When Money Talks: A History of Coins and Numismatics 
by Frank L. Holt.
Oxford, 336 pp., £25.99, October 2021, 978 0 19 751765 9
Show More
Coin Hoards and Hoarding in the Roman World 
edited by Jerome Mairat, Andrew Wilson and Chris Howgego.
Oxford, 368 pp., £90, May 2022, 978 0 19 886638 1
Show More
Show More

All the shops​ on my university campus have gone cashless. The vendors at our farmers’ market use little card readers that plug into their phones. We use cash very little these days and coins even less. By recent historical standards this is an aberration. Since the 1850s and 1860s, when banknotes were standardised in Britain and the United States respectively, most people, most of the time, have thought of money in terms of coins and paper banknotes, marked with numbers to indicate their value in relation to one another and to the goods and services for which they can be exchanged. For many centuries before that, coins alone had that role. Money can exist without leaving a trace behind. Coinage cannot, which makes it valuable for historians. The closer we approach the present, the less coins tell us what other sources don’t tell us just as well or better, but coins from antiquity are uniquely valuable, both for the information encoded on their surfaces and for the light they throw on all sorts of questions that other sources cannot readily answer.

The western history of coinage begins in southwest Asia Minor, where electrum, a naturally occurring alloy of gold and silver, is found in rivers such as the Pactolus and Maeander. In the sixth century BC, local big men began putting marks on the little blobs of electrum that were already serving as money, first punching stamps on one side, later adding more complicated designs on the other. Unlike modern coins, which are milled, or ancient Chinese coins, which were cast, western Eurasian coinages were struck. A coin blank was placed over a die secured in an anvil, engraved with what would become the obverse or ‘heads’ face of the coin. A second, reverse die was aligned above the blank and struck with a hammer to impress the image on each side. Not long after these proto-coins were invented, it was realised that electrum could be refined into its component metals and so gold and silver coins began to be struck as well. While all three metals had intrinsic value, once the markings on them were calibrated to correlate with size, the metal lumps no longer needed to be weighed. Their value as bullion was unchanged, but they had become coins.

The Greeks, with their addiction to origin stories, ascribed the invention of coinage to Croesus of Lydia, and it is certainly true that he minted the first gold coins of fairly uniform weight and purity. The Greeks adopted the new technology with alacrity, and Persian kings from Darius onwards struck coins for their western satrapies and to pay Greek mercenaries, although most of their empire continued to use older means of exchange. In island and mainland Greece, coins swiftly replaced the iron spits that had until then served as money, though Greek vocabulary preserved the memory of that origin: the large silver coin that became ubiquitous was called the drachm, derived from the word for ‘handful’, and its fraction was the obol, from obeliskoi, ‘spits’ – six obols, a handful, made a drachm. Like everything in the ancient Greek world, the coinage is so varied that it defies summary. More than 1600 city-states struck their own coins at one time or another, using scores of incompatible weight systems, with only a very few – the drachms of Athens with Athena on the obverse and her owl on the reverse; the colts of Corinth with their image of Pegasus – gaining acceptance outside local or regional markets. Alexander the Great, who struck drachms on the Athenian weight standard, made that standard universal, while his conquests monetised the Persian Empire as far as Central Asia, so that Greek traditions eventually became hybridised with those of the Indian subcontinent.

By then, Greek colonisation in southern Italy and coastal Spain and Gaul had brought coinage to the western Mediterranean, where it was slowly adopted by various Iberian, Italic and Celtic peoples. The Romans did without their own coinage much longer than many other Italian tribes, content with irregular hunks of bronze that had to be weighed to determine their value. At the beginning of the third century Bc these were replaced with more or less regular cast bronzes that weighed as much as eight ounces and were marked to show their nominal value. Only during the Second Punic War, which began in 218 Bc, did they begin to produce precious metal coinage in the manner of the Greeks. The first silver denarii appeared halfway through the war, in 211. Coins were struck in increasing quantities as the republic grew richer and richer, and were the responsibility of the tresviri monetales, junior magistrates just embarking on their political careers and with a lot to prove. Familial and ancestral achievements or episodes from Rome’s real or legendary history were depicted on these coins, in a riot of symbolism that is allusively richer than anything in the Greek tradition, though almost never as aesthetically pleasing.

