Taming the chronology of South Gaulish Samian found at Hadrian's Wall and the German Limes using Linked Open Data

Since the mid-1990s the RGZM (Römisch-Germanisches Zentralmuseum Mainz) provides web-based databases containing millions of datasets with content from many different archaeological disciplines. These relational databases were constructed in interdisciplinary transnational projects and include a lot of “hidden archaeological assumptions”. The aim of our paper is to make these hidden assumptions visible and provide them as a Linked Open Data graph to establish reproducible research as a basis for Open Science.

In particular the Samian Research database [1] at the RGZM which is based on the Names on Terra Sigillata project of Leeds and Reading Universities funded by the AHRC offers nearly 250'000 identified potter stamps, which are traditionally dated in a traditional way. In Roman archaeology this is usually expressed by establishing "absolute dates" in well known “from-to” tables, whereas in reality, the situation is much more diffuse. In fact, Samian dating is quite often based upon so-called ‘dated sites’, which are themselves dated by Samian. Occurrences of South Gaulish Samian on “historically dated sites” like Hadrian’s Wall with its absolute dating or the Wetteraulimes with its less clear dating are difficult to use as dating arguments because it is not possible to say a priori whether these were “survivals” or whether they were part of a still existing South Gaulish export industry. Therefore an RDF (Resource Description Framework) based model has been developed to deal with occurrences making the Samian dating process verifiable and transparent.

This paper focuses on modelling dating arguments using a relative chronology based on Allen’s interval algebra [2] compared with absolute dates in the prototypical build Academic Meta Tool [2] to create Linked Open Data for reproducible and transparent research. This tool enables the creation of an RDF-based ontology such as for modelling dating arguments and visualising the semantic reasoning for detecting errors.

[1] http://rgzm.de/samian

[2] http://cse.unl.edu/~choueiry/Documents/Allen-CACM1983.pdf

[3] http://academic-meta-tool.xyz