Monday, March 15, 2021


1. Over the last month or two I've been following an interesting archaeological debate over the discovery of coinage. I thought I'd share it with you.

2. It's generally accepted by archaeologists and numismatists that the first coins were invented in Lydia, modern day western Turkey, in the 7th Century B.C.E. (i.e. 610 B.C.E. or so). The idea quickly spread to Greece. The Lydians used electrum, a strange silver/gold mix, to make their discs. (I wrote about electrum coins here). I've included an example below.

Electrum coin from Ephesus, 625–600 BC [Source]

We don't know exactly why the Lydians used electrum, or even if they treated their discs in the same way that future generations would use coins. But when the Greek city states copied Lydian coinage in the 6th Century, they didn't use electrum. Their coins were pure silver.

3. Lydia's electrum coins aren't the topic of this post. The debate that I'm going to describe revolves around the belief among some archaeologists that a form of proto-coinage had been invented prior to the Lydians and their electrum coins. This proto-coinage came in the form of sealed and regulated bags of hacksilver (more on hacksilver later).

Others archaeologists disagree. They are adamant that Lydia remains ground zero for coinage.

For lack of better terminology, I'll call the first group of archaeologists, those who think there was a predecessor sort of coin, the proto-coiners.

So relax and follow along.

4. By the way, mine is an outsider's account on the archaeology of money and coinage. I am not an archaeologist, so I will certainly get a few things wrong. Nevertheless, I am hoping that my regular monetary economics readership will enjoy learning how archaeologists attack the problem of money.

5. How popular was silver in ancient society?

"Silver served as the main measure of value, the means of payment and credit, and as an indirect form of exchange in Near Eastern economies from the mid-3rd millennium onward," write archaeologists Tzilla Eshell, Ayelet Gilboa, Naama Yahalom-Mack, and Ofir Tirosh (Eshel et al) in a 2018 article entitled Four Iron Age Silver Hoards from Southern Phoenicia. The Near East is a catch-all term for modern-day Israel, Iraq, Iran, Jordan, and Syria. I will return later to Eshel and coauthors' paper.

Morris Silver, an economist who researches ancient economies, describes Mesopotamian texts of the middle of the second half of the third millennium that show silver being used by street vendors, to pay rent, purchase dates, oil, barley, animals, slaves, and real estate.

According to archaeologists Seymour Gitin & Amir Golani (2004), Assyrian economic texts from the 7th C B.C. show that the majority of all types of payments were already being made in silver, including those for tribute, craftsmen obligations, and for conscription and labor commutations.

Cuneiform tablet, loan of silver [Source: The Met]

The Old Assyrian cuneiform tablet above from around 1900 B.C.E. says that 6 minas (c. 3 kg) of silver are owed by two men to the merchant Ashur-idi. One third of the loan must be paid by the next harvest and the rest at a later date. If it is not repaid by that time it will accrue interest charged at a monthly rate.

6. If silver had already become a sort of medium of exchange sometime between 3000 B.C.E. and 2000 B.C.E., it wasn't in coin form, but as hacksilver. By hacksilver, what is usually meant by archaeologists is silver ingots, hacked pieces of ingot, silver scrap, and cut up bits of silver jewellery.

Below are some examples of hacksilver:

7. One reason for the hacking or cutting-up of silver may have been to make small change. If 2 grams of silver was required to make a payment, but a payee only had a single 10 gram ingot, then a small part of it had to be cut off.

8. Another less obvious reason for hacking, suggested by Eshel & coauthors, is that it may have been a way for merchants to check for quality. Pure silver is soft. Mixing copper into silver makes for a harder ingot. A solid smash to the ingot may have been the accepted way of verifying whether an ingot was good silver or not.

9. It wasn't till 610 BC or so that the Lydians made the first coins. So for over a thousand years, silver circulated as a medium of exchange, in hacked form.

10. The big innovation with coins is the stamp. Because we trust the issuer's brand, we needn't weigh out or assay (i.e. smash/hack) silver prior to engaging in trade. So trade was much more fluid.

If you think about it, branding metal is a pretty big step for a society to take. It means that laws, norms, and institutions have become established enough for people to be confident in something as abstract as an issuer's emblem. Too much fraud, warfare, and lawlessness, and branding breaks down—you've got to go back to weighing and hacking silver yourself.  

