1Located in a bend of the river Weser near Höxter in eastern Westphalia (Germany), Corvey is one of the largest abandoned medieval towns in Central Europe. Excavations led by Hans-Georg Stephan between the 1970s and early 2000s revealed stone-walled cellars, a paved road and remains of a large church with cemetery (Stephan, 2000). Despite his seminal work, the geography and development of the urban settlement are still poorly understood. Therefore, the archaeology department of the regional association of Westphalia-Lippe (LWL) and the Ludwig Boltzmann Institute for Archaeological Prospection and Virtual Archaeology (LBI ArchPro) initiated a high-resolution GPR and magnetometry surtentative parish church, which was partially excavated in 1988 (Coolen et al., 2020). Combining the results of the GPR survey with evidence from the excavation, we can reconstruct the building in great detail and draw further conclusions on its chronology and architectural context.
2The abbey of Corvey was founded by Benedictine monks from Corbie in France (hence the name ‘nova Corbeia’) in 822 and soon became a prominent theological and intellectual centre the newly conquered Saxon lands. Only few decades later, the priory Niggenkerken (meaning ‘new church’) was established some 500 m to the southwest. The abbey’s strategic location at a crossing of the Westphalian Hellweg, an important traffic route, and the river Weser, as well as the minting privilege issued in 833, stimulated the development of a market. In the following centuries, the lay settlements surrounding the abbey and the priory grew into a medium sized town, which was fortified with an earthen rampart and a moat in the second half of the 12th century (Fig. 1). The fortification encompassed 55 ha, including both monasteries. Besides struggles between the abbots of Corvey and other worldly and ecclesiastical rulers, the competition with Höxter, another market town located only 1.5 km upstream, caused increasing friction. In 1265, soldiers of the Paderborn Bishop Simon zur Lippe and residents of Höxter assaulted Corvey. The town seems to have rapidly declined and shrunk to a small settlement after this event, before being completely abandoned in the 15th century. Although the abbey never regained its original significance, it survived the centuries and became a private palace in the 19th century. In 2014, the abbey complex, including its unique Carolingian westwork, was declared a Unesco World Heritage Site. Along with a rather inconspicuous remnant of the rampart, the westwork is the only visible reminiscence of medieval Corvey today.
Figure 1. Extent of the medieval town of Corvey with the location of the main church buildings and the surveyed area. The black rectangle marks the extent of figure 2.
(Graphics: J. Coolen, LWL-Archäologie für Westfalen. Coordinate grid in ETRS89 UTM-Zone 32N/EPSG 25832.)
3Although field surveys, initial geophysical investigations (Stephan & Südekum, 1995; Sikorski et al., 2011), aerial archaeology and excavations allow for a basic reconstruction of the town’s geography and chronology (Stephan, 2000; König et al., 2003; Koch et al., 2006, 2015; Bergmann, 2011), many questions cannot be answered yet. Given the size and the site’s protected status, which prevents further large-scale excavations, we can only strive to answer these questions with non-destructive methods. Earlier surveys have demonstrated the potential of geophysical surveys but do not provide a holistic interpretation of the entire site. Moreover, new technical developments allow for a much more detailed recording of archaeological structures. Therefore, LWL and LBI ArchPro aim to cover the entire area of the medieval town and the abbey with high-resolution GPR and magnetometry within the next few years.
4The first campaign in 2019 focused on a former sawmill in the central area of the medieval town, which will be redeveloped in the run-up to a large horticultural show in 2023.
5The GPR survey was carried out with a 16-channel Malå Mira system with 400 MHz antennas, offering a spatial resolution of app. 8 × 8 cm (cf. Trinks et al., 2018). The data were processed in LBI ArchPro’s APRadar-software. Despite the rather loamy and moist soil, the penetration depth reached up to 1.5 m.
6After tentative remains of a church had been located in a trial trench in 1981, an excavation in 1988 revealed the choir and two side apses. Although the excavation did not produce direct evidence for the construction date, Stephan believed that the Romanesque church was erected in the first half of the 12th century or soon after 1150. The finds show that it was used until the 15th century. This is in line with historical records referring to the abolition of the parish in 1512 and imply that this was indeed the marcketkerchen (market church) mentioned in historical documents, hence providing an important clue for the location of the market square. The GPR data show the complete ground plan of the church building and reveal some hitherto unknown and important construction details (Fig. 2-3). The excavation trench of 1988 is clearly visible in the data, as the gravel backfill causes strong reflections. The data confirm that it was a three-aisled church without a transept. With a total length of 36.5 m, the church appears to be much smaller than Stephan had estimated. On the west side, the solid foundations of an almost square tower with external dimensions of app. 9.2 × 8.2 m stand out, which aligns with the central nave. Such central belfries became a typical element of parish churches in Westphalia from the 12th century onwards.
Figure 2. Combined GPR depth slice of app. 70-110 cm below ground surface, showing the foundations of the medieval parish church. Reflective deposits are shown dark.
(Graphics: J. Coolen, LWL-Archäologie für Westfalen & M. Wallner, LBI ArchPro.)
Figure 3. Archaeological interpretation of the GPR data.
(Graphics: J. Coolen, LWL–Archäologie für Westfalen.)
7In analogy to late Romanesque churches in the Weser area, the church can be reconstructed as a basilica with some certainty. The width of the nave (including aisles) is 17.6 m. The central nave has a clear width of 6.8 m and is more than twice as wide as the aisles (max. 2.8 m). The nave is separated from the aisles on both sides by a series of rectangular pier foundations. The northern row in particular reveals an alternating system of a strong central pier flanked by smaller intermediate pillars. The alternating pillars, along with the width ratio of the nave and aisles and the rather massive foundations, suggest that the nave and aisles were completely vaulted. This sets it apart from other Romanesque churches in the vicinity and implies a mid-12th century or later date.
8A narrower foundation wall runs off the north aisle, suggesting an annex, possibly a sacristy or a porch, on the northern side. The scant remains of a tentative cemetery wall, which were recorded during the excavation some 5 m north of the church, appear as a diffuse linear anomaly running to the northeast. While numerous graves were found around the church during the excavation, they unfortunately do not cause sufficient contrast to be identified by GPR.
9Apart from the northern annex, which is likely a later addition, the GPR data do not allow for a clear distinction of multiple construction phases. Indeed, they seem to confirm that the church was designed as a late Romanesque basilica from the start. However, the excavation revealed at least two phases of the choir and suggest that the aisles and tower were added later. The reconstruction of the complete church based on the GPR survey calls for a critical reconsideration of the excavation results and their original interpretation, thus providing a fine example of how non-destructive prospection can help making sense of legacy excavation data.