People across the world are increasingly tightly connected through their personal, professional and online social networks. These are studied by the thriving discipline of network science, helping us to understand how viruses and information alike spread through social networks. But we didn’t always live in a global village. How do present-day social networks differ from those in the past, and how have these structures changed over centuries? How did phenomena like the adoption of innovations change as the social networks through which they spread changed? 

The key to these important questions lies with archaeological and historical network research. This growing field adapts computational and statistical network approaches developed by social scientists and physicists, and applies them to archaeological material culture and historical textual sources. Roman inscriptions represent a crucial resource for the study of past social networks, thanks to being well-documented in large numbers and the ability to derive both material and textual information from them. These are monumental carved inscriptions, graffiti or scratched inscriptions, and dipinti or painted inscriptions, that can contain a diversity of information including personal names, professional affiliation, or interpersonal relationships. Vast numbers have been documented on objects ranging from stone temples, wall paintings in Pompeii and funerary stones, to scratchings and painted markings on ceramic amphora containers. Crucially, hundreds of thousands of inscriptions have recently become openly available in online databases and are ready to be analysed by formal methods.

But unlocking the potential of past social network studies is currently prevented by a number of challenges. First, past social network data is rarely made open access, which limits our ability to replicate published results about past social network structure and prohibits quantitative comparative studies. Second, social network reconstructions from Roman inscriptions are extremely rare, missing a crucial opportunity to explore how material and textual data in combination can reveal a range of interpersonal social networks (kinship, professional, political), how they differed across geographical contexts (the Roman world spanned at its height from Spain to Syria and from Britain to Morrocco), and how they changed over centuries-long timespans (ca. 200 BC – AD 500). Third, there is a need for a methodological framework to enable formal comparisons between past social networks, that incorporates statistical sensitivity analyses of the impact of missing or uncertain information that is a common feature of all archaeological and historical data.

We tackle the three challenges by developing the discipline-wide standards of publication and deposition historical network data, facilitated by the creation of the first Open Repository for Historical Network Data. Further, we plan to develop methods of uncertainty quantification within the historical networks and apply them in our empirical studies.