11. The proto-coiners don't agree that Lydia was the first to "brand" silver. They suggest that bagged and sealed hacksilver was already circulating in a way similar to coins. Some authority, perhaps a government administrator or a merchant, pre-weighed a certain amount of good hacksilver, bagged it, and affixed their seal to it. And so anyone who was offered the bag in trade could treat it just as they would a coin. As a verified amount of silver, it needn't be weighed or hacked. The bag would have been accepted according to whatever information was inscribed on the seal.

12. If the proto-coiners are right, that means our ancestors were better monetary innovators than we originally thought. It pushes the effective date of coinage technology back by 500 or so years.  

13. It's a fascinating debate, especially because it invokes a set of mysterious old hoards that archaeologists have discovered over the years. These hoards are typically hidden in clay jars underneath the floors of houses by their owners, probably for safekeeping. And then they were forgotten or some disaster befell their owner, only to be rediscovered thousands of years later. Who were these people? Why did their hoard get forgotten?

14. One of the key hoards around which the debate revolves is the Tel Dor hoard, which was found north of Haifa in Israel. It was excavated in the 1990s by Ephraim Stern, an Israeli archaeologist at the Hebrew University of Jerusalem. One element of the Dor hoard is an old jug filled with silver, below.

The Tel Dor hoard (a) as displayed in The Israel Museum and (b) in situ, looking east [Source: Eshel et al]

15. Stern's description of this jug (published in this 2001 paper) quickly filtered into the archaeological community. Christine Thompson, archaeologist and co-founder of the Hacksilber Project, used Stern's findings to build a proto-coinage argument. It's worth getting into the details of her argument (and subsequent rebuttals) to see how archaeologists think. You can find it in her 2003 paper, Sealed Silver in Iron Age Cisjordan and the ‘Invention’ of Coinage.

Thompson (channeling Stern) tells us that the Tel Dor hoard dates to somewhere between 1000 B.C.E and 900 B.C.E. The hoard consists of a jug containing 17 bundles of hacksilver wrapped up by linen cloth (see photo of one of the bundles below).

16. Together the silver weighs 8.5 kilograms, which at today's silver price is worth around $8,000. But that's not a great way to think about how much this hoard was worth. According to a very readable article by Tzilla Eshel, a half-gram of silver was equivalent to a day-worker's wage. So the entire hoard was the equivalent of forty-six years of labour. In modern day terms, that would value the purchasing power of the hoard at well above $1 million.

17. Stern has speculated that the hoard may have belonged to a Phoenician merchant who used the silver to build and equip ships, or to buy merchandise for eventual exchange with countries in the western Mediterranean.

18. Most of the linen wrapping in the Tel Dor hoard has long since disintegrated. Thompson notes that the bundles were closed with bullae, or clay seals. (See below). But these seals do not contain a name, just a pattern.

One of the silver bundles found at Tel Dor, and an illustration of a clay bulla, or seal. Source: Ephraim Stern in The Silver Hoard from Tel Dor [pdf]

19. If the bundles found in the Dor hoard were treated by their owner as coins, then one would expect them to weigh a standard amount, just like all modern nickels and quarters weigh the same. And that is the gist of Thompson's argument. According to her, the 17 bundles all appear to be the same weight. .
20. One of these bundles had been removed by Stern to be weighed. It registered at 490.5 grams. Thompson suggests that this 490.5 grams might align with usage of the Babylonian shekel unit of account. Under this archaic weight standard, each shekel weighed 8.3g, and 60 shekels was worth 1 mina. Thus a mina would have weighed 500 grams.

So the 490.5 gram bundle found at Dor comes close to an even mina. It's as if the bundle was a very large denomination mina coin. Thompson attributes the missing 10 or so grams to loss of a few small pieces due to disintegration of the cloth.

21. The purity of the Dor hoard is quite high, notes Thompson, suggesting that the bagged silver, like a coin, had been checked and regulated.

22. Thus Thompson has created a plausible theory about bags of silver being treated as coins. Like a coin, the sealed bags of hacksilver found at Dor had a certain purity and weight. Presumably people who received them in payment didn't have to weigh the silver. Nor did they have to assay the silver by hacking or smashing it.

