Online research of early Celtic art in Britain -- part 3

Kirkburn Sword, "logo" of 

British Museum's

Celtic 

art database

© Trustees of the British Museum

The British Museum's Celtic art database is a free, downloadable Excel spreadsheet file. The project's home page explains how it was compiled and gives directions for its use. It is introduced thus:

The project team has compiled a comprehensive database of all Celtic art found in Britain to date. This includes excavated finds, and finds recently reported to the Portable Antiquities Scheme.

Some might be surprised to see that the file contains only 2,754 entries but certain very large categories have been omitted -- coins (more than 40,000 recorded examples), and brooches (unknown number but another very large category - certainly many thousands). The latter two categories also require specialized knowledge. Despite the claim of being comprehensive, objects recorded in the trade have not been included, neither have objects recorded in other databases. For example, this important strap junction recorded in the UK detector finds database does not also appear in the Celtic art database. What remains is mostly what you could find at all of the museums in Britain (without the coins and brooches) plus the collections of those metal detectorists who report their finds to the PAS.

To complicate matters further, the data within the Celtic art database has been channelled according to the team's system of categories and not all alternative systems are mentioned. Most databases impose systems on the data to a greater or lesser degree and this is mainly why having the same sorts of things recorded in different databases is more than just useful -- as newer ways of viewing the data reveal themselves through the application of diverse systems, then all systems can adjust to encompass these views and these processes are repeated over time. If access was restricted to a single database following its own system, the chances of progress is minimal. Having two very different sources of data: museum holdings and metal detector finds, is problematical as these categories do not mix very well. The only way to improve on this situation would be to include all other possible sources. Failing that, the database's idiosyncrasies must be understood and sometimes it is more a matter of working around the way data is presented than working with the data.

I suppose that brooches were left out mainly because there are so many of them and they are recorded in so many places. Including Roman period brooches would have made for a huge project. Celtic brooches are usually classified as La Tène 1, 2, and 3. These divisions reflect aspects of the general design of the brooches focusing on the foot and its relationship to the bow: with  La Tène 1, the foot turns back toward the bow but does not make contact; with  La Tène 2, the foot attaches to the bow; with  La Tène 3, the foot ends at the catch plate and does not return toward the bow. I would have included all  La Tène 1 and 2 brooches as they are fairly rare and can provide all sorts of valuable information. On the continent, the  La Tène 1, 2, 3 system is very useful as brooches are common in graves, and help to identify and date associated items that might be unique or extremely rare. In Britain, the rarest Celtic brooch type is the  La Tène 2 involute brooch which undergoes some stylistic evolution in its life from circa 210 BC down to circa 50 BC and more precise dating is certainly required in this time span.

In summation, treat the Celtic art database as another tool but understand its idiosyncrasies. Most of what I could say about it is best left for my discussion of the Portable Antiquities Scheme which will start on Monday.