The construction of early medieval Slavic hillforts in Central Europe began more than a century earlier than previously estimated. A new chronological analysis traces the emergence of these strategic mountain settlements back to the middle of the seventh century, revising the timeline of the Great Moravian proto-state. The study was published in the journal Archaeological and Anthropological Sciences.
The region comprising present-day Moravia and western Slovakia served as the eastern frontier of the Carolingian Empire during the Early Middle Ages. This borderland eventually gave rise to Great Moravia, an influential Slavic political entity that acted as a buffer territory. The ruling elites of Great Moravia absorbed Western engineering blueprints and military strategies, eventually constructing mass fortified networks. The Bojná agglomeration, situated in the forested heights of the Inovec Mountains in western Slovakia, contains some of the most prominent ruins of these stronghold settlements.
The traditional timeline for these defensive networks places their rapid construction at the end of the eighth century or the early ninth century. Recent archaeological perspectives have pushed that date even later, proposing that the primary fortification phase did not begin until the final years of the ninth century. These conflicting timelines create an open question regarding how quickly local Slavic populations adopted advanced socio-economic structures and complex defensive architecture. Historical texts from this era are sparse and often written by outsiders, making physical evidence the primary tool for reconstructing the history of these frontier populations. Accurately dating these settlements helps anthropologists track the political stabilization of early Slavic tribes following the regional collapse of the neighboring Avar Khaganate.
Zbigniew Robak of Masaryk University led the study, working alongside co-author Karol Pieta of the Slovak Academy of Sciences. The team analyzed 79 radiocarbon samples gathered from five core sites within the Bojná agglomeration. These excavation sites include large fortified settlements, smaller outposts, and adjacent burial mounds. The researchers extracted carbon material from charcoal, iron slag, and wood that originated from rampart structures, defensive ditches, and domestic hearths.
To refine the absolute dates, the team employed Bayesian statistical modeling. This mathematical approach integrates raw radiocarbon probabilities with known archaeological layering and existing tree-ring data. The process filters out anomalies by cross-referencing multiple lines of physical evidence. The researchers evaluated every sample’s original field documentation and excluded 28 dates—some corrupted by root mold, prolonged environmental exposure, or extremely low carbon content, but most simply falling outside the relevant early medieval range (and likely reflecting genuine earlier or later phases of activity), and some bone samples set aside over concerns about their stratigraphic context. The remaining 51 reliable samples formed the foundation of their newly constructed chronological sequence.
The scientists categorized their samples by physical context to weigh their reliability. They treated scattered charcoal found in soil fill as an indicator of the earliest possible date for human activity in that immediate layer. Alternatively, carbon extracted from the load-bearing timbers inside a wall provided a highly probable date for the actual construction of that specific defensive structure.
The chronological models indicate that settlement at the Bojná complex unfolded in several distinct phases, beginning much earlier than the established eighth-century timeline. Based on the radiocarbon dates combined with artifact distributions, the earliest settlement, known as Bojná III, was established around the middle of the seventh century. The physical evidence shows continuous human activity at this site until its eventual abandonment in the early ninth century. The seventh-century ramparts coincide temporally with “Samo’s state,” an early, temporary political federation of Slavic tribes recorded in historical texts.
The data map out a secondary wave of major construction occurring near the end of the ninth century. The local inhabitants established the Bojná I and Bojná II strongholds precisely as the earliest foundational site fell out of use. Carbon samples taken directly from the massive load-bearing timbers of these newer defensive walls align closely with existing tree-ring data marking the year AD 890. This convergence points to a massive peak in complex engineering efforts during that specific decade, accompanied by repeated maintenance and expansion stretching into the early tenth century.
The revised sequence suggests that the earliest strongholds functioned primarily as symbolic centers of authority for local rulers. Advanced military-grade fortifications with complex outer walls and reinforced gates replaced the simple hilltop enclosures only as the broader Great Moravian proto-state matured. The continuous history of the Bojná complex demonstrates a gradual, century-spanning evolution of local social structures and a slow integration of Carolingian defensive techniques.
Radiocarbon dating in early medieval contexts presents persistent methodological challenges. The researchers observed a specific pattern where dates from the first half of the ninth century frequently registered as artificially older, a tendency they suspect may stem from the properties of the standard radiocarbon calibration curve, though they note it could also relate to specific characteristics of samples from this region. This statistical anomaly can make certain archaeological features appear significantly older than their true age, complicating an exact reading of the historical timeline. The practice of dating long-lived species of wood also introduces uncertainties, as the inner heartwood of an ancient tree yields an older radiocarbon date than the year the tree was actually felled and used for building.
The study also faced difficulties with poorly preserved biological samples. Several human and animal bones recovered from nearby burial mounds were excluded—a human bone sample possessed insufficient nitrogen levels and low carbon, rendering it useless for precise dating, while the animal bones, whose values were potentially consistent with reliable dating, raised concerns over their stratigraphic context. Future excavations will target new, well-preserved samples from the broader burial sites to clarify the exact transition period between the abandonment of the oldest seventh-century settlements and the rise of the massive ninth-century fortresses. The researchers suggest that obtaining a larger dataset of stratigraphically secure samples is necessary to resolve these multi-phase occupation models completely.
The paper, “Revisiting early medieval chronologies: radiocarbon dates reveal the origins and history of the Carolingian-age Great Moravian hillforts,” was authored by Zbigniew Robak and Karol Pieta.