As the republic fractured, Julius Caesar became the first living man to see his portrait on a Roman coin; his ideological heir, Mark Antony, and his legal heir, Octavian, followed suit, as did some of their rivals. After Octavian, as Augustus, cemented Caesar’s autocracy behind a façade of republican restoration, ruler portraits, long a feature of Hellenistic monarchy, came to grace the obverse of Roman coinage. The silver denarius, already two centuries old, remained the basis of the Roman monetary system, but Augustus also introduced a small, very pure gold aureus and a token coinage – where the coin’s value is greater than that of the metal itself – in orichalcum (brass of 80 per cent copper and 20 per cent zinc) and copper. The largest of these base metal ‘bronze’ coins, the sestertius, was worth a quarter of a denarius. This Augustan system survived with only minor changes for more than two centuries, leaving a permanent imprint on our vocabularies: the denarius gives us the French denier, the Spanish dinero and the Arabic dinar, as well as the ‘d’ that stood for a pre-decimal British penny. (The £ symbol is a stylised capital L, from the Latin libra via the French livre, the Roman pound from which 96 denarii were meant to be struck.)

Modern collectors tend to see this early imperial tri-metallic system as the apogee of ancient coinage, conditioned in part by the huge impact of imperial coin portraits on the Renaissance tastes that still shape our aesthetic preferences. It is true that coins of the early empire can be very attractive, especially the large sestertii, struck twelve to the pound, many of which have acquired a lovely patina over the centuries. But for scholars, what followed the collapse of the Augustan system is every bit as interesting. There were relatively minor debasements of gold and silver coinage in the first and second centuries AD, and the smallest of the fractional bronzes ceased to be struck, but a major debasement at the very end of the second century was especially damaging. Perhaps to counter inflation, in the 210s the emperor Caracalla introduced a new coin tariffed as a double denarius, although it weighed less than twice as much as an old silver coin and contained only half as much again silver. We don’t know what contemporaries called these coins (our use of ‘antoninianus’ is a guess), but because the imperial portraits on them wear a radiate crown, all these third-century coins are known as radiates, which is practically synonymous with ‘cheap’ and ‘mucky’. They were struck in very large quantities across the third century, shrinking all the time in size, weight and purity, driving the denarius out of production and making any token bronze coinage uneconomical to mint. By mid-century a theoretically silver coinage had become base metal, pure silver and bronze had ceased to be coined, and a rare gold coinage circulated as bullion whatever its nominal tariff. People remembered the tri-metallic system, but every attempt at restoring it failed, because all the bad money in circulation immediately drove new, good money out of the market and into hoards and savings.

The emperor Diocletian, who came to power in 284 and brought to an end half a century of near continuous civil war, attempted to reform the currency, as he did much else, but a stable monetary system was only achieved under his successor, Constantine. Rather than being concentrated in Rome, mints were distributed around the empire, striking enormous numbers of bronze coins, a small amount of silver and a plentiful new gold coin, struck 72 to the Roman pound. Like the old denarius before it, the Constantinian gold solidus was both very pure and linguistically productive, hence the modern Spanish sueldo, the Italian soldi, and the English soldier, a man paid in coin for military service. The bronze coin or nummus was struck to a very imprecise weight standard and came in different sizes, the relationships between which are so poorly understood that we simply number them from one to four based on their diameter. But Constantine’s efforts did not restore a true tri-metallic system. Instead, the solidus served as a standard unit of measure, but floated against bronze and silver depending on its value as bullion; this left a lot of space for arbitrage, because pay packets were in bronze but taxes had to be paid in gold. This dispensation lasted for a century or so, until the western Roman Empire limped to its end in the course of one short generation. Both silver and small change ceased to be minted, but the solidus and especially its one-third fraction, the tremissis, formed a thread between classical antiquity and the Latin Middle Ages. The story is similar, if somewhat more complicated, in the eastern Roman Empire, which bequeathed its sixth-century monetary traditions to the early Islamic world.