23. It's a convincing theory. Now for the counter-theory.

24. Raz Kletter, an archaeologist at University of Helsinki, is not convinced by the idea of a proto-coinage. In a 2004 paper, he points to the nearby Tell Keisan hoard, dated to around 1000 B.C., which also contained wrapped hacksilver bundles. The hoard includes 6 or 7 bags of cloth, says Kletter. Two of them weighed in at 24.5 and 25 grams, suggesting that they may have conformed to the same denomination. But two of the other bags measured at a 32 and 100 grams respectfully, which muddies the waters. Moreover, Kletter says that the weights of the bags does not correspond clearly to any known standard of weights and measures.

25. Tel Dor hadn't finished telling its story, either. Recall that at the time Thompson was writing, 2003, only one of Dor's 17 bundles had been weighed. It came in at 490.5 grams, and Thompson ascribed to this bundle the possible value of a clean mina of 500 grams, less a few grams due to thousands of years of wear and tear.

But in 2018, Eshel & co-authors opened a second Dor bundle. They found that it measured just 420.6 grams, which doesn't conform as closely to a mina. So that undermined some of the arguments in favor of proto-coinage.

26. That's not all. Eshel & co-authors did a chemical analysis of four hoards including both Tel Dor and Tell Keisan. Recall that Thompson suggested that the purity of Tel Dor silver indicated a degree of regulation, much like a mint controls the silver content of a coin. But Eshel & co-authors found that the silver from Tell Keisan, though "piously" packed in sealed bundles, was not so pure. It contained large amounts of copper, suggesting that it was a forgery. (See photo below). If the whole point of bagging and sealing was to create a trustworthy medium of payment, the deliberately-alloyed Tell Keisan silver seems to contradict this.

A bundle of hacksilver from Tel Keisan. Its green colour betrays its copper content. When silver corrodes it tarnishes black, but copper produces a green rust. Source: The Torah

27. This collection of counter-observations somewhat weakens the argument for an early proto-coinage in the Near East. But there are probably plenty of yet-to-be discovered hoards. Who knows, perhaps the next one will contain bundles of provably standardized hacksilver. It certainly is a provocative idea.

28. If the role of bundling and sealing of hacksilver wasn't to create a proto-form of coinage, than what was its function? Eshel & co-authors suggest that bagging was little more than a convenient manner of storing one’s wealth. Taking out a single cloth bundle and weighing it would have been much less awkward than removing individual pieces one-by-one and weighing them. 

If you're interested in learning more about the ancient hacksilver economy, I'd suggest reading How Silver Was Used for Payment, recently published in The Torah. Tzilla Eshel, the archaeologist who co-authored one of the papers I cite in my blog post, is the author and has written it with the lay-person in mind.


  1. "Morris Silver, an economist who researches ancient economies"

    Nominative determinism strikes again

    1. I believe he's referring to the idea that people do a job that corresponds with their name - eg someone called Mason who works as a potter.

  2. hello Ann Pettifor here. If you are really interested in learning more, read Michael Hudson on the Origins of Money and Credit...

    1. That article looks quite interesting (just as this post was interesting and enjoyable). The topic reminds me of Graeber's Debt, a highly readable book which I enjoyed--though as an amateur I'm in no position to verify its scholarship--and which I see in the Hudson bibliography.

  3. Even if the bags were all pure silver in exact weights that hardly proves coinage. I would be inclined to weigh and sort my silver in bags just so I know how much there is. What you need is some way to prove that these bags could be sealed in a way that's both tamper proof and could only be done by the mint.

    1. Yes, you bring up some good points. The bar to proving coinage is very high. Thompson suggests that the bullae that sealed the bags may have been an effort to authenticate their silver content. I don't know enough about bullae to be able to say how tamper proof they were.

  4. I recall reading about (terrible citation, I know) something about ancient near eastern units of value, consisting of a sealed container with an authoritative description that both asserts the value of its contents, and a curse on anyone bold enough to check.

    Likewise possibly a later debased lead-silver coinage about one in ten examples of which contain a reasonable quantity of silver, and all of which curse the person who dares test the coinage.

    In all cases, the curse is self-fulfilling by destroying the value.

    1. OK, the latter has real citations, but 12th century AD.

    2. Interesting nonetheless. Who issued these things?

    3. The "curse dirham" gets a couple page of discussion here, describing copper or silver-plated copper coins potentially intended to pass as the continuation of a traditional silver coinage.

    4. And here's a decent image

    5. Thanks, I'll check those links out.

  5. In the book "5,000 years of debt” or something like that, the author provides evidence that tallying credit was FAR more prevalent than the actual exchange of units of value (i.e. coins or proto-coins). Just like the tablet above. The earliest writing pretty much consists of tablets tallying who owes what to whom. If that's the case, the need for proto-coins would be pretty limited. And perhaps only large transactions (by merchants) would require actual silver for payment. This seems to make a lot more sense for the silver bag findings, then the theory of some standard-issue medium of exchange.

  6. Thanks for the post, JP! The timing is interesting, as I have just received multiple books on Sumer, ancient Near East and archaic bookkeeping :-) I will have more to say later, but let me start with some thoughts.

    Silver was no doubt used as a medium of exchange/payment by some people, in some transactions. But I think, like Kevin above, that there might be a risk of seeing it used more widely than it was, due to it being used as a unit of account. Cuneiform tablets are recordkeeping/accounting, and the sums there are, of course, denominated in a common unit of account, which happened to be a specific weight of silver. Generally, this kind of records are kept because there is credit/debt involved. Now, if the tablet says that Abraham owed 10 shekels of silver, how will we ever know if that meant that he had to deliver silver, or that he could as well deliver a goat and some barley worth 10 shekels of silver? I guess we don't get to know it, unless there is a tablet where it was recorded what Abraham delivered.

    Regarding the bullae/seals, those kind of things were used even earlier (8000-3000 B.C, I think) to tamper-proof records consisting of intrinsically worthless clay tokens. That kind of tokens (with a hole) could be hanged on a string and the ends of the string would be sealed with a bulla. The other option would be to have a large bulla (called a clay ball or clay "envelope" by some) that would hide inside it many smaller tokens, as records of debts, that could only be accessed by breaking the bulla.

    Anyway, I think the archeologists should welcome your effort to understand and interpret their findings. Many of them cannot be at the same time monetary historians and might thus be "prisoners" of the layman way of understanding money, which will then be reflected in their interpretations.

    1. Hi Antti,

      Some commenters seemed to want to push the discussion to a different debate, the more famous one about credit vs money. But I don't want to go there; my post was an effort to unveil a different debate about the history of coinage, one that most of us (myself included) don't know much about.

      You make a good point about whether silver was actually used as a generally-accepted medium of exchange or more as a popular unit of account in ancient Near Eastern economies. In my post I am only echoing what the archaeologists that I read had to say, i.e. Eshel et al: "Silver served as the main measure of value, the means of payment and credit, and as an indirect form of exchange in Near Eastern economies from the mid-3rd millennium onward."

      Perhaps in future posts I can get into the popular credit v money debate, although I doubt I can add much. I suspect I'll probably do some more hacksilver posts in 2021.

  7. Great discussion. Just to note, that the second Dor bundle which came in at 420.6 g if divided by 50 units rather than 60 you get 8.412 which is totally in line with a 50 shekel Mina. Standard‘s were established according to inflation and availability with different cities having different standards than the one of the city next to it only a few kilometers away, so, the argument of another bag weighing only 50 shekels rather than 60 is an insufficient argument to refute Thompson’s theory. Regarding The purity of the metal, well, this particular hoard dates to the dark ages of that period when Silver was not that readily available it may have been either that it was the norm to pass on lower quality silver just like the later Roman silver Antonioni that were nothing more than silver washed bronze coins but passed on as the days denarii by order of the empire or, perhaps this particular Dor assayer may not have been sophisticated enough to detect the greater amount of arsenic contained within. It is interesting why anti proto coiners do so easily away with the well known discovery of the three Bar Rakab inscribed ingots found at samal that are coins in every sense regardless of its size and pre date well over a century and a half to any of those of the earliest Lydian marked lumps/coins.