FLAME Numismatic Introduction

Coinage was the principal physical expression of the pre-modern economy, bearing information on the fiscal policies of states and the patterns of monetary activity. The minting of coins was a principal activity of all late antique and early medieval states of Europe, central and western Asia, and northern Africa. The discovery of coins at a distance from their site of manufacture offers evidence for the military and diplomatic outlays of governments and the nature and degree of commercial exchange. The FLAME project seeks to offer a framework for understanding these phenomena by documenting the issue and find contexts of coins in this area for the period from about CE 325 (the end of the reforms of Diocletian and Constantine) to about CE 725 (the establishment of new monetary regimes in most of these areas). The project is divided into two phases, studying minting and circulation respectively.

Phase I: Minting

It was the stamping of the identity of the issuing state onto a specially prepared piece of metal that guaranteed the contents of each coin and the willingness of the state to recognize it in payments to it. There were also imitative issues, often issued by a state to produce an identifiable denomination, as well as contemporary counterfeits which are, by there very nature, difficult to assign to a place of manufacture. States benefitted from their coin issues by having standard currency to pay their employees (military and civil), to buy materials for infrastructure, and as payment to mercenaries and other states. They often also benefitted from a mint charge that produced income for the state treasury. The quantity of coins minted at a given time and place could depend entirely on state decisions and actions or might be subject to the supply of metal and demand of coinage on the part of merchants.

The information stamped on a coin, both in the type [image] and legend [inscription] often provides more information than just the identity of the issuing authority. On a Byzantine copper coin of the sixth century, in addition to the emperor’s name, one can also read the denomination (value in the monetary system), the year of minting (in terms of the emperor’s regnal years), the name of the mint, and even the number of the workshop that produced it. On Sasanian silver coins of the same period, the ruler appears with distinctive identifying regalia and the mint is identified, but some mint names, including that of the most prolific one, cannot be identified with modern locations. Gold coins of western Europe from this century can often be identified only tentatively as Visigothic or Merovingian. In Phase I of the FLAME project we use the results of centuries of numismatic scholarship in identifying specific issues to produce an overview of minting activity in the four centuries studied on as detailed a level as the evidence permits.

Coin issues of the late antique and early medieval periods were usually in one of three metals: gold, silver, and copper, and occasionally in alloys of two or three of these. These metals had different rarities, different capabilities of striking and alloy detection, and different roles within the monetary system.


Gold was the rarest of the coin metals, the most malleable in striking and impervious to chemical change (though in its pure state most subject to wear and damage). Contemporary technology based on the touchstone made it easy for users to recognize debasements of gold coin metal by silver or copper. Most of the gold in coins minted in the period under investigation probably derived from bullion that had been refined countless times in the region since the Neolithic and Bronze ages, in the form of earlier coins, jewelry or cult objects, but sources of alluvial gold were exploited throughout the period, and towards its end newly-mined supplies from sub-Saharan Africa began to reach the Mediterranean.

The principal denominations of gold were derived ultimately from the Roman aureus of late Republican and early Imperial times, which was altered in the beginning of the fourth century to make the solidus, of pure gold, weighing 1/72 of the Roman pound (about 4.5 modern grams), and weighing 24 carats or siliquae. Gold usually served for the payment of large expenses such as diplomatic gifts and wholesale trade. In general, one solidus could be thought of as enough value to supply subsistence living to a small family for a month.

The solidus maintained its standards through the early Byzantine period, though at periods solidi were issued at weights of 20, 22, and 23 grams respectively (an enigmatic phenomenon that we hope to offer new evidence for). The weight and fabric (general relationship of diameter to thickness) of the Islamic dinar represented direct continuity from those of the Roman and Byzantine solidus. The third of the solidus, known as the triens or tremissis, served as the basis for the gold coinage of most of the successor states in the West in the sixth and seventh centuries.


Silver was much more commonly mined in the late antique and early medieval world than gold; in general it had a value of about one/tenth that of gold on a weight for weight basis, but this bimetallic ratio was constantly shifting as a result of supply and demand. Silver was almost as easy to refine and strike as gold, but was more subject to oxidation and other chemical processes. There were no ancient technologies for estimating the purity of a silver coin, other than melting a quantity of them, dividing out the elements, and weighting the remaining pure silver. As a consequence, debasement both in official mints and by counterfeiters was a constant source of concern with silver coinages. Because of the relatively lower value of silver than gold, coins made of it were more likely to carry a surplus value (the difference between the value of the coins and that of their bullion when melted) to cover the cost of minting [called brassage] and, often also some profit to the state [called seigneurage].

The principal silver coinage of the period studied by the FLAME project is the Sasanian drachm, weighing about four grams of relatively pure silver. This denomination was a direct continuation of the Greek drachma introduced into Asia by Alexander the Great and continued by his Seleucid and Parthian successors. The Sasanian drachm served as the prototype for the Islamic dirham, in overall appearance at first, and after the reforms of ‘Abd al-Malik in AH 79 (CE 698-99) in the fabric and general layout of external rings around a central type, now purely epigraphic. Silver coinage also appears to have played a role in the monetary systems of the relatively short-lived Vandal and Ostrogothic states. In the course of the seventh century, the Merovingian, Frisian and Anglo-Saxon gold tremisses went through a debasement that resulted in a purely silver coinage throughout northwestern Europe by the year 700. In general, a silver coin weighing a few grams, was sufficient for subsistence for a day or two.