Agreat deal​ remains to be learned, particularly where the technical study of coinage intersects with larger social and economic questions, but the work is hampered by the lowly status of numismatics in contemporary academia. Among the so-called auxiliary sciences of history, only heraldry and genealogy are taken less seriously, still carrying the stigma of antiquarianism from which epigraphy and archaeology long ago liberated themselves. All of these fields stemmed from the curiosity of Renaissance humanists and the collecting mania of noblemen with their Kunstkammern. The first serious studies of ancient coinage, on the Romans by the French humanist Guillaume Budé at the beginning of the 16th century, and on the Greeks by the Austrian Wolfgang Lazius fifty years later, were analytical and remain worth reading. But they were outshone and far outnumbered by essentially compilatory works, the long lists and catalogues that culminated in the work of Joseph Eckhel, another Austrian and curator of the imperial coin cabinet in Vienna. His Doctrina numorum veterum invented some of the classificatory principles with which we still work, for instance the sequence in which Greek coinages are presented in catalogues, clockwise from Spain to India and then back around to North Africa. Although Eckhel’s work was a masterpiece, the abiding image of the numismatist was that of the antiquarian worker ant, hoarding details but not very interested in what they meant. Not even Theodor Mommsen could rescue the field, although he managed to place epigraphy at the centre of ancient historical studies.

As well as this antiquarian taint, there is also the whiff of trade: academic numismatists have to rub shoulders with amateurs, many of whom possess great technical knowledge, and professionals who work for commercial concerns. The same can be true of art historians, but at least they often have dealings with the donor class whose private collections, through gift or loan, grace the world’s museums. With coins, all but the most splendid rarities cost less than a mediocre Roman bust, while at the low end of the market, Pokémon cards are a better investment. So not just sordid but déclassé. On top of that, there is a very real problem with provenance. While dealers and auction houses keep the existing stock of ancient coins recirculating, the supply of new coins, from great rarities to the most common late Roman bronzes, comes overwhelmingly from war zones or from countries where clandestine excavation is enabled by lax law enforcement and venal officialdom. Bulgaria has for decades been the source of huge caches of Greek and Roman coins, while Syria has been disgorging bronzes by the tens of thousands with the distinctive patina that sand gives to coins.

Even before Russia’s illegal invasion, Ukraine was a target for clandestine excavation, because in antiquity its coastline was dotted with Greek cities and its fertile interior was home to a succession of barbarian polities whose inhabitants were alternately recruited as mercenaries or paid to keep the peace. Recent unprovenanced finds have shown that third-century seaborne raiders carried off not just the treasures of Alexandria Troas, but also its mintworkers and the dies from the mint, which explains the sudden proliferation of imitative coinage in the region. In Iraq and Afghanistan, the Bush-Blair wars unearthed quantities of important evidence. Newly discovered coins from what must have been military mints have given us hard evidence and a rough chronology for the nearly continuous warfare between Persian shahs and the peoples of the Central Asian steppe. The extent, sophistication and chronology of eastern Hunnic polities in the Pamirs, Hindu Kush and Sind, which point up how flimsy and transitory Attila’s European empire was, are also much clearer thanks to new coin evidence.

But none of these coins is provenanced. One in a thousand at best will have been legally exported. How can we use their data ethically? We can set aside the obvious no-nos: don’t deal directly with clandestine excavators; avoid compromised auction houses; don’t obfuscate the provenance of an important coin to protect a collector. But beyond that lies a large grey area. Dispersed hoards filtered into the trade via multiple dealers and auction houses aren’t hard to spot in the catalogues, and it seems wilful to ignore the information: it’s out there whether one uses it or not. But some argue that every single find whose archaeological context is damaged or destroyed is a loss to science, and that to use the data is to be complicit. Who’s right? It’s not just that there are no answers that will satisfy everyone, it’s that a satisfactory answer does not exist.

