Medieval castles are among the most well-studied subjects in European historiography (Figure 1), not least for their role as material evidence for the consolidation of seigneurial power between the Early and Late Middle Ages (tenth–thirteenth centuries). Within the Castles project, we chose to focus on sites from three Italian regions, Tuscany, Liguria and Piedmont, as the development of castles in this part of the peninsula appears to have followed shared characteristics and patterns (Figure 2). These include the presence of pre-existing early medieval (seventh–ninth centuries) timber settlements, an initial burst of castle construction in the ninth–tenth centuries, and a prolonged phase of gradual transformation into fortified settlements, some of which persist into the modern era (Francovich & Hodges 2003).

Figure 1.

Aerial view of Rocca San Silvestro in Tuscany, one of the case studies in the project (photograph by authors).

The Castles project involved four research units, each with its own distinct disciplinary focus (Figure 3). The initial research strategy focused on identifying castles with structures that could plausibly date to the eleventh–twelfth centuries, based on pre-existing textual and archaeological evidence. After a comprehensive review of available sources, we conducted field surveys at each identified site, together with preliminary geological analyses of the locations and surrounding areas.

Figure 3.

Research units and disciplinary focus in the project (figure by authors).

The project successfully identified new key chronological trends, revealing a novel picture of the development of castles and of the phenomenon of incastellamento as a whole, both before and during the twelfth century.

In the three regions examined, we identified a strong phase of castle building between the tenth and the first half of the eleventh century. Castles from this phase were generally sizable and associated with specific economic activities, reflecting the interests of elites operating in the orbit of central royal authority. This finding diverges from the prevailing view that castles from this period were of limited economic and material significance.

Between the eleventh and twelfth centuries, a time marked by a crisis of central authority and by the consolidation of seigneurial powers, castles underwent processes of transformation. In most cases, these involved the reuse of pre-existing structures or the construction of new elite residences, buildings with strong symbolic significance. This offers tangible evidence of the broader seigneurial reorganisation. However, this phase left material traces that are less visible and less imposing than expected and which do not seem to feature the construction of fortified villages within the castle walls, contrary to earlier historiographical claims (Carocci 2021). It is only at the very end of the twelfth century that castles become veritable fortified settlements. The project has therefore highlighted a new and important chronological shift that leads us to reconsider the very nature of seigneurial change in Italy around 1100.

Indeed, the project results show that the construction of new castle walls, of houses and of structures serving the castle community all date to late periods. These findings represent a novelty within current historiography and prompt a reconsideration of established hypotheses regarding not only the fabric of power in the Italian countryside, but also settlement dynamics, castle building chronologies and, more broadly, the history of rural landscapes.

The project’s originality rests on four key aspects: its multidisciplinary approach; its scale, both in terms of its extensive territorial scope (among the largest in the Mediterranean) and the large number of mortar samples analysed; the integration of analyses with key historical questions; and the creation of a new protocol for analysis. The discovery of large stone castle structures dating back to before the year 1000 and the absence of fortified villages before the late twelfth century are important findings that pave the way for broader historical investigations, which will require further modelling and comparison across larger territorial contexts.