De Heilige Kruiskapel te Utrecht. Die Tatsachen bleiben, die Interpretation schwänkt
As is generally assumed, there was a small church within the walls of the Roman castellum as early as the seventh century. It is likely that it was founded by the Merovingian king Theutbert II (586-612) or his successor Chlotarius II (584-629) and destroyed by the Frisians. Around 695 Willibrord rebuilt the destroyed church and consecrated it to St Maarten (St Martin), possibly as a sign of gratitude towards the Frankish king, whose mayor of the palace, Pippijn II, supported him in his missionary work. Next to it he built a second church, St Salvator. He connected this church with a monastery. Until 1829 the Heilige Kruiskapel, a single-nave building with two modest transepts and a strikingly long choir was situated at Dom square. To the north of it was the collegiate church of St Maarten (Dom cathedral) and until 1582 the collegiate church of St Salvator was to be found to the south of the chapel.
The special situation of three church buildings occurring at a very short distance from one another, linked to the question where the first church of Utrecht was situated, has intrigued many researchers. There is general agreement on the history and situation of St Salvator, however, the location of the small St Maarten's church built by Willibrord is contested. Within the boundaries of the Gothic St Maarten's church no trace of an older church that could date back to the time of Willibrord has been found so far. Nor has any trace ever been found of the small Merovingian church.
Before the Second World War excavations at Dom Square made the remnants visible of the Heilige Kruiskapel and the Roman main building situated directly under it. A.E. van Giffen, the leader of the excavation, dated the chapel to the tenth century. There are archaeological arguments to assume that the oldest church of Utrecht must have consisted of a converted (?) Roman Principia. The Heilige Kruiskapel was built directly upon this converted, and subsequently destroyed, main building.
Van Giffen's dating of the chapel appears to be ill-founded, and moreover, the locations of the finds are scarcely documented. Re-evaluation of the old documentation on the excavation has made a dating of the chapel to around 700 more likely. In order to test this hypothesis part of the chapel was excavated once again in 1993. Apart from one exception, the observations in this research have confirmed the hypothesis which had been developed on the basis of the old data.
This exception concerns radiocarbon dating. 14 C-datings of charcoal, found in the mortar of the brickwork of the chapel, result in a dating to the late ninth or tenth century. The datings are: UtC-2975: 1100 BP ± 60, which corresponds to a dating between 888 and 1011; UtC-2976: 1110 BP ± 70 (883-1011 AD); UtC-2977: 1150 BP ± 60 (820-842 and 859-979 AD).
The renewed interpretation of the old archaeological data, the reconstruction of ground-level heights derived from it and the architectural and topographical arguments, however, make it more likely that the Heilige Kruiskapel is to be identified with the little church rebuilt by Willibrord after all. The (written) historical sources also fit in with this conception. However, if the 14 C-datings are correct, bishop Balderik (917-976) would have been the building commissioner. In that case he would have built a completely new church, according to an outdated model, at the site of its destroyed predecessor. The written sources record that he restored the two heavily damaged churches of St Maarten and St Salvator: no mention is made of new construction, nor of a chapel. Moreover, Balderik decided to proclaim St Maarten's church a cemetery for the bishops. The small chapel seems to be inconsistent with this, although it is frequently regarded as a burial chapel.
To the south of the chapel two parallel foundations have been found, which must have been part of a predecessor of the Roman St Salvator and which may date back to the same period as the chapel. They are the oldest building traces found within Salvator church. For that reason it is not likely that the Kruiskapel is to be identified with an older St Salvator. But the relative chronology, too, both of the reconstructed ground-level heights and of the walling found to the west and south of the chapel would present serious problems in case of a dating to the tenth century. Moreover, this would create a great problem in respect of the ground-level height which must have gone together with the sarcophagi of the eighth and ninth centuries: in the tenth century the walking level would have been at least half a metre lower and this seems unlikely.
However, when we assume that the Kruiskapel was built at the end of the seventh or beginning of the eighth century, all the arguments and data can be explained. The relative chronology and the identical site of the Roman main building and the conversion of it on the one hand, and the chapel on the other do not present any problems then, and neither does the building material used in the various periods and the way in which it was applied. This also applies to the walls to the west of the chapel and those of St Salvator. The shape of the chapel also points to an early dating, and the strikingly large choir possibly to a use for collegial purposes.But the Kruiskapel as old St Maarten's church also fits in easily with the development causing St Maarten's church to become much more important than St Salvator, a pursuit started by Balderik. The fact that the chapel was situated on the territory of Salvator is not inconsistent with this. On the contrary, this very fact combined with the uniform manner of construction of the oldest foundations within St Salvator and those of the Kruiskapel indicate that here we are dealing with the two oldest churches of Willibrord. Consequently, for the time being, we are of the opinion that in spite of the contradiction presented by the radiocarbon datings, the Heilige Kruiskapel is the small church (re)built and (re?)consecrated to St Maarten by Willibrord.