Frank L. Holt tackles the problem in a chapter he frames as a classroom discussion between a lecturer, a preternaturally learned coin collector and students playing devil’s advocate. His verdict is that we should take the curatorial approach used with dinosaur bones, where paleontologists and fossil hunters have a symbiotic and co-operative relationship. This seems less realistic in Helmand than in Wyoming. Holt has written several excellent books on Alexander the Great and his successors in Central Asia, all of them with substantial numismatic content. When Money Talks, by contrast, is rather eccentric, with puns and dad jokes and a chapter on ‘the coin’s point of view’. Philip Grierson’s magisterial Numismatics (1975), sadly out of print, remains a better introduction to the subject, though Holt’s book is more likeable. It also conveys something very important: holding a coin that someone else held two thousand years ago creates a special feeling of connectedness. Anyone who has handed a bag of cheap Roman bronzes around a room of bored undergraduates will have seen first-hand the way it electrifies the atmosphere. Coins make history feel real. And there are an awful lot of them out there.

Ancient Roman coins in particular were mass-produced on a scale that beggars the imagination and shipped to every corner of the empire, ending up as far afield as East Africa and southern India. The Mildenhall Hoard, found in Wiltshire in 1978, probably the most famous Romano-British find, weighed 180 kg and contained more than fifty thousand low-value coins from the third century. The 1918 Komin hoard, found near the mouth of the Neretva river in Croatia, contained around three hundred thousand coins from the same era. Right now, half a million Roman coins from Egyptian hoards, mostly uncleaned and unconserved, are stored between museums in Cairo, Alexandria, Toronto and Ann Arbor. More turn up all the time. A Spanish badger made international news last year when it uncovered a hoard of several hundred coins in an Asturian cave. Even for scholars, the fact of the badger was more interesting than the utterly predictable fourth-century bronzes that made up the hoard.

What survives is thought to represent about 1 per cent of the coinage originally struck; almost all of them were part of a hoard. Stray finds – single coins lost by the roadside or dropped at the bazaar – make up a tiny percentage of survivals. Yet for most of the history of numismatics, coins have been studied individually, for their iconography, their contributions to political geography (there are, for instance, Greek cities attested only by their coins), or for what their inscriptions tell us about the structure of government or the development of rulers’ titulature. Less happily, they have been used to build elaborate historical narratives of places and periods for which written evidence is virtually non-existent, like the Greco-Bactrian kingdom, most of whose rulers are known solely from coin portraits and legends. Third-century usurpers attract similar excesses. Jotapian, Uranius Antoninus, Ingenuus and Regalianus are all barely known apart from their coinages. To this list we may perhaps now add the Sponsianus whose anomalous gold coins, housed in Glasgow’s Hunterian and long dismissed as modern forgeries, sparked a Twitterstorm in December, when an earth scientist suggested that mineral deposits and surface abrasion authenticated their antiquity. The jury is out – the case for forgery either modern or ancient is very strong – but even if real, the lost history of Roman Dacia they are purported to reveal is in fact pure fantasy.

However valuable the information ancient coins provide as individual artefacts, even more valuable is what they teach us in the aggregate. A hoard can be defined as two or more coins deposited together, usually but not always intentionally, and not subsequently recovered. Since a hoard’s owner usually intended to retrieve it, the non-recovery acts as proxy evidence for its historical circumstances. Broadly speaking, there are four categories of hoard, each defined by the way it was created: accidental loss, sudden response to an emergency, storage of savings, or deliberate abandonment. Each tells us different things. Accidental hoards are things like dropped purses, or the basket of coins discovered on the docks of Herculaneum, now a fused mass of bronze and silver bearing the imprint of the container that was incinerated along with its owner in the pyroclastic surge of Vesuvius. Emergency hoards are what they sound like: collections secreted to prevent discovery in a moment of crisis, often war or the threat of it. Savings hoards are surplus wealth kept in hidden reserve by their owner. Finally, there are deliberately abandoned hoards, such as grave deposits, sacrifices and votive offerings. Each type reveals different things. Accidental and emergency hoards can be dated, usually closely, by the youngest coin they contain, and offer a snapshot of the currency circulating in a particular place at a particular time. Savings hoards represent deliberate selection over time. This generally means higher rather than lower value coins, and better quality specimens, whether measured by weight, level of wear, or precious metal content. These hoards are harder to date, because their youngest coin may come from long before the time of deposition. For that reason, they don’t provide a snapshot of monetary history, but they do reveal social choices about relative monetary values and economic decision-making. Finally, abandoned deposits represent another kind of deliberate selection, and were more likely to contain demonetised pieces which were no longer usable as currency.

