Estimating the volume of ancient coin production is a topic which, once hotly debated 1 2 3 4, is now been generally accepted as worth to be pursued 5.
The common strategy is a two-step process: first, to determine the number of original dies from the number of recorded dies in the sample. Two dozens of statistical methods have been proposed for this task whose results are generally considered as unproblematic as long as the ratio number of cois/number of dies is superior to 3 (which is the case for most ancient coinages) [[1]]. The second step is to multiply the original number of dies by what we think was the average production of a die. This is much more debatable. While it is not uneasy to produce a long list of reasons for which this average should have varied giving the illusion of a desperate level of uncertainties, things are hopefully less unsecure looking at coins. The nature of the metal and the size of the coins make indeed significant differences (but keeping inside the same level of magnitude), but it generally turns out that every time one may exercise some control, cheking through coin hoards, results are similar. Minor differences (as those related to designs, tools, and training for example) are not identifiable as such between coinages of similar metal and size.
Now what is the average production for a die? Explicit written sources are nearly non existent on the question (Delphi) while we do have a dozen of favourable cases for which we may combine extrapolated volumes of struck coins with highly precisely dated contexts. It should be noted that the results provided by this scanty evidence looks strikingly coherent. Another way to estimate the average production of ancient dies is to strike new coins trying to reproduce the exact conditions of ancient minting. Serious attempts so far have been few. While this is full of interest in terms of metallurgy and thus relevant to some extend for the issue here (mainly to look at defective dies) under study, it is doubtful that it will ever securely inform us about highly productive dies. The average production of a population of coin dies clearly depends on how these dies will behave once put into production. Questions are: do we have to take into account a large infantile mortality ? In other words: a lot of ides with a limited production which, while spectacular in our die-studies, have feebly contributed to the entire volume.
What about the average dies with an average production ? And what about the possibility to have some extra resistant dies with a tremendously large production. These questions have been differently adressed by the past. It was first and for a long time opted for a binomial (i.e. Gaussian, or symmetric) distribution (Good, Carter), with no infantile mortality (A). This is clearly an unlikely model. In the 1980s, one moved to a Poisson distribution with a negative asymetric curve (B). Around the same time, one of us, looking at hundreds of real distributions, proposed a combination of first a negative distribution (for the infantile mortality) followed by a binomial distributon (for the surviving specimens)
The Greek Overstrikes Database (GOD) aims .
References
- ^ Buttrey, Theodore V. (1993), "Calculating ancient coin production : facts and fantasies," Numismatic Chronicle, 153, p. 335-351
- ^ Buttrey, Theodore V. - Cooper, Denis (1994), "Calculating ancient coin production II : Why it cannot be done," Numismatic Chronicle, 154, p. 341-352.
- ^ Callataÿ, François de (1995), "Calculating ancient coin production : Seeking a balance," Numismatic Chronicle, 155, p. 40-54.
- ^ Buttrey, S. E. - Buttrey, Theodore V. (1997), "Calculating ancient coin production, again," American Journal of Numismatics, 2nd ser., 9, p. 113-135.
- ^ Callataÿ, François de (2005), “A quantitative survey of Hellenistic coinages: recent achievements”, in Zosia H. Archibald, John K. Davies and Vincent Gabrielsen (eds.), Making, Moving and Managing. The New World of Ancient Economies, 323-31BC, Oxford, Oxbow, p. 73-91.