Changes in the relative values of gold and silver rendered the maintenance of a bimetallic monetary system fraught with difficulties. The example of the coinage of the early Roman Empire was probably known to many of the minters of our period. Augustus set up a system with a fixed value relationship of the basic gold and silver coinage; within a century this system could not be sustained, even within the large and relatively closed Roman economic sphere, and the second and third centuries witnessed significant alteration of the amount of metal in the coins to keep their intrinsic values close enough to their official monetary values to avoid the melting or export of the overvalued coins. This reached a crisis with the drastic debasement of the silver coinage in the course of the third centuries. After the monetary reforms that initiate the period under study, silver coinage was an ephemeral part of the Roman-Byzantine monetary system.


Copper was a relatively common substance in most parts of the area in question, to the point to which a given weight of copper was worth only about one-hundredth that of the same weight of silver and a thousandth that of gold. Copper was usually too soft to be used for durable coinage, so it was frequently alloyed with tin or zinc to produce a coin metal generally called bronze. Bronze was, however, more difficult to cast into blanks and to strike than gold or silver, so the amount of labor used to produce a bronze coin was almost invariably more costly than the value of the metal within the coin. In order for the minting of bronze to be practical, the value of the resulting coins had to be set considerably higher than that of the metal, with the result that their value depended mostly on the anticipation of their being interchangeable with gold and silver issues at a state-supported official rate, producing a coinage whose value was mainly fiduciary, that is based on trust. Because of the expense of reforming bronze flans (blanks), dies for copper coins were sometimes overstruck on existing images or on trimmed down old flans.

As the Late Roman and Byzantine mints issued few silver coins, bronze coinage played a major role in supplying coins for retail use. At least four difference denominations of bronze coins of the fourth and fifth century can be distinguished mainly on the basis of their diameters, but their value in relation to the gold solidus is generally unknown. With the reforms of Anastasius at the end of the fifth century, the bronze coinage bore marks of denomination in a system of 40 nummi and its subdivisions, but the relation of the nummus to the solidus is not always documented. In Egypt under Justinian and his successors, bronze coins bore marks of denomination of 6 and 12 nummi respectively. Bronze appears sporadically in the other coinages of the period, but with the exception of that of the Vandals and Ostrogoths, appears to have had little importance.

Phase II: Circulation

It is the goal of Phase II of the FLAME project to document the find spots of the coins that comprise the issues studied in Phase I. This is based primarily on a thorough search of the literature in archaeological and numismatic publications, but also incorporates information from online databases and also well-provenanced unpublished coins in museums and other collections.

The monetary circulation of pre-modern economies is known primarily from a study of the contexts in which various coin issues have been found. Virtually all existing coins of the period under study have been below ground, though only a small percentage of those that have been found have had their find context recorded before entering modern collections. There were two basic conditions by which coins ended up in the ground: intentional burial with the intention of recovery (hoarding), and unintentional loss. There are a few modes of loss intermediate to these categories, such as inclusion in human burials and deposit in sacred sites. Not only is the survival of ancient coins because of burial in the ground not a random sample of those issued and circulating, the knowledge of finds of such coins depends greatly on such factors as the antiquity laws and numismatic infrastructure of the areas in which the finds occur.

Hoarding appears to have been a virtually universal practice by individuals with coins of high value and no access to banks or other ways to protect their wealth from theft. It is the lack of retrieval by the burier that introduces a non-random aspect to the process; unrecovered hoards are generally more common in contexts of war or natural disaster. Individuals generally buried mainly coins of high value, so hoards generally represent the highest denominations available to a given population. The coins an individual chose to hoard might include mainly those acquired over the course of years or decades (a ‘treasure hoard’), or might be primarily those received in recent transactions (a ‘circulation hoard’); it is rare for the modern numismatist to be able to reconstruct the history of a group of coins buried together with any degree of certainty. Hoards are generally published when their contents are communicated to a governmental agency or a scholar; in many cases, however, they are dispersed into the collector market without such documentation.

The second main way for coins to make their way into the ground is through inadvertent loss, usually of individual pieces. As people are more likely to be aware of the loss of a high-denomination coin and search for it, and as gold and silver coins present more of a visual contrast to earth than do copper-based ones, it is predominately small coins, especially bronze ones, that enter the ground as inadvertent losses. Another common circumstance for the loss of copper-based coins is as a result of their having lost fiduciary value through demonetization of old issues or recognition as being counterfeit; such coins were generally thrown to the ground, sometimes after being cut or otherwise marked.

Coins that enter the ground as stray losses are frequently recovered in archaeological excavations, which can retrieve thousands of coins from times and places where bronze coinage was prevalent, typically Roman and Byzantine sites in our field of study, but may be infrequent at such sites as those of Sasanian Persia or Merovingian France, where copper coinage played a minor role if any.

As intentionally hoarded coins generally represent the highest denominations available, and excavated ones generally the lowest, coins of the middle range, silver or billon (a low alloy of silver and copper), may well be underrepresented in ancient loss and in modern retrieval. A promising effort to fill this gap appears to result from the willingness of some modern governments to recognize and record the finds of individuals operating with metal detectors. Such finds are generally of single pieces or small groups, and if located along roads rather than in habitation sites, are more likely to include intermediate denominations than hoards or excavation coins. The recording and use of information from such finds raises significant ethical and political questions, and at present is limited to only specific modern nations included in our area of study.

In Phase II of the FLAME project, we are recording documented finds of the issues enumerated in our Phase I, in the realization that such distributions are far from random due to ancient monetary systems and the policies and resources of modern states. We hope that the data compiled can serve as the basis of analyses that will use the same degree of critical scrutiny as is applied to other categories of evidence to allow the construction of a monetary framework for the understanding of the Late Antique and Early Medieval economy.

Alan M. Stahl

Chair, FLAME Project