Collections of scholarly articles are rarely coherent, but Coin Hoards and Hoarding in the Roman World contains plenty to enlighten the non-specialist and surprise the expert. Both emergency and savings hoards can tell us a lot about coins’ longevity. There was, for instance, a five to tenfold increase in circulating silver between 200 and 50 Bc and no mechanism for withdrawal save natural wear and tear. Because Republican denarii were so pure, even very worn examples were hoarded well into the second century AD. Particularly large issues from the last century Bc – for instance, the massive issue of slightly underweight ‘legionary’ denarii struck by Mark Antony in the run-up to his final war with Octavian – distorted the record for a hundred years or more. It’s not surprising that the huge sums needed to pay for a standing army of several hundred thousand men should have a dominating impact on supply in the most heavily garrisoned provinces, but consideration of things like die-links can show us how likely coins from the same issue were to stay together over time. (Die-link studies, tedious but valuable, require meticulous comparison to match coins struck from the same dies, either on both sides or on one: reverse dies, which absorbed most of the force of the hammer, wore out faster than obverses.) There are some unexpected findings when we use hoards to trace how coins circulated. We know for a fact that the emperor Postumus, who ruled a separatist Gallic regime for a decade in the third century, was recognised in Spain because a great many inscriptions tell us so. But while coins bearing Postumus’ image are common in Gallic and British hoards, they are hardly found in Spanish ones. This opens up new lines of inquiry about how, or whether, Postumus paid the legionaries garrisoned at León.

Finally, evidence from accidental and emergency hoards in particular helps us understand people’s actual economic lives rather than the normative lives implicit in government issues. The proliferation of unofficial coinage and the imitations of official coinage in various parts of the Roman Empire in the second, third and fourth centuries used to be interpreted as a sign of crisis (it was seen as an example of ‘siege money’, or Notmünzen). Hoard evidence now shows the reason for the existence of such coins. Ancient states coined money to make payments, to assert prestige and identity, and occasionally, as in Ptolemaic Egypt, to create a closed economy, but never deliberately to enable a market economy among private individuals. This meant severe coin shortages were a frequent problem. As with the commercial tokens used in Victorian industrial cities or the boiled sweets used in lieu of small change in 1970s Italy, Roman imitative coinages weren’t a sign of crisis but of a highly monetised market economy that demanded more specie than the state was able to provide. The hoard evidence from third-century Gaul and fourth-century Spain, now well studied, shows definitively that official and unofficial coinages were accepted at par and that there was no selection bias in either direction. This completely recentres the way we think about the later Roman economy.

Coin Hoards and Hoarding shows how far from antiquarian stereotypes modern numismatics have come. More important, though, it’s a reminder that cheap, ugly and plentiful coins have as much to teach us as rare and priceless specimens. Like Holt’s amble, this book manages to evoke the pleasures of handling ancient coins and thinking about and with them. It also reacquaints us with a way of using money with which we are rapidly losing touch.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN


Please include name, address, and a telephone number.


Vol. 45 No. 5 · 2 March 2023

Michael Kulikowski complains about the lowly status of numismatics in contemporary academia (LRB, 2 February). There’s certainly truth in that. He doesn’t mention, though, that there are hierarchies even among numismatists, and many of them, focused on classical Greek and Roman coins, look down on paper numismatics. Yet old money isn’t just old coins but old banknotes too.

Paper money first joined metal money in seventh-century China, but didn’t really feature in the West until well into the 17th century. Kulikowski writes that banknotes were standardised in Britain in the 1850s, but in fact England, Scotland and Ireland each have distinct histories in this respect. The 1844 Banking Act did provide a framework leading to the eventual note-issuing monopoly of the Bank of England, but the equivalent Acts in Scotland and Ireland preserved the multiplicity of commercial note-issuing banks we still see today. An earlier Act in 1826 allowed Scottish and Irish banks to continue to issue £1 notes, whereas in England citizens had to make do with gold sovereigns, at least until 1914.

Jonathan Callaway
London SW15

send letters to

The Editor
London Review of Books
28 Little Russell Street
London, WC1A 2HN


Please include name, address and a telephone number

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences