great deal of loot, much of which consists of gold and silver, and will always seek a way to trade it for the better things in life. On the other, a heavily armed itinerant soldier is the very definition of a poor credit risk. The econo mists' barter scenario might be absurd when applied to transactions between neighbors in the same small rural community, but when deal ing with a transaction between the resident of such a community and a passing mercenary, it suddenly begins to make a great deal of sense.
[1] This is a simplification, the real optimal rate of inflation is really -0.8% as shown in the seminal paper The Optimum Quantity of Money.