Never bored of hoards: creating the first database of ancient Anatolian coin hoards

I’m Leah Lazar, postdoctoral researcher on the CHANGE project. Since joining the project in September, my first task has been to create a database of Anatolian coin hoards. It has been a busy and varied few months: I’ve had to think quite a bit about database design, I’ve delved into archival material on coins long since sold at auction, and I’ve become very, very familiar with the composition of over 700 relevant hoards. I’m finally nearing the end, and the fruits of my labour will be publicly accessible via www.coinhoards.org in the coming months.

High-value coins from the ancient world are usually found in hoards, that is to say collections of coins (and possibly other valuable objects) sometimes deposited deliberately for safe-keeping, sometimes accidentally lost. Hoards can provide valuable information for numismatists and historians. Take Ancient Anatolia, our region of focus. It was an overlapping patchwork of varied political entities of different sizes, from cities, to regional dynasts, to kingdoms, to empires. Many of these entities produced coinage, on different scales, for different purposes. So how do we make sense of such a complex picture of monetary interaction? From the evidence of hoards, it is possible to see how coinages actually moved from place to place – or not. Hoards show how some coinages were only used locally (like the low-value bronze coinages of cities), while others travelled much further afield (like the higher-value silver coins with the widely recognisable types of Alexander the Great).

But, until recently, it has been sometimes slow and difficult to access information about hoards, particularly groups of hoards, as it was published in unwieldy print publications. A number of projects have now begun to make hoard data more accessible and searchable through online databases, including the American Numismatic Society’s digitisation of the important 1973 Inventory of Greek Coin Hoards at www.coinhoards.org. Over 2000 hoards are included, an amazing resource, but one that is far from comprehensive. Our plan is to make available information on all hoards either found in Anatolia or containing coins produced in Anatolia, so that a full regional analysis is possible. So, in addition to the 264 hoards found in Anatolia, published in IGCH, and already online, I’ve been readying data on over seven hundred relevant hoards published in the ten volumes of Coin Hoards, a publication which catalogued ancient hoards from 1975 to 2010. We might miss a few more recently published hoards, but our database should be fairly complete.

Luckily, we had basic electronic database records for the hoards in Coin Hoards, so I wasn’t starting from scratch. The data, however, was incomplete and not suitable for the digital environment of www.coinhoards.org, which uses a Linked Open Data framework to make its records talk to other online resources on coinage and the ancient world [maybe insert a link to something on Linked Open Data here?]. The records in Coin Hoards often, necessarily, provide only minimal information, and so I had to go back to the original publications (in languages from Albanian to Turkish) to get full details about hoards. Sometimes, the only published record of a hoard is the Coin Hoards entry, as a scholar briefly got eyes on a hoard before it was sold on the market. In those cases, I tried to find more information in the editors’ archive, in the collection of the American Numismatic Society. I then had to tidy all the data, and adjust it for the Linked Open Data framework.

We hope that the database will be available for your use in the next few months (thanks to the efforts of Ethan Gruber, the Director of Data Science at the American Numismatic Society). You’ll be able to go to www.coinhoards.org, search for particular Anatolian hoards, or browse the full data-set by a number of different criteria: findspot, minting authority, denomination, metal… Once you click on a hoard, a complete record will appear, listing all the coins found in the hoard, along with a map showing its findspot and the location of the mints which produced the coins. You’ll also be able to click through to other linked resources. And you’ll be able to download data in various formats, too.

For my part, I’ve had my eyes on the ultimate research prize all along, even while bogged down in massive spreadsheets. We can get a lot from looking not just at individual hoards, but through consideration of multiple hoards together ­­– and now, for the first time, large-scale statistical analysis of Anatolian coin hoards is possible. So I’m mugging up on R (the programming language used by statisticians), and I’m starting off with a very high-level breakdown of hoarding patterns in the region over the seven centuries covered by the project. What are the geographical patterns in hoarding behaviour in ancient Anatolia? How did the frequency and size of hoards vary over time – and why? When and where did people hoard gold, silver, bronze and electrum coins – and how were different metals hoarded together? How far, on average, did coins travel from their mints, over the centuries? Give me a few months, and I should be able to provide some answers, or at least find some new questions to ask.